diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -10,21776 +10,21788 @@
   "log_history": [
     {
       "epoch": 0.0009652509652509653,
-      "grad_norm": 37.57219314575195,
+      "grad_norm": 30.452482223510742,
       "learning_rate": 2.1276595744680852e-07,
-      "loss": 2.9127,
+      "loss": 2.801,
       "step": 1
     },
     {
       "epoch": 0.0019305019305019305,
-      "grad_norm": 34.530269622802734,
+      "grad_norm": 29.30376434326172,
       "learning_rate": 4.2553191489361704e-07,
-      "loss": 2.9575,
+      "loss": 2.8243,
       "step": 2
     },
     {
       "epoch": 0.0028957528957528956,
-      "grad_norm": 31.692672729492188,
+      "grad_norm": 27.31394386291504,
       "learning_rate": 6.382978723404255e-07,
-      "loss": 2.8706,
+      "loss": 2.7423,
       "step": 3
     },
     {
       "epoch": 0.003861003861003861,
-      "grad_norm": 31.64986801147461,
+      "grad_norm": 29.70621681213379,
       "learning_rate": 8.510638297872341e-07,
-      "loss": 2.9322,
+      "loss": 2.8103,
       "step": 4
     },
     {
       "epoch": 0.004826254826254826,
-      "grad_norm": 27.894390106201172,
+      "grad_norm": 25.657827377319336,
       "learning_rate": 1.0638297872340427e-06,
-      "loss": 2.7927,
+      "loss": 2.6754,
       "step": 5
     },
     {
       "epoch": 0.005791505791505791,
-      "grad_norm": 24.181001663208008,
+      "grad_norm": 23.523008346557617,
       "learning_rate": 1.276595744680851e-06,
-      "loss": 2.8115,
+      "loss": 2.7307,
       "step": 6
     },
     {
       "epoch": 0.006756756756756757,
-      "grad_norm": 20.626800537109375,
+      "grad_norm": 19.207117080688477,
       "learning_rate": 1.4893617021276596e-06,
-      "loss": 2.6941,
+      "loss": 2.6095,
       "step": 7
     },
     {
       "epoch": 0.007722007722007722,
-      "grad_norm": 13.248446464538574,
+      "grad_norm": 15.107217788696289,
       "learning_rate": 1.7021276595744682e-06,
-      "loss": 2.5758,
+      "loss": 2.5142,
       "step": 8
     },
     {
       "epoch": 0.008687258687258687,
-      "grad_norm": 15.200055122375488,
+      "grad_norm": 16.746768951416016,
       "learning_rate": 1.9148936170212767e-06,
-      "loss": 2.5315,
+      "loss": 2.4693,
       "step": 9
     },
     {
       "epoch": 0.009652509652509652,
-      "grad_norm": 13.172714233398438,
+      "grad_norm": 14.718594551086426,
       "learning_rate": 2.1276595744680853e-06,
-      "loss": 2.5467,
+      "loss": 2.4966,
       "step": 10
     },
     {
       "epoch": 0.010617760617760617,
-      "grad_norm": 16.15408706665039,
+      "grad_norm": 18.184022903442383,
       "learning_rate": 2.340425531914894e-06,
-      "loss": 2.4509,
+      "loss": 2.4024,
       "step": 11
     },
     {
       "epoch": 0.011583011583011582,
-      "grad_norm": 17.439682006835938,
+      "grad_norm": 20.275991439819336,
       "learning_rate": 2.553191489361702e-06,
-      "loss": 2.5752,
+      "loss": 2.5471,
       "step": 12
     },
     {
       "epoch": 0.012548262548262547,
-      "grad_norm": 15.113956451416016,
+      "grad_norm": 15.218214988708496,
       "learning_rate": 2.765957446808511e-06,
-      "loss": 2.4805,
+      "loss": 2.4185,
       "step": 13
     },
     {
       "epoch": 0.013513513513513514,
-      "grad_norm": 13.25255012512207,
+      "grad_norm": 11.890456199645996,
       "learning_rate": 2.978723404255319e-06,
-      "loss": 2.4514,
+      "loss": 2.3615,
       "step": 14
     },
     {
       "epoch": 0.01447876447876448,
-      "grad_norm": 10.589703559875488,
+      "grad_norm": 13.308670043945312,
       "learning_rate": 3.191489361702128e-06,
-      "loss": 2.3685,
+      "loss": 2.3115,
       "step": 15
     },
     {
       "epoch": 0.015444015444015444,
-      "grad_norm": 8.39791488647461,
+      "grad_norm": 9.292367935180664,
       "learning_rate": 3.4042553191489363e-06,
-      "loss": 2.2548,
+      "loss": 2.2016,
       "step": 16
     },
     {
       "epoch": 0.016409266409266408,
-      "grad_norm": 11.414656639099121,
+      "grad_norm": 15.527050018310547,
       "learning_rate": 3.6170212765957453e-06,
-      "loss": 2.2782,
+      "loss": 2.2574,
       "step": 17
     },
     {
       "epoch": 0.017374517374517374,
-      "grad_norm": 17.181034088134766,
+      "grad_norm": 14.601978302001953,
       "learning_rate": 3.8297872340425535e-06,
-      "loss": 2.2162,
+      "loss": 2.1816,
       "step": 18
     },
     {
       "epoch": 0.01833976833976834,
-      "grad_norm": 9.750289916992188,
+      "grad_norm": 9.399457931518555,
       "learning_rate": 4.042553191489362e-06,
-      "loss": 2.2657,
+      "loss": 2.2297,
       "step": 19
     },
     {
       "epoch": 0.019305019305019305,
-      "grad_norm": 6.164453506469727,
+      "grad_norm": 8.317047119140625,
       "learning_rate": 4.255319148936171e-06,
-      "loss": 2.1871,
+      "loss": 2.1502,
       "step": 20
     },
     {
       "epoch": 0.02027027027027027,
-      "grad_norm": 7.542515754699707,
+      "grad_norm": 8.657756805419922,
       "learning_rate": 4.468085106382979e-06,
-      "loss": 2.2052,
+      "loss": 2.1741,
       "step": 21
     },
     {
       "epoch": 0.021235521235521235,
-      "grad_norm": 9.258981704711914,
+      "grad_norm": 7.558076858520508,
       "learning_rate": 4.680851063829788e-06,
-      "loss": 2.1841,
+      "loss": 2.1509,
       "step": 22
     },
     {
       "epoch": 0.0222007722007722,
-      "grad_norm": 8.692361831665039,
+      "grad_norm": 7.5618462562561035,
       "learning_rate": 4.893617021276596e-06,
-      "loss": 2.1692,
+      "loss": 2.1297,
       "step": 23
     },
     {
       "epoch": 0.023166023166023165,
-      "grad_norm": 8.432202339172363,
+      "grad_norm": 7.576826572418213,
       "learning_rate": 5.106382978723404e-06,
-      "loss": 2.1326,
+      "loss": 2.1115,
       "step": 24
     },
     {
       "epoch": 0.02413127413127413,
-      "grad_norm": 6.449735164642334,
+      "grad_norm": 6.072277069091797,
       "learning_rate": 5.319148936170213e-06,
-      "loss": 2.1137,
+      "loss": 2.0881,
       "step": 25
     },
     {
       "epoch": 0.025096525096525095,
-      "grad_norm": 6.142323017120361,
+      "grad_norm": 6.855001926422119,
       "learning_rate": 5.531914893617022e-06,
-      "loss": 2.0913,
+      "loss": 2.0745,
       "step": 26
     },
     {
       "epoch": 0.026061776061776062,
-      "grad_norm": 7.893868923187256,
+      "grad_norm": 7.938473224639893,
       "learning_rate": 5.744680851063831e-06,
-      "loss": 2.0388,
+      "loss": 2.0194,
       "step": 27
     },
     {
       "epoch": 0.02702702702702703,
-      "grad_norm": 7.176388263702393,
+      "grad_norm": 7.5625104904174805,
       "learning_rate": 5.957446808510638e-06,
-      "loss": 2.0789,
+      "loss": 2.0497,
       "step": 28
     },
     {
       "epoch": 0.027992277992277992,
-      "grad_norm": 6.881705284118652,
+      "grad_norm": 6.296346187591553,
       "learning_rate": 6.170212765957447e-06,
-      "loss": 2.128,
+      "loss": 2.0922,
       "step": 29
     },
     {
       "epoch": 0.02895752895752896,
-      "grad_norm": 5.771139621734619,
+      "grad_norm": 5.285174369812012,
       "learning_rate": 6.382978723404256e-06,
-      "loss": 2.0329,
+      "loss": 2.0119,
       "step": 30
     },
     {
       "epoch": 0.029922779922779922,
-      "grad_norm": 5.3250627517700195,
+      "grad_norm": 5.163051128387451,
       "learning_rate": 6.595744680851064e-06,
-      "loss": 2.0776,
+      "loss": 2.052,
       "step": 31
     },
     {
       "epoch": 0.03088803088803089,
-      "grad_norm": 5.021093845367432,
+      "grad_norm": 5.281077861785889,
       "learning_rate": 6.808510638297873e-06,
-      "loss": 2.0388,
+      "loss": 2.025,
       "step": 32
     },
     {
       "epoch": 0.03185328185328185,
-      "grad_norm": 4.986810684204102,
+      "grad_norm": 5.157776832580566,
       "learning_rate": 7.021276595744682e-06,
-      "loss": 2.0429,
+      "loss": 2.0158,
       "step": 33
     },
     {
       "epoch": 0.032818532818532815,
-      "grad_norm": 5.753068447113037,
+      "grad_norm": 5.721765518188477,
       "learning_rate": 7.234042553191491e-06,
-      "loss": 2.0471,
+      "loss": 2.0227,
       "step": 34
     },
     {
       "epoch": 0.033783783783783786,
-      "grad_norm": 6.062497615814209,
+      "grad_norm": 5.7648396492004395,
       "learning_rate": 7.446808510638298e-06,
-      "loss": 2.0988,
+      "loss": 2.0674,
       "step": 35
     },
     {
       "epoch": 0.03474903474903475,
-      "grad_norm": 5.042311191558838,
+      "grad_norm": 5.641718864440918,
       "learning_rate": 7.659574468085107e-06,
-      "loss": 2.0457,
+      "loss": 2.0285,
       "step": 36
     },
     {
       "epoch": 0.03571428571428571,
-      "grad_norm": 5.41193151473999,
+      "grad_norm": 4.963710308074951,
       "learning_rate": 7.872340425531916e-06,
-      "loss": 1.972,
+      "loss": 1.9503,
       "step": 37
     },
     {
       "epoch": 0.03667953667953668,
-      "grad_norm": 4.624916076660156,
+      "grad_norm": 4.700796127319336,
       "learning_rate": 8.085106382978723e-06,
-      "loss": 1.9425,
+      "loss": 1.9279,
       "step": 38
     },
     {
       "epoch": 0.037644787644787646,
-      "grad_norm": 5.057003498077393,
+      "grad_norm": 5.015736103057861,
       "learning_rate": 8.297872340425532e-06,
-      "loss": 1.9985,
+      "loss": 1.9753,
       "step": 39
     },
     {
       "epoch": 0.03861003861003861,
-      "grad_norm": 5.198980331420898,
+      "grad_norm": 4.837634563446045,
       "learning_rate": 8.510638297872341e-06,
-      "loss": 1.9534,
+      "loss": 1.9365,
       "step": 40
     },
     {
       "epoch": 0.03957528957528957,
-      "grad_norm": 5.4440836906433105,
+      "grad_norm": 5.602592945098877,
       "learning_rate": 8.72340425531915e-06,
-      "loss": 1.9787,
+      "loss": 1.9507,
       "step": 41
     },
     {
       "epoch": 0.04054054054054054,
-      "grad_norm": 4.812963008880615,
+      "grad_norm": 4.94753885269165,
       "learning_rate": 8.936170212765958e-06,
-      "loss": 1.9238,
+      "loss": 1.9048,
       "step": 42
     },
     {
       "epoch": 0.041505791505791506,
-      "grad_norm": 5.2607831954956055,
+      "grad_norm": 5.38804817199707,
       "learning_rate": 9.148936170212767e-06,
-      "loss": 2.0111,
+      "loss": 1.9968,
       "step": 43
     },
     {
       "epoch": 0.04247104247104247,
-      "grad_norm": 4.675182819366455,
+      "grad_norm": 4.502268314361572,
       "learning_rate": 9.361702127659576e-06,
-      "loss": 1.9967,
+      "loss": 1.9843,
       "step": 44
     },
     {
       "epoch": 0.04343629343629344,
-      "grad_norm": 5.039648532867432,
+      "grad_norm": 4.7795610427856445,
       "learning_rate": 9.574468085106385e-06,
-      "loss": 1.8747,
+      "loss": 1.8541,
       "step": 45
     },
     {
       "epoch": 0.0444015444015444,
-      "grad_norm": 5.738279819488525,
+      "grad_norm": 5.490640163421631,
       "learning_rate": 9.787234042553192e-06,
-      "loss": 1.9386,
+      "loss": 1.9126,
       "step": 46
     },
     {
       "epoch": 0.045366795366795366,
-      "grad_norm": 5.404884338378906,
+      "grad_norm": 5.236721038818359,
       "learning_rate": 1e-05,
-      "loss": 1.9091,
+      "loss": 1.8818,
       "step": 47
     },
     {
       "epoch": 0.04633204633204633,
-      "grad_norm": 5.5826735496521,
+      "grad_norm": 6.176106929779053,
       "learning_rate": 1.0212765957446808e-05,
-      "loss": 1.9017,
+      "loss": 1.8794,
       "step": 48
     },
     {
       "epoch": 0.0472972972972973,
-      "grad_norm": 4.8203020095825195,
+      "grad_norm": 4.548673629760742,
       "learning_rate": 1.0425531914893619e-05,
-      "loss": 1.951,
+      "loss": 1.9356,
       "step": 49
     },
     {
       "epoch": 0.04826254826254826,
-      "grad_norm": 5.30726957321167,
+      "grad_norm": 5.27844762802124,
       "learning_rate": 1.0638297872340426e-05,
-      "loss": 1.9703,
+      "loss": 1.9502,
       "step": 50
     },
     {
       "epoch": 0.04922779922779923,
-      "grad_norm": 4.507512092590332,
+      "grad_norm": 4.929182529449463,
       "learning_rate": 1.0851063829787233e-05,
-      "loss": 1.857,
+      "loss": 1.8334,
       "step": 51
     },
     {
       "epoch": 0.05019305019305019,
-      "grad_norm": 5.3983869552612305,
+      "grad_norm": 5.639317512512207,
       "learning_rate": 1.1063829787234044e-05,
-      "loss": 1.967,
+      "loss": 1.9478,
       "step": 52
     },
     {
       "epoch": 0.05115830115830116,
-      "grad_norm": 5.134077072143555,
+      "grad_norm": 5.549023628234863,
       "learning_rate": 1.1276595744680851e-05,
-      "loss": 1.8909,
+      "loss": 1.87,
       "step": 53
     },
     {
       "epoch": 0.052123552123552123,
-      "grad_norm": 5.454338073730469,
+      "grad_norm": 5.856292247772217,
       "learning_rate": 1.1489361702127662e-05,
-      "loss": 1.8437,
+      "loss": 1.8231,
       "step": 54
     },
     {
       "epoch": 0.05308880308880309,
-      "grad_norm": 5.207176685333252,
+      "grad_norm": 5.818023681640625,
       "learning_rate": 1.170212765957447e-05,
-      "loss": 1.9095,
+      "loss": 1.9004,
       "step": 55
     },
     {
       "epoch": 0.05405405405405406,
-      "grad_norm": 5.950668811798096,
+      "grad_norm": 6.716310501098633,
       "learning_rate": 1.1914893617021277e-05,
-      "loss": 1.9094,
+      "loss": 1.8996,
       "step": 56
     },
     {
       "epoch": 0.05501930501930502,
-      "grad_norm": 4.508608341217041,
+      "grad_norm": 5.162848472595215,
       "learning_rate": 1.2127659574468087e-05,
-      "loss": 1.8993,
+      "loss": 1.8827,
       "step": 57
     },
     {
       "epoch": 0.055984555984555984,
-      "grad_norm": 5.785097122192383,
+      "grad_norm": 6.15037727355957,
       "learning_rate": 1.2340425531914895e-05,
-      "loss": 1.8405,
+      "loss": 1.8408,
       "step": 58
     },
     {
       "epoch": 0.05694980694980695,
-      "grad_norm": 4.732987880706787,
+      "grad_norm": 5.433175563812256,
       "learning_rate": 1.2553191489361702e-05,
-      "loss": 1.9891,
+      "loss": 1.9734,
       "step": 59
     },
     {
       "epoch": 0.05791505791505792,
-      "grad_norm": 4.726958751678467,
+      "grad_norm": 4.555079460144043,
       "learning_rate": 1.2765957446808513e-05,
-      "loss": 1.9444,
+      "loss": 1.9286,
       "step": 60
     },
     {
       "epoch": 0.05888030888030888,
-      "grad_norm": 5.205662727355957,
+      "grad_norm": 4.90121603012085,
       "learning_rate": 1.297872340425532e-05,
-      "loss": 1.838,
+      "loss": 1.8213,
       "step": 61
     },
     {
       "epoch": 0.059845559845559844,
-      "grad_norm": 5.053771495819092,
+      "grad_norm": 5.712584018707275,
       "learning_rate": 1.3191489361702127e-05,
-      "loss": 1.9335,
+      "loss": 1.9165,
       "step": 62
     },
     {
       "epoch": 0.060810810810810814,
-      "grad_norm": 4.497619152069092,
+      "grad_norm": 5.807231426239014,
       "learning_rate": 1.3404255319148938e-05,
-      "loss": 1.8424,
+      "loss": 1.824,
       "step": 63
     },
     {
       "epoch": 0.06177606177606178,
-      "grad_norm": 4.811068534851074,
+      "grad_norm": 4.542278289794922,
       "learning_rate": 1.3617021276595745e-05,
-      "loss": 1.8365,
+      "loss": 1.8216,
       "step": 64
     },
     {
       "epoch": 0.06274131274131274,
-      "grad_norm": 4.842855930328369,
+      "grad_norm": 5.654561519622803,
       "learning_rate": 1.3829787234042556e-05,
-      "loss": 1.825,
+      "loss": 1.8205,
       "step": 65
     },
     {
       "epoch": 0.0637065637065637,
-      "grad_norm": 5.438333988189697,
+      "grad_norm": 5.201457977294922,
       "learning_rate": 1.4042553191489363e-05,
-      "loss": 1.848,
+      "loss": 1.8257,
       "step": 66
     },
     {
       "epoch": 0.06467181467181467,
-      "grad_norm": 5.783076286315918,
+      "grad_norm": 6.281363487243652,
       "learning_rate": 1.425531914893617e-05,
-      "loss": 1.8859,
+      "loss": 1.87,
       "step": 67
     },
     {
       "epoch": 0.06563706563706563,
-      "grad_norm": 6.11845588684082,
+      "grad_norm": 6.0109639167785645,
       "learning_rate": 1.4468085106382981e-05,
-      "loss": 1.8958,
+      "loss": 1.8878,
       "step": 68
     },
     {
       "epoch": 0.06660231660231661,
-      "grad_norm": 6.755260467529297,
+      "grad_norm": 7.37456750869751,
       "learning_rate": 1.4680851063829789e-05,
-      "loss": 1.8746,
+      "loss": 1.8677,
       "step": 69
     },
     {
       "epoch": 0.06756756756756757,
-      "grad_norm": 4.893709659576416,
+      "grad_norm": 4.581382751464844,
       "learning_rate": 1.4893617021276596e-05,
-      "loss": 1.872,
+      "loss": 1.8564,
       "step": 70
     },
     {
       "epoch": 0.06853281853281853,
-      "grad_norm": 5.599435806274414,
+      "grad_norm": 6.746163845062256,
       "learning_rate": 1.5106382978723407e-05,
-      "loss": 1.8361,
+      "loss": 1.8258,
       "step": 71
     },
     {
       "epoch": 0.0694980694980695,
-      "grad_norm": 6.107633113861084,
+      "grad_norm": 4.722505569458008,
       "learning_rate": 1.5319148936170214e-05,
-      "loss": 1.8572,
+      "loss": 1.8477,
       "step": 72
     },
     {
       "epoch": 0.07046332046332046,
-      "grad_norm": 4.610060691833496,
+      "grad_norm": 5.170559406280518,
       "learning_rate": 1.5531914893617023e-05,
-      "loss": 1.9096,
+      "loss": 1.9016,
       "step": 73
     },
     {
       "epoch": 0.07142857142857142,
-      "grad_norm": 5.756293296813965,
+      "grad_norm": 5.489589214324951,
       "learning_rate": 1.5744680851063832e-05,
-      "loss": 1.9358,
+      "loss": 1.9315,
       "step": 74
     },
     {
       "epoch": 0.07239382239382239,
-      "grad_norm": 4.8217573165893555,
+      "grad_norm": 4.888907432556152,
       "learning_rate": 1.595744680851064e-05,
-      "loss": 1.839,
+      "loss": 1.8294,
       "step": 75
     },
     {
       "epoch": 0.07335907335907337,
-      "grad_norm": 4.5468010902404785,
+      "grad_norm": 5.045965194702148,
       "learning_rate": 1.6170212765957446e-05,
-      "loss": 1.8764,
+      "loss": 1.8679,
       "step": 76
     },
     {
       "epoch": 0.07432432432432433,
-      "grad_norm": 4.610559940338135,
+      "grad_norm": 4.98272180557251,
       "learning_rate": 1.6382978723404255e-05,
-      "loss": 1.7964,
+      "loss": 1.7847,
       "step": 77
     },
     {
       "epoch": 0.07528957528957529,
-      "grad_norm": 4.943978786468506,
+      "grad_norm": 5.741695404052734,
       "learning_rate": 1.6595744680851064e-05,
-      "loss": 1.8379,
+      "loss": 1.8303,
       "step": 78
     },
     {
       "epoch": 0.07625482625482626,
-      "grad_norm": 4.75681209564209,
+      "grad_norm": 5.094552040100098,
       "learning_rate": 1.6808510638297873e-05,
-      "loss": 1.8266,
+      "loss": 1.8114,
       "step": 79
     },
     {
       "epoch": 0.07722007722007722,
-      "grad_norm": 5.278621196746826,
+      "grad_norm": 5.305397987365723,
       "learning_rate": 1.7021276595744682e-05,
-      "loss": 1.8639,
+      "loss": 1.8558,
       "step": 80
     },
     {
       "epoch": 0.07818532818532818,
-      "grad_norm": 4.590597629547119,
+      "grad_norm": 4.539999961853027,
       "learning_rate": 1.723404255319149e-05,
-      "loss": 1.8454,
+      "loss": 1.8245,
       "step": 81
     },
     {
       "epoch": 0.07915057915057915,
-      "grad_norm": 4.689551830291748,
+      "grad_norm": 4.374635696411133,
       "learning_rate": 1.74468085106383e-05,
-      "loss": 1.8563,
+      "loss": 1.8495,
       "step": 82
     },
     {
       "epoch": 0.08011583011583012,
-      "grad_norm": 5.145448207855225,
+      "grad_norm": 5.315585136413574,
       "learning_rate": 1.765957446808511e-05,
-      "loss": 1.91,
+      "loss": 1.8935,
       "step": 83
     },
     {
       "epoch": 0.08108108108108109,
-      "grad_norm": 5.2121968269348145,
+      "grad_norm": 5.383888244628906,
       "learning_rate": 1.7872340425531915e-05,
-      "loss": 1.8853,
+      "loss": 1.8763,
       "step": 84
     },
     {
       "epoch": 0.08204633204633205,
-      "grad_norm": 6.532495021820068,
+      "grad_norm": 6.251614093780518,
       "learning_rate": 1.8085106382978724e-05,
-      "loss": 1.923,
+      "loss": 1.9084,
       "step": 85
     },
     {
       "epoch": 0.08301158301158301,
-      "grad_norm": 6.2555251121521,
+      "grad_norm": 7.0370869636535645,
       "learning_rate": 1.8297872340425533e-05,
-      "loss": 1.8334,
+      "loss": 1.8285,
       "step": 86
     },
     {
       "epoch": 0.08397683397683398,
-      "grad_norm": 5.264360427856445,
+      "grad_norm": 5.763238906860352,
       "learning_rate": 1.8510638297872342e-05,
-      "loss": 1.8977,
+      "loss": 1.8899,
       "step": 87
     },
     {
       "epoch": 0.08494208494208494,
-      "grad_norm": 6.3229756355285645,
+      "grad_norm": 6.281044960021973,
       "learning_rate": 1.872340425531915e-05,
-      "loss": 1.9133,
+      "loss": 1.9079,
       "step": 88
     },
     {
       "epoch": 0.0859073359073359,
-      "grad_norm": 4.155460834503174,
+      "grad_norm": 5.231175422668457,
       "learning_rate": 1.893617021276596e-05,
-      "loss": 1.8972,
+      "loss": 1.883,
       "step": 89
     },
     {
       "epoch": 0.08687258687258688,
-      "grad_norm": 4.465296268463135,
+      "grad_norm": 4.401829719543457,
       "learning_rate": 1.914893617021277e-05,
-      "loss": 1.7892,
+      "loss": 1.7821,
       "step": 90
     },
     {
       "epoch": 0.08783783783783784,
-      "grad_norm": 5.7993388175964355,
+      "grad_norm": 7.732595920562744,
       "learning_rate": 1.9361702127659575e-05,
-      "loss": 1.8372,
+      "loss": 1.842,
       "step": 91
     },
     {
       "epoch": 0.0888030888030888,
-      "grad_norm": 4.296961784362793,
+      "grad_norm": 4.505218029022217,
       "learning_rate": 1.9574468085106384e-05,
-      "loss": 1.8955,
+      "loss": 1.8854,
       "step": 92
     },
     {
       "epoch": 0.08976833976833977,
-      "grad_norm": 6.245906829833984,
+      "grad_norm": 7.1610188484191895,
       "learning_rate": 1.9787234042553193e-05,
-      "loss": 1.8269,
+      "loss": 1.8168,
       "step": 93
     },
     {
       "epoch": 0.09073359073359073,
-      "grad_norm": 4.545825004577637,
+      "grad_norm": 5.883890628814697,
       "learning_rate": 2e-05,
-      "loss": 1.9008,
+      "loss": 1.903,
       "step": 94
     },
     {
       "epoch": 0.0916988416988417,
-      "grad_norm": 7.551859378814697,
+      "grad_norm": 8.624926567077637,
       "learning_rate": 1.999999456770665e-05,
-      "loss": 1.8902,
+      "loss": 1.8826,
       "step": 95
     },
     {
       "epoch": 0.09266409266409266,
-      "grad_norm": 4.8623504638671875,
+      "grad_norm": 5.367443561553955,
       "learning_rate": 1.9999978270832498e-05,
-      "loss": 1.798,
+      "loss": 1.794,
       "step": 96
     },
     {
       "epoch": 0.09362934362934362,
-      "grad_norm": 6.84691047668457,
+      "grad_norm": 6.668648719787598,
       "learning_rate": 1.9999951109395245e-05,
-      "loss": 1.8671,
+      "loss": 1.8552,
       "step": 97
     },
     {
       "epoch": 0.0945945945945946,
-      "grad_norm": 5.37424373626709,
+      "grad_norm": 7.84791374206543,
       "learning_rate": 1.999991308342441e-05,
-      "loss": 1.928,
+      "loss": 1.9246,
       "step": 98
     },
     {
       "epoch": 0.09555984555984556,
-      "grad_norm": 7.096020221710205,
+      "grad_norm": 6.741469383239746,
       "learning_rate": 1.99998641929613e-05,
-      "loss": 1.7758,
+      "loss": 1.7685,
       "step": 99
     },
     {
       "epoch": 0.09652509652509653,
-      "grad_norm": 4.507432460784912,
+      "grad_norm": 8.800213813781738,
       "learning_rate": 1.9999804438059042e-05,
-      "loss": 1.7964,
+      "loss": 1.7994,
       "step": 100
     },
     {
       "epoch": 0.09749034749034749,
-      "grad_norm": 7.82309103012085,
+      "grad_norm": 6.200466156005859,
       "learning_rate": 1.9999733818782543e-05,
-      "loss": 1.8715,
+      "loss": 1.8709,
       "step": 101
     },
     {
       "epoch": 0.09845559845559845,
-      "grad_norm": 4.694272518157959,
+      "grad_norm": 7.305761337280273,
       "learning_rate": 1.999965233520854e-05,
-      "loss": 1.814,
+      "loss": 1.8111,
       "step": 102
     },
     {
       "epoch": 0.09942084942084942,
-      "grad_norm": 6.026909828186035,
+      "grad_norm": 4.827910900115967,
       "learning_rate": 1.9999559987425553e-05,
-      "loss": 1.876,
+      "loss": 1.8631,
       "step": 103
     },
     {
       "epoch": 0.10038610038610038,
-      "grad_norm": 4.93021297454834,
+      "grad_norm": 6.841015815734863,
       "learning_rate": 1.999945677553392e-05,
-      "loss": 1.846,
+      "loss": 1.8463,
       "step": 104
     },
     {
       "epoch": 0.10135135135135136,
-      "grad_norm": 5.0024333000183105,
+      "grad_norm": 5.998527526855469,
       "learning_rate": 1.9999342699645774e-05,
-      "loss": 1.8466,
+      "loss": 1.84,
       "step": 105
     },
     {
       "epoch": 0.10231660231660232,
-      "grad_norm": 5.216198444366455,
+      "grad_norm": 5.6192169189453125,
       "learning_rate": 1.999921775988505e-05,
-      "loss": 1.9478,
+      "loss": 1.9397,
       "step": 106
     },
     {
       "epoch": 0.10328185328185328,
-      "grad_norm": 4.252792835235596,
+      "grad_norm": 6.623202800750732,
       "learning_rate": 1.99990819563875e-05,
-      "loss": 1.8523,
+      "loss": 1.8531,
       "step": 107
     },
     {
       "epoch": 0.10424710424710425,
-      "grad_norm": 5.397847652435303,
+      "grad_norm": 4.153651237487793,
       "learning_rate": 1.999893528930066e-05,
-      "loss": 1.8539,
+      "loss": 1.8391,
       "step": 108
     },
     {
       "epoch": 0.10521235521235521,
-      "grad_norm": 4.538277626037598,
+      "grad_norm": 5.859450340270996,
       "learning_rate": 1.999877775878388e-05,
-      "loss": 1.8018,
+      "loss": 1.7947,
       "step": 109
     },
     {
       "epoch": 0.10617760617760617,
-      "grad_norm": 4.547168731689453,
+      "grad_norm": 5.183798313140869,
       "learning_rate": 1.9998609365008313e-05,
-      "loss": 1.7968,
+      "loss": 1.7853,
       "step": 110
     },
     {
       "epoch": 0.10714285714285714,
-      "grad_norm": 4.570284366607666,
+      "grad_norm": 4.710840225219727,
       "learning_rate": 1.999843010815691e-05,
-      "loss": 1.737,
+      "loss": 1.7262,
       "step": 111
     },
     {
       "epoch": 0.10810810810810811,
-      "grad_norm": 5.048513412475586,
+      "grad_norm": 5.227701187133789,
       "learning_rate": 1.9998239988424425e-05,
-      "loss": 1.8962,
+      "loss": 1.8857,
       "step": 112
     },
     {
       "epoch": 0.10907335907335908,
-      "grad_norm": 4.87770938873291,
+      "grad_norm": 4.914846897125244,
       "learning_rate": 1.9998039006017414e-05,
-      "loss": 1.7879,
+      "loss": 1.7841,
       "step": 113
     },
     {
       "epoch": 0.11003861003861004,
-      "grad_norm": 4.417995929718018,
+      "grad_norm": 5.27622127532959,
       "learning_rate": 1.999782716115424e-05,
-      "loss": 1.8342,
+      "loss": 1.8224,
       "step": 114
     },
     {
       "epoch": 0.111003861003861,
-      "grad_norm": 5.057574272155762,
+      "grad_norm": 4.174304008483887,
       "learning_rate": 1.9997604454065063e-05,
-      "loss": 1.8057,
+      "loss": 1.8011,
       "step": 115
     },
     {
       "epoch": 0.11196911196911197,
-      "grad_norm": 4.49242639541626,
+      "grad_norm": 5.09800910949707,
       "learning_rate": 1.9997370884991842e-05,
-      "loss": 1.9251,
+      "loss": 1.9171,
       "step": 116
     },
     {
       "epoch": 0.11293436293436293,
-      "grad_norm": 8.180551528930664,
+      "grad_norm": 4.4967217445373535,
       "learning_rate": 1.9997126454188343e-05,
-      "loss": 1.8771,
+      "loss": 1.8608,
       "step": 117
     },
     {
       "epoch": 0.1138996138996139,
-      "grad_norm": 4.551602840423584,
+      "grad_norm": 4.238484859466553,
       "learning_rate": 1.999687116192013e-05,
-      "loss": 1.8492,
+      "loss": 1.8401,
       "step": 118
     },
     {
       "epoch": 0.11486486486486487,
-      "grad_norm": 8.785762786865234,
+      "grad_norm": 5.3905510902404785,
       "learning_rate": 1.9996605008464564e-05,
-      "loss": 1.8796,
+      "loss": 1.8566,
       "step": 119
     },
     {
       "epoch": 0.11583011583011583,
-      "grad_norm": 4.232022762298584,
+      "grad_norm": 4.837869167327881,
       "learning_rate": 1.999632799411081e-05,
-      "loss": 1.85,
+      "loss": 1.8418,
       "step": 120
     },
     {
       "epoch": 0.1167953667953668,
-      "grad_norm": 8.051064491271973,
+      "grad_norm": 4.974925518035889,
       "learning_rate": 1.9996040119159842e-05,
-      "loss": 1.7762,
+      "loss": 1.7553,
       "step": 121
     },
     {
       "epoch": 0.11776061776061776,
-      "grad_norm": 4.3684000968933105,
+      "grad_norm": 4.875973224639893,
       "learning_rate": 1.9995741383924413e-05,
-      "loss": 1.8294,
+      "loss": 1.8202,
       "step": 122
     },
     {
       "epoch": 0.11872586872586872,
-      "grad_norm": 7.8162841796875,
+      "grad_norm": 5.863516807556152,
       "learning_rate": 1.999543178872909e-05,
-      "loss": 1.8098,
+      "loss": 1.8035,
       "step": 123
     },
     {
       "epoch": 0.11969111969111969,
-      "grad_norm": 4.503332138061523,
+      "grad_norm": 4.682519435882568,
       "learning_rate": 1.9995111333910238e-05,
-      "loss": 1.84,
+      "loss": 1.8362,
       "step": 124
     },
     {
       "epoch": 0.12065637065637065,
-      "grad_norm": 8.560187339782715,
+      "grad_norm": 7.652477264404297,
       "learning_rate": 1.999478001981602e-05,
-      "loss": 1.8667,
+      "loss": 1.8581,
       "step": 125
     },
     {
       "epoch": 0.12162162162162163,
-      "grad_norm": 4.4390788078308105,
+      "grad_norm": 4.367491245269775,
       "learning_rate": 1.9994437846806384e-05,
-      "loss": 1.8341,
+      "loss": 1.8287,
       "step": 126
     },
     {
       "epoch": 0.12258687258687259,
-      "grad_norm": 5.524573802947998,
+      "grad_norm": 5.3426899909973145,
       "learning_rate": 1.99940848152531e-05,
-      "loss": 1.8746,
+      "loss": 1.8678,
       "step": 127
     },
     {
       "epoch": 0.12355212355212356,
-      "grad_norm": 4.953189849853516,
+      "grad_norm": 4.7820611000061035,
       "learning_rate": 1.999372092553971e-05,
-      "loss": 1.8883,
+      "loss": 1.8896,
       "step": 128
     },
     {
       "epoch": 0.12451737451737452,
-      "grad_norm": 4.29864501953125,
+      "grad_norm": 4.320755958557129,
       "learning_rate": 1.9993346178061572e-05,
-      "loss": 1.7821,
+      "loss": 1.7866,
       "step": 129
     },
     {
       "epoch": 0.12548262548262548,
-      "grad_norm": 5.282871246337891,
+      "grad_norm": 4.578153610229492,
       "learning_rate": 1.9992960573225837e-05,
-      "loss": 1.7862,
+      "loss": 1.7751,
       "step": 130
     },
     {
       "epoch": 0.12644787644787644,
-      "grad_norm": 4.665815353393555,
+      "grad_norm": 4.501065254211426,
       "learning_rate": 1.9992564111451444e-05,
-      "loss": 1.8474,
+      "loss": 1.8403,
       "step": 131
     },
     {
       "epoch": 0.1274131274131274,
-      "grad_norm": 4.723055839538574,
+      "grad_norm": 4.429023265838623,
       "learning_rate": 1.999215679316913e-05,
-      "loss": 1.838,
+      "loss": 1.8273,
       "step": 132
     },
     {
       "epoch": 0.12837837837837837,
-      "grad_norm": 4.902134418487549,
+      "grad_norm": 4.49923849105835,
       "learning_rate": 1.9991738618821435e-05,
-      "loss": 1.8074,
+      "loss": 1.7947,
       "step": 133
     },
     {
       "epoch": 0.12934362934362933,
-      "grad_norm": 4.841723918914795,
+      "grad_norm": 4.509472846984863,
       "learning_rate": 1.9991309588862683e-05,
-      "loss": 1.8605,
+      "loss": 1.8452,
       "step": 134
     },
     {
       "epoch": 0.1303088803088803,
-      "grad_norm": 4.6323723793029785,
+      "grad_norm": 4.3350090980529785,
       "learning_rate": 1.9990869703759e-05,
-      "loss": 1.8063,
+      "loss": 1.7964,
       "step": 135
     },
     {
       "epoch": 0.13127413127413126,
-      "grad_norm": 4.535574436187744,
+      "grad_norm": 4.3472676277160645,
       "learning_rate": 1.9990418963988306e-05,
-      "loss": 1.881,
+      "loss": 1.8745,
       "step": 136
     },
     {
       "epoch": 0.13223938223938225,
-      "grad_norm": 4.749152183532715,
+      "grad_norm": 4.407988548278809,
       "learning_rate": 1.9989957370040305e-05,
-      "loss": 1.7487,
+      "loss": 1.7383,
       "step": 137
     },
     {
       "epoch": 0.13320463320463322,
-      "grad_norm": 4.237002849578857,
+      "grad_norm": 4.147645473480225,
       "learning_rate": 1.9989484922416503e-05,
-      "loss": 1.8002,
+      "loss": 1.79,
       "step": 138
     },
     {
       "epoch": 0.13416988416988418,
-      "grad_norm": 4.75019645690918,
+      "grad_norm": 4.456454753875732,
       "learning_rate": 1.9989001621630196e-05,
-      "loss": 1.8117,
+      "loss": 1.8061,
       "step": 139
     },
     {
       "epoch": 0.13513513513513514,
-      "grad_norm": 5.378240585327148,
+      "grad_norm": 4.734296798706055,
       "learning_rate": 1.998850746820647e-05,
-      "loss": 1.8826,
+      "loss": 1.8744,
       "step": 140
     },
     {
       "epoch": 0.1361003861003861,
-      "grad_norm": 5.388830661773682,
+      "grad_norm": 4.757154941558838,
       "learning_rate": 1.9988002462682198e-05,
-      "loss": 1.8002,
+      "loss": 1.7902,
       "step": 141
     },
     {
       "epoch": 0.13706563706563707,
-      "grad_norm": 5.59326171875,
+      "grad_norm": 5.0651092529296875,
       "learning_rate": 1.998748660560605e-05,
-      "loss": 1.8897,
+      "loss": 1.8781,
       "step": 142
     },
     {
       "epoch": 0.13803088803088803,
-      "grad_norm": 5.555648326873779,
+      "grad_norm": 5.7025532722473145,
       "learning_rate": 1.9986959897538488e-05,
-      "loss": 1.8162,
+      "loss": 1.7791,
       "step": 143
     },
     {
       "epoch": 0.138996138996139,
-      "grad_norm": 7.1515350341796875,
+      "grad_norm": 5.915772914886475,
       "learning_rate": 1.9986422339051753e-05,
-      "loss": 1.741,
+      "loss": 1.7375,
       "step": 144
     },
     {
       "epoch": 0.13996138996138996,
-      "grad_norm": 5.384510040283203,
+      "grad_norm": 6.178494930267334,
       "learning_rate": 1.9985873930729883e-05,
-      "loss": 1.8948,
+      "loss": 1.887,
       "step": 145
     },
     {
       "epoch": 0.14092664092664092,
-      "grad_norm": 7.778863430023193,
+      "grad_norm": 7.490489482879639,
       "learning_rate": 1.9985314673168696e-05,
-      "loss": 1.8543,
+      "loss": 1.8418,
       "step": 146
     },
     {
       "epoch": 0.14189189189189189,
-      "grad_norm": 6.057487964630127,
+      "grad_norm": 5.003859043121338,
       "learning_rate": 1.998474456697581e-05,
-      "loss": 1.8448,
+      "loss": 1.824,
       "step": 147
     },
     {
       "epoch": 0.14285714285714285,
-      "grad_norm": 5.474917888641357,
+      "grad_norm": 7.53109073638916,
       "learning_rate": 1.9984163612770617e-05,
-      "loss": 1.7643,
+      "loss": 1.765,
       "step": 148
     },
     {
       "epoch": 0.1438223938223938,
-      "grad_norm": 7.61667013168335,
+      "grad_norm": 5.186059951782227,
       "learning_rate": 1.9983571811184297e-05,
-      "loss": 1.8911,
+      "loss": 1.8715,
       "step": 149
     },
     {
       "epoch": 0.14478764478764478,
-      "grad_norm": 4.5976643562316895,
+      "grad_norm": 4.9835357666015625,
       "learning_rate": 1.9982969162859828e-05,
-      "loss": 1.815,
+      "loss": 1.8022,
       "step": 150
     },
     {
       "epoch": 0.14575289575289574,
-      "grad_norm": 6.352999687194824,
+      "grad_norm": 6.683533191680908,
       "learning_rate": 1.9982355668451948e-05,
-      "loss": 1.8259,
+      "loss": 1.8176,
       "step": 151
     },
     {
       "epoch": 0.14671814671814673,
-      "grad_norm": 5.403079509735107,
+      "grad_norm": 4.464743614196777,
       "learning_rate": 1.9981731328627208e-05,
-      "loss": 1.7196,
+      "loss": 1.711,
       "step": 152
     },
     {
       "epoch": 0.1476833976833977,
-      "grad_norm": 4.54139518737793,
+      "grad_norm": 7.062692642211914,
       "learning_rate": 1.9981096144063916e-05,
-      "loss": 1.8072,
+      "loss": 1.8094,
       "step": 153
     },
     {
       "epoch": 0.14864864864864866,
-      "grad_norm": 4.9963507652282715,
+      "grad_norm": 4.416406631469727,
       "learning_rate": 1.9980450115452183e-05,
-      "loss": 1.8265,
+      "loss": 1.8192,
       "step": 154
     },
     {
       "epoch": 0.14961389961389962,
-      "grad_norm": 4.345424175262451,
+      "grad_norm": 5.021982192993164,
       "learning_rate": 1.9979793243493887e-05,
-      "loss": 1.8207,
+      "loss": 1.8204,
       "step": 155
     },
     {
       "epoch": 0.15057915057915058,
-      "grad_norm": 4.619804859161377,
+      "grad_norm": 5.424167156219482,
       "learning_rate": 1.9979125528902688e-05,
-      "loss": 1.7745,
+      "loss": 1.7607,
       "step": 156
     },
     {
       "epoch": 0.15154440154440155,
-      "grad_norm": 5.153815746307373,
+      "grad_norm": 4.938490867614746,
       "learning_rate": 1.9978446972404043e-05,
-      "loss": 1.8312,
+      "loss": 1.822,
       "step": 157
     },
     {
       "epoch": 0.1525096525096525,
-      "grad_norm": 4.700374126434326,
+      "grad_norm": 6.113457679748535,
       "learning_rate": 1.9977757574735164e-05,
-      "loss": 1.8808,
+      "loss": 1.871,
       "step": 158
     },
     {
       "epoch": 0.15347490347490347,
-      "grad_norm": 4.401470184326172,
+      "grad_norm": 5.008244514465332,
       "learning_rate": 1.9977057336645057e-05,
-      "loss": 1.8824,
+      "loss": 1.8726,
       "step": 159
     },
     {
       "epoch": 0.15444015444015444,
-      "grad_norm": 4.030679225921631,
+      "grad_norm": 4.41956901550293,
       "learning_rate": 1.9976346258894502e-05,
-      "loss": 1.8622,
+      "loss": 1.8546,
       "step": 160
     },
     {
       "epoch": 0.1554054054054054,
-      "grad_norm": 5.170019149780273,
+      "grad_norm": 4.555746078491211,
       "learning_rate": 1.9975624342256057e-05,
-      "loss": 1.8153,
+      "loss": 1.8035,
       "step": 161
     },
     {
       "epoch": 0.15637065637065636,
-      "grad_norm": 4.371926307678223,
+      "grad_norm": 4.574209690093994,
       "learning_rate": 1.997489158751405e-05,
-      "loss": 1.8585,
+      "loss": 1.8562,
       "step": 162
     },
     {
       "epoch": 0.15733590733590733,
-      "grad_norm": 5.159310340881348,
+      "grad_norm": 4.344732284545898,
       "learning_rate": 1.9974147995464593e-05,
-      "loss": 1.838,
+      "loss": 1.8295,
       "step": 163
     },
     {
       "epoch": 0.1583011583011583,
-      "grad_norm": 4.847809314727783,
+      "grad_norm": 4.424563407897949,
       "learning_rate": 1.997339356691557e-05,
-      "loss": 1.8734,
+      "loss": 1.8641,
       "step": 164
     },
     {
       "epoch": 0.15926640926640925,
-      "grad_norm": 4.6886305809021,
+      "grad_norm": 4.062377452850342,
       "learning_rate": 1.997262830268663e-05,
-      "loss": 1.7985,
+      "loss": 1.7869,
       "step": 165
     },
     {
       "epoch": 0.16023166023166024,
-      "grad_norm": 4.294079780578613,
+      "grad_norm": 4.342826843261719,
       "learning_rate": 1.9971852203609204e-05,
-      "loss": 1.8025,
+      "loss": 1.793,
       "step": 166
     },
     {
       "epoch": 0.1611969111969112,
-      "grad_norm": 5.942923069000244,
+      "grad_norm": 4.555099964141846,
       "learning_rate": 1.9971065270526494e-05,
-      "loss": 1.8206,
+      "loss": 1.8094,
       "step": 167
     },
     {
       "epoch": 0.16216216216216217,
-      "grad_norm": 4.145232200622559,
+      "grad_norm": 4.240360736846924,
       "learning_rate": 1.9970267504293466e-05,
-      "loss": 1.8218,
+      "loss": 1.8171,
       "step": 168
     },
     {
       "epoch": 0.16312741312741313,
-      "grad_norm": 5.085411548614502,
+      "grad_norm": 4.916861534118652,
       "learning_rate": 1.996945890577686e-05,
-      "loss": 1.8176,
+      "loss": 1.8028,
       "step": 169
     },
     {
       "epoch": 0.1640926640926641,
-      "grad_norm": 5.151740074157715,
+      "grad_norm": 4.524811267852783,
       "learning_rate": 1.9968639475855192e-05,
-      "loss": 1.8552,
+      "loss": 1.8516,
       "step": 170
     },
     {
       "epoch": 0.16505791505791506,
-      "grad_norm": 4.554969310760498,
+      "grad_norm": 4.570613861083984,
       "learning_rate": 1.996780921541873e-05,
-      "loss": 1.8348,
+      "loss": 1.8224,
       "step": 171
     },
     {
       "epoch": 0.16602316602316602,
-      "grad_norm": 5.52618408203125,
+      "grad_norm": 4.703306674957275,
       "learning_rate": 1.9966968125369522e-05,
-      "loss": 1.8799,
+      "loss": 1.8648,
       "step": 172
     },
     {
       "epoch": 0.166988416988417,
-      "grad_norm": 4.414917469024658,
+      "grad_norm": 4.607660293579102,
       "learning_rate": 1.996611620662138e-05,
-      "loss": 1.7495,
+      "loss": 1.7432,
       "step": 173
     },
     {
       "epoch": 0.16795366795366795,
-      "grad_norm": 4.095779895782471,
+      "grad_norm": 3.978327512741089,
       "learning_rate": 1.9965253460099872e-05,
-      "loss": 1.7518,
+      "loss": 1.7449,
       "step": 174
     },
     {
       "epoch": 0.16891891891891891,
-      "grad_norm": 4.854794979095459,
+      "grad_norm": 4.030558109283447,
       "learning_rate": 1.9964379886742344e-05,
-      "loss": 1.7869,
+      "loss": 1.7754,
       "step": 175
     },
     {
       "epoch": 0.16988416988416988,
-      "grad_norm": 5.24516487121582,
+      "grad_norm": 5.352463245391846,
       "learning_rate": 1.996349548749789e-05,
-      "loss": 1.8371,
+      "loss": 1.8339,
       "step": 176
     },
     {
       "epoch": 0.17084942084942084,
-      "grad_norm": 4.558859825134277,
+      "grad_norm": 4.64077091217041,
       "learning_rate": 1.9962600263327376e-05,
-      "loss": 1.8612,
+      "loss": 1.8595,
       "step": 177
     },
     {
       "epoch": 0.1718146718146718,
-      "grad_norm": 4.1520538330078125,
+      "grad_norm": 4.549771308898926,
       "learning_rate": 1.9961694215203424e-05,
-      "loss": 1.859,
+      "loss": 1.8583,
       "step": 178
     },
     {
       "epoch": 0.17277992277992277,
-      "grad_norm": 5.53354024887085,
+      "grad_norm": 4.953789234161377,
       "learning_rate": 1.9960777344110426e-05,
-      "loss": 1.8372,
+      "loss": 1.8327,
       "step": 179
     },
     {
       "epoch": 0.17374517374517376,
-      "grad_norm": 4.267213821411133,
+      "grad_norm": 5.258177757263184,
       "learning_rate": 1.9959849651044515e-05,
-      "loss": 1.7665,
+      "loss": 1.7596,
       "step": 180
     },
     {
       "epoch": 0.17471042471042472,
-      "grad_norm": 4.841174602508545,
+      "grad_norm": 4.6174516677856445,
       "learning_rate": 1.9958911137013596e-05,
-      "loss": 1.8199,
+      "loss": 1.8102,
       "step": 181
     },
     {
       "epoch": 0.17567567567567569,
-      "grad_norm": 5.410240173339844,
+      "grad_norm": 5.28028678894043,
       "learning_rate": 1.9957961803037325e-05,
-      "loss": 1.8309,
+      "loss": 1.8249,
       "step": 182
     },
     {
       "epoch": 0.17664092664092665,
-      "grad_norm": 5.265180587768555,
+      "grad_norm": 5.261460781097412,
       "learning_rate": 1.9957001650147116e-05,
-      "loss": 1.8111,
+      "loss": 1.7908,
       "step": 183
     },
     {
       "epoch": 0.1776061776061776,
-      "grad_norm": 4.3163251876831055,
+      "grad_norm": 4.200467109680176,
       "learning_rate": 1.9956030679386132e-05,
-      "loss": 1.8032,
+      "loss": 1.7972,
       "step": 184
     },
     {
       "epoch": 0.17857142857142858,
-      "grad_norm": 5.663985252380371,
+      "grad_norm": 6.062233924865723,
       "learning_rate": 1.9955048891809293e-05,
-      "loss": 1.7645,
+      "loss": 1.7592,
       "step": 185
     },
     {
       "epoch": 0.17953667953667954,
-      "grad_norm": 5.375406742095947,
+      "grad_norm": 4.287155628204346,
       "learning_rate": 1.9954056288483273e-05,
-      "loss": 1.8664,
+      "loss": 1.8521,
       "step": 186
     },
     {
       "epoch": 0.1805019305019305,
-      "grad_norm": 5.803011894226074,
+      "grad_norm": 6.471398830413818,
       "learning_rate": 1.9953052870486492e-05,
-      "loss": 1.69,
+      "loss": 1.6832,
       "step": 187
     },
     {
       "epoch": 0.18146718146718147,
-      "grad_norm": 5.244572162628174,
+      "grad_norm": 4.670758247375488,
       "learning_rate": 1.995203863890912e-05,
-      "loss": 1.8374,
+      "loss": 1.8153,
       "step": 188
     },
     {
       "epoch": 0.18243243243243243,
-      "grad_norm": 5.326415061950684,
+      "grad_norm": 6.0269293785095215,
       "learning_rate": 1.9951013594853087e-05,
-      "loss": 1.8369,
+      "loss": 1.8321,
       "step": 189
     },
     {
       "epoch": 0.1833976833976834,
-      "grad_norm": 3.968106985092163,
+      "grad_norm": 4.165921688079834,
       "learning_rate": 1.9949977739432052e-05,
-      "loss": 1.7869,
+      "loss": 1.7842,
       "step": 190
     },
     {
       "epoch": 0.18436293436293436,
-      "grad_norm": 6.265425205230713,
+      "grad_norm": 5.985523223876953,
       "learning_rate": 1.9948931073771433e-05,
-      "loss": 1.8038,
+      "loss": 1.8011,
       "step": 191
     },
     {
       "epoch": 0.18532818532818532,
-      "grad_norm": 4.435483932495117,
+      "grad_norm": 4.428592681884766,
       "learning_rate": 1.9947873599008387e-05,
-      "loss": 1.7796,
+      "loss": 1.7724,
       "step": 192
     },
     {
       "epoch": 0.18629343629343628,
-      "grad_norm": 4.4179887771606445,
+      "grad_norm": 5.086099147796631,
       "learning_rate": 1.9946805316291817e-05,
-      "loss": 1.7582,
+      "loss": 1.7536,
       "step": 193
     },
     {
       "epoch": 0.18725868725868725,
-      "grad_norm": 5.019647121429443,
+      "grad_norm": 5.060903072357178,
       "learning_rate": 1.994572622678237e-05,
-      "loss": 1.7806,
+      "loss": 1.7784,
       "step": 194
     },
     {
       "epoch": 0.18822393822393824,
-      "grad_norm": 4.530157566070557,
+      "grad_norm": 4.4722747802734375,
       "learning_rate": 1.994463633165243e-05,
-      "loss": 1.8291,
+      "loss": 1.8375,
       "step": 195
     },
     {
       "epoch": 0.1891891891891892,
-      "grad_norm": 4.435397624969482,
+      "grad_norm": 4.909644603729248,
       "learning_rate": 1.9943535632086125e-05,
-      "loss": 1.8541,
+      "loss": 1.8479,
       "step": 196
     },
     {
       "epoch": 0.19015444015444016,
-      "grad_norm": 4.224645137786865,
+      "grad_norm": 4.263146877288818,
       "learning_rate": 1.9942424129279318e-05,
-      "loss": 1.7747,
+      "loss": 1.7631,
       "step": 197
     },
     {
       "epoch": 0.19111969111969113,
-      "grad_norm": 4.534023284912109,
+      "grad_norm": 4.199649333953857,
       "learning_rate": 1.9941301824439608e-05,
-      "loss": 1.821,
+      "loss": 1.8072,
       "step": 198
     },
     {
       "epoch": 0.1920849420849421,
-      "grad_norm": 4.295060634613037,
+      "grad_norm": 4.226156711578369,
       "learning_rate": 1.994016871878634e-05,
-      "loss": 1.7816,
+      "loss": 1.7705,
       "step": 199
     },
     {
       "epoch": 0.19305019305019305,
-      "grad_norm": 4.451807498931885,
+      "grad_norm": 4.450245380401611,
       "learning_rate": 1.993902481355058e-05,
-      "loss": 1.8481,
+      "loss": 1.847,
       "step": 200
     },
     {
       "epoch": 0.19401544401544402,
-      "grad_norm": 4.234887599945068,
+      "grad_norm": 3.9832916259765625,
       "learning_rate": 1.9937870109975138e-05,
-      "loss": 1.8334,
+      "loss": 1.8176,
       "step": 201
     },
     {
       "epoch": 0.19498069498069498,
-      "grad_norm": 4.383005619049072,
+      "grad_norm": 4.105828762054443,
       "learning_rate": 1.9936704609314553e-05,
-      "loss": 1.85,
+      "loss": 1.8439,
       "step": 202
     },
     {
       "epoch": 0.19594594594594594,
-      "grad_norm": 4.373894691467285,
+      "grad_norm": 4.667481899261475,
       "learning_rate": 1.9935528312835087e-05,
-      "loss": 1.844,
+      "loss": 1.8369,
       "step": 203
     },
     {
       "epoch": 0.1969111969111969,
-      "grad_norm": 4.654743671417236,
+      "grad_norm": 4.9099345207214355,
       "learning_rate": 1.993434122181474e-05,
-      "loss": 1.893,
+      "loss": 1.8803,
       "step": 204
     },
     {
       "epoch": 0.19787644787644787,
-      "grad_norm": 7.665431499481201,
+      "grad_norm": 5.737005233764648,
       "learning_rate": 1.9933143337543238e-05,
-      "loss": 1.7608,
+      "loss": 1.7546,
       "step": 205
     },
     {
       "epoch": 0.19884169884169883,
-      "grad_norm": 4.686855792999268,
+      "grad_norm": 4.749411106109619,
       "learning_rate": 1.9931934661322035e-05,
-      "loss": 1.8406,
+      "loss": 1.8395,
       "step": 206
     },
     {
       "epoch": 0.1998069498069498,
-      "grad_norm": 6.368899822235107,
+      "grad_norm": 5.484257698059082,
       "learning_rate": 1.9930715194464304e-05,
-      "loss": 1.7996,
+      "loss": 1.7914,
       "step": 207
     },
     {
       "epoch": 0.20077220077220076,
-      "grad_norm": 5.041531085968018,
+      "grad_norm": 4.37596321105957,
       "learning_rate": 1.992948493829495e-05,
-      "loss": 1.8371,
+      "loss": 1.8365,
       "step": 208
     },
     {
       "epoch": 0.20173745173745175,
-      "grad_norm": 3.9813947677612305,
+      "grad_norm": 4.335319519042969,
       "learning_rate": 1.9928243894150586e-05,
       "loss": 1.8108,
       "step": 209
     },
     {
       "epoch": 0.20270270270270271,
-      "grad_norm": 4.69705867767334,
+      "grad_norm": 4.358355522155762,
       "learning_rate": 1.9926992063379568e-05,
-      "loss": 1.788,
+      "loss": 1.7811,
       "step": 210
     },
     {
       "epoch": 0.20366795366795368,
-      "grad_norm": 5.991170883178711,
+      "grad_norm": 4.676483631134033,
       "learning_rate": 1.992572944734195e-05,
-      "loss": 1.8376,
+      "loss": 1.8313,
       "step": 211
     },
     {
       "epoch": 0.20463320463320464,
-      "grad_norm": 4.168196201324463,
+      "grad_norm": 4.090336799621582,
       "learning_rate": 1.9924456047409517e-05,
-      "loss": 1.8825,
+      "loss": 1.8763,
       "step": 212
     },
     {
       "epoch": 0.2055984555984556,
-      "grad_norm": 5.6147284507751465,
+      "grad_norm": 4.695727825164795,
       "learning_rate": 1.992317186496576e-05,
-      "loss": 1.8339,
+      "loss": 1.8221,
       "step": 213
     },
     {
       "epoch": 0.20656370656370657,
-      "grad_norm": 4.425781726837158,
+      "grad_norm": 4.306410789489746,
       "learning_rate": 1.9921876901405894e-05,
-      "loss": 1.7542,
+      "loss": 1.744,
       "step": 214
     },
     {
       "epoch": 0.20752895752895753,
-      "grad_norm": 5.102436065673828,
+      "grad_norm": 4.832817077636719,
       "learning_rate": 1.9920571158136837e-05,
-      "loss": 1.8345,
+      "loss": 1.828,
       "step": 215
     },
     {
       "epoch": 0.2084942084942085,
-      "grad_norm": 4.2920379638671875,
+      "grad_norm": 4.055543899536133,
       "learning_rate": 1.9919254636577235e-05,
-      "loss": 1.7899,
+      "loss": 1.7823,
       "step": 216
     },
     {
       "epoch": 0.20945945945945946,
-      "grad_norm": 4.998841285705566,
+      "grad_norm": 4.983643054962158,
       "learning_rate": 1.9917927338157428e-05,
-      "loss": 1.7901,
+      "loss": 1.7826,
       "step": 217
     },
     {
       "epoch": 0.21042471042471042,
-      "grad_norm": 4.545851707458496,
+      "grad_norm": 4.454193115234375,
       "learning_rate": 1.9916589264319475e-05,
-      "loss": 1.8788,
+      "loss": 1.8724,
       "step": 218
     },
     {
       "epoch": 0.21138996138996138,
-      "grad_norm": 4.769996643066406,
+      "grad_norm": 4.9410014152526855,
       "learning_rate": 1.9915240416517134e-05,
-      "loss": 1.7339,
+      "loss": 1.7248,
       "step": 219
     },
     {
       "epoch": 0.21235521235521235,
-      "grad_norm": 5.561061859130859,
+      "grad_norm": 5.156111240386963,
       "learning_rate": 1.991388079621587e-05,
-      "loss": 1.8551,
+      "loss": 1.8371,
       "step": 220
     },
     {
       "epoch": 0.2133204633204633,
-      "grad_norm": 5.276782035827637,
+      "grad_norm": 5.075760841369629,
       "learning_rate": 1.9912510404892865e-05,
-      "loss": 1.7999,
+      "loss": 1.7924,
       "step": 221
     },
     {
       "epoch": 0.21428571428571427,
-      "grad_norm": 6.434167861938477,
+      "grad_norm": 5.788449764251709,
       "learning_rate": 1.991112924403698e-05,
-      "loss": 1.8323,
+      "loss": 1.8221,
       "step": 222
     },
     {
       "epoch": 0.21525096525096524,
-      "grad_norm": 5.014518737792969,
+      "grad_norm": 5.00429630279541,
       "learning_rate": 1.9909737315148798e-05,
-      "loss": 1.7433,
+      "loss": 1.7391,
       "step": 223
     },
     {
       "epoch": 0.21621621621621623,
-      "grad_norm": 6.130193710327148,
+      "grad_norm": 5.11868143081665,
       "learning_rate": 1.9908334619740587e-05,
-      "loss": 1.7799,
+      "loss": 1.7661,
       "step": 224
     },
     {
       "epoch": 0.2171814671814672,
-      "grad_norm": 3.872027635574341,
+      "grad_norm": 3.729013681411743,
       "learning_rate": 1.9906921159336318e-05,
-      "loss": 1.8152,
+      "loss": 1.8091,
       "step": 225
     },
     {
       "epoch": 0.21814671814671815,
-      "grad_norm": 6.342243671417236,
+      "grad_norm": 5.980942249298096,
       "learning_rate": 1.990549693547166e-05,
-      "loss": 1.8283,
+      "loss": 1.8184,
       "step": 226
     },
     {
       "epoch": 0.21911196911196912,
-      "grad_norm": 4.687882423400879,
+      "grad_norm": 4.190624237060547,
       "learning_rate": 1.9904061949693972e-05,
-      "loss": 1.8143,
+      "loss": 1.8044,
       "step": 227
     },
     {
       "epoch": 0.22007722007722008,
-      "grad_norm": 4.088676929473877,
+      "grad_norm": 4.453525066375732,
       "learning_rate": 1.9902616203562307e-05,
-      "loss": 1.7731,
+      "loss": 1.771,
       "step": 228
     },
     {
       "epoch": 0.22104247104247104,
-      "grad_norm": 5.394243240356445,
+      "grad_norm": 4.810642719268799,
       "learning_rate": 1.990115969864741e-05,
-      "loss": 1.834,
+      "loss": 1.8217,
       "step": 229
     },
     {
       "epoch": 0.222007722007722,
-      "grad_norm": 4.424513816833496,
+      "grad_norm": 4.197521209716797,
       "learning_rate": 1.989969243653171e-05,
-      "loss": 1.8217,
+      "loss": 1.8154,
       "step": 230
     },
     {
       "epoch": 0.22297297297297297,
-      "grad_norm": 4.55377197265625,
+      "grad_norm": 4.5611348152160645,
       "learning_rate": 1.989821441880933e-05,
-      "loss": 1.7464,
+      "loss": 1.7374,
       "step": 231
     },
     {
       "epoch": 0.22393822393822393,
-      "grad_norm": 4.709877014160156,
+      "grad_norm": 4.286485195159912,
       "learning_rate": 1.9896725647086074e-05,
-      "loss": 1.7462,
+      "loss": 1.7354,
       "step": 232
     },
     {
       "epoch": 0.2249034749034749,
-      "grad_norm": 4.184065818786621,
+      "grad_norm": 3.962979316711426,
       "learning_rate": 1.9895226122979426e-05,
-      "loss": 1.8002,
+      "loss": 1.7948,
       "step": 233
     },
     {
       "epoch": 0.22586872586872586,
-      "grad_norm": 4.604548931121826,
+      "grad_norm": 4.495889663696289,
       "learning_rate": 1.9893715848118566e-05,
-      "loss": 1.7847,
+      "loss": 1.7806,
       "step": 234
     },
     {
       "epoch": 0.22683397683397682,
-      "grad_norm": 5.295757293701172,
+      "grad_norm": 3.8808040618896484,
       "learning_rate": 1.9892194824144338e-05,
-      "loss": 1.817,
+      "loss": 1.8062,
       "step": 235
     },
     {
       "epoch": 0.2277992277992278,
-      "grad_norm": 4.350464344024658,
+      "grad_norm": 4.682729721069336,
       "learning_rate": 1.9890663052709276e-05,
-      "loss": 1.7864,
+      "loss": 1.7804,
       "step": 236
     },
     {
       "epoch": 0.22876447876447875,
-      "grad_norm": 6.103113651275635,
+      "grad_norm": 4.928658962249756,
       "learning_rate": 1.9889120535477584e-05,
-      "loss": 1.8386,
+      "loss": 1.8244,
       "step": 237
     },
     {
       "epoch": 0.22972972972972974,
-      "grad_norm": 4.705803871154785,
+      "grad_norm": 6.492860794067383,
       "learning_rate": 1.9887567274125143e-05,
-      "loss": 1.7247,
+      "loss": 1.7256,
       "step": 238
     },
     {
       "epoch": 0.2306949806949807,
-      "grad_norm": 4.0369768142700195,
+      "grad_norm": 3.87528657913208,
       "learning_rate": 1.988600327033951e-05,
-      "loss": 1.811,
+      "loss": 1.8017,
       "step": 239
     },
     {
       "epoch": 0.23166023166023167,
-      "grad_norm": 4.492372512817383,
+      "grad_norm": 4.363749027252197,
       "learning_rate": 1.9884428525819906e-05,
-      "loss": 1.808,
+      "loss": 1.8015,
       "step": 240
     },
     {
       "epoch": 0.23262548262548263,
-      "grad_norm": 4.9545722007751465,
+      "grad_norm": 5.678363800048828,
       "learning_rate": 1.988284304227723e-05,
-      "loss": 1.7934,
+      "loss": 1.7929,
       "step": 241
     },
     {
       "epoch": 0.2335907335907336,
-      "grad_norm": 3.98966646194458,
+      "grad_norm": 3.910369634628296,
       "learning_rate": 1.9881246821434047e-05,
-      "loss": 1.7916,
+      "loss": 1.7841,
       "step": 242
     },
     {
       "epoch": 0.23455598455598456,
-      "grad_norm": 4.651267051696777,
+      "grad_norm": 5.802454471588135,
       "learning_rate": 1.9879639865024573e-05,
-      "loss": 1.7568,
+      "loss": 1.7575,
       "step": 243
     },
     {
       "epoch": 0.23552123552123552,
-      "grad_norm": 4.708108901977539,
+      "grad_norm": 4.077709197998047,
       "learning_rate": 1.9878022174794714e-05,
-      "loss": 1.711,
+      "loss": 1.7051,
       "step": 244
     },
     {
       "epoch": 0.23648648648648649,
-      "grad_norm": 4.429472923278809,
+      "grad_norm": 4.2989959716796875,
       "learning_rate": 1.9876393752502013e-05,
-      "loss": 1.8032,
+      "loss": 1.7983,
       "step": 245
     },
     {
       "epoch": 0.23745173745173745,
-      "grad_norm": 4.605119228363037,
+      "grad_norm": 5.0080885887146,
       "learning_rate": 1.987475459991569e-05,
-      "loss": 1.8229,
+      "loss": 1.8153,
       "step": 246
     },
     {
       "epoch": 0.2384169884169884,
-      "grad_norm": 5.334287166595459,
+      "grad_norm": 4.696242332458496,
       "learning_rate": 1.9873104718816613e-05,
-      "loss": 1.7261,
+      "loss": 1.7212,
       "step": 247
     },
     {
       "epoch": 0.23938223938223938,
-      "grad_norm": 3.8895561695098877,
+      "grad_norm": 4.104733467102051,
       "learning_rate": 1.987144411099731e-05,
-      "loss": 1.7948,
+      "loss": 1.7874,
       "step": 248
     },
     {
       "epoch": 0.24034749034749034,
-      "grad_norm": 5.1954522132873535,
+      "grad_norm": 5.331149578094482,
       "learning_rate": 1.9869772778261967e-05,
-      "loss": 1.7504,
+      "loss": 1.7445,
       "step": 249
     },
     {
       "epoch": 0.2413127413127413,
-      "grad_norm": 4.352097988128662,
+      "grad_norm": 3.845475673675537,
       "learning_rate": 1.986809072242641e-05,
-      "loss": 1.7795,
+      "loss": 1.777,
       "step": 250
     },
     {
       "epoch": 0.24227799227799227,
-      "grad_norm": 3.8593056201934814,
+      "grad_norm": 3.929105758666992,
       "learning_rate": 1.9866397945318133e-05,
-      "loss": 1.83,
+      "loss": 1.8233,
       "step": 251
     },
     {
       "epoch": 0.24324324324324326,
-      "grad_norm": 4.728880882263184,
+      "grad_norm": 4.144432067871094,
       "learning_rate": 1.986469444877626e-05,
-      "loss": 1.7474,
+      "loss": 1.7436,
       "step": 252
     },
     {
       "epoch": 0.24420849420849422,
-      "grad_norm": 5.145429611206055,
+      "grad_norm": 4.668457984924316,
       "learning_rate": 1.9862980234651574e-05,
-      "loss": 1.7845,
+      "loss": 1.7751,
       "step": 253
     },
     {
       "epoch": 0.24517374517374518,
-      "grad_norm": 3.842902183532715,
+      "grad_norm": 3.8014016151428223,
       "learning_rate": 1.9861255304806493e-05,
-      "loss": 1.7645,
+      "loss": 1.7682,
       "step": 254
     },
     {
       "epoch": 0.24613899613899615,
-      "grad_norm": 4.171161651611328,
+      "grad_norm": 4.098991870880127,
       "learning_rate": 1.9859519661115087e-05,
-      "loss": 1.7954,
+      "loss": 1.7838,
       "step": 255
     },
     {
       "epoch": 0.2471042471042471,
-      "grad_norm": 4.398767471313477,
+      "grad_norm": 4.3869757652282715,
       "learning_rate": 1.985777330546306e-05,
-      "loss": 1.8025,
+      "loss": 1.8019,
       "step": 256
     },
     {
       "epoch": 0.24806949806949807,
-      "grad_norm": 4.435847282409668,
+      "grad_norm": 4.22720193862915,
       "learning_rate": 1.9856016239747754e-05,
-      "loss": 1.7602,
+      "loss": 1.7458,
       "step": 257
     },
     {
       "epoch": 0.24903474903474904,
-      "grad_norm": 3.882884979248047,
+      "grad_norm": 4.103383541107178,
       "learning_rate": 1.985424846587815e-05,
-      "loss": 1.7644,
+      "loss": 1.7551,
       "step": 258
     },
     {
       "epoch": 0.25,
-      "grad_norm": 4.172622203826904,
+      "grad_norm": 4.036272048950195,
       "learning_rate": 1.985246998577486e-05,
-      "loss": 1.7587,
+      "loss": 1.7533,
       "step": 259
     },
     {
       "epoch": 0.25096525096525096,
-      "grad_norm": 4.639303207397461,
+      "grad_norm": 4.93221378326416,
       "learning_rate": 1.985068080137013e-05,
-      "loss": 1.7984,
+      "loss": 1.7878,
       "step": 260
     },
     {
       "epoch": 0.2519305019305019,
-      "grad_norm": 3.98333477973938,
+      "grad_norm": 4.076192855834961,
       "learning_rate": 1.9848880914607834e-05,
-      "loss": 1.7535,
+      "loss": 1.7491,
       "step": 261
     },
     {
       "epoch": 0.2528957528957529,
-      "grad_norm": 4.069552898406982,
+      "grad_norm": 3.999781847000122,
       "learning_rate": 1.9847070327443476e-05,
-      "loss": 1.6939,
+      "loss": 1.6894,
       "step": 262
     },
     {
       "epoch": 0.25386100386100385,
-      "grad_norm": 5.007661819458008,
+      "grad_norm": 4.890449047088623,
       "learning_rate": 1.984524904184418e-05,
-      "loss": 1.7565,
+      "loss": 1.7524,
       "step": 263
     },
     {
       "epoch": 0.2548262548262548,
-      "grad_norm": 4.196667194366455,
+      "grad_norm": 4.391631603240967,
       "learning_rate": 1.9843417059788702e-05,
-      "loss": 1.7631,
+      "loss": 1.7688,
       "step": 264
     },
     {
       "epoch": 0.2557915057915058,
-      "grad_norm": 5.928943157196045,
+      "grad_norm": 6.642754077911377,
       "learning_rate": 1.9841574383267415e-05,
-      "loss": 1.756,
+      "loss": 1.7502,
       "step": 265
     },
     {
       "epoch": 0.25675675675675674,
-      "grad_norm": 3.984234571456909,
+      "grad_norm": 4.223073959350586,
       "learning_rate": 1.9839721014282307e-05,
-      "loss": 1.7994,
+      "loss": 1.7967,
       "step": 266
     },
     {
       "epoch": 0.2577220077220077,
-      "grad_norm": 3.9803266525268555,
+      "grad_norm": 4.617069244384766,
       "learning_rate": 1.983785695484699e-05,
-      "loss": 1.7556,
+      "loss": 1.7483,
       "step": 267
     },
     {
       "epoch": 0.25868725868725867,
-      "grad_norm": 4.372330665588379,
+      "grad_norm": 4.370770454406738,
       "learning_rate": 1.9835982206986685e-05,
-      "loss": 1.6999,
+      "loss": 1.6963,
       "step": 268
     },
     {
       "epoch": 0.25965250965250963,
-      "grad_norm": 3.8054566383361816,
+      "grad_norm": 4.219081401824951,
       "learning_rate": 1.9834096772738233e-05,
-      "loss": 1.72,
+      "loss": 1.7143,
       "step": 269
     },
     {
       "epoch": 0.2606177606177606,
-      "grad_norm": 4.114311218261719,
+      "grad_norm": 4.230241298675537,
       "learning_rate": 1.9832200654150077e-05,
-      "loss": 1.8302,
+      "loss": 1.8241,
       "step": 270
     },
     {
       "epoch": 0.26158301158301156,
-      "grad_norm": 4.022603988647461,
+      "grad_norm": 4.8200883865356445,
       "learning_rate": 1.983029385328227e-05,
-      "loss": 1.7972,
+      "loss": 1.7919,
       "step": 271
     },
     {
       "epoch": 0.2625482625482625,
-      "grad_norm": 4.409243583679199,
+      "grad_norm": 4.727015018463135,
       "learning_rate": 1.982837637220647e-05,
-      "loss": 1.7628,
+      "loss": 1.7553,
       "step": 272
     },
     {
       "epoch": 0.2635135135135135,
-      "grad_norm": 4.133562088012695,
+      "grad_norm": 4.672626495361328,
       "learning_rate": 1.982644821300595e-05,
-      "loss": 1.7827,
+      "loss": 1.7819,
       "step": 273
     },
     {
       "epoch": 0.2644787644787645,
-      "grad_norm": 5.861300945281982,
+      "grad_norm": 6.530807971954346,
       "learning_rate": 1.9824509377775567e-05,
-      "loss": 1.7821,
+      "loss": 1.7776,
       "step": 274
     },
     {
       "epoch": 0.26544401544401547,
-      "grad_norm": 4.026004791259766,
+      "grad_norm": 4.291546821594238,
       "learning_rate": 1.982255986862179e-05,
-      "loss": 1.753,
+      "loss": 1.7469,
       "step": 275
     },
     {
       "epoch": 0.26640926640926643,
-      "grad_norm": 6.394083023071289,
+      "grad_norm": 7.1855878829956055,
       "learning_rate": 1.9820599687662677e-05,
-      "loss": 1.8002,
+      "loss": 1.791,
       "step": 276
     },
     {
       "epoch": 0.2673745173745174,
-      "grad_norm": 3.9767532348632812,
+      "grad_norm": 4.064293384552002,
       "learning_rate": 1.981862883702789e-05,
-      "loss": 1.7948,
+      "loss": 1.7844,
       "step": 277
     },
     {
       "epoch": 0.26833976833976836,
-      "grad_norm": 7.201776504516602,
+      "grad_norm": 7.966151714324951,
       "learning_rate": 1.9816647318858666e-05,
-      "loss": 1.777,
+      "loss": 1.7693,
       "step": 278
     },
     {
       "epoch": 0.2693050193050193,
-      "grad_norm": 3.9289729595184326,
+      "grad_norm": 3.978511095046997,
       "learning_rate": 1.981465513530785e-05,
-      "loss": 1.7838,
+      "loss": 1.7786,
       "step": 279
     },
     {
       "epoch": 0.2702702702702703,
-      "grad_norm": 6.732141494750977,
+      "grad_norm": 7.43217134475708,
       "learning_rate": 1.981265228853987e-05,
-      "loss": 1.7321,
+      "loss": 1.7261,
       "step": 280
     },
     {
       "epoch": 0.27123552123552125,
-      "grad_norm": 5.770628929138184,
+      "grad_norm": 5.40924596786499,
       "learning_rate": 1.981063878073073e-05,
-      "loss": 1.7905,
+      "loss": 1.791,
       "step": 281
     },
     {
       "epoch": 0.2722007722007722,
-      "grad_norm": 4.88922119140625,
+      "grad_norm": 5.001474857330322,
       "learning_rate": 1.980861461406802e-05,
-      "loss": 1.7441,
+      "loss": 1.7385,
       "step": 282
     },
     {
       "epoch": 0.2731660231660232,
-      "grad_norm": 5.953951358795166,
+      "grad_norm": 6.0322136878967285,
       "learning_rate": 1.9806579790750922e-05,
-      "loss": 1.7527,
+      "loss": 1.7441,
       "step": 283
     },
     {
       "epoch": 0.27413127413127414,
-      "grad_norm": 4.962509632110596,
+      "grad_norm": 4.800754547119141,
       "learning_rate": 1.9804534312990183e-05,
-      "loss": 1.7342,
+      "loss": 1.7262,
       "step": 284
     },
     {
       "epoch": 0.2750965250965251,
-      "grad_norm": 6.674172401428223,
+      "grad_norm": 7.34820556640625,
       "learning_rate": 1.980247818300813e-05,
-      "loss": 1.7753,
+      "loss": 1.7695,
       "step": 285
     },
     {
       "epoch": 0.27606177606177607,
-      "grad_norm": 6.852287769317627,
+      "grad_norm": 6.524095058441162,
       "learning_rate": 1.9800411403038664e-05,
-      "loss": 1.7277,
+      "loss": 1.7201,
       "step": 286
     },
     {
       "epoch": 0.27702702702702703,
-      "grad_norm": 5.453775882720947,
+      "grad_norm": 6.248870849609375,
       "learning_rate": 1.9798333975327258e-05,
-      "loss": 1.8125,
+      "loss": 1.8091,
       "step": 287
     },
     {
       "epoch": 0.277992277992278,
-      "grad_norm": 7.326210021972656,
+      "grad_norm": 7.1641435623168945,
       "learning_rate": 1.9796245902130948e-05,
-      "loss": 1.7298,
+      "loss": 1.7247,
       "step": 288
     },
     {
       "epoch": 0.27895752895752896,
-      "grad_norm": 4.287469863891602,
+      "grad_norm": 4.247777462005615,
       "learning_rate": 1.979414718571834e-05,
-      "loss": 1.7778,
+      "loss": 1.7753,
       "step": 289
     },
     {
       "epoch": 0.2799227799227799,
-      "grad_norm": 6.503158092498779,
+      "grad_norm": 6.73613166809082,
       "learning_rate": 1.9792037828369607e-05,
-      "loss": 1.7942,
+      "loss": 1.7866,
       "step": 290
     },
     {
       "epoch": 0.2808880308880309,
-      "grad_norm": 6.972282886505127,
+      "grad_norm": 6.153137683868408,
       "learning_rate": 1.978991783237647e-05,
-      "loss": 1.7601,
+      "loss": 1.7547,
       "step": 291
     },
     {
       "epoch": 0.28185328185328185,
-      "grad_norm": 6.6554460525512695,
+      "grad_norm": 7.561403274536133,
       "learning_rate": 1.9787787200042224e-05,
-      "loss": 1.7281,
+      "loss": 1.7232,
       "step": 292
     },
     {
       "epoch": 0.2828185328185328,
-      "grad_norm": 4.683424472808838,
+      "grad_norm": 4.631079196929932,
       "learning_rate": 1.9785645933681713e-05,
-      "loss": 1.7916,
+      "loss": 1.7908,
       "step": 293
     },
     {
       "epoch": 0.28378378378378377,
-      "grad_norm": 3.806309700012207,
+      "grad_norm": 4.228327751159668,
       "learning_rate": 1.978349403562133e-05,
-      "loss": 1.8447,
+      "loss": 1.839,
       "step": 294
     },
     {
       "epoch": 0.28474903474903474,
-      "grad_norm": 5.332488059997559,
+      "grad_norm": 5.91856050491333,
       "learning_rate": 1.978133150819902e-05,
-      "loss": 1.7481,
+      "loss": 1.7449,
       "step": 295
     },
     {
       "epoch": 0.2857142857142857,
-      "grad_norm": 4.274888515472412,
+      "grad_norm": 4.431145668029785,
       "learning_rate": 1.9779158353764293e-05,
-      "loss": 1.7075,
+      "loss": 1.7006,
       "step": 296
     },
     {
       "epoch": 0.28667953667953666,
-      "grad_norm": 4.067286491394043,
+      "grad_norm": 4.71598482131958,
       "learning_rate": 1.977697457467818e-05,
-      "loss": 1.7307,
+      "loss": 1.7278,
       "step": 297
     },
     {
       "epoch": 0.2876447876447876,
-      "grad_norm": 4.854985237121582,
+      "grad_norm": 4.46198844909668,
       "learning_rate": 1.9774780173313273e-05,
-      "loss": 1.7905,
+      "loss": 1.7822,
       "step": 298
     },
     {
       "epoch": 0.2886100386100386,
-      "grad_norm": 4.431771278381348,
+      "grad_norm": 5.589810371398926,
       "learning_rate": 1.9772575152053696e-05,
-      "loss": 1.7558,
+      "loss": 1.747,
       "step": 299
     },
     {
       "epoch": 0.28957528957528955,
-      "grad_norm": 7.292600154876709,
+      "grad_norm": 6.826247215270996,
       "learning_rate": 1.977035951329511e-05,
-      "loss": 1.7599,
+      "loss": 1.7505,
       "step": 300
     },
     {
       "epoch": 0.2905405405405405,
-      "grad_norm": 4.510945796966553,
+      "grad_norm": 5.011523723602295,
       "learning_rate": 1.976813325944472e-05,
-      "loss": 1.8491,
+      "loss": 1.8394,
       "step": 301
     },
     {
       "epoch": 0.2915057915057915,
-      "grad_norm": 5.671480178833008,
+      "grad_norm": 6.30789852142334,
       "learning_rate": 1.9765896392921253e-05,
-      "loss": 1.7796,
+      "loss": 1.7731,
       "step": 302
     },
     {
       "epoch": 0.2924710424710425,
-      "grad_norm": 5.7752227783203125,
+      "grad_norm": 5.324947357177734,
       "learning_rate": 1.9763648916154982e-05,
-      "loss": 1.8094,
+      "loss": 1.8005,
       "step": 303
     },
     {
       "epoch": 0.29343629343629346,
-      "grad_norm": 4.039534568786621,
+      "grad_norm": 4.851571083068848,
       "learning_rate": 1.9761390831587687e-05,
-      "loss": 1.7191,
+      "loss": 1.719,
       "step": 304
     },
     {
       "epoch": 0.2944015444015444,
-      "grad_norm": 7.9651312828063965,
+      "grad_norm": 8.98678207397461,
       "learning_rate": 1.975912214167269e-05,
-      "loss": 1.7937,
+      "loss": 1.7892,
       "step": 305
     },
     {
       "epoch": 0.2953667953667954,
-      "grad_norm": 4.313594341278076,
+      "grad_norm": 3.633197546005249,
       "learning_rate": 1.975684284887483e-05,
-      "loss": 1.7014,
+      "loss": 1.6937,
       "step": 306
     },
     {
       "epoch": 0.29633204633204635,
-      "grad_norm": 4.699955463409424,
+      "grad_norm": 5.538055896759033,
       "learning_rate": 1.975455295567046e-05,
-      "loss": 1.738,
+      "loss": 1.7383,
       "step": 307
     },
     {
       "epoch": 0.2972972972972973,
-      "grad_norm": 5.305117130279541,
+      "grad_norm": 5.335306167602539,
       "learning_rate": 1.9752252464547454e-05,
-      "loss": 1.8219,
+      "loss": 1.8089,
       "step": 308
     },
     {
       "epoch": 0.2982625482625483,
-      "grad_norm": 5.326988220214844,
+      "grad_norm": 4.801578998565674,
       "learning_rate": 1.97499413780052e-05,
-      "loss": 1.8626,
+      "loss": 1.8493,
       "step": 309
     },
     {
       "epoch": 0.29922779922779924,
-      "grad_norm": 4.539814472198486,
+      "grad_norm": 5.033331394195557,
       "learning_rate": 1.9747619698554606e-05,
-      "loss": 1.7771,
+      "loss": 1.7715,
       "step": 310
     },
     {
       "epoch": 0.3001930501930502,
-      "grad_norm": 4.68597412109375,
+      "grad_norm": 4.851049423217773,
       "learning_rate": 1.9745287428718072e-05,
-      "loss": 1.8303,
+      "loss": 1.8273,
       "step": 311
     },
     {
       "epoch": 0.30115830115830117,
-      "grad_norm": 6.816437721252441,
+      "grad_norm": 5.847753524780273,
       "learning_rate": 1.9742944571029515e-05,
-      "loss": 1.8103,
+      "loss": 1.8037,
       "step": 312
     },
     {
       "epoch": 0.30212355212355213,
-      "grad_norm": 3.888559341430664,
+      "grad_norm": 4.01953649520874,
       "learning_rate": 1.9740591128034357e-05,
-      "loss": 1.7313,
+      "loss": 1.7274,
       "step": 313
     },
     {
       "epoch": 0.3030888030888031,
-      "grad_norm": 5.622889041900635,
+      "grad_norm": 5.382868766784668,
       "learning_rate": 1.973822710228951e-05,
-      "loss": 1.7976,
+      "loss": 1.7918,
       "step": 314
     },
     {
       "epoch": 0.30405405405405406,
-      "grad_norm": 5.591082572937012,
+      "grad_norm": 5.3062005043029785,
       "learning_rate": 1.9735852496363395e-05,
-      "loss": 1.7817,
+      "loss": 1.776,
       "step": 315
     },
     {
       "epoch": 0.305019305019305,
-      "grad_norm": 5.452985763549805,
+      "grad_norm": 5.371334075927734,
       "learning_rate": 1.973346731283592e-05,
-      "loss": 1.7994,
+      "loss": 1.7891,
       "step": 316
     },
     {
       "epoch": 0.305984555984556,
-      "grad_norm": 5.106136322021484,
+      "grad_norm": 5.160740375518799,
       "learning_rate": 1.973107155429849e-05,
-      "loss": 1.7853,
+      "loss": 1.7796,
       "step": 317
     },
     {
       "epoch": 0.30694980694980695,
-      "grad_norm": 4.438343524932861,
+      "grad_norm": 4.5635833740234375,
       "learning_rate": 1.9728665223354003e-05,
-      "loss": 1.7399,
+      "loss": 1.7378,
       "step": 318
     },
     {
       "epoch": 0.3079150579150579,
-      "grad_norm": 4.4550251960754395,
+      "grad_norm": 4.520266056060791,
       "learning_rate": 1.972624832261683e-05,
-      "loss": 1.7319,
+      "loss": 1.7274,
       "step": 319
     },
     {
       "epoch": 0.3088803088803089,
-      "grad_norm": 5.767579555511475,
+      "grad_norm": 5.553651809692383,
       "learning_rate": 1.9723820854712836e-05,
-      "loss": 1.6821,
+      "loss": 1.6794,
       "step": 320
     },
     {
       "epoch": 0.30984555984555984,
-      "grad_norm": 4.572739601135254,
+      "grad_norm": 4.72226095199585,
       "learning_rate": 1.972138282227937e-05,
-      "loss": 1.8463,
+      "loss": 1.8433,
       "step": 321
     },
     {
       "epoch": 0.3108108108108108,
-      "grad_norm": 6.5078020095825195,
+      "grad_norm": 6.390082836151123,
       "learning_rate": 1.971893422796524e-05,
-      "loss": 1.7614,
+      "loss": 1.752,
       "step": 322
     },
     {
       "epoch": 0.31177606177606176,
-      "grad_norm": 3.8346452713012695,
+      "grad_norm": 3.9620985984802246,
       "learning_rate": 1.9716475074430762e-05,
-      "loss": 1.7916,
+      "loss": 1.7878,
       "step": 323
     },
     {
       "epoch": 0.3127413127413127,
-      "grad_norm": 4.194652557373047,
+      "grad_norm": 4.554835319519043,
       "learning_rate": 1.9714005364347687e-05,
-      "loss": 1.6699,
+      "loss": 1.6737,
       "step": 324
     },
     {
       "epoch": 0.3137065637065637,
-      "grad_norm": 5.606922149658203,
+      "grad_norm": 5.261370658874512,
       "learning_rate": 1.971152510039926e-05,
-      "loss": 1.7607,
+      "loss": 1.7561,
       "step": 325
     },
     {
       "epoch": 0.31467181467181465,
-      "grad_norm": 3.6330485343933105,
+      "grad_norm": 3.943145275115967,
       "learning_rate": 1.9709034285280185e-05,
-      "loss": 1.7754,
+      "loss": 1.7652,
       "step": 326
     },
     {
       "epoch": 0.3156370656370656,
-      "grad_norm": 5.310985565185547,
+      "grad_norm": 5.0705976486206055,
       "learning_rate": 1.9706532921696634e-05,
-      "loss": 1.6881,
+      "loss": 1.6777,
       "step": 327
     },
     {
       "epoch": 0.3166023166023166,
-      "grad_norm": 4.781001567840576,
+      "grad_norm": 4.380019664764404,
       "learning_rate": 1.9704021012366228e-05,
-      "loss": 1.7821,
+      "loss": 1.7746,
       "step": 328
     },
     {
       "epoch": 0.31756756756756754,
-      "grad_norm": 6.116867542266846,
+      "grad_norm": 6.164519309997559,
       "learning_rate": 1.9701498560018055e-05,
-      "loss": 1.8737,
+      "loss": 1.8715,
       "step": 329
     },
     {
       "epoch": 0.3185328185328185,
-      "grad_norm": 5.587736129760742,
+      "grad_norm": 5.041706085205078,
       "learning_rate": 1.9698965567392656e-05,
-      "loss": 1.7625,
+      "loss": 1.7579,
       "step": 330
     },
     {
       "epoch": 0.3194980694980695,
-      "grad_norm": 3.6453404426574707,
+      "grad_norm": 4.0172200202941895,
       "learning_rate": 1.9696422037242023e-05,
-      "loss": 1.7638,
+      "loss": 1.7596,
       "step": 331
     },
     {
       "epoch": 0.3204633204633205,
-      "grad_norm": 5.7377214431762695,
+      "grad_norm": 5.907834529876709,
       "learning_rate": 1.96938679723296e-05,
-      "loss": 1.7712,
+      "loss": 1.767,
       "step": 332
     },
     {
       "epoch": 0.32142857142857145,
-      "grad_norm": 5.419057369232178,
+      "grad_norm": 4.533480167388916,
       "learning_rate": 1.9691303375430265e-05,
-      "loss": 1.7522,
+      "loss": 1.7476,
       "step": 333
     },
     {
       "epoch": 0.3223938223938224,
-      "grad_norm": 3.678049087524414,
+      "grad_norm": 4.188773155212402,
       "learning_rate": 1.968872824933035e-05,
-      "loss": 1.7662,
+      "loss": 1.7681,
       "step": 334
     },
     {
       "epoch": 0.3233590733590734,
-      "grad_norm": 5.382504940032959,
+      "grad_norm": 5.2045979499816895,
       "learning_rate": 1.9686142596827623e-05,
-      "loss": 1.784,
+      "loss": 1.7754,
       "step": 335
     },
     {
       "epoch": 0.32432432432432434,
-      "grad_norm": 4.854099750518799,
+      "grad_norm": 4.066795825958252,
       "learning_rate": 1.9683546420731292e-05,
-      "loss": 1.7126,
+      "loss": 1.7117,
       "step": 336
     },
     {
       "epoch": 0.3252895752895753,
-      "grad_norm": 4.074610233306885,
+      "grad_norm": 4.666566371917725,
       "learning_rate": 1.9680939723861993e-05,
-      "loss": 1.7813,
+      "loss": 1.7703,
       "step": 337
     },
     {
       "epoch": 0.32625482625482627,
-      "grad_norm": 4.782020568847656,
+      "grad_norm": 4.8976569175720215,
       "learning_rate": 1.967832250905179e-05,
-      "loss": 1.7093,
+      "loss": 1.7007,
       "step": 338
     },
     {
       "epoch": 0.32722007722007723,
-      "grad_norm": 4.857923984527588,
+      "grad_norm": 4.366602420806885,
       "learning_rate": 1.9675694779144184e-05,
-      "loss": 1.7251,
+      "loss": 1.7167,
       "step": 339
     },
     {
       "epoch": 0.3281853281853282,
-      "grad_norm": 4.356438636779785,
+      "grad_norm": 4.61339807510376,
       "learning_rate": 1.9673056536994094e-05,
-      "loss": 1.7315,
+      "loss": 1.7251,
       "step": 340
     },
     {
       "epoch": 0.32915057915057916,
-      "grad_norm": 4.294323444366455,
+      "grad_norm": 4.267677307128906,
       "learning_rate": 1.967040778546786e-05,
-      "loss": 1.7349,
+      "loss": 1.7253,
       "step": 341
     },
     {
       "epoch": 0.3301158301158301,
-      "grad_norm": 4.033375263214111,
+      "grad_norm": 4.421290397644043,
       "learning_rate": 1.9667748527443248e-05,
-      "loss": 1.6029,
+      "loss": 1.5961,
       "step": 342
     },
     {
       "epoch": 0.3310810810810811,
-      "grad_norm": 3.696950912475586,
+      "grad_norm": 3.7843995094299316,
       "learning_rate": 1.966507876580942e-05,
-      "loss": 1.7534,
+      "loss": 1.7466,
       "step": 343
     },
     {
       "epoch": 0.33204633204633205,
-      "grad_norm": 3.718315839767456,
+      "grad_norm": 3.6765830516815186,
       "learning_rate": 1.9662398503466963e-05,
-      "loss": 1.7581,
+      "loss": 1.7488,
       "step": 344
     },
     {
       "epoch": 0.333011583011583,
-      "grad_norm": 3.952453851699829,
+      "grad_norm": 3.747260808944702,
       "learning_rate": 1.9659707743327878e-05,
-      "loss": 1.7004,
+      "loss": 1.696,
       "step": 345
     },
     {
       "epoch": 0.333976833976834,
-      "grad_norm": 4.828619956970215,
+      "grad_norm": 4.562520980834961,
       "learning_rate": 1.9657006488315563e-05,
-      "loss": 1.7381,
+      "loss": 1.73,
       "step": 346
     },
     {
       "epoch": 0.33494208494208494,
-      "grad_norm": 4.196254730224609,
+      "grad_norm": 4.384054660797119,
       "learning_rate": 1.965429474136482e-05,
-      "loss": 1.7943,
+      "loss": 1.7889,
       "step": 347
     },
     {
       "epoch": 0.3359073359073359,
-      "grad_norm": 4.142396926879883,
+      "grad_norm": 4.2236504554748535,
       "learning_rate": 1.965157250542185e-05,
-      "loss": 1.7326,
+      "loss": 1.7273,
       "step": 348
     },
     {
       "epoch": 0.33687258687258687,
-      "grad_norm": 4.161625385284424,
+      "grad_norm": 3.887948989868164,
       "learning_rate": 1.9648839783444243e-05,
-      "loss": 1.7341,
+      "loss": 1.7241,
       "step": 349
     },
     {
       "epoch": 0.33783783783783783,
-      "grad_norm": 4.203663349151611,
+      "grad_norm": 4.439640522003174,
       "learning_rate": 1.9646096578401e-05,
-      "loss": 1.7165,
+      "loss": 1.7126,
       "step": 350
     },
     {
       "epoch": 0.3388030888030888,
-      "grad_norm": 5.153465747833252,
+      "grad_norm": 5.132867336273193,
       "learning_rate": 1.9643342893272493e-05,
-      "loss": 1.803,
+      "loss": 1.7935,
       "step": 351
     },
     {
       "epoch": 0.33976833976833976,
-      "grad_norm": 3.938094139099121,
+      "grad_norm": 4.038984298706055,
       "learning_rate": 1.9640578731050487e-05,
-      "loss": 1.7499,
+      "loss": 1.7375,
       "step": 352
     },
     {
       "epoch": 0.3407335907335907,
-      "grad_norm": 5.379260063171387,
+      "grad_norm": 5.575202941894531,
       "learning_rate": 1.9637804094738133e-05,
-      "loss": 1.7332,
+      "loss": 1.7341,
       "step": 353
     },
     {
       "epoch": 0.3416988416988417,
-      "grad_norm": 3.7647342681884766,
+      "grad_norm": 3.7973532676696777,
       "learning_rate": 1.9635018987349955e-05,
-      "loss": 1.7723,
+      "loss": 1.7649,
       "step": 354
     },
     {
       "epoch": 0.34266409266409265,
-      "grad_norm": 4.630080223083496,
+      "grad_norm": 5.167145729064941,
       "learning_rate": 1.963222341191186e-05,
-      "loss": 1.7612,
+      "loss": 1.7435,
       "step": 355
     },
     {
       "epoch": 0.3436293436293436,
-      "grad_norm": 4.819670677185059,
+      "grad_norm": 4.491520404815674,
       "learning_rate": 1.9629417371461124e-05,
-      "loss": 1.7614,
+      "loss": 1.7575,
       "step": 356
     },
     {
       "epoch": 0.34459459459459457,
-      "grad_norm": 5.193247318267822,
+      "grad_norm": 6.588226795196533,
       "learning_rate": 1.962660086904639e-05,
-      "loss": 1.7174,
+      "loss": 1.7158,
       "step": 357
     },
     {
       "epoch": 0.34555984555984554,
-      "grad_norm": 4.070240020751953,
+      "grad_norm": 4.113042831420898,
       "learning_rate": 1.9623773907727682e-05,
-      "loss": 1.7737,
+      "loss": 1.769,
       "step": 358
     },
     {
       "epoch": 0.3465250965250965,
-      "grad_norm": 3.897179126739502,
+      "grad_norm": 4.43924617767334,
       "learning_rate": 1.962093649057637e-05,
-      "loss": 1.738,
+      "loss": 1.7387,
       "step": 359
     },
     {
       "epoch": 0.3474903474903475,
-      "grad_norm": 3.6528360843658447,
+      "grad_norm": 3.7788119316101074,
       "learning_rate": 1.9618088620675188e-05,
-      "loss": 1.6711,
+      "loss": 1.6662,
       "step": 360
     },
     {
       "epoch": 0.3484555984555985,
-      "grad_norm": 3.6807613372802734,
+      "grad_norm": 3.710517406463623,
       "learning_rate": 1.9615230301118234e-05,
-      "loss": 1.7994,
+      "loss": 1.7983,
       "step": 361
     },
     {
       "epoch": 0.34942084942084944,
-      "grad_norm": 4.079460620880127,
+      "grad_norm": 3.7616159915924072,
       "learning_rate": 1.961236153501095e-05,
-      "loss": 1.7327,
+      "loss": 1.7266,
       "step": 362
     },
     {
       "epoch": 0.3503861003861004,
-      "grad_norm": 4.715167999267578,
+      "grad_norm": 4.776553630828857,
       "learning_rate": 1.9609482325470137e-05,
-      "loss": 1.7292,
+      "loss": 1.7247,
       "step": 363
     },
     {
       "epoch": 0.35135135135135137,
-      "grad_norm": 3.709078073501587,
+      "grad_norm": 3.8136777877807617,
       "learning_rate": 1.960659267562393e-05,
-      "loss": 1.7232,
+      "loss": 1.7153,
       "step": 364
     },
     {
       "epoch": 0.35231660231660233,
-      "grad_norm": 3.9571263790130615,
+      "grad_norm": 4.289064884185791,
       "learning_rate": 1.9603692588611816e-05,
-      "loss": 1.733,
+      "loss": 1.7312,
       "step": 365
     },
     {
       "epoch": 0.3532818532818533,
-      "grad_norm": 4.878260612487793,
+      "grad_norm": 4.563814640045166,
       "learning_rate": 1.9600782067584625e-05,
-      "loss": 1.7165,
+      "loss": 1.7149,
       "step": 366
     },
     {
       "epoch": 0.35424710424710426,
-      "grad_norm": 3.7100722789764404,
+      "grad_norm": 3.8826053142547607,
       "learning_rate": 1.9597861115704512e-05,
-      "loss": 1.7196,
+      "loss": 1.7151,
       "step": 367
     },
     {
       "epoch": 0.3552123552123552,
-      "grad_norm": 5.340725898742676,
+      "grad_norm": 5.331918239593506,
       "learning_rate": 1.9594929736144978e-05,
-      "loss": 1.7328,
+      "loss": 1.7312,
       "step": 368
     },
     {
       "epoch": 0.3561776061776062,
-      "grad_norm": 3.7168092727661133,
+      "grad_norm": 3.751333713531494,
       "learning_rate": 1.9591987932090836e-05,
-      "loss": 1.7269,
+      "loss": 1.7169,
       "step": 369
     },
     {
       "epoch": 0.35714285714285715,
-      "grad_norm": 4.261868476867676,
+      "grad_norm": 4.4523468017578125,
       "learning_rate": 1.9589035706738237e-05,
-      "loss": 1.7852,
+      "loss": 1.7736,
       "step": 370
     },
     {
       "epoch": 0.3581081081081081,
-      "grad_norm": 4.013570308685303,
+      "grad_norm": 3.9097135066986084,
       "learning_rate": 1.958607306329466e-05,
-      "loss": 1.7234,
+      "loss": 1.7211,
       "step": 371
     },
     {
       "epoch": 0.3590733590733591,
-      "grad_norm": 3.7619221210479736,
+      "grad_norm": 3.9183874130249023,
       "learning_rate": 1.9583100004978886e-05,
-      "loss": 1.6764,
+      "loss": 1.6799,
       "step": 372
     },
     {
       "epoch": 0.36003861003861004,
-      "grad_norm": 4.159510612487793,
+      "grad_norm": 4.234004497528076,
       "learning_rate": 1.9580116535021022e-05,
-      "loss": 1.7341,
+      "loss": 1.7307,
       "step": 373
     },
     {
       "epoch": 0.361003861003861,
-      "grad_norm": 3.8162317276000977,
+      "grad_norm": 3.7126665115356445,
       "learning_rate": 1.9577122656662485e-05,
-      "loss": 1.7593,
+      "loss": 1.7539,
       "step": 374
     },
     {
       "epoch": 0.36196911196911197,
-      "grad_norm": 4.361527919769287,
+      "grad_norm": 4.241182327270508,
       "learning_rate": 1.9574118373156e-05,
-      "loss": 1.7437,
+      "loss": 1.7375,
       "step": 375
     },
     {
       "epoch": 0.36293436293436293,
-      "grad_norm": 4.29148006439209,
+      "grad_norm": 4.18535041809082,
       "learning_rate": 1.95711036877656e-05,
-      "loss": 1.8134,
+      "loss": 1.8111,
       "step": 376
     },
     {
       "epoch": 0.3638996138996139,
-      "grad_norm": 3.7081241607666016,
+      "grad_norm": 3.734405755996704,
       "learning_rate": 1.9568078603766613e-05,
-      "loss": 1.7224,
+      "loss": 1.7205,
       "step": 377
     },
     {
       "epoch": 0.36486486486486486,
-      "grad_norm": 4.411427974700928,
+      "grad_norm": 4.161388397216797,
       "learning_rate": 1.9565043124445665e-05,
-      "loss": 1.753,
+      "loss": 1.7436,
       "step": 378
     },
     {
       "epoch": 0.3658301158301158,
-      "grad_norm": 4.272647380828857,
+      "grad_norm": 4.304963111877441,
       "learning_rate": 1.9561997253100684e-05,
-      "loss": 1.7252,
+      "loss": 1.7221,
       "step": 379
     },
     {
       "epoch": 0.3667953667953668,
-      "grad_norm": 3.9347710609436035,
+      "grad_norm": 3.8452253341674805,
       "learning_rate": 1.9558940993040885e-05,
-      "loss": 1.7363,
+      "loss": 1.7314,
       "step": 380
     },
     {
       "epoch": 0.36776061776061775,
-      "grad_norm": 3.8959333896636963,
+      "grad_norm": 3.8398406505584717,
       "learning_rate": 1.955587434758676e-05,
-      "loss": 1.7543,
+      "loss": 1.746,
       "step": 381
     },
     {
       "epoch": 0.3687258687258687,
-      "grad_norm": 3.861565113067627,
+      "grad_norm": 3.8568508625030518,
       "learning_rate": 1.9552797320070097e-05,
-      "loss": 1.7333,
+      "loss": 1.7265,
       "step": 382
     },
     {
       "epoch": 0.3696911196911197,
-      "grad_norm": 4.173425197601318,
+      "grad_norm": 4.175405025482178,
       "learning_rate": 1.954970991383396e-05,
-      "loss": 1.749,
+      "loss": 1.7465,
       "step": 383
     },
     {
       "epoch": 0.37065637065637064,
-      "grad_norm": 5.294433116912842,
+      "grad_norm": 4.90998649597168,
       "learning_rate": 1.9546612132232688e-05,
-      "loss": 1.7576,
+      "loss": 1.752,
       "step": 384
     },
     {
       "epoch": 0.3716216216216216,
-      "grad_norm": 4.082633972167969,
+      "grad_norm": 4.095590591430664,
       "learning_rate": 1.9543503978631894e-05,
-      "loss": 1.7569,
+      "loss": 1.7522,
       "step": 385
     },
     {
       "epoch": 0.37258687258687256,
-      "grad_norm": 4.692306995391846,
+      "grad_norm": 4.905577182769775,
       "learning_rate": 1.9540385456408456e-05,
-      "loss": 1.6946,
+      "loss": 1.6862,
       "step": 386
     },
     {
       "epoch": 0.3735521235521235,
-      "grad_norm": 4.214600563049316,
+      "grad_norm": 4.0402607917785645,
       "learning_rate": 1.953725656895052e-05,
-      "loss": 1.7236,
+      "loss": 1.7172,
       "step": 387
     },
     {
       "epoch": 0.3745173745173745,
-      "grad_norm": 4.012565612792969,
+      "grad_norm": 3.8445563316345215,
       "learning_rate": 1.953411731965749e-05,
-      "loss": 1.7263,
+      "loss": 1.7173,
       "step": 388
     },
     {
       "epoch": 0.3754826254826255,
-      "grad_norm": 4.862763404846191,
+      "grad_norm": 4.813847541809082,
       "learning_rate": 1.9530967711940038e-05,
-      "loss": 1.7547,
+      "loss": 1.7531,
       "step": 389
     },
     {
       "epoch": 0.3764478764478765,
-      "grad_norm": 4.253438472747803,
+      "grad_norm": 3.916043519973755,
       "learning_rate": 1.952780774922008e-05,
-      "loss": 1.6917,
+      "loss": 1.6902,
       "step": 390
     },
     {
       "epoch": 0.37741312741312744,
-      "grad_norm": 3.990596055984497,
+      "grad_norm": 4.006529331207275,
       "learning_rate": 1.952463743493078e-05,
-      "loss": 1.7482,
+      "loss": 1.7376,
       "step": 391
     },
     {
       "epoch": 0.3783783783783784,
-      "grad_norm": 4.393350601196289,
+      "grad_norm": 4.091460227966309,
       "learning_rate": 1.9521456772516554e-05,
-      "loss": 1.778,
+      "loss": 1.7727,
       "step": 392
     },
     {
       "epoch": 0.37934362934362936,
-      "grad_norm": 4.936257362365723,
+      "grad_norm": 4.620602130889893,
       "learning_rate": 1.9518265765433066e-05,
-      "loss": 1.7288,
+      "loss": 1.7312,
       "step": 393
     },
     {
       "epoch": 0.3803088803088803,
-      "grad_norm": 4.040771007537842,
+      "grad_norm": 3.850342273712158,
       "learning_rate": 1.951506441714721e-05,
-      "loss": 1.6901,
+      "loss": 1.6855,
       "step": 394
     },
     {
       "epoch": 0.3812741312741313,
-      "grad_norm": 4.402784824371338,
+      "grad_norm": 4.028168201446533,
       "learning_rate": 1.951185273113712e-05,
-      "loss": 1.7557,
+      "loss": 1.7515,
       "step": 395
     },
     {
       "epoch": 0.38223938223938225,
-      "grad_norm": 4.547336578369141,
+      "grad_norm": 4.427280902862549,
       "learning_rate": 1.9508630710892157e-05,
-      "loss": 1.7385,
+      "loss": 1.7391,
       "step": 396
     },
     {
       "epoch": 0.3832046332046332,
-      "grad_norm": 5.552551746368408,
+      "grad_norm": 4.896625995635986,
       "learning_rate": 1.9505398359912914e-05,
-      "loss": 1.7842,
+      "loss": 1.7786,
       "step": 397
     },
     {
       "epoch": 0.3841698841698842,
-      "grad_norm": 3.7976114749908447,
+      "grad_norm": 3.9859886169433594,
       "learning_rate": 1.950215568171121e-05,
-      "loss": 1.6636,
+      "loss": 1.6556,
       "step": 398
     },
     {
       "epoch": 0.38513513513513514,
-      "grad_norm": 3.8602499961853027,
+      "grad_norm": 3.8103716373443604,
       "learning_rate": 1.9498902679810083e-05,
-      "loss": 1.7281,
+      "loss": 1.7185,
       "step": 399
     },
     {
       "epoch": 0.3861003861003861,
-      "grad_norm": 3.753833055496216,
+      "grad_norm": 3.961416244506836,
       "learning_rate": 1.9495639357743773e-05,
-      "loss": 1.7545,
+      "loss": 1.7487,
       "step": 400
     },
     {
       "epoch": 0.38706563706563707,
-      "grad_norm": 3.642061471939087,
+      "grad_norm": 3.638380765914917,
       "learning_rate": 1.9492365719057752e-05,
-      "loss": 1.738,
+      "loss": 1.7313,
       "step": 401
     },
     {
       "epoch": 0.38803088803088803,
-      "grad_norm": 3.9402945041656494,
+      "grad_norm": 3.931870460510254,
       "learning_rate": 1.9489081767308696e-05,
-      "loss": 1.7693,
+      "loss": 1.7581,
       "step": 402
     },
     {
       "epoch": 0.388996138996139,
-      "grad_norm": 3.8130996227264404,
+      "grad_norm": 3.9259908199310303,
       "learning_rate": 1.9485787506064483e-05,
-      "loss": 1.7868,
+      "loss": 1.7808,
       "step": 403
     },
     {
       "epoch": 0.38996138996138996,
-      "grad_norm": 3.8527071475982666,
+      "grad_norm": 3.833301305770874,
       "learning_rate": 1.9482482938904183e-05,
-      "loss": 1.7396,
+      "loss": 1.7306,
       "step": 404
     },
     {
       "epoch": 0.3909266409266409,
-      "grad_norm": 4.455222129821777,
+      "grad_norm": 4.573353290557861,
       "learning_rate": 1.947916806941808e-05,
-      "loss": 1.7663,
+      "loss": 1.7627,
       "step": 405
     },
     {
       "epoch": 0.3918918918918919,
-      "grad_norm": 3.8062307834625244,
+      "grad_norm": 3.8645942211151123,
       "learning_rate": 1.947584290120764e-05,
-      "loss": 1.6825,
+      "loss": 1.6778,
       "step": 406
     },
     {
       "epoch": 0.39285714285714285,
-      "grad_norm": 3.9247851371765137,
+      "grad_norm": 4.101409912109375,
       "learning_rate": 1.9472507437885523e-05,
-      "loss": 1.7211,
+      "loss": 1.7141,
       "step": 407
     },
     {
       "epoch": 0.3938223938223938,
-      "grad_norm": 4.017999649047852,
+      "grad_norm": 4.171138286590576,
       "learning_rate": 1.9469161683075568e-05,
-      "loss": 1.6961,
+      "loss": 1.6858,
       "step": 408
     },
     {
       "epoch": 0.3947876447876448,
-      "grad_norm": 3.875453233718872,
+      "grad_norm": 3.890751361846924,
       "learning_rate": 1.94658056404128e-05,
-      "loss": 1.8132,
+      "loss": 1.8115,
       "step": 409
     },
     {
       "epoch": 0.39575289575289574,
-      "grad_norm": 4.510627269744873,
+      "grad_norm": 4.988679885864258,
       "learning_rate": 1.9462439313543423e-05,
-      "loss": 1.7492,
+      "loss": 1.7451,
       "step": 410
     },
     {
       "epoch": 0.3967181467181467,
-      "grad_norm": 4.282536506652832,
+      "grad_norm": 4.12715482711792,
       "learning_rate": 1.9459062706124813e-05,
-      "loss": 1.6867,
+      "loss": 1.673,
       "step": 411
     },
     {
       "epoch": 0.39768339768339767,
-      "grad_norm": 4.016504764556885,
+      "grad_norm": 4.202396392822266,
       "learning_rate": 1.945567582182551e-05,
-      "loss": 1.6929,
+      "loss": 1.6866,
       "step": 412
     },
     {
       "epoch": 0.39864864864864863,
-      "grad_norm": 3.6763112545013428,
+      "grad_norm": 3.6307177543640137,
       "learning_rate": 1.9452278664325227e-05,
-      "loss": 1.7274,
+      "loss": 1.7235,
       "step": 413
     },
     {
       "epoch": 0.3996138996138996,
-      "grad_norm": 4.243254661560059,
+      "grad_norm": 4.151615142822266,
       "learning_rate": 1.9448871237314834e-05,
-      "loss": 1.7558,
+      "loss": 1.7541,
       "step": 414
     },
     {
       "epoch": 0.40057915057915056,
-      "grad_norm": 3.863778591156006,
+      "grad_norm": 3.9506571292877197,
       "learning_rate": 1.9445453544496363e-05,
-      "loss": 1.7452,
+      "loss": 1.7447,
       "step": 415
     },
     {
       "epoch": 0.4015444015444015,
-      "grad_norm": 3.8378090858459473,
+      "grad_norm": 3.893714189529419,
       "learning_rate": 1.9442025589582988e-05,
-      "loss": 1.8414,
+      "loss": 1.8338,
       "step": 416
     },
     {
       "epoch": 0.4025096525096525,
-      "grad_norm": 3.61639404296875,
+      "grad_norm": 3.644362449645996,
       "learning_rate": 1.9438587376299047e-05,
-      "loss": 1.6905,
+      "loss": 1.6808,
       "step": 417
     },
     {
       "epoch": 0.4034749034749035,
-      "grad_norm": 4.8080902099609375,
+      "grad_norm": 4.656255722045898,
       "learning_rate": 1.9435138908380017e-05,
-      "loss": 1.7115,
+      "loss": 1.7084,
       "step": 418
     },
     {
       "epoch": 0.40444015444015446,
-      "grad_norm": 4.0947418212890625,
+      "grad_norm": 3.973984956741333,
       "learning_rate": 1.9431680189572514e-05,
-      "loss": 1.7023,
+      "loss": 1.6942,
       "step": 419
     },
     {
       "epoch": 0.40540540540540543,
-      "grad_norm": 4.377869606018066,
+      "grad_norm": 4.226536750793457,
       "learning_rate": 1.942821122363429e-05,
-      "loss": 1.7552,
+      "loss": 1.7397,
       "step": 420
     },
     {
       "epoch": 0.4063706563706564,
-      "grad_norm": 4.720925331115723,
+      "grad_norm": 4.542608737945557,
       "learning_rate": 1.942473201433424e-05,
-      "loss": 1.7333,
+      "loss": 1.7276,
       "step": 421
     },
     {
       "epoch": 0.40733590733590735,
-      "grad_norm": 3.5857903957366943,
+      "grad_norm": 3.5939407348632812,
       "learning_rate": 1.9421242565452373e-05,
-      "loss": 1.711,
+      "loss": 1.7035,
       "step": 422
     },
     {
       "epoch": 0.4083011583011583,
-      "grad_norm": 4.641952991485596,
+      "grad_norm": 4.464782238006592,
       "learning_rate": 1.9417742880779835e-05,
-      "loss": 1.7562,
+      "loss": 1.7499,
       "step": 423
     },
     {
       "epoch": 0.4092664092664093,
-      "grad_norm": 3.7906808853149414,
+      "grad_norm": 3.890183210372925,
       "learning_rate": 1.9414232964118893e-05,
-      "loss": 1.7231,
+      "loss": 1.7216,
       "step": 424
     },
     {
       "epoch": 0.41023166023166024,
-      "grad_norm": 3.551534414291382,
+      "grad_norm": 3.5962111949920654,
       "learning_rate": 1.941071281928292e-05,
-      "loss": 1.6716,
+      "loss": 1.6674,
       "step": 425
     },
     {
       "epoch": 0.4111969111969112,
-      "grad_norm": 4.0085930824279785,
+      "grad_norm": 3.900623083114624,
       "learning_rate": 1.940718245009641e-05,
-      "loss": 1.6725,
+      "loss": 1.6677,
       "step": 426
     },
     {
       "epoch": 0.41216216216216217,
-      "grad_norm": 3.807429790496826,
+      "grad_norm": 3.743797779083252,
       "learning_rate": 1.940364186039496e-05,
-      "loss": 1.7575,
+      "loss": 1.7542,
       "step": 427
     },
     {
       "epoch": 0.41312741312741313,
-      "grad_norm": 3.671699047088623,
+      "grad_norm": 3.5732181072235107,
       "learning_rate": 1.9400091054025286e-05,
-      "loss": 1.686,
+      "loss": 1.687,
       "step": 428
     },
     {
       "epoch": 0.4140926640926641,
-      "grad_norm": 3.4760184288024902,
+      "grad_norm": 3.623192548751831,
       "learning_rate": 1.939653003484518e-05,
-      "loss": 1.7228,
+      "loss": 1.7166,
       "step": 429
     },
     {
       "epoch": 0.41505791505791506,
-      "grad_norm": 3.6061408519744873,
+      "grad_norm": 3.582862377166748,
       "learning_rate": 1.9392958806723544e-05,
-      "loss": 1.7131,
+      "loss": 1.7066,
       "step": 430
     },
     {
       "epoch": 0.416023166023166,
-      "grad_norm": 5.025588512420654,
+      "grad_norm": 5.185995101928711,
       "learning_rate": 1.9389377373540373e-05,
-      "loss": 1.6722,
+      "loss": 1.6629,
       "step": 431
     },
     {
       "epoch": 0.416988416988417,
-      "grad_norm": 5.2907280921936035,
+      "grad_norm": 5.050741672515869,
       "learning_rate": 1.9385785739186746e-05,
-      "loss": 1.7654,
+      "loss": 1.7608,
       "step": 432
     },
     {
       "epoch": 0.41795366795366795,
-      "grad_norm": 4.422195911407471,
+      "grad_norm": 4.399065971374512,
       "learning_rate": 1.9382183907564824e-05,
-      "loss": 1.6597,
+      "loss": 1.6515,
       "step": 433
     },
     {
       "epoch": 0.4189189189189189,
-      "grad_norm": 5.675929546356201,
+      "grad_norm": 5.3179779052734375,
       "learning_rate": 1.9378571882587848e-05,
-      "loss": 1.7137,
+      "loss": 1.705,
       "step": 434
     },
     {
       "epoch": 0.4198841698841699,
-      "grad_norm": 4.700954914093018,
+      "grad_norm": 4.73206901550293,
       "learning_rate": 1.937494966818014e-05,
-      "loss": 1.7297,
+      "loss": 1.7309,
       "step": 435
     },
     {
       "epoch": 0.42084942084942084,
-      "grad_norm": 3.9680557250976562,
+      "grad_norm": 3.979621648788452,
       "learning_rate": 1.9371317268277075e-05,
-      "loss": 1.6652,
+      "loss": 1.6601,
       "step": 436
     },
     {
       "epoch": 0.4218146718146718,
-      "grad_norm": 6.557024955749512,
+      "grad_norm": 6.588276386260986,
       "learning_rate": 1.936767468682511e-05,
-      "loss": 1.7385,
+      "loss": 1.7328,
       "step": 437
     },
     {
       "epoch": 0.42277992277992277,
-      "grad_norm": 3.592010498046875,
+      "grad_norm": 3.691559076309204,
       "learning_rate": 1.9364021927781764e-05,
-      "loss": 1.7594,
+      "loss": 1.7581,
       "step": 438
     },
     {
       "epoch": 0.42374517374517373,
-      "grad_norm": 5.884945392608643,
+      "grad_norm": 5.932117938995361,
       "learning_rate": 1.9360358995115605e-05,
-      "loss": 1.7047,
+      "loss": 1.6938,
       "step": 439
     },
     {
       "epoch": 0.4247104247104247,
-      "grad_norm": 4.135279655456543,
+      "grad_norm": 4.174450397491455,
       "learning_rate": 1.9356685892806255e-05,
-      "loss": 1.7335,
+      "loss": 1.7252,
       "step": 440
     },
     {
       "epoch": 0.42567567567567566,
-      "grad_norm": 4.141030311584473,
+      "grad_norm": 4.358636379241943,
       "learning_rate": 1.9353002624844392e-05,
-      "loss": 1.7431,
+      "loss": 1.7296,
       "step": 441
     },
     {
       "epoch": 0.4266409266409266,
-      "grad_norm": 5.444819450378418,
+      "grad_norm": 5.36053991317749,
       "learning_rate": 1.9349309195231732e-05,
-      "loss": 1.6846,
+      "loss": 1.6827,
       "step": 442
     },
     {
       "epoch": 0.4276061776061776,
-      "grad_norm": 3.9959070682525635,
+      "grad_norm": 4.094240665435791,
       "learning_rate": 1.934560560798104e-05,
-      "loss": 1.748,
+      "loss": 1.7387,
       "step": 443
     },
     {
       "epoch": 0.42857142857142855,
-      "grad_norm": 6.770045757293701,
+      "grad_norm": 6.735424995422363,
       "learning_rate": 1.9341891867116102e-05,
-      "loss": 1.7336,
+      "loss": 1.7279,
       "step": 444
     },
     {
       "epoch": 0.4295366795366795,
-      "grad_norm": 4.172303676605225,
+      "grad_norm": 4.079831600189209,
       "learning_rate": 1.9338167976671747e-05,
-      "loss": 1.7244,
+      "loss": 1.723,
       "step": 445
     },
     {
       "epoch": 0.4305019305019305,
-      "grad_norm": 5.299330234527588,
+      "grad_norm": 5.601644515991211,
       "learning_rate": 1.933443394069383e-05,
       "loss": 1.6752,
       "step": 446
     },
     {
       "epoch": 0.4314671814671815,
-      "grad_norm": 5.519097805023193,
+      "grad_norm": 5.604174613952637,
       "learning_rate": 1.9330689763239225e-05,
-      "loss": 1.7469,
+      "loss": 1.7416,
       "step": 447
     },
     {
       "epoch": 0.43243243243243246,
-      "grad_norm": 4.377777099609375,
+      "grad_norm": 4.963226318359375,
       "learning_rate": 1.9326935448375822e-05,
-      "loss": 1.739,
+      "loss": 1.7359,
       "step": 448
     },
     {
       "epoch": 0.4333976833976834,
-      "grad_norm": 6.823882579803467,
+      "grad_norm": 6.457501411437988,
       "learning_rate": 1.9323171000182538e-05,
-      "loss": 1.7708,
+      "loss": 1.7589,
       "step": 449
     },
     {
       "epoch": 0.4343629343629344,
-      "grad_norm": 3.7352969646453857,
+      "grad_norm": 3.651960611343384,
       "learning_rate": 1.9319396422749284e-05,
-      "loss": 1.7035,
+      "loss": 1.6957,
       "step": 450
     },
     {
       "epoch": 0.43532818532818535,
-      "grad_norm": 5.113969802856445,
+      "grad_norm": 5.249007701873779,
       "learning_rate": 1.9315611720176982e-05,
-      "loss": 1.6949,
+      "loss": 1.6885,
       "step": 451
     },
     {
       "epoch": 0.4362934362934363,
-      "grad_norm": 5.587355613708496,
+      "grad_norm": 5.065216541290283,
       "learning_rate": 1.931181689657756e-05,
-      "loss": 1.7613,
+      "loss": 1.7545,
       "step": 452
     },
     {
       "epoch": 0.4372586872586873,
-      "grad_norm": 4.409136772155762,
+      "grad_norm": 4.777818202972412,
       "learning_rate": 1.9308011956073933e-05,
-      "loss": 1.7631,
+      "loss": 1.7609,
       "step": 453
     },
     {
       "epoch": 0.43822393822393824,
-      "grad_norm": 5.3545074462890625,
+      "grad_norm": 5.120543003082275,
       "learning_rate": 1.9304196902800014e-05,
-      "loss": 1.7496,
+      "loss": 1.7411,
       "step": 454
     },
     {
       "epoch": 0.4391891891891892,
-      "grad_norm": 3.6010336875915527,
+      "grad_norm": 3.7689151763916016,
       "learning_rate": 1.9300371740900698e-05,
-      "loss": 1.7381,
+      "loss": 1.727,
       "step": 455
     },
     {
       "epoch": 0.44015444015444016,
-      "grad_norm": 3.643707513809204,
+      "grad_norm": 3.6061551570892334,
       "learning_rate": 1.9296536474531864e-05,
-      "loss": 1.7991,
+      "loss": 1.7934,
       "step": 456
     },
     {
       "epoch": 0.4411196911196911,
-      "grad_norm": 4.69381856918335,
+      "grad_norm": 4.489399433135986,
       "learning_rate": 1.9292691107860374e-05,
-      "loss": 1.6745,
+      "loss": 1.6648,
       "step": 457
     },
     {
       "epoch": 0.4420849420849421,
-      "grad_norm": 3.876347780227661,
+      "grad_norm": 3.8636281490325928,
       "learning_rate": 1.928883564506406e-05,
-      "loss": 1.7045,
+      "loss": 1.6874,
       "step": 458
     },
     {
       "epoch": 0.44305019305019305,
-      "grad_norm": 3.8134143352508545,
+      "grad_norm": 3.8556625843048096,
       "learning_rate": 1.928497009033172e-05,
-      "loss": 1.6769,
+      "loss": 1.6825,
       "step": 459
     },
     {
       "epoch": 0.444015444015444,
-      "grad_norm": 3.9049108028411865,
+      "grad_norm": 3.9801862239837646,
       "learning_rate": 1.9281094447863126e-05,
-      "loss": 1.7107,
+      "loss": 1.7126,
       "step": 460
     },
     {
       "epoch": 0.444980694980695,
-      "grad_norm": 3.940433979034424,
+      "grad_norm": 4.001309871673584,
       "learning_rate": 1.927720872186899e-05,
-      "loss": 1.712,
+      "loss": 1.7113,
       "step": 461
     },
     {
       "epoch": 0.44594594594594594,
-      "grad_norm": 3.828815221786499,
+      "grad_norm": 3.7829551696777344,
       "learning_rate": 1.9273312916571008e-05,
-      "loss": 1.765,
+      "loss": 1.7644,
       "step": 462
     },
     {
       "epoch": 0.4469111969111969,
-      "grad_norm": 3.92582106590271,
+      "grad_norm": 3.8276846408843994,
       "learning_rate": 1.9269407036201803e-05,
-      "loss": 1.7118,
+      "loss": 1.7061,
       "step": 463
     },
     {
       "epoch": 0.44787644787644787,
-      "grad_norm": 4.062038421630859,
+      "grad_norm": 4.0219340324401855,
       "learning_rate": 1.9265491085004956e-05,
-      "loss": 1.7196,
+      "loss": 1.7146,
       "step": 464
     },
     {
       "epoch": 0.44884169884169883,
-      "grad_norm": 4.037832736968994,
+      "grad_norm": 4.084554195404053,
       "learning_rate": 1.926156506723498e-05,
-      "loss": 1.7286,
+      "loss": 1.721,
       "step": 465
     },
     {
       "epoch": 0.4498069498069498,
-      "grad_norm": 4.069302082061768,
+      "grad_norm": 4.197999954223633,
       "learning_rate": 1.9257628987157343e-05,
-      "loss": 1.692,
+      "loss": 1.6819,
       "step": 466
     },
     {
       "epoch": 0.45077220077220076,
-      "grad_norm": 4.033402442932129,
+      "grad_norm": 3.9873580932617188,
       "learning_rate": 1.9253682849048417e-05,
-      "loss": 1.717,
+      "loss": 1.7143,
       "step": 467
     },
     {
       "epoch": 0.4517374517374517,
-      "grad_norm": 4.400399684906006,
+      "grad_norm": 4.612418174743652,
       "learning_rate": 1.9249726657195534e-05,
-      "loss": 1.7185,
+      "loss": 1.715,
       "step": 468
     },
     {
       "epoch": 0.4527027027027027,
-      "grad_norm": 3.828836679458618,
+      "grad_norm": 4.0667314529418945,
       "learning_rate": 1.9245760415896923e-05,
-      "loss": 1.6929,
+      "loss": 1.6846,
       "step": 469
     },
     {
       "epoch": 0.45366795366795365,
-      "grad_norm": 5.331608772277832,
+      "grad_norm": 5.316004753112793,
       "learning_rate": 1.924178412946174e-05,
-      "loss": 1.7378,
+      "loss": 1.7374,
       "step": 470
     },
     {
       "epoch": 0.4546332046332046,
-      "grad_norm": 4.237404823303223,
+      "grad_norm": 4.403238773345947,
       "learning_rate": 1.9237797802210067e-05,
-      "loss": 1.7607,
+      "loss": 1.7511,
       "step": 471
     },
     {
       "epoch": 0.4555984555984556,
-      "grad_norm": 4.649611949920654,
+      "grad_norm": 4.905179977416992,
       "learning_rate": 1.9233801438472875e-05,
-      "loss": 1.7569,
+      "loss": 1.7494,
       "step": 472
     },
     {
       "epoch": 0.45656370656370654,
-      "grad_norm": 6.196529865264893,
+      "grad_norm": 6.052628517150879,
       "learning_rate": 1.922979504259205e-05,
-      "loss": 1.7296,
+      "loss": 1.7169,
       "step": 473
     },
     {
       "epoch": 0.4575289575289575,
-      "grad_norm": 3.977447748184204,
+      "grad_norm": 3.9463837146759033,
       "learning_rate": 1.922577861892037e-05,
-      "loss": 1.7504,
+      "loss": 1.7441,
       "step": 474
     },
     {
       "epoch": 0.4584942084942085,
-      "grad_norm": 9.30162525177002,
+      "grad_norm": 9.358197212219238,
       "learning_rate": 1.9221752171821523e-05,
-      "loss": 1.6945,
+      "loss": 1.6897,
       "step": 475
     },
     {
       "epoch": 0.4594594594594595,
-      "grad_norm": 3.6693522930145264,
+      "grad_norm": 3.676985263824463,
       "learning_rate": 1.9217715705670072e-05,
-      "loss": 1.6373,
+      "loss": 1.6303,
       "step": 476
     },
     {
       "epoch": 0.46042471042471045,
-      "grad_norm": 6.122697830200195,
+      "grad_norm": 6.051453113555908,
       "learning_rate": 1.9213669224851473e-05,
-      "loss": 1.6249,
+      "loss": 1.6154,
       "step": 477
     },
     {
       "epoch": 0.4613899613899614,
-      "grad_norm": 5.823074817657471,
+      "grad_norm": 5.9097771644592285,
       "learning_rate": 1.9209612733762056e-05,
-      "loss": 1.6743,
+      "loss": 1.6698,
       "step": 478
     },
     {
       "epoch": 0.4623552123552124,
-      "grad_norm": 3.764169216156006,
+      "grad_norm": 3.943181037902832,
       "learning_rate": 1.9205546236809037e-05,
-      "loss": 1.7555,
+      "loss": 1.745,
       "step": 479
     },
     {
       "epoch": 0.46332046332046334,
-      "grad_norm": 4.683211326599121,
+      "grad_norm": 4.891289710998535,
       "learning_rate": 1.9201469738410493e-05,
-      "loss": 1.7276,
+      "loss": 1.7197,
       "step": 480
     },
     {
       "epoch": 0.4642857142857143,
-      "grad_norm": 5.23283576965332,
+      "grad_norm": 4.775295257568359,
       "learning_rate": 1.919738324299537e-05,
-      "loss": 1.6407,
+      "loss": 1.6373,
       "step": 481
     },
     {
       "epoch": 0.46525096525096526,
-      "grad_norm": 4.238620281219482,
+      "grad_norm": 4.169955253601074,
       "learning_rate": 1.9193286755003475e-05,
-      "loss": 1.6999,
+      "loss": 1.6943,
       "step": 482
     },
     {
       "epoch": 0.46621621621621623,
-      "grad_norm": 4.695125579833984,
+      "grad_norm": 4.452609062194824,
       "learning_rate": 1.918918027888548e-05,
-      "loss": 1.7013,
+      "loss": 1.6909,
       "step": 483
     },
     {
       "epoch": 0.4671814671814672,
-      "grad_norm": 5.553123474121094,
+      "grad_norm": 5.130198001861572,
       "learning_rate": 1.9185063819102896e-05,
-      "loss": 1.7287,
+      "loss": 1.732,
       "step": 484
     },
     {
       "epoch": 0.46814671814671815,
-      "grad_norm": 4.4962077140808105,
+      "grad_norm": 4.7012152671813965,
       "learning_rate": 1.918093738012809e-05,
-      "loss": 1.7472,
+      "loss": 1.7429,
       "step": 485
     },
     {
       "epoch": 0.4691119691119691,
-      "grad_norm": 5.718868255615234,
+      "grad_norm": 5.2393598556518555,
       "learning_rate": 1.9176800966444263e-05,
-      "loss": 1.717,
+      "loss": 1.7047,
       "step": 486
     },
     {
       "epoch": 0.4700772200772201,
-      "grad_norm": 4.19191837310791,
+      "grad_norm": 4.199380874633789,
       "learning_rate": 1.917265458254546e-05,
-      "loss": 1.6781,
+      "loss": 1.6745,
       "step": 487
     },
     {
       "epoch": 0.47104247104247104,
-      "grad_norm": 4.743502616882324,
+      "grad_norm": 4.6788506507873535,
       "learning_rate": 1.9168498232936557e-05,
-      "loss": 1.7112,
+      "loss": 1.6977,
       "step": 488
     },
     {
       "epoch": 0.472007722007722,
-      "grad_norm": 3.346381902694702,
+      "grad_norm": 3.401582956314087,
       "learning_rate": 1.9164331922133254e-05,
-      "loss": 1.7331,
+      "loss": 1.7218,
       "step": 489
     },
     {
       "epoch": 0.47297297297297297,
-      "grad_norm": 3.966032028198242,
+      "grad_norm": 4.121973991394043,
       "learning_rate": 1.9160155654662075e-05,
-      "loss": 1.69,
+      "loss": 1.683,
       "step": 490
     },
     {
       "epoch": 0.47393822393822393,
-      "grad_norm": 4.457366943359375,
+      "grad_norm": 4.16146993637085,
       "learning_rate": 1.9155969435060363e-05,
-      "loss": 1.6704,
+      "loss": 1.6628,
       "step": 491
     },
     {
       "epoch": 0.4749034749034749,
-      "grad_norm": 4.079778671264648,
+      "grad_norm": 4.121311664581299,
       "learning_rate": 1.915177326787627e-05,
-      "loss": 1.6572,
+      "loss": 1.6515,
       "step": 492
     },
     {
       "epoch": 0.47586872586872586,
-      "grad_norm": 5.250837802886963,
+      "grad_norm": 5.128288745880127,
       "learning_rate": 1.914756715766877e-05,
-      "loss": 1.7621,
+      "loss": 1.7567,
       "step": 493
     },
     {
       "epoch": 0.4768339768339768,
-      "grad_norm": 4.468987464904785,
+      "grad_norm": 4.360013008117676,
       "learning_rate": 1.914335110900761e-05,
-      "loss": 1.6141,
+      "loss": 1.6086,
       "step": 494
     },
     {
       "epoch": 0.4777992277992278,
-      "grad_norm": 5.575342178344727,
+      "grad_norm": 5.781803131103516,
       "learning_rate": 1.9139125126473365e-05,
-      "loss": 1.6715,
+      "loss": 1.6658,
       "step": 495
     },
     {
       "epoch": 0.47876447876447875,
-      "grad_norm": 6.624782562255859,
+      "grad_norm": 6.578468322753906,
       "learning_rate": 1.9134889214657385e-05,
-      "loss": 1.7033,
+      "loss": 1.7018,
       "step": 496
     },
     {
       "epoch": 0.4797297297297297,
-      "grad_norm": 4.400557518005371,
+      "grad_norm": 4.954802513122559,
       "learning_rate": 1.9130643378161817e-05,
-      "loss": 1.649,
+      "loss": 1.6483,
       "step": 497
     },
     {
       "epoch": 0.4806949806949807,
-      "grad_norm": 5.939478397369385,
+      "grad_norm": 6.14895486831665,
       "learning_rate": 1.9126387621599585e-05,
-      "loss": 1.7254,
+      "loss": 1.7206,
       "step": 498
     },
     {
       "epoch": 0.48166023166023164,
-      "grad_norm": 4.109172344207764,
+      "grad_norm": 3.6972389221191406,
       "learning_rate": 1.9122121949594392e-05,
-      "loss": 1.6279,
+      "loss": 1.6248,
       "step": 499
     },
     {
       "epoch": 0.4826254826254826,
-      "grad_norm": 4.4112701416015625,
+      "grad_norm": 4.924983978271484,
       "learning_rate": 1.9117846366780714e-05,
-      "loss": 1.7164,
+      "loss": 1.7134,
       "step": 500
     },
     {
       "epoch": 0.48359073359073357,
-      "grad_norm": 6.475155353546143,
+      "grad_norm": 5.9817399978637695,
       "learning_rate": 1.9113560877803798e-05,
-      "loss": 1.663,
+      "loss": 1.6605,
       "step": 501
     },
     {
       "epoch": 0.48455598455598453,
-      "grad_norm": 4.766849040985107,
+      "grad_norm": 5.171220302581787,
       "learning_rate": 1.9109265487319645e-05,
-      "loss": 1.6338,
+      "loss": 1.626,
       "step": 502
     },
     {
       "epoch": 0.4855212355212355,
-      "grad_norm": 5.895227432250977,
+      "grad_norm": 6.184906482696533,
       "learning_rate": 1.910496019999502e-05,
-      "loss": 1.7569,
+      "loss": 1.7431,
       "step": 503
     },
     {
       "epoch": 0.4864864864864865,
-      "grad_norm": 4.210174083709717,
+      "grad_norm": 4.513047218322754,
       "learning_rate": 1.9100645020507448e-05,
-      "loss": 1.7424,
+      "loss": 1.7412,
       "step": 504
     },
     {
       "epoch": 0.4874517374517375,
-      "grad_norm": 4.535717487335205,
+      "grad_norm": 4.442533493041992,
       "learning_rate": 1.9096319953545186e-05,
-      "loss": 1.7557,
+      "loss": 1.7491,
       "step": 505
     },
     {
       "epoch": 0.48841698841698844,
-      "grad_norm": 4.628802299499512,
+      "grad_norm": 4.440882205963135,
       "learning_rate": 1.9091985003807242e-05,
-      "loss": 1.6228,
+      "loss": 1.6215,
       "step": 506
     },
     {
       "epoch": 0.4893822393822394,
-      "grad_norm": 4.0775017738342285,
+      "grad_norm": 4.3051605224609375,
       "learning_rate": 1.9087640176003356e-05,
-      "loss": 1.6476,
+      "loss": 1.6369,
       "step": 507
     },
     {
       "epoch": 0.49034749034749037,
-      "grad_norm": 5.471363067626953,
+      "grad_norm": 5.43113899230957,
       "learning_rate": 1.9083285474854012e-05,
-      "loss": 1.6965,
+      "loss": 1.6972,
       "step": 508
     },
     {
       "epoch": 0.49131274131274133,
-      "grad_norm": 3.517397880554199,
+      "grad_norm": 3.441530704498291,
       "learning_rate": 1.9078920905090408e-05,
-      "loss": 1.6821,
+      "loss": 1.685,
       "step": 509
     },
     {
       "epoch": 0.4922779922779923,
-      "grad_norm": 5.442646503448486,
+      "grad_norm": 5.869041919708252,
       "learning_rate": 1.907454647145447e-05,
-      "loss": 1.7628,
+      "loss": 1.7553,
       "step": 510
     },
     {
       "epoch": 0.49324324324324326,
-      "grad_norm": 3.913710355758667,
+      "grad_norm": 3.6109423637390137,
       "learning_rate": 1.9070162178698838e-05,
-      "loss": 1.7004,
+      "loss": 1.6937,
       "step": 511
     },
     {
       "epoch": 0.4942084942084942,
-      "grad_norm": 3.85947322845459,
+      "grad_norm": 3.9249019622802734,
       "learning_rate": 1.9065768031586864e-05,
-      "loss": 1.6542,
+      "loss": 1.6456,
       "step": 512
     },
     {
       "epoch": 0.4951737451737452,
-      "grad_norm": 4.821897983551025,
+      "grad_norm": 4.960227012634277,
       "learning_rate": 1.906136403489261e-05,
-      "loss": 1.6969,
+      "loss": 1.6931,
       "step": 513
     },
     {
       "epoch": 0.49613899613899615,
-      "grad_norm": 3.840449094772339,
+      "grad_norm": 3.654670238494873,
       "learning_rate": 1.905695019340083e-05,
-      "loss": 1.7016,
+      "loss": 1.6944,
       "step": 514
     },
     {
       "epoch": 0.4971042471042471,
-      "grad_norm": 5.120670795440674,
+      "grad_norm": 5.541905879974365,
       "learning_rate": 1.9052526511906993e-05,
-      "loss": 1.7121,
+      "loss": 1.7063,
       "step": 515
     },
     {
       "epoch": 0.4980694980694981,
-      "grad_norm": 4.671112060546875,
+      "grad_norm": 4.81696891784668,
       "learning_rate": 1.9048092995217234e-05,
-      "loss": 1.7133,
+      "loss": 1.7075,
       "step": 516
     },
     {
       "epoch": 0.49903474903474904,
-      "grad_norm": 4.036538600921631,
+      "grad_norm": 4.262349605560303,
       "learning_rate": 1.904364964814839e-05,
-      "loss": 1.6991,
+      "loss": 1.6987,
       "step": 517
     },
     {
       "epoch": 0.5,
-      "grad_norm": 4.0499701499938965,
+      "grad_norm": 4.317120552062988,
       "learning_rate": 1.9039196475527974e-05,
-      "loss": 1.7296,
+      "loss": 1.7243,
       "step": 518
     },
     {
       "epoch": 0.500965250965251,
-      "grad_norm": 4.240586757659912,
+      "grad_norm": 4.049501895904541,
       "learning_rate": 1.9034733482194176e-05,
-      "loss": 1.6821,
+      "loss": 1.6702,
       "step": 519
     },
     {
       "epoch": 0.5019305019305019,
-      "grad_norm": 3.5308685302734375,
+      "grad_norm": 3.7106359004974365,
       "learning_rate": 1.903026067299585e-05,
-      "loss": 1.6224,
+      "loss": 1.6154,
       "step": 520
     },
     {
       "epoch": 0.502895752895753,
-      "grad_norm": 3.4554286003112793,
+      "grad_norm": 3.5182018280029297,
       "learning_rate": 1.9025778052792518e-05,
-      "loss": 1.7154,
+      "loss": 1.7104,
       "step": 521
     },
     {
       "epoch": 0.5038610038610039,
-      "grad_norm": 3.995757818222046,
+      "grad_norm": 4.057737827301025,
       "learning_rate": 1.902128562645437e-05,
-      "loss": 1.7159,
+      "loss": 1.7152,
       "step": 522
     },
     {
       "epoch": 0.5048262548262549,
-      "grad_norm": 3.843982696533203,
+      "grad_norm": 3.68337345123291,
       "learning_rate": 1.901678339886223e-05,
-      "loss": 1.7729,
+      "loss": 1.7668,
       "step": 523
     },
     {
       "epoch": 0.5057915057915058,
-      "grad_norm": 3.784400463104248,
+      "grad_norm": 3.687596559524536,
       "learning_rate": 1.901227137490759e-05,
-      "loss": 1.6779,
+      "loss": 1.6681,
       "step": 524
     },
     {
       "epoch": 0.5067567567567568,
-      "grad_norm": 4.4152350425720215,
+      "grad_norm": 4.363684177398682,
       "learning_rate": 1.9007749559492575e-05,
-      "loss": 1.6983,
+      "loss": 1.6884,
       "step": 525
     },
     {
       "epoch": 0.5077220077220077,
-      "grad_norm": 3.8352880477905273,
+      "grad_norm": 3.909970998764038,
       "learning_rate": 1.9003217957529948e-05,
-      "loss": 1.6381,
+      "loss": 1.6366,
       "step": 526
     },
     {
       "epoch": 0.5086872586872587,
-      "grad_norm": 3.8873074054718018,
+      "grad_norm": 3.9115724563598633,
       "learning_rate": 1.899867657394311e-05,
-      "loss": 1.7291,
+      "loss": 1.7157,
       "step": 527
     },
     {
       "epoch": 0.5096525096525096,
-      "grad_norm": 4.419820785522461,
+      "grad_norm": 4.275258541107178,
       "learning_rate": 1.899412541366609e-05,
-      "loss": 1.6595,
+      "loss": 1.6545,
       "step": 528
     },
     {
       "epoch": 0.5106177606177607,
-      "grad_norm": 3.6705615520477295,
+      "grad_norm": 3.6352553367614746,
       "learning_rate": 1.898956448164353e-05,
-      "loss": 1.6919,
+      "loss": 1.6813,
       "step": 529
     },
     {
       "epoch": 0.5115830115830116,
-      "grad_norm": 3.7511744499206543,
+      "grad_norm": 3.6823346614837646,
       "learning_rate": 1.8984993782830695e-05,
-      "loss": 1.6396,
+      "loss": 1.6346,
       "step": 530
     },
     {
       "epoch": 0.5125482625482626,
-      "grad_norm": 4.202239990234375,
+      "grad_norm": 4.081632614135742,
       "learning_rate": 1.8980413322193462e-05,
-      "loss": 1.6633,
+      "loss": 1.6609,
       "step": 531
     },
     {
       "epoch": 0.5135135135135135,
-      "grad_norm": 3.8342583179473877,
+      "grad_norm": 3.9719972610473633,
       "learning_rate": 1.8975823104708314e-05,
-      "loss": 1.6892,
+      "loss": 1.6837,
       "step": 532
     },
     {
       "epoch": 0.5144787644787645,
-      "grad_norm": 3.90169095993042,
+      "grad_norm": 3.83481502532959,
       "learning_rate": 1.8971223135362328e-05,
-      "loss": 1.7225,
+      "loss": 1.7171,
       "step": 533
     },
     {
       "epoch": 0.5154440154440154,
-      "grad_norm": 4.5637688636779785,
+      "grad_norm": 4.9827561378479,
       "learning_rate": 1.896661341915318e-05,
-      "loss": 1.7099,
+      "loss": 1.7133,
       "step": 534
     },
     {
       "epoch": 0.5164092664092664,
-      "grad_norm": 3.929818868637085,
+      "grad_norm": 3.8905863761901855,
       "learning_rate": 1.8961993961089145e-05,
-      "loss": 1.7169,
+      "loss": 1.7157,
       "step": 535
     },
     {
       "epoch": 0.5173745173745173,
-      "grad_norm": 4.224812984466553,
+      "grad_norm": 4.3186187744140625,
       "learning_rate": 1.895736476618906e-05,
-      "loss": 1.7094,
+      "loss": 1.7132,
       "step": 536
     },
     {
       "epoch": 0.5183397683397684,
-      "grad_norm": 4.2393317222595215,
+      "grad_norm": 4.754609107971191,
       "learning_rate": 1.8952725839482368e-05,
-      "loss": 1.6806,
+      "loss": 1.6875,
       "step": 537
     },
     {
       "epoch": 0.5193050193050193,
-      "grad_norm": 4.0183305740356445,
+      "grad_norm": 4.35208797454834,
       "learning_rate": 1.894807718600906e-05,
-      "loss": 1.7971,
+      "loss": 1.7965,
       "step": 538
     },
     {
       "epoch": 0.5202702702702703,
-      "grad_norm": 4.289072513580322,
+      "grad_norm": 4.698247909545898,
       "learning_rate": 1.8943418810819715e-05,
-      "loss": 1.6686,
+      "loss": 1.6687,
       "step": 539
     },
     {
       "epoch": 0.5212355212355212,
-      "grad_norm": 4.444197654724121,
+      "grad_norm": 4.267046928405762,
       "learning_rate": 1.8938750718975456e-05,
-      "loss": 1.7366,
+      "loss": 1.7285,
       "step": 540
     },
     {
       "epoch": 0.5222007722007722,
-      "grad_norm": 3.5191142559051514,
+      "grad_norm": 3.8613996505737305,
       "learning_rate": 1.8934072915547978e-05,
-      "loss": 1.7118,
+      "loss": 1.7066,
       "step": 541
     },
     {
       "epoch": 0.5231660231660231,
-      "grad_norm": 3.8551297187805176,
+      "grad_norm": 3.9286131858825684,
       "learning_rate": 1.892938540561952e-05,
-      "loss": 1.6944,
+      "loss": 1.6912,
       "step": 542
     },
     {
       "epoch": 0.5241312741312741,
-      "grad_norm": 4.650771141052246,
+      "grad_norm": 4.836663246154785,
       "learning_rate": 1.8924688194282862e-05,
-      "loss": 1.6372,
+      "loss": 1.6302,
       "step": 543
     },
     {
       "epoch": 0.525096525096525,
-      "grad_norm": 4.20098352432251,
+      "grad_norm": 4.330646991729736,
       "learning_rate": 1.8919981286641336e-05,
-      "loss": 1.6815,
+      "loss": 1.6782,
       "step": 544
     },
     {
       "epoch": 0.5260617760617761,
-      "grad_norm": 3.853182315826416,
+      "grad_norm": 3.8098952770233154,
       "learning_rate": 1.891526468780881e-05,
-      "loss": 1.6536,
+      "loss": 1.6517,
       "step": 545
     },
     {
       "epoch": 0.527027027027027,
-      "grad_norm": 3.648350954055786,
+      "grad_norm": 3.9491610527038574,
       "learning_rate": 1.8910538402909658e-05,
-      "loss": 1.6387,
+      "loss": 1.6386,
       "step": 546
     },
     {
       "epoch": 0.527992277992278,
-      "grad_norm": 3.5757272243499756,
+      "grad_norm": 3.6207921504974365,
       "learning_rate": 1.89058024370788e-05,
-      "loss": 1.6526,
+      "loss": 1.6443,
       "step": 547
     },
     {
       "epoch": 0.528957528957529,
-      "grad_norm": 3.996739149093628,
+      "grad_norm": 4.105937957763672,
       "learning_rate": 1.890105679546167e-05,
-      "loss": 1.7263,
+      "loss": 1.7159,
       "step": 548
     },
     {
       "epoch": 0.5299227799227799,
-      "grad_norm": 3.7277631759643555,
+      "grad_norm": 3.756416082382202,
       "learning_rate": 1.8896301483214207e-05,
-      "loss": 1.6864,
+      "loss": 1.6858,
       "step": 549
     },
     {
       "epoch": 0.5308880308880309,
-      "grad_norm": 3.406763792037964,
+      "grad_norm": 3.4839091300964355,
       "learning_rate": 1.8891536505502865e-05,
-      "loss": 1.6984,
+      "loss": 1.699,
       "step": 550
     },
     {
       "epoch": 0.5318532818532818,
-      "grad_norm": 4.051969051361084,
+      "grad_norm": 4.1251420974731445,
       "learning_rate": 1.8886761867504595e-05,
-      "loss": 1.657,
+      "loss": 1.6538,
       "step": 551
     },
     {
       "epoch": 0.5328185328185329,
-      "grad_norm": 3.5501787662506104,
+      "grad_norm": 3.564776659011841,
       "learning_rate": 1.8881977574406842e-05,
-      "loss": 1.6886,
+      "loss": 1.682,
       "step": 552
     },
     {
       "epoch": 0.5337837837837838,
-      "grad_norm": 3.881225109100342,
+      "grad_norm": 4.057106971740723,
       "learning_rate": 1.887718363140754e-05,
-      "loss": 1.7374,
+      "loss": 1.7306,
       "step": 553
     },
     {
       "epoch": 0.5347490347490348,
-      "grad_norm": 4.034554481506348,
+      "grad_norm": 4.495352745056152,
       "learning_rate": 1.8872380043715117e-05,
-      "loss": 1.6904,
+      "loss": 1.6874,
       "step": 554
     },
     {
       "epoch": 0.5357142857142857,
-      "grad_norm": 3.6535398960113525,
+      "grad_norm": 3.6222147941589355,
       "learning_rate": 1.8867566816548468e-05,
-      "loss": 1.7026,
+      "loss": 1.6988,
       "step": 555
     },
     {
       "epoch": 0.5366795366795367,
-      "grad_norm": 3.7659504413604736,
+      "grad_norm": 4.525403022766113,
       "learning_rate": 1.8862743955136966e-05,
-      "loss": 1.6495,
+      "loss": 1.6341,
       "step": 556
     },
     {
       "epoch": 0.5376447876447876,
-      "grad_norm": 4.140193939208984,
+      "grad_norm": 3.9802591800689697,
       "learning_rate": 1.885791146472045e-05,
-      "loss": 1.7337,
+      "loss": 1.7349,
       "step": 557
     },
     {
       "epoch": 0.5386100386100386,
-      "grad_norm": 3.7351629734039307,
+      "grad_norm": 4.020228385925293,
       "learning_rate": 1.8853069350549223e-05,
-      "loss": 1.7096,
+      "loss": 1.7073,
       "step": 558
     },
     {
       "epoch": 0.5395752895752896,
-      "grad_norm": 5.067814826965332,
+      "grad_norm": 5.290850639343262,
       "learning_rate": 1.884821761788404e-05,
-      "loss": 1.6834,
+      "loss": 1.6853,
       "step": 559
     },
     {
       "epoch": 0.5405405405405406,
-      "grad_norm": 3.812654733657837,
+      "grad_norm": 3.7491772174835205,
       "learning_rate": 1.8843356271996106e-05,
-      "loss": 1.6414,
+      "loss": 1.6317,
       "step": 560
     },
     {
       "epoch": 0.5415057915057915,
-      "grad_norm": 4.1293158531188965,
+      "grad_norm": 4.447864532470703,
       "learning_rate": 1.8838485318167078e-05,
-      "loss": 1.7375,
+      "loss": 1.7302,
       "step": 561
     },
     {
       "epoch": 0.5424710424710425,
-      "grad_norm": 4.6960625648498535,
+      "grad_norm": 4.68576192855835,
       "learning_rate": 1.8833604761689045e-05,
-      "loss": 1.6657,
+      "loss": 1.6605,
       "step": 562
     },
     {
       "epoch": 0.5434362934362934,
-      "grad_norm": 3.738614320755005,
+      "grad_norm": 4.18660306930542,
       "learning_rate": 1.8828714607864524e-05,
-      "loss": 1.6944,
+      "loss": 1.6828,
       "step": 563
     },
     {
       "epoch": 0.5444015444015444,
-      "grad_norm": 4.760887145996094,
+      "grad_norm": 4.8297553062438965,
       "learning_rate": 1.882381486200647e-05,
-      "loss": 1.6449,
+      "loss": 1.643,
       "step": 564
     },
     {
       "epoch": 0.5453667953667953,
-      "grad_norm": 3.6925551891326904,
+      "grad_norm": 3.8056528568267822,
       "learning_rate": 1.881890552943826e-05,
-      "loss": 1.6279,
+      "loss": 1.6223,
       "step": 565
     },
     {
       "epoch": 0.5463320463320464,
-      "grad_norm": 3.980975389480591,
+      "grad_norm": 4.114663600921631,
       "learning_rate": 1.8813986615493664e-05,
-      "loss": 1.7143,
+      "loss": 1.7098,
       "step": 566
     },
     {
       "epoch": 0.5472972972972973,
-      "grad_norm": 3.7689743041992188,
+      "grad_norm": 3.889599084854126,
       "learning_rate": 1.8809058125516894e-05,
-      "loss": 1.6662,
+      "loss": 1.6607,
       "step": 567
     },
     {
       "epoch": 0.5482625482625483,
-      "grad_norm": 3.663180351257324,
+      "grad_norm": 3.799654483795166,
       "learning_rate": 1.8804120064862547e-05,
-      "loss": 1.671,
+      "loss": 1.6657,
       "step": 568
     },
     {
       "epoch": 0.5492277992277992,
-      "grad_norm": 4.199854373931885,
+      "grad_norm": 4.369001388549805,
       "learning_rate": 1.879917243889562e-05,
-      "loss": 1.7025,
+      "loss": 1.7019,
       "step": 569
     },
     {
       "epoch": 0.5501930501930502,
-      "grad_norm": 3.8860297203063965,
+      "grad_norm": 3.728576898574829,
       "learning_rate": 1.8794215252991504e-05,
-      "loss": 1.7047,
+      "loss": 1.6988,
       "step": 570
     },
     {
       "epoch": 0.5511583011583011,
-      "grad_norm": 4.648730754852295,
+      "grad_norm": 4.725655555725098,
       "learning_rate": 1.8789248512535978e-05,
-      "loss": 1.6904,
+      "loss": 1.6872,
       "step": 571
     },
     {
       "epoch": 0.5521235521235521,
-      "grad_norm": 4.810901641845703,
+      "grad_norm": 4.884721279144287,
       "learning_rate": 1.87842722229252e-05,
-      "loss": 1.7054,
+      "loss": 1.6962,
       "step": 572
     },
     {
       "epoch": 0.553088803088803,
-      "grad_norm": 4.9697041511535645,
+      "grad_norm": 4.998650550842285,
       "learning_rate": 1.87792863895657e-05,
-      "loss": 1.6567,
+      "loss": 1.6493,
       "step": 573
     },
     {
       "epoch": 0.5540540540540541,
-      "grad_norm": 3.7920072078704834,
+      "grad_norm": 3.9220736026763916,
       "learning_rate": 1.8774291017874384e-05,
-      "loss": 1.6878,
+      "loss": 1.682,
       "step": 574
     },
     {
       "epoch": 0.555019305019305,
-      "grad_norm": 3.894486904144287,
+      "grad_norm": 3.899780511856079,
       "learning_rate": 1.8769286113278515e-05,
-      "loss": 1.6685,
+      "loss": 1.6528,
       "step": 575
     },
     {
       "epoch": 0.555984555984556,
-      "grad_norm": 3.9627492427825928,
+      "grad_norm": 4.038690567016602,
       "learning_rate": 1.8764271681215714e-05,
-      "loss": 1.6961,
+      "loss": 1.6871,
       "step": 576
     },
     {
       "epoch": 0.556949806949807,
-      "grad_norm": 4.0938591957092285,
+      "grad_norm": 3.8037397861480713,
       "learning_rate": 1.8759247727133956e-05,
-      "loss": 1.6871,
+      "loss": 1.6784,
       "step": 577
     },
     {
       "epoch": 0.5579150579150579,
-      "grad_norm": 3.893789768218994,
+      "grad_norm": 3.8892714977264404,
       "learning_rate": 1.8754214256491564e-05,
-      "loss": 1.6215,
+      "loss": 1.616,
       "step": 578
     },
     {
       "epoch": 0.5588803088803089,
-      "grad_norm": 3.9757490158081055,
+      "grad_norm": 4.161351680755615,
       "learning_rate": 1.8749171274757184e-05,
-      "loss": 1.6701,
+      "loss": 1.6576,
       "step": 579
     },
     {
       "epoch": 0.5598455598455598,
-      "grad_norm": 4.111543655395508,
+      "grad_norm": 4.06419563293457,
       "learning_rate": 1.874411878740982e-05,
-      "loss": 1.7002,
+      "loss": 1.702,
       "step": 580
     },
     {
       "epoch": 0.5608108108108109,
-      "grad_norm": 4.3574299812316895,
+      "grad_norm": 4.575520992279053,
       "learning_rate": 1.873905679993878e-05,
-      "loss": 1.656,
+      "loss": 1.6532,
       "step": 581
     },
     {
       "epoch": 0.5617760617760618,
-      "grad_norm": 4.900210857391357,
+      "grad_norm": 4.708689212799072,
       "learning_rate": 1.873398531784371e-05,
-      "loss": 1.6724,
+      "loss": 1.6634,
       "step": 582
     },
     {
       "epoch": 0.5627413127413128,
-      "grad_norm": 4.666040420532227,
+      "grad_norm": 4.68616247177124,
       "learning_rate": 1.8728904346634563e-05,
-      "loss": 1.6004,
+      "loss": 1.5927,
       "step": 583
     },
     {
       "epoch": 0.5637065637065637,
-      "grad_norm": 4.129469871520996,
+      "grad_norm": 4.331949234008789,
       "learning_rate": 1.872381389183161e-05,
-      "loss": 1.7019,
+      "loss": 1.6948,
       "step": 584
     },
     {
       "epoch": 0.5646718146718147,
-      "grad_norm": 4.4677934646606445,
+      "grad_norm": 4.408545017242432,
       "learning_rate": 1.8718713958965413e-05,
-      "loss": 1.7334,
+      "loss": 1.7376,
       "step": 585
     },
     {
       "epoch": 0.5656370656370656,
-      "grad_norm": 3.5727407932281494,
+      "grad_norm": 3.5919249057769775,
       "learning_rate": 1.8713604553576842e-05,
-      "loss": 1.6449,
+      "loss": 1.6406,
       "step": 586
     },
     {
       "epoch": 0.5666023166023166,
-      "grad_norm": 3.8256876468658447,
+      "grad_norm": 3.7751920223236084,
       "learning_rate": 1.870848568121705e-05,
-      "loss": 1.6638,
+      "loss": 1.6609,
       "step": 587
     },
     {
       "epoch": 0.5675675675675675,
-      "grad_norm": 3.4252405166625977,
+      "grad_norm": 3.493919849395752,
       "learning_rate": 1.8703357347447487e-05,
-      "loss": 1.6841,
+      "loss": 1.6825,
       "step": 588
     },
     {
       "epoch": 0.5685328185328186,
-      "grad_norm": 4.170039176940918,
+      "grad_norm": 4.1167707443237305,
       "learning_rate": 1.8698219557839875e-05,
-      "loss": 1.5711,
+      "loss": 1.5643,
       "step": 589
     },
     {
       "epoch": 0.5694980694980695,
-      "grad_norm": 3.6318793296813965,
+      "grad_norm": 3.659196138381958,
       "learning_rate": 1.8693072317976205e-05,
-      "loss": 1.6643,
+      "loss": 1.6545,
       "step": 590
     },
     {
       "epoch": 0.5704633204633205,
-      "grad_norm": 3.745554208755493,
+      "grad_norm": 3.9532222747802734,
       "learning_rate": 1.868791563344874e-05,
-      "loss": 1.6813,
+      "loss": 1.6778,
       "step": 591
     },
     {
       "epoch": 0.5714285714285714,
-      "grad_norm": 5.6081461906433105,
+      "grad_norm": 5.388474941253662,
       "learning_rate": 1.868274950986001e-05,
-      "loss": 1.6829,
+      "loss": 1.676,
       "step": 592
     },
     {
       "epoch": 0.5723938223938224,
-      "grad_norm": 4.0746541023254395,
+      "grad_norm": 4.04116678237915,
       "learning_rate": 1.8677573952822793e-05,
-      "loss": 1.7016,
+      "loss": 1.6961,
       "step": 593
     },
     {
       "epoch": 0.5733590733590733,
-      "grad_norm": 5.50642728805542,
+      "grad_norm": 4.922682762145996,
       "learning_rate": 1.867238896796012e-05,
-      "loss": 1.6679,
+      "loss": 1.6584,
       "step": 594
     },
     {
       "epoch": 0.5743243243243243,
-      "grad_norm": 4.05189847946167,
+      "grad_norm": 3.9228298664093018,
       "learning_rate": 1.8667194560905255e-05,
-      "loss": 1.6469,
+      "loss": 1.641,
       "step": 595
     },
     {
       "epoch": 0.5752895752895753,
-      "grad_norm": 4.773981094360352,
+      "grad_norm": 4.6174774169921875,
       "learning_rate": 1.8661990737301714e-05,
-      "loss": 1.6685,
+      "loss": 1.6689,
       "step": 596
     },
     {
       "epoch": 0.5762548262548263,
-      "grad_norm": 4.233147144317627,
+      "grad_norm": 4.087060928344727,
       "learning_rate": 1.865677750280323e-05,
-      "loss": 1.7635,
+      "loss": 1.7592,
       "step": 597
     },
     {
       "epoch": 0.5772200772200772,
-      "grad_norm": 5.221375465393066,
+      "grad_norm": 5.132890701293945,
       "learning_rate": 1.8651554863073776e-05,
-      "loss": 1.6962,
+      "loss": 1.6936,
       "step": 598
     },
     {
       "epoch": 0.5781853281853282,
-      "grad_norm": 3.8712682723999023,
+      "grad_norm": 3.6421070098876953,
       "learning_rate": 1.864632282378753e-05,
-      "loss": 1.6369,
+      "loss": 1.6374,
       "step": 599
     },
     {
       "epoch": 0.5791505791505791,
-      "grad_norm": 4.542074203491211,
+      "grad_norm": 4.667139530181885,
       "learning_rate": 1.864108139062888e-05,
-      "loss": 1.6512,
+      "loss": 1.6481,
       "step": 600
     },
     {
       "epoch": 0.5801158301158301,
-      "grad_norm": 5.4050374031066895,
+      "grad_norm": 4.830512046813965,
       "learning_rate": 1.8635830569292434e-05,
-      "loss": 1.7831,
+      "loss": 1.7709,
       "step": 601
     },
     {
       "epoch": 0.581081081081081,
-      "grad_norm": 4.166981220245361,
+      "grad_norm": 4.243442058563232,
       "learning_rate": 1.863057036548299e-05,
-      "loss": 1.7131,
+      "loss": 1.7026,
       "step": 602
     },
     {
       "epoch": 0.582046332046332,
-      "grad_norm": 5.486953258514404,
+      "grad_norm": 5.475966930389404,
       "learning_rate": 1.862530078491554e-05,
-      "loss": 1.6969,
+      "loss": 1.6967,
       "step": 603
     },
     {
       "epoch": 0.583011583011583,
-      "grad_norm": 4.638899326324463,
+      "grad_norm": 4.3399858474731445,
       "learning_rate": 1.862002183331527e-05,
-      "loss": 1.6511,
+      "loss": 1.6432,
       "step": 604
     },
     {
       "epoch": 0.583976833976834,
-      "grad_norm": 4.579814910888672,
+      "grad_norm": 4.829336166381836,
       "learning_rate": 1.8614733516417543e-05,
-      "loss": 1.6889,
+      "loss": 1.6811,
       "step": 605
     },
     {
       "epoch": 0.584942084942085,
-      "grad_norm": 5.79071044921875,
+      "grad_norm": 5.62183141708374,
       "learning_rate": 1.860943583996789e-05,
-      "loss": 1.654,
+      "loss": 1.6441,
       "step": 606
     },
     {
       "epoch": 0.5859073359073359,
-      "grad_norm": 3.7253856658935547,
+      "grad_norm": 3.679896354675293,
       "learning_rate": 1.8604128809722024e-05,
-      "loss": 1.6732,
+      "loss": 1.6679,
       "step": 607
     },
     {
       "epoch": 0.5868725868725869,
-      "grad_norm": 4.581823825836182,
+      "grad_norm": 4.416443824768066,
       "learning_rate": 1.859881243144581e-05,
-      "loss": 1.6502,
+      "loss": 1.6415,
       "step": 608
     },
     {
       "epoch": 0.5878378378378378,
-      "grad_norm": 4.885397911071777,
+      "grad_norm": 4.440003871917725,
       "learning_rate": 1.8593486710915277e-05,
-      "loss": 1.7279,
+      "loss": 1.7181,
       "step": 609
     },
     {
       "epoch": 0.5888030888030888,
-      "grad_norm": 3.970855236053467,
+      "grad_norm": 3.8060402870178223,
       "learning_rate": 1.8588151653916596e-05,
-      "loss": 1.6674,
+      "loss": 1.6629,
       "step": 610
     },
     {
       "epoch": 0.5897683397683398,
-      "grad_norm": 4.714148044586182,
+      "grad_norm": 4.427528381347656,
       "learning_rate": 1.858280726624609e-05,
-      "loss": 1.6763,
+      "loss": 1.6689,
       "step": 611
     },
     {
       "epoch": 0.5907335907335908,
-      "grad_norm": 6.035184860229492,
+      "grad_norm": 5.576624393463135,
       "learning_rate": 1.8577453553710213e-05,
-      "loss": 1.7452,
+      "loss": 1.7389,
       "step": 612
     },
     {
       "epoch": 0.5916988416988417,
-      "grad_norm": 4.146158695220947,
+      "grad_norm": 3.950612783432007,
       "learning_rate": 1.8572090522125553e-05,
-      "loss": 1.6846,
+      "loss": 1.6746,
       "step": 613
     },
     {
       "epoch": 0.5926640926640927,
-      "grad_norm": 5.308503150939941,
+      "grad_norm": 5.045098304748535,
       "learning_rate": 1.8566718177318825e-05,
-      "loss": 1.6529,
+      "loss": 1.6472,
       "step": 614
     },
     {
       "epoch": 0.5936293436293436,
-      "grad_norm": 4.472123146057129,
+      "grad_norm": 4.416777610778809,
       "learning_rate": 1.856133652512685e-05,
-      "loss": 1.734,
+      "loss": 1.7312,
       "step": 615
     },
     {
       "epoch": 0.5945945945945946,
-      "grad_norm": 3.907163143157959,
+      "grad_norm": 3.9182238578796387,
       "learning_rate": 1.8555945571396584e-05,
-      "loss": 1.6416,
+      "loss": 1.641,
       "step": 616
     },
     {
       "epoch": 0.5955598455598455,
-      "grad_norm": 4.636082172393799,
+      "grad_norm": 4.311014175415039,
       "learning_rate": 1.8550545321985065e-05,
-      "loss": 1.6213,
+      "loss": 1.6134,
       "step": 617
     },
     {
       "epoch": 0.5965250965250966,
-      "grad_norm": 3.847663640975952,
+      "grad_norm": 3.72153639793396,
       "learning_rate": 1.8545135782759446e-05,
-      "loss": 1.6378,
+      "loss": 1.6339,
       "step": 618
     },
     {
       "epoch": 0.5974903474903475,
-      "grad_norm": 4.3884501457214355,
+      "grad_norm": 4.565605640411377,
       "learning_rate": 1.8539716959596967e-05,
-      "loss": 1.7255,
+      "loss": 1.7145,
       "step": 619
     },
     {
       "epoch": 0.5984555984555985,
-      "grad_norm": 3.7094476222991943,
+      "grad_norm": 3.6757571697235107,
       "learning_rate": 1.853428885838495e-05,
-      "loss": 1.6741,
+      "loss": 1.6705,
       "step": 620
     },
     {
       "epoch": 0.5994208494208494,
-      "grad_norm": 3.8970134258270264,
+      "grad_norm": 4.00325870513916,
       "learning_rate": 1.852885148502081e-05,
-      "loss": 1.6916,
+      "loss": 1.6918,
       "step": 621
     },
     {
       "epoch": 0.6003861003861004,
-      "grad_norm": 3.7777106761932373,
+      "grad_norm": 3.7909343242645264,
       "learning_rate": 1.8523404845412028e-05,
-      "loss": 1.6407,
+      "loss": 1.6311,
       "step": 622
     },
     {
       "epoch": 0.6013513513513513,
-      "grad_norm": 3.556906223297119,
+      "grad_norm": 3.77614688873291,
       "learning_rate": 1.851794894547615e-05,
-      "loss": 1.6999,
+      "loss": 1.6967,
       "step": 623
     },
     {
       "epoch": 0.6023166023166023,
-      "grad_norm": 4.2413105964660645,
+      "grad_norm": 4.4654412269592285,
       "learning_rate": 1.8512483791140787e-05,
-      "loss": 1.6502,
+      "loss": 1.6446,
       "step": 624
     },
     {
       "epoch": 0.6032818532818532,
-      "grad_norm": 3.9186458587646484,
+      "grad_norm": 4.21907901763916,
       "learning_rate": 1.85070093883436e-05,
-      "loss": 1.5961,
+      "loss": 1.5894,
       "step": 625
     },
     {
       "epoch": 0.6042471042471043,
-      "grad_norm": 3.8897972106933594,
+      "grad_norm": 4.456218242645264,
       "learning_rate": 1.8501525743032305e-05,
-      "loss": 1.7424,
+      "loss": 1.7427,
       "step": 626
     },
     {
       "epoch": 0.6052123552123552,
-      "grad_norm": 4.06066370010376,
+      "grad_norm": 4.269322395324707,
       "learning_rate": 1.8496032861164657e-05,
-      "loss": 1.6369,
+      "loss": 1.6397,
       "step": 627
     },
     {
       "epoch": 0.6061776061776062,
-      "grad_norm": 3.7971489429473877,
+      "grad_norm": 3.7456743717193604,
       "learning_rate": 1.8490530748708442e-05,
-      "loss": 1.7034,
+      "loss": 1.6994,
       "step": 628
     },
     {
       "epoch": 0.6071428571428571,
-      "grad_norm": 3.7843589782714844,
+      "grad_norm": 3.7403011322021484,
       "learning_rate": 1.8485019411641477e-05,
-      "loss": 1.6269,
+      "loss": 1.6166,
       "step": 629
     },
     {
       "epoch": 0.6081081081081081,
-      "grad_norm": 3.9630510807037354,
+      "grad_norm": 3.929677963256836,
       "learning_rate": 1.8479498855951607e-05,
-      "loss": 1.7404,
+      "loss": 1.7351,
       "step": 630
     },
     {
       "epoch": 0.609073359073359,
-      "grad_norm": 3.554232120513916,
+      "grad_norm": 3.499879837036133,
       "learning_rate": 1.8473969087636683e-05,
-      "loss": 1.7156,
+      "loss": 1.7089,
       "step": 631
     },
     {
       "epoch": 0.61003861003861,
-      "grad_norm": 4.078527450561523,
+      "grad_norm": 3.9015913009643555,
       "learning_rate": 1.8468430112704574e-05,
-      "loss": 1.6219,
+      "loss": 1.6126,
       "step": 632
     },
     {
       "epoch": 0.611003861003861,
-      "grad_norm": 3.7438254356384277,
+      "grad_norm": 3.8820247650146484,
       "learning_rate": 1.8462881937173144e-05,
-      "loss": 1.6589,
+      "loss": 1.6572,
       "step": 633
     },
     {
       "epoch": 0.611969111969112,
-      "grad_norm": 3.675246238708496,
+      "grad_norm": 3.781386613845825,
       "learning_rate": 1.8457324567070255e-05,
-      "loss": 1.7171,
+      "loss": 1.7147,
       "step": 634
     },
     {
       "epoch": 0.612934362934363,
-      "grad_norm": 4.8114519119262695,
+      "grad_norm": 4.752954483032227,
       "learning_rate": 1.845175800843376e-05,
-      "loss": 1.6456,
+      "loss": 1.6447,
       "step": 635
     },
     {
       "epoch": 0.6138996138996139,
-      "grad_norm": 4.08369255065918,
+      "grad_norm": 4.235104084014893,
       "learning_rate": 1.84461822673115e-05,
-      "loss": 1.7574,
+      "loss": 1.7549,
       "step": 636
     },
     {
       "epoch": 0.6148648648648649,
-      "grad_norm": 4.7406206130981445,
+      "grad_norm": 4.64028787612915,
       "learning_rate": 1.8440597349761286e-05,
-      "loss": 1.6294,
+      "loss": 1.6236,
       "step": 637
     },
     {
       "epoch": 0.6158301158301158,
-      "grad_norm": 4.751537322998047,
+      "grad_norm": 4.94498872756958,
       "learning_rate": 1.8435003261850895e-05,
-      "loss": 1.641,
+      "loss": 1.6375,
       "step": 638
     },
     {
       "epoch": 0.6167953667953668,
-      "grad_norm": 4.042719841003418,
+      "grad_norm": 4.195744037628174,
       "learning_rate": 1.8429400009658076e-05,
-      "loss": 1.5904,
+      "loss": 1.6153,
       "step": 639
     },
     {
       "epoch": 0.6177606177606177,
-      "grad_norm": 3.7075490951538086,
+      "grad_norm": 4.0266032218933105,
       "learning_rate": 1.842378759927053e-05,
-      "loss": 1.6742,
+      "loss": 1.6733,
       "step": 640
     },
     {
       "epoch": 0.6187258687258688,
-      "grad_norm": 3.8176772594451904,
+      "grad_norm": 3.9950990676879883,
       "learning_rate": 1.8418166036785912e-05,
-      "loss": 1.6862,
+      "loss": 1.6775,
       "step": 641
     },
     {
       "epoch": 0.6196911196911197,
-      "grad_norm": 3.418301820755005,
+      "grad_norm": 3.339768409729004,
       "learning_rate": 1.8412535328311813e-05,
-      "loss": 1.6231,
+      "loss": 1.6232,
       "step": 642
     },
     {
       "epoch": 0.6206563706563707,
-      "grad_norm": 3.5319151878356934,
+      "grad_norm": 3.5728282928466797,
       "learning_rate": 1.8406895479965766e-05,
-      "loss": 1.7387,
+      "loss": 1.7386,
       "step": 643
     },
     {
       "epoch": 0.6216216216216216,
-      "grad_norm": 4.254901885986328,
+      "grad_norm": 4.101585388183594,
       "learning_rate": 1.8401246497875238e-05,
-      "loss": 1.6421,
+      "loss": 1.6379,
       "step": 644
     },
     {
       "epoch": 0.6225868725868726,
-      "grad_norm": 4.021042823791504,
+      "grad_norm": 4.004106044769287,
       "learning_rate": 1.8395588388177606e-05,
-      "loss": 1.6828,
+      "loss": 1.6794,
       "step": 645
     },
     {
       "epoch": 0.6235521235521235,
-      "grad_norm": 3.8519716262817383,
+      "grad_norm": 3.664290428161621,
       "learning_rate": 1.838992115702018e-05,
-      "loss": 1.613,
+      "loss": 1.6022,
       "step": 646
     },
     {
       "epoch": 0.6245173745173745,
-      "grad_norm": 5.556555271148682,
+      "grad_norm": 5.374147415161133,
       "learning_rate": 1.8384244810560166e-05,
-      "loss": 1.6835,
+      "loss": 1.672,
       "step": 647
     },
     {
       "epoch": 0.6254826254826255,
-      "grad_norm": 5.064825057983398,
+      "grad_norm": 4.9376726150512695,
       "learning_rate": 1.8378559354964686e-05,
-      "loss": 1.6579,
+      "loss": 1.6519,
       "step": 648
     },
     {
       "epoch": 0.6264478764478765,
-      "grad_norm": 3.874995470046997,
+      "grad_norm": 3.877178907394409,
       "learning_rate": 1.8372864796410748e-05,
-      "loss": 1.6936,
+      "loss": 1.6815,
       "step": 649
     },
     {
       "epoch": 0.6274131274131274,
-      "grad_norm": 3.5412216186523438,
+      "grad_norm": 3.5297791957855225,
       "learning_rate": 1.8367161141085256e-05,
-      "loss": 1.6235,
+      "loss": 1.6173,
       "step": 650
     },
     {
       "epoch": 0.6283783783783784,
-      "grad_norm": 3.8917672634124756,
+      "grad_norm": 4.228473663330078,
       "learning_rate": 1.8361448395184995e-05,
-      "loss": 1.7078,
+      "loss": 1.7003,
       "step": 651
     },
     {
       "epoch": 0.6293436293436293,
-      "grad_norm": 3.8688552379608154,
+      "grad_norm": 3.884369373321533,
       "learning_rate": 1.835572656491663e-05,
-      "loss": 1.6606,
+      "loss": 1.6539,
       "step": 652
     },
     {
       "epoch": 0.6303088803088803,
-      "grad_norm": 4.006204605102539,
+      "grad_norm": 4.309850692749023,
       "learning_rate": 1.8349995656496692e-05,
-      "loss": 1.66,
+      "loss": 1.6479,
       "step": 653
     },
     {
       "epoch": 0.6312741312741312,
-      "grad_norm": 4.492037773132324,
+      "grad_norm": 4.515528678894043,
       "learning_rate": 1.8344255676151572e-05,
-      "loss": 1.6344,
+      "loss": 1.6308,
       "step": 654
     },
     {
       "epoch": 0.6322393822393823,
-      "grad_norm": 5.894545555114746,
+      "grad_norm": 5.842560291290283,
       "learning_rate": 1.8338506630117527e-05,
-      "loss": 1.7011,
+      "loss": 1.6924,
       "step": 655
     },
     {
       "epoch": 0.6332046332046332,
-      "grad_norm": 3.9362146854400635,
+      "grad_norm": 3.946831464767456,
       "learning_rate": 1.8332748524640657e-05,
-      "loss": 1.5894,
+      "loss": 1.5733,
       "step": 656
     },
     {
       "epoch": 0.6341698841698842,
-      "grad_norm": 4.3061113357543945,
+      "grad_norm": 4.146152019500732,
       "learning_rate": 1.8326981365976903e-05,
-      "loss": 1.6885,
+      "loss": 1.6886,
       "step": 657
     },
     {
       "epoch": 0.6351351351351351,
-      "grad_norm": 5.677443504333496,
+      "grad_norm": 5.751983642578125,
       "learning_rate": 1.832120516039205e-05,
-      "loss": 1.6736,
+      "loss": 1.6761,
       "step": 658
     },
     {
       "epoch": 0.6361003861003861,
-      "grad_norm": 3.6564252376556396,
+      "grad_norm": 3.7259654998779297,
       "learning_rate": 1.8315419914161697e-05,
-      "loss": 1.5815,
+      "loss": 1.5717,
       "step": 659
     },
     {
       "epoch": 0.637065637065637,
-      "grad_norm": 8.331316947937012,
+      "grad_norm": 8.071708679199219,
       "learning_rate": 1.8309625633571283e-05,
-      "loss": 1.6649,
+      "loss": 1.6622,
       "step": 660
     },
     {
       "epoch": 0.638030888030888,
-      "grad_norm": 4.027668476104736,
+      "grad_norm": 4.02902364730835,
       "learning_rate": 1.830382232491605e-05,
-      "loss": 1.645,
+      "loss": 1.6399,
       "step": 661
     },
     {
       "epoch": 0.638996138996139,
-      "grad_norm": 5.594932556152344,
+      "grad_norm": 5.5180230140686035,
       "learning_rate": 1.829800999450106e-05,
-      "loss": 1.687,
+      "loss": 1.6739,
       "step": 662
     },
     {
       "epoch": 0.63996138996139,
-      "grad_norm": 4.846828937530518,
+      "grad_norm": 4.736966133117676,
       "learning_rate": 1.829218864864116e-05,
-      "loss": 1.6671,
+      "loss": 1.6545,
       "step": 663
     },
     {
       "epoch": 0.640926640926641,
-      "grad_norm": 3.602202892303467,
+      "grad_norm": 3.6216139793395996,
       "learning_rate": 1.8286358293661007e-05,
-      "loss": 1.6521,
+      "loss": 1.6485,
       "step": 664
     },
     {
       "epoch": 0.6418918918918919,
-      "grad_norm": 5.479918956756592,
+      "grad_norm": 5.401447772979736,
       "learning_rate": 1.828051893589504e-05,
-      "loss": 1.6227,
+      "loss": 1.6226,
       "step": 665
     },
     {
       "epoch": 0.6428571428571429,
-      "grad_norm": 4.227159023284912,
+      "grad_norm": 4.094205379486084,
       "learning_rate": 1.827467058168748e-05,
-      "loss": 1.6611,
+      "loss": 1.6556,
       "step": 666
     },
     {
       "epoch": 0.6438223938223938,
-      "grad_norm": 4.943421840667725,
+      "grad_norm": 4.560113430023193,
       "learning_rate": 1.8268813237392323e-05,
-      "loss": 1.6425,
+      "loss": 1.6401,
       "step": 667
     },
     {
       "epoch": 0.6447876447876448,
-      "grad_norm": 3.53085994720459,
+      "grad_norm": 3.7017579078674316,
       "learning_rate": 1.826294690937333e-05,
-      "loss": 1.6765,
+      "loss": 1.6654,
       "step": 668
     },
     {
       "epoch": 0.6457528957528957,
-      "grad_norm": 3.979865312576294,
+      "grad_norm": 3.8793370723724365,
       "learning_rate": 1.8257071604004025e-05,
-      "loss": 1.7101,
+      "loss": 1.7068,
       "step": 669
     },
     {
       "epoch": 0.6467181467181468,
-      "grad_norm": 3.7377078533172607,
+      "grad_norm": 3.8790905475616455,
       "learning_rate": 1.8251187327667682e-05,
-      "loss": 1.6569,
+      "loss": 1.6505,
       "step": 670
     },
     {
       "epoch": 0.6476833976833977,
-      "grad_norm": 3.7672817707061768,
+      "grad_norm": 3.7488832473754883,
       "learning_rate": 1.824529408675733e-05,
-      "loss": 1.6771,
+      "loss": 1.668,
       "step": 671
     },
     {
       "epoch": 0.6486486486486487,
-      "grad_norm": 3.766531467437744,
+      "grad_norm": 3.658513069152832,
       "learning_rate": 1.8239391887675725e-05,
-      "loss": 1.6386,
+      "loss": 1.6286,
       "step": 672
     },
     {
       "epoch": 0.6496138996138996,
-      "grad_norm": 4.239291667938232,
+      "grad_norm": 4.241335391998291,
       "learning_rate": 1.8233480736835366e-05,
-      "loss": 1.7249,
+      "loss": 1.7069,
       "step": 673
     },
     {
       "epoch": 0.6505791505791506,
-      "grad_norm": 6.503885269165039,
+      "grad_norm": 5.861279487609863,
       "learning_rate": 1.822756064065847e-05,
-      "loss": 1.6037,
+      "loss": 1.6031,
       "step": 674
     },
     {
       "epoch": 0.6515444015444015,
-      "grad_norm": 3.618699073791504,
+      "grad_norm": 3.548616409301758,
       "learning_rate": 1.8221631605576983e-05,
-      "loss": 1.6653,
+      "loss": 1.6618,
       "step": 675
     },
     {
       "epoch": 0.6525096525096525,
-      "grad_norm": 4.299498081207275,
+      "grad_norm": 4.0124430656433105,
       "learning_rate": 1.8215693638032555e-05,
-      "loss": 1.6083,
+      "loss": 1.5946,
       "step": 676
     },
     {
       "epoch": 0.6534749034749034,
-      "grad_norm": 6.6480631828308105,
+      "grad_norm": 6.381943225860596,
       "learning_rate": 1.8209746744476538e-05,
-      "loss": 1.7064,
+      "loss": 1.7002,
       "step": 677
     },
     {
       "epoch": 0.6544401544401545,
-      "grad_norm": 3.544414520263672,
+      "grad_norm": 3.6448354721069336,
       "learning_rate": 1.820379093136999e-05,
-      "loss": 1.6489,
+      "loss": 1.6402,
       "step": 678
     },
     {
       "epoch": 0.6554054054054054,
-      "grad_norm": 5.126187324523926,
+      "grad_norm": 4.546553611755371,
       "learning_rate": 1.8197826205183657e-05,
-      "loss": 1.6937,
+      "loss": 1.6911,
       "step": 679
     },
     {
       "epoch": 0.6563706563706564,
-      "grad_norm": 5.806528091430664,
+      "grad_norm": 6.178905010223389,
       "learning_rate": 1.819185257239796e-05,
-      "loss": 1.7678,
+      "loss": 1.7585,
       "step": 680
     },
     {
       "epoch": 0.6573359073359073,
-      "grad_norm": 4.069647312164307,
+      "grad_norm": 4.14552116394043,
       "learning_rate": 1.8185870039503014e-05,
-      "loss": 1.6307,
+      "loss": 1.6276,
       "step": 681
     },
     {
       "epoch": 0.6583011583011583,
-      "grad_norm": 4.525701522827148,
+      "grad_norm": 4.552587985992432,
       "learning_rate": 1.8179878612998586e-05,
-      "loss": 1.6497,
+      "loss": 1.6402,
       "step": 682
     },
     {
       "epoch": 0.6592664092664092,
-      "grad_norm": 5.206305027008057,
+      "grad_norm": 5.610740661621094,
       "learning_rate": 1.8173878299394118e-05,
-      "loss": 1.7508,
+      "loss": 1.744,
       "step": 683
     },
     {
       "epoch": 0.6602316602316602,
-      "grad_norm": 3.8882358074188232,
+      "grad_norm": 3.845806837081909,
       "learning_rate": 1.8167869105208703e-05,
-      "loss": 1.6425,
+      "loss": 1.6328,
       "step": 684
     },
     {
       "epoch": 0.6611969111969112,
-      "grad_norm": 5.368058681488037,
+      "grad_norm": 5.7050461769104,
       "learning_rate": 1.816185103697108e-05,
-      "loss": 1.6902,
+      "loss": 1.6831,
       "step": 685
     },
     {
       "epoch": 0.6621621621621622,
-      "grad_norm": 3.8239119052886963,
+      "grad_norm": 3.79701566696167,
       "learning_rate": 1.8155824101219626e-05,
-      "loss": 1.5893,
+      "loss": 1.577,
       "step": 686
     },
     {
       "epoch": 0.6631274131274131,
-      "grad_norm": 5.2074198722839355,
+      "grad_norm": 5.665323734283447,
       "learning_rate": 1.814978830450237e-05,
-      "loss": 1.7116,
+      "loss": 1.7001,
       "step": 687
     },
     {
       "epoch": 0.6640926640926641,
-      "grad_norm": 3.5977511405944824,
+      "grad_norm": 3.6753861904144287,
       "learning_rate": 1.8143743653376944e-05,
-      "loss": 1.6007,
+      "loss": 1.5898,
       "step": 688
     },
     {
       "epoch": 0.665057915057915,
-      "grad_norm": 4.01193380355835,
+      "grad_norm": 3.9668290615081787,
       "learning_rate": 1.8137690154410617e-05,
-      "loss": 1.649,
+      "loss": 1.6461,
       "step": 689
     },
     {
       "epoch": 0.666023166023166,
-      "grad_norm": 4.035015106201172,
+      "grad_norm": 4.2066826820373535,
       "learning_rate": 1.8131627814180264e-05,
-      "loss": 1.6641,
+      "loss": 1.658,
       "step": 690
     },
     {
       "epoch": 0.666988416988417,
-      "grad_norm": 3.439990282058716,
+      "grad_norm": 3.485719680786133,
       "learning_rate": 1.8125556639272375e-05,
-      "loss": 1.627,
+      "loss": 1.6083,
       "step": 691
     },
     {
       "epoch": 0.667953667953668,
-      "grad_norm": 3.9323132038116455,
+      "grad_norm": 3.8689022064208984,
       "learning_rate": 1.8119476636283018e-05,
-      "loss": 1.688,
+      "loss": 1.668,
       "step": 692
     },
     {
       "epoch": 0.668918918918919,
-      "grad_norm": 4.216135025024414,
+      "grad_norm": 4.516170024871826,
       "learning_rate": 1.811338781181787e-05,
-      "loss": 1.6948,
+      "loss": 1.6995,
       "step": 693
     },
     {
       "epoch": 0.6698841698841699,
-      "grad_norm": 3.865110158920288,
+      "grad_norm": 3.772052049636841,
       "learning_rate": 1.8107290172492193e-05,
-      "loss": 1.6122,
+      "loss": 1.6103,
       "step": 694
     },
     {
       "epoch": 0.6708494208494209,
-      "grad_norm": 4.306641578674316,
+      "grad_norm": 5.059944152832031,
       "learning_rate": 1.8101183724930812e-05,
-      "loss": 1.7233,
+      "loss": 1.7232,
       "step": 695
     },
     {
       "epoch": 0.6718146718146718,
-      "grad_norm": 3.8325014114379883,
+      "grad_norm": 3.8924083709716797,
       "learning_rate": 1.809506847576813e-05,
-      "loss": 1.6368,
+      "loss": 1.6245,
       "step": 696
     },
     {
       "epoch": 0.6727799227799228,
-      "grad_norm": 5.356362819671631,
+      "grad_norm": 5.451060771942139,
       "learning_rate": 1.8088944431648118e-05,
       "loss": 1.651,
       "step": 697
     },
     {
       "epoch": 0.6737451737451737,
-      "grad_norm": 4.689187526702881,
+      "grad_norm": 4.8589372634887695,
       "learning_rate": 1.8082811599224293e-05,
-      "loss": 1.6605,
+      "loss": 1.6533,
       "step": 698
     },
     {
       "epoch": 0.6747104247104247,
-      "grad_norm": 4.608354568481445,
+      "grad_norm": 4.4537763595581055,
       "learning_rate": 1.8076669985159726e-05,
-      "loss": 1.6396,
+      "loss": 1.6324,
       "step": 699
     },
     {
       "epoch": 0.6756756756756757,
-      "grad_norm": 5.5505757331848145,
+      "grad_norm": 5.503808975219727,
       "learning_rate": 1.8070519596127027e-05,
-      "loss": 1.7028,
+      "loss": 1.6978,
       "step": 700
     },
     {
       "epoch": 0.6766409266409267,
-      "grad_norm": 3.8463668823242188,
+      "grad_norm": 3.595211982727051,
       "learning_rate": 1.8064360438808335e-05,
-      "loss": 1.6099,
+      "loss": 1.6047,
       "step": 701
     },
     {
       "epoch": 0.6776061776061776,
-      "grad_norm": 5.928972244262695,
+      "grad_norm": 5.901825904846191,
       "learning_rate": 1.805819251989533e-05,
-      "loss": 1.6517,
+      "loss": 1.6368,
       "step": 702
     },
     {
       "epoch": 0.6785714285714286,
-      "grad_norm": 3.801046848297119,
+      "grad_norm": 3.6578152179718018,
       "learning_rate": 1.8052015846089187e-05,
-      "loss": 1.6492,
+      "loss": 1.6417,
       "step": 703
     },
     {
       "epoch": 0.6795366795366795,
-      "grad_norm": 5.3330302238464355,
+      "grad_norm": 5.410284042358398,
       "learning_rate": 1.804583042410062e-05,
-      "loss": 1.6213,
+      "loss": 1.6116,
       "step": 704
     },
     {
       "epoch": 0.6805019305019305,
-      "grad_norm": 4.596437931060791,
+      "grad_norm": 4.585605621337891,
       "learning_rate": 1.8039636260649823e-05,
-      "loss": 1.6368,
+      "loss": 1.6281,
       "step": 705
     },
     {
       "epoch": 0.6814671814671814,
-      "grad_norm": 4.642145156860352,
+      "grad_norm": 4.398314476013184,
       "learning_rate": 1.8033433362466506e-05,
-      "loss": 1.618,
+      "loss": 1.6081,
       "step": 706
     },
     {
       "epoch": 0.6824324324324325,
-      "grad_norm": 5.829033374786377,
+      "grad_norm": 6.003376483917236,
       "learning_rate": 1.802722173628986e-05,
-      "loss": 1.6833,
+      "loss": 1.6677,
       "step": 707
     },
     {
       "epoch": 0.6833976833976834,
-      "grad_norm": 4.1263017654418945,
+      "grad_norm": 4.240137577056885,
       "learning_rate": 1.802100138886856e-05,
-      "loss": 1.717,
+      "loss": 1.7063,
       "step": 708
     },
     {
       "epoch": 0.6843629343629344,
-      "grad_norm": 5.381926536560059,
+      "grad_norm": 5.325348377227783,
       "learning_rate": 1.8014772326960758e-05,
-      "loss": 1.6017,
+      "loss": 1.5992,
       "step": 709
     },
     {
       "epoch": 0.6853281853281853,
-      "grad_norm": 6.798498630523682,
+      "grad_norm": 7.179402828216553,
       "learning_rate": 1.8008534557334064e-05,
-      "loss": 1.6535,
+      "loss": 1.6477,
       "step": 710
     },
     {
       "epoch": 0.6862934362934363,
-      "grad_norm": 3.4682328701019287,
+      "grad_norm": 3.5598175525665283,
       "learning_rate": 1.800228808676557e-05,
-      "loss": 1.5921,
+      "loss": 1.5865,
       "step": 711
     },
     {
       "epoch": 0.6872586872586872,
-      "grad_norm": 5.673734188079834,
+      "grad_norm": 6.021512985229492,
       "learning_rate": 1.7996032922041797e-05,
-      "loss": 1.6136,
+      "loss": 1.6111,
       "step": 712
     },
     {
       "epoch": 0.6882239382239382,
-      "grad_norm": 6.470987319946289,
+      "grad_norm": 6.4079790115356445,
       "learning_rate": 1.7989769069958728e-05,
-      "loss": 1.7331,
+      "loss": 1.7285,
       "step": 713
     },
     {
       "epoch": 0.6891891891891891,
-      "grad_norm": 3.7526819705963135,
+      "grad_norm": 4.205503463745117,
       "learning_rate": 1.798349653732178e-05,
-      "loss": 1.6889,
+      "loss": 1.6799,
       "step": 714
     },
     {
       "epoch": 0.6901544401544402,
-      "grad_norm": 7.26838493347168,
+      "grad_norm": 6.947514057159424,
       "learning_rate": 1.79772153309458e-05,
-      "loss": 1.6973,
+      "loss": 1.6983,
       "step": 715
     },
     {
       "epoch": 0.6911196911196911,
-      "grad_norm": 3.74743390083313,
+      "grad_norm": 3.834216833114624,
       "learning_rate": 1.7970925457655055e-05,
-      "loss": 1.682,
+      "loss": 1.6859,
       "step": 716
     },
     {
       "epoch": 0.6920849420849421,
-      "grad_norm": 4.3754191398620605,
+      "grad_norm": 4.444882392883301,
       "learning_rate": 1.7964626924283238e-05,
-      "loss": 1.5958,
+      "loss": 1.5848,
       "step": 717
     },
     {
       "epoch": 0.693050193050193,
-      "grad_norm": 4.73179292678833,
+      "grad_norm": 4.370157718658447,
       "learning_rate": 1.7958319737673444e-05,
-      "loss": 1.6401,
+      "loss": 1.6339,
       "step": 718
     },
     {
       "epoch": 0.694015444015444,
-      "grad_norm": 4.429830074310303,
+      "grad_norm": 4.75160026550293,
       "learning_rate": 1.795200390467817e-05,
-      "loss": 1.6231,
+      "loss": 1.6249,
       "step": 719
     },
     {
       "epoch": 0.694980694980695,
-      "grad_norm": 3.9136927127838135,
+      "grad_norm": 4.057354927062988,
       "learning_rate": 1.7945679432159302e-05,
-      "loss": 1.5997,
+      "loss": 1.5894,
       "step": 720
     },
     {
       "epoch": 0.6959459459459459,
-      "grad_norm": 4.052205562591553,
+      "grad_norm": 4.3197340965271,
       "learning_rate": 1.7939346326988127e-05,
-      "loss": 1.6286,
+      "loss": 1.6242,
       "step": 721
     },
     {
       "epoch": 0.696911196911197,
-      "grad_norm": 4.2305450439453125,
+      "grad_norm": 4.26387357711792,
       "learning_rate": 1.7933004596045294e-05,
-      "loss": 1.6407,
+      "loss": 1.6345,
       "step": 722
     },
     {
       "epoch": 0.6978764478764479,
-      "grad_norm": 3.89178729057312,
+      "grad_norm": 3.884181022644043,
       "learning_rate": 1.7926654246220843e-05,
-      "loss": 1.7152,
+      "loss": 1.7073,
       "step": 723
     },
     {
       "epoch": 0.6988416988416989,
-      "grad_norm": 4.049131870269775,
+      "grad_norm": 3.9794468879699707,
       "learning_rate": 1.7920295284414153e-05,
-      "loss": 1.7144,
+      "loss": 1.7097,
       "step": 724
     },
     {
       "epoch": 0.6998069498069498,
-      "grad_norm": 3.665480136871338,
+      "grad_norm": 3.9345505237579346,
       "learning_rate": 1.791392771753398e-05,
-      "loss": 1.6139,
+      "loss": 1.6044,
       "step": 725
     },
     {
       "epoch": 0.7007722007722008,
-      "grad_norm": 4.4121928215026855,
+      "grad_norm": 4.391848087310791,
       "learning_rate": 1.7907551552498423e-05,
-      "loss": 1.6958,
+      "loss": 1.6866,
       "step": 726
     },
     {
       "epoch": 0.7017374517374517,
-      "grad_norm": 4.521336078643799,
+      "grad_norm": 4.846347808837891,
       "learning_rate": 1.7901166796234922e-05,
-      "loss": 1.7486,
+      "loss": 1.7464,
       "step": 727
     },
     {
       "epoch": 0.7027027027027027,
-      "grad_norm": 3.2848575115203857,
+      "grad_norm": 3.301159381866455,
       "learning_rate": 1.789477345568025e-05,
-      "loss": 1.6371,
+      "loss": 1.6317,
       "step": 728
     },
     {
       "epoch": 0.7036679536679536,
-      "grad_norm": 3.7929630279541016,
+      "grad_norm": 3.796570301055908,
       "learning_rate": 1.78883715377805e-05,
       "loss": 1.5856,
       "step": 729
     },
     {
       "epoch": 0.7046332046332047,
-      "grad_norm": 3.8128833770751953,
+      "grad_norm": 3.7487294673919678,
       "learning_rate": 1.78819610494911e-05,
-      "loss": 1.6768,
+      "loss": 1.6722,
       "step": 730
     },
     {
       "epoch": 0.7055984555984556,
-      "grad_norm": 3.6518142223358154,
+      "grad_norm": 3.8048343658447266,
       "learning_rate": 1.7875541997776783e-05,
-      "loss": 1.5678,
+      "loss": 1.5727,
       "step": 731
     },
     {
       "epoch": 0.7065637065637066,
-      "grad_norm": 3.605537176132202,
+      "grad_norm": 3.8091816902160645,
       "learning_rate": 1.7869114389611574e-05,
-      "loss": 1.6645,
+      "loss": 1.6654,
       "step": 732
     },
     {
       "epoch": 0.7075289575289575,
-      "grad_norm": 3.7172515392303467,
+      "grad_norm": 3.8063769340515137,
       "learning_rate": 1.786267823197881e-05,
-      "loss": 1.6761,
+      "loss": 1.6722,
       "step": 733
     },
     {
       "epoch": 0.7084942084942085,
-      "grad_norm": 3.7993557453155518,
+      "grad_norm": 3.874403953552246,
       "learning_rate": 1.785623353187111e-05,
-      "loss": 1.5896,
+      "loss": 1.5982,
       "step": 734
     },
     {
       "epoch": 0.7094594594594594,
-      "grad_norm": 3.8158602714538574,
+      "grad_norm": 3.765474319458008,
       "learning_rate": 1.7849780296290368e-05,
-      "loss": 1.5756,
+      "loss": 1.5771,
       "step": 735
     },
     {
       "epoch": 0.7104247104247104,
-      "grad_norm": 3.8875603675842285,
+      "grad_norm": 3.943721294403076,
       "learning_rate": 1.7843318532247767e-05,
-      "loss": 1.6682,
+      "loss": 1.6647,
       "step": 736
     },
     {
       "epoch": 0.7113899613899614,
-      "grad_norm": 4.242062568664551,
+      "grad_norm": 4.203096389770508,
       "learning_rate": 1.783684824676374e-05,
-      "loss": 1.6656,
+      "loss": 1.6669,
       "step": 737
     },
     {
       "epoch": 0.7123552123552124,
-      "grad_norm": 3.6864781379699707,
+      "grad_norm": 3.8499934673309326,
       "learning_rate": 1.783036944686799e-05,
-      "loss": 1.6468,
+      "loss": 1.6395,
       "step": 738
     },
     {
       "epoch": 0.7133204633204633,
-      "grad_norm": 4.057651042938232,
+      "grad_norm": 3.983306884765625,
       "learning_rate": 1.782388213959946e-05,
-      "loss": 1.6584,
+      "loss": 1.6569,
       "step": 739
     },
     {
       "epoch": 0.7142857142857143,
-      "grad_norm": 4.905232906341553,
+      "grad_norm": 5.203009128570557,
       "learning_rate": 1.7817386332006346e-05,
-      "loss": 1.6772,
+      "loss": 1.6661,
       "step": 740
     },
     {
       "epoch": 0.7152509652509652,
-      "grad_norm": 3.8806796073913574,
+      "grad_norm": 4.16640043258667,
       "learning_rate": 1.781088203114607e-05,
-      "loss": 1.6806,
+      "loss": 1.6674,
       "step": 741
     },
     {
       "epoch": 0.7162162162162162,
-      "grad_norm": 5.752893447875977,
+      "grad_norm": 6.012908935546875,
       "learning_rate": 1.7804369244085283e-05,
-      "loss": 1.6034,
+      "loss": 1.6045,
       "step": 742
     },
     {
       "epoch": 0.7171814671814671,
-      "grad_norm": 4.513430118560791,
+      "grad_norm": 4.994954586029053,
       "learning_rate": 1.7797847977899873e-05,
-      "loss": 1.6523,
+      "loss": 1.6465,
       "step": 743
     },
     {
       "epoch": 0.7181467181467182,
-      "grad_norm": 4.028271198272705,
+      "grad_norm": 3.761045455932617,
       "learning_rate": 1.7791318239674914e-05,
-      "loss": 1.6314,
+      "loss": 1.621,
       "step": 744
     },
     {
       "epoch": 0.7191119691119691,
-      "grad_norm": 6.31402063369751,
+      "grad_norm": 6.32354211807251,
       "learning_rate": 1.7784780036504696e-05,
-      "loss": 1.7083,
+      "loss": 1.7018,
       "step": 745
     },
     {
       "epoch": 0.7200772200772201,
-      "grad_norm": 3.543903350830078,
+      "grad_norm": 3.732147455215454,
       "learning_rate": 1.7778233375492716e-05,
-      "loss": 1.6528,
+      "loss": 1.6479,
       "step": 746
     },
     {
       "epoch": 0.721042471042471,
-      "grad_norm": 5.082342624664307,
+      "grad_norm": 4.837199687957764,
       "learning_rate": 1.777167826375164e-05,
-      "loss": 1.679,
+      "loss": 1.6761,
       "step": 747
     },
     {
       "epoch": 0.722007722007722,
-      "grad_norm": 4.935264587402344,
+      "grad_norm": 5.0045671463012695,
       "learning_rate": 1.7765114708403332e-05,
-      "loss": 1.6521,
+      "loss": 1.642,
       "step": 748
     },
     {
       "epoch": 0.722972972972973,
-      "grad_norm": 3.5462920665740967,
+      "grad_norm": 3.548366069793701,
       "learning_rate": 1.7758542716578826e-05,
-      "loss": 1.6503,
+      "loss": 1.6391,
       "step": 749
     },
     {
       "epoch": 0.7239382239382239,
-      "grad_norm": 6.806006908416748,
+      "grad_norm": 6.565161228179932,
       "learning_rate": 1.7751962295418314e-05,
-      "loss": 1.6455,
+      "loss": 1.6357,
       "step": 750
     },
     {
       "epoch": 0.724903474903475,
-      "grad_norm": 3.723386526107788,
+      "grad_norm": 3.677649736404419,
       "learning_rate": 1.7745373452071154e-05,
-      "loss": 1.7378,
+      "loss": 1.7312,
       "step": 751
     },
     {
       "epoch": 0.7258687258687259,
-      "grad_norm": 4.961573600769043,
+      "grad_norm": 4.6709465980529785,
       "learning_rate": 1.7738776193695853e-05,
-      "loss": 1.6893,
+      "loss": 1.6812,
       "step": 752
     },
     {
       "epoch": 0.7268339768339769,
-      "grad_norm": 3.9472904205322266,
+      "grad_norm": 4.0239057540893555,
       "learning_rate": 1.7732170527460058e-05,
-      "loss": 1.6363,
+      "loss": 1.6367,
       "step": 753
     },
     {
       "epoch": 0.7277992277992278,
-      "grad_norm": 4.425893783569336,
+      "grad_norm": 4.505486011505127,
       "learning_rate": 1.7725556460540553e-05,
-      "loss": 1.6562,
+      "loss": 1.6482,
       "step": 754
     },
     {
       "epoch": 0.7287644787644788,
-      "grad_norm": 5.396843433380127,
+      "grad_norm": 5.364902973175049,
       "learning_rate": 1.7718934000123247e-05,
-      "loss": 1.6127,
+      "loss": 1.6102,
       "step": 755
     },
     {
       "epoch": 0.7297297297297297,
-      "grad_norm": 4.616225719451904,
+      "grad_norm": 4.719703674316406,
       "learning_rate": 1.7712303153403167e-05,
-      "loss": 1.6582,
+      "loss": 1.6548,
       "step": 756
     },
     {
       "epoch": 0.7306949806949807,
-      "grad_norm": 4.454399585723877,
+      "grad_norm": 4.107908725738525,
       "learning_rate": 1.7705663927584464e-05,
-      "loss": 1.6997,
+      "loss": 1.6994,
       "step": 757
     },
     {
       "epoch": 0.7316602316602316,
-      "grad_norm": 4.703960418701172,
+      "grad_norm": 4.762269973754883,
       "learning_rate": 1.769901632988037e-05,
-      "loss": 1.6392,
+      "loss": 1.6387,
       "step": 758
     },
     {
       "epoch": 0.7326254826254827,
-      "grad_norm": 3.2728753089904785,
+      "grad_norm": 3.364326238632202,
       "learning_rate": 1.7692360367513238e-05,
-      "loss": 1.6287,
+      "loss": 1.6233,
       "step": 759
     },
     {
       "epoch": 0.7335907335907336,
-      "grad_norm": 5.88960075378418,
+      "grad_norm": 5.190091609954834,
       "learning_rate": 1.7685696047714486e-05,
-      "loss": 1.6408,
+      "loss": 1.6292,
       "step": 760
     },
     {
       "epoch": 0.7345559845559846,
-      "grad_norm": 3.615349054336548,
+      "grad_norm": 3.807145118713379,
       "learning_rate": 1.7679023377724623e-05,
-      "loss": 1.61,
+      "loss": 1.6057,
       "step": 761
     },
     {
       "epoch": 0.7355212355212355,
-      "grad_norm": 4.272396087646484,
+      "grad_norm": 3.8820934295654297,
       "learning_rate": 1.7672342364793233e-05,
-      "loss": 1.7013,
+      "loss": 1.6935,
       "step": 762
     },
     {
       "epoch": 0.7364864864864865,
-      "grad_norm": 5.247366905212402,
+      "grad_norm": 5.047033786773682,
       "learning_rate": 1.7665653016178957e-05,
-      "loss": 1.5842,
+      "loss": 1.591,
       "step": 763
     },
     {
       "epoch": 0.7374517374517374,
-      "grad_norm": 3.7255361080169678,
+      "grad_norm": 3.6449434757232666,
       "learning_rate": 1.7658955339149497e-05,
-      "loss": 1.6795,
+      "loss": 1.676,
       "step": 764
     },
     {
       "epoch": 0.7384169884169884,
-      "grad_norm": 5.505873203277588,
+      "grad_norm": 4.963261127471924,
       "learning_rate": 1.7652249340981608e-05,
-      "loss": 1.6291,
+      "loss": 1.6217,
       "step": 765
     },
     {
       "epoch": 0.7393822393822393,
-      "grad_norm": 3.6880943775177,
+      "grad_norm": 3.691138744354248,
       "learning_rate": 1.764553502896107e-05,
-      "loss": 1.6307,
+      "loss": 1.6257,
       "step": 766
     },
     {
       "epoch": 0.7403474903474904,
-      "grad_norm": 4.854920864105225,
+      "grad_norm": 4.815586090087891,
       "learning_rate": 1.7638812410382714e-05,
-      "loss": 1.6176,
+      "loss": 1.6158,
       "step": 767
     },
     {
       "epoch": 0.7413127413127413,
-      "grad_norm": 4.184886455535889,
+      "grad_norm": 3.8075175285339355,
       "learning_rate": 1.763208149255038e-05,
-      "loss": 1.5853,
+      "loss": 1.5804,
       "step": 768
     },
     {
       "epoch": 0.7422779922779923,
-      "grad_norm": 3.6255054473876953,
+      "grad_norm": 3.687127113342285,
       "learning_rate": 1.7625342282776935e-05,
-      "loss": 1.5887,
+      "loss": 1.5802,
       "step": 769
     },
     {
       "epoch": 0.7432432432432432,
-      "grad_norm": 4.314204692840576,
+      "grad_norm": 3.7649049758911133,
       "learning_rate": 1.7618594788384252e-05,
-      "loss": 1.672,
+      "loss": 1.6649,
       "step": 770
     },
     {
       "epoch": 0.7442084942084942,
-      "grad_norm": 3.8578593730926514,
+      "grad_norm": 3.7448785305023193,
       "learning_rate": 1.7611839016703208e-05,
-      "loss": 1.6656,
+      "loss": 1.6658,
       "step": 771
     },
     {
       "epoch": 0.7451737451737451,
-      "grad_norm": 3.6460721492767334,
+      "grad_norm": 3.6919028759002686,
       "learning_rate": 1.7605074975073667e-05,
-      "loss": 1.589,
+      "loss": 1.5909,
       "step": 772
     },
     {
       "epoch": 0.7461389961389961,
-      "grad_norm": 3.87589430809021,
+      "grad_norm": 3.5019888877868652,
       "learning_rate": 1.759830267084448e-05,
-      "loss": 1.5959,
+      "loss": 1.5975,
       "step": 773
     },
     {
       "epoch": 0.747104247104247,
-      "grad_norm": 3.7357277870178223,
+      "grad_norm": 3.7080700397491455,
       "learning_rate": 1.7591522111373475e-05,
-      "loss": 1.6224,
+      "loss": 1.6076,
       "step": 774
     },
     {
       "epoch": 0.7480694980694981,
-      "grad_norm": 3.794887065887451,
+      "grad_norm": 3.8176276683807373,
       "learning_rate": 1.758473330402745e-05,
-      "loss": 1.6611,
+      "loss": 1.6527,
       "step": 775
     },
     {
       "epoch": 0.749034749034749,
-      "grad_norm": 3.853482961654663,
+      "grad_norm": 3.767636299133301,
       "learning_rate": 1.757793625618217e-05,
-      "loss": 1.676,
+      "loss": 1.6704,
       "step": 776
     },
     {
       "epoch": 0.75,
-      "grad_norm": 4.004652500152588,
+      "grad_norm": 4.114348411560059,
       "learning_rate": 1.7571130975222335e-05,
-      "loss": 1.5552,
+      "loss": 1.556,
       "step": 777
     },
     {
       "epoch": 0.750965250965251,
-      "grad_norm": 4.246206283569336,
+      "grad_norm": 4.0248565673828125,
       "learning_rate": 1.7564317468541608e-05,
-      "loss": 1.6533,
+      "loss": 1.65,
       "step": 778
     },
     {
       "epoch": 0.7519305019305019,
-      "grad_norm": 4.195562839508057,
+      "grad_norm": 4.275343894958496,
       "learning_rate": 1.7557495743542586e-05,
-      "loss": 1.6361,
+      "loss": 1.6303,
       "step": 779
     },
     {
       "epoch": 0.752895752895753,
-      "grad_norm": 4.303326606750488,
+      "grad_norm": 4.1412811279296875,
       "learning_rate": 1.7550665807636783e-05,
-      "loss": 1.6806,
+      "loss": 1.6786,
       "step": 780
     },
     {
       "epoch": 0.7538610038610039,
-      "grad_norm": 3.555691719055176,
+      "grad_norm": 3.600876808166504,
       "learning_rate": 1.7543827668244648e-05,
-      "loss": 1.67,
+      "loss": 1.6625,
       "step": 781
     },
     {
       "epoch": 0.7548262548262549,
-      "grad_norm": 3.820612907409668,
+      "grad_norm": 3.781980276107788,
       "learning_rate": 1.7536981332795535e-05,
-      "loss": 1.6068,
+      "loss": 1.5994,
       "step": 782
     },
     {
       "epoch": 0.7557915057915058,
-      "grad_norm": 3.809476375579834,
+      "grad_norm": 4.197249412536621,
       "learning_rate": 1.7530126808727703e-05,
-      "loss": 1.6098,
+      "loss": 1.5972,
       "step": 783
     },
     {
       "epoch": 0.7567567567567568,
-      "grad_norm": 4.754481792449951,
+      "grad_norm": 4.471271991729736,
       "learning_rate": 1.7523264103488314e-05,
-      "loss": 1.635,
+      "loss": 1.6273,
       "step": 784
     },
     {
       "epoch": 0.7577220077220077,
-      "grad_norm": 3.7754509449005127,
+      "grad_norm": 3.6252808570861816,
       "learning_rate": 1.7516393224533406e-05,
-      "loss": 1.6649,
+      "loss": 1.661,
       "step": 785
     },
     {
       "epoch": 0.7586872586872587,
-      "grad_norm": 5.550960540771484,
+      "grad_norm": 4.8447747230529785,
       "learning_rate": 1.7509514179327914e-05,
-      "loss": 1.6158,
+      "loss": 1.6092,
       "step": 786
     },
     {
       "epoch": 0.7596525096525096,
-      "grad_norm": 3.930851459503174,
+      "grad_norm": 3.906669855117798,
       "learning_rate": 1.750262697534563e-05,
-      "loss": 1.6389,
+      "loss": 1.6341,
       "step": 787
     },
     {
       "epoch": 0.7606177606177607,
-      "grad_norm": 6.380732536315918,
+      "grad_norm": 5.845431804656982,
       "learning_rate": 1.7495731620069216e-05,
-      "loss": 1.694,
+      "loss": 1.6823,
       "step": 788
     },
     {
       "epoch": 0.7615830115830116,
-      "grad_norm": 3.5392532348632812,
+      "grad_norm": 3.5788626670837402,
       "learning_rate": 1.7488828120990194e-05,
-      "loss": 1.5979,
+      "loss": 1.5904,
       "step": 789
     },
     {
       "epoch": 0.7625482625482626,
-      "grad_norm": 6.182634353637695,
+      "grad_norm": 5.991710186004639,
       "learning_rate": 1.748191648560893e-05,
-      "loss": 1.6096,
+      "loss": 1.607,
       "step": 790
     },
     {
       "epoch": 0.7635135135135135,
-      "grad_norm": 4.290624618530273,
+      "grad_norm": 4.137788772583008,
       "learning_rate": 1.7474996721434628e-05,
-      "loss": 1.6527,
+      "loss": 1.6416,
       "step": 791
     },
     {
       "epoch": 0.7644787644787645,
-      "grad_norm": 4.417576313018799,
+      "grad_norm": 4.646829605102539,
       "learning_rate": 1.7468068835985326e-05,
-      "loss": 1.5916,
+      "loss": 1.5875,
       "step": 792
     },
     {
       "epoch": 0.7654440154440154,
-      "grad_norm": 4.773061275482178,
+      "grad_norm": 4.539814472198486,
       "learning_rate": 1.746113283678789e-05,
-      "loss": 1.5791,
+      "loss": 1.5757,
       "step": 793
     },
     {
       "epoch": 0.7664092664092664,
-      "grad_norm": 3.7683379650115967,
+      "grad_norm": 3.8905677795410156,
       "learning_rate": 1.7454188731377988e-05,
-      "loss": 1.601,
+      "loss": 1.5935,
       "step": 794
     },
     {
       "epoch": 0.7673745173745173,
-      "grad_norm": 5.138625144958496,
+      "grad_norm": 5.144592761993408,
       "learning_rate": 1.744723652730011e-05,
-      "loss": 1.5937,
+      "loss": 1.5882,
       "step": 795
     },
     {
       "epoch": 0.7683397683397684,
-      "grad_norm": 3.958002805709839,
+      "grad_norm": 3.892186164855957,
       "learning_rate": 1.7440276232107537e-05,
-      "loss": 1.6804,
+      "loss": 1.6733,
       "step": 796
     },
     {
       "epoch": 0.7693050193050193,
-      "grad_norm": 3.6182544231414795,
+      "grad_norm": 3.7866945266723633,
       "learning_rate": 1.7433307853362344e-05,
-      "loss": 1.6123,
+      "loss": 1.6044,
       "step": 797
     },
     {
       "epoch": 0.7702702702702703,
-      "grad_norm": 4.057281970977783,
+      "grad_norm": 3.852226734161377,
       "learning_rate": 1.742633139863538e-05,
-      "loss": 1.6334,
+      "loss": 1.6346,
       "step": 798
     },
     {
       "epoch": 0.7712355212355212,
-      "grad_norm": 3.439039707183838,
+      "grad_norm": 3.567687749862671,
       "learning_rate": 1.7419346875506282e-05,
-      "loss": 1.5937,
+      "loss": 1.5886,
       "step": 799
     },
     {
       "epoch": 0.7722007722007722,
-      "grad_norm": 3.634779453277588,
+      "grad_norm": 3.6340935230255127,
       "learning_rate": 1.7412354291563442e-05,
-      "loss": 1.587,
+      "loss": 1.5877,
       "step": 800
     },
     {
       "epoch": 0.7731660231660231,
-      "grad_norm": 3.980450391769409,
+      "grad_norm": 3.8008763790130615,
       "learning_rate": 1.7405353654404017e-05,
-      "loss": 1.6383,
+      "loss": 1.6288,
       "step": 801
     },
     {
       "epoch": 0.7741312741312741,
-      "grad_norm": 3.7642734050750732,
+      "grad_norm": 3.8893048763275146,
       "learning_rate": 1.7398344971633902e-05,
-      "loss": 1.6375,
+      "loss": 1.6279,
       "step": 802
     },
     {
       "epoch": 0.775096525096525,
-      "grad_norm": 3.817166566848755,
+      "grad_norm": 3.905496835708618,
       "learning_rate": 1.7391328250867748e-05,
-      "loss": 1.6397,
+      "loss": 1.6627,
       "step": 803
     },
     {
       "epoch": 0.7760617760617761,
-      "grad_norm": 5.095615863800049,
+      "grad_norm": 4.917308807373047,
       "learning_rate": 1.738430349972893e-05,
-      "loss": 1.6001,
+      "loss": 1.5884,
       "step": 804
     },
     {
       "epoch": 0.777027027027027,
-      "grad_norm": 3.6880061626434326,
+      "grad_norm": 3.723726511001587,
       "learning_rate": 1.737727072584955e-05,
-      "loss": 1.6828,
+      "loss": 1.6755,
       "step": 805
     },
     {
       "epoch": 0.777992277992278,
-      "grad_norm": 3.768186569213867,
+      "grad_norm": 3.767915725708008,
       "learning_rate": 1.7370229936870425e-05,
-      "loss": 1.5754,
+      "loss": 1.5658,
       "step": 806
     },
     {
       "epoch": 0.778957528957529,
-      "grad_norm": 4.330945014953613,
+      "grad_norm": 4.204976558685303,
       "learning_rate": 1.7363181140441086e-05,
-      "loss": 1.513,
+      "loss": 1.5127,
       "step": 807
     },
     {
       "epoch": 0.7799227799227799,
-      "grad_norm": 4.447457790374756,
+      "grad_norm": 4.545008659362793,
       "learning_rate": 1.7356124344219757e-05,
-      "loss": 1.6115,
+      "loss": 1.6085,
       "step": 808
     },
     {
       "epoch": 0.7808880308880309,
-      "grad_norm": 4.9061279296875,
+      "grad_norm": 4.675474643707275,
       "learning_rate": 1.7349059555873348e-05,
-      "loss": 1.6968,
+      "loss": 1.6818,
       "step": 809
     },
     {
       "epoch": 0.7818532818532818,
-      "grad_norm": 5.54481840133667,
+      "grad_norm": 5.643557071685791,
       "learning_rate": 1.7341986783077466e-05,
-      "loss": 1.6252,
+      "loss": 1.6191,
       "step": 810
     },
     {
       "epoch": 0.7828185328185329,
-      "grad_norm": 3.801086664199829,
+      "grad_norm": 3.6744368076324463,
       "learning_rate": 1.7334906033516388e-05,
-      "loss": 1.5357,
+      "loss": 1.529,
       "step": 811
     },
     {
       "epoch": 0.7837837837837838,
-      "grad_norm": 3.63675594329834,
+      "grad_norm": 3.7496156692504883,
       "learning_rate": 1.7327817314883055e-05,
-      "loss": 1.5803,
+      "loss": 1.579,
       "step": 812
     },
     {
       "epoch": 0.7847490347490348,
-      "grad_norm": 4.226805686950684,
+      "grad_norm": 4.134372234344482,
       "learning_rate": 1.732072063487906e-05,
-      "loss": 1.6519,
+      "loss": 1.6477,
       "step": 813
     },
     {
       "epoch": 0.7857142857142857,
-      "grad_norm": 4.242192268371582,
+      "grad_norm": 4.090663433074951,
       "learning_rate": 1.7313616001214666e-05,
-      "loss": 1.629,
+      "loss": 1.6239,
       "step": 814
     },
     {
       "epoch": 0.7866795366795367,
-      "grad_norm": 4.957813739776611,
+      "grad_norm": 5.143884658813477,
       "learning_rate": 1.730650342160875e-05,
-      "loss": 1.6566,
+      "loss": 1.65,
       "step": 815
     },
     {
       "epoch": 0.7876447876447876,
-      "grad_norm": 4.036060333251953,
+      "grad_norm": 4.16240930557251,
       "learning_rate": 1.7299382903788844e-05,
-      "loss": 1.604,
+      "loss": 1.5953,
       "step": 816
     },
     {
       "epoch": 0.7886100386100386,
-      "grad_norm": 4.0332794189453125,
+      "grad_norm": 4.07451868057251,
       "learning_rate": 1.7292254455491095e-05,
-      "loss": 1.639,
+      "loss": 1.6349,
       "step": 817
     },
     {
       "epoch": 0.7895752895752896,
-      "grad_norm": 5.2410807609558105,
+      "grad_norm": 5.421199321746826,
       "learning_rate": 1.7285118084460265e-05,
-      "loss": 1.6602,
+      "loss": 1.6532,
       "step": 818
     },
     {
       "epoch": 0.7905405405405406,
-      "grad_norm": 3.6184473037719727,
+      "grad_norm": 3.5609583854675293,
       "learning_rate": 1.727797379844972e-05,
-      "loss": 1.6424,
+      "loss": 1.6358,
       "step": 819
     },
     {
       "epoch": 0.7915057915057915,
-      "grad_norm": 4.824613571166992,
+      "grad_norm": 5.00728702545166,
       "learning_rate": 1.7270821605221448e-05,
-      "loss": 1.655,
+      "loss": 1.6409,
       "step": 820
     },
     {
       "epoch": 0.7924710424710425,
-      "grad_norm": 5.858402252197266,
+      "grad_norm": 5.608322620391846,
       "learning_rate": 1.7263661512546e-05,
-      "loss": 1.6421,
+      "loss": 1.6339,
       "step": 821
     },
     {
       "epoch": 0.7934362934362934,
-      "grad_norm": 3.8016343116760254,
+      "grad_norm": 4.07835578918457,
       "learning_rate": 1.7256493528202523e-05,
-      "loss": 1.5408,
+      "loss": 1.5385,
       "step": 822
     },
     {
       "epoch": 0.7944015444015444,
-      "grad_norm": 5.814817428588867,
+      "grad_norm": 5.650109767913818,
       "learning_rate": 1.7249317659978736e-05,
-      "loss": 1.7092,
+      "loss": 1.7093,
       "step": 823
     },
     {
       "epoch": 0.7953667953667953,
-      "grad_norm": 3.719472885131836,
+      "grad_norm": 3.550572633743286,
       "learning_rate": 1.7242133915670923e-05,
-      "loss": 1.6248,
+      "loss": 1.6167,
       "step": 824
     },
     {
       "epoch": 0.7963320463320464,
-      "grad_norm": 4.978057861328125,
+      "grad_norm": 5.315227031707764,
       "learning_rate": 1.7234942303083925e-05,
-      "loss": 1.6871,
+      "loss": 1.6725,
       "step": 825
     },
     {
       "epoch": 0.7972972972972973,
-      "grad_norm": 4.768524646759033,
+      "grad_norm": 4.4534592628479,
       "learning_rate": 1.7227742830031132e-05,
-      "loss": 1.6106,
+      "loss": 1.6007,
       "step": 826
     },
     {
       "epoch": 0.7982625482625483,
-      "grad_norm": 3.8294737339019775,
+      "grad_norm": 4.165935039520264,
       "learning_rate": 1.7220535504334477e-05,
-      "loss": 1.6351,
+      "loss": 1.6366,
       "step": 827
     },
     {
       "epoch": 0.7992277992277992,
-      "grad_norm": 4.684476375579834,
+      "grad_norm": 4.738185405731201,
       "learning_rate": 1.7213320333824417e-05,
-      "loss": 1.5997,
+      "loss": 1.5923,
       "step": 828
     },
     {
       "epoch": 0.8001930501930502,
-      "grad_norm": 3.5390968322753906,
+      "grad_norm": 3.5068016052246094,
       "learning_rate": 1.7206097326339937e-05,
-      "loss": 1.6052,
+      "loss": 1.5954,
       "step": 829
     },
     {
       "epoch": 0.8011583011583011,
-      "grad_norm": 3.9698216915130615,
+      "grad_norm": 4.283023834228516,
       "learning_rate": 1.7198866489728537e-05,
-      "loss": 1.5969,
+      "loss": 1.5926,
       "step": 830
     },
     {
       "epoch": 0.8021235521235521,
-      "grad_norm": 4.555296421051025,
+      "grad_norm": 4.736050605773926,
       "learning_rate": 1.7191627831846226e-05,
-      "loss": 1.6549,
+      "loss": 1.6501,
       "step": 831
     },
     {
       "epoch": 0.803088803088803,
-      "grad_norm": 4.702564239501953,
+      "grad_norm": 4.952388763427734,
       "learning_rate": 1.71843813605575e-05,
-      "loss": 1.5909,
+      "loss": 1.594,
       "step": 832
     },
     {
       "epoch": 0.8040540540540541,
-      "grad_norm": 3.884777307510376,
+      "grad_norm": 4.1485371589660645,
       "learning_rate": 1.7177127083735352e-05,
-      "loss": 1.6226,
+      "loss": 1.6126,
       "step": 833
     },
     {
       "epoch": 0.805019305019305,
-      "grad_norm": 3.7954657077789307,
+      "grad_norm": 3.6254308223724365,
       "learning_rate": 1.716986500926126e-05,
-      "loss": 1.6062,
+      "loss": 1.6015,
       "step": 834
     },
     {
       "epoch": 0.805984555984556,
-      "grad_norm": 4.036071300506592,
+      "grad_norm": 4.123758316040039,
       "learning_rate": 1.716259514502516e-05,
-      "loss": 1.6093,
+      "loss": 1.6029,
       "step": 835
     },
     {
       "epoch": 0.806949806949807,
-      "grad_norm": 4.345009803771973,
+      "grad_norm": 4.5040740966796875,
       "learning_rate": 1.715531749892546e-05,
-      "loss": 1.5864,
+      "loss": 1.5841,
       "step": 836
     },
     {
       "epoch": 0.8079150579150579,
-      "grad_norm": 3.6161437034606934,
+      "grad_norm": 3.673802375793457,
       "learning_rate": 1.7148032078869027e-05,
-      "loss": 1.6097,
+      "loss": 1.5996,
       "step": 837
     },
     {
       "epoch": 0.8088803088803089,
-      "grad_norm": 3.721926689147949,
+      "grad_norm": 3.717745304107666,
       "learning_rate": 1.714073889277117e-05,
-      "loss": 1.6151,
+      "loss": 1.6095,
       "step": 838
     },
     {
       "epoch": 0.8098455598455598,
-      "grad_norm": 4.898469924926758,
+      "grad_norm": 4.739195823669434,
       "learning_rate": 1.713343794855562e-05,
-      "loss": 1.6506,
+      "loss": 1.6444,
       "step": 839
     },
     {
       "epoch": 0.8108108108108109,
-      "grad_norm": 4.193187236785889,
+      "grad_norm": 4.159494400024414,
       "learning_rate": 1.7126129254154568e-05,
-      "loss": 1.5663,
+      "loss": 1.563,
       "step": 840
     },
     {
       "epoch": 0.8117760617760618,
-      "grad_norm": 5.0784406661987305,
+      "grad_norm": 5.345032691955566,
       "learning_rate": 1.7118812817508603e-05,
-      "loss": 1.6037,
+      "loss": 1.5964,
       "step": 841
     },
     {
       "epoch": 0.8127413127413128,
-      "grad_norm": 3.959920644760132,
+      "grad_norm": 4.040149211883545,
       "learning_rate": 1.7111488646566728e-05,
-      "loss": 1.6617,
+      "loss": 1.6553,
       "step": 842
     },
     {
       "epoch": 0.8137065637065637,
-      "grad_norm": 4.190287113189697,
+      "grad_norm": 4.579443454742432,
       "learning_rate": 1.7104156749286352e-05,
-      "loss": 1.5909,
+      "loss": 1.5832,
       "step": 843
     },
     {
       "epoch": 0.8146718146718147,
-      "grad_norm": 3.578152894973755,
+      "grad_norm": 4.017465591430664,
       "learning_rate": 1.7096817133633283e-05,
-      "loss": 1.6875,
+      "loss": 1.6599,
       "step": 844
     },
     {
       "epoch": 0.8156370656370656,
-      "grad_norm": 3.6228625774383545,
+      "grad_norm": 3.5751612186431885,
       "learning_rate": 1.7089469807581705e-05,
-      "loss": 1.6502,
+      "loss": 1.6349,
       "step": 845
     },
     {
       "epoch": 0.8166023166023166,
-      "grad_norm": 3.968024730682373,
+      "grad_norm": 4.302496910095215,
       "learning_rate": 1.708211477911419e-05,
-      "loss": 1.562,
+      "loss": 1.564,
       "step": 846
     },
     {
       "epoch": 0.8175675675675675,
-      "grad_norm": 4.099898815155029,
+      "grad_norm": 4.013199806213379,
       "learning_rate": 1.7074752056221663e-05,
-      "loss": 1.6169,
+      "loss": 1.6035,
       "step": 847
     },
     {
       "epoch": 0.8185328185328186,
-      "grad_norm": 4.45789098739624,
+      "grad_norm": 5.070469379425049,
       "learning_rate": 1.7067381646903426e-05,
-      "loss": 1.6571,
+      "loss": 1.6447,
       "step": 848
     },
     {
       "epoch": 0.8194980694980695,
-      "grad_norm": 3.8516762256622314,
+      "grad_norm": 3.641484022140503,
       "learning_rate": 1.706000355916712e-05,
-      "loss": 1.5939,
+      "loss": 1.5899,
       "step": 849
     },
     {
       "epoch": 0.8204633204633205,
-      "grad_norm": 3.3639023303985596,
+      "grad_norm": 3.3147683143615723,
       "learning_rate": 1.705261780102874e-05,
-      "loss": 1.6121,
+      "loss": 1.6102,
       "step": 850
     },
     {
       "epoch": 0.8214285714285714,
-      "grad_norm": 4.166355609893799,
+      "grad_norm": 4.635789394378662,
       "learning_rate": 1.7045224380512597e-05,
-      "loss": 1.534,
+      "loss": 1.529,
       "step": 851
     },
     {
       "epoch": 0.8223938223938224,
-      "grad_norm": 3.9245617389678955,
+      "grad_norm": 4.032971382141113,
       "learning_rate": 1.7037823305651345e-05,
-      "loss": 1.6817,
+      "loss": 1.6809,
       "step": 852
     },
     {
       "epoch": 0.8233590733590733,
-      "grad_norm": 4.970295429229736,
+      "grad_norm": 5.239378452301025,
       "learning_rate": 1.7030414584485938e-05,
-      "loss": 1.6512,
+      "loss": 1.6413,
       "step": 853
     },
     {
       "epoch": 0.8243243243243243,
-      "grad_norm": 3.748103380203247,
+      "grad_norm": 3.5098118782043457,
       "learning_rate": 1.702299822506565e-05,
-      "loss": 1.6636,
+      "loss": 1.6292,
       "step": 854
     },
     {
       "epoch": 0.8252895752895753,
-      "grad_norm": 4.198969841003418,
+      "grad_norm": 4.286452770233154,
       "learning_rate": 1.701557423544805e-05,
-      "loss": 1.6339,
+      "loss": 1.6266,
       "step": 855
     },
     {
       "epoch": 0.8262548262548263,
-      "grad_norm": 3.630988359451294,
+      "grad_norm": 3.809485673904419,
       "learning_rate": 1.700814262369899e-05,
-      "loss": 1.6347,
+      "loss": 1.6319,
       "step": 856
     },
     {
       "epoch": 0.8272200772200772,
-      "grad_norm": 3.7846217155456543,
+      "grad_norm": 3.7764649391174316,
       "learning_rate": 1.700070339789262e-05,
-      "loss": 1.5582,
+      "loss": 1.5507,
       "step": 857
     },
     {
       "epoch": 0.8281853281853282,
-      "grad_norm": 4.339964866638184,
+      "grad_norm": 4.4860358238220215,
       "learning_rate": 1.699325656611134e-05,
-      "loss": 1.631,
+      "loss": 1.6215,
       "step": 858
     },
     {
       "epoch": 0.8291505791505791,
-      "grad_norm": 3.7031326293945312,
+      "grad_norm": 3.692535877227783,
       "learning_rate": 1.6985802136445832e-05,
-      "loss": 1.5525,
+      "loss": 1.5429,
       "step": 859
     },
     {
       "epoch": 0.8301158301158301,
-      "grad_norm": 4.129475116729736,
+      "grad_norm": 3.9702510833740234,
       "learning_rate": 1.6978340116995022e-05,
-      "loss": 1.6242,
+      "loss": 1.6161,
       "step": 860
     },
     {
       "epoch": 0.831081081081081,
-      "grad_norm": 3.5283150672912598,
+      "grad_norm": 3.5276710987091064,
       "learning_rate": 1.6970870515866086e-05,
       "loss": 1.5839,
       "step": 861
     },
     {
       "epoch": 0.832046332046332,
-      "grad_norm": 3.6117632389068604,
+      "grad_norm": 3.6565020084381104,
       "learning_rate": 1.6963393341174438e-05,
-      "loss": 1.6413,
+      "loss": 1.6349,
       "step": 862
     },
     {
       "epoch": 0.833011583011583,
-      "grad_norm": 3.7649621963500977,
+      "grad_norm": 3.8464736938476562,
       "learning_rate": 1.6955908601043724e-05,
-      "loss": 1.5737,
+      "loss": 1.5653,
       "step": 863
     },
     {
       "epoch": 0.833976833976834,
-      "grad_norm": 3.8086862564086914,
+      "grad_norm": 4.159819602966309,
       "learning_rate": 1.6948416303605796e-05,
-      "loss": 1.6382,
+      "loss": 1.6304,
       "step": 864
     },
     {
       "epoch": 0.834942084942085,
-      "grad_norm": 4.5638508796691895,
+      "grad_norm": 4.597322463989258,
       "learning_rate": 1.694091645700073e-05,
-      "loss": 1.5265,
+      "loss": 1.5229,
       "step": 865
     },
     {
       "epoch": 0.8359073359073359,
-      "grad_norm": 3.99662184715271,
+      "grad_norm": 3.767566680908203,
       "learning_rate": 1.6933409069376797e-05,
-      "loss": 1.6353,
+      "loss": 1.6235,
       "step": 866
     },
     {
       "epoch": 0.8368725868725869,
-      "grad_norm": 4.2375407218933105,
+      "grad_norm": 4.109277725219727,
       "learning_rate": 1.6925894148890473e-05,
-      "loss": 1.6229,
+      "loss": 1.6136,
       "step": 867
     },
     {
       "epoch": 0.8378378378378378,
-      "grad_norm": 3.9651808738708496,
+      "grad_norm": 3.950861692428589,
       "learning_rate": 1.691837170370639e-05,
-      "loss": 1.6518,
+      "loss": 1.6397,
       "step": 868
     },
     {
       "epoch": 0.8388030888030888,
-      "grad_norm": 4.371737480163574,
+      "grad_norm": 3.912580966949463,
       "learning_rate": 1.6910841741997395e-05,
-      "loss": 1.6,
+      "loss": 1.5916,
       "step": 869
     },
     {
       "epoch": 0.8397683397683398,
-      "grad_norm": 3.907066583633423,
+      "grad_norm": 3.833831787109375,
       "learning_rate": 1.6903304271944467e-05,
-      "loss": 1.5402,
+      "loss": 1.5308,
       "step": 870
     },
     {
       "epoch": 0.8407335907335908,
-      "grad_norm": 4.892452239990234,
+      "grad_norm": 4.4868927001953125,
       "learning_rate": 1.6895759301736757e-05,
-      "loss": 1.6335,
+      "loss": 1.6288,
       "step": 871
     },
     {
       "epoch": 0.8416988416988417,
-      "grad_norm": 4.123228073120117,
+      "grad_norm": 4.106043815612793,
       "learning_rate": 1.688820683957156e-05,
-      "loss": 1.6499,
+      "loss": 1.6427,
       "step": 872
     },
     {
       "epoch": 0.8426640926640927,
-      "grad_norm": 4.737846374511719,
+      "grad_norm": 4.748254776000977,
       "learning_rate": 1.6880646893654327e-05,
-      "loss": 1.6568,
+      "loss": 1.6417,
       "step": 873
     },
     {
       "epoch": 0.8436293436293436,
-      "grad_norm": 5.440667152404785,
+      "grad_norm": 5.299671173095703,
       "learning_rate": 1.6873079472198613e-05,
-      "loss": 1.6075,
+      "loss": 1.6014,
       "step": 874
     },
     {
       "epoch": 0.8445945945945946,
-      "grad_norm": 4.515224456787109,
+      "grad_norm": 4.233623504638672,
       "learning_rate": 1.6865504583426117e-05,
-      "loss": 1.586,
+      "loss": 1.5797,
       "step": 875
     },
     {
       "epoch": 0.8455598455598455,
-      "grad_norm": 6.032504081726074,
+      "grad_norm": 5.539729118347168,
       "learning_rate": 1.6857922235566642e-05,
-      "loss": 1.6334,
+      "loss": 1.6239,
       "step": 876
     },
     {
       "epoch": 0.8465250965250966,
-      "grad_norm": 4.806612968444824,
+      "grad_norm": 5.368149757385254,
       "learning_rate": 1.6850332436858087e-05,
-      "loss": 1.6048,
+      "loss": 1.607,
       "step": 877
     },
     {
       "epoch": 0.8474903474903475,
-      "grad_norm": 5.170681476593018,
+      "grad_norm": 4.759030818939209,
       "learning_rate": 1.6842735195546466e-05,
-      "loss": 1.6263,
+      "loss": 1.6117,
       "step": 878
     },
     {
       "epoch": 0.8484555984555985,
-      "grad_norm": 8.467127799987793,
+      "grad_norm": 8.628740310668945,
       "learning_rate": 1.683513051988586e-05,
-      "loss": 1.6341,
+      "loss": 1.6204,
       "step": 879
     },
     {
       "epoch": 0.8494208494208494,
-      "grad_norm": 4.348141670227051,
+      "grad_norm": 4.25000524520874,
       "learning_rate": 1.682751841813844e-05,
-      "loss": 1.5846,
+      "loss": 1.5715,
       "step": 880
     },
     {
       "epoch": 0.8503861003861004,
-      "grad_norm": 5.740167140960693,
+      "grad_norm": 5.508338928222656,
       "learning_rate": 1.681989889857444e-05,
-      "loss": 1.598,
+      "loss": 1.594,
       "step": 881
     },
     {
       "epoch": 0.8513513513513513,
-      "grad_norm": 4.303435802459717,
+      "grad_norm": 4.011481761932373,
       "learning_rate": 1.6812271969472145e-05,
-      "loss": 1.626,
+      "loss": 1.6202,
       "step": 882
     },
     {
       "epoch": 0.8523166023166023,
-      "grad_norm": 4.012930393218994,
+      "grad_norm": 3.8110480308532715,
       "learning_rate": 1.680463763911791e-05,
-      "loss": 1.6634,
+      "loss": 1.6587,
       "step": 883
     },
     {
       "epoch": 0.8532818532818532,
-      "grad_norm": 5.010847568511963,
+      "grad_norm": 4.410029411315918,
       "learning_rate": 1.679699591580611e-05,
-      "loss": 1.6119,
+      "loss": 1.6051,
       "step": 884
     },
     {
       "epoch": 0.8542471042471043,
-      "grad_norm": 5.053982734680176,
+      "grad_norm": 5.200742721557617,
       "learning_rate": 1.6789346807839165e-05,
-      "loss": 1.5977,
+      "loss": 1.572,
       "step": 885
     },
     {
       "epoch": 0.8552123552123552,
-      "grad_norm": 4.631558418273926,
+      "grad_norm": 4.255461692810059,
       "learning_rate": 1.6781690323527512e-05,
-      "loss": 1.6513,
+      "loss": 1.6463,
       "step": 886
     },
     {
       "epoch": 0.8561776061776062,
-      "grad_norm": 4.911116600036621,
+      "grad_norm": 4.77396821975708,
       "learning_rate": 1.677402647118961e-05,
-      "loss": 1.6718,
+      "loss": 1.6664,
       "step": 887
     },
     {
       "epoch": 0.8571428571428571,
-      "grad_norm": 4.912961006164551,
+      "grad_norm": 5.016380310058594,
       "learning_rate": 1.676635525915191e-05,
-      "loss": 1.6031,
+      "loss": 1.5932,
       "step": 888
     },
     {
       "epoch": 0.8581081081081081,
-      "grad_norm": 6.79826021194458,
+      "grad_norm": 6.258908271789551,
       "learning_rate": 1.675867669574888e-05,
-      "loss": 1.6854,
+      "loss": 1.6557,
       "step": 889
     },
     {
       "epoch": 0.859073359073359,
-      "grad_norm": 3.9823529720306396,
+      "grad_norm": 4.18447732925415,
       "learning_rate": 1.6750990789322948e-05,
-      "loss": 1.6156,
+      "loss": 1.6113,
       "step": 890
     },
     {
       "epoch": 0.86003861003861,
-      "grad_norm": 3.468554735183716,
+      "grad_norm": 3.354567050933838,
       "learning_rate": 1.6743297548224543e-05,
-      "loss": 1.5787,
+      "loss": 1.5788,
       "step": 891
     },
     {
       "epoch": 0.861003861003861,
-      "grad_norm": 5.073884010314941,
+      "grad_norm": 5.06559944152832,
       "learning_rate": 1.673559698081205e-05,
-      "loss": 1.5174,
+      "loss": 1.5079,
       "step": 892
     },
     {
       "epoch": 0.861969111969112,
-      "grad_norm": 4.256384372711182,
+      "grad_norm": 4.420685768127441,
       "learning_rate": 1.6727889095451814e-05,
-      "loss": 1.6093,
+      "loss": 1.6068,
       "step": 893
     },
     {
       "epoch": 0.862934362934363,
-      "grad_norm": 3.446455478668213,
+      "grad_norm": 3.6816649436950684,
       "learning_rate": 1.6720173900518137e-05,
-      "loss": 1.5703,
+      "loss": 1.5635,
       "step": 894
     },
     {
       "epoch": 0.8638996138996139,
-      "grad_norm": 6.296940803527832,
+      "grad_norm": 6.583624839782715,
       "learning_rate": 1.671245140439326e-05,
-      "loss": 1.6112,
+      "loss": 1.6136,
       "step": 895
     },
     {
       "epoch": 0.8648648648648649,
-      "grad_norm": 3.773432493209839,
+      "grad_norm": 3.8659839630126953,
       "learning_rate": 1.670472161546736e-05,
-      "loss": 1.6364,
+      "loss": 1.6356,
       "step": 896
     },
     {
       "epoch": 0.8658301158301158,
-      "grad_norm": 5.01470422744751,
+      "grad_norm": 5.110574722290039,
       "learning_rate": 1.669698454213852e-05,
-      "loss": 1.5684,
+      "loss": 1.5641,
       "step": 897
     },
     {
       "epoch": 0.8667953667953668,
-      "grad_norm": 4.782303810119629,
+      "grad_norm": 4.352468013763428,
       "learning_rate": 1.6689240192812765e-05,
-      "loss": 1.5945,
+      "loss": 1.5894,
       "step": 898
     },
     {
       "epoch": 0.8677606177606177,
-      "grad_norm": 4.499415397644043,
+      "grad_norm": 4.626100540161133,
       "learning_rate": 1.6681488575904003e-05,
-      "loss": 1.6817,
+      "loss": 1.6824,
       "step": 899
     },
     {
       "epoch": 0.8687258687258688,
-      "grad_norm": 4.399867057800293,
+      "grad_norm": 4.722452640533447,
       "learning_rate": 1.6673729699834046e-05,
-      "loss": 1.6501,
+      "loss": 1.6439,
       "step": 900
     },
     {
       "epoch": 0.8696911196911197,
-      "grad_norm": 3.5721189975738525,
+      "grad_norm": 3.519948720932007,
       "learning_rate": 1.6665963573032593e-05,
-      "loss": 1.5793,
+      "loss": 1.568,
       "step": 901
     },
     {
       "epoch": 0.8706563706563707,
-      "grad_norm": 4.1614274978637695,
+      "grad_norm": 3.9670755863189697,
       "learning_rate": 1.6658190203937218e-05,
-      "loss": 1.6342,
+      "loss": 1.6315,
       "step": 902
     },
     {
       "epoch": 0.8716216216216216,
-      "grad_norm": 4.05487585067749,
+      "grad_norm": 4.5356621742248535,
       "learning_rate": 1.6650409600993373e-05,
-      "loss": 1.6461,
+      "loss": 1.641,
       "step": 903
     },
     {
       "epoch": 0.8725868725868726,
-      "grad_norm": 5.089929103851318,
+      "grad_norm": 4.9804558753967285,
       "learning_rate": 1.6642621772654347e-05,
-      "loss": 1.5484,
+      "loss": 1.5491,
       "step": 904
     },
     {
       "epoch": 0.8735521235521235,
-      "grad_norm": 3.659691572189331,
+      "grad_norm": 3.856940746307373,
       "learning_rate": 1.663482672738131e-05,
-      "loss": 1.6263,
+      "loss": 1.6211,
       "step": 905
     },
     {
       "epoch": 0.8745173745173745,
-      "grad_norm": 5.445408344268799,
+      "grad_norm": 5.476229667663574,
       "learning_rate": 1.6627024473643244e-05,
-      "loss": 1.6315,
+      "loss": 1.63,
       "step": 906
     },
     {
       "epoch": 0.8754826254826255,
-      "grad_norm": 3.7327609062194824,
+      "grad_norm": 3.610341787338257,
       "learning_rate": 1.6619215019916983e-05,
-      "loss": 1.6925,
+      "loss": 1.691,
       "step": 907
     },
     {
       "epoch": 0.8764478764478765,
-      "grad_norm": 4.179985046386719,
+      "grad_norm": 4.48106575012207,
       "learning_rate": 1.6611398374687172e-05,
-      "loss": 1.5642,
+      "loss": 1.5623,
       "step": 908
     },
     {
       "epoch": 0.8774131274131274,
-      "grad_norm": 3.8566882610321045,
+      "grad_norm": 3.830092430114746,
       "learning_rate": 1.6603574546446274e-05,
-      "loss": 1.5556,
+      "loss": 1.5507,
       "step": 909
     },
     {
       "epoch": 0.8783783783783784,
-      "grad_norm": 3.8949801921844482,
+      "grad_norm": 4.035863876342773,
       "learning_rate": 1.659574354369455e-05,
-      "loss": 1.6518,
+      "loss": 1.6493,
       "step": 910
     },
     {
       "epoch": 0.8793436293436293,
-      "grad_norm": 3.5222058296203613,
+      "grad_norm": 3.491997241973877,
       "learning_rate": 1.6587905374940067e-05,
-      "loss": 1.6216,
+      "loss": 1.6165,
       "step": 911
     },
     {
       "epoch": 0.8803088803088803,
-      "grad_norm": 3.6448230743408203,
+      "grad_norm": 3.7057418823242188,
       "learning_rate": 1.6580060048698673e-05,
-      "loss": 1.6862,
+      "loss": 1.6828,
       "step": 912
     },
     {
       "epoch": 0.8812741312741312,
-      "grad_norm": 3.809236526489258,
+      "grad_norm": 3.911670207977295,
       "learning_rate": 1.6572207573493987e-05,
-      "loss": 1.549,
+      "loss": 1.5536,
       "step": 913
     },
     {
       "epoch": 0.8822393822393823,
-      "grad_norm": 4.041685581207275,
+      "grad_norm": 4.163541793823242,
       "learning_rate": 1.6564347957857395e-05,
-      "loss": 1.5944,
+      "loss": 1.591,
       "step": 914
     },
     {
       "epoch": 0.8832046332046332,
-      "grad_norm": 3.648710012435913,
+      "grad_norm": 3.6439478397369385,
       "learning_rate": 1.6556481210328053e-05,
-      "loss": 1.6113,
+      "loss": 1.6149,
       "step": 915
     },
     {
       "epoch": 0.8841698841698842,
-      "grad_norm": 3.8146796226501465,
+      "grad_norm": 3.894913911819458,
       "learning_rate": 1.6548607339452853e-05,
-      "loss": 1.602,
+      "loss": 1.5985,
       "step": 916
     },
     {
       "epoch": 0.8851351351351351,
-      "grad_norm": 4.380678653717041,
+      "grad_norm": 4.2380900382995605,
       "learning_rate": 1.6540726353786428e-05,
-      "loss": 1.6612,
+      "loss": 1.6495,
       "step": 917
     },
     {
       "epoch": 0.8861003861003861,
-      "grad_norm": 4.221372127532959,
+      "grad_norm": 4.298264503479004,
       "learning_rate": 1.6532838261891145e-05,
-      "loss": 1.6558,
+      "loss": 1.6495,
       "step": 918
     },
     {
       "epoch": 0.887065637065637,
-      "grad_norm": 5.726127624511719,
+      "grad_norm": 5.919624328613281,
       "learning_rate": 1.6524943072337094e-05,
-      "loss": 1.5827,
+      "loss": 1.572,
       "step": 919
     },
     {
       "epoch": 0.888030888030888,
-      "grad_norm": 4.030428409576416,
+      "grad_norm": 3.921599864959717,
       "learning_rate": 1.6517040793702066e-05,
-      "loss": 1.5973,
+      "loss": 1.5919,
       "step": 920
     },
     {
       "epoch": 0.888996138996139,
-      "grad_norm": 5.082960605621338,
+      "grad_norm": 5.421648979187012,
       "learning_rate": 1.6509131434571564e-05,
-      "loss": 1.6047,
+      "loss": 1.5943,
       "step": 921
     },
     {
       "epoch": 0.88996138996139,
-      "grad_norm": 4.497354984283447,
+      "grad_norm": 4.719018936157227,
       "learning_rate": 1.6501215003538776e-05,
-      "loss": 1.582,
+      "loss": 1.5708,
       "step": 922
     },
     {
       "epoch": 0.890926640926641,
-      "grad_norm": 4.039971828460693,
+      "grad_norm": 4.276191711425781,
       "learning_rate": 1.6493291509204582e-05,
-      "loss": 1.5584,
+      "loss": 1.5486,
       "step": 923
     },
     {
       "epoch": 0.8918918918918919,
-      "grad_norm": 4.771897792816162,
+      "grad_norm": 5.185001850128174,
       "learning_rate": 1.6485360960177525e-05,
-      "loss": 1.567,
+      "loss": 1.5544,
       "step": 924
     },
     {
       "epoch": 0.8928571428571429,
-      "grad_norm": 3.8797128200531006,
+      "grad_norm": 3.831700325012207,
       "learning_rate": 1.6477423365073825e-05,
-      "loss": 1.6056,
+      "loss": 1.604,
       "step": 925
     },
     {
       "epoch": 0.8938223938223938,
-      "grad_norm": 5.256024360656738,
+      "grad_norm": 5.699657917022705,
       "learning_rate": 1.646947873251735e-05,
-      "loss": 1.61,
+      "loss": 1.5994,
       "step": 926
     },
     {
       "epoch": 0.8947876447876448,
-      "grad_norm": 4.378505706787109,
+      "grad_norm": 4.4281134605407715,
       "learning_rate": 1.6461527071139613e-05,
-      "loss": 1.5766,
+      "loss": 1.5652,
       "step": 927
     },
     {
       "epoch": 0.8957528957528957,
-      "grad_norm": 3.816833257675171,
+      "grad_norm": 3.7446024417877197,
       "learning_rate": 1.645356838957976e-05,
-      "loss": 1.6357,
+      "loss": 1.6282,
       "step": 928
     },
     {
       "epoch": 0.8967181467181468,
-      "grad_norm": 3.6121904850006104,
+      "grad_norm": 3.7554051876068115,
       "learning_rate": 1.644560269648458e-05,
-      "loss": 1.5561,
+      "loss": 1.5531,
       "step": 929
     },
     {
       "epoch": 0.8976833976833977,
-      "grad_norm": 4.310316562652588,
+      "grad_norm": 4.249633312225342,
       "learning_rate": 1.6437630000508466e-05,
-      "loss": 1.5876,
+      "loss": 1.5819,
       "step": 930
     },
     {
       "epoch": 0.8986486486486487,
-      "grad_norm": 3.828895092010498,
+      "grad_norm": 3.800032138824463,
       "learning_rate": 1.642965031031342e-05,
-      "loss": 1.611,
+      "loss": 1.605,
       "step": 931
     },
     {
       "epoch": 0.8996138996138996,
-      "grad_norm": 3.4154212474823,
+      "grad_norm": 3.407468795776367,
       "learning_rate": 1.6421663634569045e-05,
-      "loss": 1.6266,
+      "loss": 1.6296,
       "step": 932
     },
     {
       "epoch": 0.9005791505791506,
-      "grad_norm": 4.432816028594971,
+      "grad_norm": 4.52619743347168,
       "learning_rate": 1.641366998195254e-05,
-      "loss": 1.5617,
+      "loss": 1.5542,
       "step": 933
     },
     {
       "epoch": 0.9015444015444015,
-      "grad_norm": 4.561981201171875,
+      "grad_norm": 4.676898956298828,
       "learning_rate": 1.6405669361148668e-05,
-      "loss": 1.5748,
+      "loss": 1.5707,
       "step": 934
     },
     {
       "epoch": 0.9025096525096525,
-      "grad_norm": 3.6001009941101074,
+      "grad_norm": 3.588561773300171,
       "learning_rate": 1.6397661780849785e-05,
-      "loss": 1.5985,
+      "loss": 1.5963,
       "step": 935
     },
     {
       "epoch": 0.9034749034749034,
-      "grad_norm": 3.6477224826812744,
+      "grad_norm": 3.5512993335723877,
       "learning_rate": 1.6389647249755788e-05,
-      "loss": 1.5923,
+      "loss": 1.5825,
       "step": 936
     },
     {
       "epoch": 0.9044401544401545,
-      "grad_norm": 4.120946407318115,
+      "grad_norm": 3.9842844009399414,
       "learning_rate": 1.6381625776574135e-05,
-      "loss": 1.6292,
+      "loss": 1.6153,
       "step": 937
     },
     {
       "epoch": 0.9054054054054054,
-      "grad_norm": 3.8856008052825928,
+      "grad_norm": 3.767000675201416,
       "learning_rate": 1.6373597370019825e-05,
-      "loss": 1.5703,
+      "loss": 1.5662,
       "step": 938
     },
     {
       "epoch": 0.9063706563706564,
-      "grad_norm": 4.099815368652344,
+      "grad_norm": 3.807650327682495,
       "learning_rate": 1.6365562038815394e-05,
-      "loss": 1.5469,
+      "loss": 1.5449,
       "step": 939
     },
     {
       "epoch": 0.9073359073359073,
-      "grad_norm": 3.5519697666168213,
+      "grad_norm": 3.6314284801483154,
       "learning_rate": 1.6357519791690893e-05,
-      "loss": 1.6077,
+      "loss": 1.6018,
       "step": 940
     },
     {
       "epoch": 0.9083011583011583,
-      "grad_norm": 4.991375923156738,
+      "grad_norm": 4.356168746948242,
       "learning_rate": 1.634947063738389e-05,
-      "loss": 1.6249,
+      "loss": 1.6252,
       "step": 941
     },
     {
       "epoch": 0.9092664092664092,
-      "grad_norm": 4.817924976348877,
+      "grad_norm": 4.603969097137451,
       "learning_rate": 1.6341414584639464e-05,
-      "loss": 1.6458,
+      "loss": 1.6382,
       "step": 942
     },
     {
       "epoch": 0.9102316602316602,
-      "grad_norm": 4.194146156311035,
+      "grad_norm": 3.9769392013549805,
       "learning_rate": 1.633335164221018e-05,
-      "loss": 1.5919,
+      "loss": 1.5845,
       "step": 943
     },
     {
       "epoch": 0.9111969111969112,
-      "grad_norm": 6.744943618774414,
+      "grad_norm": 6.792914867401123,
       "learning_rate": 1.632528181885609e-05,
-      "loss": 1.5598,
+      "loss": 1.5547,
       "step": 944
     },
     {
       "epoch": 0.9121621621621622,
-      "grad_norm": 3.4473798274993896,
+      "grad_norm": 3.4049627780914307,
       "learning_rate": 1.6317205123344725e-05,
-      "loss": 1.6495,
+      "loss": 1.6468,
       "step": 945
     },
     {
       "epoch": 0.9131274131274131,
-      "grad_norm": 6.090734004974365,
+      "grad_norm": 5.710813999176025,
       "learning_rate": 1.6309121564451084e-05,
-      "loss": 1.611,
+      "loss": 1.6003,
       "step": 946
     },
     {
       "epoch": 0.9140926640926641,
-      "grad_norm": 4.007740497589111,
+      "grad_norm": 3.9127869606018066,
       "learning_rate": 1.6301031150957616e-05,
-      "loss": 1.5825,
+      "loss": 1.5785,
       "step": 947
     },
     {
       "epoch": 0.915057915057915,
-      "grad_norm": 3.920323133468628,
+      "grad_norm": 3.8964462280273438,
       "learning_rate": 1.6292933891654217e-05,
-      "loss": 1.5947,
+      "loss": 1.5921,
       "step": 948
     },
     {
       "epoch": 0.916023166023166,
-      "grad_norm": 5.101447582244873,
+      "grad_norm": 5.1625165939331055,
       "learning_rate": 1.6284829795338235e-05,
-      "loss": 1.5809,
+      "loss": 1.5721,
       "step": 949
     },
     {
       "epoch": 0.916988416988417,
-      "grad_norm": 4.230952262878418,
+      "grad_norm": 4.036871433258057,
       "learning_rate": 1.627671887081443e-05,
-      "loss": 1.5773,
+      "loss": 1.5696,
       "step": 950
     },
     {
       "epoch": 0.917953667953668,
-      "grad_norm": 4.483916759490967,
+      "grad_norm": 4.538830757141113,
       "learning_rate": 1.6268601126894986e-05,
-      "loss": 1.6321,
+      "loss": 1.6267,
       "step": 951
     },
     {
       "epoch": 0.918918918918919,
-      "grad_norm": 3.5368592739105225,
+      "grad_norm": 3.6375904083251953,
       "learning_rate": 1.6260476572399494e-05,
-      "loss": 1.6531,
+      "loss": 1.647,
       "step": 952
     },
     {
       "epoch": 0.9198841698841699,
-      "grad_norm": 4.395543098449707,
+      "grad_norm": 4.266676425933838,
       "learning_rate": 1.625234521615495e-05,
-      "loss": 1.5416,
+      "loss": 1.533,
       "step": 953
     },
     {
       "epoch": 0.9208494208494209,
-      "grad_norm": 3.7641806602478027,
+      "grad_norm": 3.898174524307251,
       "learning_rate": 1.6244207066995737e-05,
-      "loss": 1.6085,
+      "loss": 1.6063,
       "step": 954
     },
     {
       "epoch": 0.9218146718146718,
-      "grad_norm": 4.261475086212158,
+      "grad_norm": 4.291110038757324,
       "learning_rate": 1.6236062133763617e-05,
-      "loss": 1.5874,
+      "loss": 1.5805,
       "step": 955
     },
     {
       "epoch": 0.9227799227799228,
-      "grad_norm": 5.187191009521484,
+      "grad_norm": 5.286990165710449,
       "learning_rate": 1.6227910425307723e-05,
-      "loss": 1.5936,
+      "loss": 1.5885,
       "step": 956
     },
     {
       "epoch": 0.9237451737451737,
-      "grad_norm": 3.669710159301758,
+      "grad_norm": 3.479762315750122,
       "learning_rate": 1.6219751950484547e-05,
-      "loss": 1.549,
+      "loss": 1.5394,
       "step": 957
     },
     {
       "epoch": 0.9247104247104247,
-      "grad_norm": 5.148052215576172,
+      "grad_norm": 5.072516441345215,
       "learning_rate": 1.6211586718157935e-05,
-      "loss": 1.5487,
+      "loss": 1.5459,
       "step": 958
     },
     {
       "epoch": 0.9256756756756757,
-      "grad_norm": 3.8471758365631104,
+      "grad_norm": 3.8459577560424805,
       "learning_rate": 1.6203414737199083e-05,
-      "loss": 1.601,
+      "loss": 1.5983,
       "step": 959
     },
     {
       "epoch": 0.9266409266409267,
-      "grad_norm": 4.013639450073242,
+      "grad_norm": 3.908803939819336,
       "learning_rate": 1.61952360164865e-05,
-      "loss": 1.6314,
+      "loss": 1.6207,
       "step": 960
     },
     {
       "epoch": 0.9276061776061776,
-      "grad_norm": 3.7651102542877197,
+      "grad_norm": 3.7441136837005615,
       "learning_rate": 1.618705056490603e-05,
-      "loss": 1.5831,
+      "loss": 1.576,
       "step": 961
     },
     {
       "epoch": 0.9285714285714286,
-      "grad_norm": 4.085816860198975,
+      "grad_norm": 4.197796821594238,
       "learning_rate": 1.6178858391350835e-05,
-      "loss": 1.6087,
+      "loss": 1.6013,
       "step": 962
     },
     {
       "epoch": 0.9295366795366795,
-      "grad_norm": 3.8041088581085205,
+      "grad_norm": 3.815641403198242,
       "learning_rate": 1.6170659504721365e-05,
-      "loss": 1.6063,
+      "loss": 1.5926,
       "step": 963
     },
     {
       "epoch": 0.9305019305019305,
-      "grad_norm": 5.450389385223389,
+      "grad_norm": 5.319431781768799,
       "learning_rate": 1.6162453913925373e-05,
-      "loss": 1.5322,
+      "loss": 1.5118,
       "step": 964
     },
     {
       "epoch": 0.9314671814671814,
-      "grad_norm": 4.215041637420654,
+      "grad_norm": 4.213870525360107,
       "learning_rate": 1.6154241627877894e-05,
-      "loss": 1.6165,
+      "loss": 1.6052,
       "step": 965
     },
     {
       "epoch": 0.9324324324324325,
-      "grad_norm": 3.923663377761841,
+      "grad_norm": 3.9244916439056396,
       "learning_rate": 1.614602265550124e-05,
-      "loss": 1.5717,
+      "loss": 1.5591,
       "step": 966
     },
     {
       "epoch": 0.9333976833976834,
-      "grad_norm": 3.8256661891937256,
+      "grad_norm": 3.917710781097412,
       "learning_rate": 1.613779700572499e-05,
-      "loss": 1.6302,
+      "loss": 1.6285,
       "step": 967
     },
     {
       "epoch": 0.9343629343629344,
-      "grad_norm": 4.588598251342773,
+      "grad_norm": 4.8031463623046875,
       "learning_rate": 1.612956468748596e-05,
-      "loss": 1.6229,
+      "loss": 1.6168,
       "step": 968
     },
     {
       "epoch": 0.9353281853281853,
-      "grad_norm": 4.931879043579102,
+      "grad_norm": 5.04961633682251,
       "learning_rate": 1.6121325709728228e-05,
-      "loss": 1.5349,
+      "loss": 1.5339,
       "step": 969
     },
     {
       "epoch": 0.9362934362934363,
-      "grad_norm": 3.869826555252075,
+      "grad_norm": 4.07681941986084,
       "learning_rate": 1.611308008140311e-05,
-      "loss": 1.6005,
+      "loss": 1.5918,
       "step": 970
     },
     {
       "epoch": 0.9372586872586872,
-      "grad_norm": 4.035933971405029,
+      "grad_norm": 3.8145267963409424,
       "learning_rate": 1.610482781146913e-05,
-      "loss": 1.6232,
+      "loss": 1.621,
       "step": 971
     },
     {
       "epoch": 0.9382239382239382,
-      "grad_norm": 3.962907075881958,
+      "grad_norm": 4.189996719360352,
       "learning_rate": 1.6096568908892048e-05,
-      "loss": 1.6566,
+      "loss": 1.6558,
       "step": 972
     },
     {
       "epoch": 0.9391891891891891,
-      "grad_norm": 3.8459584712982178,
+      "grad_norm": 3.9039154052734375,
       "learning_rate": 1.608830338264481e-05,
-      "loss": 1.5392,
+      "loss": 1.5338,
       "step": 973
     },
     {
       "epoch": 0.9401544401544402,
-      "grad_norm": 3.968818187713623,
+      "grad_norm": 3.976487398147583,
       "learning_rate": 1.608003124170758e-05,
-      "loss": 1.5504,
+      "loss": 1.5454,
       "step": 974
     },
     {
       "epoch": 0.9411196911196911,
-      "grad_norm": 3.915367364883423,
+      "grad_norm": 3.9862024784088135,
       "learning_rate": 1.6071752495067684e-05,
-      "loss": 1.57,
+      "loss": 1.5674,
       "step": 975
     },
     {
       "epoch": 0.9420849420849421,
-      "grad_norm": 3.6851232051849365,
+      "grad_norm": 3.6938111782073975,
       "learning_rate": 1.6063467151719653e-05,
-      "loss": 1.569,
+      "loss": 1.5642,
       "step": 976
     },
     {
       "epoch": 0.943050193050193,
-      "grad_norm": 4.471630096435547,
+      "grad_norm": 4.355929851531982,
       "learning_rate": 1.6055175220665158e-05,
-      "loss": 1.6378,
+      "loss": 1.6381,
       "step": 977
     },
     {
       "epoch": 0.944015444015444,
-      "grad_norm": 4.199156761169434,
+      "grad_norm": 4.148268222808838,
       "learning_rate": 1.6046876710913043e-05,
-      "loss": 1.5974,
+      "loss": 1.5848,
       "step": 978
     },
     {
       "epoch": 0.944980694980695,
-      "grad_norm": 5.314173221588135,
+      "grad_norm": 5.121153354644775,
       "learning_rate": 1.60385716314793e-05,
-      "loss": 1.5,
+      "loss": 1.4897,
       "step": 979
     },
     {
       "epoch": 0.9459459459459459,
-      "grad_norm": 4.639838695526123,
+      "grad_norm": 4.730162620544434,
       "learning_rate": 1.603025999138705e-05,
-      "loss": 1.601,
+      "loss": 1.5813,
       "step": 980
     },
     {
       "epoch": 0.946911196911197,
-      "grad_norm": 5.4433159828186035,
+      "grad_norm": 5.042290210723877,
       "learning_rate": 1.602194179966655e-05,
-      "loss": 1.5841,
+      "loss": 1.5739,
       "step": 981
     },
     {
       "epoch": 0.9478764478764479,
-      "grad_norm": 5.501407623291016,
+      "grad_norm": 5.478323459625244,
       "learning_rate": 1.6013617065355167e-05,
-      "loss": 1.6584,
+      "loss": 1.6587,
       "step": 982
     },
     {
       "epoch": 0.9488416988416989,
-      "grad_norm": 3.7589755058288574,
+      "grad_norm": 3.7047712802886963,
       "learning_rate": 1.6005285797497383e-05,
-      "loss": 1.5948,
+      "loss": 1.5854,
       "step": 983
     },
     {
       "epoch": 0.9498069498069498,
-      "grad_norm": 3.967992067337036,
+      "grad_norm": 4.2115349769592285,
       "learning_rate": 1.5996948005144774e-05,
-      "loss": 1.6061,
+      "loss": 1.6018,
       "step": 984
     },
     {
       "epoch": 0.9507722007722008,
-      "grad_norm": 5.262543678283691,
+      "grad_norm": 5.228156089782715,
       "learning_rate": 1.598860369735601e-05,
-      "loss": 1.5848,
+      "loss": 1.5768,
       "step": 985
     },
     {
       "epoch": 0.9517374517374517,
-      "grad_norm": 3.765303611755371,
+      "grad_norm": 3.791060209274292,
       "learning_rate": 1.598025288319684e-05,
-      "loss": 1.5626,
+      "loss": 1.5674,
       "step": 986
     },
     {
       "epoch": 0.9527027027027027,
-      "grad_norm": 4.409090995788574,
+      "grad_norm": 4.410627365112305,
       "learning_rate": 1.5971895571740072e-05,
-      "loss": 1.6094,
+      "loss": 1.5957,
       "step": 987
     },
     {
       "epoch": 0.9536679536679536,
-      "grad_norm": 5.229574203491211,
+      "grad_norm": 5.231039524078369,
       "learning_rate": 1.596353177206558e-05,
-      "loss": 1.6579,
+      "loss": 1.6578,
       "step": 988
     },
     {
       "epoch": 0.9546332046332047,
-      "grad_norm": 4.5636887550354,
+      "grad_norm": 4.446424961090088,
       "learning_rate": 1.5955161493260288e-05,
-      "loss": 1.6182,
+      "loss": 1.6103,
       "step": 989
     },
     {
       "epoch": 0.9555984555984556,
-      "grad_norm": 3.9794015884399414,
+      "grad_norm": 3.95806884765625,
       "learning_rate": 1.5946784744418162e-05,
-      "loss": 1.612,
+      "loss": 1.6047,
       "step": 990
     },
     {
       "epoch": 0.9565637065637066,
-      "grad_norm": 3.775864601135254,
+      "grad_norm": 3.802037239074707,
       "learning_rate": 1.5938401534640185e-05,
-      "loss": 1.6424,
+      "loss": 1.637,
       "step": 991
     },
     {
       "epoch": 0.9575289575289575,
-      "grad_norm": 3.8312482833862305,
+      "grad_norm": 3.933542013168335,
       "learning_rate": 1.5930011873034377e-05,
-      "loss": 1.6274,
+      "loss": 1.6263,
       "step": 992
     },
     {
       "epoch": 0.9584942084942085,
-      "grad_norm": 3.593261480331421,
+      "grad_norm": 3.6130356788635254,
       "learning_rate": 1.592161576871575e-05,
-      "loss": 1.6042,
+      "loss": 1.5996,
       "step": 993
     },
     {
       "epoch": 0.9594594594594594,
-      "grad_norm": 3.6945395469665527,
+      "grad_norm": 3.6092867851257324,
       "learning_rate": 1.591321323080633e-05,
-      "loss": 1.5665,
+      "loss": 1.5648,
       "step": 994
     },
     {
       "epoch": 0.9604247104247104,
-      "grad_norm": 4.436593532562256,
+      "grad_norm": 4.57637357711792,
       "learning_rate": 1.5904804268435127e-05,
-      "loss": 1.6183,
+      "loss": 1.6157,
       "step": 995
     },
     {
       "epoch": 0.9613899613899614,
-      "grad_norm": 5.163026332855225,
+      "grad_norm": 4.85537052154541,
       "learning_rate": 1.589638889073813e-05,
-      "loss": 1.4779,
+      "loss": 1.4741,
       "step": 996
     },
     {
       "epoch": 0.9623552123552124,
-      "grad_norm": 3.6680943965911865,
+      "grad_norm": 3.605579137802124,
       "learning_rate": 1.5887967106858296e-05,
-      "loss": 1.5125,
+      "loss": 1.5084,
       "step": 997
     },
     {
       "epoch": 0.9633204633204633,
-      "grad_norm": 3.6280436515808105,
+      "grad_norm": 3.572723865509033,
       "learning_rate": 1.587953892594555e-05,
-      "loss": 1.5655,
+      "loss": 1.5552,
       "step": 998
     },
     {
       "epoch": 0.9642857142857143,
-      "grad_norm": 3.6882362365722656,
+      "grad_norm": 3.594184160232544,
       "learning_rate": 1.587110435715676e-05,
-      "loss": 1.6317,
+      "loss": 1.6242,
       "step": 999
     },
     {
       "epoch": 0.9652509652509652,
-      "grad_norm": 4.365917205810547,
+      "grad_norm": 4.265100955963135,
       "learning_rate": 1.5862663409655733e-05,
-      "loss": 1.5443,
+      "loss": 1.5316,
       "step": 1000
     },
     {
       "epoch": 0.9662162162162162,
-      "grad_norm": 3.6691746711730957,
+      "grad_norm": 3.563755512237549,
       "learning_rate": 1.585421609261322e-05,
-      "loss": 1.6013,
+      "loss": 1.5942,
       "step": 1001
     },
     {
       "epoch": 0.9671814671814671,
-      "grad_norm": 4.16331672668457,
+      "grad_norm": 3.8315529823303223,
       "learning_rate": 1.584576241520687e-05,
-      "loss": 1.4935,
+      "loss": 1.4837,
       "step": 1002
     },
     {
       "epoch": 0.9681467181467182,
-      "grad_norm": 4.23085355758667,
+      "grad_norm": 4.154335021972656,
       "learning_rate": 1.583730238662126e-05,
-      "loss": 1.6189,
+      "loss": 1.6204,
       "step": 1003
     },
     {
       "epoch": 0.9691119691119691,
-      "grad_norm": 4.181711673736572,
+      "grad_norm": 4.1938958168029785,
       "learning_rate": 1.582883601604786e-05,
-      "loss": 1.6259,
+      "loss": 1.615,
       "step": 1004
     },
     {
       "epoch": 0.9700772200772201,
-      "grad_norm": 4.299831867218018,
+      "grad_norm": 4.04390811920166,
       "learning_rate": 1.5820363312685032e-05,
-      "loss": 1.5885,
+      "loss": 1.5845,
       "step": 1005
     },
     {
       "epoch": 0.971042471042471,
-      "grad_norm": 3.6799392700195312,
+      "grad_norm": 3.6144485473632812,
       "learning_rate": 1.581188428573802e-05,
-      "loss": 1.608,
+      "loss": 1.6016,
       "step": 1006
     },
     {
       "epoch": 0.972007722007722,
-      "grad_norm": 4.73622989654541,
+      "grad_norm": 4.518500328063965,
       "learning_rate": 1.5803398944418934e-05,
-      "loss": 1.5512,
+      "loss": 1.53,
       "step": 1007
     },
     {
       "epoch": 0.972972972972973,
-      "grad_norm": 3.8058416843414307,
+      "grad_norm": 3.9674909114837646,
       "learning_rate": 1.579490729794675e-05,
-      "loss": 1.5421,
+      "loss": 1.5374,
       "step": 1008
     },
     {
       "epoch": 0.9739382239382239,
-      "grad_norm": 3.8185997009277344,
+      "grad_norm": 3.770488977432251,
       "learning_rate": 1.578640935554728e-05,
-      "loss": 1.7025,
+      "loss": 1.6959,
       "step": 1009
     },
     {
       "epoch": 0.974903474903475,
-      "grad_norm": 3.7489326000213623,
+      "grad_norm": 3.7466063499450684,
       "learning_rate": 1.5777905126453204e-05,
-      "loss": 1.5785,
+      "loss": 1.5769,
       "step": 1010
     },
     {
       "epoch": 0.9758687258687259,
-      "grad_norm": 4.161064147949219,
+      "grad_norm": 4.373013973236084,
       "learning_rate": 1.5769394619904002e-05,
-      "loss": 1.5742,
+      "loss": 1.5758,
       "step": 1011
     },
     {
       "epoch": 0.9768339768339769,
-      "grad_norm": 4.197792053222656,
+      "grad_norm": 4.130446910858154,
       "learning_rate": 1.5760877845145996e-05,
-      "loss": 1.5541,
+      "loss": 1.5448,
       "step": 1012
     },
     {
       "epoch": 0.9777992277992278,
-      "grad_norm": 5.174560546875,
+      "grad_norm": 5.252704620361328,
       "learning_rate": 1.57523548114323e-05,
-      "loss": 1.5785,
+      "loss": 1.5677,
       "step": 1013
     },
     {
       "epoch": 0.9787644787644788,
-      "grad_norm": 6.082184314727783,
+      "grad_norm": 6.149196147918701,
       "learning_rate": 1.5743825528022848e-05,
-      "loss": 1.616,
+      "loss": 1.6078,
       "step": 1014
     },
     {
       "epoch": 0.9797297297297297,
-      "grad_norm": 4.4992899894714355,
+      "grad_norm": 4.310429096221924,
       "learning_rate": 1.5735290004184354e-05,
-      "loss": 1.5737,
+      "loss": 1.579,
       "step": 1015
     },
     {
       "epoch": 0.9806949806949807,
-      "grad_norm": 8.02248764038086,
+      "grad_norm": 7.682847499847412,
       "learning_rate": 1.57267482491903e-05,
-      "loss": 1.6128,
+      "loss": 1.5894,
       "step": 1016
     },
     {
       "epoch": 0.9816602316602316,
-      "grad_norm": 3.4630367755889893,
+      "grad_norm": 3.4865612983703613,
       "learning_rate": 1.5718200272320965e-05,
-      "loss": 1.5705,
+      "loss": 1.5655,
       "step": 1017
     },
     {
       "epoch": 0.9826254826254827,
-      "grad_norm": 5.955785751342773,
+      "grad_norm": 5.759737014770508,
       "learning_rate": 1.570964608286336e-05,
-      "loss": 1.5853,
+      "loss": 1.5828,
       "step": 1018
     },
     {
       "epoch": 0.9835907335907336,
-      "grad_norm": 5.194217681884766,
+      "grad_norm": 5.4913177490234375,
       "learning_rate": 1.5701085690111268e-05,
-      "loss": 1.6157,
+      "loss": 1.6211,
       "step": 1019
     },
     {
       "epoch": 0.9845559845559846,
-      "grad_norm": 3.750546932220459,
+      "grad_norm": 3.7356483936309814,
       "learning_rate": 1.5692519103365194e-05,
-      "loss": 1.5675,
+      "loss": 1.5534,
       "step": 1020
     },
     {
       "epoch": 0.9855212355212355,
-      "grad_norm": 5.692684173583984,
+      "grad_norm": 5.908417224884033,
       "learning_rate": 1.5683946331932385e-05,
-      "loss": 1.6369,
+      "loss": 1.6378,
       "step": 1021
     },
     {
       "epoch": 0.9864864864864865,
-      "grad_norm": 4.506756782531738,
+      "grad_norm": 4.4945149421691895,
       "learning_rate": 1.56753673851268e-05,
-      "loss": 1.5216,
+      "loss": 1.5123,
       "step": 1022
     },
     {
       "epoch": 0.9874517374517374,
-      "grad_norm": 4.121953964233398,
+      "grad_norm": 4.160107135772705,
       "learning_rate": 1.5666782272269113e-05,
-      "loss": 1.6066,
+      "loss": 1.6016,
       "step": 1023
     },
     {
       "epoch": 0.9884169884169884,
-      "grad_norm": 4.941500186920166,
+      "grad_norm": 5.2991557121276855,
       "learning_rate": 1.5658191002686694e-05,
-      "loss": 1.6255,
+      "loss": 1.6104,
       "step": 1024
     },
     {
       "epoch": 0.9893822393822393,
-      "grad_norm": 3.950207471847534,
+      "grad_norm": 3.833500862121582,
       "learning_rate": 1.5649593585713604e-05,
-      "loss": 1.5147,
+      "loss": 1.5077,
       "step": 1025
     },
     {
       "epoch": 0.9903474903474904,
-      "grad_norm": 4.301034927368164,
+      "grad_norm": 4.387840747833252,
       "learning_rate": 1.5640990030690576e-05,
-      "loss": 1.5777,
+      "loss": 1.5679,
       "step": 1026
     },
     {
       "epoch": 0.9913127413127413,
-      "grad_norm": 4.856922149658203,
+      "grad_norm": 4.732850074768066,
       "learning_rate": 1.563238034696502e-05,
-      "loss": 1.536,
+      "loss": 1.5305,
       "step": 1027
     },
     {
       "epoch": 0.9922779922779923,
-      "grad_norm": 3.9627559185028076,
+      "grad_norm": 4.133497714996338,
       "learning_rate": 1.5623764543891e-05,
-      "loss": 1.5217,
+      "loss": 1.517,
       "step": 1028
     },
     {
       "epoch": 0.9932432432432432,
-      "grad_norm": 4.766280174255371,
+      "grad_norm": 4.6533050537109375,
       "learning_rate": 1.561514263082923e-05,
-      "loss": 1.558,
+      "loss": 1.5402,
       "step": 1029
     },
     {
       "epoch": 0.9942084942084942,
-      "grad_norm": 4.64534330368042,
+      "grad_norm": 4.38959264755249,
       "learning_rate": 1.560651461714707e-05,
-      "loss": 1.6071,
+      "loss": 1.6006,
       "step": 1030
     },
     {
       "epoch": 0.9951737451737451,
-      "grad_norm": 3.7165379524230957,
+      "grad_norm": 3.6689014434814453,
       "learning_rate": 1.5597880512218486e-05,
-      "loss": 1.5449,
+      "loss": 1.5453,
       "step": 1031
     },
     {
       "epoch": 0.9961389961389961,
-      "grad_norm": 4.039422512054443,
+      "grad_norm": 3.918680429458618,
       "learning_rate": 1.5589240325424088e-05,
-      "loss": 1.6113,
+      "loss": 1.618,
       "step": 1032
     },
     {
       "epoch": 0.997104247104247,
-      "grad_norm": 4.411838531494141,
+      "grad_norm": 4.30108118057251,
       "learning_rate": 1.5580594066151076e-05,
-      "loss": 1.5712,
+      "loss": 1.5623,
       "step": 1033
     },
     {
       "epoch": 0.9980694980694981,
-      "grad_norm": 3.8884987831115723,
+      "grad_norm": 3.7327120304107666,
       "learning_rate": 1.5571941743793256e-05,
-      "loss": 1.5246,
+      "loss": 1.5147,
       "step": 1034
     },
     {
       "epoch": 0.999034749034749,
-      "grad_norm": 3.771638870239258,
+      "grad_norm": 3.5460550785064697,
       "learning_rate": 1.556328336775102e-05,
-      "loss": 1.5772,
+      "loss": 1.569,
       "step": 1035
     },
     {
       "epoch": 1.0,
-      "grad_norm": 3.959677219390869,
+      "grad_norm": 3.960181713104248,
       "learning_rate": 1.5554618947431326e-05,
-      "loss": 1.5756,
+      "loss": 1.5627,
       "step": 1036
     },
     {
       "epoch": 1.000965250965251,
-      "grad_norm": 3.3843886852264404,
+      "grad_norm": 3.2626802921295166,
       "learning_rate": 1.5545948492247723e-05,
-      "loss": 1.2667,
+      "loss": 1.2561,
       "step": 1037
     },
     {
       "epoch": 1.001930501930502,
-      "grad_norm": 3.923689603805542,
+      "grad_norm": 3.771458387374878,
       "learning_rate": 1.5537272011620294e-05,
-      "loss": 1.2317,
+      "loss": 1.2285,
       "step": 1038
     },
     {
       "epoch": 1.002895752895753,
-      "grad_norm": 3.9356627464294434,
+      "grad_norm": 4.010953426361084,
       "learning_rate": 1.5528589514975675e-05,
-      "loss": 1.1738,
+      "loss": 1.1695,
       "step": 1039
     },
     {
       "epoch": 1.0038610038610039,
-      "grad_norm": 3.580815553665161,
+      "grad_norm": 3.549330949783325,
       "learning_rate": 1.5519901011747046e-05,
-      "loss": 1.2487,
+      "loss": 1.2426,
       "step": 1040
     },
     {
       "epoch": 1.0048262548262548,
-      "grad_norm": 4.160267353057861,
+      "grad_norm": 4.1646881103515625,
       "learning_rate": 1.55112065113741e-05,
-      "loss": 1.1911,
+      "loss": 1.1848,
       "step": 1041
     },
     {
       "epoch": 1.005791505791506,
-      "grad_norm": 4.202048301696777,
+      "grad_norm": 4.075333595275879,
       "learning_rate": 1.5502506023303058e-05,
-      "loss": 1.2276,
+      "loss": 1.2153,
       "step": 1042
     },
     {
       "epoch": 1.0067567567567568,
-      "grad_norm": 5.078566074371338,
+      "grad_norm": 5.0098652839660645,
       "learning_rate": 1.5493799556986634e-05,
-      "loss": 1.262,
+      "loss": 1.2558,
       "step": 1043
     },
     {
       "epoch": 1.0077220077220077,
-      "grad_norm": 4.21602201461792,
+      "grad_norm": 4.199959754943848,
       "learning_rate": 1.5485087121884056e-05,
-      "loss": 1.19,
+      "loss": 1.167,
       "step": 1044
     },
     {
       "epoch": 1.0086872586872586,
-      "grad_norm": 4.7379608154296875,
+      "grad_norm": 4.603140354156494,
       "learning_rate": 1.547636872746101e-05,
-      "loss": 1.1399,
+      "loss": 1.1323,
       "step": 1045
     },
     {
       "epoch": 1.0096525096525097,
-      "grad_norm": 4.72409200668335,
+      "grad_norm": 4.621537208557129,
       "learning_rate": 1.546764438318968e-05,
-      "loss": 1.1955,
+      "loss": 1.1866,
       "step": 1046
     },
     {
       "epoch": 1.0106177606177607,
-      "grad_norm": 3.7184345722198486,
+      "grad_norm": 3.802692174911499,
       "learning_rate": 1.5458914098548702e-05,
-      "loss": 1.1271,
+      "loss": 1.1045,
       "step": 1047
     },
     {
       "epoch": 1.0115830115830116,
-      "grad_norm": 5.027003765106201,
+      "grad_norm": 5.206601142883301,
       "learning_rate": 1.5450177883023174e-05,
-      "loss": 1.21,
+      "loss": 1.1994,
       "step": 1048
     },
     {
       "epoch": 1.0125482625482625,
-      "grad_norm": 5.185225009918213,
+      "grad_norm": 4.882866382598877,
       "learning_rate": 1.5441435746104628e-05,
-      "loss": 1.2206,
+      "loss": 1.2103,
       "step": 1049
     },
     {
       "epoch": 1.0135135135135136,
-      "grad_norm": 3.985189437866211,
+      "grad_norm": 4.108059406280518,
       "learning_rate": 1.5432687697291037e-05,
-      "loss": 1.1847,
+      "loss": 1.1713,
       "step": 1050
     },
     {
       "epoch": 1.0144787644787645,
-      "grad_norm": 6.368143081665039,
+      "grad_norm": 6.5675554275512695,
       "learning_rate": 1.5423933746086793e-05,
-      "loss": 1.138,
+      "loss": 1.1336,
       "step": 1051
     },
     {
       "epoch": 1.0154440154440154,
-      "grad_norm": 3.65655517578125,
+      "grad_norm": 3.694398880004883,
       "learning_rate": 1.5415173902002703e-05,
-      "loss": 1.1519,
+      "loss": 1.1366,
       "step": 1052
     },
     {
       "epoch": 1.0164092664092663,
-      "grad_norm": 4.059977054595947,
+      "grad_norm": 4.288461685180664,
       "learning_rate": 1.5406408174555978e-05,
-      "loss": 1.2186,
+      "loss": 1.2132,
       "step": 1053
     },
     {
       "epoch": 1.0173745173745175,
-      "grad_norm": 5.151540279388428,
+      "grad_norm": 5.404181003570557,
       "learning_rate": 1.5397636573270213e-05,
-      "loss": 1.1636,
+      "loss": 1.1583,
       "step": 1054
     },
     {
       "epoch": 1.0183397683397684,
-      "grad_norm": 4.020742893218994,
+      "grad_norm": 4.181848049163818,
       "learning_rate": 1.5388859107675396e-05,
-      "loss": 1.1955,
+      "loss": 1.1798,
       "step": 1055
     },
     {
       "epoch": 1.0193050193050193,
-      "grad_norm": 4.228591442108154,
+      "grad_norm": 4.34151554107666,
       "learning_rate": 1.5380075787307875e-05,
-      "loss": 1.1862,
+      "loss": 1.1824,
       "step": 1056
     },
     {
       "epoch": 1.0202702702702702,
-      "grad_norm": 5.486896991729736,
+      "grad_norm": 5.72661018371582,
       "learning_rate": 1.5371286621710374e-05,
-      "loss": 1.1797,
+      "loss": 1.1711,
       "step": 1057
     },
     {
       "epoch": 1.0212355212355213,
-      "grad_norm": 4.261332035064697,
+      "grad_norm": 4.3028483390808105,
       "learning_rate": 1.536249162043194e-05,
-      "loss": 1.2069,
+      "loss": 1.1907,
       "step": 1058
     },
     {
       "epoch": 1.0222007722007722,
-      "grad_norm": 5.16762113571167,
+      "grad_norm": 5.094120502471924,
       "learning_rate": 1.5353690793028e-05,
-      "loss": 1.2037,
+      "loss": 1.1889,
       "step": 1059
     },
     {
       "epoch": 1.0231660231660231,
-      "grad_norm": 5.293707370758057,
+      "grad_norm": 5.464247226715088,
       "learning_rate": 1.534488414906028e-05,
-      "loss": 1.1758,
+      "loss": 1.1752,
       "step": 1060
     },
     {
       "epoch": 1.024131274131274,
-      "grad_norm": 4.2502336502075195,
+      "grad_norm": 3.999506950378418,
       "learning_rate": 1.533607169809683e-05,
-      "loss": 1.1617,
+      "loss": 1.1542,
       "step": 1061
     },
     {
       "epoch": 1.0250965250965252,
-      "grad_norm": 4.9595441818237305,
+      "grad_norm": 4.701669216156006,
       "learning_rate": 1.532725344971202e-05,
-      "loss": 1.1538,
+      "loss": 1.1453,
       "step": 1062
     },
     {
       "epoch": 1.026061776061776,
-      "grad_norm": 4.1361541748046875,
+      "grad_norm": 4.289036750793457,
       "learning_rate": 1.5318429413486507e-05,
-      "loss": 1.0488,
+      "loss": 1.0387,
       "step": 1063
     },
     {
       "epoch": 1.027027027027027,
-      "grad_norm": 3.9857394695281982,
+      "grad_norm": 3.8738110065460205,
       "learning_rate": 1.530959959900725e-05,
-      "loss": 1.103,
+      "loss": 1.097,
       "step": 1064
     },
     {
       "epoch": 1.0279922779922779,
-      "grad_norm": 3.9534642696380615,
+      "grad_norm": 4.054632663726807,
       "learning_rate": 1.5300764015867467e-05,
-      "loss": 1.1534,
+      "loss": 1.1391,
       "step": 1065
     },
     {
       "epoch": 1.028957528957529,
-      "grad_norm": 5.401212692260742,
+      "grad_norm": 5.859082221984863,
       "learning_rate": 1.5291922673666664e-05,
-      "loss": 1.2082,
+      "loss": 1.2035,
       "step": 1066
     },
     {
       "epoch": 1.02992277992278,
-      "grad_norm": 3.9076457023620605,
+      "grad_norm": 3.886763334274292,
       "learning_rate": 1.5283075582010586e-05,
-      "loss": 1.1735,
+      "loss": 1.1681,
       "step": 1067
     },
     {
       "epoch": 1.0308880308880308,
-      "grad_norm": 4.63938045501709,
+      "grad_norm": 4.557432651519775,
       "learning_rate": 1.527422275051124e-05,
-      "loss": 1.1487,
+      "loss": 1.1441,
       "step": 1068
     },
     {
       "epoch": 1.031853281853282,
-      "grad_norm": 4.2600836753845215,
+      "grad_norm": 4.5361738204956055,
       "learning_rate": 1.5265364188786852e-05,
-      "loss": 1.205,
+      "loss": 1.1995,
       "step": 1069
     },
     {
       "epoch": 1.0328185328185329,
-      "grad_norm": 4.081650733947754,
+      "grad_norm": 3.997636079788208,
       "learning_rate": 1.5256499906461893e-05,
-      "loss": 1.1628,
+      "loss": 1.1502,
       "step": 1070
     },
     {
       "epoch": 1.0337837837837838,
-      "grad_norm": 4.65988826751709,
+      "grad_norm": 4.658125877380371,
       "learning_rate": 1.5247629913167034e-05,
-      "loss": 1.1916,
+      "loss": 1.1782,
       "step": 1071
     },
     {
       "epoch": 1.0347490347490347,
-      "grad_norm": 4.270849227905273,
+      "grad_norm": 4.281274795532227,
       "learning_rate": 1.5238754218539155e-05,
-      "loss": 1.2091,
+      "loss": 1.1977,
       "step": 1072
     },
     {
       "epoch": 1.0357142857142858,
-      "grad_norm": 4.084417343139648,
+      "grad_norm": 3.9605226516723633,
       "learning_rate": 1.5229872832221336e-05,
-      "loss": 1.1886,
+      "loss": 1.1761,
       "step": 1073
     },
     {
       "epoch": 1.0366795366795367,
-      "grad_norm": 4.581521034240723,
+      "grad_norm": 4.473110198974609,
       "learning_rate": 1.522098576386283e-05,
-      "loss": 1.1868,
+      "loss": 1.1766,
       "step": 1074
     },
     {
       "epoch": 1.0376447876447876,
-      "grad_norm": 5.107059955596924,
+      "grad_norm": 4.724252700805664,
       "learning_rate": 1.5212093023119075e-05,
-      "loss": 1.1687,
+      "loss": 1.1585,
       "step": 1075
     },
     {
       "epoch": 1.0386100386100385,
-      "grad_norm": 4.153777599334717,
+      "grad_norm": 4.2992472648620605,
       "learning_rate": 1.5203194619651664e-05,
-      "loss": 1.0946,
+      "loss": 1.0876,
       "step": 1076
     },
     {
       "epoch": 1.0395752895752897,
-      "grad_norm": 4.025907516479492,
+      "grad_norm": 4.029987812042236,
       "learning_rate": 1.5194290563128346e-05,
-      "loss": 1.1327,
+      "loss": 1.1254,
       "step": 1077
     },
     {
       "epoch": 1.0405405405405406,
-      "grad_norm": 4.304230213165283,
+      "grad_norm": 4.306401252746582,
       "learning_rate": 1.5185380863223007e-05,
-      "loss": 1.2248,
+      "loss": 1.2101,
       "step": 1078
     },
     {
       "epoch": 1.0415057915057915,
-      "grad_norm": 4.770325183868408,
+      "grad_norm": 4.9542412757873535,
       "learning_rate": 1.5176465529615672e-05,
-      "loss": 1.2188,
+      "loss": 1.2043,
       "step": 1079
     },
     {
       "epoch": 1.0424710424710424,
-      "grad_norm": 4.624781608581543,
+      "grad_norm": 4.995790004730225,
       "learning_rate": 1.5167544571992478e-05,
-      "loss": 1.2264,
+      "loss": 1.2086,
       "step": 1080
     },
     {
       "epoch": 1.0434362934362935,
-      "grad_norm": 4.707468032836914,
+      "grad_norm": 4.516042232513428,
       "learning_rate": 1.5158618000045679e-05,
-      "loss": 1.2513,
+      "loss": 1.2393,
       "step": 1081
     },
     {
       "epoch": 1.0444015444015444,
-      "grad_norm": 4.72183895111084,
+      "grad_norm": 4.76840877532959,
       "learning_rate": 1.5149685823473633e-05,
-      "loss": 1.2276,
+      "loss": 1.22,
       "step": 1082
     },
     {
       "epoch": 1.0453667953667953,
-      "grad_norm": 4.330811023712158,
+      "grad_norm": 4.727954387664795,
       "learning_rate": 1.5140748051980767e-05,
-      "loss": 1.2213,
+      "loss": 1.2105,
       "step": 1083
     },
     {
       "epoch": 1.0463320463320462,
-      "grad_norm": 4.455382823944092,
+      "grad_norm": 3.8478519916534424,
       "learning_rate": 1.5131804695277612e-05,
-      "loss": 1.1809,
+      "loss": 1.1717,
       "step": 1084
     },
     {
       "epoch": 1.0472972972972974,
-      "grad_norm": 3.680363178253174,
+      "grad_norm": 3.8120574951171875,
       "learning_rate": 1.5122855763080748e-05,
-      "loss": 1.2191,
+      "loss": 1.2034,
       "step": 1085
     },
     {
       "epoch": 1.0482625482625483,
-      "grad_norm": 3.820376396179199,
+      "grad_norm": 4.666320323944092,
       "learning_rate": 1.5113901265112829e-05,
-      "loss": 1.1232,
+      "loss": 1.1112,
       "step": 1086
     },
     {
       "epoch": 1.0492277992277992,
-      "grad_norm": 4.394763469696045,
+      "grad_norm": 4.422824382781982,
       "learning_rate": 1.5104941211102536e-05,
-      "loss": 1.26,
+      "loss": 1.2447,
       "step": 1087
     },
     {
       "epoch": 1.05019305019305,
-      "grad_norm": 4.531584739685059,
+      "grad_norm": 5.103439807891846,
       "learning_rate": 1.5095975610784603e-05,
-      "loss": 1.2006,
+      "loss": 1.1819,
       "step": 1088
     },
     {
       "epoch": 1.0511583011583012,
-      "grad_norm": 3.9240989685058594,
+      "grad_norm": 4.5219340324401855,
       "learning_rate": 1.5087004473899783e-05,
-      "loss": 1.2255,
+      "loss": 1.2201,
       "step": 1089
     },
     {
       "epoch": 1.0521235521235521,
-      "grad_norm": 4.53143835067749,
+      "grad_norm": 4.326501846313477,
       "learning_rate": 1.5078027810194847e-05,
-      "loss": 1.2436,
+      "loss": 1.2263,
       "step": 1090
     },
     {
       "epoch": 1.053088803088803,
-      "grad_norm": 4.420854091644287,
+      "grad_norm": 4.6630988121032715,
       "learning_rate": 1.5069045629422569e-05,
-      "loss": 1.1336,
+      "loss": 1.1294,
       "step": 1091
     },
     {
       "epoch": 1.054054054054054,
-      "grad_norm": 4.524373531341553,
+      "grad_norm": 4.566372871398926,
       "learning_rate": 1.5060057941341717e-05,
-      "loss": 1.2893,
+      "loss": 1.2735,
       "step": 1092
     },
     {
       "epoch": 1.055019305019305,
-      "grad_norm": 4.0564188957214355,
+      "grad_norm": 4.139376640319824,
       "learning_rate": 1.5051064755717041e-05,
-      "loss": 1.0964,
+      "loss": 1.0875,
       "step": 1093
     },
     {
       "epoch": 1.055984555984556,
-      "grad_norm": 4.545648574829102,
+      "grad_norm": 4.8554558753967285,
       "learning_rate": 1.5042066082319266e-05,
-      "loss": 1.2065,
+      "loss": 1.1926,
       "step": 1094
     },
     {
       "epoch": 1.056949806949807,
-      "grad_norm": 3.6719343662261963,
+      "grad_norm": 4.095548629760742,
       "learning_rate": 1.5033061930925081e-05,
-      "loss": 1.1443,
+      "loss": 1.1382,
       "step": 1095
     },
     {
       "epoch": 1.057915057915058,
-      "grad_norm": 5.457347393035889,
+      "grad_norm": 4.951704025268555,
       "learning_rate": 1.5024052311317122e-05,
-      "loss": 1.1244,
+      "loss": 1.1059,
       "step": 1096
     },
     {
       "epoch": 1.058880308880309,
-      "grad_norm": 4.415368556976318,
+      "grad_norm": 4.424997806549072,
       "learning_rate": 1.5015037233283969e-05,
-      "loss": 1.2113,
+      "loss": 1.1939,
       "step": 1097
     },
     {
       "epoch": 1.0598455598455598,
-      "grad_norm": 4.450148582458496,
+      "grad_norm": 4.328413486480713,
       "learning_rate": 1.5006016706620132e-05,
-      "loss": 1.2133,
+      "loss": 1.1929,
       "step": 1098
     },
     {
       "epoch": 1.0608108108108107,
-      "grad_norm": 4.177610397338867,
+      "grad_norm": 4.0156168937683105,
       "learning_rate": 1.499699074112604e-05,
-      "loss": 1.2148,
+      "loss": 1.1978,
       "step": 1099
     },
     {
       "epoch": 1.0617760617760619,
-      "grad_norm": 4.037389755249023,
+      "grad_norm": 3.960352659225464,
       "learning_rate": 1.4987959346608029e-05,
-      "loss": 1.1619,
+      "loss": 1.151,
       "step": 1100
     },
     {
       "epoch": 1.0627413127413128,
-      "grad_norm": 4.289162635803223,
+      "grad_norm": 4.663453578948975,
       "learning_rate": 1.497892253287834e-05,
-      "loss": 1.1953,
+      "loss": 1.1874,
       "step": 1101
     },
     {
       "epoch": 1.0637065637065637,
-      "grad_norm": 4.018803119659424,
+      "grad_norm": 3.9500622749328613,
       "learning_rate": 1.4969880309755098e-05,
-      "loss": 1.1564,
+      "loss": 1.1437,
       "step": 1102
     },
     {
       "epoch": 1.0646718146718146,
-      "grad_norm": 4.385936737060547,
+      "grad_norm": 4.493472576141357,
       "learning_rate": 1.4960832687062303e-05,
-      "loss": 1.1924,
+      "loss": 1.1794,
       "step": 1103
     },
     {
       "epoch": 1.0656370656370657,
-      "grad_norm": 3.9476888179779053,
+      "grad_norm": 4.155457973480225,
       "learning_rate": 1.495177967462982e-05,
-      "loss": 1.1805,
+      "loss": 1.1688,
       "step": 1104
     },
     {
       "epoch": 1.0666023166023166,
-      "grad_norm": 3.9989492893218994,
+      "grad_norm": 4.080810546875,
       "learning_rate": 1.4942721282293378e-05,
-      "loss": 1.1573,
+      "loss": 1.1417,
       "step": 1105
     },
     {
       "epoch": 1.0675675675675675,
-      "grad_norm": 3.8703742027282715,
+      "grad_norm": 4.098885536193848,
       "learning_rate": 1.4933657519894542e-05,
-      "loss": 1.15,
+      "loss": 1.1471,
       "step": 1106
     },
     {
       "epoch": 1.0685328185328185,
-      "grad_norm": 4.762238025665283,
+      "grad_norm": 4.593235492706299,
       "learning_rate": 1.4924588397280717e-05,
-      "loss": 1.2102,
+      "loss": 1.2065,
       "step": 1107
     },
     {
       "epoch": 1.0694980694980696,
-      "grad_norm": 4.119187355041504,
+      "grad_norm": 3.9634385108947754,
       "learning_rate": 1.4915513924305132e-05,
-      "loss": 1.1715,
+      "loss": 1.1443,
       "step": 1108
     },
     {
       "epoch": 1.0704633204633205,
-      "grad_norm": 4.430844306945801,
+      "grad_norm": 4.316656112670898,
       "learning_rate": 1.4906434110826822e-05,
-      "loss": 1.1914,
+      "loss": 1.1868,
       "step": 1109
     },
     {
       "epoch": 1.0714285714285714,
-      "grad_norm": 4.662785053253174,
+      "grad_norm": 4.0473713874816895,
       "learning_rate": 1.4897348966710632e-05,
-      "loss": 1.237,
+      "loss": 1.2411,
       "step": 1110
     },
     {
       "epoch": 1.0723938223938223,
-      "grad_norm": 4.0754547119140625,
+      "grad_norm": 4.139842510223389,
       "learning_rate": 1.4888258501827197e-05,
-      "loss": 1.1837,
+      "loss": 1.1757,
       "step": 1111
     },
     {
       "epoch": 1.0733590733590734,
-      "grad_norm": 4.096596717834473,
+      "grad_norm": 3.9712462425231934,
       "learning_rate": 1.4879162726052926e-05,
-      "loss": 1.1665,
+      "loss": 1.1477,
       "step": 1112
     },
     {
       "epoch": 1.0743243243243243,
-      "grad_norm": 4.2532782554626465,
+      "grad_norm": 4.107168197631836,
       "learning_rate": 1.4870061649270014e-05,
-      "loss": 1.1448,
+      "loss": 1.1265,
       "step": 1113
     },
     {
       "epoch": 1.0752895752895753,
-      "grad_norm": 4.306217193603516,
+      "grad_norm": 4.1352949142456055,
       "learning_rate": 1.4860955281366392e-05,
-      "loss": 1.1413,
+      "loss": 1.1379,
       "step": 1114
     },
     {
       "epoch": 1.0762548262548262,
-      "grad_norm": 4.049156665802002,
+      "grad_norm": 3.948075532913208,
       "learning_rate": 1.4851843632235761e-05,
-      "loss": 1.1803,
+      "loss": 1.1742,
       "step": 1115
     },
     {
       "epoch": 1.0772200772200773,
-      "grad_norm": 4.349161624908447,
+      "grad_norm": 4.552215099334717,
       "learning_rate": 1.4842726711777546e-05,
-      "loss": 1.202,
+      "loss": 1.1799,
       "step": 1116
     },
     {
       "epoch": 1.0781853281853282,
-      "grad_norm": 4.6802778244018555,
+      "grad_norm": 4.6967082023620605,
       "learning_rate": 1.483360452989691e-05,
-      "loss": 1.2489,
+      "loss": 1.2296,
       "step": 1117
     },
     {
       "epoch": 1.079150579150579,
-      "grad_norm": 4.26700496673584,
+      "grad_norm": 4.111815452575684,
       "learning_rate": 1.482447709650472e-05,
-      "loss": 1.1868,
+      "loss": 1.1778,
       "step": 1118
     },
     {
       "epoch": 1.08011583011583,
-      "grad_norm": 4.153388500213623,
+      "grad_norm": 4.523697376251221,
       "learning_rate": 1.4815344421517561e-05,
-      "loss": 1.1798,
+      "loss": 1.1631,
       "step": 1119
     },
     {
       "epoch": 1.0810810810810811,
-      "grad_norm": 4.498805046081543,
+      "grad_norm": 4.757104873657227,
       "learning_rate": 1.4806206514857703e-05,
-      "loss": 1.2535,
+      "loss": 1.2415,
       "step": 1120
     },
     {
       "epoch": 1.082046332046332,
-      "grad_norm": 4.3766655921936035,
+      "grad_norm": 4.252121448516846,
       "learning_rate": 1.4797063386453107e-05,
-      "loss": 1.2405,
+      "loss": 1.2303,
       "step": 1121
     },
     {
       "epoch": 1.083011583011583,
-      "grad_norm": 4.181150913238525,
+      "grad_norm": 4.185622692108154,
       "learning_rate": 1.4787915046237401e-05,
-      "loss": 1.2012,
+      "loss": 1.1958,
       "step": 1122
     },
     {
       "epoch": 1.083976833976834,
-      "grad_norm": 4.513302326202393,
+      "grad_norm": 4.487340927124023,
       "learning_rate": 1.4778761504149882e-05,
-      "loss": 1.2228,
+      "loss": 1.2162,
       "step": 1123
     },
     {
       "epoch": 1.084942084942085,
-      "grad_norm": 3.881986618041992,
+      "grad_norm": 3.950575113296509,
       "learning_rate": 1.4769602770135492e-05,
-      "loss": 1.2317,
+      "loss": 1.2188,
       "step": 1124
     },
     {
       "epoch": 1.085907335907336,
-      "grad_norm": 4.63210916519165,
+      "grad_norm": 4.660755157470703,
       "learning_rate": 1.4760438854144816e-05,
-      "loss": 1.2549,
+      "loss": 1.2447,
       "step": 1125
     },
     {
       "epoch": 1.0868725868725868,
-      "grad_norm": 4.1304473876953125,
+      "grad_norm": 4.179408550262451,
       "learning_rate": 1.4751269766134075e-05,
-      "loss": 1.196,
+      "loss": 1.1925,
       "step": 1126
     },
     {
       "epoch": 1.087837837837838,
-      "grad_norm": 3.830836534500122,
+      "grad_norm": 3.8522393703460693,
       "learning_rate": 1.47420955160651e-05,
-      "loss": 1.2011,
+      "loss": 1.1928,
       "step": 1127
     },
     {
       "epoch": 1.0888030888030888,
-      "grad_norm": 4.23577880859375,
+      "grad_norm": 4.039646625518799,
       "learning_rate": 1.4732916113905336e-05,
-      "loss": 1.1752,
+      "loss": 1.1582,
       "step": 1128
     },
     {
       "epoch": 1.0897683397683398,
-      "grad_norm": 4.022891521453857,
+      "grad_norm": 3.991276741027832,
       "learning_rate": 1.4723731569627827e-05,
-      "loss": 1.1643,
+      "loss": 1.1466,
       "step": 1129
     },
     {
       "epoch": 1.0907335907335907,
-      "grad_norm": 4.168032169342041,
+      "grad_norm": 4.065153121948242,
       "learning_rate": 1.4714541893211195e-05,
-      "loss": 1.1538,
+      "loss": 1.1464,
       "step": 1130
     },
     {
       "epoch": 1.0916988416988418,
-      "grad_norm": 4.2090044021606445,
+      "grad_norm": 4.263173580169678,
       "learning_rate": 1.4705347094639643e-05,
-      "loss": 1.2151,
+      "loss": 1.2057,
       "step": 1131
     },
     {
       "epoch": 1.0926640926640927,
-      "grad_norm": 4.411066055297852,
+      "grad_norm": 4.323197841644287,
       "learning_rate": 1.4696147183902949e-05,
-      "loss": 1.1558,
+      "loss": 1.1519,
       "step": 1132
     },
     {
       "epoch": 1.0936293436293436,
-      "grad_norm": 4.503549575805664,
+      "grad_norm": 4.411391258239746,
       "learning_rate": 1.4686942170996426e-05,
-      "loss": 1.1777,
+      "loss": 1.1652,
       "step": 1133
     },
     {
       "epoch": 1.0945945945945945,
-      "grad_norm": 4.157337188720703,
+      "grad_norm": 4.219332695007324,
       "learning_rate": 1.4677732065920945e-05,
-      "loss": 1.213,
+      "loss": 1.1974,
       "step": 1134
     },
     {
       "epoch": 1.0955598455598456,
-      "grad_norm": 4.277044773101807,
+      "grad_norm": 4.295759677886963,
       "learning_rate": 1.4668516878682902e-05,
-      "loss": 1.2097,
+      "loss": 1.1944,
       "step": 1135
     },
     {
       "epoch": 1.0965250965250966,
-      "grad_norm": 4.421609878540039,
+      "grad_norm": 4.382910251617432,
       "learning_rate": 1.465929661929422e-05,
-      "loss": 1.19,
+      "loss": 1.1581,
       "step": 1136
     },
     {
       "epoch": 1.0974903474903475,
-      "grad_norm": 4.0304436683654785,
+      "grad_norm": 4.184013366699219,
       "learning_rate": 1.4650071297772326e-05,
-      "loss": 1.2228,
+      "loss": 1.2092,
       "step": 1137
     },
     {
       "epoch": 1.0984555984555984,
-      "grad_norm": 4.247618198394775,
+      "grad_norm": 3.945744752883911,
       "learning_rate": 1.4640840924140155e-05,
-      "loss": 1.1725,
+      "loss": 1.1602,
       "step": 1138
     },
     {
       "epoch": 1.0994208494208495,
-      "grad_norm": 4.0964813232421875,
+      "grad_norm": 4.078369617462158,
       "learning_rate": 1.4631605508426124e-05,
-      "loss": 1.1491,
+      "loss": 1.1298,
       "step": 1139
     },
     {
       "epoch": 1.1003861003861004,
-      "grad_norm": 4.424109935760498,
+      "grad_norm": 4.5597429275512695,
       "learning_rate": 1.4622365060664132e-05,
-      "loss": 1.1315,
+      "loss": 1.1119,
       "step": 1140
     },
     {
       "epoch": 1.1013513513513513,
-      "grad_norm": 3.909346342086792,
+      "grad_norm": 4.027429103851318,
       "learning_rate": 1.4613119590893536e-05,
-      "loss": 1.1786,
+      "loss": 1.164,
       "step": 1141
     },
     {
       "epoch": 1.1023166023166022,
-      "grad_norm": 4.109957218170166,
+      "grad_norm": 4.14832878112793,
       "learning_rate": 1.4603869109159172e-05,
-      "loss": 1.2014,
+      "loss": 1.1925,
       "step": 1142
     },
     {
       "epoch": 1.1032818532818534,
-      "grad_norm": 4.767273902893066,
+      "grad_norm": 4.6476054191589355,
       "learning_rate": 1.4594613625511288e-05,
-      "loss": 1.2175,
+      "loss": 1.2012,
       "step": 1143
     },
     {
       "epoch": 1.1042471042471043,
-      "grad_norm": 4.3955559730529785,
+      "grad_norm": 4.334449291229248,
       "learning_rate": 1.45853531500056e-05,
-      "loss": 1.1938,
+      "loss": 1.1688,
       "step": 1144
     },
     {
       "epoch": 1.1052123552123552,
-      "grad_norm": 4.092763900756836,
+      "grad_norm": 4.2845001220703125,
       "learning_rate": 1.4576087692703223e-05,
-      "loss": 1.1766,
+      "loss": 1.1722,
       "step": 1145
     },
     {
       "epoch": 1.106177606177606,
-      "grad_norm": 6.0021538734436035,
+      "grad_norm": 5.637049198150635,
       "learning_rate": 1.4566817263670698e-05,
-      "loss": 1.1753,
+      "loss": 1.166,
       "step": 1146
     },
     {
       "epoch": 1.1071428571428572,
-      "grad_norm": 4.584158420562744,
+      "grad_norm": 4.823710918426514,
       "learning_rate": 1.455754187297996e-05,
-      "loss": 1.2373,
+      "loss": 1.2109,
       "step": 1147
     },
     {
       "epoch": 1.1081081081081081,
-      "grad_norm": 4.043362140655518,
+      "grad_norm": 3.9147346019744873,
       "learning_rate": 1.4548261530708336e-05,
-      "loss": 1.1766,
+      "loss": 1.1559,
       "step": 1148
     },
     {
       "epoch": 1.109073359073359,
-      "grad_norm": 4.277347564697266,
+      "grad_norm": 4.267472743988037,
       "learning_rate": 1.4538976246938541e-05,
-      "loss": 1.1801,
+      "loss": 1.175,
       "step": 1149
     },
     {
       "epoch": 1.1100386100386102,
-      "grad_norm": 4.641721248626709,
+      "grad_norm": 4.329949855804443,
       "learning_rate": 1.4529686031758642e-05,
-      "loss": 1.2534,
+      "loss": 1.2416,
       "step": 1150
     },
     {
       "epoch": 1.111003861003861,
-      "grad_norm": 4.423067092895508,
+      "grad_norm": 4.4839324951171875,
       "learning_rate": 1.4520390895262089e-05,
-      "loss": 1.1882,
+      "loss": 1.1613,
       "step": 1151
     },
     {
       "epoch": 1.111969111969112,
-      "grad_norm": 4.20358943939209,
+      "grad_norm": 4.150601387023926,
       "learning_rate": 1.4511090847547643e-05,
-      "loss": 1.1374,
+      "loss": 1.1297,
       "step": 1152
     },
     {
       "epoch": 1.1129343629343629,
-      "grad_norm": 4.543899059295654,
+      "grad_norm": 4.371912002563477,
       "learning_rate": 1.4501785898719443e-05,
-      "loss": 1.1742,
+      "loss": 1.1544,
       "step": 1153
     },
     {
       "epoch": 1.1138996138996138,
-      "grad_norm": 4.224864482879639,
+      "grad_norm": 4.355582237243652,
       "learning_rate": 1.4492476058886915e-05,
-      "loss": 1.1377,
+      "loss": 1.1262,
       "step": 1154
     },
     {
       "epoch": 1.114864864864865,
-      "grad_norm": 4.120309352874756,
+      "grad_norm": 3.978856325149536,
       "learning_rate": 1.4483161338164827e-05,
-      "loss": 1.1776,
+      "loss": 1.1584,
       "step": 1155
     },
     {
       "epoch": 1.1158301158301158,
-      "grad_norm": 4.584604263305664,
+      "grad_norm": 4.848778247833252,
       "learning_rate": 1.4473841746673225e-05,
-      "loss": 1.1792,
+      "loss": 1.1582,
       "step": 1156
     },
     {
       "epoch": 1.1167953667953667,
-      "grad_norm": 4.499202728271484,
+      "grad_norm": 4.5610432624816895,
       "learning_rate": 1.4464517294537472e-05,
-      "loss": 1.1798,
+      "loss": 1.1687,
       "step": 1157
     },
     {
       "epoch": 1.1177606177606179,
-      "grad_norm": 4.687706470489502,
+      "grad_norm": 4.56625509262085,
       "learning_rate": 1.4455187991888195e-05,
-      "loss": 1.1704,
+      "loss": 1.1499,
       "step": 1158
     },
     {
       "epoch": 1.1187258687258688,
-      "grad_norm": 5.001594543457031,
+      "grad_norm": 5.1750102043151855,
       "learning_rate": 1.4445853848861295e-05,
-      "loss": 1.2071,
+      "loss": 1.1982,
       "step": 1159
     },
     {
       "epoch": 1.1196911196911197,
-      "grad_norm": 4.183671474456787,
+      "grad_norm": 4.172297954559326,
       "learning_rate": 1.4436514875597935e-05,
-      "loss": 1.1657,
+      "loss": 1.1409,
       "step": 1160
     },
     {
       "epoch": 1.1206563706563706,
-      "grad_norm": 4.838707447052002,
+      "grad_norm": 4.369122505187988,
       "learning_rate": 1.4427171082244523e-05,
-      "loss": 1.151,
+      "loss": 1.1287,
       "step": 1161
     },
     {
       "epoch": 1.1216216216216217,
-      "grad_norm": 5.016688823699951,
+      "grad_norm": 5.256588935852051,
       "learning_rate": 1.4417822478952701e-05,
-      "loss": 1.1944,
+      "loss": 1.1782,
       "step": 1162
     },
     {
       "epoch": 1.1225868725868726,
-      "grad_norm": 4.297835826873779,
+      "grad_norm": 4.0972161293029785,
       "learning_rate": 1.4408469075879344e-05,
-      "loss": 1.1759,
+      "loss": 1.1598,
       "step": 1163
     },
     {
       "epoch": 1.1235521235521235,
-      "grad_norm": 4.285523414611816,
+      "grad_norm": 4.100122928619385,
       "learning_rate": 1.4399110883186538e-05,
-      "loss": 1.2142,
+      "loss": 1.1974,
       "step": 1164
     },
     {
       "epoch": 1.1245173745173744,
-      "grad_norm": 4.139347553253174,
+      "grad_norm": 4.142644882202148,
       "learning_rate": 1.4389747911041568e-05,
-      "loss": 1.216,
+      "loss": 1.189,
       "step": 1165
     },
     {
       "epoch": 1.1254826254826256,
-      "grad_norm": 4.459876537322998,
+      "grad_norm": 5.117334842681885,
       "learning_rate": 1.438038016961692e-05,
-      "loss": 1.1942,
+      "loss": 1.1675,
       "step": 1166
     },
     {
       "epoch": 1.1264478764478765,
-      "grad_norm": 4.0485124588012695,
+      "grad_norm": 4.032501697540283,
       "learning_rate": 1.4371007669090257e-05,
-      "loss": 1.2403,
+      "loss": 1.2156,
       "step": 1167
     },
     {
       "epoch": 1.1274131274131274,
-      "grad_norm": 4.578690528869629,
+      "grad_norm": 5.053812503814697,
       "learning_rate": 1.4361630419644417e-05,
-      "loss": 1.182,
+      "loss": 1.1686,
       "step": 1168
     },
     {
       "epoch": 1.1283783783783783,
-      "grad_norm": 4.086171627044678,
+      "grad_norm": 4.541929721832275,
       "learning_rate": 1.4352248431467386e-05,
-      "loss": 1.1811,
+      "loss": 1.1652,
       "step": 1169
     },
     {
       "epoch": 1.1293436293436294,
-      "grad_norm": 4.123401165008545,
+      "grad_norm": 4.125700950622559,
       "learning_rate": 1.4342861714752315e-05,
-      "loss": 1.223,
+      "loss": 1.2027,
       "step": 1170
     },
     {
       "epoch": 1.1303088803088803,
-      "grad_norm": 4.041556358337402,
+      "grad_norm": 4.251321315765381,
       "learning_rate": 1.433347027969748e-05,
-      "loss": 1.1837,
+      "loss": 1.1532,
       "step": 1171
     },
     {
       "epoch": 1.1312741312741312,
-      "grad_norm": 4.314547538757324,
+      "grad_norm": 4.759225845336914,
       "learning_rate": 1.4324074136506283e-05,
-      "loss": 1.226,
+      "loss": 1.211,
       "step": 1172
     },
     {
       "epoch": 1.1322393822393821,
-      "grad_norm": 3.764941692352295,
+      "grad_norm": 3.837709665298462,
       "learning_rate": 1.4314673295387251e-05,
-      "loss": 1.1753,
+      "loss": 1.1583,
       "step": 1173
     },
     {
       "epoch": 1.1332046332046333,
-      "grad_norm": 4.116872787475586,
+      "grad_norm": 4.6579270362854,
       "learning_rate": 1.4305267766554007e-05,
-      "loss": 1.203,
+      "loss": 1.1926,
       "step": 1174
     },
     {
       "epoch": 1.1341698841698842,
-      "grad_norm": 3.960906505584717,
+      "grad_norm": 3.997992992401123,
       "learning_rate": 1.4295857560225268e-05,
-      "loss": 1.1122,
+      "loss": 1.0946,
       "step": 1175
     },
     {
       "epoch": 1.135135135135135,
-      "grad_norm": 4.113468170166016,
+      "grad_norm": 4.128835678100586,
       "learning_rate": 1.4286442686624842e-05,
-      "loss": 1.2211,
+      "loss": 1.2034,
       "step": 1176
     },
     {
       "epoch": 1.1361003861003862,
-      "grad_norm": 4.512895584106445,
+      "grad_norm": 4.765344142913818,
       "learning_rate": 1.4277023155981587e-05,
-      "loss": 1.1981,
+      "loss": 1.1898,
       "step": 1177
     },
     {
       "epoch": 1.1370656370656371,
-      "grad_norm": 4.439307689666748,
+      "grad_norm": 4.9625244140625,
       "learning_rate": 1.4267598978529447e-05,
-      "loss": 1.231,
+      "loss": 1.2123,
       "step": 1178
     },
     {
       "epoch": 1.138030888030888,
-      "grad_norm": 4.276900291442871,
+      "grad_norm": 4.048965930938721,
       "learning_rate": 1.425817016450739e-05,
-      "loss": 1.1545,
+      "loss": 1.1419,
       "step": 1179
     },
     {
       "epoch": 1.138996138996139,
-      "grad_norm": 4.67651891708374,
+      "grad_norm": 4.758865833282471,
       "learning_rate": 1.4248736724159442e-05,
-      "loss": 1.2395,
+      "loss": 1.2231,
       "step": 1180
     },
     {
       "epoch": 1.1399613899613898,
-      "grad_norm": 4.468369007110596,
+      "grad_norm": 4.948935508728027,
       "learning_rate": 1.4239298667734638e-05,
-      "loss": 1.1793,
+      "loss": 1.1687,
       "step": 1181
     },
     {
       "epoch": 1.140926640926641,
-      "grad_norm": 4.818966388702393,
+      "grad_norm": 4.252557754516602,
       "learning_rate": 1.4229856005487044e-05,
-      "loss": 1.3307,
+      "loss": 1.3225,
       "step": 1182
     },
     {
       "epoch": 1.1418918918918919,
-      "grad_norm": 4.063793182373047,
+      "grad_norm": 4.158505439758301,
       "learning_rate": 1.4220408747675714e-05,
-      "loss": 1.214,
+      "loss": 1.1951,
       "step": 1183
     },
     {
       "epoch": 1.1428571428571428,
-      "grad_norm": 4.494610786437988,
+      "grad_norm": 4.033731937408447,
       "learning_rate": 1.421095690456471e-05,
-      "loss": 1.1376,
+      "loss": 1.1172,
       "step": 1184
     },
     {
       "epoch": 1.143822393822394,
-      "grad_norm": 4.65255069732666,
+      "grad_norm": 4.397652626037598,
       "learning_rate": 1.4201500486423067e-05,
-      "loss": 1.2266,
+      "loss": 1.2104,
       "step": 1185
     },
     {
       "epoch": 1.1447876447876448,
-      "grad_norm": 4.087616443634033,
+      "grad_norm": 4.114090442657471,
       "learning_rate": 1.4192039503524792e-05,
-      "loss": 1.1821,
+      "loss": 1.1579,
       "step": 1186
     },
     {
       "epoch": 1.1457528957528957,
-      "grad_norm": 4.112333297729492,
+      "grad_norm": 4.2401227951049805,
       "learning_rate": 1.4182573966148851e-05,
-      "loss": 1.2152,
+      "loss": 1.2056,
       "step": 1187
     },
     {
       "epoch": 1.1467181467181466,
-      "grad_norm": 3.8481669425964355,
+      "grad_norm": 3.8388912677764893,
       "learning_rate": 1.4173103884579163e-05,
-      "loss": 1.1877,
+      "loss": 1.1559,
       "step": 1188
     },
     {
       "epoch": 1.1476833976833978,
-      "grad_norm": 4.059640407562256,
+      "grad_norm": 4.154800891876221,
       "learning_rate": 1.4163629269104573e-05,
-      "loss": 1.2537,
+      "loss": 1.2426,
       "step": 1189
     },
     {
       "epoch": 1.1486486486486487,
-      "grad_norm": 4.0942230224609375,
+      "grad_norm": 4.539668560028076,
       "learning_rate": 1.4154150130018867e-05,
-      "loss": 1.238,
+      "loss": 1.2119,
       "step": 1190
     },
     {
       "epoch": 1.1496138996138996,
-      "grad_norm": 4.230532169342041,
+      "grad_norm": 4.19478702545166,
       "learning_rate": 1.414466647762073e-05,
-      "loss": 1.1883,
+      "loss": 1.1752,
       "step": 1191
     },
     {
       "epoch": 1.1505791505791505,
-      "grad_norm": 4.073517322540283,
+      "grad_norm": 4.2998833656311035,
       "learning_rate": 1.4135178322213765e-05,
-      "loss": 1.1941,
+      "loss": 1.1645,
       "step": 1192
     },
     {
       "epoch": 1.1515444015444016,
-      "grad_norm": 4.37415885925293,
+      "grad_norm": 5.111591339111328,
       "learning_rate": 1.412568567410646e-05,
-      "loss": 1.1625,
+      "loss": 1.149,
       "step": 1193
     },
     {
       "epoch": 1.1525096525096525,
-      "grad_norm": 4.088471412658691,
+      "grad_norm": 4.316877841949463,
       "learning_rate": 1.4116188543612182e-05,
-      "loss": 1.1843,
+      "loss": 1.1706,
       "step": 1194
     },
     {
       "epoch": 1.1534749034749034,
-      "grad_norm": 4.376766204833984,
+      "grad_norm": 4.4887375831604,
       "learning_rate": 1.4106686941049171e-05,
-      "loss": 1.1423,
+      "loss": 1.1224,
       "step": 1195
     },
     {
       "epoch": 1.1544401544401544,
-      "grad_norm": 4.660998821258545,
+      "grad_norm": 5.803928852081299,
       "learning_rate": 1.4097180876740525e-05,
-      "loss": 1.271,
+      "loss": 1.2469,
       "step": 1196
     },
     {
       "epoch": 1.1554054054054055,
-      "grad_norm": 4.198565483093262,
+      "grad_norm": 4.27832555770874,
       "learning_rate": 1.4087670361014194e-05,
-      "loss": 1.2037,
+      "loss": 1.1895,
       "step": 1197
     },
     {
       "epoch": 1.1563706563706564,
-      "grad_norm": 4.214802265167236,
+      "grad_norm": 4.386936187744141,
       "learning_rate": 1.4078155404202957e-05,
-      "loss": 1.2044,
+      "loss": 1.1874,
       "step": 1198
     },
     {
       "epoch": 1.1573359073359073,
-      "grad_norm": 4.1454362869262695,
+      "grad_norm": 4.474199295043945,
       "learning_rate": 1.4068636016644424e-05,
-      "loss": 1.1671,
+      "loss": 1.1429,
       "step": 1199
     },
     {
       "epoch": 1.1583011583011582,
-      "grad_norm": 4.182562351226807,
+      "grad_norm": 4.2635111808776855,
       "learning_rate": 1.4059112208681012e-05,
-      "loss": 1.1981,
+      "loss": 1.1816,
       "step": 1200
     },
     {
       "epoch": 1.1592664092664093,
-      "grad_norm": 4.076030731201172,
+      "grad_norm": 4.184596061706543,
       "learning_rate": 1.4049583990659947e-05,
-      "loss": 1.2083,
+      "loss": 1.19,
       "step": 1201
     },
     {
       "epoch": 1.1602316602316602,
-      "grad_norm": 4.5406270027160645,
+      "grad_norm": 5.020327568054199,
       "learning_rate": 1.4040051372933244e-05,
-      "loss": 1.2541,
+      "loss": 1.2206,
       "step": 1202
     },
     {
       "epoch": 1.1611969111969112,
-      "grad_norm": 4.326791286468506,
+      "grad_norm": 4.363395690917969,
       "learning_rate": 1.4030514365857696e-05,
-      "loss": 1.1372,
+      "loss": 1.1227,
       "step": 1203
     },
     {
       "epoch": 1.1621621621621623,
-      "grad_norm": 4.059686183929443,
+      "grad_norm": 4.144587516784668,
       "learning_rate": 1.402097297979487e-05,
-      "loss": 1.1778,
+      "loss": 1.1523,
       "step": 1204
     },
     {
       "epoch": 1.1631274131274132,
-      "grad_norm": 3.9642210006713867,
+      "grad_norm": 4.188613414764404,
       "learning_rate": 1.4011427225111091e-05,
-      "loss": 1.1207,
+      "loss": 1.1011,
       "step": 1205
     },
     {
       "epoch": 1.164092664092664,
-      "grad_norm": 4.094119071960449,
+      "grad_norm": 4.3513994216918945,
       "learning_rate": 1.4001877112177418e-05,
-      "loss": 1.1544,
+      "loss": 1.1425,
       "step": 1206
     },
     {
       "epoch": 1.165057915057915,
-      "grad_norm": 4.555078506469727,
+      "grad_norm": 4.5945820808410645,
       "learning_rate": 1.3992322651369663e-05,
-      "loss": 1.1621,
+      "loss": 1.1494,
       "step": 1207
     },
     {
       "epoch": 1.166023166023166,
-      "grad_norm": 4.552582740783691,
+      "grad_norm": 4.368198871612549,
       "learning_rate": 1.3982763853068345e-05,
-      "loss": 1.175,
+      "loss": 1.1598,
       "step": 1208
     },
     {
       "epoch": 1.166988416988417,
-      "grad_norm": 4.766264915466309,
+      "grad_norm": 4.568612575531006,
       "learning_rate": 1.3973200727658709e-05,
-      "loss": 1.1716,
+      "loss": 1.1413,
       "step": 1209
     },
     {
       "epoch": 1.167953667953668,
-      "grad_norm": 4.566944599151611,
+      "grad_norm": 4.223705291748047,
       "learning_rate": 1.3963633285530688e-05,
-      "loss": 1.2204,
+      "loss": 1.1999,
       "step": 1210
     },
     {
       "epoch": 1.1689189189189189,
-      "grad_norm": 4.437086582183838,
+      "grad_norm": 3.972006320953369,
       "learning_rate": 1.3954061537078927e-05,
-      "loss": 1.1336,
+      "loss": 1.1076,
       "step": 1211
     },
     {
       "epoch": 1.16988416988417,
-      "grad_norm": 4.496171474456787,
+      "grad_norm": 4.4944844245910645,
       "learning_rate": 1.3944485492702717e-05,
-      "loss": 1.2572,
+      "loss": 1.2287,
       "step": 1212
     },
     {
       "epoch": 1.170849420849421,
-      "grad_norm": 4.487252235412598,
+      "grad_norm": 4.583246231079102,
       "learning_rate": 1.393490516280605e-05,
-      "loss": 1.2293,
+      "loss": 1.2068,
       "step": 1213
     },
     {
       "epoch": 1.1718146718146718,
-      "grad_norm": 4.405369281768799,
+      "grad_norm": 4.425785064697266,
       "learning_rate": 1.3925320557797549e-05,
-      "loss": 1.2419,
+      "loss": 1.2198,
       "step": 1214
     },
     {
       "epoch": 1.1727799227799227,
-      "grad_norm": 4.267056941986084,
+      "grad_norm": 4.002625942230225,
       "learning_rate": 1.3915731688090492e-05,
-      "loss": 1.1946,
+      "loss": 1.1638,
       "step": 1215
     },
     {
       "epoch": 1.1737451737451738,
-      "grad_norm": 4.19378137588501,
+      "grad_norm": 3.9422833919525146,
       "learning_rate": 1.3906138564102794e-05,
-      "loss": 1.1232,
+      "loss": 1.0935,
       "step": 1216
     },
     {
       "epoch": 1.1747104247104247,
-      "grad_norm": 4.049899101257324,
+      "grad_norm": 4.325109481811523,
       "learning_rate": 1.3896541196256985e-05,
-      "loss": 1.2202,
+      "loss": 1.201,
       "step": 1217
     },
     {
       "epoch": 1.1756756756756757,
-      "grad_norm": 4.774388313293457,
+      "grad_norm": 4.450918674468994,
       "learning_rate": 1.3886939594980209e-05,
-      "loss": 1.2143,
+      "loss": 1.1858,
       "step": 1218
     },
     {
       "epoch": 1.1766409266409266,
-      "grad_norm": 3.9391682147979736,
+      "grad_norm": 3.939162015914917,
       "learning_rate": 1.3877333770704208e-05,
-      "loss": 1.259,
+      "loss": 1.234,
       "step": 1219
     },
     {
       "epoch": 1.1776061776061777,
-      "grad_norm": 4.408590793609619,
+      "grad_norm": 4.265638828277588,
       "learning_rate": 1.3867723733865314e-05,
-      "loss": 1.2118,
+      "loss": 1.1765,
       "step": 1220
     },
     {
       "epoch": 1.1785714285714286,
-      "grad_norm": 4.427639007568359,
+      "grad_norm": 4.398950099945068,
       "learning_rate": 1.3858109494904437e-05,
-      "loss": 1.1951,
+      "loss": 1.1729,
       "step": 1221
     },
     {
       "epoch": 1.1795366795366795,
-      "grad_norm": 4.049868106842041,
+      "grad_norm": 4.100317001342773,
       "learning_rate": 1.3848491064267046e-05,
-      "loss": 1.187,
+      "loss": 1.1682,
       "step": 1222
     },
     {
       "epoch": 1.1805019305019304,
-      "grad_norm": 3.7526743412017822,
+      "grad_norm": 3.8548827171325684,
       "learning_rate": 1.3838868452403172e-05,
-      "loss": 1.1953,
+      "loss": 1.1712,
       "step": 1223
     },
     {
       "epoch": 1.1814671814671815,
-      "grad_norm": 4.315376281738281,
+      "grad_norm": 4.3090434074401855,
       "learning_rate": 1.3829241669767381e-05,
-      "loss": 1.1979,
+      "loss": 1.172,
       "step": 1224
     },
     {
       "epoch": 1.1824324324324325,
-      "grad_norm": 4.181025505065918,
+      "grad_norm": 4.033326148986816,
       "learning_rate": 1.381961072681878e-05,
-      "loss": 1.1545,
+      "loss": 1.1368,
       "step": 1225
     },
     {
       "epoch": 1.1833976833976834,
-      "grad_norm": 4.2237548828125,
+      "grad_norm": 4.149012088775635,
       "learning_rate": 1.3809975634020985e-05,
-      "loss": 1.1436,
+      "loss": 1.1245,
       "step": 1226
     },
     {
       "epoch": 1.1843629343629343,
-      "grad_norm": 4.464808464050293,
+      "grad_norm": 4.336759090423584,
       "learning_rate": 1.380033640184213e-05,
-      "loss": 1.1657,
+      "loss": 1.1351,
       "step": 1227
     },
     {
       "epoch": 1.1853281853281854,
-      "grad_norm": 4.275068759918213,
+      "grad_norm": 4.1892266273498535,
       "learning_rate": 1.3790693040754839e-05,
-      "loss": 1.2498,
+      "loss": 1.2171,
       "step": 1228
     },
     {
       "epoch": 1.1862934362934363,
-      "grad_norm": 4.397349834442139,
+      "grad_norm": 4.195492267608643,
       "learning_rate": 1.3781045561236228e-05,
-      "loss": 1.2087,
+      "loss": 1.1913,
       "step": 1229
     },
     {
       "epoch": 1.1872586872586872,
-      "grad_norm": 4.3661980628967285,
+      "grad_norm": 4.193690299987793,
       "learning_rate": 1.3771393973767886e-05,
-      "loss": 1.2416,
+      "loss": 1.2298,
       "step": 1230
     },
     {
       "epoch": 1.1882239382239383,
-      "grad_norm": 4.0517802238464355,
+      "grad_norm": 3.944406509399414,
       "learning_rate": 1.376173828883586e-05,
-      "loss": 1.1691,
+      "loss": 1.1553,
       "step": 1231
     },
     {
       "epoch": 1.1891891891891893,
-      "grad_norm": 4.048524856567383,
+      "grad_norm": 4.0007171630859375,
       "learning_rate": 1.3752078516930653e-05,
-      "loss": 1.1568,
+      "loss": 1.1328,
       "step": 1232
     },
     {
       "epoch": 1.1901544401544402,
-      "grad_norm": 4.841455936431885,
+      "grad_norm": 4.685445308685303,
       "learning_rate": 1.374241466854721e-05,
-      "loss": 1.1866,
+      "loss": 1.1635,
       "step": 1233
     },
     {
       "epoch": 1.191119691119691,
-      "grad_norm": 4.119144439697266,
+      "grad_norm": 3.9061317443847656,
       "learning_rate": 1.3732746754184905e-05,
-      "loss": 1.2027,
+      "loss": 1.1763,
       "step": 1234
     },
     {
       "epoch": 1.192084942084942,
-      "grad_norm": 4.689269542694092,
+      "grad_norm": 4.485651016235352,
       "learning_rate": 1.3723074784347521e-05,
-      "loss": 1.2116,
+      "loss": 1.1921,
       "step": 1235
     },
     {
       "epoch": 1.193050193050193,
-      "grad_norm": 5.033204078674316,
+      "grad_norm": 4.581057071685791,
       "learning_rate": 1.371339876954326e-05,
-      "loss": 1.2976,
+      "loss": 1.2709,
       "step": 1236
     },
     {
       "epoch": 1.194015444015444,
-      "grad_norm": 4.489188194274902,
+      "grad_norm": 4.367702484130859,
       "learning_rate": 1.3703718720284707e-05,
-      "loss": 1.0913,
+      "loss": 1.0733,
       "step": 1237
     },
     {
       "epoch": 1.194980694980695,
-      "grad_norm": 4.7308669090271,
+      "grad_norm": 4.548466205596924,
       "learning_rate": 1.369403464708884e-05,
-      "loss": 1.216,
+      "loss": 1.1856,
       "step": 1238
     },
     {
       "epoch": 1.195945945945946,
-      "grad_norm": 4.222895622253418,
+      "grad_norm": 4.026926040649414,
       "learning_rate": 1.3684346560477e-05,
-      "loss": 1.1872,
+      "loss": 1.1783,
       "step": 1239
     },
     {
       "epoch": 1.196911196911197,
-      "grad_norm": 3.997390031814575,
+      "grad_norm": 3.880228281021118,
       "learning_rate": 1.3674654470974897e-05,
-      "loss": 1.1965,
+      "loss": 1.1843,
       "step": 1240
     },
     {
       "epoch": 1.1978764478764479,
-      "grad_norm": 4.557868003845215,
+      "grad_norm": 4.35034704208374,
       "learning_rate": 1.3664958389112581e-05,
-      "loss": 1.2309,
+      "loss": 1.2073,
       "step": 1241
     },
     {
       "epoch": 1.1988416988416988,
-      "grad_norm": 4.424615383148193,
+      "grad_norm": 4.249173641204834,
       "learning_rate": 1.3655258325424447e-05,
-      "loss": 1.2487,
+      "loss": 1.2268,
       "step": 1242
     },
     {
       "epoch": 1.1998069498069497,
-      "grad_norm": 4.087277889251709,
+      "grad_norm": 3.9031732082366943,
       "learning_rate": 1.3645554290449215e-05,
-      "loss": 1.2148,
+      "loss": 1.1931,
       "step": 1243
     },
     {
       "epoch": 1.2007722007722008,
-      "grad_norm": 3.749039649963379,
+      "grad_norm": 3.7276415824890137,
       "learning_rate": 1.3635846294729915e-05,
-      "loss": 1.1904,
+      "loss": 1.1635,
       "step": 1244
     },
     {
       "epoch": 1.2017374517374517,
-      "grad_norm": 4.259382724761963,
+      "grad_norm": 4.315927028656006,
       "learning_rate": 1.3626134348813885e-05,
-      "loss": 1.2536,
+      "loss": 1.2191,
       "step": 1245
     },
     {
       "epoch": 1.2027027027027026,
-      "grad_norm": 4.148220062255859,
+      "grad_norm": 4.035200119018555,
       "learning_rate": 1.3616418463252747e-05,
-      "loss": 1.2717,
+      "loss": 1.2413,
       "step": 1246
     },
     {
       "epoch": 1.2036679536679538,
-      "grad_norm": 4.54502534866333,
+      "grad_norm": 4.93955659866333,
       "learning_rate": 1.360669864860242e-05,
-      "loss": 1.2153,
+      "loss": 1.1855,
       "step": 1247
     },
     {
       "epoch": 1.2046332046332047,
-      "grad_norm": 3.793149471282959,
+      "grad_norm": 4.154675006866455,
       "learning_rate": 1.3596974915423071e-05,
-      "loss": 1.1592,
+      "loss": 1.1374,
       "step": 1248
     },
     {
       "epoch": 1.2055984555984556,
-      "grad_norm": 4.2187652587890625,
+      "grad_norm": 3.961106061935425,
       "learning_rate": 1.358724727427914e-05,
-      "loss": 1.1378,
+      "loss": 1.1127,
       "step": 1249
     },
     {
       "epoch": 1.2065637065637065,
-      "grad_norm": 4.317118167877197,
+      "grad_norm": 4.575501441955566,
       "learning_rate": 1.3577515735739302e-05,
-      "loss": 1.2069,
+      "loss": 1.1876,
       "step": 1250
     },
     {
       "epoch": 1.2075289575289576,
-      "grad_norm": 4.604206562042236,
+      "grad_norm": 4.848623752593994,
       "learning_rate": 1.3567780310376476e-05,
-      "loss": 1.1417,
+      "loss": 1.1133,
       "step": 1251
     },
     {
       "epoch": 1.2084942084942085,
-      "grad_norm": 4.524065017700195,
+      "grad_norm": 4.288299083709717,
       "learning_rate": 1.3558041008767799e-05,
-      "loss": 1.1692,
+      "loss": 1.1365,
       "step": 1252
     },
     {
       "epoch": 1.2094594594594594,
-      "grad_norm": 4.630565166473389,
+      "grad_norm": 4.583204746246338,
       "learning_rate": 1.354829784149462e-05,
-      "loss": 1.217,
+      "loss": 1.1825,
       "step": 1253
     },
     {
       "epoch": 1.2104247104247103,
-      "grad_norm": 4.320188045501709,
+      "grad_norm": 4.781323432922363,
       "learning_rate": 1.3538550819142484e-05,
-      "loss": 1.2048,
+      "loss": 1.1742,
       "step": 1254
     },
     {
       "epoch": 1.2113899613899615,
-      "grad_norm": 4.694369792938232,
+      "grad_norm": 4.196986675262451,
       "learning_rate": 1.3528799952301128e-05,
-      "loss": 1.2499,
+      "loss": 1.2165,
       "step": 1255
     },
     {
       "epoch": 1.2123552123552124,
-      "grad_norm": 4.20485782623291,
+      "grad_norm": 4.358056545257568,
       "learning_rate": 1.351904525156447e-05,
-      "loss": 1.2518,
+      "loss": 1.222,
       "step": 1256
     },
     {
       "epoch": 1.2133204633204633,
-      "grad_norm": 4.132609844207764,
+      "grad_norm": 4.231851577758789,
       "learning_rate": 1.3509286727530588e-05,
-      "loss": 1.1469,
+      "loss": 1.1361,
       "step": 1257
     },
     {
       "epoch": 1.2142857142857142,
-      "grad_norm": 4.2905592918396,
+      "grad_norm": 4.035199165344238,
       "learning_rate": 1.3499524390801716e-05,
-      "loss": 1.2003,
+      "loss": 1.1788,
       "step": 1258
     },
     {
       "epoch": 1.2152509652509653,
-      "grad_norm": 4.843905448913574,
+      "grad_norm": 4.5761332511901855,
       "learning_rate": 1.3489758251984222e-05,
-      "loss": 1.2131,
+      "loss": 1.1826,
       "step": 1259
     },
     {
       "epoch": 1.2162162162162162,
-      "grad_norm": 4.21188497543335,
+      "grad_norm": 4.045440196990967,
       "learning_rate": 1.3479988321688619e-05,
-      "loss": 1.2113,
+      "loss": 1.1842,
       "step": 1260
     },
     {
       "epoch": 1.2171814671814671,
-      "grad_norm": 4.492689609527588,
+      "grad_norm": 4.254245758056641,
       "learning_rate": 1.3470214610529536e-05,
-      "loss": 1.2315,
+      "loss": 1.1995,
       "step": 1261
     },
     {
       "epoch": 1.218146718146718,
-      "grad_norm": 4.724948883056641,
+      "grad_norm": 4.792616367340088,
       "learning_rate": 1.3460437129125694e-05,
-      "loss": 1.2238,
+      "loss": 1.2062,
       "step": 1262
     },
     {
       "epoch": 1.2191119691119692,
-      "grad_norm": 4.601674556732178,
+      "grad_norm": 4.571348667144775,
       "learning_rate": 1.3450655888099935e-05,
-      "loss": 1.1761,
+      "loss": 1.1473,
       "step": 1263
     },
     {
       "epoch": 1.22007722007722,
-      "grad_norm": 5.123815059661865,
+      "grad_norm": 4.655112266540527,
       "learning_rate": 1.3440870898079166e-05,
-      "loss": 1.2684,
+      "loss": 1.2398,
       "step": 1264
     },
     {
       "epoch": 1.221042471042471,
-      "grad_norm": 4.135692119598389,
+      "grad_norm": 4.464754104614258,
       "learning_rate": 1.3431082169694376e-05,
-      "loss": 1.1713,
+      "loss": 1.1454,
       "step": 1265
     },
     {
       "epoch": 1.2220077220077221,
-      "grad_norm": 4.55994176864624,
+      "grad_norm": 4.222874641418457,
       "learning_rate": 1.3421289713580616e-05,
-      "loss": 1.2257,
+      "loss": 1.194,
       "step": 1266
     },
     {
       "epoch": 1.222972972972973,
-      "grad_norm": 4.352341175079346,
+      "grad_norm": 4.210389614105225,
       "learning_rate": 1.341149354037698e-05,
-      "loss": 1.223,
+      "loss": 1.189,
       "step": 1267
     },
     {
       "epoch": 1.223938223938224,
-      "grad_norm": 4.42564058303833,
+      "grad_norm": 5.461085796356201,
       "learning_rate": 1.340169366072661e-05,
-      "loss": 1.1589,
+      "loss": 1.1427,
       "step": 1268
     },
     {
       "epoch": 1.2249034749034748,
-      "grad_norm": 4.8221611976623535,
+      "grad_norm": 4.228640556335449,
       "learning_rate": 1.3391890085276669e-05,
-      "loss": 1.2191,
+      "loss": 1.1864,
       "step": 1269
     },
     {
       "epoch": 1.2258687258687258,
-      "grad_norm": 4.877390384674072,
+      "grad_norm": 4.811310291290283,
       "learning_rate": 1.3382082824678336e-05,
-      "loss": 1.2135,
+      "loss": 1.1854,
       "step": 1270
     },
     {
       "epoch": 1.2268339768339769,
-      "grad_norm": 4.110681533813477,
+      "grad_norm": 4.434942245483398,
       "learning_rate": 1.337227188958679e-05,
-      "loss": 1.172,
+      "loss": 1.1406,
       "step": 1271
     },
     {
       "epoch": 1.2277992277992278,
-      "grad_norm": 5.047957420349121,
+      "grad_norm": 4.15158224105835,
       "learning_rate": 1.3362457290661215e-05,
-      "loss": 1.2651,
+      "loss": 1.2343,
       "step": 1272
     },
     {
       "epoch": 1.2287644787644787,
-      "grad_norm": 4.183019161224365,
+      "grad_norm": 4.299196243286133,
       "learning_rate": 1.335263903856476e-05,
-      "loss": 1.2434,
+      "loss": 1.2124,
       "step": 1273
     },
     {
       "epoch": 1.2297297297297298,
-      "grad_norm": 5.047949314117432,
+      "grad_norm": 4.241453647613525,
       "learning_rate": 1.3342817143964557e-05,
-      "loss": 1.2027,
+      "loss": 1.174,
       "step": 1274
     },
     {
       "epoch": 1.2306949806949807,
-      "grad_norm": 4.732578277587891,
+      "grad_norm": 4.511514186859131,
       "learning_rate": 1.333299161753168e-05,
-      "loss": 1.2546,
+      "loss": 1.2138,
       "step": 1275
     },
     {
       "epoch": 1.2316602316602316,
-      "grad_norm": 4.286832809448242,
+      "grad_norm": 4.485677719116211,
       "learning_rate": 1.3323162469941164e-05,
-      "loss": 1.2295,
+      "loss": 1.2135,
       "step": 1276
     },
     {
       "epoch": 1.2326254826254825,
-      "grad_norm": 4.886010646820068,
+      "grad_norm": 4.350978851318359,
       "learning_rate": 1.3313329711871969e-05,
-      "loss": 1.1589,
+      "loss": 1.1412,
       "step": 1277
     },
     {
       "epoch": 1.2335907335907337,
-      "grad_norm": 4.348087310791016,
+      "grad_norm": 4.344076633453369,
       "learning_rate": 1.3303493354006985e-05,
-      "loss": 1.2417,
+      "loss": 1.2185,
       "step": 1278
     },
     {
       "epoch": 1.2345559845559846,
-      "grad_norm": 4.194649696350098,
+      "grad_norm": 4.518075466156006,
       "learning_rate": 1.3293653407032997e-05,
-      "loss": 1.197,
+      "loss": 1.1668,
       "step": 1279
     },
     {
       "epoch": 1.2355212355212355,
-      "grad_norm": 4.147604942321777,
+      "grad_norm": 4.878932952880859,
       "learning_rate": 1.328380988164071e-05,
-      "loss": 1.2041,
+      "loss": 1.1847,
       "step": 1280
     },
     {
       "epoch": 1.2364864864864864,
-      "grad_norm": 4.729609489440918,
+      "grad_norm": 4.558157920837402,
       "learning_rate": 1.3273962788524705e-05,
-      "loss": 1.2053,
+      "loss": 1.1844,
       "step": 1281
     },
     {
       "epoch": 1.2374517374517375,
-      "grad_norm": 5.589478969573975,
+      "grad_norm": 6.208391189575195,
       "learning_rate": 1.3264112138383445e-05,
-      "loss": 1.262,
+      "loss": 1.2344,
       "step": 1282
     },
     {
       "epoch": 1.2384169884169884,
-      "grad_norm": 3.983586072921753,
+      "grad_norm": 4.216725826263428,
       "learning_rate": 1.3254257941919249e-05,
-      "loss": 1.1723,
+      "loss": 1.1433,
       "step": 1283
     },
     {
       "epoch": 1.2393822393822393,
-      "grad_norm": 4.140334606170654,
+      "grad_norm": 4.023191928863525,
       "learning_rate": 1.3244400209838297e-05,
-      "loss": 1.2247,
+      "loss": 1.1944,
       "step": 1284
     },
     {
       "epoch": 1.2403474903474903,
-      "grad_norm": 4.859688758850098,
+      "grad_norm": 4.910943031311035,
       "learning_rate": 1.3234538952850606e-05,
-      "loss": 1.1928,
+      "loss": 1.1664,
       "step": 1285
     },
     {
       "epoch": 1.2413127413127414,
-      "grad_norm": 4.2499189376831055,
+      "grad_norm": 4.637852191925049,
       "learning_rate": 1.3224674181670025e-05,
-      "loss": 1.158,
+      "loss": 1.1216,
       "step": 1286
     },
     {
       "epoch": 1.2422779922779923,
-      "grad_norm": 3.908384323120117,
+      "grad_norm": 4.14736795425415,
       "learning_rate": 1.3214805907014223e-05,
-      "loss": 1.2375,
+      "loss": 1.2027,
       "step": 1287
     },
     {
       "epoch": 1.2432432432432432,
-      "grad_norm": 4.4078168869018555,
+      "grad_norm": 4.636044502258301,
       "learning_rate": 1.3204934139604668e-05,
-      "loss": 1.1594,
+      "loss": 1.144,
       "step": 1288
     },
     {
       "epoch": 1.244208494208494,
-      "grad_norm": 4.318722724914551,
+      "grad_norm": 4.665686130523682,
       "learning_rate": 1.3195058890166628e-05,
-      "loss": 1.2081,
+      "loss": 1.1804,
       "step": 1289
     },
     {
       "epoch": 1.2451737451737452,
-      "grad_norm": 4.253969192504883,
+      "grad_norm": 4.089426040649414,
       "learning_rate": 1.3185180169429155e-05,
-      "loss": 1.1388,
+      "loss": 1.1087,
       "step": 1290
     },
     {
       "epoch": 1.2461389961389961,
-      "grad_norm": 4.060301780700684,
+      "grad_norm": 4.140326023101807,
       "learning_rate": 1.3175297988125071e-05,
-      "loss": 1.2017,
+      "loss": 1.1779,
       "step": 1291
     },
     {
       "epoch": 1.247104247104247,
-      "grad_norm": 4.686599254608154,
+      "grad_norm": 5.166257858276367,
       "learning_rate": 1.3165412356990954e-05,
-      "loss": 1.1764,
+      "loss": 1.1393,
       "step": 1292
     },
     {
       "epoch": 1.2480694980694982,
-      "grad_norm": 4.2792067527771,
+      "grad_norm": 4.020298957824707,
       "learning_rate": 1.315552328676714e-05,
-      "loss": 1.1698,
+      "loss": 1.1379,
       "step": 1293
     },
     {
       "epoch": 1.249034749034749,
-      "grad_norm": 4.322442531585693,
+      "grad_norm": 4.3975043296813965,
       "learning_rate": 1.314563078819769e-05,
-      "loss": 1.1632,
+      "loss": 1.1376,
       "step": 1294
     },
     {
       "epoch": 1.25,
-      "grad_norm": 4.505197525024414,
+      "grad_norm": 4.989447116851807,
       "learning_rate": 1.3135734872030397e-05,
-      "loss": 1.2016,
+      "loss": 1.1643,
       "step": 1295
     },
     {
       "epoch": 1.250965250965251,
-      "grad_norm": 4.2458906173706055,
+      "grad_norm": 4.167351722717285,
       "learning_rate": 1.3125835549016763e-05,
-      "loss": 1.1946,
+      "loss": 1.1712,
       "step": 1296
     },
     {
       "epoch": 1.2519305019305018,
-      "grad_norm": 4.345571517944336,
+      "grad_norm": 4.720897674560547,
       "learning_rate": 1.3115932829911997e-05,
-      "loss": 1.2077,
+      "loss": 1.1767,
       "step": 1297
     },
     {
       "epoch": 1.252895752895753,
-      "grad_norm": 4.5000433921813965,
+      "grad_norm": 5.669806003570557,
       "learning_rate": 1.3106026725474987e-05,
-      "loss": 1.1879,
+      "loss": 1.1663,
       "step": 1298
     },
     {
       "epoch": 1.2538610038610039,
-      "grad_norm": 4.399386405944824,
+      "grad_norm": 4.570425987243652,
       "learning_rate": 1.3096117246468317e-05,
-      "loss": 1.2398,
+      "loss": 1.2082,
       "step": 1299
     },
     {
       "epoch": 1.2548262548262548,
-      "grad_norm": 4.267080307006836,
+      "grad_norm": 4.390018939971924,
       "learning_rate": 1.3086204403658216e-05,
-      "loss": 1.2535,
+      "loss": 1.2185,
       "step": 1300
     },
     {
       "epoch": 1.255791505791506,
-      "grad_norm": 4.634487628936768,
+      "grad_norm": 6.695591449737549,
       "learning_rate": 1.3076288207814585e-05,
-      "loss": 1.3211,
+      "loss": 1.2923,
       "step": 1301
     },
     {
       "epoch": 1.2567567567567568,
-      "grad_norm": 3.9481966495513916,
+      "grad_norm": 4.250987529754639,
       "learning_rate": 1.3066368669710953e-05,
-      "loss": 1.1489,
+      "loss": 1.1191,
       "step": 1302
     },
     {
       "epoch": 1.2577220077220077,
-      "grad_norm": 4.762670993804932,
+      "grad_norm": 4.857234001159668,
       "learning_rate": 1.3056445800124497e-05,
-      "loss": 1.217,
+      "loss": 1.1836,
       "step": 1303
     },
     {
       "epoch": 1.2586872586872586,
-      "grad_norm": 4.1158766746521,
+      "grad_norm": 5.951472282409668,
       "learning_rate": 1.3046519609836002e-05,
-      "loss": 1.1996,
+      "loss": 1.1763,
       "step": 1304
     },
     {
       "epoch": 1.2596525096525095,
-      "grad_norm": 4.694605827331543,
+      "grad_norm": 4.68471097946167,
       "learning_rate": 1.303659010962986e-05,
-      "loss": 1.1895,
+      "loss": 1.1574,
       "step": 1305
     },
     {
       "epoch": 1.2606177606177607,
-      "grad_norm": 5.370503902435303,
+      "grad_norm": 5.1641764640808105,
       "learning_rate": 1.3026657310294067e-05,
-      "loss": 1.2432,
+      "loss": 1.2092,
       "step": 1306
     },
     {
       "epoch": 1.2615830115830116,
-      "grad_norm": 4.1840362548828125,
+      "grad_norm": 5.911360740661621,
       "learning_rate": 1.3016721222620197e-05,
-      "loss": 1.2045,
+      "loss": 1.1798,
       "step": 1307
     },
     {
       "epoch": 1.2625482625482625,
-      "grad_norm": 5.977203845977783,
+      "grad_norm": 4.218881130218506,
       "learning_rate": 1.3006781857403394e-05,
-      "loss": 1.2218,
+      "loss": 1.1853,
       "step": 1308
     },
     {
       "epoch": 1.2635135135135136,
-      "grad_norm": 4.654899597167969,
+      "grad_norm": 4.663739204406738,
       "learning_rate": 1.2996839225442377e-05,
-      "loss": 1.3054,
+      "loss": 1.2706,
       "step": 1309
     },
     {
       "epoch": 1.2644787644787645,
-      "grad_norm": 4.7028727531433105,
+      "grad_norm": 5.501549243927002,
       "learning_rate": 1.2986893337539397e-05,
-      "loss": 1.2712,
+      "loss": 1.243,
       "step": 1310
     },
     {
       "epoch": 1.2654440154440154,
-      "grad_norm": 5.325465679168701,
+      "grad_norm": 4.621982097625732,
       "learning_rate": 1.2976944204500255e-05,
-      "loss": 1.1635,
+      "loss": 1.1274,
       "step": 1311
     },
     {
       "epoch": 1.2664092664092665,
-      "grad_norm": 4.675345420837402,
+      "grad_norm": 4.860998153686523,
       "learning_rate": 1.2966991837134271e-05,
-      "loss": 1.267,
+      "loss": 1.2317,
       "step": 1312
     },
     {
       "epoch": 1.2673745173745175,
-      "grad_norm": 4.558609962463379,
+      "grad_norm": 5.091229438781738,
       "learning_rate": 1.295703624625428e-05,
-      "loss": 1.2272,
+      "loss": 1.197,
       "step": 1313
     },
     {
       "epoch": 1.2683397683397684,
-      "grad_norm": 4.645543575286865,
+      "grad_norm": 4.268128395080566,
       "learning_rate": 1.294707744267662e-05,
-      "loss": 1.2366,
+      "loss": 1.2101,
       "step": 1314
     },
     {
       "epoch": 1.2693050193050193,
-      "grad_norm": 4.8329620361328125,
+      "grad_norm": 5.04547119140625,
       "learning_rate": 1.2937115437221119e-05,
-      "loss": 1.193,
+      "loss": 1.1632,
       "step": 1315
     },
     {
       "epoch": 1.2702702702702702,
-      "grad_norm": 4.300382137298584,
+      "grad_norm": 4.460942268371582,
       "learning_rate": 1.2927150240711089e-05,
-      "loss": 1.1907,
+      "loss": 1.1589,
       "step": 1316
     },
     {
       "epoch": 1.2712355212355213,
-      "grad_norm": 4.584596633911133,
+      "grad_norm": 4.028692245483398,
       "learning_rate": 1.2917181863973298e-05,
-      "loss": 1.2358,
+      "loss": 1.2034,
       "step": 1317
     },
     {
       "epoch": 1.2722007722007722,
-      "grad_norm": 4.23142671585083,
+      "grad_norm": 4.381527900695801,
       "learning_rate": 1.290721031783798e-05,
-      "loss": 1.2462,
+      "loss": 1.2081,
       "step": 1318
     },
     {
       "epoch": 1.2731660231660231,
-      "grad_norm": 4.2517828941345215,
+      "grad_norm": 4.259705066680908,
       "learning_rate": 1.2897235613138804e-05,
-      "loss": 1.2021,
+      "loss": 1.1663,
       "step": 1319
     },
     {
       "epoch": 1.2741312741312742,
-      "grad_norm": 4.605195999145508,
+      "grad_norm": 4.889810085296631,
       "learning_rate": 1.2887257760712875e-05,
-      "loss": 1.246,
+      "loss": 1.2141,
       "step": 1320
     },
     {
       "epoch": 1.2750965250965252,
-      "grad_norm": 4.4199347496032715,
+      "grad_norm": 4.390961647033691,
       "learning_rate": 1.287727677140072e-05,
-      "loss": 1.1999,
+      "loss": 1.1663,
       "step": 1321
     },
     {
       "epoch": 1.276061776061776,
-      "grad_norm": 3.882155656814575,
+      "grad_norm": 4.501543998718262,
       "learning_rate": 1.2867292656046268e-05,
-      "loss": 1.1948,
+      "loss": 1.1642,
       "step": 1322
     },
     {
       "epoch": 1.277027027027027,
-      "grad_norm": 4.638187885284424,
+      "grad_norm": 5.317819118499756,
       "learning_rate": 1.285730542549685e-05,
-      "loss": 1.2795,
+      "loss": 1.2599,
       "step": 1323
     },
     {
       "epoch": 1.2779922779922779,
-      "grad_norm": 4.155126571655273,
+      "grad_norm": 4.074085712432861,
       "learning_rate": 1.284731509060318e-05,
-      "loss": 1.1946,
+      "loss": 1.1575,
       "step": 1324
     },
     {
       "epoch": 1.278957528957529,
-      "grad_norm": 4.521946907043457,
+      "grad_norm": 6.189610481262207,
       "learning_rate": 1.2837321662219342e-05,
-      "loss": 1.2178,
+      "loss": 1.1843,
       "step": 1325
     },
     {
       "epoch": 1.27992277992278,
-      "grad_norm": 4.1087212562561035,
+      "grad_norm": 4.341475009918213,
       "learning_rate": 1.2827325151202783e-05,
-      "loss": 1.2762,
+      "loss": 1.2388,
       "step": 1326
     },
     {
       "epoch": 1.2808880308880308,
-      "grad_norm": 3.956718683242798,
+      "grad_norm": 3.9079694747924805,
       "learning_rate": 1.2817325568414299e-05,
-      "loss": 1.2021,
+      "loss": 1.1684,
       "step": 1327
     },
     {
       "epoch": 1.281853281853282,
-      "grad_norm": 4.165088176727295,
+      "grad_norm": 5.74494743347168,
       "learning_rate": 1.2807322924718024e-05,
-      "loss": 1.1841,
+      "loss": 1.1505,
       "step": 1328
     },
     {
       "epoch": 1.2828185328185329,
-      "grad_norm": 4.392886161804199,
+      "grad_norm": 5.171532154083252,
       "learning_rate": 1.2797317230981416e-05,
-      "loss": 1.1817,
+      "loss": 1.1424,
       "step": 1329
     },
     {
       "epoch": 1.2837837837837838,
-      "grad_norm": 4.641284465789795,
+      "grad_norm": 4.2881364822387695,
       "learning_rate": 1.2787308498075253e-05,
-      "loss": 1.1638,
+      "loss": 1.1259,
       "step": 1330
     },
     {
       "epoch": 1.2847490347490347,
-      "grad_norm": 4.319555759429932,
+      "grad_norm": 5.986820220947266,
       "learning_rate": 1.27772967368736e-05,
-      "loss": 1.268,
+      "loss": 1.2442,
       "step": 1331
     },
     {
       "epoch": 1.2857142857142856,
-      "grad_norm": 4.382452964782715,
+      "grad_norm": 4.524091720581055,
       "learning_rate": 1.276728195825383e-05,
-      "loss": 1.1895,
+      "loss": 1.1433,
       "step": 1332
     },
     {
       "epoch": 1.2866795366795367,
-      "grad_norm": 4.566013813018799,
+      "grad_norm": 4.656999111175537,
       "learning_rate": 1.2757264173096582e-05,
-      "loss": 1.213,
+      "loss": 1.1871,
       "step": 1333
     },
     {
       "epoch": 1.2876447876447876,
-      "grad_norm": 4.029094219207764,
+      "grad_norm": 5.563279628753662,
       "learning_rate": 1.2747243392285769e-05,
-      "loss": 1.2524,
+      "loss": 1.2231,
       "step": 1334
     },
     {
       "epoch": 1.2886100386100385,
-      "grad_norm": 4.885570049285889,
+      "grad_norm": 4.622396469116211,
       "learning_rate": 1.2737219626708549e-05,
-      "loss": 1.2848,
+      "loss": 1.2503,
       "step": 1335
     },
     {
       "epoch": 1.2895752895752897,
-      "grad_norm": 4.629604816436768,
+      "grad_norm": 4.35022497177124,
       "learning_rate": 1.2727192887255332e-05,
-      "loss": 1.1951,
+      "loss": 1.1563,
       "step": 1336
     },
     {
       "epoch": 1.2905405405405406,
-      "grad_norm": 4.667120933532715,
+      "grad_norm": 4.997153282165527,
       "learning_rate": 1.2717163184819761e-05,
-      "loss": 1.1794,
+      "loss": 1.1495,
       "step": 1337
     },
     {
       "epoch": 1.2915057915057915,
-      "grad_norm": 4.037268161773682,
+      "grad_norm": 5.1063151359558105,
       "learning_rate": 1.2707130530298688e-05,
-      "loss": 1.2396,
+      "loss": 1.1963,
       "step": 1338
     },
     {
       "epoch": 1.2924710424710426,
-      "grad_norm": 4.660993576049805,
+      "grad_norm": 4.5562310218811035,
       "learning_rate": 1.2697094934592177e-05,
-      "loss": 1.2385,
+      "loss": 1.2099,
       "step": 1339
     },
     {
       "epoch": 1.2934362934362935,
-      "grad_norm": 4.749906539916992,
+      "grad_norm": 4.840873718261719,
       "learning_rate": 1.2687056408603493e-05,
-      "loss": 1.1782,
+      "loss": 1.1301,
       "step": 1340
     },
     {
       "epoch": 1.2944015444015444,
-      "grad_norm": 4.700593948364258,
+      "grad_norm": 4.116574764251709,
       "learning_rate": 1.2677014963239075e-05,
-      "loss": 1.2046,
+      "loss": 1.1614,
       "step": 1341
     },
     {
       "epoch": 1.2953667953667953,
-      "grad_norm": 4.356866359710693,
+      "grad_norm": 4.393024921417236,
       "learning_rate": 1.2666970609408535e-05,
-      "loss": 1.2227,
+      "loss": 1.199,
       "step": 1342
     },
     {
       "epoch": 1.2963320463320462,
-      "grad_norm": 5.177303791046143,
+      "grad_norm": 4.218954563140869,
       "learning_rate": 1.2656923358024659e-05,
-      "loss": 1.2335,
+      "loss": 1.1929,
       "step": 1343
     },
     {
       "epoch": 1.2972972972972974,
-      "grad_norm": 5.240119934082031,
+      "grad_norm": 4.200098037719727,
       "learning_rate": 1.264687322000336e-05,
-      "loss": 1.1615,
+      "loss": 1.1198,
       "step": 1344
     },
     {
       "epoch": 1.2982625482625483,
-      "grad_norm": 4.722900867462158,
+      "grad_norm": 4.5496506690979,
       "learning_rate": 1.2636820206263704e-05,
-      "loss": 1.211,
+      "loss": 1.1736,
       "step": 1345
     },
     {
       "epoch": 1.2992277992277992,
-      "grad_norm": 5.814818859100342,
+      "grad_norm": 4.075380802154541,
       "learning_rate": 1.2626764327727871e-05,
-      "loss": 1.2365,
+      "loss": 1.204,
       "step": 1346
     },
     {
       "epoch": 1.3001930501930503,
-      "grad_norm": 4.204340934753418,
+      "grad_norm": 3.8272788524627686,
       "learning_rate": 1.261670559532116e-05,
-      "loss": 1.1819,
+      "loss": 1.1378,
       "step": 1347
     },
     {
       "epoch": 1.3011583011583012,
-      "grad_norm": 4.929001331329346,
+      "grad_norm": 4.229931354522705,
       "learning_rate": 1.2606644019971967e-05,
-      "loss": 1.2352,
+      "loss": 1.2021,
       "step": 1348
     },
     {
       "epoch": 1.3021235521235521,
-      "grad_norm": 4.845025539398193,
+      "grad_norm": 4.255991458892822,
       "learning_rate": 1.259657961261178e-05,
-      "loss": 1.2067,
+      "loss": 1.1553,
       "step": 1349
     },
     {
       "epoch": 1.303088803088803,
-      "grad_norm": 4.197154998779297,
+      "grad_norm": 4.27820348739624,
       "learning_rate": 1.2586512384175156e-05,
-      "loss": 1.2472,
+      "loss": 1.2019,
       "step": 1350
     },
     {
       "epoch": 1.304054054054054,
-      "grad_norm": 4.3102922439575195,
+      "grad_norm": 4.52101993560791,
       "learning_rate": 1.2576442345599729e-05,
-      "loss": 1.2445,
+      "loss": 1.1998,
       "step": 1351
     },
     {
       "epoch": 1.305019305019305,
-      "grad_norm": 4.416947364807129,
+      "grad_norm": 4.122433185577393,
       "learning_rate": 1.2566369507826175e-05,
-      "loss": 1.2305,
+      "loss": 1.1921,
       "step": 1352
     },
     {
       "epoch": 1.305984555984556,
-      "grad_norm": 4.655396461486816,
+      "grad_norm": 4.375638008117676,
       "learning_rate": 1.2556293881798218e-05,
-      "loss": 1.2124,
+      "loss": 1.1764,
       "step": 1353
     },
     {
       "epoch": 1.306949806949807,
-      "grad_norm": 4.648707866668701,
+      "grad_norm": 4.206446170806885,
       "learning_rate": 1.2546215478462611e-05,
-      "loss": 1.1722,
+      "loss": 1.1206,
       "step": 1354
     },
     {
       "epoch": 1.307915057915058,
-      "grad_norm": 4.69895076751709,
+      "grad_norm": 4.147001266479492,
       "learning_rate": 1.2536134308769118e-05,
-      "loss": 1.1696,
+      "loss": 1.1407,
       "step": 1355
     },
     {
       "epoch": 1.308880308880309,
-      "grad_norm": 4.79047966003418,
+      "grad_norm": 4.764206409454346,
       "learning_rate": 1.2526050383670516e-05,
-      "loss": 1.1772,
+      "loss": 1.1285,
       "step": 1356
     },
     {
       "epoch": 1.3098455598455598,
-      "grad_norm": 4.19055700302124,
+      "grad_norm": 4.390753269195557,
       "learning_rate": 1.2515963714122577e-05,
-      "loss": 1.1856,
+      "loss": 1.1401,
       "step": 1357
     },
     {
       "epoch": 1.3108108108108107,
-      "grad_norm": 4.9402241706848145,
+      "grad_norm": 4.43456506729126,
       "learning_rate": 1.2505874311084041e-05,
-      "loss": 1.2117,
+      "loss": 1.1816,
       "step": 1358
     },
     {
       "epoch": 1.3117760617760617,
-      "grad_norm": 5.0677924156188965,
+      "grad_norm": 4.900396823883057,
       "learning_rate": 1.2495782185516638e-05,
-      "loss": 1.2339,
+      "loss": 1.2048,
       "step": 1359
     },
     {
       "epoch": 1.3127413127413128,
-      "grad_norm": 4.164126396179199,
+      "grad_norm": 4.672094821929932,
       "learning_rate": 1.2485687348385038e-05,
-      "loss": 1.1588,
+      "loss": 1.1118,
       "step": 1360
     },
     {
       "epoch": 1.3137065637065637,
-      "grad_norm": 5.482615947723389,
+      "grad_norm": 4.272616863250732,
       "learning_rate": 1.2475589810656868e-05,
-      "loss": 1.1945,
+      "loss": 1.1602,
       "step": 1361
     },
     {
       "epoch": 1.3146718146718146,
-      "grad_norm": 5.1199750900268555,
+      "grad_norm": 5.021596908569336,
       "learning_rate": 1.246548958330268e-05,
-      "loss": 1.1531,
+      "loss": 1.1141,
       "step": 1362
     },
     {
       "epoch": 1.3156370656370657,
-      "grad_norm": 5.075857162475586,
+      "grad_norm": 4.822278022766113,
       "learning_rate": 1.2455386677295962e-05,
-      "loss": 1.1786,
+      "loss": 1.1493,
       "step": 1363
     },
     {
       "epoch": 1.3166023166023166,
-      "grad_norm": 6.397689342498779,
+      "grad_norm": 4.732657432556152,
       "learning_rate": 1.2445281103613096e-05,
-      "loss": 1.2092,
+      "loss": 1.1734,
       "step": 1364
     },
     {
       "epoch": 1.3175675675675675,
-      "grad_norm": 4.174447059631348,
+      "grad_norm": 4.594897270202637,
       "learning_rate": 1.2435172873233371e-05,
-      "loss": 1.1784,
+      "loss": 1.1489,
       "step": 1365
     },
     {
       "epoch": 1.3185328185328185,
-      "grad_norm": 5.897645950317383,
+      "grad_norm": 4.135673999786377,
       "learning_rate": 1.2425061997138972e-05,
-      "loss": 1.2342,
+      "loss": 1.1848,
       "step": 1366
     },
     {
       "epoch": 1.3194980694980696,
-      "grad_norm": 5.2558794021606445,
+      "grad_norm": 4.408533573150635,
       "learning_rate": 1.2414948486314932e-05,
-      "loss": 1.263,
+      "loss": 1.223,
       "step": 1367
     },
     {
       "epoch": 1.3204633204633205,
-      "grad_norm": 4.3572845458984375,
+      "grad_norm": 4.819552421569824,
       "learning_rate": 1.2404832351749179e-05,
-      "loss": 1.2517,
+      "loss": 1.2112,
       "step": 1368
     },
     {
       "epoch": 1.3214285714285714,
-      "grad_norm": 5.881748676300049,
+      "grad_norm": 4.18181848526001,
       "learning_rate": 1.239471360443246e-05,
-      "loss": 1.2184,
+      "loss": 1.1771,
       "step": 1369
     },
     {
       "epoch": 1.3223938223938223,
-      "grad_norm": 4.852202415466309,
+      "grad_norm": 4.222750186920166,
       "learning_rate": 1.2384592255358385e-05,
-      "loss": 1.176,
+      "loss": 1.1348,
       "step": 1370
     },
     {
       "epoch": 1.3233590733590734,
-      "grad_norm": 4.252193450927734,
+      "grad_norm": 5.044025421142578,
       "learning_rate": 1.2374468315523375e-05,
-      "loss": 1.2621,
+      "loss": 1.2271,
       "step": 1371
     },
     {
       "epoch": 1.3243243243243243,
-      "grad_norm": 6.008481025695801,
+      "grad_norm": 4.4775471687316895,
       "learning_rate": 1.2364341795926684e-05,
-      "loss": 1.1949,
+      "loss": 1.1541,
       "step": 1372
     },
     {
       "epoch": 1.3252895752895753,
-      "grad_norm": 5.349253177642822,
+      "grad_norm": 4.7761759757995605,
       "learning_rate": 1.2354212707570344e-05,
-      "loss": 1.217,
+      "loss": 1.1805,
       "step": 1373
     },
     {
       "epoch": 1.3262548262548264,
-      "grad_norm": 5.8846435546875,
+      "grad_norm": 4.018557071685791,
       "learning_rate": 1.2344081061459195e-05,
-      "loss": 1.2277,
+      "loss": 1.1827,
       "step": 1374
     },
     {
       "epoch": 1.3272200772200773,
-      "grad_norm": 6.014350891113281,
+      "grad_norm": 4.207991123199463,
       "learning_rate": 1.2333946868600854e-05,
-      "loss": 1.1879,
+      "loss": 1.1419,
       "step": 1375
     },
     {
       "epoch": 1.3281853281853282,
-      "grad_norm": 3.9860541820526123,
+      "grad_norm": 4.1578545570373535,
       "learning_rate": 1.2323810140005701e-05,
-      "loss": 1.1859,
+      "loss": 1.1469,
       "step": 1376
     },
     {
       "epoch": 1.329150579150579,
-      "grad_norm": 6.001570701599121,
+      "grad_norm": 4.198439598083496,
       "learning_rate": 1.2313670886686874e-05,
-      "loss": 1.1648,
+      "loss": 1.1236,
       "step": 1377
     },
     {
       "epoch": 1.33011583011583,
-      "grad_norm": 5.547324180603027,
+      "grad_norm": 4.622223854064941,
       "learning_rate": 1.230352911966025e-05,
-      "loss": 1.2331,
+      "loss": 1.1804,
       "step": 1378
     },
     {
       "epoch": 1.3310810810810811,
-      "grad_norm": 4.244880199432373,
+      "grad_norm": 4.129556655883789,
       "learning_rate": 1.2293384849944445e-05,
-      "loss": 1.1715,
+      "loss": 1.1223,
       "step": 1379
     },
     {
       "epoch": 1.332046332046332,
-      "grad_norm": 5.324029445648193,
+      "grad_norm": 4.312408447265625,
       "learning_rate": 1.2283238088560783e-05,
-      "loss": 1.2098,
+      "loss": 1.1635,
       "step": 1380
     },
     {
       "epoch": 1.333011583011583,
-      "grad_norm": 5.186968803405762,
+      "grad_norm": 4.0953803062438965,
       "learning_rate": 1.2273088846533303e-05,
-      "loss": 1.1639,
+      "loss": 1.1227,
       "step": 1381
     },
     {
       "epoch": 1.333976833976834,
-      "grad_norm": 4.450983047485352,
+      "grad_norm": 4.319990634918213,
       "learning_rate": 1.226293713488874e-05,
-      "loss": 1.253,
+      "loss": 1.2086,
       "step": 1382
     },
     {
       "epoch": 1.334942084942085,
-      "grad_norm": 4.1214680671691895,
+      "grad_norm": 3.9820897579193115,
       "learning_rate": 1.2252782964656502e-05,
-      "loss": 1.1587,
+      "loss": 1.1149,
       "step": 1383
     },
     {
       "epoch": 1.335907335907336,
-      "grad_norm": 4.0626349449157715,
+      "grad_norm": 4.439387798309326,
       "learning_rate": 1.2242626346868682e-05,
-      "loss": 1.1852,
+      "loss": 1.1389,
       "step": 1384
     },
     {
       "epoch": 1.3368725868725868,
-      "grad_norm": 4.6084723472595215,
+      "grad_norm": 4.244980335235596,
       "learning_rate": 1.2232467292560025e-05,
-      "loss": 1.2039,
+      "loss": 1.1621,
       "step": 1385
     },
     {
       "epoch": 1.3378378378378377,
-      "grad_norm": 4.472030162811279,
+      "grad_norm": 5.179893970489502,
       "learning_rate": 1.2222305812767919e-05,
-      "loss": 1.2213,
+      "loss": 1.1771,
       "step": 1386
     },
     {
       "epoch": 1.3388030888030888,
-      "grad_norm": 4.617482662200928,
+      "grad_norm": 5.248672008514404,
       "learning_rate": 1.2212141918532394e-05,
-      "loss": 1.2015,
+      "loss": 1.1618,
       "step": 1387
     },
     {
       "epoch": 1.3397683397683398,
-      "grad_norm": 4.271813869476318,
+      "grad_norm": 4.305422306060791,
       "learning_rate": 1.2201975620896102e-05,
-      "loss": 1.1907,
+      "loss": 1.1497,
       "step": 1388
     },
     {
       "epoch": 1.3407335907335907,
-      "grad_norm": 4.236230373382568,
+      "grad_norm": 4.861572742462158,
       "learning_rate": 1.2191806930904304e-05,
-      "loss": 1.2442,
+      "loss": 1.2069,
       "step": 1389
     },
     {
       "epoch": 1.3416988416988418,
-      "grad_norm": 4.1355204582214355,
+      "grad_norm": 4.439981460571289,
       "learning_rate": 1.2181635859604864e-05,
-      "loss": 1.1606,
+      "loss": 1.1188,
       "step": 1390
     },
     {
       "epoch": 1.3426640926640927,
-      "grad_norm": 4.975465297698975,
+      "grad_norm": 4.996955871582031,
       "learning_rate": 1.2171462418048228e-05,
-      "loss": 1.1792,
+      "loss": 1.1434,
       "step": 1391
     },
     {
       "epoch": 1.3436293436293436,
-      "grad_norm": 4.425232410430908,
+      "grad_norm": 4.684093952178955,
       "learning_rate": 1.216128661728742e-05,
-      "loss": 1.2268,
+      "loss": 1.2046,
       "step": 1392
     },
     {
       "epoch": 1.3445945945945945,
-      "grad_norm": 4.223307132720947,
+      "grad_norm": 4.13955020904541,
       "learning_rate": 1.2151108468378028e-05,
-      "loss": 1.2575,
+      "loss": 1.2145,
       "step": 1393
     },
     {
       "epoch": 1.3455598455598454,
-      "grad_norm": 4.804199695587158,
+      "grad_norm": 4.376919269561768,
       "learning_rate": 1.2140927982378189e-05,
-      "loss": 1.2522,
+      "loss": 1.2038,
       "step": 1394
     },
     {
       "epoch": 1.3465250965250966,
-      "grad_norm": 4.646818161010742,
+      "grad_norm": 5.41073751449585,
       "learning_rate": 1.2130745170348584e-05,
-      "loss": 1.2433,
+      "loss": 1.2084,
       "step": 1395
     },
     {
       "epoch": 1.3474903474903475,
-      "grad_norm": 3.9267654418945312,
+      "grad_norm": 4.079817295074463,
       "learning_rate": 1.212056004335241e-05,
-      "loss": 1.1664,
+      "loss": 1.1151,
       "step": 1396
     },
     {
       "epoch": 1.3484555984555984,
-      "grad_norm": 4.747241497039795,
+      "grad_norm": 4.423679828643799,
       "learning_rate": 1.2110372612455394e-05,
-      "loss": 1.2441,
+      "loss": 1.1937,
       "step": 1397
     },
     {
       "epoch": 1.3494208494208495,
-      "grad_norm": 4.446130275726318,
+      "grad_norm": 5.345794677734375,
       "learning_rate": 1.2100182888725751e-05,
-      "loss": 1.1992,
+      "loss": 1.155,
       "step": 1398
     },
     {
       "epoch": 1.3503861003861004,
-      "grad_norm": 4.628722190856934,
+      "grad_norm": 4.360614776611328,
       "learning_rate": 1.2089990883234206e-05,
-      "loss": 1.207,
+      "loss": 1.1549,
       "step": 1399
     },
     {
       "epoch": 1.3513513513513513,
-      "grad_norm": 5.560851573944092,
+      "grad_norm": 4.595292568206787,
       "learning_rate": 1.2079796607053939e-05,
-      "loss": 1.2201,
+      "loss": 1.1731,
       "step": 1400
     },
     {
       "epoch": 1.3523166023166024,
-      "grad_norm": 3.9680185317993164,
+      "grad_norm": 4.903984069824219,
       "learning_rate": 1.2069600071260621e-05,
-      "loss": 1.1057,
+      "loss": 1.0593,
       "step": 1401
     },
     {
       "epoch": 1.3532818532818534,
-      "grad_norm": 5.056230068206787,
+      "grad_norm": 4.108474254608154,
       "learning_rate": 1.2059401286932354e-05,
-      "loss": 1.1756,
+      "loss": 1.1263,
       "step": 1402
     },
     {
       "epoch": 1.3542471042471043,
-      "grad_norm": 5.653619766235352,
+      "grad_norm": 4.674413204193115,
       "learning_rate": 1.204920026514971e-05,
-      "loss": 1.2339,
+      "loss": 1.1967,
       "step": 1403
     },
     {
       "epoch": 1.3552123552123552,
-      "grad_norm": 4.381045818328857,
+      "grad_norm": 5.17887544631958,
       "learning_rate": 1.2038997016995668e-05,
-      "loss": 1.2332,
+      "loss": 1.1898,
       "step": 1404
     },
     {
       "epoch": 1.356177606177606,
-      "grad_norm": 5.91451358795166,
+      "grad_norm": 4.410851955413818,
       "learning_rate": 1.202879155355564e-05,
-      "loss": 1.2509,
+      "loss": 1.2002,
       "step": 1405
     },
     {
       "epoch": 1.3571428571428572,
-      "grad_norm": 5.2406535148620605,
+      "grad_norm": 4.492305278778076,
       "learning_rate": 1.2018583885917434e-05,
-      "loss": 1.1803,
+      "loss": 1.135,
       "step": 1406
     },
     {
       "epoch": 1.3581081081081081,
-      "grad_norm": 3.9339377880096436,
+      "grad_norm": 4.695481777191162,
       "learning_rate": 1.2008374025171265e-05,
-      "loss": 1.142,
+      "loss": 1.0964,
       "step": 1407
     },
     {
       "epoch": 1.359073359073359,
-      "grad_norm": 5.352104187011719,
+      "grad_norm": 4.7396345138549805,
       "learning_rate": 1.1998161982409724e-05,
-      "loss": 1.2281,
+      "loss": 1.1932,
       "step": 1408
     },
     {
       "epoch": 1.3600386100386102,
-      "grad_norm": 5.339706897735596,
+      "grad_norm": 4.986920356750488,
       "learning_rate": 1.1987947768727773e-05,
-      "loss": 1.1884,
+      "loss": 1.1442,
       "step": 1409
     },
     {
       "epoch": 1.361003861003861,
-      "grad_norm": 4.254153728485107,
+      "grad_norm": 5.392637729644775,
       "learning_rate": 1.1977731395222731e-05,
-      "loss": 1.2148,
+      "loss": 1.1662,
       "step": 1410
     },
     {
       "epoch": 1.361969111969112,
-      "grad_norm": 4.582601547241211,
+      "grad_norm": 4.2345967292785645,
       "learning_rate": 1.1967512872994266e-05,
-      "loss": 1.198,
+      "loss": 1.1664,
       "step": 1411
     },
     {
       "epoch": 1.3629343629343629,
-      "grad_norm": 4.7897844314575195,
+      "grad_norm": 4.198776721954346,
       "learning_rate": 1.1957292213144386e-05,
-      "loss": 1.1731,
+      "loss": 1.1319,
       "step": 1412
     },
     {
       "epoch": 1.3638996138996138,
-      "grad_norm": 4.098704814910889,
+      "grad_norm": 5.643971920013428,
       "learning_rate": 1.1947069426777404e-05,
-      "loss": 1.1872,
+      "loss": 1.1608,
       "step": 1413
     },
     {
       "epoch": 1.364864864864865,
-      "grad_norm": 5.669726371765137,
+      "grad_norm": 4.070546627044678,
       "learning_rate": 1.1936844524999966e-05,
-      "loss": 1.1926,
+      "loss": 1.1422,
       "step": 1414
     },
     {
       "epoch": 1.3658301158301158,
-      "grad_norm": 4.694494247436523,
+      "grad_norm": 4.592640399932861,
       "learning_rate": 1.1926617518921e-05,
-      "loss": 1.189,
+      "loss": 1.1489,
       "step": 1415
     },
     {
       "epoch": 1.3667953667953667,
-      "grad_norm": 4.791158199310303,
+      "grad_norm": 4.745957851409912,
       "learning_rate": 1.1916388419651725e-05,
-      "loss": 1.179,
+      "loss": 1.1334,
       "step": 1416
     },
     {
       "epoch": 1.3677606177606179,
-      "grad_norm": 5.525454521179199,
+      "grad_norm": 4.4757981300354,
       "learning_rate": 1.1906157238305635e-05,
-      "loss": 1.2432,
+      "loss": 1.2058,
       "step": 1417
     },
     {
       "epoch": 1.3687258687258688,
-      "grad_norm": 4.639889717102051,
+      "grad_norm": 4.9787068367004395,
       "learning_rate": 1.189592398599849e-05,
-      "loss": 1.189,
+      "loss": 1.1481,
       "step": 1418
     },
     {
       "epoch": 1.3696911196911197,
-      "grad_norm": 5.080594539642334,
+      "grad_norm": 4.848492622375488,
       "learning_rate": 1.1885688673848289e-05,
-      "loss": 1.1942,
+      "loss": 1.1541,
       "step": 1419
     },
     {
       "epoch": 1.3706563706563706,
-      "grad_norm": 4.258522033691406,
+      "grad_norm": 4.776654243469238,
       "learning_rate": 1.1875451312975278e-05,
-      "loss": 1.2235,
+      "loss": 1.1835,
       "step": 1420
     },
     {
       "epoch": 1.3716216216216215,
-      "grad_norm": 4.478529453277588,
+      "grad_norm": 4.728508472442627,
       "learning_rate": 1.1865211914501926e-05,
-      "loss": 1.1379,
+      "loss": 1.0935,
       "step": 1421
     },
     {
       "epoch": 1.3725868725868726,
-      "grad_norm": 4.141099452972412,
+      "grad_norm": 4.560262203216553,
       "learning_rate": 1.1854970489552924e-05,
-      "loss": 1.1214,
+      "loss": 1.0671,
       "step": 1422
     },
     {
       "epoch": 1.3735521235521235,
-      "grad_norm": 4.646026611328125,
+      "grad_norm": 5.6196818351745605,
       "learning_rate": 1.1844727049255143e-05,
-      "loss": 1.229,
+      "loss": 1.1855,
       "step": 1423
     },
     {
       "epoch": 1.3745173745173744,
-      "grad_norm": 4.416215419769287,
+      "grad_norm": 4.482076644897461,
       "learning_rate": 1.1834481604737671e-05,
-      "loss": 1.1703,
+      "loss": 1.1244,
       "step": 1424
     },
     {
       "epoch": 1.3754826254826256,
-      "grad_norm": 4.077454090118408,
+      "grad_norm": 4.705634593963623,
       "learning_rate": 1.1824234167131748e-05,
-      "loss": 1.1772,
+      "loss": 1.1394,
       "step": 1425
     },
     {
       "epoch": 1.3764478764478765,
-      "grad_norm": 4.6979660987854,
+      "grad_norm": 5.486203193664551,
       "learning_rate": 1.1813984747570802e-05,
-      "loss": 1.2484,
+      "loss": 1.2025,
       "step": 1426
     },
     {
       "epoch": 1.3774131274131274,
-      "grad_norm": 4.771971225738525,
+      "grad_norm": 4.540385723114014,
       "learning_rate": 1.1803733357190396e-05,
-      "loss": 1.1678,
+      "loss": 1.1225,
       "step": 1427
     },
     {
       "epoch": 1.3783783783783785,
-      "grad_norm": 4.0661420822143555,
+      "grad_norm": 4.256995677947998,
       "learning_rate": 1.1793480007128248e-05,
-      "loss": 1.2631,
+      "loss": 1.2168,
       "step": 1428
     },
     {
       "epoch": 1.3793436293436294,
-      "grad_norm": 4.452084541320801,
+      "grad_norm": 4.8880133628845215,
       "learning_rate": 1.1783224708524192e-05,
-      "loss": 1.2458,
+      "loss": 1.1951,
       "step": 1429
     },
     {
       "epoch": 1.3803088803088803,
-      "grad_norm": 4.443688869476318,
+      "grad_norm": 4.52516508102417,
       "learning_rate": 1.1772967472520196e-05,
-      "loss": 1.2118,
+      "loss": 1.1726,
       "step": 1430
     },
     {
       "epoch": 1.3812741312741312,
-      "grad_norm": 4.165465354919434,
+      "grad_norm": 4.318268775939941,
       "learning_rate": 1.1762708310260312e-05,
-      "loss": 1.1932,
+      "loss": 1.1427,
       "step": 1431
     },
     {
       "epoch": 1.3822393822393821,
-      "grad_norm": 4.817536354064941,
+      "grad_norm": 5.500685691833496,
       "learning_rate": 1.1752447232890702e-05,
-      "loss": 1.1686,
+      "loss": 1.1319,
       "step": 1432
     },
     {
       "epoch": 1.3832046332046333,
-      "grad_norm": 4.203273296356201,
+      "grad_norm": 4.424462795257568,
       "learning_rate": 1.1742184251559602e-05,
-      "loss": 1.2121,
+      "loss": 1.171,
       "step": 1433
     },
     {
       "epoch": 1.3841698841698842,
-      "grad_norm": 4.254453182220459,
+      "grad_norm": 4.141973495483398,
       "learning_rate": 1.1731919377417318e-05,
-      "loss": 1.2411,
+      "loss": 1.1885,
       "step": 1434
     },
     {
       "epoch": 1.385135135135135,
-      "grad_norm": 3.992889404296875,
+      "grad_norm": 4.2509002685546875,
       "learning_rate": 1.1721652621616208e-05,
-      "loss": 1.1499,
+      "loss": 1.1136,
       "step": 1435
     },
     {
       "epoch": 1.3861003861003862,
-      "grad_norm": 4.445054054260254,
+      "grad_norm": 4.857095241546631,
       "learning_rate": 1.171138399531068e-05,
-      "loss": 1.2199,
+      "loss": 1.1738,
       "step": 1436
     },
     {
       "epoch": 1.3870656370656371,
-      "grad_norm": 4.198911666870117,
+      "grad_norm": 4.440127849578857,
       "learning_rate": 1.1701113509657174e-05,
-      "loss": 1.1976,
+      "loss": 1.1532,
       "step": 1437
     },
     {
       "epoch": 1.388030888030888,
-      "grad_norm": 4.214139938354492,
+      "grad_norm": 4.397274971008301,
       "learning_rate": 1.1690841175814145e-05,
-      "loss": 1.2287,
+      "loss": 1.1855,
       "step": 1438
     },
     {
       "epoch": 1.388996138996139,
-      "grad_norm": 4.2203369140625,
+      "grad_norm": 4.500129222869873,
       "learning_rate": 1.1680567004942062e-05,
-      "loss": 1.2446,
+      "loss": 1.197,
       "step": 1439
     },
     {
       "epoch": 1.3899613899613898,
-      "grad_norm": 4.274831295013428,
+      "grad_norm": 4.71156120300293,
       "learning_rate": 1.1670291008203383e-05,
-      "loss": 1.2148,
+      "loss": 1.1678,
       "step": 1440
     },
     {
       "epoch": 1.390926640926641,
-      "grad_norm": 4.5262675285339355,
+      "grad_norm": 4.279789924621582,
       "learning_rate": 1.1660013196762556e-05,
-      "loss": 1.1929,
+      "loss": 1.1427,
       "step": 1441
     },
     {
       "epoch": 1.3918918918918919,
-      "grad_norm": 4.146603584289551,
+      "grad_norm": 4.351563930511475,
       "learning_rate": 1.1649733581786001e-05,
-      "loss": 1.2134,
+      "loss": 1.1703,
       "step": 1442
     },
     {
       "epoch": 1.3928571428571428,
-      "grad_norm": 4.600409984588623,
+      "grad_norm": 4.700047969818115,
       "learning_rate": 1.163945217444209e-05,
-      "loss": 1.1481,
+      "loss": 1.1021,
       "step": 1443
     },
     {
       "epoch": 1.393822393822394,
-      "grad_norm": 5.549932956695557,
+      "grad_norm": 5.407989501953125,
       "learning_rate": 1.1629168985901149e-05,
-      "loss": 1.1883,
+      "loss": 1.1446,
       "step": 1444
     },
     {
       "epoch": 1.3947876447876448,
-      "grad_norm": 4.512418746948242,
+      "grad_norm": 4.365533351898193,
       "learning_rate": 1.1618884027335438e-05,
-      "loss": 1.2328,
+      "loss": 1.1758,
       "step": 1445
     },
     {
       "epoch": 1.3957528957528957,
-      "grad_norm": 4.620266914367676,
+      "grad_norm": 4.361486434936523,
       "learning_rate": 1.1608597309919141e-05,
-      "loss": 1.2499,
+      "loss": 1.1973,
       "step": 1446
     },
     {
       "epoch": 1.3967181467181466,
-      "grad_norm": 4.4497528076171875,
+      "grad_norm": 5.271634101867676,
       "learning_rate": 1.1598308844828348e-05,
-      "loss": 1.2454,
+      "loss": 1.2034,
       "step": 1447
     },
     {
       "epoch": 1.3976833976833976,
-      "grad_norm": 4.136813163757324,
+      "grad_norm": 4.105298042297363,
       "learning_rate": 1.1588018643241053e-05,
-      "loss": 1.2569,
+      "loss": 1.2013,
       "step": 1448
     },
     {
       "epoch": 1.3986486486486487,
-      "grad_norm": 5.220927715301514,
+      "grad_norm": 4.918732166290283,
       "learning_rate": 1.1577726716337134e-05,
-      "loss": 1.2982,
+      "loss": 1.2425,
       "step": 1449
     },
     {
       "epoch": 1.3996138996138996,
-      "grad_norm": 4.751706123352051,
+      "grad_norm": 5.4311652183532715,
       "learning_rate": 1.1567433075298341e-05,
-      "loss": 1.1566,
+      "loss": 1.1131,
       "step": 1450
     },
     {
       "epoch": 1.4005791505791505,
-      "grad_norm": 5.164363384246826,
+      "grad_norm": 3.812173366546631,
       "learning_rate": 1.1557137731308302e-05,
-      "loss": 1.1564,
+      "loss": 1.0982,
       "step": 1451
     },
     {
       "epoch": 1.4015444015444016,
-      "grad_norm": 4.481420040130615,
+      "grad_norm": 4.50590181350708,
       "learning_rate": 1.1546840695552465e-05,
-      "loss": 1.2552,
+      "loss": 1.2075,
       "step": 1452
     },
     {
       "epoch": 1.4025096525096525,
-      "grad_norm": 4.251123428344727,
+      "grad_norm": 5.012929439544678,
       "learning_rate": 1.153654197921815e-05,
-      "loss": 1.166,
+      "loss": 1.113,
       "step": 1453
     },
     {
       "epoch": 1.4034749034749034,
-      "grad_norm": 5.288816928863525,
+      "grad_norm": 4.4866204261779785,
       "learning_rate": 1.1526241593494474e-05,
-      "loss": 1.1846,
+      "loss": 1.1366,
       "step": 1454
     },
     {
       "epoch": 1.4044401544401546,
-      "grad_norm": 4.537790775299072,
+      "grad_norm": 4.030855178833008,
       "learning_rate": 1.1515939549572389e-05,
-      "loss": 1.2219,
+      "loss": 1.1616,
       "step": 1455
     },
     {
       "epoch": 1.4054054054054055,
-      "grad_norm": 4.063901901245117,
+      "grad_norm": 4.634772300720215,
       "learning_rate": 1.1505635858644635e-05,
-      "loss": 1.2029,
+      "loss": 1.1581,
       "step": 1456
     },
     {
       "epoch": 1.4063706563706564,
-      "grad_norm": 5.178050994873047,
+      "grad_norm": 4.370773792266846,
       "learning_rate": 1.1495330531905749e-05,
-      "loss": 1.1267,
+      "loss": 1.0788,
       "step": 1457
     },
     {
       "epoch": 1.4073359073359073,
-      "grad_norm": 4.601864337921143,
+      "grad_norm": 4.727386951446533,
       "learning_rate": 1.1485023580552039e-05,
-      "loss": 1.2514,
+      "loss": 1.2011,
       "step": 1458
     },
     {
       "epoch": 1.4083011583011582,
-      "grad_norm": 3.957143545150757,
+      "grad_norm": 5.378551006317139,
       "learning_rate": 1.1474715015781584e-05,
-      "loss": 1.194,
+      "loss": 1.1537,
       "step": 1459
     },
     {
       "epoch": 1.4092664092664093,
-      "grad_norm": 4.590023994445801,
+      "grad_norm": 5.11268949508667,
       "learning_rate": 1.1464404848794218e-05,
-      "loss": 1.2605,
+      "loss": 1.2253,
       "step": 1460
     },
     {
       "epoch": 1.4102316602316602,
-      "grad_norm": 4.938705921173096,
+      "grad_norm": 4.316988945007324,
       "learning_rate": 1.1454093090791503e-05,
-      "loss": 1.2199,
+      "loss": 1.1667,
       "step": 1461
     },
     {
       "epoch": 1.4111969111969112,
-      "grad_norm": 5.405674457550049,
+      "grad_norm": 6.1982927322387695,
       "learning_rate": 1.1443779752976745e-05,
-      "loss": 1.1818,
+      "loss": 1.1292,
       "step": 1462
     },
     {
       "epoch": 1.4121621621621623,
-      "grad_norm": 6.0514817237854,
+      "grad_norm": 3.923567771911621,
       "learning_rate": 1.1433464846554953e-05,
-      "loss": 1.1749,
+      "loss": 1.1184,
       "step": 1463
     },
     {
       "epoch": 1.4131274131274132,
-      "grad_norm": 6.1509904861450195,
+      "grad_norm": 4.213250637054443,
       "learning_rate": 1.1423148382732854e-05,
-      "loss": 1.2601,
+      "loss": 1.202,
       "step": 1464
     },
     {
       "epoch": 1.414092664092664,
-      "grad_norm": 6.008955001831055,
+      "grad_norm": 4.268097400665283,
       "learning_rate": 1.1412830372718851e-05,
-      "loss": 1.2095,
+      "loss": 1.1582,
       "step": 1465
     },
     {
       "epoch": 1.415057915057915,
-      "grad_norm": 7.071529388427734,
+      "grad_norm": 4.222315311431885,
       "learning_rate": 1.1402510827723043e-05,
-      "loss": 1.1539,
+      "loss": 1.0828,
       "step": 1466
     },
     {
       "epoch": 1.416023166023166,
-      "grad_norm": 4.7585954666137695,
+      "grad_norm": 4.981771945953369,
       "learning_rate": 1.1392189758957183e-05,
-      "loss": 1.1916,
+      "loss": 1.131,
       "step": 1467
     },
     {
       "epoch": 1.416988416988417,
-      "grad_norm": 5.787749290466309,
+      "grad_norm": 4.141545295715332,
       "learning_rate": 1.1381867177634693e-05,
-      "loss": 1.1798,
+      "loss": 1.1259,
       "step": 1468
     },
     {
       "epoch": 1.417953667953668,
-      "grad_norm": 4.924343109130859,
+      "grad_norm": 4.17612361907959,
       "learning_rate": 1.1371543094970624e-05,
-      "loss": 1.2515,
+      "loss": 1.1937,
       "step": 1469
     },
     {
       "epoch": 1.4189189189189189,
-      "grad_norm": 4.688542366027832,
+      "grad_norm": 4.499458312988281,
       "learning_rate": 1.1361217522181667e-05,
-      "loss": 1.1731,
+      "loss": 1.1318,
       "step": 1470
     },
     {
       "epoch": 1.41988416988417,
-      "grad_norm": 5.041671276092529,
+      "grad_norm": 4.074406623840332,
       "learning_rate": 1.135089047048613e-05,
-      "loss": 1.2632,
+      "loss": 1.2056,
       "step": 1471
     },
     {
       "epoch": 1.420849420849421,
-      "grad_norm": 4.252134799957275,
+      "grad_norm": 4.3268656730651855,
       "learning_rate": 1.134056195110393e-05,
-      "loss": 1.1982,
+      "loss": 1.1518,
       "step": 1472
     },
     {
       "epoch": 1.4218146718146718,
-      "grad_norm": 4.020081520080566,
+      "grad_norm": 4.040134906768799,
       "learning_rate": 1.1330231975256576e-05,
-      "loss": 1.2558,
+      "loss": 1.2041,
       "step": 1473
     },
     {
       "epoch": 1.4227799227799227,
-      "grad_norm": 5.2743024826049805,
+      "grad_norm": 4.151791095733643,
       "learning_rate": 1.1319900554167158e-05,
-      "loss": 1.228,
+      "loss": 1.1616,
       "step": 1474
     },
     {
       "epoch": 1.4237451737451736,
-      "grad_norm": 4.727425575256348,
+      "grad_norm": 4.448980331420898,
       "learning_rate": 1.130956769906034e-05,
-      "loss": 1.1643,
+      "loss": 1.1148,
       "step": 1475
     },
     {
       "epoch": 1.4247104247104247,
-      "grad_norm": 4.66657829284668,
+      "grad_norm": 4.366015434265137,
       "learning_rate": 1.1299233421162339e-05,
-      "loss": 1.2035,
+      "loss": 1.1542,
       "step": 1476
     },
     {
       "epoch": 1.4256756756756757,
-      "grad_norm": 4.683762550354004,
+      "grad_norm": 4.1559157371521,
       "learning_rate": 1.1288897731700922e-05,
-      "loss": 1.2453,
+      "loss": 1.1799,
       "step": 1477
     },
     {
       "epoch": 1.4266409266409266,
-      "grad_norm": 4.349846839904785,
+      "grad_norm": 4.251609802246094,
       "learning_rate": 1.1278560641905392e-05,
-      "loss": 1.2206,
+      "loss": 1.1668,
       "step": 1478
     },
     {
       "epoch": 1.4276061776061777,
-      "grad_norm": 4.372389793395996,
+      "grad_norm": 4.060056686401367,
       "learning_rate": 1.1268222163006566e-05,
-      "loss": 1.141,
+      "loss": 1.0783,
       "step": 1479
     },
     {
       "epoch": 1.4285714285714286,
-      "grad_norm": 4.824862957000732,
+      "grad_norm": 4.067476749420166,
       "learning_rate": 1.1257882306236776e-05,
-      "loss": 1.2172,
+      "loss": 1.1579,
       "step": 1480
     },
     {
       "epoch": 1.4295366795366795,
-      "grad_norm": 4.0703887939453125,
+      "grad_norm": 4.082293510437012,
       "learning_rate": 1.1247541082829847e-05,
-      "loss": 1.1931,
+      "loss": 1.1385,
       "step": 1481
     },
     {
       "epoch": 1.4305019305019304,
-      "grad_norm": 5.336752414703369,
+      "grad_norm": 4.715116024017334,
       "learning_rate": 1.1237198504021091e-05,
-      "loss": 1.2588,
+      "loss": 1.1958,
       "step": 1482
     },
     {
       "epoch": 1.4314671814671815,
-      "grad_norm": 4.761801719665527,
+      "grad_norm": 4.337517738342285,
       "learning_rate": 1.1226854581047295e-05,
-      "loss": 1.3055,
+      "loss": 1.2637,
       "step": 1483
     },
     {
       "epoch": 1.4324324324324325,
-      "grad_norm": 4.233315467834473,
+      "grad_norm": 4.6156768798828125,
       "learning_rate": 1.12165093251467e-05,
-      "loss": 1.1884,
+      "loss": 1.145,
       "step": 1484
     },
     {
       "epoch": 1.4333976833976834,
-      "grad_norm": 5.077378749847412,
+      "grad_norm": 4.192505359649658,
       "learning_rate": 1.1206162747559001e-05,
-      "loss": 1.1707,
+      "loss": 1.1149,
       "step": 1485
     },
     {
       "epoch": 1.4343629343629343,
-      "grad_norm": 4.529723644256592,
+      "grad_norm": 4.702146053314209,
       "learning_rate": 1.1195814859525332e-05,
-      "loss": 1.2067,
+      "loss": 1.1503,
       "step": 1486
     },
     {
       "epoch": 1.4353281853281854,
-      "grad_norm": 4.297308444976807,
+      "grad_norm": 4.742667198181152,
       "learning_rate": 1.1185465672288237e-05,
-      "loss": 1.1901,
+      "loss": 1.1327,
       "step": 1487
     },
     {
       "epoch": 1.4362934362934363,
-      "grad_norm": 4.214022636413574,
+      "grad_norm": 4.254252910614014,
       "learning_rate": 1.1175115197091682e-05,
-      "loss": 1.1859,
+      "loss": 1.1409,
       "step": 1488
     },
     {
       "epoch": 1.4372586872586872,
-      "grad_norm": 5.2057881355285645,
+      "grad_norm": 4.709556579589844,
       "learning_rate": 1.1164763445181034e-05,
-      "loss": 1.1926,
+      "loss": 1.1394,
       "step": 1489
     },
     {
       "epoch": 1.4382239382239383,
-      "grad_norm": 4.289044380187988,
+      "grad_norm": 5.285987854003906,
       "learning_rate": 1.1154410427803042e-05,
-      "loss": 1.1837,
+      "loss": 1.1272,
       "step": 1490
     },
     {
       "epoch": 1.4391891891891893,
-      "grad_norm": 5.051332473754883,
+      "grad_norm": 4.330855369567871,
       "learning_rate": 1.1144056156205834e-05,
-      "loss": 1.2201,
+      "loss": 1.169,
       "step": 1491
     },
     {
       "epoch": 1.4401544401544402,
-      "grad_norm": 4.821202754974365,
+      "grad_norm": 4.015356540679932,
       "learning_rate": 1.1133700641638892e-05,
-      "loss": 1.1524,
+      "loss": 1.1019,
       "step": 1492
     },
     {
       "epoch": 1.441119691119691,
-      "grad_norm": 4.25753116607666,
+      "grad_norm": 4.381031513214111,
       "learning_rate": 1.1123343895353063e-05,
-      "loss": 1.199,
+      "loss": 1.1333,
       "step": 1493
     },
     {
       "epoch": 1.442084942084942,
-      "grad_norm": 4.807949542999268,
+      "grad_norm": 4.608134746551514,
       "learning_rate": 1.1112985928600513e-05,
-      "loss": 1.1852,
+      "loss": 1.1279,
       "step": 1494
     },
     {
       "epoch": 1.443050193050193,
-      "grad_norm": 4.423308849334717,
+      "grad_norm": 4.329056739807129,
       "learning_rate": 1.1102626752634758e-05,
-      "loss": 1.2182,
+      "loss": 1.1627,
       "step": 1495
     },
     {
       "epoch": 1.444015444015444,
-      "grad_norm": 4.054159641265869,
+      "grad_norm": 4.169540882110596,
       "learning_rate": 1.1092266378710602e-05,
-      "loss": 1.2056,
+      "loss": 1.1505,
       "step": 1496
     },
     {
       "epoch": 1.444980694980695,
-      "grad_norm": 4.658977508544922,
+      "grad_norm": 4.2645463943481445,
       "learning_rate": 1.1081904818084169e-05,
-      "loss": 1.26,
+      "loss": 1.2132,
       "step": 1497
     },
     {
       "epoch": 1.445945945945946,
-      "grad_norm": 4.359392166137695,
+      "grad_norm": 4.297389030456543,
       "learning_rate": 1.1071542082012865e-05,
-      "loss": 1.2023,
+      "loss": 1.1392,
       "step": 1498
     },
     {
       "epoch": 1.446911196911197,
-      "grad_norm": 4.363769054412842,
+      "grad_norm": 4.436461448669434,
       "learning_rate": 1.106117818175538e-05,
-      "loss": 1.2128,
+      "loss": 1.1629,
       "step": 1499
     },
     {
       "epoch": 1.4478764478764479,
-      "grad_norm": 4.644874095916748,
+      "grad_norm": 4.619462013244629,
       "learning_rate": 1.1050813128571654e-05,
-      "loss": 1.2804,
+      "loss": 1.2191,
       "step": 1500
     },
     {
       "epoch": 1.4488416988416988,
-      "grad_norm": 3.9422566890716553,
+      "grad_norm": 4.060195446014404,
       "learning_rate": 1.1040446933722894e-05,
-      "loss": 1.1319,
+      "loss": 1.0934,
       "step": 1501
     },
     {
       "epoch": 1.4498069498069497,
-      "grad_norm": 4.872453689575195,
+      "grad_norm": 4.1281418800354,
       "learning_rate": 1.1030079608471544e-05,
-      "loss": 1.2062,
+      "loss": 1.1496,
       "step": 1502
     },
     {
       "epoch": 1.4507722007722008,
-      "grad_norm": 4.18463659286499,
+      "grad_norm": 4.344203472137451,
       "learning_rate": 1.1019711164081272e-05,
-      "loss": 1.223,
+      "loss": 1.1793,
       "step": 1503
     },
     {
       "epoch": 1.4517374517374517,
-      "grad_norm": 4.284641265869141,
+      "grad_norm": 4.460936546325684,
       "learning_rate": 1.1009341611816963e-05,
-      "loss": 1.1816,
+      "loss": 1.1298,
       "step": 1504
     },
     {
       "epoch": 1.4527027027027026,
-      "grad_norm": 4.588273525238037,
+      "grad_norm": 4.23732328414917,
       "learning_rate": 1.099897096294471e-05,
-      "loss": 1.2189,
+      "loss": 1.1582,
       "step": 1505
     },
     {
       "epoch": 1.4536679536679538,
-      "grad_norm": 4.396002292633057,
+      "grad_norm": 4.307660102844238,
       "learning_rate": 1.0988599228731794e-05,
-      "loss": 1.2246,
+      "loss": 1.1824,
       "step": 1506
     },
     {
       "epoch": 1.4546332046332047,
-      "grad_norm": 4.360711574554443,
+      "grad_norm": 4.70295524597168,
       "learning_rate": 1.0978226420446671e-05,
-      "loss": 1.1902,
+      "loss": 1.1395,
       "step": 1507
     },
     {
       "epoch": 1.4555984555984556,
-      "grad_norm": 4.676750183105469,
+      "grad_norm": 4.1898980140686035,
       "learning_rate": 1.0967852549358975e-05,
-      "loss": 1.237,
+      "loss": 1.1756,
       "step": 1508
     },
     {
       "epoch": 1.4565637065637065,
-      "grad_norm": 5.116564750671387,
+      "grad_norm": 4.131186008453369,
       "learning_rate": 1.0957477626739483e-05,
-      "loss": 1.1937,
+      "loss": 1.136,
       "step": 1509
     },
     {
       "epoch": 1.4575289575289574,
-      "grad_norm": 4.32762336730957,
+      "grad_norm": 4.482240676879883,
       "learning_rate": 1.0947101663860123e-05,
-      "loss": 1.1836,
+      "loss": 1.1314,
       "step": 1510
     },
     {
       "epoch": 1.4584942084942085,
-      "grad_norm": 4.377365589141846,
+      "grad_norm": 4.456892967224121,
       "learning_rate": 1.0936724671993948e-05,
-      "loss": 1.1516,
+      "loss": 1.0902,
       "step": 1511
     },
     {
       "epoch": 1.4594594594594594,
-      "grad_norm": 4.5395002365112305,
+      "grad_norm": 4.1208176612854,
       "learning_rate": 1.092634666241513e-05,
-      "loss": 1.1631,
+      "loss": 1.1102,
       "step": 1512
     },
     {
       "epoch": 1.4604247104247103,
-      "grad_norm": 4.971916198730469,
+      "grad_norm": 4.820292949676514,
       "learning_rate": 1.091596764639895e-05,
-      "loss": 1.2129,
+      "loss": 1.1553,
       "step": 1513
     },
     {
       "epoch": 1.4613899613899615,
-      "grad_norm": 4.18382453918457,
+      "grad_norm": 4.327723979949951,
       "learning_rate": 1.0905587635221782e-05,
-      "loss": 1.1798,
+      "loss": 1.1335,
       "step": 1514
     },
     {
       "epoch": 1.4623552123552124,
-      "grad_norm": 4.545231342315674,
+      "grad_norm": 4.346757888793945,
       "learning_rate": 1.089520664016107e-05,
-      "loss": 1.1676,
+      "loss": 1.1169,
       "step": 1515
     },
     {
       "epoch": 1.4633204633204633,
-      "grad_norm": 4.12399435043335,
+      "grad_norm": 4.473211288452148,
       "learning_rate": 1.0884824672495346e-05,
-      "loss": 1.194,
+      "loss": 1.1424,
       "step": 1516
     },
     {
       "epoch": 1.4642857142857144,
-      "grad_norm": 5.3278489112854,
+      "grad_norm": 4.884164810180664,
       "learning_rate": 1.0874441743504182e-05,
-      "loss": 1.2275,
+      "loss": 1.1589,
       "step": 1517
     },
     {
       "epoch": 1.4652509652509653,
-      "grad_norm": 4.356189727783203,
+      "grad_norm": 4.29329252243042,
       "learning_rate": 1.086405786446821e-05,
-      "loss": 1.1639,
+      "loss": 1.101,
       "step": 1518
     },
     {
       "epoch": 1.4662162162162162,
-      "grad_norm": 4.2782793045043945,
+      "grad_norm": 4.260406017303467,
       "learning_rate": 1.0853673046669071e-05,
-      "loss": 1.2363,
+      "loss": 1.1651,
       "step": 1519
     },
     {
       "epoch": 1.4671814671814671,
-      "grad_norm": 4.346798896789551,
+      "grad_norm": 4.4287333488464355,
       "learning_rate": 1.0843287301389455e-05,
-      "loss": 1.2136,
+      "loss": 1.1504,
       "step": 1520
     },
     {
       "epoch": 1.468146718146718,
-      "grad_norm": 4.449827194213867,
+      "grad_norm": 4.118264675140381,
       "learning_rate": 1.0832900639913035e-05,
-      "loss": 1.2552,
+      "loss": 1.1744,
       "step": 1521
     },
     {
       "epoch": 1.4691119691119692,
-      "grad_norm": 4.275735855102539,
+      "grad_norm": 4.0781073570251465,
       "learning_rate": 1.0822513073524498e-05,
-      "loss": 1.1877,
+      "loss": 1.1353,
       "step": 1522
     },
     {
       "epoch": 1.47007722007722,
-      "grad_norm": 5.334815502166748,
+      "grad_norm": 4.914519309997559,
       "learning_rate": 1.0812124613509496e-05,
-      "loss": 1.2357,
+      "loss": 1.1746,
       "step": 1523
     },
     {
       "epoch": 1.471042471042471,
-      "grad_norm": 5.0692853927612305,
+      "grad_norm": 4.1767778396606445,
       "learning_rate": 1.080173527115467e-05,
-      "loss": 1.2511,
+      "loss": 1.191,
       "step": 1524
     },
     {
       "epoch": 1.4720077220077221,
-      "grad_norm": 5.489389896392822,
+      "grad_norm": 4.390262126922607,
       "learning_rate": 1.0791345057747604e-05,
-      "loss": 1.2368,
+      "loss": 1.1703,
       "step": 1525
     },
     {
       "epoch": 1.472972972972973,
-      "grad_norm": 4.0504469871521,
+      "grad_norm": 4.194972991943359,
       "learning_rate": 1.0780953984576841e-05,
-      "loss": 1.146,
+      "loss": 1.0877,
       "step": 1526
     },
     {
       "epoch": 1.473938223938224,
-      "grad_norm": 4.052753448486328,
+      "grad_norm": 4.326572418212891,
       "learning_rate": 1.0770562062931849e-05,
-      "loss": 1.1287,
+      "loss": 1.078,
       "step": 1527
     },
     {
       "epoch": 1.4749034749034748,
-      "grad_norm": 5.282724380493164,
+      "grad_norm": 4.477307319641113,
       "learning_rate": 1.0760169304103023e-05,
-      "loss": 1.2229,
+      "loss": 1.1696,
       "step": 1528
     },
     {
       "epoch": 1.4758687258687258,
-      "grad_norm": 4.906439304351807,
+      "grad_norm": 5.135695457458496,
       "learning_rate": 1.0749775719381665e-05,
-      "loss": 1.2455,
+      "loss": 1.2008,
       "step": 1529
     },
     {
       "epoch": 1.4768339768339769,
-      "grad_norm": 5.4240031242370605,
+      "grad_norm": 4.345811367034912,
       "learning_rate": 1.0739381320059978e-05,
-      "loss": 1.1586,
+      "loss": 1.0852,
       "step": 1530
     },
     {
       "epoch": 1.4777992277992278,
-      "grad_norm": 4.4138875007629395,
+      "grad_norm": 4.496689796447754,
       "learning_rate": 1.0728986117431044e-05,
-      "loss": 1.2136,
+      "loss": 1.1606,
       "step": 1531
     },
     {
       "epoch": 1.4787644787644787,
-      "grad_norm": 4.2003350257873535,
+      "grad_norm": 4.814667224884033,
       "learning_rate": 1.0718590122788822e-05,
-      "loss": 1.2395,
+      "loss": 1.1827,
       "step": 1532
     },
     {
       "epoch": 1.4797297297297298,
-      "grad_norm": 4.415201187133789,
+      "grad_norm": 3.9193756580352783,
       "learning_rate": 1.0708193347428128e-05,
-      "loss": 1.2035,
+      "loss": 1.1519,
       "step": 1533
     },
     {
       "epoch": 1.4806949806949807,
-      "grad_norm": 3.981343984603882,
+      "grad_norm": 3.9438159465789795,
       "learning_rate": 1.0697795802644635e-05,
-      "loss": 1.1869,
+      "loss": 1.1312,
       "step": 1534
     },
     {
       "epoch": 1.4816602316602316,
-      "grad_norm": 4.455201148986816,
+      "grad_norm": 4.318869113922119,
       "learning_rate": 1.0687397499734842e-05,
-      "loss": 1.1693,
+      "loss": 1.0999,
       "step": 1535
     },
     {
       "epoch": 1.4826254826254825,
-      "grad_norm": 4.045246124267578,
+      "grad_norm": 4.359642028808594,
       "learning_rate": 1.0676998449996075e-05,
-      "loss": 1.2343,
+      "loss": 1.1735,
       "step": 1536
     },
     {
       "epoch": 1.4835907335907335,
-      "grad_norm": 4.094812393188477,
+      "grad_norm": 4.17499303817749,
       "learning_rate": 1.0666598664726471e-05,
-      "loss": 1.2074,
+      "loss": 1.1493,
       "step": 1537
     },
     {
       "epoch": 1.4845559845559846,
-      "grad_norm": 5.180531978607178,
+      "grad_norm": 4.2197585105896,
       "learning_rate": 1.0656198155224968e-05,
-      "loss": 1.2136,
+      "loss": 1.1459,
       "step": 1538
     },
     {
       "epoch": 1.4855212355212355,
-      "grad_norm": 4.258422374725342,
+      "grad_norm": 4.487946033477783,
       "learning_rate": 1.064579693279129e-05,
-      "loss": 1.1904,
+      "loss": 1.1204,
       "step": 1539
     },
     {
       "epoch": 1.4864864864864864,
-      "grad_norm": 4.394831657409668,
+      "grad_norm": 4.359161853790283,
       "learning_rate": 1.0635395008725936e-05,
-      "loss": 1.2433,
+      "loss": 1.1997,
       "step": 1540
     },
     {
       "epoch": 1.4874517374517375,
-      "grad_norm": 4.647140979766846,
+      "grad_norm": 4.590597152709961,
       "learning_rate": 1.0624992394330165e-05,
-      "loss": 1.1996,
+      "loss": 1.1329,
       "step": 1541
     },
     {
       "epoch": 1.4884169884169884,
-      "grad_norm": 4.8291521072387695,
+      "grad_norm": 4.68887996673584,
       "learning_rate": 1.061458910090599e-05,
-      "loss": 1.2498,
+      "loss": 1.1885,
       "step": 1542
     },
     {
       "epoch": 1.4893822393822393,
-      "grad_norm": 4.660562038421631,
+      "grad_norm": 4.476648330688477,
       "learning_rate": 1.0604185139756155e-05,
-      "loss": 1.2261,
+      "loss": 1.1573,
       "step": 1543
     },
     {
       "epoch": 1.4903474903474905,
-      "grad_norm": 4.097156524658203,
+      "grad_norm": 4.049508571624756,
       "learning_rate": 1.0593780522184139e-05,
-      "loss": 1.2191,
+      "loss": 1.1564,
       "step": 1544
     },
     {
       "epoch": 1.4913127413127414,
-      "grad_norm": 4.38627815246582,
+      "grad_norm": 4.354706764221191,
       "learning_rate": 1.0583375259494125e-05,
-      "loss": 1.2229,
+      "loss": 1.1503,
       "step": 1545
     },
     {
       "epoch": 1.4922779922779923,
-      "grad_norm": 4.089046001434326,
+      "grad_norm": 4.169605255126953,
       "learning_rate": 1.0572969362991e-05,
-      "loss": 1.2135,
+      "loss": 1.1406,
       "step": 1546
     },
     {
       "epoch": 1.4932432432432432,
-      "grad_norm": 4.13787317276001,
+      "grad_norm": 4.14640474319458,
       "learning_rate": 1.0562562843980346e-05,
-      "loss": 1.1843,
+      "loss": 1.1195,
       "step": 1547
     },
     {
       "epoch": 1.494208494208494,
-      "grad_norm": 4.3745832443237305,
+      "grad_norm": 4.165933609008789,
       "learning_rate": 1.0552155713768409e-05,
-      "loss": 1.2013,
+      "loss": 1.1366,
       "step": 1548
     },
     {
       "epoch": 1.4951737451737452,
-      "grad_norm": 4.971855640411377,
+      "grad_norm": 5.470155239105225,
       "learning_rate": 1.0541747983662114e-05,
-      "loss": 1.162,
+      "loss": 1.0978,
       "step": 1549
     },
     {
       "epoch": 1.4961389961389961,
-      "grad_norm": 4.968060493469238,
+      "grad_norm": 4.212482452392578,
       "learning_rate": 1.0531339664969021e-05,
-      "loss": 1.2469,
+      "loss": 1.1941,
       "step": 1550
     },
     {
       "epoch": 1.497104247104247,
-      "grad_norm": 4.692889213562012,
+      "grad_norm": 4.2750043869018555,
       "learning_rate": 1.0520930768997347e-05,
-      "loss": 1.23,
+      "loss": 1.1776,
       "step": 1551
     },
     {
       "epoch": 1.4980694980694982,
-      "grad_norm": 4.44988489151001,
+      "grad_norm": 4.54524564743042,
       "learning_rate": 1.0510521307055914e-05,
-      "loss": 1.1305,
+      "loss": 1.0655,
       "step": 1552
     },
     {
       "epoch": 1.499034749034749,
-      "grad_norm": 4.482110023498535,
+      "grad_norm": 4.774360179901123,
       "learning_rate": 1.0500111290454189e-05,
-      "loss": 1.1734,
+      "loss": 1.1104,
       "step": 1553
     },
     {
       "epoch": 1.5,
-      "grad_norm": 4.157884120941162,
+      "grad_norm": 4.129669189453125,
       "learning_rate": 1.0489700730502212e-05,
-      "loss": 1.198,
+      "loss": 1.158,
       "step": 1554
     },
     {
       "epoch": 1.5009652509652511,
-      "grad_norm": 4.152207851409912,
+      "grad_norm": 4.169137477874756,
       "learning_rate": 1.047928963851063e-05,
-      "loss": 1.205,
+      "loss": 1.1389,
       "step": 1555
     },
     {
       "epoch": 1.5019305019305018,
-      "grad_norm": 4.259824275970459,
+      "grad_norm": 5.092118263244629,
       "learning_rate": 1.046887802579067e-05,
-      "loss": 1.2947,
+      "loss": 1.2401,
       "step": 1556
     },
     {
       "epoch": 1.502895752895753,
-      "grad_norm": 4.260583877563477,
+      "grad_norm": 4.930356025695801,
       "learning_rate": 1.0458465903654107e-05,
-      "loss": 1.1686,
+      "loss": 1.1127,
       "step": 1557
     },
     {
       "epoch": 1.5038610038610039,
-      "grad_norm": 4.16087007522583,
+      "grad_norm": 4.276647567749023,
       "learning_rate": 1.0448053283413293e-05,
-      "loss": 1.1988,
+      "loss": 1.1405,
       "step": 1558
     },
     {
       "epoch": 1.5048262548262548,
-      "grad_norm": 3.9509360790252686,
+      "grad_norm": 4.032553195953369,
       "learning_rate": 1.0437640176381099e-05,
-      "loss": 1.1873,
+      "loss": 1.1258,
       "step": 1559
     },
     {
       "epoch": 1.505791505791506,
-      "grad_norm": 4.118860244750977,
+      "grad_norm": 5.094645023345947,
       "learning_rate": 1.0427226593870947e-05,
-      "loss": 1.1767,
+      "loss": 1.1206,
       "step": 1560
     },
     {
       "epoch": 1.5067567567567568,
-      "grad_norm": 4.576681137084961,
+      "grad_norm": 5.243322849273682,
       "learning_rate": 1.0416812547196753e-05,
-      "loss": 1.1545,
+      "loss": 1.0751,
       "step": 1561
     },
     {
       "epoch": 1.5077220077220077,
-      "grad_norm": 3.929964542388916,
+      "grad_norm": 4.005675792694092,
       "learning_rate": 1.0406398047672956e-05,
-      "loss": 1.209,
+      "loss": 1.1567,
       "step": 1562
     },
     {
       "epoch": 1.5086872586872588,
-      "grad_norm": 4.916624069213867,
+      "grad_norm": 5.26769495010376,
       "learning_rate": 1.0395983106614474e-05,
-      "loss": 1.1555,
+      "loss": 1.0985,
       "step": 1563
     },
     {
       "epoch": 1.5096525096525095,
-      "grad_norm": 4.800134181976318,
+      "grad_norm": 5.525345802307129,
       "learning_rate": 1.0385567735336712e-05,
-      "loss": 1.2684,
+      "loss": 1.2066,
       "step": 1564
     },
     {
       "epoch": 1.5106177606177607,
-      "grad_norm": 5.12754487991333,
+      "grad_norm": 4.7366180419921875,
       "learning_rate": 1.0375151945155544e-05,
-      "loss": 1.1961,
+      "loss": 1.1338,
       "step": 1565
     },
     {
       "epoch": 1.5115830115830116,
-      "grad_norm": 4.375315189361572,
+      "grad_norm": 4.658515930175781,
       "learning_rate": 1.0364735747387292e-05,
-      "loss": 1.2706,
+      "loss": 1.2021,
       "step": 1566
     },
     {
       "epoch": 1.5125482625482625,
-      "grad_norm": 3.993865489959717,
+      "grad_norm": 4.9155073165893555,
       "learning_rate": 1.0354319153348724e-05,
-      "loss": 1.2037,
+      "loss": 1.1551,
       "step": 1567
     },
     {
       "epoch": 1.5135135135135136,
-      "grad_norm": 4.187656402587891,
+      "grad_norm": 5.192066669464111,
       "learning_rate": 1.034390217435704e-05,
-      "loss": 1.2099,
+      "loss": 1.1619,
       "step": 1568
     },
     {
       "epoch": 1.5144787644787645,
-      "grad_norm": 4.1158447265625,
+      "grad_norm": 4.221813201904297,
       "learning_rate": 1.0333484821729857e-05,
-      "loss": 1.1932,
+      "loss": 1.1284,
       "step": 1569
     },
     {
       "epoch": 1.5154440154440154,
-      "grad_norm": 4.137932300567627,
+      "grad_norm": 4.860891819000244,
       "learning_rate": 1.0323067106785199e-05,
-      "loss": 1.2271,
+      "loss": 1.1678,
       "step": 1570
     },
     {
       "epoch": 1.5164092664092665,
-      "grad_norm": 3.928439140319824,
+      "grad_norm": 5.030066013336182,
       "learning_rate": 1.031264904084148e-05,
-      "loss": 1.1639,
+      "loss": 1.1105,
       "step": 1571
     },
     {
       "epoch": 1.5173745173745172,
-      "grad_norm": 4.347318172454834,
+      "grad_norm": 4.641828536987305,
       "learning_rate": 1.03022306352175e-05,
-      "loss": 1.2631,
+      "loss": 1.1946,
       "step": 1572
     },
     {
       "epoch": 1.5183397683397684,
-      "grad_norm": 4.589171886444092,
+      "grad_norm": 4.662938594818115,
       "learning_rate": 1.0291811901232428e-05,
-      "loss": 1.2417,
+      "loss": 1.1627,
       "step": 1573
     },
     {
       "epoch": 1.5193050193050193,
-      "grad_norm": 4.574836254119873,
+      "grad_norm": 5.428515434265137,
       "learning_rate": 1.0281392850205783e-05,
-      "loss": 1.1509,
+      "loss": 1.0804,
       "step": 1574
     },
     {
       "epoch": 1.5202702702702702,
-      "grad_norm": 4.554994106292725,
+      "grad_norm": 4.41550874710083,
       "learning_rate": 1.0270973493457437e-05,
-      "loss": 1.2337,
+      "loss": 1.1668,
       "step": 1575
     },
     {
       "epoch": 1.5212355212355213,
-      "grad_norm": 4.198336124420166,
+      "grad_norm": 4.422458648681641,
       "learning_rate": 1.026055384230759e-05,
-      "loss": 1.2079,
+      "loss": 1.1462,
       "step": 1576
     },
     {
       "epoch": 1.5222007722007722,
-      "grad_norm": 4.500290870666504,
+      "grad_norm": 4.613282203674316,
       "learning_rate": 1.025013390807676e-05,
-      "loss": 1.1258,
+      "loss": 1.0616,
       "step": 1577
     },
     {
       "epoch": 1.5231660231660231,
-      "grad_norm": 4.623536586761475,
+      "grad_norm": 5.357022285461426,
       "learning_rate": 1.0239713702085781e-05,
-      "loss": 1.2186,
+      "loss": 1.1619,
       "step": 1578
     },
     {
       "epoch": 1.5241312741312742,
-      "grad_norm": 4.773224830627441,
+      "grad_norm": 5.389514923095703,
       "learning_rate": 1.0229293235655768e-05,
-      "loss": 1.1922,
+      "loss": 1.1371,
       "step": 1579
     },
     {
       "epoch": 1.525096525096525,
-      "grad_norm": 3.701310396194458,
+      "grad_norm": 4.0096564292907715,
       "learning_rate": 1.0218872520108135e-05,
-      "loss": 1.2281,
+      "loss": 1.1721,
       "step": 1580
     },
     {
       "epoch": 1.526061776061776,
-      "grad_norm": 4.196020603179932,
+      "grad_norm": 5.416517734527588,
       "learning_rate": 1.020845156676455e-05,
-      "loss": 1.1747,
+      "loss": 1.1118,
       "step": 1581
     },
     {
       "epoch": 1.527027027027027,
-      "grad_norm": 4.553520202636719,
+      "grad_norm": 5.518841743469238,
       "learning_rate": 1.0198030386946956e-05,
-      "loss": 1.3491,
+      "loss": 1.2844,
       "step": 1582
     },
     {
       "epoch": 1.5279922779922779,
-      "grad_norm": 4.003087997436523,
+      "grad_norm": 3.9605185985565186,
       "learning_rate": 1.0187608991977531e-05,
-      "loss": 1.2327,
+      "loss": 1.1685,
       "step": 1583
     },
     {
       "epoch": 1.528957528957529,
-      "grad_norm": 5.1173882484436035,
+      "grad_norm": 5.4257121086120605,
       "learning_rate": 1.0177187393178687e-05,
-      "loss": 1.178,
+      "loss": 1.1073,
       "step": 1584
     },
     {
       "epoch": 1.52992277992278,
-      "grad_norm": 4.88187313079834,
+      "grad_norm": 4.151060581207275,
       "learning_rate": 1.0166765601873067e-05,
-      "loss": 1.1399,
+      "loss": 1.0851,
       "step": 1585
     },
     {
       "epoch": 1.5308880308880308,
-      "grad_norm": 4.126231670379639,
+      "grad_norm": 4.885241985321045,
       "learning_rate": 1.0156343629383509e-05,
-      "loss": 1.2101,
+      "loss": 1.1599,
       "step": 1586
     },
     {
       "epoch": 1.531853281853282,
-      "grad_norm": 4.237364292144775,
+      "grad_norm": 4.206613540649414,
       "learning_rate": 1.0145921487033065e-05,
-      "loss": 1.2188,
+      "loss": 1.1528,
       "step": 1587
     },
     {
       "epoch": 1.5328185328185329,
-      "grad_norm": 3.97042179107666,
+      "grad_norm": 5.2064666748046875,
       "learning_rate": 1.0135499186144952e-05,
-      "loss": 1.1823,
+      "loss": 1.117,
       "step": 1588
     },
     {
       "epoch": 1.5337837837837838,
-      "grad_norm": 5.60618257522583,
+      "grad_norm": 4.374683380126953,
       "learning_rate": 1.0125076738042577e-05,
-      "loss": 1.2448,
+      "loss": 1.1766,
       "step": 1589
     },
     {
       "epoch": 1.534749034749035,
-      "grad_norm": 4.790480136871338,
+      "grad_norm": 4.799027442932129,
       "learning_rate": 1.011465415404949e-05,
-      "loss": 1.1589,
+      "loss": 1.1046,
       "step": 1590
     },
     {
       "epoch": 1.5357142857142856,
-      "grad_norm": 4.756139278411865,
+      "grad_norm": 4.340676784515381,
       "learning_rate": 1.010423144548941e-05,
-      "loss": 1.1934,
+      "loss": 1.117,
       "step": 1591
     },
     {
       "epoch": 1.5366795366795367,
-      "grad_norm": 5.73478889465332,
+      "grad_norm": 4.433998107910156,
       "learning_rate": 1.0093808623686165e-05,
-      "loss": 1.1871,
+      "loss": 1.1125,
       "step": 1592
     },
     {
       "epoch": 1.5376447876447876,
-      "grad_norm": 3.9583542346954346,
+      "grad_norm": 4.142754077911377,
       "learning_rate": 1.0083385699963733e-05,
-      "loss": 1.2149,
+      "loss": 1.1431,
       "step": 1593
     },
     {
       "epoch": 1.5386100386100385,
-      "grad_norm": 4.312107086181641,
+      "grad_norm": 4.5632452964782715,
       "learning_rate": 1.0072962685646182e-05,
-      "loss": 1.1849,
+      "loss": 1.1245,
       "step": 1594
     },
     {
       "epoch": 1.5395752895752897,
-      "grad_norm": 5.653817653656006,
+      "grad_norm": 4.112701892852783,
       "learning_rate": 1.0062539592057689e-05,
-      "loss": 1.2231,
+      "loss": 1.1619,
       "step": 1595
     },
     {
       "epoch": 1.5405405405405406,
-      "grad_norm": 4.556292533874512,
+      "grad_norm": 4.4833221435546875,
       "learning_rate": 1.0052116430522514e-05,
-      "loss": 1.2432,
+      "loss": 1.1778,
       "step": 1596
     },
     {
       "epoch": 1.5415057915057915,
-      "grad_norm": 5.225686550140381,
+      "grad_norm": 4.250216960906982,
       "learning_rate": 1.0041693212364992e-05,
-      "loss": 1.1649,
+      "loss": 1.1025,
       "step": 1597
     },
     {
       "epoch": 1.5424710424710426,
-      "grad_norm": 5.1191606521606445,
+      "grad_norm": 4.717977046966553,
       "learning_rate": 1.0031269948909517e-05,
-      "loss": 1.1819,
+      "loss": 1.1098,
       "step": 1598
     },
     {
       "epoch": 1.5434362934362933,
-      "grad_norm": 4.36271333694458,
+      "grad_norm": 4.512975692749023,
       "learning_rate": 1.0020846651480534e-05,
-      "loss": 1.2678,
+      "loss": 1.2038,
       "step": 1599
     },
     {
       "epoch": 1.5444015444015444,
-      "grad_norm": 4.28533411026001,
+      "grad_norm": 5.88560676574707,
       "learning_rate": 1.0010423331402528e-05,
-      "loss": 1.1961,
+      "loss": 1.1409,
       "step": 1600
     },
     {
       "epoch": 1.5453667953667953,
-      "grad_norm": 6.018980503082275,
+      "grad_norm": 4.14727258682251,
       "learning_rate": 1e-05,
-      "loss": 1.187,
+      "loss": 1.1108,
       "step": 1601
     },
     {
       "epoch": 1.5463320463320462,
-      "grad_norm": 4.648509502410889,
+      "grad_norm": 5.062203407287598,
       "learning_rate": 9.989576668597476e-06,
-      "loss": 1.1882,
+      "loss": 1.1239,
       "step": 1602
     },
     {
       "epoch": 1.5472972972972974,
-      "grad_norm": 6.939119815826416,
+      "grad_norm": 4.5702385902404785,
       "learning_rate": 9.979153348519466e-06,
-      "loss": 1.2215,
+      "loss": 1.1504,
       "step": 1603
     },
     {
       "epoch": 1.5482625482625483,
-      "grad_norm": 6.063055992126465,
+      "grad_norm": 4.5437912940979,
       "learning_rate": 9.968730051090487e-06,
-      "loss": 1.2178,
+      "loss": 1.1399,
       "step": 1604
     },
     {
       "epoch": 1.5492277992277992,
-      "grad_norm": 5.186483860015869,
+      "grad_norm": 4.552403926849365,
       "learning_rate": 9.958306787635011e-06,
-      "loss": 1.3228,
+      "loss": 1.2498,
       "step": 1605
     },
     {
       "epoch": 1.5501930501930503,
-      "grad_norm": 5.201595783233643,
+      "grad_norm": 4.708113670349121,
       "learning_rate": 9.94788356947749e-06,
-      "loss": 1.1799,
+      "loss": 1.1088,
       "step": 1606
     },
     {
       "epoch": 1.551158301158301,
-      "grad_norm": 4.202732563018799,
+      "grad_norm": 4.467221736907959,
       "learning_rate": 9.937460407942312e-06,
-      "loss": 1.1754,
+      "loss": 1.1084,
       "step": 1607
     },
     {
       "epoch": 1.5521235521235521,
-      "grad_norm": 4.17235803604126,
+      "grad_norm": 4.112290859222412,
       "learning_rate": 9.927037314353821e-06,
-      "loss": 1.1695,
+      "loss": 1.093,
       "step": 1608
     },
     {
       "epoch": 1.553088803088803,
-      "grad_norm": 4.822647571563721,
+      "grad_norm": 5.706310749053955,
       "learning_rate": 9.91661430003627e-06,
-      "loss": 1.2279,
+      "loss": 1.1597,
       "step": 1609
     },
     {
       "epoch": 1.554054054054054,
-      "grad_norm": 4.211466312408447,
+      "grad_norm": 3.9913737773895264,
       "learning_rate": 9.906191376313837e-06,
-      "loss": 1.1714,
+      "loss": 1.0954,
       "step": 1610
     },
     {
       "epoch": 1.555019305019305,
-      "grad_norm": 4.621142387390137,
+      "grad_norm": 4.569804668426514,
       "learning_rate": 9.895768554510593e-06,
-      "loss": 1.1973,
+      "loss": 1.1402,
       "step": 1611
     },
     {
       "epoch": 1.555984555984556,
-      "grad_norm": 5.649501323699951,
+      "grad_norm": 4.494530200958252,
       "learning_rate": 9.88534584595051e-06,
-      "loss": 1.1692,
+      "loss": 1.0984,
       "step": 1612
     },
     {
       "epoch": 1.556949806949807,
-      "grad_norm": 4.738320350646973,
+      "grad_norm": 4.013968467712402,
       "learning_rate": 9.874923261957428e-06,
-      "loss": 1.1858,
+      "loss": 1.1037,
       "step": 1613
     },
     {
       "epoch": 1.557915057915058,
-      "grad_norm": 4.468520641326904,
+      "grad_norm": 4.380873203277588,
       "learning_rate": 9.864500813855051e-06,
-      "loss": 1.2527,
+      "loss": 1.1842,
       "step": 1614
     },
     {
       "epoch": 1.558880308880309,
-      "grad_norm": 5.1193037033081055,
+      "grad_norm": 4.219939231872559,
       "learning_rate": 9.854078512966937e-06,
-      "loss": 1.1502,
+      "loss": 1.0838,
       "step": 1615
     },
     {
       "epoch": 1.5598455598455598,
-      "grad_norm": 5.412008285522461,
+      "grad_norm": 4.27462911605835,
       "learning_rate": 9.84365637061649e-06,
-      "loss": 1.3047,
+      "loss": 1.2464,
       "step": 1616
     },
     {
       "epoch": 1.560810810810811,
-      "grad_norm": 5.719449996948242,
+      "grad_norm": 4.720950126647949,
       "learning_rate": 9.833234398126938e-06,
-      "loss": 1.2183,
+      "loss": 1.1584,
       "step": 1617
     },
     {
       "epoch": 1.5617760617760617,
-      "grad_norm": 4.692526817321777,
+      "grad_norm": 3.9838640689849854,
       "learning_rate": 9.822812606821317e-06,
-      "loss": 1.1853,
+      "loss": 1.1129,
       "step": 1618
     },
     {
       "epoch": 1.5627413127413128,
-      "grad_norm": 4.056661605834961,
+      "grad_norm": 4.235706806182861,
       "learning_rate": 9.812391008022472e-06,
-      "loss": 1.1964,
+      "loss": 1.133,
       "step": 1619
     },
     {
       "epoch": 1.5637065637065637,
-      "grad_norm": 5.884062767028809,
+      "grad_norm": 4.799947261810303,
       "learning_rate": 9.801969613053047e-06,
-      "loss": 1.1945,
+      "loss": 1.1211,
       "step": 1620
     },
     {
       "epoch": 1.5646718146718146,
-      "grad_norm": 4.1736626625061035,
+      "grad_norm": 4.2059149742126465,
       "learning_rate": 9.791548433235452e-06,
-      "loss": 1.2004,
+      "loss": 1.1336,
       "step": 1621
     },
     {
       "epoch": 1.5656370656370657,
-      "grad_norm": 4.439320087432861,
+      "grad_norm": 4.099020004272461,
       "learning_rate": 9.781127479891872e-06,
-      "loss": 1.1776,
+      "loss": 1.1021,
       "step": 1622
     },
     {
       "epoch": 1.5666023166023166,
-      "grad_norm": 4.904225826263428,
+      "grad_norm": 4.475094318389893,
       "learning_rate": 9.770706764344235e-06,
-      "loss": 1.1812,
+      "loss": 1.1041,
       "step": 1623
     },
     {
       "epoch": 1.5675675675675675,
-      "grad_norm": 4.182514667510986,
+      "grad_norm": 4.472807884216309,
       "learning_rate": 9.760286297914224e-06,
-      "loss": 1.2016,
+      "loss": 1.1385,
       "step": 1624
     },
     {
       "epoch": 1.5685328185328187,
-      "grad_norm": 4.227596282958984,
+      "grad_norm": 4.300841808319092,
       "learning_rate": 9.749866091923241e-06,
-      "loss": 1.2117,
+      "loss": 1.1443,
       "step": 1625
     },
     {
       "epoch": 1.5694980694980694,
-      "grad_norm": 4.459351062774658,
+      "grad_norm": 4.1261725425720215,
       "learning_rate": 9.739446157692416e-06,
-      "loss": 1.1541,
+      "loss": 1.0733,
       "step": 1626
     },
     {
       "epoch": 1.5704633204633205,
-      "grad_norm": 4.430898666381836,
+      "grad_norm": 4.3626837730407715,
       "learning_rate": 9.729026506542566e-06,
-      "loss": 1.1422,
+      "loss": 1.0744,
       "step": 1627
     },
     {
       "epoch": 1.5714285714285714,
-      "grad_norm": 4.316422939300537,
+      "grad_norm": 4.418444633483887,
       "learning_rate": 9.71860714979422e-06,
-      "loss": 1.1945,
+      "loss": 1.1144,
       "step": 1628
     },
     {
       "epoch": 1.5723938223938223,
-      "grad_norm": 4.788571357727051,
+      "grad_norm": 4.496374607086182,
       "learning_rate": 9.708188098767577e-06,
-      "loss": 1.1744,
+      "loss": 1.1029,
       "step": 1629
     },
     {
       "epoch": 1.5733590733590734,
-      "grad_norm": 4.579430103302002,
+      "grad_norm": 4.690560817718506,
       "learning_rate": 9.697769364782502e-06,
-      "loss": 1.2298,
+      "loss": 1.1621,
       "step": 1630
     },
     {
       "epoch": 1.5743243243243243,
-      "grad_norm": 4.8498687744140625,
+      "grad_norm": 4.257440567016602,
       "learning_rate": 9.687350959158525e-06,
-      "loss": 1.2028,
+      "loss": 1.131,
       "step": 1631
     },
     {
       "epoch": 1.5752895752895753,
-      "grad_norm": 4.816117763519287,
+      "grad_norm": 4.154414176940918,
       "learning_rate": 9.676932893214806e-06,
-      "loss": 1.1988,
+      "loss": 1.1182,
       "step": 1632
     },
     {
       "epoch": 1.5762548262548264,
-      "grad_norm": 4.2359514236450195,
+      "grad_norm": 4.160477638244629,
       "learning_rate": 9.666515178270146e-06,
-      "loss": 1.1664,
+      "loss": 1.091,
       "step": 1633
     },
     {
       "epoch": 1.577220077220077,
-      "grad_norm": 4.571994781494141,
+      "grad_norm": 4.293689250946045,
       "learning_rate": 9.65609782564296e-06,
-      "loss": 1.1995,
+      "loss": 1.1176,
       "step": 1634
     },
     {
       "epoch": 1.5781853281853282,
-      "grad_norm": 4.250718116760254,
+      "grad_norm": 4.275992393493652,
       "learning_rate": 9.645680846651281e-06,
-      "loss": 1.1907,
+      "loss": 1.1134,
       "step": 1635
     },
     {
       "epoch": 1.579150579150579,
-      "grad_norm": 4.135900497436523,
+      "grad_norm": 4.237476825714111,
       "learning_rate": 9.635264252612712e-06,
-      "loss": 1.1788,
+      "loss": 1.1154,
       "step": 1636
     },
     {
       "epoch": 1.58011583011583,
-      "grad_norm": 4.255411148071289,
+      "grad_norm": 4.2550177574157715,
       "learning_rate": 9.62484805484446e-06,
-      "loss": 1.2615,
+      "loss": 1.1824,
       "step": 1637
     },
     {
       "epoch": 1.5810810810810811,
-      "grad_norm": 4.202883243560791,
+      "grad_norm": 4.229808807373047,
       "learning_rate": 9.614432264663288e-06,
-      "loss": 1.2089,
+      "loss": 1.1323,
       "step": 1638
     },
     {
       "epoch": 1.582046332046332,
-      "grad_norm": 4.191743850708008,
+      "grad_norm": 3.95070481300354,
       "learning_rate": 9.604016893385528e-06,
-      "loss": 1.2093,
+      "loss": 1.1064,
       "step": 1639
     },
     {
       "epoch": 1.583011583011583,
-      "grad_norm": 5.0460686683654785,
+      "grad_norm": 4.345361232757568,
       "learning_rate": 9.593601952327049e-06,
-      "loss": 1.168,
+      "loss": 1.0897,
       "step": 1640
     },
     {
       "epoch": 1.583976833976834,
-      "grad_norm": 4.46348237991333,
+      "grad_norm": 4.326444625854492,
       "learning_rate": 9.58318745280325e-06,
-      "loss": 1.188,
+      "loss": 1.1118,
       "step": 1641
     },
     {
       "epoch": 1.584942084942085,
-      "grad_norm": 5.175261497497559,
+      "grad_norm": 4.796785831451416,
       "learning_rate": 9.572773406129056e-06,
-      "loss": 1.185,
+      "loss": 1.114,
       "step": 1642
     },
     {
       "epoch": 1.585907335907336,
-      "grad_norm": 4.250487804412842,
+      "grad_norm": 4.3166399002075195,
       "learning_rate": 9.562359823618901e-06,
-      "loss": 1.167,
+      "loss": 1.1074,
       "step": 1643
     },
     {
       "epoch": 1.586872586872587,
-      "grad_norm": 4.666817665100098,
+      "grad_norm": 4.569392204284668,
       "learning_rate": 9.551946716586712e-06,
-      "loss": 1.1994,
+      "loss": 1.1393,
       "step": 1644
     },
     {
       "epoch": 1.5878378378378377,
-      "grad_norm": 6.0936713218688965,
+      "grad_norm": 5.153756618499756,
       "learning_rate": 9.541534096345896e-06,
-      "loss": 1.2028,
+      "loss": 1.1175,
       "step": 1645
     },
     {
       "epoch": 1.5888030888030888,
-      "grad_norm": 4.434663772583008,
+      "grad_norm": 4.340869426727295,
       "learning_rate": 9.531121974209334e-06,
-      "loss": 1.2667,
+      "loss": 1.1776,
       "step": 1646
     },
     {
       "epoch": 1.5897683397683398,
-      "grad_norm": 5.738936901092529,
+      "grad_norm": 4.514755725860596,
       "learning_rate": 9.52071036148937e-06,
-      "loss": 1.192,
+      "loss": 1.1294,
       "step": 1647
     },
     {
       "epoch": 1.5907335907335907,
-      "grad_norm": 4.914812088012695,
+      "grad_norm": 4.548101425170898,
       "learning_rate": 9.510299269497791e-06,
-      "loss": 1.2331,
+      "loss": 1.1469,
       "step": 1648
     },
     {
       "epoch": 1.5916988416988418,
-      "grad_norm": 4.0906901359558105,
+      "grad_norm": 4.447849273681641,
       "learning_rate": 9.499888709545816e-06,
-      "loss": 1.2696,
+      "loss": 1.2029,
       "step": 1649
     },
     {
       "epoch": 1.5926640926640927,
-      "grad_norm": 5.1826324462890625,
+      "grad_norm": 4.387144565582275,
       "learning_rate": 9.489478692944087e-06,
-      "loss": 1.1439,
+      "loss": 1.0621,
       "step": 1650
     },
     {
       "epoch": 1.5936293436293436,
-      "grad_norm": 4.445993423461914,
+      "grad_norm": 4.189123630523682,
       "learning_rate": 9.479069231002658e-06,
-      "loss": 1.1957,
+      "loss": 1.1209,
       "step": 1651
     },
     {
       "epoch": 1.5945945945945947,
-      "grad_norm": 4.667313098907471,
+      "grad_norm": 4.776906967163086,
       "learning_rate": 9.46866033503098e-06,
-      "loss": 1.2407,
+      "loss": 1.1679,
       "step": 1652
     },
     {
       "epoch": 1.5955598455598454,
-      "grad_norm": 4.472392559051514,
+      "grad_norm": 4.434809684753418,
       "learning_rate": 9.458252016337891e-06,
-      "loss": 1.2802,
+      "loss": 1.2169,
       "step": 1653
     },
     {
       "epoch": 1.5965250965250966,
-      "grad_norm": 4.763835906982422,
+      "grad_norm": 4.32791805267334,
       "learning_rate": 9.447844286231594e-06,
-      "loss": 1.2293,
+      "loss": 1.1627,
       "step": 1654
     },
     {
       "epoch": 1.5974903474903475,
-      "grad_norm": 3.701159715652466,
+      "grad_norm": 4.149640083312988,
       "learning_rate": 9.437437156019657e-06,
-      "loss": 1.1311,
+      "loss": 1.0608,
       "step": 1655
     },
     {
       "epoch": 1.5984555984555984,
-      "grad_norm": 5.122166633605957,
+      "grad_norm": 4.14592981338501,
       "learning_rate": 9.427030637009002e-06,
-      "loss": 1.1632,
+      "loss": 1.0814,
       "step": 1656
     },
     {
       "epoch": 1.5994208494208495,
-      "grad_norm": 4.4899678230285645,
+      "grad_norm": 4.505517482757568,
       "learning_rate": 9.416624740505882e-06,
-      "loss": 1.2298,
+      "loss": 1.1515,
       "step": 1657
     },
     {
       "epoch": 1.6003861003861004,
-      "grad_norm": 4.307829856872559,
+      "grad_norm": 4.343052864074707,
       "learning_rate": 9.406219477815868e-06,
-      "loss": 1.2088,
+      "loss": 1.1275,
       "step": 1658
     },
     {
       "epoch": 1.6013513513513513,
-      "grad_norm": 4.44758939743042,
+      "grad_norm": 4.818719863891602,
       "learning_rate": 9.395814860243848e-06,
-      "loss": 1.204,
+      "loss": 1.1311,
       "step": 1659
     },
     {
       "epoch": 1.6023166023166024,
-      "grad_norm": 4.317739963531494,
+      "grad_norm": 4.469165802001953,
       "learning_rate": 9.385410899094014e-06,
-      "loss": 1.2388,
+      "loss": 1.1718,
       "step": 1660
     },
     {
       "epoch": 1.6032818532818531,
-      "grad_norm": 4.105050563812256,
+      "grad_norm": 4.762367248535156,
       "learning_rate": 9.375007605669835e-06,
-      "loss": 1.2187,
+      "loss": 1.1424,
       "step": 1661
     },
     {
       "epoch": 1.6042471042471043,
-      "grad_norm": 4.454011917114258,
+      "grad_norm": 4.280526638031006,
       "learning_rate": 9.364604991274069e-06,
-      "loss": 1.2077,
+      "loss": 1.1296,
       "step": 1662
     },
     {
       "epoch": 1.6052123552123552,
-      "grad_norm": 4.6989426612854,
+      "grad_norm": 4.139819622039795,
       "learning_rate": 9.354203067208711e-06,
-      "loss": 1.1345,
+      "loss": 1.0605,
       "step": 1663
     },
     {
       "epoch": 1.606177606177606,
-      "grad_norm": 4.412756443023682,
+      "grad_norm": 4.427341938018799,
       "learning_rate": 9.343801844775035e-06,
-      "loss": 1.198,
+      "loss": 1.1126,
       "step": 1664
     },
     {
       "epoch": 1.6071428571428572,
-      "grad_norm": 4.618637561798096,
+      "grad_norm": 4.6331281661987305,
       "learning_rate": 9.33340133527353e-06,
-      "loss": 1.2647,
+      "loss": 1.189,
       "step": 1665
     },
     {
       "epoch": 1.6081081081081081,
-      "grad_norm": 4.564658164978027,
+      "grad_norm": 4.600239276885986,
       "learning_rate": 9.32300155000393e-06,
-      "loss": 1.224,
+      "loss": 1.1449,
       "step": 1666
     },
     {
       "epoch": 1.609073359073359,
-      "grad_norm": 4.380929470062256,
+      "grad_norm": 4.7197489738464355,
       "learning_rate": 9.312602500265162e-06,
-      "loss": 1.1752,
+      "loss": 1.1022,
       "step": 1667
     },
     {
       "epoch": 1.6100386100386102,
-      "grad_norm": 4.754830360412598,
+      "grad_norm": 4.397529125213623,
       "learning_rate": 9.302204197355368e-06,
-      "loss": 1.2209,
+      "loss": 1.1265,
       "step": 1668
     },
     {
       "epoch": 1.6110038610038608,
-      "grad_norm": 3.8590304851531982,
+      "grad_norm": 4.1421027183532715,
       "learning_rate": 9.291806652571873e-06,
-      "loss": 1.171,
+      "loss": 1.1129,
       "step": 1669
     },
     {
       "epoch": 1.611969111969112,
-      "grad_norm": 4.317570209503174,
+      "grad_norm": 4.418212413787842,
       "learning_rate": 9.28140987721118e-06,
-      "loss": 1.136,
+      "loss": 1.0583,
       "step": 1670
     },
     {
       "epoch": 1.612934362934363,
-      "grad_norm": 4.725559234619141,
+      "grad_norm": 4.703913688659668,
       "learning_rate": 9.27101388256896e-06,
-      "loss": 1.1823,
+      "loss": 1.1005,
       "step": 1671
     },
     {
       "epoch": 1.6138996138996138,
-      "grad_norm": 4.458336353302002,
+      "grad_norm": 4.7624006271362305,
       "learning_rate": 9.260618679940025e-06,
-      "loss": 1.2175,
+      "loss": 1.1345,
       "step": 1672
     },
     {
       "epoch": 1.614864864864865,
-      "grad_norm": 4.144418239593506,
+      "grad_norm": 4.401514053344727,
       "learning_rate": 9.250224280618337e-06,
-      "loss": 1.2245,
+      "loss": 1.1435,
       "step": 1673
     },
     {
       "epoch": 1.6158301158301158,
-      "grad_norm": 4.354918956756592,
+      "grad_norm": 4.4396209716796875,
       "learning_rate": 9.239830695896978e-06,
-      "loss": 1.1639,
+      "loss": 1.086,
       "step": 1674
     },
     {
       "epoch": 1.6167953667953667,
-      "grad_norm": 4.767379283905029,
+      "grad_norm": 4.730851650238037,
       "learning_rate": 9.229437937068155e-06,
-      "loss": 1.2238,
+      "loss": 1.1394,
       "step": 1675
     },
     {
       "epoch": 1.6177606177606179,
-      "grad_norm": 4.249136924743652,
+      "grad_norm": 4.259762763977051,
       "learning_rate": 9.219046015423162e-06,
-      "loss": 1.209,
+      "loss": 1.1234,
       "step": 1676
     },
     {
       "epoch": 1.6187258687258688,
-      "grad_norm": 4.206948280334473,
+      "grad_norm": 4.35798454284668,
       "learning_rate": 9.2086549422524e-06,
-      "loss": 1.1854,
+      "loss": 1.1243,
       "step": 1677
     },
     {
       "epoch": 1.6196911196911197,
-      "grad_norm": 4.812625408172607,
+      "grad_norm": 5.463016510009766,
       "learning_rate": 9.198264728845332e-06,
-      "loss": 1.174,
+      "loss": 1.104,
       "step": 1678
     },
     {
       "epoch": 1.6206563706563708,
-      "grad_norm": 4.244357109069824,
+      "grad_norm": 4.12177848815918,
       "learning_rate": 9.187875386490504e-06,
-      "loss": 1.1699,
+      "loss": 1.0983,
       "step": 1679
     },
     {
       "epoch": 1.6216216216216215,
-      "grad_norm": 5.580759048461914,
+      "grad_norm": 5.262537956237793,
       "learning_rate": 9.177486926475507e-06,
-      "loss": 1.1849,
+      "loss": 1.1063,
       "step": 1680
     },
     {
       "epoch": 1.6225868725868726,
-      "grad_norm": 4.3491692543029785,
+      "grad_norm": 4.460806369781494,
       "learning_rate": 9.167099360086967e-06,
-      "loss": 1.1835,
+      "loss": 1.1058,
       "step": 1681
     },
     {
       "epoch": 1.6235521235521235,
-      "grad_norm": 5.0731940269470215,
+      "grad_norm": 4.351454257965088,
       "learning_rate": 9.156712698610546e-06,
-      "loss": 1.1614,
+      "loss": 1.0882,
       "step": 1682
     },
     {
       "epoch": 1.6245173745173744,
-      "grad_norm": 4.256218910217285,
+      "grad_norm": 4.466792106628418,
       "learning_rate": 9.146326953330929e-06,
-      "loss": 1.2258,
+      "loss": 1.16,
       "step": 1683
     },
     {
       "epoch": 1.6254826254826256,
-      "grad_norm": 4.389143943786621,
+      "grad_norm": 4.746945381164551,
       "learning_rate": 9.135942135531797e-06,
-      "loss": 1.2423,
+      "loss": 1.1656,
       "step": 1684
     },
     {
       "epoch": 1.6264478764478765,
-      "grad_norm": 5.3131561279296875,
+      "grad_norm": 4.435761451721191,
       "learning_rate": 9.12555825649582e-06,
-      "loss": 1.208,
+      "loss": 1.1336,
       "step": 1685
     },
     {
       "epoch": 1.6274131274131274,
-      "grad_norm": 4.0794548988342285,
+      "grad_norm": 4.212765693664551,
       "learning_rate": 9.115175327504656e-06,
-      "loss": 1.2044,
+      "loss": 1.1249,
       "step": 1686
     },
     {
       "epoch": 1.6283783783783785,
-      "grad_norm": 5.013195037841797,
+      "grad_norm": 4.521017551422119,
       "learning_rate": 9.104793359838931e-06,
-      "loss": 1.2365,
+      "loss": 1.1549,
       "step": 1687
     },
     {
       "epoch": 1.6293436293436292,
-      "grad_norm": 3.932877779006958,
+      "grad_norm": 5.166461944580078,
       "learning_rate": 9.094412364778222e-06,
-      "loss": 1.2031,
+      "loss": 1.1252,
       "step": 1688
     },
     {
       "epoch": 1.6303088803088803,
-      "grad_norm": 4.152234077453613,
+      "grad_norm": 4.069167613983154,
       "learning_rate": 9.084032353601053e-06,
-      "loss": 1.1921,
+      "loss": 1.1115,
       "step": 1689
     },
     {
       "epoch": 1.6312741312741312,
-      "grad_norm": 4.233832836151123,
+      "grad_norm": 5.020195484161377,
       "learning_rate": 9.073653337584872e-06,
-      "loss": 1.2151,
+      "loss": 1.1396,
       "step": 1690
     },
     {
       "epoch": 1.6322393822393821,
-      "grad_norm": 4.629663467407227,
+      "grad_norm": 4.627904415130615,
       "learning_rate": 9.063275328006055e-06,
-      "loss": 1.1999,
+      "loss": 1.1196,
       "step": 1691
     },
     {
       "epoch": 1.6332046332046333,
-      "grad_norm": 4.575491905212402,
+      "grad_norm": 4.439563274383545,
       "learning_rate": 9.05289833613988e-06,
-      "loss": 1.183,
+      "loss": 1.1005,
       "step": 1692
     },
     {
       "epoch": 1.6341698841698842,
-      "grad_norm": 4.473489284515381,
+      "grad_norm": 4.267497539520264,
       "learning_rate": 9.042522373260522e-06,
-      "loss": 1.2094,
+      "loss": 1.1495,
       "step": 1693
     },
     {
       "epoch": 1.635135135135135,
-      "grad_norm": 4.254086494445801,
+      "grad_norm": 4.575072765350342,
       "learning_rate": 9.03214745064103e-06,
-      "loss": 1.2092,
+      "loss": 1.1228,
       "step": 1694
     },
     {
       "epoch": 1.6361003861003862,
-      "grad_norm": 4.169577598571777,
+      "grad_norm": 4.169126987457275,
       "learning_rate": 9.02177357955333e-06,
-      "loss": 1.1868,
+      "loss": 1.1038,
       "step": 1695
     },
     {
       "epoch": 1.637065637065637,
-      "grad_norm": 4.918846607208252,
+      "grad_norm": 4.373612403869629,
       "learning_rate": 9.01140077126821e-06,
-      "loss": 1.214,
+      "loss": 1.1407,
       "step": 1696
     },
     {
       "epoch": 1.638030888030888,
-      "grad_norm": 4.24448823928833,
+      "grad_norm": 4.4685845375061035,
       "learning_rate": 9.00102903705529e-06,
-      "loss": 1.1783,
+      "loss": 1.1118,
       "step": 1697
     },
     {
       "epoch": 1.6389961389961392,
-      "grad_norm": 4.124368667602539,
+      "grad_norm": 4.164231300354004,
       "learning_rate": 8.990658388183038e-06,
-      "loss": 1.148,
+      "loss": 1.0657,
       "step": 1698
     },
     {
       "epoch": 1.6399613899613898,
-      "grad_norm": 4.327722072601318,
+      "grad_norm": 4.416538238525391,
       "learning_rate": 8.98028883591873e-06,
-      "loss": 1.2095,
+      "loss": 1.1293,
       "step": 1699
     },
     {
       "epoch": 1.640926640926641,
-      "grad_norm": 4.191107749938965,
+      "grad_norm": 4.1006293296813965,
       "learning_rate": 8.969920391528459e-06,
-      "loss": 1.1779,
+      "loss": 1.0894,
       "step": 1700
     },
     {
       "epoch": 1.6418918918918919,
-      "grad_norm": 4.929723739624023,
+      "grad_norm": 4.815096855163574,
       "learning_rate": 8.959553066277104e-06,
-      "loss": 1.1678,
+      "loss": 1.0728,
       "step": 1701
     },
     {
       "epoch": 1.6428571428571428,
-      "grad_norm": 4.273260593414307,
+      "grad_norm": 4.367018699645996,
       "learning_rate": 8.949186871428347e-06,
-      "loss": 1.1998,
+      "loss": 1.1277,
       "step": 1702
     },
     {
       "epoch": 1.643822393822394,
-      "grad_norm": 5.487876892089844,
+      "grad_norm": 4.297732830047607,
       "learning_rate": 8.938821818244626e-06,
-      "loss": 1.2464,
+      "loss": 1.1564,
       "step": 1703
     },
     {
       "epoch": 1.6447876447876448,
-      "grad_norm": 3.923842191696167,
+      "grad_norm": 4.397197723388672,
       "learning_rate": 8.928457917987137e-06,
-      "loss": 1.2016,
+      "loss": 1.1159,
       "step": 1704
     },
     {
       "epoch": 1.6457528957528957,
-      "grad_norm": 4.900607585906982,
+      "grad_norm": 4.462916374206543,
       "learning_rate": 8.918095181915833e-06,
-      "loss": 1.2306,
+      "loss": 1.1645,
       "step": 1705
     },
     {
       "epoch": 1.6467181467181469,
-      "grad_norm": 5.042302131652832,
+      "grad_norm": 4.433206558227539,
       "learning_rate": 8.9077336212894e-06,
-      "loss": 1.1821,
+      "loss": 1.1093,
       "step": 1706
     },
     {
       "epoch": 1.6476833976833976,
-      "grad_norm": 4.501278400421143,
+      "grad_norm": 4.508075714111328,
       "learning_rate": 8.897373247365249e-06,
-      "loss": 1.2297,
+      "loss": 1.1388,
       "step": 1707
     },
     {
       "epoch": 1.6486486486486487,
-      "grad_norm": 4.577908039093018,
+      "grad_norm": 4.344179630279541,
       "learning_rate": 8.887014071399488e-06,
-      "loss": 1.231,
+      "loss": 1.1555,
       "step": 1708
     },
     {
       "epoch": 1.6496138996138996,
-      "grad_norm": 5.391479015350342,
+      "grad_norm": 4.900021553039551,
       "learning_rate": 8.876656104646942e-06,
-      "loss": 1.2098,
+      "loss": 1.1383,
       "step": 1709
     },
     {
       "epoch": 1.6505791505791505,
-      "grad_norm": 4.688028812408447,
+      "grad_norm": 4.218349933624268,
       "learning_rate": 8.86629935836111e-06,
-      "loss": 1.1722,
+      "loss": 1.0828,
       "step": 1710
     },
     {
       "epoch": 1.6515444015444016,
-      "grad_norm": 5.170119762420654,
+      "grad_norm": 4.315882205963135,
       "learning_rate": 8.855943843794171e-06,
-      "loss": 1.2354,
+      "loss": 1.1588,
       "step": 1711
     },
     {
       "epoch": 1.6525096525096525,
-      "grad_norm": 4.34480619430542,
+      "grad_norm": 4.179636478424072,
       "learning_rate": 8.84558957219696e-06,
-      "loss": 1.1562,
+      "loss": 1.07,
       "step": 1712
     },
     {
       "epoch": 1.6534749034749034,
-      "grad_norm": 4.477398872375488,
+      "grad_norm": 4.671791076660156,
       "learning_rate": 8.835236554818967e-06,
-      "loss": 1.2139,
+      "loss": 1.1304,
       "step": 1713
     },
     {
       "epoch": 1.6544401544401546,
-      "grad_norm": 4.357746124267578,
+      "grad_norm": 4.379439353942871,
       "learning_rate": 8.82488480290832e-06,
-      "loss": 1.1659,
+      "loss": 1.0929,
       "step": 1714
     },
     {
       "epoch": 1.6554054054054053,
-      "grad_norm": 4.401113033294678,
+      "grad_norm": 4.533711910247803,
       "learning_rate": 8.814534327711768e-06,
-      "loss": 1.2111,
+      "loss": 1.1255,
       "step": 1715
     },
     {
       "epoch": 1.6563706563706564,
-      "grad_norm": 4.574923992156982,
+      "grad_norm": 5.072481632232666,
       "learning_rate": 8.804185140474675e-06,
-      "loss": 1.1184,
+      "loss": 1.0455,
       "step": 1716
     },
     {
       "epoch": 1.6573359073359073,
-      "grad_norm": 4.733505725860596,
+      "grad_norm": 3.9613430500030518,
       "learning_rate": 8.793837252441e-06,
-      "loss": 1.182,
+      "loss": 1.0953,
       "step": 1717
     },
     {
       "epoch": 1.6583011583011582,
-      "grad_norm": 4.71107816696167,
+      "grad_norm": 4.845931053161621,
       "learning_rate": 8.783490674853302e-06,
-      "loss": 1.2215,
+      "loss": 1.1469,
       "step": 1718
     },
     {
       "epoch": 1.6592664092664093,
-      "grad_norm": 5.102229595184326,
+      "grad_norm": 4.363834381103516,
       "learning_rate": 8.773145418952708e-06,
-      "loss": 1.1866,
+      "loss": 1.0949,
       "step": 1719
     },
     {
       "epoch": 1.6602316602316602,
-      "grad_norm": 4.5466132164001465,
+      "grad_norm": 4.350069999694824,
       "learning_rate": 8.762801495978914e-06,
-      "loss": 1.2404,
+      "loss": 1.1663,
       "step": 1720
     },
     {
       "epoch": 1.6611969111969112,
-      "grad_norm": 4.194196701049805,
+      "grad_norm": 4.199606895446777,
       "learning_rate": 8.752458917170157e-06,
-      "loss": 1.1256,
+      "loss": 1.0552,
       "step": 1721
     },
     {
       "epoch": 1.6621621621621623,
-      "grad_norm": 5.514886379241943,
+      "grad_norm": 4.5061445236206055,
       "learning_rate": 8.742117693763229e-06,
-      "loss": 1.2062,
+      "loss": 1.1062,
       "step": 1722
     },
     {
       "epoch": 1.663127413127413,
-      "grad_norm": 4.242153167724609,
+      "grad_norm": 4.399968147277832,
       "learning_rate": 8.731777836993437e-06,
-      "loss": 1.1739,
+      "loss": 1.1021,
       "step": 1723
     },
     {
       "epoch": 1.664092664092664,
-      "grad_norm": 4.451455116271973,
+      "grad_norm": 4.2255120277404785,
       "learning_rate": 8.721439358094613e-06,
-      "loss": 1.2389,
+      "loss": 1.1482,
       "step": 1724
     },
     {
       "epoch": 1.665057915057915,
-      "grad_norm": 5.862081527709961,
+      "grad_norm": 4.252571105957031,
       "learning_rate": 8.71110226829908e-06,
-      "loss": 1.2433,
+      "loss": 1.1426,
       "step": 1725
     },
     {
       "epoch": 1.666023166023166,
-      "grad_norm": 4.439873695373535,
+      "grad_norm": 4.2842793464660645,
       "learning_rate": 8.700766578837664e-06,
-      "loss": 1.1859,
+      "loss": 1.1041,
       "step": 1726
     },
     {
       "epoch": 1.666988416988417,
-      "grad_norm": 4.343622207641602,
+      "grad_norm": 4.363378047943115,
       "learning_rate": 8.690432300939664e-06,
-      "loss": 1.1274,
+      "loss": 1.0564,
       "step": 1727
     },
     {
       "epoch": 1.667953667953668,
-      "grad_norm": 4.739264011383057,
+      "grad_norm": 4.431227684020996,
       "learning_rate": 8.680099445832843e-06,
-      "loss": 1.149,
+      "loss": 1.0544,
       "step": 1728
     },
     {
       "epoch": 1.6689189189189189,
-      "grad_norm": 4.081526756286621,
+      "grad_norm": 4.420811176300049,
       "learning_rate": 8.669768024743429e-06,
-      "loss": 1.1458,
+      "loss": 1.0754,
       "step": 1729
     },
     {
       "epoch": 1.66988416988417,
-      "grad_norm": 4.270155906677246,
+      "grad_norm": 4.0607380867004395,
       "learning_rate": 8.659438048896072e-06,
-      "loss": 1.2105,
+      "loss": 1.1183,
       "step": 1730
     },
     {
       "epoch": 1.670849420849421,
-      "grad_norm": 4.286427021026611,
+      "grad_norm": 4.604260444641113,
       "learning_rate": 8.649109529513873e-06,
-      "loss": 1.1492,
+      "loss": 1.0673,
       "step": 1731
     },
     {
       "epoch": 1.6718146718146718,
-      "grad_norm": 4.289936065673828,
+      "grad_norm": 4.217810153961182,
       "learning_rate": 8.638782477818335e-06,
-      "loss": 1.1017,
+      "loss": 1.033,
       "step": 1732
     },
     {
       "epoch": 1.672779922779923,
-      "grad_norm": 4.637660980224609,
+      "grad_norm": 4.3615498542785645,
       "learning_rate": 8.628456905029383e-06,
-      "loss": 1.2084,
+      "loss": 1.114,
       "step": 1733
     },
     {
       "epoch": 1.6737451737451736,
-      "grad_norm": 4.21026086807251,
+      "grad_norm": 4.17683744430542,
       "learning_rate": 8.618132822365312e-06,
-      "loss": 1.2537,
+      "loss": 1.1669,
       "step": 1734
     },
     {
       "epoch": 1.6747104247104247,
-      "grad_norm": 3.7908935546875,
+      "grad_norm": 3.789360284805298,
       "learning_rate": 8.607810241042819e-06,
-      "loss": 1.1979,
+      "loss": 1.1117,
       "step": 1735
     },
     {
       "epoch": 1.6756756756756757,
-      "grad_norm": 4.276378154754639,
+      "grad_norm": 4.2690839767456055,
       "learning_rate": 8.597489172276959e-06,
-      "loss": 1.1713,
+      "loss": 1.0907,
       "step": 1736
     },
     {
       "epoch": 1.6766409266409266,
-      "grad_norm": 4.769236087799072,
+      "grad_norm": 4.702624320983887,
       "learning_rate": 8.587169627281149e-06,
-      "loss": 1.204,
+      "loss": 1.1254,
       "step": 1737
     },
     {
       "epoch": 1.6776061776061777,
-      "grad_norm": 4.216571807861328,
+      "grad_norm": 4.081230163574219,
       "learning_rate": 8.576851617267151e-06,
-      "loss": 1.2026,
+      "loss": 1.1095,
       "step": 1738
     },
     {
       "epoch": 1.6785714285714286,
-      "grad_norm": 4.589860916137695,
+      "grad_norm": 4.545328140258789,
       "learning_rate": 8.566535153445049e-06,
-      "loss": 1.2139,
+      "loss": 1.1423,
       "step": 1739
     },
     {
       "epoch": 1.6795366795366795,
-      "grad_norm": 4.3696722984313965,
+      "grad_norm": 4.471418380737305,
       "learning_rate": 8.556220247023257e-06,
-      "loss": 1.1702,
+      "loss": 1.0898,
       "step": 1740
     },
     {
       "epoch": 1.6805019305019306,
-      "grad_norm": 4.10421085357666,
+      "grad_norm": 4.347437381744385,
       "learning_rate": 8.545906909208498e-06,
-      "loss": 1.1449,
+      "loss": 1.0664,
       "step": 1741
     },
     {
       "epoch": 1.6814671814671813,
-      "grad_norm": 4.313359260559082,
+      "grad_norm": 4.383377552032471,
       "learning_rate": 8.535595151205787e-06,
-      "loss": 1.2174,
+      "loss": 1.1242,
       "step": 1742
     },
     {
       "epoch": 1.6824324324324325,
-      "grad_norm": 4.988412857055664,
+      "grad_norm": 5.136453151702881,
       "learning_rate": 8.525284984218418e-06,
-      "loss": 1.2489,
+      "loss": 1.1747,
       "step": 1743
     },
     {
       "epoch": 1.6833976833976834,
-      "grad_norm": 5.241071701049805,
+      "grad_norm": 4.427022457122803,
       "learning_rate": 8.514976419447963e-06,
-      "loss": 1.2148,
+      "loss": 1.1339,
       "step": 1744
     },
     {
       "epoch": 1.6843629343629343,
-      "grad_norm": 5.049296855926514,
+      "grad_norm": 4.6567063331604,
       "learning_rate": 8.504669468094256e-06,
-      "loss": 1.1645,
+      "loss": 1.0861,
       "step": 1745
     },
     {
       "epoch": 1.6853281853281854,
-      "grad_norm": 5.042121887207031,
+      "grad_norm": 4.6290669441223145,
       "learning_rate": 8.494364141355368e-06,
-      "loss": 1.1522,
+      "loss": 1.0663,
       "step": 1746
     },
     {
       "epoch": 1.6862934362934363,
-      "grad_norm": 4.717390060424805,
+      "grad_norm": 4.114441871643066,
       "learning_rate": 8.484060450427616e-06,
-      "loss": 1.1816,
+      "loss": 1.094,
       "step": 1747
     },
     {
       "epoch": 1.6872586872586872,
-      "grad_norm": 4.598116874694824,
+      "grad_norm": 4.702686786651611,
       "learning_rate": 8.47375840650553e-06,
-      "loss": 1.2413,
+      "loss": 1.1524,
       "step": 1748
     },
     {
       "epoch": 1.6882239382239383,
-      "grad_norm": 4.408492088317871,
+      "grad_norm": 4.158077716827393,
       "learning_rate": 8.463458020781855e-06,
-      "loss": 1.1495,
+      "loss": 1.0629,
       "step": 1749
     },
     {
       "epoch": 1.689189189189189,
-      "grad_norm": 4.6580586433410645,
+      "grad_norm": 4.464077949523926,
       "learning_rate": 8.453159304447535e-06,
-      "loss": 1.2852,
+      "loss": 1.177,
       "step": 1750
     },
     {
       "epoch": 1.6901544401544402,
-      "grad_norm": 4.525180816650391,
+      "grad_norm": 4.717648506164551,
       "learning_rate": 8.442862268691703e-06,
-      "loss": 1.1859,
+      "loss": 1.1127,
       "step": 1751
     },
     {
       "epoch": 1.691119691119691,
-      "grad_norm": 4.853588104248047,
+      "grad_norm": 4.74327278137207,
       "learning_rate": 8.43256692470166e-06,
-      "loss": 1.2238,
+      "loss": 1.1343,
       "step": 1752
     },
     {
       "epoch": 1.692084942084942,
-      "grad_norm": 5.207756519317627,
+      "grad_norm": 4.823689937591553,
       "learning_rate": 8.422273283662869e-06,
-      "loss": 1.1487,
+      "loss": 1.0627,
       "step": 1753
     },
     {
       "epoch": 1.693050193050193,
-      "grad_norm": 4.643857002258301,
+      "grad_norm": 4.651237487792969,
       "learning_rate": 8.41198135675895e-06,
-      "loss": 1.1702,
+      "loss": 1.0887,
       "step": 1754
     },
     {
       "epoch": 1.694015444015444,
-      "grad_norm": 4.395835876464844,
+      "grad_norm": 4.5374836921691895,
       "learning_rate": 8.401691155171654e-06,
-      "loss": 1.2509,
+      "loss": 1.1664,
       "step": 1755
     },
     {
       "epoch": 1.694980694980695,
-      "grad_norm": 4.321347713470459,
+      "grad_norm": 4.5740861892700195,
       "learning_rate": 8.391402690080864e-06,
-      "loss": 1.1961,
+      "loss": 1.1036,
       "step": 1756
     },
     {
       "epoch": 1.695945945945946,
-      "grad_norm": 4.597218990325928,
+      "grad_norm": 4.997374057769775,
       "learning_rate": 8.381115972664563e-06,
-      "loss": 1.2019,
+      "loss": 1.1022,
       "step": 1757
     },
     {
       "epoch": 1.696911196911197,
-      "grad_norm": 4.090878486633301,
+      "grad_norm": 4.214781761169434,
       "learning_rate": 8.370831014098854e-06,
-      "loss": 1.2323,
+      "loss": 1.143,
       "step": 1758
     },
     {
       "epoch": 1.6978764478764479,
-      "grad_norm": 4.209794521331787,
+      "grad_norm": 4.494675159454346,
       "learning_rate": 8.360547825557912e-06,
-      "loss": 1.1784,
+      "loss": 1.0798,
       "step": 1759
     },
     {
       "epoch": 1.698841698841699,
-      "grad_norm": 4.162314414978027,
+      "grad_norm": 4.681112766265869,
       "learning_rate": 8.350266418214006e-06,
-      "loss": 1.1674,
+      "loss": 1.0807,
       "step": 1760
     },
     {
       "epoch": 1.6998069498069497,
-      "grad_norm": 3.905789852142334,
+      "grad_norm": 4.20454740524292,
       "learning_rate": 8.339986803237446e-06,
-      "loss": 1.1917,
+      "loss": 1.0986,
       "step": 1761
     },
     {
       "epoch": 1.7007722007722008,
-      "grad_norm": 4.145606994628906,
+      "grad_norm": 4.132695198059082,
       "learning_rate": 8.32970899179662e-06,
-      "loss": 1.2348,
+      "loss": 1.1412,
       "step": 1762
     },
     {
       "epoch": 1.7017374517374517,
-      "grad_norm": 3.782750368118286,
+      "grad_norm": 4.2170586585998535,
       "learning_rate": 8.31943299505794e-06,
-      "loss": 1.2221,
+      "loss": 1.126,
       "step": 1763
     },
     {
       "epoch": 1.7027027027027026,
-      "grad_norm": 4.163419723510742,
+      "grad_norm": 4.435767650604248,
       "learning_rate": 8.309158824185855e-06,
-      "loss": 1.1523,
+      "loss": 1.0566,
       "step": 1764
     },
     {
       "epoch": 1.7036679536679538,
-      "grad_norm": 4.2257771492004395,
+      "grad_norm": 4.509848117828369,
       "learning_rate": 8.298886490342827e-06,
-      "loss": 1.1298,
+      "loss": 1.0364,
       "step": 1765
     },
     {
       "epoch": 1.7046332046332047,
-      "grad_norm": 4.363615989685059,
+      "grad_norm": 4.331830978393555,
       "learning_rate": 8.288616004689321e-06,
-      "loss": 1.1604,
+      "loss": 1.0731,
       "step": 1766
     },
     {
       "epoch": 1.7055984555984556,
-      "grad_norm": 4.496586799621582,
+      "grad_norm": 4.496250152587891,
       "learning_rate": 8.278347378383794e-06,
-      "loss": 1.2354,
+      "loss": 1.1528,
       "step": 1767
     },
     {
       "epoch": 1.7065637065637067,
-      "grad_norm": 4.663826942443848,
+      "grad_norm": 4.950375556945801,
       "learning_rate": 8.268080622582682e-06,
-      "loss": 1.221,
+      "loss": 1.1438,
       "step": 1768
     },
     {
       "epoch": 1.7075289575289574,
-      "grad_norm": 4.14992094039917,
+      "grad_norm": 4.427841663360596,
       "learning_rate": 8.2578157484404e-06,
-      "loss": 1.2565,
+      "loss": 1.175,
       "step": 1769
     },
     {
       "epoch": 1.7084942084942085,
-      "grad_norm": 3.9128541946411133,
+      "grad_norm": 4.198677062988281,
       "learning_rate": 8.247552767109301e-06,
-      "loss": 1.1902,
+      "loss": 1.1043,
       "step": 1770
     },
     {
       "epoch": 1.7094594594594594,
-      "grad_norm": 4.58083438873291,
+      "grad_norm": 4.571022033691406,
       "learning_rate": 8.237291689739692e-06,
-      "loss": 1.2451,
+      "loss": 1.1647,
       "step": 1771
     },
     {
       "epoch": 1.7104247104247103,
-      "grad_norm": 4.241626262664795,
+      "grad_norm": 4.588777542114258,
       "learning_rate": 8.227032527479807e-06,
-      "loss": 1.1586,
+      "loss": 1.072,
       "step": 1772
     },
     {
       "epoch": 1.7113899613899615,
-      "grad_norm": 5.163703441619873,
+      "grad_norm": 4.4870734214782715,
       "learning_rate": 8.216775291475808e-06,
-      "loss": 1.1513,
+      "loss": 1.0699,
       "step": 1773
     },
     {
       "epoch": 1.7123552123552124,
-      "grad_norm": 3.928961992263794,
+      "grad_norm": 3.9703915119171143,
       "learning_rate": 8.206519992871757e-06,
-      "loss": 1.2216,
+      "loss": 1.1317,
       "step": 1774
     },
     {
       "epoch": 1.7133204633204633,
-      "grad_norm": 4.547893047332764,
+      "grad_norm": 4.011623859405518,
       "learning_rate": 8.196266642809607e-06,
-      "loss": 1.2212,
+      "loss": 1.1181,
       "step": 1775
     },
     {
       "epoch": 1.7142857142857144,
-      "grad_norm": 4.903558254241943,
+      "grad_norm": 4.904574871063232,
       "learning_rate": 8.1860152524292e-06,
-      "loss": 1.2437,
+      "loss": 1.1218,
       "step": 1776
     },
     {
       "epoch": 1.715250965250965,
-      "grad_norm": 4.899398326873779,
+      "grad_norm": 4.900587558746338,
       "learning_rate": 8.175765832868252e-06,
-      "loss": 1.2081,
+      "loss": 1.1172,
       "step": 1777
     },
     {
       "epoch": 1.7162162162162162,
-      "grad_norm": 4.3335442543029785,
+      "grad_norm": 4.507471561431885,
       "learning_rate": 8.165518395262335e-06,
-      "loss": 1.1931,
+      "loss": 1.1127,
       "step": 1778
     },
     {
       "epoch": 1.7171814671814671,
-      "grad_norm": 4.198944091796875,
+      "grad_norm": 4.117510795593262,
       "learning_rate": 8.155272950744859e-06,
-      "loss": 1.1742,
+      "loss": 1.0954,
       "step": 1779
     },
     {
       "epoch": 1.718146718146718,
-      "grad_norm": 4.206162452697754,
+      "grad_norm": 4.188136100769043,
       "learning_rate": 8.14502951044708e-06,
-      "loss": 1.1192,
+      "loss": 1.0334,
       "step": 1780
     },
     {
       "epoch": 1.7191119691119692,
-      "grad_norm": 4.380505561828613,
+      "grad_norm": 4.472517967224121,
       "learning_rate": 8.134788085498073e-06,
-      "loss": 1.1848,
+      "loss": 1.0976,
       "step": 1781
     },
     {
       "epoch": 1.72007722007722,
-      "grad_norm": 4.550119400024414,
+      "grad_norm": 4.673532962799072,
       "learning_rate": 8.124548687024723e-06,
-      "loss": 1.2425,
+      "loss": 1.161,
       "step": 1782
     },
     {
       "epoch": 1.721042471042471,
-      "grad_norm": 4.7923479080200195,
+      "grad_norm": 4.261319637298584,
       "learning_rate": 8.114311326151716e-06,
-      "loss": 1.2471,
+      "loss": 1.1482,
       "step": 1783
     },
     {
       "epoch": 1.7220077220077221,
-      "grad_norm": 6.3003058433532715,
+      "grad_norm": 4.546080112457275,
       "learning_rate": 8.104076014001514e-06,
-      "loss": 1.1869,
+      "loss": 1.0952,
       "step": 1784
     },
     {
       "epoch": 1.722972972972973,
-      "grad_norm": 4.867679119110107,
+      "grad_norm": 4.736138820648193,
       "learning_rate": 8.093842761694366e-06,
-      "loss": 1.2735,
+      "loss": 1.1885,
       "step": 1785
     },
     {
       "epoch": 1.723938223938224,
-      "grad_norm": 5.90767240524292,
+      "grad_norm": 4.569971084594727,
       "learning_rate": 8.083611580348278e-06,
-      "loss": 1.1858,
+      "loss": 1.0889,
       "step": 1786
     },
     {
       "epoch": 1.724903474903475,
-      "grad_norm": 4.679843425750732,
+      "grad_norm": 4.716104984283447,
       "learning_rate": 8.073382481079005e-06,
-      "loss": 1.1917,
+      "loss": 1.0987,
       "step": 1787
     },
     {
       "epoch": 1.7258687258687258,
-      "grad_norm": 4.496260643005371,
+      "grad_norm": 4.197176456451416,
       "learning_rate": 8.063155475000037e-06,
-      "loss": 1.1404,
+      "loss": 1.0444,
       "step": 1788
     },
     {
       "epoch": 1.7268339768339769,
-      "grad_norm": 4.13546085357666,
+      "grad_norm": 4.3186936378479,
       "learning_rate": 8.052930573222597e-06,
-      "loss": 1.1757,
+      "loss": 1.0906,
       "step": 1789
     },
     {
       "epoch": 1.7277992277992278,
-      "grad_norm": 4.124424934387207,
+      "grad_norm": 4.364828109741211,
       "learning_rate": 8.042707786855618e-06,
-      "loss": 1.1352,
+      "loss": 1.0376,
       "step": 1790
     },
     {
       "epoch": 1.7287644787644787,
-      "grad_norm": 4.252573013305664,
+      "grad_norm": 4.607423305511475,
       "learning_rate": 8.032487127005732e-06,
-      "loss": 1.2342,
+      "loss": 1.133,
       "step": 1791
     },
     {
       "epoch": 1.7297297297297298,
-      "grad_norm": 4.479787349700928,
+      "grad_norm": 4.915359973907471,
       "learning_rate": 8.02226860477727e-06,
-      "loss": 1.2001,
+      "loss": 1.1043,
       "step": 1792
     },
     {
       "epoch": 1.7306949806949807,
-      "grad_norm": 4.56220006942749,
+      "grad_norm": 4.3487019538879395,
       "learning_rate": 8.01205223127223e-06,
-      "loss": 1.1628,
+      "loss": 1.0633,
       "step": 1793
     },
     {
       "epoch": 1.7316602316602316,
-      "grad_norm": 4.39263916015625,
+      "grad_norm": 4.667450904846191,
       "learning_rate": 8.001838017590278e-06,
-      "loss": 1.1549,
+      "loss": 1.0708,
       "step": 1794
     },
     {
       "epoch": 1.7326254826254828,
-      "grad_norm": 4.32893705368042,
+      "grad_norm": 5.286256790161133,
       "learning_rate": 7.991625974828735e-06,
-      "loss": 1.1817,
+      "loss": 1.1036,
       "step": 1795
     },
     {
       "epoch": 1.7335907335907335,
-      "grad_norm": 5.986409664154053,
+      "grad_norm": 4.660265922546387,
       "learning_rate": 7.981416114082568e-06,
-      "loss": 1.1799,
+      "loss": 1.0964,
       "step": 1796
     },
     {
       "epoch": 1.7345559845559846,
-      "grad_norm": 5.146429061889648,
+      "grad_norm": 4.694789886474609,
       "learning_rate": 7.971208446444366e-06,
-      "loss": 1.1745,
+      "loss": 1.082,
       "step": 1797
     },
     {
       "epoch": 1.7355212355212355,
-      "grad_norm": 4.632083892822266,
+      "grad_norm": 4.473088264465332,
       "learning_rate": 7.961002983004335e-06,
-      "loss": 1.2565,
+      "loss": 1.1783,
       "step": 1798
     },
     {
       "epoch": 1.7364864864864864,
-      "grad_norm": 5.8181962966918945,
+      "grad_norm": 4.547810077667236,
       "learning_rate": 7.950799734850292e-06,
-      "loss": 1.1844,
+      "loss": 1.0966,
       "step": 1799
     },
     {
       "epoch": 1.7374517374517375,
-      "grad_norm": 5.046428680419922,
+      "grad_norm": 4.548643112182617,
       "learning_rate": 7.940598713067647e-06,
-      "loss": 1.1407,
+      "loss": 1.0413,
       "step": 1800
     },
     {
       "epoch": 1.7384169884169884,
-      "grad_norm": 3.968395709991455,
+      "grad_norm": 4.4184699058532715,
       "learning_rate": 7.930399928739385e-06,
-      "loss": 1.1586,
+      "loss": 1.0676,
       "step": 1801
     },
     {
       "epoch": 1.7393822393822393,
-      "grad_norm": 5.228263854980469,
+      "grad_norm": 4.427945613861084,
       "learning_rate": 7.920203392946064e-06,
-      "loss": 1.2388,
+      "loss": 1.1377,
       "step": 1802
     },
     {
       "epoch": 1.7403474903474905,
-      "grad_norm": 4.870275020599365,
+      "grad_norm": 4.846949577331543,
       "learning_rate": 7.910009116765796e-06,
-      "loss": 1.1788,
+      "loss": 1.1036,
       "step": 1803
     },
     {
       "epoch": 1.7413127413127412,
-      "grad_norm": 5.449280261993408,
+      "grad_norm": 4.154278755187988,
       "learning_rate": 7.899817111274249e-06,
-      "loss": 1.2277,
+      "loss": 1.1173,
       "step": 1804
     },
     {
       "epoch": 1.7422779922779923,
-      "grad_norm": 4.336848258972168,
+      "grad_norm": 4.608570575714111,
       "learning_rate": 7.889627387544611e-06,
-      "loss": 1.2175,
+      "loss": 1.1393,
       "step": 1805
     },
     {
       "epoch": 1.7432432432432432,
-      "grad_norm": 4.2892374992370605,
+      "grad_norm": 4.238851070404053,
       "learning_rate": 7.879439956647594e-06,
-      "loss": 1.1233,
+      "loss": 1.0311,
       "step": 1806
     },
     {
       "epoch": 1.744208494208494,
-      "grad_norm": 4.4046244621276855,
+      "grad_norm": 4.809391021728516,
       "learning_rate": 7.86925482965142e-06,
-      "loss": 1.2245,
+      "loss": 1.1298,
       "step": 1807
     },
     {
       "epoch": 1.7451737451737452,
-      "grad_norm": 4.251980304718018,
+      "grad_norm": 4.499746322631836,
       "learning_rate": 7.859072017621811e-06,
-      "loss": 1.1628,
+      "loss": 1.0793,
       "step": 1808
     },
     {
       "epoch": 1.7461389961389961,
-      "grad_norm": 4.342248916625977,
+      "grad_norm": 4.434761047363281,
       "learning_rate": 7.848891531621977e-06,
-      "loss": 1.2555,
+      "loss": 1.1538,
       "step": 1809
     },
     {
       "epoch": 1.747104247104247,
-      "grad_norm": 4.569221019744873,
+      "grad_norm": 4.726895809173584,
       "learning_rate": 7.838713382712583e-06,
-      "loss": 1.1232,
+      "loss": 1.0363,
       "step": 1810
     },
     {
       "epoch": 1.7480694980694982,
-      "grad_norm": 4.1841044425964355,
+      "grad_norm": 4.580567359924316,
       "learning_rate": 7.828537581951773e-06,
-      "loss": 1.2064,
+      "loss": 1.1128,
       "step": 1811
     },
     {
       "epoch": 1.7490347490347489,
-      "grad_norm": 4.37945556640625,
+      "grad_norm": 4.698511600494385,
       "learning_rate": 7.818364140395138e-06,
-      "loss": 1.1761,
+      "loss": 1.0842,
       "step": 1812
     },
     {
       "epoch": 1.75,
-      "grad_norm": 4.35389518737793,
+      "grad_norm": 4.618523597717285,
       "learning_rate": 7.808193069095698e-06,
-      "loss": 1.1817,
+      "loss": 1.0867,
       "step": 1813
     },
     {
       "epoch": 1.7509652509652511,
-      "grad_norm": 4.284439563751221,
+      "grad_norm": 4.588217735290527,
       "learning_rate": 7.798024379103903e-06,
-      "loss": 1.1517,
+      "loss": 1.0558,
       "step": 1814
     },
     {
       "epoch": 1.7519305019305018,
-      "grad_norm": 4.608835220336914,
+      "grad_norm": 4.88592529296875,
       "learning_rate": 7.787858081467609e-06,
-      "loss": 1.2335,
+      "loss": 1.137,
       "step": 1815
     },
     {
       "epoch": 1.752895752895753,
-      "grad_norm": 4.245322227478027,
+      "grad_norm": 4.6594343185424805,
       "learning_rate": 7.777694187232085e-06,
-      "loss": 1.1761,
+      "loss": 1.0791,
       "step": 1816
     },
     {
       "epoch": 1.7538610038610039,
-      "grad_norm": 4.203044891357422,
+      "grad_norm": 4.289200782775879,
       "learning_rate": 7.76753270743998e-06,
-      "loss": 1.1422,
+      "loss": 1.0394,
       "step": 1817
     },
     {
       "epoch": 1.7548262548262548,
-      "grad_norm": 4.512558460235596,
+      "grad_norm": 4.543483257293701,
       "learning_rate": 7.757373653131323e-06,
-      "loss": 1.1788,
+      "loss": 1.0752,
       "step": 1818
     },
     {
       "epoch": 1.755791505791506,
-      "grad_norm": 4.879763603210449,
+      "grad_norm": 5.076117515563965,
       "learning_rate": 7.7472170353435e-06,
-      "loss": 1.179,
+      "loss": 1.0823,
       "step": 1819
     },
     {
       "epoch": 1.7567567567567568,
-      "grad_norm": 4.212410926818848,
+      "grad_norm": 4.989683151245117,
       "learning_rate": 7.737062865111265e-06,
-      "loss": 1.2456,
+      "loss": 1.1537,
       "step": 1820
     },
     {
       "epoch": 1.7577220077220077,
-      "grad_norm": 4.991441249847412,
+      "grad_norm": 4.435257434844971,
       "learning_rate": 7.726911153466699e-06,
-      "loss": 1.1862,
+      "loss": 1.0824,
       "step": 1821
     },
     {
       "epoch": 1.7586872586872588,
-      "grad_norm": 3.9241816997528076,
+      "grad_norm": 4.071902275085449,
       "learning_rate": 7.716761911439217e-06,
-      "loss": 1.2542,
+      "loss": 1.1744,
       "step": 1822
     },
     {
       "epoch": 1.7596525096525095,
-      "grad_norm": 3.840327024459839,
+      "grad_norm": 4.728675842285156,
       "learning_rate": 7.706615150055557e-06,
-      "loss": 1.1657,
+      "loss": 1.0818,
       "step": 1823
     },
     {
       "epoch": 1.7606177606177607,
-      "grad_norm": 4.376457214355469,
+      "grad_norm": 4.96465539932251,
       "learning_rate": 7.696470880339751e-06,
-      "loss": 1.2348,
+      "loss": 1.133,
       "step": 1824
     },
     {
       "epoch": 1.7615830115830116,
-      "grad_norm": 4.3103508949279785,
+      "grad_norm": 4.410280704498291,
       "learning_rate": 7.686329113313129e-06,
-      "loss": 1.1787,
+      "loss": 1.0876,
       "step": 1825
     },
     {
       "epoch": 1.7625482625482625,
-      "grad_norm": 4.218089580535889,
+      "grad_norm": 4.113475799560547,
       "learning_rate": 7.676189859994299e-06,
-      "loss": 1.1962,
+      "loss": 1.1088,
       "step": 1826
     },
     {
       "epoch": 1.7635135135135136,
-      "grad_norm": 4.236190319061279,
+      "grad_norm": 4.467721462249756,
       "learning_rate": 7.666053131399151e-06,
-      "loss": 1.1673,
+      "loss": 1.0715,
       "step": 1827
     },
     {
       "epoch": 1.7644787644787645,
-      "grad_norm": 4.285384178161621,
+      "grad_norm": 4.452345371246338,
       "learning_rate": 7.655918938540808e-06,
-      "loss": 1.1574,
+      "loss": 1.077,
       "step": 1828
     },
     {
       "epoch": 1.7654440154440154,
-      "grad_norm": 4.414580345153809,
+      "grad_norm": 4.524196147918701,
       "learning_rate": 7.64578729242966e-06,
-      "loss": 1.1686,
+      "loss": 1.0885,
       "step": 1829
     },
     {
       "epoch": 1.7664092664092665,
-      "grad_norm": 4.828358173370361,
+      "grad_norm": 4.458215713500977,
       "learning_rate": 7.63565820407332e-06,
-      "loss": 1.2469,
+      "loss": 1.1609,
       "step": 1830
     },
     {
       "epoch": 1.7673745173745172,
-      "grad_norm": 4.314386367797852,
+      "grad_norm": 4.458557605743408,
       "learning_rate": 7.625531684476624e-06,
-      "loss": 1.1844,
+      "loss": 1.1034,
       "step": 1831
     },
     {
       "epoch": 1.7683397683397684,
-      "grad_norm": 4.553681373596191,
+      "grad_norm": 4.35870885848999,
       "learning_rate": 7.615407744641618e-06,
-      "loss": 1.1458,
+      "loss": 1.0433,
       "step": 1832
     },
     {
       "epoch": 1.7693050193050193,
-      "grad_norm": 4.936324119567871,
+      "grad_norm": 4.589725971221924,
       "learning_rate": 7.605286395567544e-06,
-      "loss": 1.2037,
+      "loss": 1.1208,
       "step": 1833
     },
     {
       "epoch": 1.7702702702702702,
-      "grad_norm": 4.830184459686279,
+      "grad_norm": 4.443562030792236,
       "learning_rate": 7.595167648250825e-06,
-      "loss": 1.2002,
+      "loss": 1.1007,
       "step": 1834
     },
     {
       "epoch": 1.7712355212355213,
-      "grad_norm": 4.841872692108154,
+      "grad_norm": 4.534771919250488,
       "learning_rate": 7.585051513685067e-06,
-      "loss": 1.2497,
+      "loss": 1.1365,
       "step": 1835
     },
     {
       "epoch": 1.7722007722007722,
-      "grad_norm": 4.141880512237549,
+      "grad_norm": 4.094799518585205,
       "learning_rate": 7.574938002861033e-06,
-      "loss": 1.2083,
+      "loss": 1.112,
       "step": 1836
     },
     {
       "epoch": 1.7731660231660231,
-      "grad_norm": 4.67802095413208,
+      "grad_norm": 4.340986251831055,
       "learning_rate": 7.5648271267666295e-06,
-      "loss": 1.2196,
+      "loss": 1.1125,
       "step": 1837
     },
     {
       "epoch": 1.7741312741312742,
-      "grad_norm": 4.651124954223633,
+      "grad_norm": 4.395476341247559,
       "learning_rate": 7.554718896386907e-06,
-      "loss": 1.1569,
+      "loss": 1.0577,
       "step": 1838
     },
     {
       "epoch": 1.775096525096525,
-      "grad_norm": 5.021517276763916,
+      "grad_norm": 5.3276495933532715,
       "learning_rate": 7.5446133227040415e-06,
-      "loss": 1.2042,
+      "loss": 1.1256,
       "step": 1839
     },
     {
       "epoch": 1.776061776061776,
-      "grad_norm": 5.392575263977051,
+      "grad_norm": 4.54580020904541,
       "learning_rate": 7.534510416697322e-06,
-      "loss": 1.2277,
+      "loss": 1.1291,
       "step": 1840
     },
     {
       "epoch": 1.777027027027027,
-      "grad_norm": 4.053293704986572,
+      "grad_norm": 4.0677361488342285,
       "learning_rate": 7.524410189343139e-06,
-      "loss": 1.1792,
+      "loss": 1.0783,
       "step": 1841
     },
     {
       "epoch": 1.7779922779922779,
-      "grad_norm": 4.763055801391602,
+      "grad_norm": 4.757392406463623,
       "learning_rate": 7.514312651614964e-06,
-      "loss": 1.209,
+      "loss": 1.1092,
       "step": 1842
     },
     {
       "epoch": 1.778957528957529,
-      "grad_norm": 4.666579723358154,
+      "grad_norm": 4.349710464477539,
       "learning_rate": 7.504217814483364e-06,
-      "loss": 1.211,
+      "loss": 1.1147,
       "step": 1843
     },
     {
       "epoch": 1.77992277992278,
-      "grad_norm": 4.091919898986816,
+      "grad_norm": 4.218860626220703,
       "learning_rate": 7.494125688915959e-06,
-      "loss": 1.2149,
+      "loss": 1.1164,
       "step": 1844
     },
     {
       "epoch": 1.7808880308880308,
-      "grad_norm": 5.1892595291137695,
+      "grad_norm": 4.540984630584717,
       "learning_rate": 7.484036285877429e-06,
-      "loss": 1.1575,
+      "loss": 1.0672,
       "step": 1845
     },
     {
       "epoch": 1.781853281853282,
-      "grad_norm": 4.62443733215332,
+      "grad_norm": 4.098612308502197,
       "learning_rate": 7.4739496163294875e-06,
-      "loss": 1.1507,
+      "loss": 1.0437,
       "step": 1846
     },
     {
       "epoch": 1.7828185328185329,
-      "grad_norm": 4.525411605834961,
+      "grad_norm": 4.540050983428955,
       "learning_rate": 7.463865691230885e-06,
-      "loss": 1.1297,
+      "loss": 1.0448,
       "step": 1847
     },
     {
       "epoch": 1.7837837837837838,
-      "grad_norm": 4.4371724128723145,
+      "grad_norm": 4.98792028427124,
       "learning_rate": 7.453784521537393e-06,
-      "loss": 1.217,
+      "loss": 1.1263,
       "step": 1848
     },
     {
       "epoch": 1.784749034749035,
-      "grad_norm": 4.973574638366699,
+      "grad_norm": 4.034694194793701,
       "learning_rate": 7.443706118201782e-06,
-      "loss": 1.207,
+      "loss": 1.1031,
       "step": 1849
     },
     {
       "epoch": 1.7857142857142856,
-      "grad_norm": 5.308924674987793,
+      "grad_norm": 4.361145973205566,
       "learning_rate": 7.433630492173831e-06,
-      "loss": 1.1676,
+      "loss": 1.0677,
       "step": 1850
     },
     {
       "epoch": 1.7866795366795367,
-      "grad_norm": 4.378430366516113,
+      "grad_norm": 4.718203067779541,
       "learning_rate": 7.423557654400275e-06,
-      "loss": 1.1859,
+      "loss": 1.092,
       "step": 1851
     },
     {
       "epoch": 1.7876447876447876,
-      "grad_norm": 5.219233512878418,
+      "grad_norm": 4.036832332611084,
       "learning_rate": 7.413487615824847e-06,
-      "loss": 1.1863,
+      "loss": 1.0898,
       "step": 1852
     },
     {
       "epoch": 1.7886100386100385,
-      "grad_norm": 4.620556831359863,
+      "grad_norm": 4.084123134613037,
       "learning_rate": 7.403420387388223e-06,
-      "loss": 1.1951,
+      "loss": 1.0888,
       "step": 1853
     },
     {
       "epoch": 1.7895752895752897,
-      "grad_norm": 4.507449626922607,
+      "grad_norm": 4.417779922485352,
       "learning_rate": 7.393355980028039e-06,
-      "loss": 1.1938,
+      "loss": 1.1139,
       "step": 1854
     },
     {
       "epoch": 1.7905405405405406,
-      "grad_norm": 4.811728000640869,
+      "grad_norm": 4.593133926391602,
       "learning_rate": 7.383294404678842e-06,
-      "loss": 1.1331,
+      "loss": 1.0397,
       "step": 1855
     },
     {
       "epoch": 1.7915057915057915,
-      "grad_norm": 4.3953375816345215,
+      "grad_norm": 4.37566614151001,
       "learning_rate": 7.373235672272132e-06,
-      "loss": 1.135,
+      "loss": 1.0375,
       "step": 1856
     },
     {
       "epoch": 1.7924710424710426,
-      "grad_norm": 5.127755165100098,
+      "grad_norm": 4.741438865661621,
       "learning_rate": 7.363179793736297e-06,
-      "loss": 1.2286,
+      "loss": 1.136,
       "step": 1857
     },
     {
       "epoch": 1.7934362934362933,
-      "grad_norm": 4.2956132888793945,
+      "grad_norm": 4.672888278961182,
       "learning_rate": 7.35312677999664e-06,
-      "loss": 1.1705,
+      "loss": 1.0729,
       "step": 1858
     },
     {
       "epoch": 1.7944015444015444,
-      "grad_norm": 4.386326313018799,
+      "grad_norm": 4.404036998748779,
       "learning_rate": 7.343076641975345e-06,
-      "loss": 1.2242,
+      "loss": 1.1051,
       "step": 1859
     },
     {
       "epoch": 1.7953667953667953,
-      "grad_norm": 4.163666725158691,
+      "grad_norm": 4.249918460845947,
       "learning_rate": 7.333029390591466e-06,
-      "loss": 1.1547,
+      "loss": 1.0548,
       "step": 1860
     },
     {
       "epoch": 1.7963320463320462,
-      "grad_norm": 4.53470516204834,
+      "grad_norm": 4.097388744354248,
       "learning_rate": 7.322985036760928e-06,
-      "loss": 1.1772,
+      "loss": 1.0665,
       "step": 1861
     },
     {
       "epoch": 1.7972972972972974,
-      "grad_norm": 4.408353805541992,
+      "grad_norm": 4.483771324157715,
       "learning_rate": 7.312943591396508e-06,
-      "loss": 1.154,
+      "loss": 1.0495,
       "step": 1862
     },
     {
       "epoch": 1.7982625482625483,
-      "grad_norm": 4.333704948425293,
+      "grad_norm": 4.476767063140869,
       "learning_rate": 7.3029050654078235e-06,
-      "loss": 1.2424,
+      "loss": 1.133,
       "step": 1863
     },
     {
       "epoch": 1.7992277992277992,
-      "grad_norm": 4.769482135772705,
+      "grad_norm": 4.328971862792969,
       "learning_rate": 7.292869469701315e-06,
-      "loss": 1.2207,
+      "loss": 1.1088,
       "step": 1864
     },
     {
       "epoch": 1.8001930501930503,
-      "grad_norm": 4.205842971801758,
+      "grad_norm": 3.9663944244384766,
       "learning_rate": 7.282836815180241e-06,
-      "loss": 1.1742,
+      "loss": 1.0826,
       "step": 1865
     },
     {
       "epoch": 1.801158301158301,
-      "grad_norm": 4.177207946777344,
+      "grad_norm": 4.498886585235596,
       "learning_rate": 7.272807112744666e-06,
-      "loss": 1.1845,
+      "loss": 1.0826,
       "step": 1866
     },
     {
       "epoch": 1.8021235521235521,
-      "grad_norm": 4.250517845153809,
+      "grad_norm": 4.45244836807251,
       "learning_rate": 7.262780373291452e-06,
-      "loss": 1.2649,
+      "loss": 1.1673,
       "step": 1867
     },
     {
       "epoch": 1.803088803088803,
-      "grad_norm": 4.70888614654541,
+      "grad_norm": 4.822850704193115,
       "learning_rate": 7.252756607714238e-06,
-      "loss": 1.1456,
+      "loss": 1.0413,
       "step": 1868
     },
     {
       "epoch": 1.804054054054054,
-      "grad_norm": 4.450895309448242,
+      "grad_norm": 4.562388896942139,
       "learning_rate": 7.242735826903422e-06,
-      "loss": 1.1862,
+      "loss": 1.0877,
       "step": 1869
     },
     {
       "epoch": 1.805019305019305,
-      "grad_norm": 4.835483074188232,
+      "grad_norm": 4.8845601081848145,
       "learning_rate": 7.2327180417461716e-06,
-      "loss": 1.1659,
+      "loss": 1.0718,
       "step": 1870
     },
     {
       "epoch": 1.805984555984556,
-      "grad_norm": 4.576693534851074,
+      "grad_norm": 4.154755592346191,
       "learning_rate": 7.222703263126402e-06,
-      "loss": 1.1301,
+      "loss": 1.0289,
       "step": 1871
     },
     {
       "epoch": 1.806949806949807,
-      "grad_norm": 4.504087924957275,
+      "grad_norm": 4.303405284881592,
       "learning_rate": 7.212691501924753e-06,
-      "loss": 1.118,
+      "loss": 1.0147,
       "step": 1872
     },
     {
       "epoch": 1.807915057915058,
-      "grad_norm": 4.55885124206543,
+      "grad_norm": 4.373027801513672,
       "learning_rate": 7.2026827690185874e-06,
-      "loss": 1.2146,
+      "loss": 1.0973,
       "step": 1873
     },
     {
       "epoch": 1.808880308880309,
-      "grad_norm": 5.224696636199951,
+      "grad_norm": 4.582155704498291,
       "learning_rate": 7.192677075281978e-06,
-      "loss": 1.1357,
+      "loss": 1.0319,
       "step": 1874
     },
     {
       "epoch": 1.8098455598455598,
-      "grad_norm": 4.089498519897461,
+      "grad_norm": 4.253002166748047,
       "learning_rate": 7.182674431585703e-06,
-      "loss": 1.2009,
+      "loss": 1.0988,
       "step": 1875
     },
     {
       "epoch": 1.810810810810811,
-      "grad_norm": 5.025662899017334,
+      "grad_norm": 3.9734129905700684,
       "learning_rate": 7.172674848797218e-06,
-      "loss": 1.2124,
+      "loss": 1.1127,
       "step": 1876
     },
     {
       "epoch": 1.8117760617760617,
-      "grad_norm": 4.546243667602539,
+      "grad_norm": 4.511058330535889,
       "learning_rate": 7.162678337780664e-06,
-      "loss": 1.2037,
+      "loss": 1.1078,
       "step": 1877
     },
     {
       "epoch": 1.8127413127413128,
-      "grad_norm": 4.327416896820068,
+      "grad_norm": 4.5919880867004395,
       "learning_rate": 7.1526849093968234e-06,
-      "loss": 1.2691,
+      "loss": 1.1658,
       "step": 1878
     },
     {
       "epoch": 1.8137065637065637,
-      "grad_norm": 4.666914939880371,
+      "grad_norm": 4.430997371673584,
       "learning_rate": 7.142694574503151e-06,
-      "loss": 1.2732,
+      "loss": 1.1628,
       "step": 1879
     },
     {
       "epoch": 1.8146718146718146,
-      "grad_norm": 4.3911333084106445,
+      "grad_norm": 4.466015815734863,
       "learning_rate": 7.132707343953732e-06,
-      "loss": 1.2423,
+      "loss": 1.1315,
       "step": 1880
     },
     {
       "epoch": 1.8156370656370657,
-      "grad_norm": 4.461322784423828,
+      "grad_norm": 4.208050727844238,
       "learning_rate": 7.1227232285992855e-06,
-      "loss": 1.1411,
+      "loss": 1.0419,
       "step": 1881
     },
     {
       "epoch": 1.8166023166023166,
-      "grad_norm": 4.10672664642334,
+      "grad_norm": 4.288403511047363,
       "learning_rate": 7.112742239287127e-06,
-      "loss": 1.2356,
+      "loss": 1.1275,
       "step": 1882
     },
     {
       "epoch": 1.8175675675675675,
-      "grad_norm": 4.363541126251221,
+      "grad_norm": 4.659818172454834,
       "learning_rate": 7.1027643868612e-06,
-      "loss": 1.1945,
+      "loss": 1.1002,
       "step": 1883
     },
     {
       "epoch": 1.8185328185328187,
-      "grad_norm": 4.521150588989258,
+      "grad_norm": 4.638820171356201,
       "learning_rate": 7.092789682162024e-06,
-      "loss": 1.24,
+      "loss": 1.1446,
       "step": 1884
     },
     {
       "epoch": 1.8194980694980694,
-      "grad_norm": 4.348193168640137,
+      "grad_norm": 4.353371620178223,
       "learning_rate": 7.082818136026706e-06,
-      "loss": 1.1318,
+      "loss": 1.0263,
       "step": 1885
     },
     {
       "epoch": 1.8204633204633205,
-      "grad_norm": 4.012368202209473,
+      "grad_norm": 4.657031059265137,
       "learning_rate": 7.072849759288914e-06,
-      "loss": 1.2535,
+      "loss": 1.1491,
       "step": 1886
     },
     {
       "epoch": 1.8214285714285714,
-      "grad_norm": 4.3002471923828125,
+      "grad_norm": 4.196220874786377,
       "learning_rate": 7.062884562778883e-06,
-      "loss": 1.1929,
+      "loss": 1.0924,
       "step": 1887
     },
     {
       "epoch": 1.8223938223938223,
-      "grad_norm": 4.529809474945068,
+      "grad_norm": 4.557463645935059,
       "learning_rate": 7.052922557323383e-06,
-      "loss": 1.1595,
+      "loss": 1.0426,
       "step": 1888
     },
     {
       "epoch": 1.8233590733590734,
-      "grad_norm": 4.302705764770508,
+      "grad_norm": 4.708265781402588,
       "learning_rate": 7.042963753745722e-06,
-      "loss": 1.2402,
+      "loss": 1.1352,
       "step": 1889
     },
     {
       "epoch": 1.8243243243243243,
-      "grad_norm": 4.311285495758057,
+      "grad_norm": 4.458983421325684,
       "learning_rate": 7.033008162865732e-06,
-      "loss": 1.1861,
+      "loss": 1.0904,
       "step": 1890
     },
     {
       "epoch": 1.8252895752895753,
-      "grad_norm": 4.44135046005249,
+      "grad_norm": 4.328872203826904,
       "learning_rate": 7.023055795499747e-06,
-      "loss": 1.2708,
+      "loss": 1.1764,
       "step": 1891
     },
     {
       "epoch": 1.8262548262548264,
-      "grad_norm": 4.508293151855469,
+      "grad_norm": 4.6338653564453125,
       "learning_rate": 7.013106662460605e-06,
-      "loss": 1.1778,
+      "loss": 1.0806,
       "step": 1892
     },
     {
       "epoch": 1.827220077220077,
-      "grad_norm": 4.0937395095825195,
+      "grad_norm": 4.135181903839111,
       "learning_rate": 7.003160774557625e-06,
-      "loss": 1.184,
+      "loss": 1.0653,
       "step": 1893
     },
     {
       "epoch": 1.8281853281853282,
-      "grad_norm": 4.576756954193115,
+      "grad_norm": 4.6565446853637695,
       "learning_rate": 6.993218142596609e-06,
-      "loss": 1.171,
+      "loss": 1.0702,
       "step": 1894
     },
     {
       "epoch": 1.829150579150579,
-      "grad_norm": 4.2449164390563965,
+      "grad_norm": 4.680014133453369,
       "learning_rate": 6.98327877737981e-06,
-      "loss": 1.1839,
+      "loss": 1.1038,
       "step": 1895
     },
     {
       "epoch": 1.83011583011583,
-      "grad_norm": 4.2238688468933105,
+      "grad_norm": 4.211447715759277,
       "learning_rate": 6.9733426897059375e-06,
-      "loss": 1.2014,
+      "loss": 1.095,
       "step": 1896
     },
     {
       "epoch": 1.8310810810810811,
-      "grad_norm": 3.9330780506134033,
+      "grad_norm": 4.234224319458008,
       "learning_rate": 6.963409890370141e-06,
-      "loss": 1.1652,
+      "loss": 1.0687,
       "step": 1897
     },
     {
       "epoch": 1.832046332046332,
-      "grad_norm": 4.291113376617432,
+      "grad_norm": 4.310862064361572,
       "learning_rate": 6.953480390164001e-06,
-      "loss": 1.1836,
+      "loss": 1.0922,
       "step": 1898
     },
     {
       "epoch": 1.833011583011583,
-      "grad_norm": 3.825181484222412,
+      "grad_norm": 4.0152740478515625,
       "learning_rate": 6.943554199875506e-06,
-      "loss": 1.1542,
+      "loss": 1.0661,
       "step": 1899
     },
     {
       "epoch": 1.833976833976834,
-      "grad_norm": 4.255434513092041,
+      "grad_norm": 4.483767032623291,
       "learning_rate": 6.93363133028905e-06,
-      "loss": 1.2163,
+      "loss": 1.1205,
       "step": 1900
     },
     {
       "epoch": 1.834942084942085,
-      "grad_norm": 4.05705451965332,
+      "grad_norm": 4.232025623321533,
       "learning_rate": 6.923711792185419e-06,
-      "loss": 1.2347,
+      "loss": 1.1266,
       "step": 1901
     },
     {
       "epoch": 1.835907335907336,
-      "grad_norm": 4.2765631675720215,
+      "grad_norm": 4.578000545501709,
       "learning_rate": 6.913795596341786e-06,
-      "loss": 1.2278,
+      "loss": 1.118,
       "step": 1902
     },
     {
       "epoch": 1.836872586872587,
-      "grad_norm": 3.9429051876068115,
+      "grad_norm": 4.123725891113281,
       "learning_rate": 6.903882753531687e-06,
-      "loss": 1.2926,
+      "loss": 1.1889,
       "step": 1903
     },
     {
       "epoch": 1.8378378378378377,
-      "grad_norm": 4.436086654663086,
+      "grad_norm": 4.445766448974609,
       "learning_rate": 6.8939732745250145e-06,
-      "loss": 1.2279,
+      "loss": 1.1193,
       "step": 1904
     },
     {
       "epoch": 1.8388030888030888,
-      "grad_norm": 4.028366565704346,
+      "grad_norm": 3.8740384578704834,
       "learning_rate": 6.884067170088007e-06,
-      "loss": 1.1804,
+      "loss": 1.0634,
       "step": 1905
     },
     {
       "epoch": 1.8397683397683398,
-      "grad_norm": 4.2818427085876465,
+      "grad_norm": 4.166007041931152,
       "learning_rate": 6.874164450983238e-06,
-      "loss": 1.1952,
+      "loss": 1.0863,
       "step": 1906
     },
     {
       "epoch": 1.8407335907335907,
-      "grad_norm": 4.631438732147217,
+      "grad_norm": 4.73367166519165,
       "learning_rate": 6.864265127969606e-06,
-      "loss": 1.2501,
+      "loss": 1.1475,
       "step": 1907
     },
     {
       "epoch": 1.8416988416988418,
-      "grad_norm": 4.176527976989746,
+      "grad_norm": 4.328803539276123,
       "learning_rate": 6.854369211802314e-06,
-      "loss": 1.2307,
+      "loss": 1.1163,
       "step": 1908
     },
     {
       "epoch": 1.8426640926640927,
-      "grad_norm": 4.664109230041504,
+      "grad_norm": 4.501825332641602,
       "learning_rate": 6.844476713232863e-06,
-      "loss": 1.1214,
+      "loss": 1.0145,
       "step": 1909
     },
     {
       "epoch": 1.8436293436293436,
-      "grad_norm": 5.156432151794434,
+      "grad_norm": 5.807786464691162,
       "learning_rate": 6.834587643009047e-06,
-      "loss": 1.2163,
+      "loss": 1.1275,
       "step": 1910
     },
     {
       "epoch": 1.8445945945945947,
-      "grad_norm": 3.8033342361450195,
+      "grad_norm": 4.774611473083496,
       "learning_rate": 6.824702011874933e-06,
-      "loss": 1.14,
+      "loss": 1.0433,
       "step": 1911
     },
     {
       "epoch": 1.8455598455598454,
-      "grad_norm": 5.293600082397461,
+      "grad_norm": 4.472540855407715,
       "learning_rate": 6.81481983057085e-06,
-      "loss": 1.1455,
+      "loss": 1.0632,
       "step": 1912
     },
     {
       "epoch": 1.8465250965250966,
-      "grad_norm": 5.740131855010986,
+      "grad_norm": 4.965783596038818,
       "learning_rate": 6.804941109833375e-06,
-      "loss": 1.2523,
+      "loss": 1.1631,
       "step": 1913
     },
     {
       "epoch": 1.8474903474903475,
-      "grad_norm": 4.465753555297852,
+      "grad_norm": 5.703163146972656,
       "learning_rate": 6.7950658603953355e-06,
-      "loss": 1.1601,
+      "loss": 1.0621,
       "step": 1914
     },
     {
       "epoch": 1.8484555984555984,
-      "grad_norm": 6.7862324714660645,
+      "grad_norm": 5.004845142364502,
       "learning_rate": 6.78519409298578e-06,
-      "loss": 1.1743,
+      "loss": 1.0757,
       "step": 1915
     },
     {
       "epoch": 1.8494208494208495,
-      "grad_norm": 5.846525192260742,
+      "grad_norm": 4.555800437927246,
       "learning_rate": 6.775325818329974e-06,
-      "loss": 1.2319,
+      "loss": 1.1218,
       "step": 1916
     },
     {
       "epoch": 1.8503861003861004,
-      "grad_norm": 5.151252269744873,
+      "grad_norm": 4.466085433959961,
       "learning_rate": 6.765461047149397e-06,
-      "loss": 1.2479,
+      "loss": 1.1413,
       "step": 1917
     },
     {
       "epoch": 1.8513513513513513,
-      "grad_norm": 6.499340057373047,
+      "grad_norm": 4.801868915557861,
       "learning_rate": 6.755599790161707e-06,
-      "loss": 1.2277,
+      "loss": 1.1178,
       "step": 1918
     },
     {
       "epoch": 1.8523166023166024,
-      "grad_norm": 4.451925754547119,
+      "grad_norm": 4.868246078491211,
       "learning_rate": 6.745742058080753e-06,
-      "loss": 1.1884,
+      "loss": 1.0826,
       "step": 1919
     },
     {
       "epoch": 1.8532818532818531,
-      "grad_norm": 6.6447319984436035,
+      "grad_norm": 4.53539514541626,
       "learning_rate": 6.735887861616555e-06,
-      "loss": 1.1883,
+      "loss": 1.096,
       "step": 1920
     },
     {
       "epoch": 1.8542471042471043,
-      "grad_norm": 4.626460075378418,
+      "grad_norm": 5.0313334465026855,
       "learning_rate": 6.726037211475296e-06,
-      "loss": 1.1998,
+      "loss": 1.1138,
       "step": 1921
     },
     {
       "epoch": 1.8552123552123552,
-      "grad_norm": 5.135448932647705,
+      "grad_norm": 4.479605674743652,
       "learning_rate": 6.716190118359293e-06,
-      "loss": 1.1687,
+      "loss": 1.0712,
       "step": 1922
     },
     {
       "epoch": 1.856177606177606,
-      "grad_norm": 5.576619625091553,
+      "grad_norm": 4.531142234802246,
       "learning_rate": 6.706346592967006e-06,
-      "loss": 1.1574,
+      "loss": 1.0552,
       "step": 1923
     },
     {
       "epoch": 1.8571428571428572,
-      "grad_norm": 5.447540760040283,
+      "grad_norm": 4.71004581451416,
       "learning_rate": 6.6965066459930195e-06,
-      "loss": 1.1473,
+      "loss": 1.0511,
       "step": 1924
     },
     {
       "epoch": 1.8581081081081081,
-      "grad_norm": 7.530019283294678,
+      "grad_norm": 4.819362640380859,
       "learning_rate": 6.686670288128032e-06,
-      "loss": 1.1573,
+      "loss": 1.0441,
       "step": 1925
     },
     {
       "epoch": 1.859073359073359,
-      "grad_norm": 6.108087539672852,
+      "grad_norm": 4.649701118469238,
       "learning_rate": 6.67683753005884e-06,
-      "loss": 1.245,
+      "loss": 1.1377,
       "step": 1926
     },
     {
       "epoch": 1.8600386100386102,
-      "grad_norm": 4.211674690246582,
+      "grad_norm": 4.213528156280518,
       "learning_rate": 6.667008382468323e-06,
-      "loss": 1.1678,
+      "loss": 1.0585,
       "step": 1927
     },
     {
       "epoch": 1.8610038610038608,
-      "grad_norm": 6.624056339263916,
+      "grad_norm": 4.4071879386901855,
       "learning_rate": 6.657182856035447e-06,
-      "loss": 1.2613,
+      "loss": 1.1415,
       "step": 1928
     },
     {
       "epoch": 1.861969111969112,
-      "grad_norm": 4.441121578216553,
+      "grad_norm": 4.086228370666504,
       "learning_rate": 6.6473609614352405e-06,
-      "loss": 1.1993,
+      "loss": 1.0829,
       "step": 1929
     },
     {
       "epoch": 1.862934362934363,
-      "grad_norm": 4.745296478271484,
+      "grad_norm": 4.687366962432861,
       "learning_rate": 6.637542709338789e-06,
-      "loss": 1.2113,
+      "loss": 1.0971,
       "step": 1930
     },
     {
       "epoch": 1.8638996138996138,
-      "grad_norm": 4.846886157989502,
+      "grad_norm": 4.519179344177246,
       "learning_rate": 6.627728110413214e-06,
-      "loss": 1.159,
+      "loss": 1.0524,
       "step": 1931
     },
     {
       "epoch": 1.864864864864865,
-      "grad_norm": 4.237720012664795,
+      "grad_norm": 4.328845500946045,
       "learning_rate": 6.617917175321668e-06,
-      "loss": 1.1422,
+      "loss": 1.0369,
       "step": 1932
     },
     {
       "epoch": 1.8658301158301158,
-      "grad_norm": 5.267391681671143,
+      "grad_norm": 4.249301433563232,
       "learning_rate": 6.6081099147233344e-06,
-      "loss": 1.2238,
+      "loss": 1.105,
       "step": 1933
     },
     {
       "epoch": 1.8667953667953667,
-      "grad_norm": 4.453833103179932,
+      "grad_norm": 4.560779094696045,
       "learning_rate": 6.598306339273393e-06,
-      "loss": 1.1406,
+      "loss": 1.031,
       "step": 1934
     },
     {
       "epoch": 1.8677606177606179,
-      "grad_norm": 5.526133060455322,
+      "grad_norm": 5.267545223236084,
       "learning_rate": 6.588506459623023e-06,
-      "loss": 1.2287,
+      "loss": 1.1139,
       "step": 1935
     },
     {
       "epoch": 1.8687258687258688,
-      "grad_norm": 4.279199600219727,
+      "grad_norm": 4.230781078338623,
       "learning_rate": 6.578710286419388e-06,
-      "loss": 1.2199,
+      "loss": 1.1102,
       "step": 1936
     },
     {
       "epoch": 1.8696911196911197,
-      "grad_norm": 4.354450225830078,
+      "grad_norm": 4.330410480499268,
       "learning_rate": 6.568917830305625e-06,
-      "loss": 1.1645,
+      "loss": 1.0479,
       "step": 1937
     },
     {
       "epoch": 1.8706563706563708,
-      "grad_norm": 5.210305690765381,
+      "grad_norm": 4.971562385559082,
       "learning_rate": 6.559129101920836e-06,
-      "loss": 1.1394,
+      "loss": 1.0441,
       "step": 1938
     },
     {
       "epoch": 1.8716216216216215,
-      "grad_norm": 4.386662483215332,
+      "grad_norm": 4.596487998962402,
       "learning_rate": 6.54934411190007e-06,
-      "loss": 1.175,
+      "loss": 1.0675,
       "step": 1939
     },
     {
       "epoch": 1.8725868725868726,
-      "grad_norm": 4.03883695602417,
+      "grad_norm": 4.320395469665527,
       "learning_rate": 6.539562870874309e-06,
-      "loss": 1.1861,
+      "loss": 1.087,
       "step": 1940
     },
     {
       "epoch": 1.8735521235521235,
-      "grad_norm": 5.33974552154541,
+      "grad_norm": 4.54929780960083,
       "learning_rate": 6.529785389470469e-06,
-      "loss": 1.1526,
+      "loss": 1.0549,
       "step": 1941
     },
     {
       "epoch": 1.8745173745173744,
-      "grad_norm": 4.374886512756348,
+      "grad_norm": 4.504304885864258,
       "learning_rate": 6.520011678311382e-06,
-      "loss": 1.2011,
+      "loss": 1.0945,
       "step": 1942
     },
     {
       "epoch": 1.8754826254826256,
-      "grad_norm": 4.7918829917907715,
+      "grad_norm": 5.361563682556152,
       "learning_rate": 6.510241748015779e-06,
-      "loss": 1.2436,
+      "loss": 1.1484,
       "step": 1943
     },
     {
       "epoch": 1.8764478764478765,
-      "grad_norm": 4.391792297363281,
+      "grad_norm": 4.176314830780029,
       "learning_rate": 6.500475609198291e-06,
-      "loss": 1.1879,
+      "loss": 1.0778,
       "step": 1944
     },
     {
       "epoch": 1.8774131274131274,
-      "grad_norm": 4.195566177368164,
+      "grad_norm": 4.257970333099365,
       "learning_rate": 6.490713272469414e-06,
-      "loss": 1.1857,
+      "loss": 1.0822,
       "step": 1945
     },
     {
       "epoch": 1.8783783783783785,
-      "grad_norm": 4.680538654327393,
+      "grad_norm": 4.881892681121826,
       "learning_rate": 6.480954748435531e-06,
-      "loss": 1.2311,
+      "loss": 1.1175,
       "step": 1946
     },
     {
       "epoch": 1.8793436293436292,
-      "grad_norm": 4.65995979309082,
+      "grad_norm": 5.3506693840026855,
       "learning_rate": 6.4712000476988715e-06,
-      "loss": 1.152,
+      "loss": 1.0414,
       "step": 1947
     },
     {
       "epoch": 1.8803088803088803,
-      "grad_norm": 4.1404829025268555,
+      "grad_norm": 4.306082248687744,
       "learning_rate": 6.4614491808575215e-06,
-      "loss": 1.215,
+      "loss": 1.1073,
       "step": 1948
     },
     {
       "epoch": 1.8812741312741312,
-      "grad_norm": 4.770052433013916,
+      "grad_norm": 4.727910995483398,
       "learning_rate": 6.451702158505384e-06,
-      "loss": 1.1552,
+      "loss": 1.0396,
       "step": 1949
     },
     {
       "epoch": 1.8822393822393821,
-      "grad_norm": 4.840427875518799,
+      "grad_norm": 4.806346416473389,
       "learning_rate": 6.4419589912322025e-06,
-      "loss": 1.2019,
+      "loss": 1.0945,
       "step": 1950
     },
     {
       "epoch": 1.8832046332046333,
-      "grad_norm": 4.4432148933410645,
+      "grad_norm": 4.897382736206055,
       "learning_rate": 6.4322196896235236e-06,
-      "loss": 1.1779,
+      "loss": 1.0617,
       "step": 1951
     },
     {
       "epoch": 1.8841698841698842,
-      "grad_norm": 4.31553316116333,
+      "grad_norm": 4.593417644500732,
       "learning_rate": 6.422484264260698e-06,
-      "loss": 1.1712,
+      "loss": 1.0832,
       "step": 1952
     },
     {
       "epoch": 1.885135135135135,
-      "grad_norm": 4.212576389312744,
+      "grad_norm": 4.377917289733887,
       "learning_rate": 6.412752725720864e-06,
-      "loss": 1.1775,
+      "loss": 1.0771,
       "step": 1953
     },
     {
       "epoch": 1.8861003861003862,
-      "grad_norm": 4.3694353103637695,
+      "grad_norm": 4.584083080291748,
       "learning_rate": 6.403025084576932e-06,
-      "loss": 1.2055,
+      "loss": 1.0994,
       "step": 1954
     },
     {
       "epoch": 1.887065637065637,
-      "grad_norm": 5.076420783996582,
+      "grad_norm": 5.19399356842041,
       "learning_rate": 6.393301351397583e-06,
-      "loss": 1.2423,
+      "loss": 1.1309,
       "step": 1955
     },
     {
       "epoch": 1.888030888030888,
-      "grad_norm": 4.373606204986572,
+      "grad_norm": 4.495306491851807,
       "learning_rate": 6.383581536747252e-06,
-      "loss": 1.2749,
+      "loss": 1.1669,
       "step": 1956
     },
     {
       "epoch": 1.8889961389961392,
-      "grad_norm": 5.308967113494873,
+      "grad_norm": 4.562784671783447,
       "learning_rate": 6.3738656511861195e-06,
-      "loss": 1.1588,
+      "loss": 1.052,
       "step": 1957
     },
     {
       "epoch": 1.8899613899613898,
-      "grad_norm": 4.475518703460693,
+      "grad_norm": 4.603917121887207,
       "learning_rate": 6.364153705270088e-06,
-      "loss": 1.1719,
+      "loss": 1.0678,
       "step": 1958
     },
     {
       "epoch": 1.890926640926641,
-      "grad_norm": 4.570695877075195,
+      "grad_norm": 4.6627278327941895,
       "learning_rate": 6.3544457095507885e-06,
-      "loss": 1.1194,
+      "loss": 1.0083,
       "step": 1959
     },
     {
       "epoch": 1.8918918918918919,
-      "grad_norm": 4.763030529022217,
+      "grad_norm": 4.266762733459473,
       "learning_rate": 6.344741674575553e-06,
-      "loss": 1.1407,
+      "loss": 1.0299,
       "step": 1960
     },
     {
       "epoch": 1.8928571428571428,
-      "grad_norm": 4.580066680908203,
+      "grad_norm": 4.794391632080078,
       "learning_rate": 6.335041610887422e-06,
-      "loss": 1.188,
+      "loss": 1.0786,
       "step": 1961
     },
     {
       "epoch": 1.893822393822394,
-      "grad_norm": 4.779040336608887,
+      "grad_norm": 4.379703044891357,
       "learning_rate": 6.325345529025107e-06,
-      "loss": 1.1896,
+      "loss": 1.0762,
       "step": 1962
     },
     {
       "epoch": 1.8947876447876448,
-      "grad_norm": 4.4460129737854,
+      "grad_norm": 3.976919174194336,
       "learning_rate": 6.315653439523003e-06,
-      "loss": 1.1016,
+      "loss": 0.989,
       "step": 1963
     },
     {
       "epoch": 1.8957528957528957,
-      "grad_norm": 4.613245487213135,
+      "grad_norm": 4.776845455169678,
       "learning_rate": 6.305965352911162e-06,
-      "loss": 1.1686,
+      "loss": 1.0665,
       "step": 1964
     },
     {
       "epoch": 1.8967181467181469,
-      "grad_norm": 5.302827835083008,
+      "grad_norm": 4.363754749298096,
       "learning_rate": 6.296281279715294e-06,
-      "loss": 1.1598,
+      "loss": 1.047,
       "step": 1965
     },
     {
       "epoch": 1.8976833976833976,
-      "grad_norm": 4.315168380737305,
+      "grad_norm": 4.491544246673584,
       "learning_rate": 6.2866012304567436e-06,
-      "loss": 1.1639,
+      "loss": 1.0628,
       "step": 1966
     },
     {
       "epoch": 1.8986486486486487,
-      "grad_norm": 5.385082721710205,
+      "grad_norm": 4.410923004150391,
       "learning_rate": 6.276925215652482e-06,
-      "loss": 1.2059,
+      "loss": 1.1032,
       "step": 1967
     },
     {
       "epoch": 1.8996138996138996,
-      "grad_norm": 3.810722589492798,
+      "grad_norm": 4.096185684204102,
       "learning_rate": 6.267253245815098e-06,
-      "loss": 1.214,
+      "loss": 1.1023,
       "step": 1968
     },
     {
       "epoch": 1.9005791505791505,
-      "grad_norm": 4.369156360626221,
+      "grad_norm": 4.966404438018799,
       "learning_rate": 6.257585331452791e-06,
-      "loss": 1.2251,
+      "loss": 1.1173,
       "step": 1969
     },
     {
       "epoch": 1.9015444015444016,
-      "grad_norm": 4.052249908447266,
+      "grad_norm": 4.253550052642822,
       "learning_rate": 6.247921483069352e-06,
-      "loss": 1.2096,
+      "loss": 1.0996,
       "step": 1970
     },
     {
       "epoch": 1.9025096525096525,
-      "grad_norm": 4.842185974121094,
+      "grad_norm": 4.91128396987915,
       "learning_rate": 6.238261711164146e-06,
-      "loss": 1.1637,
+      "loss": 1.0676,
       "step": 1971
     },
     {
       "epoch": 1.9034749034749034,
-      "grad_norm": 4.0050787925720215,
+      "grad_norm": 4.929739952087402,
       "learning_rate": 6.228606026232118e-06,
-      "loss": 1.2152,
+      "loss": 1.1175,
       "step": 1972
     },
     {
       "epoch": 1.9044401544401546,
-      "grad_norm": 4.679233074188232,
+      "grad_norm": 4.338520526885986,
       "learning_rate": 6.218954438763773e-06,
-      "loss": 1.1746,
+      "loss": 1.0498,
       "step": 1973
     },
     {
       "epoch": 1.9054054054054053,
-      "grad_norm": 4.082679748535156,
+      "grad_norm": 4.423638820648193,
       "learning_rate": 6.209306959245161e-06,
-      "loss": 1.1686,
+      "loss": 1.0472,
       "step": 1974
     },
     {
       "epoch": 1.9063706563706564,
-      "grad_norm": 4.452788352966309,
+      "grad_norm": 4.1862711906433105,
       "learning_rate": 6.1996635981578755e-06,
-      "loss": 1.1397,
+      "loss": 1.0436,
       "step": 1975
     },
     {
       "epoch": 1.9073359073359073,
-      "grad_norm": 4.730610370635986,
+      "grad_norm": 5.296685218811035,
       "learning_rate": 6.190024365979017e-06,
-      "loss": 1.2131,
+      "loss": 1.0977,
       "step": 1976
     },
     {
       "epoch": 1.9083011583011582,
-      "grad_norm": 5.205382347106934,
+      "grad_norm": 5.721356391906738,
       "learning_rate": 6.180389273181224e-06,
-      "loss": 1.1978,
+      "loss": 1.0901,
       "step": 1977
     },
     {
       "epoch": 1.9092664092664093,
-      "grad_norm": 4.563416481018066,
+      "grad_norm": 4.390848636627197,
       "learning_rate": 6.170758330232621e-06,
-      "loss": 1.2212,
+      "loss": 1.099,
       "step": 1978
     },
     {
       "epoch": 1.9102316602316602,
-      "grad_norm": 4.994668483734131,
+      "grad_norm": 4.841115474700928,
       "learning_rate": 6.161131547596833e-06,
-      "loss": 1.1968,
+      "loss": 1.0837,
       "step": 1979
     },
     {
       "epoch": 1.9111969111969112,
-      "grad_norm": 4.848903179168701,
+      "grad_norm": 5.371408939361572,
       "learning_rate": 6.1515089357329565e-06,
-      "loss": 1.2414,
+      "loss": 1.1327,
       "step": 1980
     },
     {
       "epoch": 1.9121621621621623,
-      "grad_norm": 4.824883937835693,
+      "grad_norm": 5.097964763641357,
       "learning_rate": 6.141890505095566e-06,
-      "loss": 1.1529,
+      "loss": 1.0462,
       "step": 1981
     },
     {
       "epoch": 1.913127413127413,
-      "grad_norm": 4.200015068054199,
+      "grad_norm": 4.742908954620361,
       "learning_rate": 6.1322762661346865e-06,
-      "loss": 1.2305,
+      "loss": 1.1333,
       "step": 1982
     },
     {
       "epoch": 1.914092664092664,
-      "grad_norm": 4.15867280960083,
+      "grad_norm": 4.532333850860596,
       "learning_rate": 6.122666229295792e-06,
-      "loss": 1.2336,
+      "loss": 1.105,
       "step": 1983
     },
     {
       "epoch": 1.915057915057915,
-      "grad_norm": 4.84639835357666,
+      "grad_norm": 4.3444695472717285,
       "learning_rate": 6.113060405019794e-06,
-      "loss": 1.1236,
+      "loss": 1.0204,
       "step": 1984
     },
     {
       "epoch": 1.916023166023166,
-      "grad_norm": 4.204685688018799,
+      "grad_norm": 4.432525634765625,
       "learning_rate": 6.103458803743018e-06,
-      "loss": 1.1784,
+      "loss": 1.0711,
       "step": 1985
     },
     {
       "epoch": 1.916988416988417,
-      "grad_norm": 4.868194580078125,
+      "grad_norm": 4.4756927490234375,
       "learning_rate": 6.093861435897208e-06,
-      "loss": 1.1808,
+      "loss": 1.0781,
       "step": 1986
     },
     {
       "epoch": 1.917953667953668,
-      "grad_norm": 4.526357173919678,
+      "grad_norm": 4.369234561920166,
       "learning_rate": 6.0842683119095095e-06,
-      "loss": 1.1911,
+      "loss": 1.0776,
       "step": 1987
     },
     {
       "epoch": 1.9189189189189189,
-      "grad_norm": 5.651543617248535,
+      "grad_norm": 4.778226852416992,
       "learning_rate": 6.0746794422024555e-06,
-      "loss": 1.2539,
+      "loss": 1.1376,
       "step": 1988
     },
     {
       "epoch": 1.91988416988417,
-      "grad_norm": 5.241189002990723,
+      "grad_norm": 4.648364543914795,
       "learning_rate": 6.065094837193954e-06,
-      "loss": 1.1517,
+      "loss": 1.0411,
       "step": 1989
     },
     {
       "epoch": 1.920849420849421,
-      "grad_norm": 4.834219455718994,
+      "grad_norm": 4.150259017944336,
       "learning_rate": 6.055514507297284e-06,
-      "loss": 1.1307,
+      "loss": 1.0001,
       "step": 1990
     },
     {
       "epoch": 1.9218146718146718,
-      "grad_norm": 5.599984645843506,
+      "grad_norm": 4.650816440582275,
       "learning_rate": 6.045938462921077e-06,
-      "loss": 1.2073,
+      "loss": 1.0933,
       "step": 1991
     },
     {
       "epoch": 1.922779922779923,
-      "grad_norm": 4.227385520935059,
+      "grad_norm": 4.52701473236084,
       "learning_rate": 6.03636671446931e-06,
-      "loss": 1.2211,
+      "loss": 1.1314,
       "step": 1992
     },
     {
       "epoch": 1.9237451737451736,
-      "grad_norm": 6.122756004333496,
+      "grad_norm": 4.560050964355469,
       "learning_rate": 6.026799272341296e-06,
-      "loss": 1.1712,
+      "loss": 1.0668,
       "step": 1993
     },
     {
       "epoch": 1.9247104247104247,
-      "grad_norm": 4.500845909118652,
+      "grad_norm": 4.216585636138916,
       "learning_rate": 6.0172361469316585e-06,
-      "loss": 1.1913,
+      "loss": 1.0757,
       "step": 1994
     },
     {
       "epoch": 1.9256756756756757,
-      "grad_norm": 5.436099052429199,
+      "grad_norm": 4.624998092651367,
       "learning_rate": 6.00767734863034e-06,
-      "loss": 1.0837,
+      "loss": 0.9819,
       "step": 1995
     },
     {
       "epoch": 1.9266409266409266,
-      "grad_norm": 5.637316703796387,
+      "grad_norm": 4.294211387634277,
       "learning_rate": 5.998122887822582e-06,
-      "loss": 1.188,
+      "loss": 1.0733,
       "step": 1996
     },
     {
       "epoch": 1.9276061776061777,
-      "grad_norm": 4.430753231048584,
+      "grad_norm": 4.429724216461182,
       "learning_rate": 5.988572774888913e-06,
-      "loss": 1.1335,
+      "loss": 1.0251,
       "step": 1997
     },
     {
       "epoch": 1.9285714285714286,
-      "grad_norm": 6.754420757293701,
+      "grad_norm": 4.24937105178833,
       "learning_rate": 5.979027020205131e-06,
-      "loss": 1.2139,
+      "loss": 1.085,
       "step": 1998
     },
     {
       "epoch": 1.9295366795366795,
-      "grad_norm": 4.46309232711792,
+      "grad_norm": 4.351349353790283,
       "learning_rate": 5.9694856341423045e-06,
-      "loss": 1.1497,
+      "loss": 1.0416,
       "step": 1999
     },
     {
       "epoch": 1.9305019305019306,
-      "grad_norm": 5.311478137969971,
+      "grad_norm": 4.381855487823486,
       "learning_rate": 5.9599486270667595e-06,
-      "loss": 1.2128,
+      "loss": 1.1049,
       "step": 2000
     },
     {
       "epoch": 1.9314671814671813,
-      "grad_norm": 5.918153762817383,
+      "grad_norm": 4.350614547729492,
       "learning_rate": 5.950416009340056e-06,
-      "loss": 1.233,
+      "loss": 1.1124,
       "step": 2001
     },
     {
       "epoch": 1.9324324324324325,
-      "grad_norm": 5.034909248352051,
+      "grad_norm": 4.825096607208252,
       "learning_rate": 5.940887791318993e-06,
-      "loss": 1.2033,
+      "loss": 1.0975,
       "step": 2002
     },
     {
       "epoch": 1.9333976833976834,
-      "grad_norm": 5.40713357925415,
+      "grad_norm": 4.554704189300537,
       "learning_rate": 5.93136398335558e-06,
-      "loss": 1.266,
+      "loss": 1.1428,
       "step": 2003
     },
     {
       "epoch": 1.9343629343629343,
-      "grad_norm": 6.276338577270508,
+      "grad_norm": 4.309081077575684,
       "learning_rate": 5.9218445957970435e-06,
-      "loss": 1.2289,
+      "loss": 1.1168,
       "step": 2004
     },
     {
       "epoch": 1.9353281853281854,
-      "grad_norm": 4.238287448883057,
+      "grad_norm": 4.748154640197754,
       "learning_rate": 5.912329638985808e-06,
-      "loss": 1.1408,
+      "loss": 1.0159,
       "step": 2005
     },
     {
       "epoch": 1.9362934362934363,
-      "grad_norm": 7.066354274749756,
+      "grad_norm": 4.344897747039795,
       "learning_rate": 5.902819123259478e-06,
-      "loss": 1.1703,
+      "loss": 1.0384,
       "step": 2006
     },
     {
       "epoch": 1.9372586872586872,
-      "grad_norm": 5.24070405960083,
+      "grad_norm": 4.654004096984863,
       "learning_rate": 5.8933130589508314e-06,
-      "loss": 1.2035,
+      "loss": 1.1039,
       "step": 2007
     },
     {
       "epoch": 1.9382239382239383,
-      "grad_norm": 4.5445404052734375,
+      "grad_norm": 5.057248592376709,
       "learning_rate": 5.883811456387821e-06,
-      "loss": 1.2079,
+      "loss": 1.0903,
       "step": 2008
     },
     {
       "epoch": 1.939189189189189,
-      "grad_norm": 7.246978759765625,
+      "grad_norm": 4.645206451416016,
       "learning_rate": 5.8743143258935435e-06,
-      "loss": 1.2059,
+      "loss": 1.0888,
       "step": 2009
     },
     {
       "epoch": 1.9401544401544402,
-      "grad_norm": 4.268136978149414,
+      "grad_norm": 4.048132419586182,
       "learning_rate": 5.8648216777862345e-06,
-      "loss": 1.1996,
+      "loss": 1.0919,
       "step": 2010
     },
     {
       "epoch": 1.941119691119691,
-      "grad_norm": 5.60009241104126,
+      "grad_norm": 4.1368021965026855,
       "learning_rate": 5.855333522379272e-06,
-      "loss": 1.2054,
+      "loss": 1.0857,
       "step": 2011
     },
     {
       "epoch": 1.942084942084942,
-      "grad_norm": 5.696875095367432,
+      "grad_norm": 4.471269607543945,
       "learning_rate": 5.845849869981137e-06,
-      "loss": 1.2104,
+      "loss": 1.0896,
       "step": 2012
     },
     {
       "epoch": 1.943050193050193,
-      "grad_norm": 4.394375324249268,
+      "grad_norm": 4.721922397613525,
       "learning_rate": 5.836370730895427e-06,
-      "loss": 1.1857,
+      "loss": 1.0844,
       "step": 2013
     },
     {
       "epoch": 1.944015444015444,
-      "grad_norm": 5.050985336303711,
+      "grad_norm": 4.639439105987549,
       "learning_rate": 5.8268961154208395e-06,
-      "loss": 1.2082,
+      "loss": 1.0797,
       "step": 2014
     },
     {
       "epoch": 1.944980694980695,
-      "grad_norm": 4.772952556610107,
+      "grad_norm": 4.050045490264893,
       "learning_rate": 5.817426033851152e-06,
-      "loss": 1.211,
+      "loss": 1.0925,
       "step": 2015
     },
     {
       "epoch": 1.945945945945946,
-      "grad_norm": 4.337421894073486,
+      "grad_norm": 4.1329731941223145,
       "learning_rate": 5.8079604964752126e-06,
-      "loss": 1.1849,
+      "loss": 1.0679,
       "step": 2016
     },
     {
       "epoch": 1.946911196911197,
-      "grad_norm": 5.666919231414795,
+      "grad_norm": 4.735876083374023,
       "learning_rate": 5.798499513576934e-06,
-      "loss": 1.2063,
+      "loss": 1.0947,
       "step": 2017
     },
     {
       "epoch": 1.9478764478764479,
-      "grad_norm": 4.089428424835205,
+      "grad_norm": 4.149044513702393,
       "learning_rate": 5.7890430954352895e-06,
-      "loss": 1.2211,
+      "loss": 1.0963,
       "step": 2018
     },
     {
       "epoch": 1.948841698841699,
-      "grad_norm": 5.097423553466797,
+      "grad_norm": 4.226522922515869,
       "learning_rate": 5.779591252324286e-06,
-      "loss": 1.1857,
+      "loss": 1.0751,
       "step": 2019
     },
     {
       "epoch": 1.9498069498069497,
-      "grad_norm": 4.608874797821045,
+      "grad_norm": 4.186374187469482,
       "learning_rate": 5.7701439945129625e-06,
-      "loss": 1.1794,
+      "loss": 1.0577,
       "step": 2020
     },
     {
       "epoch": 1.9507722007722008,
-      "grad_norm": 4.127477169036865,
+      "grad_norm": 4.299487590789795,
       "learning_rate": 5.7607013322653635e-06,
-      "loss": 1.2072,
+      "loss": 1.0893,
       "step": 2021
     },
     {
       "epoch": 1.9517374517374517,
-      "grad_norm": 5.385368824005127,
+      "grad_norm": 4.706480979919434,
       "learning_rate": 5.751263275840562e-06,
-      "loss": 1.1942,
+      "loss": 1.0757,
       "step": 2022
     },
     {
       "epoch": 1.9527027027027026,
-      "grad_norm": 4.392673492431641,
+      "grad_norm": 4.4907989501953125,
       "learning_rate": 5.741829835492611e-06,
-      "loss": 1.1737,
+      "loss": 1.0678,
       "step": 2023
     },
     {
       "epoch": 1.9536679536679538,
-      "grad_norm": 4.49373197555542,
+      "grad_norm": 4.749791622161865,
       "learning_rate": 5.73240102147056e-06,
-      "loss": 1.1957,
+      "loss": 1.0819,
       "step": 2024
     },
     {
       "epoch": 1.9546332046332047,
-      "grad_norm": 5.873342037200928,
+      "grad_norm": 4.52852725982666,
       "learning_rate": 5.722976844018414e-06,
-      "loss": 1.1902,
+      "loss": 1.0558,
       "step": 2025
     },
     {
       "epoch": 1.9555984555984556,
-      "grad_norm": 4.4943766593933105,
+      "grad_norm": 4.874543190002441,
       "learning_rate": 5.713557313375162e-06,
-      "loss": 1.1804,
+      "loss": 1.0757,
       "step": 2026
     },
     {
       "epoch": 1.9565637065637067,
-      "grad_norm": 4.722905158996582,
+      "grad_norm": 5.421721935272217,
       "learning_rate": 5.704142439774732e-06,
-      "loss": 1.2106,
+      "loss": 1.1061,
       "step": 2027
     },
     {
       "epoch": 1.9575289575289574,
-      "grad_norm": 5.092207908630371,
+      "grad_norm": 4.889191150665283,
       "learning_rate": 5.694732233445996e-06,
-      "loss": 1.179,
+      "loss": 1.0551,
       "step": 2028
     },
     {
       "epoch": 1.9584942084942085,
-      "grad_norm": 4.395949363708496,
+      "grad_norm": 4.305527210235596,
       "learning_rate": 5.685326704612753e-06,
-      "loss": 1.2379,
+      "loss": 1.1229,
       "step": 2029
     },
     {
       "epoch": 1.9594594594594594,
-      "grad_norm": 4.497315406799316,
+      "grad_norm": 5.3912353515625,
       "learning_rate": 5.675925863493721e-06,
-      "loss": 1.154,
+      "loss": 1.0365,
       "step": 2030
     },
     {
       "epoch": 1.9604247104247103,
-      "grad_norm": 5.878604412078857,
+      "grad_norm": 5.096640586853027,
       "learning_rate": 5.666529720302525e-06,
-      "loss": 1.1275,
+      "loss": 1.0153,
       "step": 2031
     },
     {
       "epoch": 1.9613899613899615,
-      "grad_norm": 4.746928691864014,
+      "grad_norm": 4.605041980743408,
       "learning_rate": 5.657138285247687e-06,
-      "loss": 1.2247,
+      "loss": 1.1063,
       "step": 2032
     },
     {
       "epoch": 1.9623552123552124,
-      "grad_norm": 4.125810623168945,
+      "grad_norm": 5.023714065551758,
       "learning_rate": 5.647751568532616e-06,
-      "loss": 1.2614,
+      "loss": 1.1451,
       "step": 2033
     },
     {
       "epoch": 1.9633204633204633,
-      "grad_norm": 5.916082859039307,
+      "grad_norm": 4.710081100463867,
       "learning_rate": 5.638369580355587e-06,
-      "loss": 1.2408,
+      "loss": 1.1228,
       "step": 2034
     },
     {
       "epoch": 1.9642857142857144,
-      "grad_norm": 5.335667610168457,
+      "grad_norm": 4.639779567718506,
       "learning_rate": 5.628992330909744e-06,
-      "loss": 1.2014,
+      "loss": 1.0829,
       "step": 2035
     },
     {
       "epoch": 1.965250965250965,
-      "grad_norm": 5.12222957611084,
+      "grad_norm": 4.128330230712891,
       "learning_rate": 5.619619830383083e-06,
-      "loss": 1.237,
+      "loss": 1.1126,
       "step": 2036
     },
     {
       "epoch": 1.9662162162162162,
-      "grad_norm": 4.592029094696045,
+      "grad_norm": 4.8176679611206055,
       "learning_rate": 5.610252088958434e-06,
-      "loss": 1.2112,
+      "loss": 1.0951,
       "step": 2037
     },
     {
       "epoch": 1.9671814671814671,
-      "grad_norm": 4.4699225425720215,
+      "grad_norm": 5.221687316894531,
       "learning_rate": 5.600889116813466e-06,
-      "loss": 1.2412,
+      "loss": 1.1107,
       "step": 2038
     },
     {
       "epoch": 1.968146718146718,
-      "grad_norm": 4.312219142913818,
+      "grad_norm": 4.201988697052002,
       "learning_rate": 5.591530924120659e-06,
-      "loss": 1.1981,
+      "loss": 1.103,
       "step": 2039
     },
     {
       "epoch": 1.9691119691119692,
-      "grad_norm": 5.5404815673828125,
+      "grad_norm": 4.243892669677734,
       "learning_rate": 5.5821775210473015e-06,
-      "loss": 1.1716,
+      "loss": 1.0408,
       "step": 2040
     },
     {
       "epoch": 1.97007722007722,
-      "grad_norm": 4.853655815124512,
+      "grad_norm": 4.15623664855957,
       "learning_rate": 5.5728289177554805e-06,
-      "loss": 1.0895,
+      "loss": 0.9745,
       "step": 2041
     },
     {
       "epoch": 1.971042471042471,
-      "grad_norm": 4.7461419105529785,
+      "grad_norm": 4.335485935211182,
       "learning_rate": 5.563485124402068e-06,
-      "loss": 1.2178,
+      "loss": 1.0958,
       "step": 2042
     },
     {
       "epoch": 1.9720077220077221,
-      "grad_norm": 5.6155595779418945,
+      "grad_norm": 4.681387901306152,
       "learning_rate": 5.554146151138706e-06,
-      "loss": 1.2066,
+      "loss": 1.0911,
       "step": 2043
     },
     {
       "epoch": 1.972972972972973,
-      "grad_norm": 5.168851375579834,
+      "grad_norm": 5.217894554138184,
       "learning_rate": 5.5448120081118085e-06,
-      "loss": 1.1773,
+      "loss": 1.0577,
       "step": 2044
     },
     {
       "epoch": 1.973938223938224,
-      "grad_norm": 5.844036102294922,
+      "grad_norm": 4.054657459259033,
       "learning_rate": 5.53548270546253e-06,
-      "loss": 1.1823,
+      "loss": 1.0657,
       "step": 2045
     },
     {
       "epoch": 1.974903474903475,
-      "grad_norm": 6.1245341300964355,
+      "grad_norm": 4.653079986572266,
       "learning_rate": 5.526158253326777e-06,
-      "loss": 1.2009,
+      "loss": 1.0815,
       "step": 2046
     },
     {
       "epoch": 1.9758687258687258,
-      "grad_norm": 4.416387557983398,
+      "grad_norm": 4.764430999755859,
       "learning_rate": 5.5168386618351785e-06,
-      "loss": 1.2369,
+      "loss": 1.1114,
       "step": 2047
     },
     {
       "epoch": 1.9768339768339769,
-      "grad_norm": 6.00404167175293,
+      "grad_norm": 4.4174299240112305,
       "learning_rate": 5.507523941113086e-06,
-      "loss": 1.1976,
+      "loss": 1.0906,
       "step": 2048
     },
     {
       "epoch": 1.9777992277992278,
-      "grad_norm": 7.08099365234375,
+      "grad_norm": 4.760361194610596,
       "learning_rate": 5.498214101280561e-06,
-      "loss": 1.1504,
+      "loss": 1.0267,
       "step": 2049
     },
     {
       "epoch": 1.9787644787644787,
-      "grad_norm": 4.339968204498291,
+      "grad_norm": 5.082463264465332,
       "learning_rate": 5.488909152452357e-06,
-      "loss": 1.1816,
+      "loss": 1.0851,
       "step": 2050
     },
     {
       "epoch": 1.9797297297297298,
-      "grad_norm": 6.9030537605285645,
+      "grad_norm": 4.498269557952881,
       "learning_rate": 5.479609104737917e-06,
-      "loss": 1.2805,
+      "loss": 1.158,
       "step": 2051
     },
     {
       "epoch": 1.9806949806949807,
-      "grad_norm": 6.546342372894287,
+      "grad_norm": 4.259524822235107,
       "learning_rate": 5.4703139682413585e-06,
-      "loss": 1.1604,
+      "loss": 1.0315,
       "step": 2052
     },
     {
       "epoch": 1.9816602316602316,
-      "grad_norm": 4.142830848693848,
+      "grad_norm": 5.131359577178955,
       "learning_rate": 5.461023753061463e-06,
-      "loss": 1.1628,
+      "loss": 1.0627,
       "step": 2053
     },
     {
       "epoch": 1.9826254826254828,
-      "grad_norm": 6.560509204864502,
+      "grad_norm": 5.326815128326416,
       "learning_rate": 5.451738469291666e-06,
-      "loss": 1.1804,
+      "loss": 1.0647,
       "step": 2054
     },
     {
       "epoch": 1.9835907335907335,
-      "grad_norm": 6.8526811599731445,
+      "grad_norm": 4.53480863571167,
       "learning_rate": 5.442458127020045e-06,
-      "loss": 1.1882,
+      "loss": 1.068,
       "step": 2055
     },
     {
       "epoch": 1.9845559845559846,
-      "grad_norm": 4.52562952041626,
+      "grad_norm": 4.714165687561035,
       "learning_rate": 5.433182736329305e-06,
-      "loss": 1.1942,
+      "loss": 1.0896,
       "step": 2056
     },
     {
       "epoch": 1.9855212355212355,
-      "grad_norm": 6.92567777633667,
+      "grad_norm": 4.919694900512695,
       "learning_rate": 5.423912307296779e-06,
-      "loss": 1.1957,
+      "loss": 1.0701,
       "step": 2057
     },
     {
       "epoch": 1.9864864864864864,
-      "grad_norm": 5.3068366050720215,
+      "grad_norm": 4.702139377593994,
       "learning_rate": 5.414646849994403e-06,
-      "loss": 1.1577,
+      "loss": 1.0444,
       "step": 2058
     },
     {
       "epoch": 1.9874517374517375,
-      "grad_norm": 4.130660057067871,
+      "grad_norm": 4.418658256530762,
       "learning_rate": 5.405386374488709e-06,
-      "loss": 1.1749,
+      "loss": 1.0648,
       "step": 2059
     },
     {
       "epoch": 1.9884169884169884,
-      "grad_norm": 4.9182209968566895,
+      "grad_norm": 4.333770275115967,
       "learning_rate": 5.396130890840834e-06,
-      "loss": 1.2134,
+      "loss": 1.0981,
       "step": 2060
     },
     {
       "epoch": 1.9893822393822393,
-      "grad_norm": 4.714017391204834,
+      "grad_norm": 5.3206939697265625,
       "learning_rate": 5.386880409106465e-06,
-      "loss": 1.1217,
+      "loss": 1.0144,
       "step": 2061
     },
     {
       "epoch": 1.9903474903474905,
-      "grad_norm": 4.295403480529785,
+      "grad_norm": 4.526954174041748,
       "learning_rate": 5.377634939335874e-06,
-      "loss": 1.1718,
+      "loss": 1.0656,
       "step": 2062
     },
     {
       "epoch": 1.9913127413127412,
-      "grad_norm": 4.523777484893799,
+      "grad_norm": 4.345027923583984,
       "learning_rate": 5.368394491573876e-06,
-      "loss": 1.2635,
+      "loss": 1.1521,
       "step": 2063
     },
     {
       "epoch": 1.9922779922779923,
-      "grad_norm": 5.195492744445801,
+      "grad_norm": 4.415921688079834,
       "learning_rate": 5.359159075859848e-06,
-      "loss": 1.1536,
+      "loss": 1.0323,
       "step": 2064
     },
     {
       "epoch": 1.9932432432432432,
-      "grad_norm": 3.8529839515686035,
+      "grad_norm": 4.93175745010376,
       "learning_rate": 5.349928702227677e-06,
-      "loss": 1.1507,
+      "loss": 1.0346,
       "step": 2065
     },
     {
       "epoch": 1.994208494208494,
-      "grad_norm": 4.830549716949463,
+      "grad_norm": 4.735451698303223,
       "learning_rate": 5.3407033807057855e-06,
-      "loss": 1.1928,
+      "loss": 1.0931,
       "step": 2066
     },
     {
       "epoch": 1.9951737451737452,
-      "grad_norm": 4.1817402839660645,
+      "grad_norm": 4.438542366027832,
       "learning_rate": 5.331483121317098e-06,
-      "loss": 1.1438,
+      "loss": 1.0427,
       "step": 2067
     },
     {
       "epoch": 1.9961389961389961,
-      "grad_norm": 4.0853495597839355,
+      "grad_norm": 4.329216957092285,
       "learning_rate": 5.322267934079056e-06,
-      "loss": 1.2292,
+      "loss": 1.1116,
       "step": 2068
     },
     {
       "epoch": 1.997104247104247,
-      "grad_norm": 4.8644866943359375,
+      "grad_norm": 4.581772327423096,
       "learning_rate": 5.313057829003578e-06,
-      "loss": 1.2231,
+      "loss": 1.0963,
       "step": 2069
     },
     {
       "epoch": 1.9980694980694982,
-      "grad_norm": 4.481793403625488,
+      "grad_norm": 5.217652320861816,
       "learning_rate": 5.303852816097057e-06,
-      "loss": 1.1979,
+      "loss": 1.0783,
       "step": 2070
     },
     {
       "epoch": 1.9990347490347489,
-      "grad_norm": 4.819639205932617,
+      "grad_norm": 4.746365070343018,
       "learning_rate": 5.294652905360356e-06,
-      "loss": 1.2558,
+      "loss": 1.1282,
       "step": 2071
     },
     {
       "epoch": 2.0,
-      "grad_norm": 4.601913928985596,
+      "grad_norm": 3.6739749908447266,
       "learning_rate": 5.285458106788808e-06,
-      "loss": 1.0367,
+      "loss": 0.869,
       "step": 2072
     },
     {
       "epoch": 2.000965250965251,
-      "grad_norm": 3.6043078899383545,
+      "grad_norm": 3.7369914054870605,
       "learning_rate": 5.27626843037218e-06,
-      "loss": 0.7693,
+      "loss": 0.6528,
       "step": 2073
     },
     {
       "epoch": 2.001930501930502,
-      "grad_norm": 3.638916015625,
+      "grad_norm": 4.049583435058594,
       "learning_rate": 5.267083886094668e-06,
-      "loss": 0.7657,
+      "loss": 0.6775,
       "step": 2074
     },
     {
       "epoch": 2.002895752895753,
-      "grad_norm": 4.656620025634766,
+      "grad_norm": 3.419797897338867,
       "learning_rate": 5.2579044839349e-06,
-      "loss": 0.7083,
+      "loss": 0.637,
       "step": 2075
     },
     {
       "epoch": 2.003861003861004,
-      "grad_norm": 3.8113861083984375,
+      "grad_norm": 3.5364811420440674,
       "learning_rate": 5.248730233865926e-06,
-      "loss": 0.7342,
+      "loss": 0.6611,
       "step": 2076
     },
     {
       "epoch": 2.0048262548262548,
-      "grad_norm": 4.38107442855835,
+      "grad_norm": 3.851083993911743,
       "learning_rate": 5.239561145855183e-06,
-      "loss": 0.6649,
+      "loss": 0.5855,
       "step": 2077
     },
     {
       "epoch": 2.005791505791506,
-      "grad_norm": 5.1227498054504395,
+      "grad_norm": 3.6916134357452393,
       "learning_rate": 5.230397229864515e-06,
-      "loss": 0.694,
+      "loss": 0.5965,
       "step": 2078
     },
     {
       "epoch": 2.0067567567567566,
-      "grad_norm": 5.051416873931885,
+      "grad_norm": 3.5625970363616943,
       "learning_rate": 5.221238495850121e-06,
-      "loss": 0.7027,
+      "loss": 0.5987,
       "step": 2079
     },
     {
       "epoch": 2.0077220077220077,
-      "grad_norm": 5.646663665771484,
+      "grad_norm": 4.052452087402344,
       "learning_rate": 5.212084953762599e-06,
-      "loss": 0.7197,
+      "loss": 0.6198,
       "step": 2080
     },
     {
       "epoch": 2.008687258687259,
-      "grad_norm": 5.704290390014648,
+      "grad_norm": 4.28278112411499,
       "learning_rate": 5.202936613546894e-06,
-      "loss": 0.7074,
+      "loss": 0.6191,
       "step": 2081
     },
     {
       "epoch": 2.0096525096525095,
-      "grad_norm": 5.16745662689209,
+      "grad_norm": 4.958489418029785,
       "learning_rate": 5.193793485142301e-06,
-      "loss": 0.6318,
+      "loss": 0.5759,
       "step": 2082
     },
     {
       "epoch": 2.0106177606177607,
-      "grad_norm": 5.382175445556641,
+      "grad_norm": 5.160446643829346,
       "learning_rate": 5.1846555784824404e-06,
-      "loss": 0.6801,
+      "loss": 0.5871,
       "step": 2083
     },
     {
       "epoch": 2.011583011583012,
-      "grad_norm": 5.676861763000488,
+      "grad_norm": 5.844228744506836,
       "learning_rate": 5.17552290349528e-06,
-      "loss": 0.7505,
+      "loss": 0.6402,
       "step": 2084
     },
     {
       "epoch": 2.0125482625482625,
-      "grad_norm": 4.259150505065918,
+      "grad_norm": 5.312577247619629,
       "learning_rate": 5.166395470103092e-06,
-      "loss": 0.7028,
+      "loss": 0.6317,
       "step": 2085
     },
     {
       "epoch": 2.0135135135135136,
-      "grad_norm": 4.849636077880859,
+      "grad_norm": 5.7476959228515625,
       "learning_rate": 5.157273288222455e-06,
-      "loss": 0.7002,
+      "loss": 0.6087,
       "step": 2086
     },
     {
       "epoch": 2.0144787644787643,
-      "grad_norm": 5.219135761260986,
+      "grad_norm": 5.040074348449707,
       "learning_rate": 5.1481563677642455e-06,
-      "loss": 0.7052,
+      "loss": 0.5926,
       "step": 2087
     },
     {
       "epoch": 2.0154440154440154,
-      "grad_norm": 4.577015399932861,
+      "grad_norm": 5.295816898345947,
       "learning_rate": 5.139044718633609e-06,
-      "loss": 0.6767,
+      "loss": 0.5889,
       "step": 2088
     },
     {
       "epoch": 2.0164092664092665,
-      "grad_norm": 4.252452850341797,
+      "grad_norm": 4.931952953338623,
       "learning_rate": 5.129938350729989e-06,
-      "loss": 0.6946,
+      "loss": 0.5942,
       "step": 2089
     },
     {
       "epoch": 2.0173745173745172,
-      "grad_norm": 4.529285907745361,
+      "grad_norm": 4.646266460418701,
       "learning_rate": 5.120837273947073e-06,
-      "loss": 0.6331,
+      "loss": 0.546,
       "step": 2090
     },
     {
       "epoch": 2.0183397683397684,
-      "grad_norm": 4.3095927238464355,
+      "grad_norm": 4.9664225578308105,
       "learning_rate": 5.111741498172808e-06,
-      "loss": 0.6511,
+      "loss": 0.5576,
       "step": 2091
     },
     {
       "epoch": 2.0193050193050195,
-      "grad_norm": 4.548341274261475,
+      "grad_norm": 4.919215202331543,
       "learning_rate": 5.102651033289369e-06,
-      "loss": 0.7208,
+      "loss": 0.6051,
       "step": 2092
     },
     {
       "epoch": 2.02027027027027,
-      "grad_norm": 4.232816696166992,
+      "grad_norm": 4.052984714508057,
       "learning_rate": 5.09356588917318e-06,
-      "loss": 0.687,
+      "loss": 0.6016,
       "step": 2093
     },
     {
       "epoch": 2.0212355212355213,
-      "grad_norm": 4.521832466125488,
+      "grad_norm": 4.059286117553711,
       "learning_rate": 5.084486075694871e-06,
-      "loss": 0.7281,
+      "loss": 0.6137,
       "step": 2094
     },
     {
       "epoch": 2.022200772200772,
-      "grad_norm": 4.4892754554748535,
+      "grad_norm": 4.073573112487793,
       "learning_rate": 5.075411602719284e-06,
-      "loss": 0.6967,
+      "loss": 0.594,
       "step": 2095
     },
     {
       "epoch": 2.023166023166023,
-      "grad_norm": 4.672778129577637,
+      "grad_norm": 4.092824935913086,
       "learning_rate": 5.06634248010546e-06,
-      "loss": 0.6521,
+      "loss": 0.5787,
       "step": 2096
     },
     {
       "epoch": 2.0241312741312742,
-      "grad_norm": 4.826206684112549,
+      "grad_norm": 4.274193286895752,
       "learning_rate": 5.057278717706624e-06,
-      "loss": 0.6482,
+      "loss": 0.577,
       "step": 2097
     },
     {
       "epoch": 2.025096525096525,
-      "grad_norm": 4.64094877243042,
+      "grad_norm": 3.9777979850769043,
       "learning_rate": 5.0482203253701815e-06,
-      "loss": 0.6018,
+      "loss": 0.5275,
       "step": 2098
     },
     {
       "epoch": 2.026061776061776,
-      "grad_norm": 4.916898250579834,
+      "grad_norm": 4.646378993988037,
       "learning_rate": 5.0391673129377e-06,
-      "loss": 0.6544,
+      "loss": 0.5662,
       "step": 2099
     },
     {
       "epoch": 2.027027027027027,
-      "grad_norm": 4.634807109832764,
+      "grad_norm": 4.353132724761963,
       "learning_rate": 5.030119690244903e-06,
-      "loss": 0.6411,
+      "loss": 0.5626,
       "step": 2100
     },
     {
       "epoch": 2.027992277992278,
-      "grad_norm": 4.873452186584473,
+      "grad_norm": 4.449350833892822,
       "learning_rate": 5.02107746712166e-06,
-      "loss": 0.6592,
+      "loss": 0.578,
       "step": 2101
     },
     {
       "epoch": 2.028957528957529,
-      "grad_norm": 4.586275577545166,
+      "grad_norm": 4.387383937835693,
       "learning_rate": 5.012040653391972e-06,
-      "loss": 0.6816,
+      "loss": 0.6108,
       "step": 2102
     },
     {
       "epoch": 2.02992277992278,
-      "grad_norm": 4.615923881530762,
+      "grad_norm": 4.589904308319092,
       "learning_rate": 5.003009258873963e-06,
-      "loss": 0.6463,
+      "loss": 0.5716,
       "step": 2103
     },
     {
       "epoch": 2.030888030888031,
-      "grad_norm": 4.522491931915283,
+      "grad_norm": 4.344857215881348,
       "learning_rate": 4.993983293379871e-06,
-      "loss": 0.6184,
+      "loss": 0.5435,
       "step": 2104
     },
     {
       "epoch": 2.031853281853282,
-      "grad_norm": 4.397302627563477,
+      "grad_norm": 4.595139503479004,
       "learning_rate": 4.984962766716034e-06,
-      "loss": 0.6795,
+      "loss": 0.5838,
       "step": 2105
     },
     {
       "epoch": 2.0328185328185326,
-      "grad_norm": 4.632791042327881,
+      "grad_norm": 4.197519779205322,
       "learning_rate": 4.975947688682881e-06,
-      "loss": 0.6398,
+      "loss": 0.542,
       "step": 2106
     },
     {
       "epoch": 2.0337837837837838,
-      "grad_norm": 4.260961532592773,
+      "grad_norm": 4.575899600982666,
       "learning_rate": 4.9669380690749215e-06,
-      "loss": 0.6514,
+      "loss": 0.5835,
       "step": 2107
     },
     {
       "epoch": 2.034749034749035,
-      "grad_norm": 5.073614597320557,
+      "grad_norm": 4.902754306793213,
       "learning_rate": 4.957933917680735e-06,
-      "loss": 0.6558,
+      "loss": 0.5361,
       "step": 2108
     },
     {
       "epoch": 2.0357142857142856,
-      "grad_norm": 4.8688483238220215,
+      "grad_norm": 4.742628574371338,
       "learning_rate": 4.948935244282961e-06,
-      "loss": 0.6295,
+      "loss": 0.5308,
       "step": 2109
     },
     {
       "epoch": 2.0366795366795367,
-      "grad_norm": 4.682999610900879,
+      "grad_norm": 5.0072526931762695,
       "learning_rate": 4.939942058658286e-06,
-      "loss": 0.6618,
+      "loss": 0.5459,
       "step": 2110
     },
     {
       "epoch": 2.037644787644788,
-      "grad_norm": 4.4976983070373535,
+      "grad_norm": 4.65580940246582,
       "learning_rate": 4.9309543705774335e-06,
-      "loss": 0.6809,
+      "loss": 0.5821,
       "step": 2111
     },
     {
       "epoch": 2.0386100386100385,
-      "grad_norm": 4.920668601989746,
+      "grad_norm": 4.767663955688477,
       "learning_rate": 4.921972189805154e-06,
-      "loss": 0.7048,
+      "loss": 0.5936,
       "step": 2112
     },
     {
       "epoch": 2.0395752895752897,
-      "grad_norm": 4.667609214782715,
+      "grad_norm": 4.49662971496582,
       "learning_rate": 4.912995526100216e-06,
-      "loss": 0.6312,
+      "loss": 0.5234,
       "step": 2113
     },
     {
       "epoch": 2.0405405405405403,
-      "grad_norm": 4.698695182800293,
+      "grad_norm": 4.633452415466309,
       "learning_rate": 4.904024389215401e-06,
-      "loss": 0.6086,
+      "loss": 0.5161,
       "step": 2114
     },
     {
       "epoch": 2.0415057915057915,
-      "grad_norm": 4.660233974456787,
+      "grad_norm": 4.614391803741455,
       "learning_rate": 4.895058788897469e-06,
-      "loss": 0.6518,
+      "loss": 0.5496,
       "step": 2115
     },
     {
       "epoch": 2.0424710424710426,
-      "grad_norm": 5.051577091217041,
+      "grad_norm": 4.506239891052246,
       "learning_rate": 4.886098734887177e-06,
-      "loss": 0.6578,
+      "loss": 0.5606,
       "step": 2116
     },
     {
       "epoch": 2.0434362934362933,
-      "grad_norm": 4.889238357543945,
+      "grad_norm": 4.49740743637085,
       "learning_rate": 4.87714423691925e-06,
-      "loss": 0.6245,
+      "loss": 0.5269,
       "step": 2117
     },
     {
       "epoch": 2.0444015444015444,
-      "grad_norm": 4.660045623779297,
+      "grad_norm": 4.70435905456543,
       "learning_rate": 4.868195304722391e-06,
-      "loss": 0.695,
+      "loss": 0.5903,
       "step": 2118
     },
     {
       "epoch": 2.0453667953667956,
-      "grad_norm": 4.155313968658447,
+      "grad_norm": 4.061666488647461,
       "learning_rate": 4.859251948019236e-06,
-      "loss": 0.6629,
+      "loss": 0.5457,
       "step": 2119
     },
     {
       "epoch": 2.0463320463320462,
-      "grad_norm": 4.660876750946045,
+      "grad_norm": 4.4224534034729,
       "learning_rate": 4.850314176526373e-06,
-      "loss": 0.6384,
+      "loss": 0.556,
       "step": 2120
     },
     {
       "epoch": 2.0472972972972974,
-      "grad_norm": 4.83228874206543,
+      "grad_norm": 4.47259521484375,
       "learning_rate": 4.8413819999543185e-06,
-      "loss": 0.6539,
+      "loss": 0.5488,
       "step": 2121
     },
     {
       "epoch": 2.048262548262548,
-      "grad_norm": 4.625710964202881,
+      "grad_norm": 4.704776763916016,
       "learning_rate": 4.8324554280075255e-06,
-      "loss": 0.7382,
+      "loss": 0.6405,
       "step": 2122
     },
     {
       "epoch": 2.049227799227799,
-      "grad_norm": 4.783905506134033,
+      "grad_norm": 4.780275344848633,
       "learning_rate": 4.823534470384333e-06,
-      "loss": 0.7325,
+      "loss": 0.641,
       "step": 2123
     },
     {
       "epoch": 2.0501930501930503,
-      "grad_norm": 4.6731038093566895,
+      "grad_norm": 4.860573768615723,
       "learning_rate": 4.814619136776998e-06,
-      "loss": 0.686,
+      "loss": 0.5771,
       "step": 2124
     },
     {
       "epoch": 2.051158301158301,
-      "grad_norm": 5.126330852508545,
+      "grad_norm": 4.6456403732299805,
       "learning_rate": 4.805709436871659e-06,
-      "loss": 0.65,
+      "loss": 0.5453,
       "step": 2125
     },
     {
       "epoch": 2.052123552123552,
-      "grad_norm": 5.164779186248779,
+      "grad_norm": 4.652058124542236,
       "learning_rate": 4.796805380348336e-06,
-      "loss": 0.6678,
+      "loss": 0.5491,
       "step": 2126
     },
     {
       "epoch": 2.0530888030888033,
-      "grad_norm": 4.569568157196045,
+      "grad_norm": 4.787850856781006,
       "learning_rate": 4.78790697688093e-06,
-      "loss": 0.637,
+      "loss": 0.5456,
       "step": 2127
     },
     {
       "epoch": 2.054054054054054,
-      "grad_norm": 5.218832015991211,
+      "grad_norm": 5.174369812011719,
       "learning_rate": 4.7790142361371735e-06,
-      "loss": 0.6875,
+      "loss": 0.5913,
       "step": 2128
     },
     {
       "epoch": 2.055019305019305,
-      "grad_norm": 5.137186050415039,
+      "grad_norm": 4.936854362487793,
       "learning_rate": 4.77012716777867e-06,
-      "loss": 0.6989,
+      "loss": 0.5883,
       "step": 2129
     },
     {
       "epoch": 2.0559845559845558,
-      "grad_norm": 4.738580226898193,
+      "grad_norm": 4.539516448974609,
       "learning_rate": 4.761245781460846e-06,
-      "loss": 0.6949,
+      "loss": 0.5861,
       "step": 2130
     },
     {
       "epoch": 2.056949806949807,
-      "grad_norm": 5.230045795440674,
+      "grad_norm": 4.27752161026001,
       "learning_rate": 4.752370086832971e-06,
-      "loss": 0.6575,
+      "loss": 0.5325,
       "step": 2131
     },
     {
       "epoch": 2.057915057915058,
-      "grad_norm": 4.877504348754883,
+      "grad_norm": 4.549672603607178,
       "learning_rate": 4.743500093538111e-06,
-      "loss": 0.7114,
+      "loss": 0.58,
       "step": 2132
     },
     {
       "epoch": 2.0588803088803087,
-      "grad_norm": 4.573655605316162,
+      "grad_norm": 4.883086204528809,
       "learning_rate": 4.73463581121315e-06,
-      "loss": 0.6913,
+      "loss": 0.5874,
       "step": 2133
     },
     {
       "epoch": 2.05984555984556,
-      "grad_norm": 4.723101615905762,
+      "grad_norm": 4.432310104370117,
       "learning_rate": 4.725777249488761e-06,
-      "loss": 0.674,
+      "loss": 0.556,
       "step": 2134
     },
     {
       "epoch": 2.060810810810811,
-      "grad_norm": 4.4217023849487305,
+      "grad_norm": 4.407434463500977,
       "learning_rate": 4.716924417989413e-06,
-      "loss": 0.6296,
+      "loss": 0.521,
       "step": 2135
     },
     {
       "epoch": 2.0617760617760617,
-      "grad_norm": 4.694399356842041,
+      "grad_norm": 4.895625591278076,
       "learning_rate": 4.7080773263333405e-06,
-      "loss": 0.6748,
+      "loss": 0.5852,
       "step": 2136
     },
     {
       "epoch": 2.062741312741313,
-      "grad_norm": 4.912079334259033,
+      "grad_norm": 4.728435039520264,
       "learning_rate": 4.699235984132535e-06,
-      "loss": 0.6427,
+      "loss": 0.5186,
       "step": 2137
     },
     {
       "epoch": 2.063706563706564,
-      "grad_norm": 4.5845723152160645,
+      "grad_norm": 4.9962544441223145,
       "learning_rate": 4.690400400992752e-06,
-      "loss": 0.6512,
+      "loss": 0.5616,
       "step": 2138
     },
     {
       "epoch": 2.0646718146718146,
-      "grad_norm": 4.9580512046813965,
+      "grad_norm": 4.823432922363281,
       "learning_rate": 4.681570586513492e-06,
-      "loss": 0.6706,
+      "loss": 0.5728,
       "step": 2139
     },
     {
       "epoch": 2.0656370656370657,
-      "grad_norm": 4.75042200088501,
+      "grad_norm": 4.529953956604004,
       "learning_rate": 4.672746550287985e-06,
-      "loss": 0.6449,
+      "loss": 0.539,
       "step": 2140
     },
     {
       "epoch": 2.0666023166023164,
-      "grad_norm": 5.1743245124816895,
+      "grad_norm": 4.412690162658691,
       "learning_rate": 4.663928301903175e-06,
-      "loss": 0.6604,
+      "loss": 0.5353,
       "step": 2141
     },
     {
       "epoch": 2.0675675675675675,
-      "grad_norm": 4.8098955154418945,
+      "grad_norm": 4.605326175689697,
       "learning_rate": 4.655115850939722e-06,
-      "loss": 0.6772,
+      "loss": 0.5519,
       "step": 2142
     },
     {
       "epoch": 2.0685328185328187,
-      "grad_norm": 5.1585259437561035,
+      "grad_norm": 4.746756553649902,
       "learning_rate": 4.6463092069719996e-06,
-      "loss": 0.6969,
+      "loss": 0.5693,
       "step": 2143
     },
     {
       "epoch": 2.0694980694980694,
-      "grad_norm": 4.868752956390381,
+      "grad_norm": 4.824178218841553,
       "learning_rate": 4.637508379568057e-06,
-      "loss": 0.6561,
+      "loss": 0.5764,
       "step": 2144
     },
     {
       "epoch": 2.0704633204633205,
-      "grad_norm": 4.869829177856445,
+      "grad_norm": 4.554269790649414,
       "learning_rate": 4.628713378289634e-06,
-      "loss": 0.6787,
+      "loss": 0.5633,
       "step": 2145
     },
     {
       "epoch": 2.0714285714285716,
-      "grad_norm": 4.9792633056640625,
+      "grad_norm": 4.564504623413086,
       "learning_rate": 4.619924212692125e-06,
-      "loss": 0.6789,
+      "loss": 0.5694,
       "step": 2146
     },
     {
       "epoch": 2.0723938223938223,
-      "grad_norm": 4.65444278717041,
+      "grad_norm": 4.765178680419922,
       "learning_rate": 4.611140892324606e-06,
-      "loss": 0.6409,
+      "loss": 0.5286,
       "step": 2147
     },
     {
       "epoch": 2.0733590733590734,
-      "grad_norm": 4.609977722167969,
+      "grad_norm": 4.764468669891357,
       "learning_rate": 4.602363426729788e-06,
-      "loss": 0.6254,
+      "loss": 0.538,
       "step": 2148
     },
     {
       "epoch": 2.074324324324324,
-      "grad_norm": 4.691058158874512,
+      "grad_norm": 4.465149879455566,
       "learning_rate": 4.593591825444028e-06,
-      "loss": 0.6442,
+      "loss": 0.539,
       "step": 2149
     },
     {
       "epoch": 2.0752895752895753,
-      "grad_norm": 4.959919452667236,
+      "grad_norm": 4.88388204574585,
       "learning_rate": 4.584826097997297e-06,
-      "loss": 0.6973,
+      "loss": 0.58,
       "step": 2150
     },
     {
       "epoch": 2.0762548262548264,
-      "grad_norm": 4.790925979614258,
+      "grad_norm": 4.58705997467041,
       "learning_rate": 4.576066253913209e-06,
-      "loss": 0.6681,
+      "loss": 0.5624,
       "step": 2151
     },
     {
       "epoch": 2.077220077220077,
-      "grad_norm": 5.478535175323486,
+      "grad_norm": 5.03671932220459,
       "learning_rate": 4.567312302708965e-06,
-      "loss": 0.6703,
+      "loss": 0.5436,
       "step": 2152
     },
     {
       "epoch": 2.078185328185328,
-      "grad_norm": 4.707211494445801,
+      "grad_norm": 4.8304243087768555,
       "learning_rate": 4.558564253895374e-06,
-      "loss": 0.6848,
+      "loss": 0.5673,
       "step": 2153
     },
     {
       "epoch": 2.0791505791505793,
-      "grad_norm": 5.1434006690979,
+      "grad_norm": 4.853978157043457,
       "learning_rate": 4.549822116976827e-06,
-      "loss": 0.6834,
+      "loss": 0.5507,
       "step": 2154
     },
     {
       "epoch": 2.08011583011583,
-      "grad_norm": 5.3403730392456055,
+      "grad_norm": 4.97976016998291,
       "learning_rate": 4.5410859014512975e-06,
-      "loss": 0.6803,
+      "loss": 0.5597,
       "step": 2155
     },
     {
       "epoch": 2.081081081081081,
-      "grad_norm": 4.581385135650635,
+      "grad_norm": 4.615689277648926,
       "learning_rate": 4.532355616810321e-06,
-      "loss": 0.6413,
+      "loss": 0.5256,
       "step": 2156
     },
     {
       "epoch": 2.082046332046332,
-      "grad_norm": 4.807558536529541,
+      "grad_norm": 4.652611255645752,
       "learning_rate": 4.523631272538992e-06,
-      "loss": 0.6617,
+      "loss": 0.5502,
       "step": 2157
     },
     {
       "epoch": 2.083011583011583,
-      "grad_norm": 4.22251033782959,
+      "grad_norm": 4.608834266662598,
       "learning_rate": 4.514912878115948e-06,
-      "loss": 0.6829,
+      "loss": 0.5946,
       "step": 2158
     },
     {
       "epoch": 2.083976833976834,
-      "grad_norm": 4.495012283325195,
+      "grad_norm": 4.699759483337402,
       "learning_rate": 4.5062004430133655e-06,
-      "loss": 0.7071,
+      "loss": 0.5934,
       "step": 2159
     },
     {
       "epoch": 2.0849420849420848,
-      "grad_norm": 4.16301965713501,
+      "grad_norm": 4.292397975921631,
       "learning_rate": 4.497493976696945e-06,
-      "loss": 0.6848,
+      "loss": 0.5938,
       "step": 2160
     },
     {
       "epoch": 2.085907335907336,
-      "grad_norm": 5.111375331878662,
+      "grad_norm": 5.059668064117432,
       "learning_rate": 4.4887934886259025e-06,
-      "loss": 0.6817,
+      "loss": 0.5656,
       "step": 2161
     },
     {
       "epoch": 2.086872586872587,
-      "grad_norm": 4.352403163909912,
+      "grad_norm": 4.336543560028076,
       "learning_rate": 4.480098988252958e-06,
-      "loss": 0.7081,
+      "loss": 0.5756,
       "step": 2162
     },
     {
       "epoch": 2.0878378378378377,
-      "grad_norm": 4.8227152824401855,
+      "grad_norm": 4.695181369781494,
       "learning_rate": 4.471410485024327e-06,
-      "loss": 0.6611,
+      "loss": 0.5516,
       "step": 2163
     },
     {
       "epoch": 2.088803088803089,
-      "grad_norm": 4.887004375457764,
+      "grad_norm": 4.7586588859558105,
       "learning_rate": 4.4627279883797096e-06,
-      "loss": 0.7235,
+      "loss": 0.5857,
       "step": 2164
     },
     {
       "epoch": 2.08976833976834,
-      "grad_norm": 4.713346481323242,
+      "grad_norm": 4.185433864593506,
       "learning_rate": 4.454051507752279e-06,
-      "loss": 0.6352,
+      "loss": 0.51,
       "step": 2165
     },
     {
       "epoch": 2.0907335907335907,
-      "grad_norm": 4.931870460510254,
+      "grad_norm": 4.633828163146973,
       "learning_rate": 4.445381052568674e-06,
-      "loss": 0.6878,
+      "loss": 0.5493,
       "step": 2166
     },
     {
       "epoch": 2.091698841698842,
-      "grad_norm": 4.754703044891357,
+      "grad_norm": 4.499352931976318,
       "learning_rate": 4.436716632248985e-06,
-      "loss": 0.6186,
+      "loss": 0.4784,
       "step": 2167
     },
     {
       "epoch": 2.0926640926640925,
-      "grad_norm": 5.296566486358643,
+      "grad_norm": 4.812720775604248,
       "learning_rate": 4.4280582562067465e-06,
-      "loss": 0.6907,
+      "loss": 0.5675,
       "step": 2168
     },
     {
       "epoch": 2.0936293436293436,
-      "grad_norm": 4.916853427886963,
+      "grad_norm": 4.581881999969482,
       "learning_rate": 4.419405933848927e-06,
-      "loss": 0.6636,
+      "loss": 0.537,
       "step": 2169
     },
     {
       "epoch": 2.0945945945945947,
-      "grad_norm": 4.9489288330078125,
+      "grad_norm": 5.138509273529053,
       "learning_rate": 4.410759674575915e-06,
-      "loss": 0.7377,
+      "loss": 0.6291,
       "step": 2170
     },
     {
       "epoch": 2.0955598455598454,
-      "grad_norm": 5.000365257263184,
+      "grad_norm": 4.543698310852051,
       "learning_rate": 4.4021194877815165e-06,
-      "loss": 0.6841,
+      "loss": 0.5665,
       "step": 2171
     },
     {
       "epoch": 2.0965250965250966,
-      "grad_norm": 4.6656575202941895,
+      "grad_norm": 4.8228440284729,
       "learning_rate": 4.393485382852936e-06,
-      "loss": 0.7516,
+      "loss": 0.6377,
       "step": 2172
     },
     {
       "epoch": 2.0974903474903477,
-      "grad_norm": 4.6403021812438965,
+      "grad_norm": 4.641752243041992,
       "learning_rate": 4.384857369170772e-06,
-      "loss": 0.6799,
+      "loss": 0.559,
       "step": 2173
     },
     {
       "epoch": 2.0984555984555984,
-      "grad_norm": 4.789555549621582,
+      "grad_norm": 5.043064117431641,
       "learning_rate": 4.376235456109003e-06,
-      "loss": 0.6836,
+      "loss": 0.577,
       "step": 2174
     },
     {
       "epoch": 2.0994208494208495,
-      "grad_norm": 4.911081314086914,
+      "grad_norm": 4.893931865692139,
       "learning_rate": 4.3676196530349845e-06,
-      "loss": 0.6228,
+      "loss": 0.5054,
       "step": 2175
     },
     {
       "epoch": 2.1003861003861,
-      "grad_norm": 4.8095879554748535,
+      "grad_norm": 4.954989433288574,
       "learning_rate": 4.359009969309429e-06,
-      "loss": 0.6683,
+      "loss": 0.5576,
       "step": 2176
     },
     {
       "epoch": 2.1013513513513513,
-      "grad_norm": 5.17523717880249,
+      "grad_norm": 4.9561052322387695,
       "learning_rate": 4.350406414286401e-06,
-      "loss": 0.7043,
+      "loss": 0.5629,
       "step": 2177
     },
     {
       "epoch": 2.1023166023166024,
-      "grad_norm": 4.903587818145752,
+      "grad_norm": 4.964174270629883,
       "learning_rate": 4.341808997313308e-06,
-      "loss": 0.6637,
+      "loss": 0.537,
       "step": 2178
     },
     {
       "epoch": 2.103281853281853,
-      "grad_norm": 4.904983043670654,
+      "grad_norm": 4.8784332275390625,
       "learning_rate": 4.3332177277308885e-06,
-      "loss": 0.7013,
+      "loss": 0.5987,
       "step": 2179
     },
     {
       "epoch": 2.1042471042471043,
-      "grad_norm": 5.028541564941406,
+      "grad_norm": 5.0084967613220215,
       "learning_rate": 4.324632614873199e-06,
-      "loss": 0.6973,
+      "loss": 0.5804,
       "step": 2180
     },
     {
       "epoch": 2.1052123552123554,
-      "grad_norm": 4.6789093017578125,
+      "grad_norm": 4.892880439758301,
       "learning_rate": 4.316053668067619e-06,
-      "loss": 0.6755,
+      "loss": 0.5678,
       "step": 2181
     },
     {
       "epoch": 2.106177606177606,
-      "grad_norm": 5.516636848449707,
+      "grad_norm": 5.346131324768066,
       "learning_rate": 4.3074808966348105e-06,
-      "loss": 0.6592,
+      "loss": 0.5305,
       "step": 2182
     },
     {
       "epoch": 2.107142857142857,
-      "grad_norm": 4.884287357330322,
+      "grad_norm": 4.654419898986816,
       "learning_rate": 4.298914309888736e-06,
-      "loss": 0.6969,
+      "loss": 0.5681,
       "step": 2183
     },
     {
       "epoch": 2.108108108108108,
-      "grad_norm": 4.955660343170166,
+      "grad_norm": 5.201515197753906,
       "learning_rate": 4.290353917136639e-06,
-      "loss": 0.6357,
+      "loss": 0.5218,
       "step": 2184
     },
     {
       "epoch": 2.109073359073359,
-      "grad_norm": 4.877090930938721,
+      "grad_norm": 5.0456929206848145,
       "learning_rate": 4.281799727679039e-06,
-      "loss": 0.6607,
+      "loss": 0.5488,
       "step": 2185
     },
     {
       "epoch": 2.11003861003861,
-      "grad_norm": 4.54473876953125,
+      "grad_norm": 4.442846298217773,
       "learning_rate": 4.2732517508097016e-06,
-      "loss": 0.6441,
+      "loss": 0.5099,
       "step": 2186
     },
     {
       "epoch": 2.111003861003861,
-      "grad_norm": 4.789745330810547,
+      "grad_norm": 4.640462875366211,
       "learning_rate": 4.264709995815651e-06,
-      "loss": 0.6652,
+      "loss": 0.5326,
       "step": 2187
     },
     {
       "epoch": 2.111969111969112,
-      "grad_norm": 4.660126686096191,
+      "grad_norm": 4.447720527648926,
       "learning_rate": 4.25617447197715e-06,
-      "loss": 0.6766,
+      "loss": 0.5437,
       "step": 2188
     },
     {
       "epoch": 2.112934362934363,
-      "grad_norm": 5.456777095794678,
+      "grad_norm": 5.47343111038208,
       "learning_rate": 4.247645188567698e-06,
-      "loss": 0.674,
+      "loss": 0.5425,
       "step": 2189
     },
     {
       "epoch": 2.113899613899614,
-      "grad_norm": 4.886148452758789,
+      "grad_norm": 4.517058372497559,
       "learning_rate": 4.2391221548540094e-06,
-      "loss": 0.6653,
+      "loss": 0.5439,
       "step": 2190
     },
     {
       "epoch": 2.114864864864865,
-      "grad_norm": 4.867854595184326,
+      "grad_norm": 4.955538272857666,
       "learning_rate": 4.230605380096001e-06,
-      "loss": 0.7063,
+      "loss": 0.5713,
       "step": 2191
     },
     {
       "epoch": 2.115830115830116,
-      "grad_norm": 5.060470104217529,
+      "grad_norm": 5.277383327484131,
       "learning_rate": 4.222094873546797e-06,
-      "loss": 0.6619,
+      "loss": 0.5295,
       "step": 2192
     },
     {
       "epoch": 2.1167953667953667,
-      "grad_norm": 4.72755241394043,
+      "grad_norm": 4.8949174880981445,
       "learning_rate": 4.213590644452717e-06,
-      "loss": 0.7273,
+      "loss": 0.5924,
       "step": 2193
     },
     {
       "epoch": 2.117760617760618,
-      "grad_norm": 5.060878276824951,
+      "grad_norm": 4.787299633026123,
       "learning_rate": 4.205092702053257e-06,
-      "loss": 0.685,
+      "loss": 0.5283,
       "step": 2194
     },
     {
       "epoch": 2.1187258687258685,
-      "grad_norm": 4.600476264953613,
+      "grad_norm": 4.664861679077148,
       "learning_rate": 4.1966010555810696e-06,
-      "loss": 0.7129,
+      "loss": 0.5969,
       "step": 2195
     },
     {
       "epoch": 2.1196911196911197,
-      "grad_norm": 4.683096408843994,
+      "grad_norm": 5.040746688842773,
       "learning_rate": 4.18811571426198e-06,
-      "loss": 0.7142,
+      "loss": 0.5841,
       "step": 2196
     },
     {
       "epoch": 2.120656370656371,
-      "grad_norm": 4.833113193511963,
+      "grad_norm": 5.23708438873291,
       "learning_rate": 4.179636687314967e-06,
-      "loss": 0.7439,
+      "loss": 0.6022,
       "step": 2197
     },
     {
       "epoch": 2.1216216216216215,
-      "grad_norm": 4.970816135406494,
+      "grad_norm": 4.965457439422607,
       "learning_rate": 4.171163983952144e-06,
-      "loss": 0.6971,
+      "loss": 0.5564,
       "step": 2198
     },
     {
       "epoch": 2.1225868725868726,
-      "grad_norm": 4.352797985076904,
+      "grad_norm": 4.551243782043457,
       "learning_rate": 4.1626976133787445e-06,
-      "loss": 0.7058,
+      "loss": 0.5763,
       "step": 2199
     },
     {
       "epoch": 2.1235521235521237,
-      "grad_norm": 4.643060684204102,
+      "grad_norm": 4.769243240356445,
       "learning_rate": 4.154237584793132e-06,
-      "loss": 0.6841,
+      "loss": 0.5675,
       "step": 2200
     },
     {
       "epoch": 2.1245173745173744,
-      "grad_norm": 4.573018550872803,
+      "grad_norm": 5.064882755279541,
       "learning_rate": 4.145783907386782e-06,
-      "loss": 0.7119,
+      "loss": 0.5827,
       "step": 2201
     },
     {
       "epoch": 2.1254826254826256,
-      "grad_norm": 4.699937343597412,
+      "grad_norm": 4.795884132385254,
       "learning_rate": 4.1373365903442665e-06,
-      "loss": 0.667,
+      "loss": 0.5306,
       "step": 2202
     },
     {
       "epoch": 2.1264478764478763,
-      "grad_norm": 4.921900749206543,
+      "grad_norm": 4.811540603637695,
       "learning_rate": 4.1288956428432455e-06,
-      "loss": 0.6998,
+      "loss": 0.5789,
       "step": 2203
     },
     {
       "epoch": 2.1274131274131274,
-      "grad_norm": 4.901946544647217,
+      "grad_norm": 4.671502590179443,
       "learning_rate": 4.120461074054455e-06,
-      "loss": 0.6726,
+      "loss": 0.5384,
       "step": 2204
     },
     {
       "epoch": 2.1283783783783785,
-      "grad_norm": 5.032048225402832,
+      "grad_norm": 4.902651786804199,
       "learning_rate": 4.112032893141706e-06,
-      "loss": 0.7078,
+      "loss": 0.5715,
       "step": 2205
     },
     {
       "epoch": 2.129343629343629,
-      "grad_norm": 4.5773162841796875,
+      "grad_norm": 4.854404449462891,
       "learning_rate": 4.1036111092618725e-06,
-      "loss": 0.6699,
+      "loss": 0.5613,
       "step": 2206
     },
     {
       "epoch": 2.1303088803088803,
-      "grad_norm": 5.37986421585083,
+      "grad_norm": 4.914683818817139,
       "learning_rate": 4.095195731564878e-06,
-      "loss": 0.7211,
+      "loss": 0.5955,
       "step": 2207
     },
     {
       "epoch": 2.1312741312741315,
-      "grad_norm": 4.763434410095215,
+      "grad_norm": 4.389202117919922,
       "learning_rate": 4.086786769193673e-06,
-      "loss": 0.6721,
+      "loss": 0.5406,
       "step": 2208
     },
     {
       "epoch": 2.132239382239382,
-      "grad_norm": 4.474204063415527,
+      "grad_norm": 4.381209373474121,
       "learning_rate": 4.0783842312842506e-06,
-      "loss": 0.7242,
+      "loss": 0.5769,
       "step": 2209
     },
     {
       "epoch": 2.1332046332046333,
-      "grad_norm": 4.982334136962891,
+      "grad_norm": 4.870187759399414,
       "learning_rate": 4.069988126965625e-06,
-      "loss": 0.6732,
+      "loss": 0.5389,
       "step": 2210
     },
     {
       "epoch": 2.134169884169884,
-      "grad_norm": 5.002752304077148,
+      "grad_norm": 4.981731414794922,
       "learning_rate": 4.061598465359815e-06,
-      "loss": 0.662,
+      "loss": 0.5418,
       "step": 2211
     },
     {
       "epoch": 2.135135135135135,
-      "grad_norm": 5.2900238037109375,
+      "grad_norm": 4.9232611656188965,
       "learning_rate": 4.053215255581843e-06,
-      "loss": 0.6901,
+      "loss": 0.5388,
       "step": 2212
     },
     {
       "epoch": 2.136100386100386,
-      "grad_norm": 4.898741245269775,
+      "grad_norm": 5.039803981781006,
       "learning_rate": 4.044838506739711e-06,
-      "loss": 0.7002,
+      "loss": 0.5516,
       "step": 2213
     },
     {
       "epoch": 2.137065637065637,
-      "grad_norm": 4.9963884353637695,
+      "grad_norm": 4.95274019241333,
       "learning_rate": 4.03646822793442e-06,
-      "loss": 0.6949,
+      "loss": 0.5532,
       "step": 2214
     },
     {
       "epoch": 2.138030888030888,
-      "grad_norm": 4.586988925933838,
+      "grad_norm": 4.421284198760986,
       "learning_rate": 4.028104428259929e-06,
-      "loss": 0.6962,
+      "loss": 0.5576,
       "step": 2215
     },
     {
       "epoch": 2.138996138996139,
-      "grad_norm": 4.49263334274292,
+      "grad_norm": 4.643174648284912,
       "learning_rate": 4.019747116803164e-06,
-      "loss": 0.6961,
+      "loss": 0.5513,
       "step": 2216
     },
     {
       "epoch": 2.13996138996139,
-      "grad_norm": 4.701002597808838,
+      "grad_norm": 4.833516597747803,
       "learning_rate": 4.011396302643989e-06,
-      "loss": 0.7185,
+      "loss": 0.5992,
       "step": 2217
     },
     {
       "epoch": 2.140926640926641,
-      "grad_norm": 4.800919532775879,
+      "grad_norm": 4.856517791748047,
       "learning_rate": 4.003051994855226e-06,
-      "loss": 0.6886,
+      "loss": 0.5437,
       "step": 2218
     },
     {
       "epoch": 2.141891891891892,
-      "grad_norm": 4.730530738830566,
+      "grad_norm": 4.656328201293945,
       "learning_rate": 3.99471420250262e-06,
-      "loss": 0.6515,
+      "loss": 0.5101,
       "step": 2219
     },
     {
       "epoch": 2.142857142857143,
-      "grad_norm": 5.242630958557129,
+      "grad_norm": 4.925358295440674,
       "learning_rate": 3.986382934644836e-06,
-      "loss": 0.694,
+      "loss": 0.5293,
       "step": 2220
     },
     {
       "epoch": 2.143822393822394,
-      "grad_norm": 4.6426005363464355,
+      "grad_norm": 4.599172115325928,
       "learning_rate": 3.978058200333453e-06,
-      "loss": 0.68,
+      "loss": 0.5352,
       "step": 2221
     },
     {
       "epoch": 2.1447876447876446,
-      "grad_norm": 4.800093650817871,
+      "grad_norm": 4.931187629699707,
       "learning_rate": 3.969740008612951e-06,
-      "loss": 0.6869,
+      "loss": 0.5515,
       "step": 2222
     },
     {
       "epoch": 2.1457528957528957,
-      "grad_norm": 4.5579681396484375,
+      "grad_norm": 4.603859901428223,
       "learning_rate": 3.961428368520702e-06,
-      "loss": 0.6762,
+      "loss": 0.5682,
       "step": 2223
     },
     {
       "epoch": 2.146718146718147,
-      "grad_norm": 5.217144012451172,
+      "grad_norm": 5.244414329528809,
       "learning_rate": 3.953123289086958e-06,
-      "loss": 0.6761,
+      "loss": 0.5457,
       "step": 2224
     },
     {
       "epoch": 2.1476833976833976,
-      "grad_norm": 4.891463756561279,
+      "grad_norm": 4.6841864585876465,
       "learning_rate": 3.944824779334845e-06,
-      "loss": 0.6881,
+      "loss": 0.5591,
       "step": 2225
     },
     {
       "epoch": 2.1486486486486487,
-      "grad_norm": 4.874589920043945,
+      "grad_norm": 4.961796760559082,
       "learning_rate": 3.936532848280351e-06,
-      "loss": 0.7041,
+      "loss": 0.562,
       "step": 2226
     },
     {
       "epoch": 2.1496138996139,
-      "grad_norm": 4.724242210388184,
+      "grad_norm": 5.141530990600586,
       "learning_rate": 3.928247504932317e-06,
-      "loss": 0.7238,
+      "loss": 0.592,
       "step": 2227
     },
     {
       "epoch": 2.1505791505791505,
-      "grad_norm": 4.630641460418701,
+      "grad_norm": 4.555334568023682,
       "learning_rate": 3.919968758292425e-06,
-      "loss": 0.6537,
+      "loss": 0.5156,
       "step": 2228
     },
     {
       "epoch": 2.1515444015444016,
-      "grad_norm": 4.487358093261719,
+      "grad_norm": 4.653707027435303,
       "learning_rate": 3.9116966173551915e-06,
-      "loss": 0.671,
+      "loss": 0.5404,
       "step": 2229
     },
     {
       "epoch": 2.1525096525096523,
-      "grad_norm": 4.850062370300293,
+      "grad_norm": 4.685117244720459,
       "learning_rate": 3.903431091107957e-06,
-      "loss": 0.7052,
+      "loss": 0.5522,
       "step": 2230
     },
     {
       "epoch": 2.1534749034749034,
-      "grad_norm": 4.9462504386901855,
+      "grad_norm": 4.701056957244873,
       "learning_rate": 3.895172188530871e-06,
-      "loss": 0.7052,
+      "loss": 0.57,
       "step": 2231
     },
     {
       "epoch": 2.1544401544401546,
-      "grad_norm": 4.509797096252441,
+      "grad_norm": 4.380445957183838,
       "learning_rate": 3.8869199185968944e-06,
-      "loss": 0.7541,
+      "loss": 0.6063,
       "step": 2232
     },
     {
       "epoch": 2.1554054054054053,
-      "grad_norm": 4.4127302169799805,
+      "grad_norm": 4.553746700286865,
       "learning_rate": 3.878674290271773e-06,
-      "loss": 0.6825,
+      "loss": 0.5614,
       "step": 2233
     },
     {
       "epoch": 2.1563706563706564,
-      "grad_norm": 5.578569412231445,
+      "grad_norm": 4.895527362823486,
       "learning_rate": 3.8704353125140436e-06,
-      "loss": 0.6822,
+      "loss": 0.5508,
       "step": 2234
     },
     {
       "epoch": 2.1573359073359075,
-      "grad_norm": 4.80642032623291,
+      "grad_norm": 4.923834323883057,
       "learning_rate": 3.862202994275015e-06,
-      "loss": 0.6905,
+      "loss": 0.5516,
       "step": 2235
     },
     {
       "epoch": 2.158301158301158,
-      "grad_norm": 4.7371134757995605,
+      "grad_norm": 4.739142417907715,
       "learning_rate": 3.853977344498761e-06,
-      "loss": 0.6829,
+      "loss": 0.5407,
       "step": 2236
     },
     {
       "epoch": 2.1592664092664093,
-      "grad_norm": 5.159573078155518,
+      "grad_norm": 4.9512619972229,
       "learning_rate": 3.845758372122107e-06,
-      "loss": 0.7233,
+      "loss": 0.5558,
       "step": 2237
     },
     {
       "epoch": 2.16023166023166,
-      "grad_norm": 4.449493885040283,
+      "grad_norm": 4.289952754974365,
       "learning_rate": 3.83754608607463e-06,
-      "loss": 0.6616,
+      "loss": 0.5304,
       "step": 2238
     },
     {
       "epoch": 2.161196911196911,
-      "grad_norm": 4.797781467437744,
+      "grad_norm": 4.468839168548584,
       "learning_rate": 3.82934049527864e-06,
-      "loss": 0.6864,
+      "loss": 0.5208,
       "step": 2239
     },
     {
       "epoch": 2.1621621621621623,
-      "grad_norm": 4.917870998382568,
+      "grad_norm": 4.5977020263671875,
       "learning_rate": 3.821141608649169e-06,
-      "loss": 0.7103,
+      "loss": 0.567,
       "step": 2240
     },
     {
       "epoch": 2.163127413127413,
-      "grad_norm": 4.734244346618652,
+      "grad_norm": 4.675628662109375,
       "learning_rate": 3.812949435093971e-06,
-      "loss": 0.6438,
+      "loss": 0.5109,
       "step": 2241
     },
     {
       "epoch": 2.164092664092664,
-      "grad_norm": 5.052276611328125,
+      "grad_norm": 4.8543853759765625,
       "learning_rate": 3.804763983513503e-06,
-      "loss": 0.7217,
+      "loss": 0.5825,
       "step": 2242
     },
     {
       "epoch": 2.1650579150579152,
-      "grad_norm": 4.778468608856201,
+      "grad_norm": 4.880924701690674,
       "learning_rate": 3.7965852628009204e-06,
-      "loss": 0.6653,
+      "loss": 0.5501,
       "step": 2243
     },
     {
       "epoch": 2.166023166023166,
-      "grad_norm": 5.127320289611816,
+      "grad_norm": 4.901237487792969,
       "learning_rate": 3.788413281842065e-06,
-      "loss": 0.7194,
+      "loss": 0.5601,
       "step": 2244
     },
     {
       "epoch": 2.166988416988417,
-      "grad_norm": 4.7970805168151855,
+      "grad_norm": 4.791472911834717,
       "learning_rate": 3.780248049515456e-06,
-      "loss": 0.6861,
+      "loss": 0.5489,
       "step": 2245
     },
     {
       "epoch": 2.167953667953668,
-      "grad_norm": 5.094338893890381,
+      "grad_norm": 4.789416313171387,
       "learning_rate": 3.7720895746922812e-06,
-      "loss": 0.7325,
+      "loss": 0.5968,
       "step": 2246
     },
     {
       "epoch": 2.168918918918919,
-      "grad_norm": 4.787800312042236,
+      "grad_norm": 4.547966480255127,
       "learning_rate": 3.7639378662363823e-06,
-      "loss": 0.682,
+      "loss": 0.5507,
       "step": 2247
     },
     {
       "epoch": 2.16988416988417,
-      "grad_norm": 5.029555320739746,
+      "grad_norm": 4.809093952178955,
       "learning_rate": 3.7557929330042654e-06,
-      "loss": 0.7349,
+      "loss": 0.5887,
       "step": 2248
     },
     {
       "epoch": 2.1708494208494207,
-      "grad_norm": 4.596803188323975,
+      "grad_norm": 4.981399059295654,
       "learning_rate": 3.747654783845052e-06,
-      "loss": 0.702,
+      "loss": 0.5773,
       "step": 2249
     },
     {
       "epoch": 2.171814671814672,
-      "grad_norm": 4.6137566566467285,
+      "grad_norm": 4.653560638427734,
       "learning_rate": 3.739523427600509e-06,
-      "loss": 0.6399,
+      "loss": 0.5241,
       "step": 2250
     },
     {
       "epoch": 2.172779922779923,
-      "grad_norm": 4.858090877532959,
+      "grad_norm": 4.916978359222412,
       "learning_rate": 3.7313988731050156e-06,
-      "loss": 0.7122,
+      "loss": 0.5538,
       "step": 2251
     },
     {
       "epoch": 2.1737451737451736,
-      "grad_norm": 4.909048080444336,
+      "grad_norm": 5.132083415985107,
       "learning_rate": 3.723281129185574e-06,
-      "loss": 0.7443,
+      "loss": 0.57,
       "step": 2252
     },
     {
       "epoch": 2.1747104247104247,
-      "grad_norm": 4.852362632751465,
+      "grad_norm": 4.721053600311279,
       "learning_rate": 3.7151702046617677e-06,
-      "loss": 0.6985,
+      "loss": 0.5393,
       "step": 2253
     },
     {
       "epoch": 2.175675675675676,
-      "grad_norm": 4.483128547668457,
+      "grad_norm": 4.541321754455566,
       "learning_rate": 3.7070661083457838e-06,
-      "loss": 0.6975,
+      "loss": 0.5385,
       "step": 2254
     },
     {
       "epoch": 2.1766409266409266,
-      "grad_norm": 4.893508434295654,
+      "grad_norm": 4.820981025695801,
       "learning_rate": 3.698968849042386e-06,
-      "loss": 0.7307,
+      "loss": 0.5835,
       "step": 2255
     },
     {
       "epoch": 2.1776061776061777,
-      "grad_norm": 5.04038667678833,
+      "grad_norm": 5.075002193450928,
       "learning_rate": 3.690878435548917e-06,
-      "loss": 0.7117,
+      "loss": 0.5562,
       "step": 2256
     },
     {
       "epoch": 2.1785714285714284,
-      "grad_norm": 4.4502949714660645,
+      "grad_norm": 4.626798629760742,
       "learning_rate": 3.682794876655278e-06,
-      "loss": 0.6626,
+      "loss": 0.5447,
       "step": 2257
     },
     {
       "epoch": 2.1795366795366795,
-      "grad_norm": 4.678581237792969,
+      "grad_norm": 4.658825397491455,
       "learning_rate": 3.6747181811439146e-06,
-      "loss": 0.6716,
+      "loss": 0.5289,
       "step": 2258
     },
     {
       "epoch": 2.1805019305019306,
-      "grad_norm": 4.655972480773926,
+      "grad_norm": 4.680113792419434,
       "learning_rate": 3.6666483577898215e-06,
-      "loss": 0.6374,
+      "loss": 0.5021,
       "step": 2259
     },
     {
       "epoch": 2.1814671814671813,
-      "grad_norm": 5.271326541900635,
+      "grad_norm": 5.47307014465332,
       "learning_rate": 3.658585415360537e-06,
-      "loss": 0.6766,
+      "loss": 0.5582,
       "step": 2260
     },
     {
       "epoch": 2.1824324324324325,
-      "grad_norm": 4.562073707580566,
+      "grad_norm": 4.703200817108154,
       "learning_rate": 3.650529362616113e-06,
-      "loss": 0.7137,
+      "loss": 0.5705,
       "step": 2261
     },
     {
       "epoch": 2.1833976833976836,
-      "grad_norm": 5.000593185424805,
+      "grad_norm": 4.834163188934326,
       "learning_rate": 3.6424802083091124e-06,
-      "loss": 0.6968,
+      "loss": 0.5347,
       "step": 2262
     },
     {
       "epoch": 2.1843629343629343,
-      "grad_norm": 4.962578773498535,
+      "grad_norm": 4.518887996673584,
       "learning_rate": 3.634437961184608e-06,
-      "loss": 0.6495,
+      "loss": 0.4982,
       "step": 2263
     },
     {
       "epoch": 2.1853281853281854,
-      "grad_norm": 4.9937262535095215,
+      "grad_norm": 4.86975622177124,
       "learning_rate": 3.6264026299801758e-06,
-      "loss": 0.7027,
+      "loss": 0.5722,
       "step": 2264
     },
     {
       "epoch": 2.186293436293436,
-      "grad_norm": 4.739039897918701,
+      "grad_norm": 4.99892520904541,
       "learning_rate": 3.618374223425868e-06,
-      "loss": 0.7133,
+      "loss": 0.5542,
       "step": 2265
     },
     {
       "epoch": 2.187258687258687,
-      "grad_norm": 5.114686965942383,
+      "grad_norm": 5.10707426071167,
       "learning_rate": 3.6103527502442183e-06,
-      "loss": 0.7089,
+      "loss": 0.5729,
       "step": 2266
     },
     {
       "epoch": 2.1882239382239383,
-      "grad_norm": 4.9151835441589355,
+      "grad_norm": 4.810132026672363,
       "learning_rate": 3.6023382191502175e-06,
-      "loss": 0.663,
+      "loss": 0.5323,
       "step": 2267
     },
     {
       "epoch": 2.189189189189189,
-      "grad_norm": 5.096463203430176,
+      "grad_norm": 4.99933385848999,
       "learning_rate": 3.5943306388513333e-06,
-      "loss": 0.7173,
+      "loss": 0.5557,
       "step": 2268
     },
     {
       "epoch": 2.19015444015444,
-      "grad_norm": 4.799716472625732,
+      "grad_norm": 4.704514026641846,
       "learning_rate": 3.586330018047465e-06,
-      "loss": 0.6574,
+      "loss": 0.5082,
       "step": 2269
     },
     {
       "epoch": 2.1911196911196913,
-      "grad_norm": 4.545619010925293,
+      "grad_norm": 4.4531989097595215,
       "learning_rate": 3.57833636543096e-06,
-      "loss": 0.751,
+      "loss": 0.5847,
       "step": 2270
     },
     {
       "epoch": 2.192084942084942,
-      "grad_norm": 4.805863380432129,
+      "grad_norm": 4.692183017730713,
       "learning_rate": 3.570349689686583e-06,
-      "loss": 0.6806,
+      "loss": 0.5412,
       "step": 2271
     },
     {
       "epoch": 2.193050193050193,
-      "grad_norm": 4.8336286544799805,
+      "grad_norm": 4.530494213104248,
       "learning_rate": 3.5623699994915363e-06,
-      "loss": 0.6843,
+      "loss": 0.5256,
       "step": 2272
     },
     {
       "epoch": 2.1940154440154442,
-      "grad_norm": 4.659176826477051,
+      "grad_norm": 4.462390899658203,
       "learning_rate": 3.554397303515421e-06,
-      "loss": 0.6987,
+      "loss": 0.5394,
       "step": 2273
     },
     {
       "epoch": 2.194980694980695,
-      "grad_norm": 4.7669830322265625,
+      "grad_norm": 4.854856967926025,
       "learning_rate": 3.5464316104202403e-06,
-      "loss": 0.7378,
+      "loss": 0.5903,
       "step": 2274
     },
     {
       "epoch": 2.195945945945946,
-      "grad_norm": 5.27888298034668,
+      "grad_norm": 4.391628265380859,
       "learning_rate": 3.5384729288603912e-06,
-      "loss": 0.6762,
+      "loss": 0.5194,
       "step": 2275
     },
     {
       "epoch": 2.1969111969111967,
-      "grad_norm": 4.9906535148620605,
+      "grad_norm": 4.647050857543945,
       "learning_rate": 3.5305212674826526e-06,
-      "loss": 0.7317,
+      "loss": 0.5782,
       "step": 2276
     },
     {
       "epoch": 2.197876447876448,
-      "grad_norm": 4.950737476348877,
+      "grad_norm": 4.897363185882568,
       "learning_rate": 3.522576634926176e-06,
-      "loss": 0.7051,
+      "loss": 0.5389,
       "step": 2277
     },
     {
       "epoch": 2.198841698841699,
-      "grad_norm": 5.027462005615234,
+      "grad_norm": 5.346492767333984,
       "learning_rate": 3.514639039822475e-06,
-      "loss": 0.6887,
+      "loss": 0.5521,
       "step": 2278
     },
     {
       "epoch": 2.1998069498069497,
-      "grad_norm": 5.140257358551025,
+      "grad_norm": 4.49440336227417,
       "learning_rate": 3.506708490795424e-06,
-      "loss": 0.6656,
+      "loss": 0.5126,
       "step": 2279
     },
     {
       "epoch": 2.200772200772201,
-      "grad_norm": 5.102508544921875,
+      "grad_norm": 4.93634557723999,
       "learning_rate": 3.498784996461225e-06,
-      "loss": 0.6501,
+      "loss": 0.4962,
       "step": 2280
     },
     {
       "epoch": 2.201737451737452,
-      "grad_norm": 5.576716899871826,
+      "grad_norm": 5.390356540679932,
       "learning_rate": 3.490868565428439e-06,
-      "loss": 0.7047,
+      "loss": 0.5624,
       "step": 2281
     },
     {
       "epoch": 2.2027027027027026,
-      "grad_norm": 5.020506381988525,
+      "grad_norm": 4.874907970428467,
       "learning_rate": 3.4829592062979358e-06,
-      "loss": 0.7033,
+      "loss": 0.5425,
       "step": 2282
     },
     {
       "epoch": 2.2036679536679538,
-      "grad_norm": 5.018290996551514,
+      "grad_norm": 4.814367294311523,
       "learning_rate": 3.475056927662912e-06,
-      "loss": 0.7177,
+      "loss": 0.5832,
       "step": 2283
     },
     {
       "epoch": 2.2046332046332044,
-      "grad_norm": 5.415291786193848,
+      "grad_norm": 5.183836460113525,
       "learning_rate": 3.467161738108855e-06,
-      "loss": 0.7378,
+      "loss": 0.5789,
       "step": 2284
     },
     {
       "epoch": 2.2055984555984556,
-      "grad_norm": 4.696223735809326,
+      "grad_norm": 4.565914154052734,
       "learning_rate": 3.4592736462135734e-06,
-      "loss": 0.6622,
+      "loss": 0.5097,
       "step": 2285
     },
     {
       "epoch": 2.2065637065637067,
-      "grad_norm": 5.0370635986328125,
+      "grad_norm": 4.881713390350342,
       "learning_rate": 3.4513926605471504e-06,
-      "loss": 0.7077,
+      "loss": 0.5648,
       "step": 2286
     },
     {
       "epoch": 2.2075289575289574,
-      "grad_norm": 5.2918782234191895,
+      "grad_norm": 5.019075393676758,
       "learning_rate": 3.443518789671949e-06,
-      "loss": 0.7347,
+      "loss": 0.5588,
       "step": 2287
     },
     {
       "epoch": 2.2084942084942085,
-      "grad_norm": 4.940263271331787,
+      "grad_norm": 4.97376823425293,
       "learning_rate": 3.4356520421426063e-06,
-      "loss": 0.7493,
+      "loss": 0.598,
       "step": 2288
     },
     {
       "epoch": 2.2094594594594597,
-      "grad_norm": 4.90836763381958,
+      "grad_norm": 5.205156326293945,
       "learning_rate": 3.427792426506017e-06,
-      "loss": 0.6653,
+      "loss": 0.523,
       "step": 2289
     },
     {
       "epoch": 2.2104247104247103,
-      "grad_norm": 4.963898658752441,
+      "grad_norm": 4.711502552032471,
       "learning_rate": 3.4199399513013297e-06,
-      "loss": 0.709,
+      "loss": 0.5581,
       "step": 2290
     },
     {
       "epoch": 2.2113899613899615,
-      "grad_norm": 4.532299041748047,
+      "grad_norm": 4.557409763336182,
       "learning_rate": 3.412094625059934e-06,
-      "loss": 0.6859,
+      "loss": 0.5493,
       "step": 2291
     },
     {
       "epoch": 2.212355212355212,
-      "grad_norm": 5.141048908233643,
+      "grad_norm": 5.038810729980469,
       "learning_rate": 3.4042564563054525e-06,
-      "loss": 0.68,
+      "loss": 0.5282,
       "step": 2292
     },
     {
       "epoch": 2.2133204633204633,
-      "grad_norm": 5.234963417053223,
+      "grad_norm": 5.306894779205322,
       "learning_rate": 3.3964254535537322e-06,
-      "loss": 0.7707,
+      "loss": 0.6165,
       "step": 2293
     },
     {
       "epoch": 2.2142857142857144,
-      "grad_norm": 5.121557235717773,
+      "grad_norm": 5.158482074737549,
       "learning_rate": 3.388601625312833e-06,
-      "loss": 0.7329,
+      "loss": 0.587,
       "step": 2294
     },
     {
       "epoch": 2.215250965250965,
-      "grad_norm": 4.665938377380371,
+      "grad_norm": 4.594370365142822,
       "learning_rate": 3.3807849800830207e-06,
-      "loss": 0.7215,
+      "loss": 0.558,
       "step": 2295
     },
     {
       "epoch": 2.2162162162162162,
-      "grad_norm": 5.325049877166748,
+      "grad_norm": 4.522037029266357,
       "learning_rate": 3.3729755263567577e-06,
-      "loss": 0.7204,
+      "loss": 0.5567,
       "step": 2296
     },
     {
       "epoch": 2.2171814671814674,
-      "grad_norm": 4.914840221405029,
+      "grad_norm": 4.856266498565674,
       "learning_rate": 3.3651732726186924e-06,
-      "loss": 0.7348,
+      "loss": 0.5749,
       "step": 2297
     },
     {
       "epoch": 2.218146718146718,
-      "grad_norm": 4.860520362854004,
+      "grad_norm": 4.713069438934326,
       "learning_rate": 3.3573782273456523e-06,
-      "loss": 0.7138,
+      "loss": 0.5696,
       "step": 2298
     },
     {
       "epoch": 2.219111969111969,
-      "grad_norm": 5.827704429626465,
+      "grad_norm": 4.884941577911377,
       "learning_rate": 3.3495903990066315e-06,
-      "loss": 0.6851,
+      "loss": 0.5262,
       "step": 2299
     },
     {
       "epoch": 2.2200772200772203,
-      "grad_norm": 4.654504776000977,
+      "grad_norm": 4.548620700836182,
       "learning_rate": 3.341809796062783e-06,
-      "loss": 0.6777,
+      "loss": 0.553,
       "step": 2300
     },
     {
       "epoch": 2.221042471042471,
-      "grad_norm": 5.1855268478393555,
+      "grad_norm": 4.656248569488525,
       "learning_rate": 3.33403642696741e-06,
-      "loss": 0.734,
+      "loss": 0.5789,
       "step": 2301
     },
     {
       "epoch": 2.222007722007722,
-      "grad_norm": 4.997435092926025,
+      "grad_norm": 4.80281925201416,
       "learning_rate": 3.3262703001659567e-06,
-      "loss": 0.7135,
+      "loss": 0.5551,
       "step": 2302
     },
     {
       "epoch": 2.222972972972973,
-      "grad_norm": 5.055532932281494,
+      "grad_norm": 4.523697376251221,
       "learning_rate": 3.318511424096e-06,
-      "loss": 0.7069,
+      "loss": 0.5237,
       "step": 2303
     },
     {
       "epoch": 2.223938223938224,
-      "grad_norm": 5.179004192352295,
+      "grad_norm": 4.894985675811768,
       "learning_rate": 3.310759807187238e-06,
-      "loss": 0.7401,
+      "loss": 0.5902,
       "step": 2304
     },
     {
       "epoch": 2.224903474903475,
-      "grad_norm": 6.071235656738281,
+      "grad_norm": 5.404079437255859,
       "learning_rate": 3.3030154578614783e-06,
-      "loss": 0.7607,
+      "loss": 0.6052,
       "step": 2305
     },
     {
       "epoch": 2.2258687258687258,
-      "grad_norm": 4.733060836791992,
+      "grad_norm": 4.528689861297607,
       "learning_rate": 3.2952783845326465e-06,
-      "loss": 0.6579,
+      "loss": 0.4928,
       "step": 2306
     },
     {
       "epoch": 2.226833976833977,
-      "grad_norm": 4.739913463592529,
+      "grad_norm": 4.634459972381592,
       "learning_rate": 3.287548595606742e-06,
-      "loss": 0.722,
+      "loss": 0.5334,
       "step": 2307
     },
     {
       "epoch": 2.2277992277992276,
-      "grad_norm": 5.429046154022217,
+      "grad_norm": 4.733312129974365,
       "learning_rate": 3.2798260994818663e-06,
-      "loss": 0.6672,
+      "loss": 0.5035,
       "step": 2308
     },
     {
       "epoch": 2.2287644787644787,
-      "grad_norm": 4.519412994384766,
+      "grad_norm": 4.526803016662598,
       "learning_rate": 3.2721109045481867e-06,
-      "loss": 0.6304,
+      "loss": 0.4891,
       "step": 2309
     },
     {
       "epoch": 2.22972972972973,
-      "grad_norm": 4.585423946380615,
+      "grad_norm": 4.827352523803711,
       "learning_rate": 3.2644030191879563e-06,
-      "loss": 0.6756,
+      "loss": 0.5093,
       "step": 2310
     },
     {
       "epoch": 2.2306949806949805,
-      "grad_norm": 4.876979351043701,
+      "grad_norm": 4.685091972351074,
       "learning_rate": 3.256702451775461e-06,
-      "loss": 0.7007,
+      "loss": 0.5543,
       "step": 2311
     },
     {
       "epoch": 2.2316602316602316,
-      "grad_norm": 5.468658447265625,
+      "grad_norm": 5.041427135467529,
       "learning_rate": 3.2490092106770545e-06,
-      "loss": 0.7515,
+      "loss": 0.5883,
       "step": 2312
     },
     {
       "epoch": 2.2326254826254828,
-      "grad_norm": 4.656357288360596,
+      "grad_norm": 4.662158966064453,
       "learning_rate": 3.2413233042511203e-06,
-      "loss": 0.7121,
+      "loss": 0.558,
       "step": 2313
     },
     {
       "epoch": 2.2335907335907335,
-      "grad_norm": 4.834227085113525,
+      "grad_norm": 4.692500591278076,
       "learning_rate": 3.2336447408480874e-06,
-      "loss": 0.6866,
+      "loss": 0.5473,
       "step": 2314
     },
     {
       "epoch": 2.2345559845559846,
-      "grad_norm": 4.927269458770752,
+      "grad_norm": 5.2263407707214355,
       "learning_rate": 3.2259735288103956e-06,
-      "loss": 0.7328,
+      "loss": 0.5753,
       "step": 2315
     },
     {
       "epoch": 2.2355212355212357,
-      "grad_norm": 4.478086471557617,
+      "grad_norm": 4.390286445617676,
       "learning_rate": 3.218309676472492e-06,
-      "loss": 0.651,
+      "loss": 0.5004,
       "step": 2316
     },
     {
       "epoch": 2.2364864864864864,
-      "grad_norm": 5.098201751708984,
+      "grad_norm": 5.016356468200684,
       "learning_rate": 3.210653192160842e-06,
-      "loss": 0.7066,
+      "loss": 0.5401,
       "step": 2317
     },
     {
       "epoch": 2.2374517374517375,
-      "grad_norm": 5.49979829788208,
+      "grad_norm": 4.763302326202393,
       "learning_rate": 3.2030040841938927e-06,
-      "loss": 0.7486,
+      "loss": 0.5606,
       "step": 2318
     },
     {
       "epoch": 2.238416988416988,
-      "grad_norm": 5.119729995727539,
+      "grad_norm": 4.53725528717041,
       "learning_rate": 3.195362360882096e-06,
-      "loss": 0.6676,
+      "loss": 0.5096,
       "step": 2319
     },
     {
       "epoch": 2.2393822393822393,
-      "grad_norm": 5.401087760925293,
+      "grad_norm": 5.008630752563477,
       "learning_rate": 3.1877280305278568e-06,
-      "loss": 0.7287,
+      "loss": 0.5501,
       "step": 2320
     },
     {
       "epoch": 2.2403474903474905,
-      "grad_norm": 4.851002216339111,
+      "grad_norm": 5.233468532562256,
       "learning_rate": 3.1801011014255645e-06,
-      "loss": 0.7232,
+      "loss": 0.5802,
       "step": 2321
     },
     {
       "epoch": 2.241312741312741,
-      "grad_norm": 5.39402437210083,
+      "grad_norm": 5.044122695922852,
       "learning_rate": 3.1724815818615584e-06,
-      "loss": 0.7458,
+      "loss": 0.5855,
       "step": 2322
     },
     {
       "epoch": 2.2422779922779923,
-      "grad_norm": 5.446462631225586,
+      "grad_norm": 5.23109245300293,
       "learning_rate": 3.164869480114138e-06,
-      "loss": 0.7149,
+      "loss": 0.5608,
       "step": 2323
     },
     {
       "epoch": 2.2432432432432434,
-      "grad_norm": 4.565616130828857,
+      "grad_norm": 4.674023628234863,
       "learning_rate": 3.157264804453537e-06,
-      "loss": 0.6811,
+      "loss": 0.5229,
       "step": 2324
     },
     {
       "epoch": 2.244208494208494,
-      "grad_norm": 5.289637088775635,
+      "grad_norm": 5.171153545379639,
       "learning_rate": 3.1496675631419147e-06,
-      "loss": 0.7427,
+      "loss": 0.5845,
       "step": 2325
     },
     {
       "epoch": 2.2451737451737452,
-      "grad_norm": 5.112846851348877,
+      "grad_norm": 4.84445333480835,
       "learning_rate": 3.1420777644333613e-06,
-      "loss": 0.7339,
+      "loss": 0.5626,
       "step": 2326
     },
     {
       "epoch": 2.2461389961389964,
-      "grad_norm": 5.118411540985107,
+      "grad_norm": 4.888006210327148,
       "learning_rate": 3.134495416573884e-06,
-      "loss": 0.6778,
+      "loss": 0.5123,
       "step": 2327
     },
     {
       "epoch": 2.247104247104247,
-      "grad_norm": 4.570213794708252,
+      "grad_norm": 4.726391792297363,
       "learning_rate": 3.1269205278013903e-06,
-      "loss": 0.6523,
+      "loss": 0.5165,
       "step": 2328
     },
     {
       "epoch": 2.248069498069498,
-      "grad_norm": 4.825456142425537,
+      "grad_norm": 5.236876964569092,
       "learning_rate": 3.1193531063456785e-06,
-      "loss": 0.6886,
+      "loss": 0.5502,
       "step": 2329
     },
     {
       "epoch": 2.249034749034749,
-      "grad_norm": 4.914466381072998,
+      "grad_norm": 4.663729190826416,
       "learning_rate": 3.111793160428439e-06,
-      "loss": 0.7074,
+      "loss": 0.5629,
       "step": 2330
     },
     {
       "epoch": 2.25,
-      "grad_norm": 5.233219146728516,
+      "grad_norm": 4.998804092407227,
       "learning_rate": 3.1042406982632467e-06,
-      "loss": 0.7019,
+      "loss": 0.5557,
       "step": 2331
     },
     {
       "epoch": 2.250965250965251,
-      "grad_norm": 5.194098472595215,
+      "grad_norm": 5.190543174743652,
       "learning_rate": 3.0966957280555367e-06,
-      "loss": 0.7208,
+      "loss": 0.5795,
       "step": 2332
     },
     {
       "epoch": 2.251930501930502,
-      "grad_norm": 4.804540157318115,
+      "grad_norm": 5.029326438903809,
       "learning_rate": 3.08915825800261e-06,
-      "loss": 0.7185,
+      "loss": 0.5703,
       "step": 2333
     },
     {
       "epoch": 2.252895752895753,
-      "grad_norm": 5.275886535644531,
+      "grad_norm": 4.96004581451416,
       "learning_rate": 3.0816282962936083e-06,
-      "loss": 0.7112,
+      "loss": 0.5454,
       "step": 2334
     },
     {
       "epoch": 2.2538610038610036,
-      "grad_norm": 5.222113132476807,
+      "grad_norm": 4.945067405700684,
       "learning_rate": 3.074105851109531e-06,
-      "loss": 0.6777,
+      "loss": 0.5142,
       "step": 2335
     },
     {
       "epoch": 2.2548262548262548,
-      "grad_norm": 4.733977317810059,
+      "grad_norm": 4.568992614746094,
       "learning_rate": 3.0665909306232023e-06,
-      "loss": 0.6635,
+      "loss": 0.5114,
       "step": 2336
     },
     {
       "epoch": 2.255791505791506,
-      "grad_norm": 4.772053241729736,
+      "grad_norm": 4.581787586212158,
       "learning_rate": 3.0590835429992747e-06,
-      "loss": 0.7097,
+      "loss": 0.5268,
       "step": 2337
     },
     {
       "epoch": 2.2567567567567566,
-      "grad_norm": 5.085040092468262,
+      "grad_norm": 4.966142654418945,
       "learning_rate": 3.0515836963942056e-06,
-      "loss": 0.715,
+      "loss": 0.5576,
       "step": 2338
     },
     {
       "epoch": 2.2577220077220077,
-      "grad_norm": 5.072872161865234,
+      "grad_norm": 4.931494235992432,
       "learning_rate": 3.044091398956279e-06,
-      "loss": 0.7268,
+      "loss": 0.5833,
       "step": 2339
     },
     {
       "epoch": 2.258687258687259,
-      "grad_norm": 4.757853031158447,
+      "grad_norm": 4.844551086425781,
       "learning_rate": 3.0366066588255616e-06,
-      "loss": 0.7046,
+      "loss": 0.5304,
       "step": 2340
     },
     {
       "epoch": 2.2596525096525095,
-      "grad_norm": 5.22769021987915,
+      "grad_norm": 5.411068916320801,
       "learning_rate": 3.029129484133915e-06,
-      "loss": 0.734,
+      "loss": 0.5762,
       "step": 2341
     },
     {
       "epoch": 2.2606177606177607,
-      "grad_norm": 4.824207305908203,
+      "grad_norm": 4.746107578277588,
       "learning_rate": 3.02165988300498e-06,
-      "loss": 0.7313,
+      "loss": 0.5545,
       "step": 2342
     },
     {
       "epoch": 2.261583011583012,
-      "grad_norm": 4.599949359893799,
+      "grad_norm": 4.698272705078125,
       "learning_rate": 3.014197863554171e-06,
-      "loss": 0.7436,
+      "loss": 0.5595,
       "step": 2343
     },
     {
       "epoch": 2.2625482625482625,
-      "grad_norm": 4.781702041625977,
+      "grad_norm": 5.015766620635986,
       "learning_rate": 3.0067434338886623e-06,
-      "loss": 0.7186,
+      "loss": 0.5735,
       "step": 2344
     },
     {
       "epoch": 2.2635135135135136,
-      "grad_norm": 4.790346145629883,
+      "grad_norm": 4.890917778015137,
       "learning_rate": 2.9992966021073834e-06,
-      "loss": 0.7361,
+      "loss": 0.5738,
       "step": 2345
     },
     {
       "epoch": 2.2644787644787643,
-      "grad_norm": 4.375913143157959,
+      "grad_norm": 4.693212032318115,
       "learning_rate": 2.9918573763010106e-06,
-      "loss": 0.7012,
+      "loss": 0.5509,
       "step": 2346
     },
     {
       "epoch": 2.2654440154440154,
-      "grad_norm": 4.515613079071045,
+      "grad_norm": 4.405959606170654,
       "learning_rate": 2.984425764551954e-06,
-      "loss": 0.6792,
+      "loss": 0.5178,
       "step": 2347
     },
     {
       "epoch": 2.2664092664092665,
-      "grad_norm": 5.460712909698486,
+      "grad_norm": 4.953944683074951,
       "learning_rate": 2.9770017749343526e-06,
-      "loss": 0.7358,
+      "loss": 0.5527,
       "step": 2348
     },
     {
       "epoch": 2.2673745173745172,
-      "grad_norm": 4.83195686340332,
+      "grad_norm": 4.878808498382568,
       "learning_rate": 2.9695854155140648e-06,
-      "loss": 0.6767,
+      "loss": 0.5284,
       "step": 2349
     },
     {
       "epoch": 2.2683397683397684,
-      "grad_norm": 4.955194473266602,
+      "grad_norm": 4.7936296463012695,
       "learning_rate": 2.9621766943486597e-06,
-      "loss": 0.7982,
+      "loss": 0.6135,
       "step": 2350
     },
     {
       "epoch": 2.2693050193050195,
-      "grad_norm": 4.737084865570068,
+      "grad_norm": 4.905156135559082,
       "learning_rate": 2.9547756194874046e-06,
-      "loss": 0.7212,
+      "loss": 0.5624,
       "step": 2351
     },
     {
       "epoch": 2.27027027027027,
-      "grad_norm": 4.95520544052124,
+      "grad_norm": 5.144260406494141,
       "learning_rate": 2.9473821989712626e-06,
-      "loss": 0.7618,
+      "loss": 0.5871,
       "step": 2352
     },
     {
       "epoch": 2.2712355212355213,
-      "grad_norm": 5.034097194671631,
+      "grad_norm": 4.95668888092041,
       "learning_rate": 2.9399964408328795e-06,
-      "loss": 0.7017,
+      "loss": 0.5198,
       "step": 2353
     },
     {
       "epoch": 2.2722007722007724,
-      "grad_norm": 5.054235458374023,
+      "grad_norm": 4.962818622589111,
       "learning_rate": 2.932618353096576e-06,
-      "loss": 0.7333,
+      "loss": 0.5783,
       "step": 2354
     },
     {
       "epoch": 2.273166023166023,
-      "grad_norm": 4.866247653961182,
+      "grad_norm": 5.21099853515625,
       "learning_rate": 2.925247943778339e-06,
-      "loss": 0.7135,
+      "loss": 0.531,
       "step": 2355
     },
     {
       "epoch": 2.2741312741312742,
-      "grad_norm": 4.6588640213012695,
+      "grad_norm": 4.706159591674805,
       "learning_rate": 2.9178852208858146e-06,
-      "loss": 0.7327,
+      "loss": 0.561,
       "step": 2356
     },
     {
       "epoch": 2.275096525096525,
-      "grad_norm": 5.072292327880859,
+      "grad_norm": 4.786400318145752,
       "learning_rate": 2.9105301924182972e-06,
-      "loss": 0.7306,
+      "loss": 0.559,
       "step": 2357
     },
     {
       "epoch": 2.276061776061776,
-      "grad_norm": 5.094712257385254,
+      "grad_norm": 5.721562385559082,
       "learning_rate": 2.9031828663667194e-06,
-      "loss": 0.7504,
+      "loss": 0.5842,
       "step": 2358
     },
     {
       "epoch": 2.277027027027027,
-      "grad_norm": 4.930572032928467,
+      "grad_norm": 4.93635892868042,
       "learning_rate": 2.8958432507136492e-06,
-      "loss": 0.7214,
+      "loss": 0.5625,
       "step": 2359
     },
     {
       "epoch": 2.277992277992278,
-      "grad_norm": 5.065821647644043,
+      "grad_norm": 5.138400077819824,
       "learning_rate": 2.8885113534332742e-06,
-      "loss": 0.6841,
+      "loss": 0.5295,
       "step": 2360
     },
     {
       "epoch": 2.278957528957529,
-      "grad_norm": 4.5738043785095215,
+      "grad_norm": 4.554810523986816,
       "learning_rate": 2.8811871824913993e-06,
-      "loss": 0.7253,
+      "loss": 0.5595,
       "step": 2361
     },
     {
       "epoch": 2.2799227799227797,
-      "grad_norm": 5.412510395050049,
+      "grad_norm": 5.036932468414307,
       "learning_rate": 2.873870745845433e-06,
-      "loss": 0.7534,
+      "loss": 0.5786,
       "step": 2362
     },
     {
       "epoch": 2.280888030888031,
-      "grad_norm": 5.102321624755859,
+      "grad_norm": 4.842325210571289,
       "learning_rate": 2.8665620514443805e-06,
-      "loss": 0.7576,
+      "loss": 0.6014,
       "step": 2363
     },
     {
       "epoch": 2.281853281853282,
-      "grad_norm": 4.808786869049072,
+      "grad_norm": 4.507662296295166,
       "learning_rate": 2.8592611072288356e-06,
-      "loss": 0.7234,
+      "loss": 0.5541,
       "step": 2364
     },
     {
       "epoch": 2.2828185328185326,
-      "grad_norm": 5.819774150848389,
+      "grad_norm": 4.973712921142578,
       "learning_rate": 2.8519679211309746e-06,
-      "loss": 0.7136,
+      "loss": 0.5297,
       "step": 2365
     },
     {
       "epoch": 2.2837837837837838,
-      "grad_norm": 5.263688564300537,
+      "grad_norm": 5.050879955291748,
       "learning_rate": 2.8446825010745427e-06,
-      "loss": 0.7136,
+      "loss": 0.5395,
       "step": 2366
     },
     {
       "epoch": 2.284749034749035,
-      "grad_norm": 5.062206745147705,
+      "grad_norm": 5.031006813049316,
       "learning_rate": 2.8374048549748457e-06,
-      "loss": 0.7149,
+      "loss": 0.5641,
       "step": 2367
     },
     {
       "epoch": 2.2857142857142856,
-      "grad_norm": 5.310696601867676,
+      "grad_norm": 4.888494491577148,
       "learning_rate": 2.8301349907387464e-06,
-      "loss": 0.7304,
+      "loss": 0.5502,
       "step": 2368
     },
     {
       "epoch": 2.2866795366795367,
-      "grad_norm": 5.125802516937256,
+      "grad_norm": 5.165656566619873,
       "learning_rate": 2.8228729162646518e-06,
-      "loss": 0.7395,
+      "loss": 0.5667,
       "step": 2369
     },
     {
       "epoch": 2.287644787644788,
-      "grad_norm": 4.984714508056641,
+      "grad_norm": 4.976781368255615,
       "learning_rate": 2.8156186394425054e-06,
-      "loss": 0.7306,
+      "loss": 0.5656,
       "step": 2370
     },
     {
       "epoch": 2.2886100386100385,
-      "grad_norm": 5.050137042999268,
+      "grad_norm": 4.874149322509766,
       "learning_rate": 2.80837216815378e-06,
-      "loss": 0.6995,
+      "loss": 0.5462,
       "step": 2371
     },
     {
       "epoch": 2.2895752895752897,
-      "grad_norm": 4.933598518371582,
+      "grad_norm": 4.988937854766846,
       "learning_rate": 2.801133510271463e-06,
-      "loss": 0.7144,
+      "loss": 0.552,
       "step": 2372
     },
     {
       "epoch": 2.2905405405405403,
-      "grad_norm": 4.596554279327393,
+      "grad_norm": 4.351539611816406,
       "learning_rate": 2.7939026736600673e-06,
-      "loss": 0.7064,
+      "loss": 0.5259,
       "step": 2373
     },
     {
       "epoch": 2.2915057915057915,
-      "grad_norm": 5.07573127746582,
+      "grad_norm": 4.6414055824279785,
       "learning_rate": 2.7866796661755866e-06,
-      "loss": 0.7067,
+      "loss": 0.516,
       "step": 2374
     },
     {
       "epoch": 2.2924710424710426,
-      "grad_norm": 5.1375908851623535,
+      "grad_norm": 5.183434963226318,
       "learning_rate": 2.7794644956655272e-06,
-      "loss": 0.7172,
+      "loss": 0.5414,
       "step": 2375
     },
     {
       "epoch": 2.2934362934362933,
-      "grad_norm": 4.794432163238525,
+      "grad_norm": 4.560048580169678,
       "learning_rate": 2.772257169968866e-06,
-      "loss": 0.7097,
+      "loss": 0.5285,
       "step": 2376
     },
     {
       "epoch": 2.2944015444015444,
-      "grad_norm": 4.823713779449463,
+      "grad_norm": 4.737052917480469,
       "learning_rate": 2.765057696916077e-06,
-      "loss": 0.7465,
+      "loss": 0.5669,
       "step": 2377
     },
     {
       "epoch": 2.2953667953667956,
-      "grad_norm": 4.790879249572754,
+      "grad_norm": 4.826165199279785,
       "learning_rate": 2.7578660843290796e-06,
-      "loss": 0.7635,
+      "loss": 0.5883,
       "step": 2378
     },
     {
       "epoch": 2.2963320463320462,
-      "grad_norm": 5.0206298828125,
+      "grad_norm": 5.318382263183594,
       "learning_rate": 2.7506823400212657e-06,
-      "loss": 0.7172,
+      "loss": 0.5649,
       "step": 2379
     },
     {
       "epoch": 2.2972972972972974,
-      "grad_norm": 4.932159423828125,
+      "grad_norm": 4.804963111877441,
       "learning_rate": 2.743506471797476e-06,
-      "loss": 0.7068,
+      "loss": 0.5295,
       "step": 2380
     },
     {
       "epoch": 2.2982625482625485,
-      "grad_norm": 4.736297607421875,
+      "grad_norm": 4.60273551940918,
       "learning_rate": 2.736338487453999e-06,
-      "loss": 0.6866,
+      "loss": 0.5181,
       "step": 2381
     },
     {
       "epoch": 2.299227799227799,
-      "grad_norm": 4.674778461456299,
+      "grad_norm": 4.6513543128967285,
       "learning_rate": 2.7291783947785544e-06,
-      "loss": 0.7199,
+      "loss": 0.5488,
       "step": 2382
     },
     {
       "epoch": 2.3001930501930503,
-      "grad_norm": 4.690591335296631,
+      "grad_norm": 4.668834686279297,
       "learning_rate": 2.7220262015502796e-06,
-      "loss": 0.69,
+      "loss": 0.5244,
       "step": 2383
     },
     {
       "epoch": 2.301158301158301,
-      "grad_norm": 4.7273268699646,
+      "grad_norm": 4.76549768447876,
       "learning_rate": 2.714881915539739e-06,
-      "loss": 0.7404,
+      "loss": 0.5497,
       "step": 2384
     },
     {
       "epoch": 2.302123552123552,
-      "grad_norm": 4.441293239593506,
+      "grad_norm": 4.722814083099365,
       "learning_rate": 2.7077455445089075e-06,
-      "loss": 0.7021,
+      "loss": 0.546,
       "step": 2385
     },
     {
       "epoch": 2.3030888030888033,
-      "grad_norm": 5.465465545654297,
+      "grad_norm": 5.267011642456055,
       "learning_rate": 2.70061709621116e-06,
-      "loss": 0.7992,
+      "loss": 0.6124,
       "step": 2386
     },
     {
       "epoch": 2.304054054054054,
-      "grad_norm": 4.92411470413208,
+      "grad_norm": 4.639126300811768,
       "learning_rate": 2.6934965783912524e-06,
-      "loss": 0.7286,
+      "loss": 0.5568,
       "step": 2387
     },
     {
       "epoch": 2.305019305019305,
-      "grad_norm": 4.971828460693359,
+      "grad_norm": 4.984114646911621,
       "learning_rate": 2.686383998785336e-06,
-      "loss": 0.7351,
+      "loss": 0.5583,
       "step": 2388
     },
     {
       "epoch": 2.3059845559845558,
-      "grad_norm": 4.93446159362793,
+      "grad_norm": 4.8600382804870605,
       "learning_rate": 2.679279365120937e-06,
-      "loss": 0.6786,
+      "loss": 0.5092,
       "step": 2389
     },
     {
       "epoch": 2.306949806949807,
-      "grad_norm": 4.765533924102783,
+      "grad_norm": 4.589352607727051,
       "learning_rate": 2.6721826851169462e-06,
-      "loss": 0.6873,
+      "loss": 0.5125,
       "step": 2390
     },
     {
       "epoch": 2.307915057915058,
-      "grad_norm": 5.01896858215332,
+      "grad_norm": 4.553400039672852,
       "learning_rate": 2.665093966483615e-06,
-      "loss": 0.7841,
+      "loss": 0.5616,
       "step": 2391
     },
     {
       "epoch": 2.3088803088803087,
-      "grad_norm": 4.646848678588867,
+      "grad_norm": 4.156573295593262,
       "learning_rate": 2.6580132169225335e-06,
-      "loss": 0.6952,
+      "loss": 0.5162,
       "step": 2392
     },
     {
       "epoch": 2.30984555984556,
-      "grad_norm": 5.088491439819336,
+      "grad_norm": 5.311767578125,
       "learning_rate": 2.650940444126654e-06,
-      "loss": 0.7163,
+      "loss": 0.5328,
       "step": 2393
     },
     {
       "epoch": 2.310810810810811,
-      "grad_norm": 4.857966899871826,
+      "grad_norm": 4.902195930480957,
       "learning_rate": 2.643875655780248e-06,
-      "loss": 0.728,
+      "loss": 0.5609,
       "step": 2394
     },
     {
       "epoch": 2.3117760617760617,
-      "grad_norm": 5.659574031829834,
+      "grad_norm": 5.236776351928711,
       "learning_rate": 2.6368188595589183e-06,
-      "loss": 0.7304,
+      "loss": 0.5473,
       "step": 2395
     },
     {
       "epoch": 2.312741312741313,
-      "grad_norm": 4.86353874206543,
+      "grad_norm": 4.85888147354126,
       "learning_rate": 2.6297700631295776e-06,
-      "loss": 0.7056,
+      "loss": 0.53,
       "step": 2396
     },
     {
       "epoch": 2.313706563706564,
-      "grad_norm": 4.285360813140869,
+      "grad_norm": 4.260880947113037,
       "learning_rate": 2.6227292741504517e-06,
-      "loss": 0.6679,
+      "loss": 0.5061,
       "step": 2397
     },
     {
       "epoch": 2.3146718146718146,
-      "grad_norm": 4.985572338104248,
+      "grad_norm": 4.913229942321777,
       "learning_rate": 2.615696500271072e-06,
-      "loss": 0.6983,
+      "loss": 0.5266,
       "step": 2398
     },
     {
       "epoch": 2.3156370656370657,
-      "grad_norm": 5.162213325500488,
+      "grad_norm": 5.038549900054932,
       "learning_rate": 2.6086717491322545e-06,
-      "loss": 0.714,
+      "loss": 0.5104,
       "step": 2399
     },
     {
       "epoch": 2.3166023166023164,
-      "grad_norm": 4.907494068145752,
+      "grad_norm": 4.9835052490234375,
       "learning_rate": 2.6016550283661037e-06,
-      "loss": 0.7243,
+      "loss": 0.533,
       "step": 2400
     },
     {
       "epoch": 2.3175675675675675,
-      "grad_norm": 4.992336750030518,
+      "grad_norm": 5.3861470222473145,
       "learning_rate": 2.5946463455959868e-06,
-      "loss": 0.7138,
+      "loss": 0.5448,
       "step": 2401
     },
     {
       "epoch": 2.3185328185328187,
-      "grad_norm": 4.930575370788574,
+      "grad_norm": 4.8271050453186035,
       "learning_rate": 2.5876457084365593e-06,
-      "loss": 0.7534,
+      "loss": 0.5818,
       "step": 2402
     },
     {
       "epoch": 2.3194980694980694,
-      "grad_norm": 4.708988189697266,
+      "grad_norm": 4.743216514587402,
       "learning_rate": 2.5806531244937193e-06,
-      "loss": 0.7042,
+      "loss": 0.5184,
       "step": 2403
     },
     {
       "epoch": 2.3204633204633205,
-      "grad_norm": 4.472352504730225,
+      "grad_norm": 4.764296531677246,
       "learning_rate": 2.5736686013646226e-06,
-      "loss": 0.7182,
+      "loss": 0.5614,
       "step": 2404
     },
     {
       "epoch": 2.3214285714285716,
-      "grad_norm": 4.982179164886475,
+      "grad_norm": 5.0896100997924805,
       "learning_rate": 2.5666921466376583e-06,
-      "loss": 0.6801,
+      "loss": 0.5301,
       "step": 2405
     },
     {
       "epoch": 2.3223938223938223,
-      "grad_norm": 5.021693706512451,
+      "grad_norm": 4.842288017272949,
       "learning_rate": 2.5597237678924637e-06,
-      "loss": 0.7223,
+      "loss": 0.5394,
       "step": 2406
     },
     {
       "epoch": 2.3233590733590734,
-      "grad_norm": 4.943729400634766,
+      "grad_norm": 4.7943243980407715,
       "learning_rate": 2.5527634726998906e-06,
-      "loss": 0.7029,
+      "loss": 0.5209,
       "step": 2407
     },
     {
       "epoch": 2.3243243243243246,
-      "grad_norm": 5.22763204574585,
+      "grad_norm": 5.276082992553711,
       "learning_rate": 2.5458112686220136e-06,
-      "loss": 0.7379,
+      "loss": 0.549,
       "step": 2408
     },
     {
       "epoch": 2.3252895752895753,
-      "grad_norm": 5.044573783874512,
+      "grad_norm": 4.894245624542236,
       "learning_rate": 2.5388671632121144e-06,
-      "loss": 0.7035,
+      "loss": 0.5325,
       "step": 2409
     },
     {
       "epoch": 2.3262548262548264,
-      "grad_norm": 4.836273670196533,
+      "grad_norm": 4.723193168640137,
       "learning_rate": 2.531931164014676e-06,
-      "loss": 0.7247,
+      "loss": 0.5619,
       "step": 2410
     },
     {
       "epoch": 2.327220077220077,
-      "grad_norm": 5.079002857208252,
+      "grad_norm": 4.947766304016113,
       "learning_rate": 2.5250032785653744e-06,
-      "loss": 0.6896,
+      "loss": 0.5059,
       "step": 2411
     },
     {
       "epoch": 2.328185328185328,
-      "grad_norm": 5.008070945739746,
+      "grad_norm": 4.667328834533691,
       "learning_rate": 2.518083514391073e-06,
-      "loss": 0.7174,
+      "loss": 0.5367,
       "step": 2412
     },
     {
       "epoch": 2.3291505791505793,
-      "grad_norm": 4.878530502319336,
+      "grad_norm": 4.911200523376465,
       "learning_rate": 2.511171879009808e-06,
-      "loss": 0.7195,
+      "loss": 0.542,
       "step": 2413
     },
     {
       "epoch": 2.33011583011583,
-      "grad_norm": 4.816442489624023,
+      "grad_norm": 4.748602390289307,
       "learning_rate": 2.5042683799307864e-06,
-      "loss": 0.6529,
+      "loss": 0.4838,
       "step": 2414
     },
     {
       "epoch": 2.331081081081081,
-      "grad_norm": 4.850268363952637,
+      "grad_norm": 4.994816303253174,
       "learning_rate": 2.4973730246543736e-06,
-      "loss": 0.7392,
+      "loss": 0.5237,
       "step": 2415
     },
     {
       "epoch": 2.332046332046332,
-      "grad_norm": 5.145356178283691,
+      "grad_norm": 4.80022668838501,
       "learning_rate": 2.4904858206720892e-06,
-      "loss": 0.7219,
+      "loss": 0.5485,
       "step": 2416
     },
     {
       "epoch": 2.333011583011583,
-      "grad_norm": 4.7043375968933105,
+      "grad_norm": 4.905946731567383,
       "learning_rate": 2.483606775466595e-06,
-      "loss": 0.6624,
+      "loss": 0.4904,
       "step": 2417
     },
     {
       "epoch": 2.333976833976834,
-      "grad_norm": 4.561450004577637,
+      "grad_norm": 4.4718098640441895,
       "learning_rate": 2.4767358965116893e-06,
-      "loss": 0.6903,
+      "loss": 0.5211,
       "step": 2418
     },
     {
       "epoch": 2.3349420849420848,
-      "grad_norm": 4.79498291015625,
+      "grad_norm": 4.877486228942871,
       "learning_rate": 2.4698731912722983e-06,
-      "loss": 0.749,
+      "loss": 0.5781,
       "step": 2419
     },
     {
       "epoch": 2.335907335907336,
-      "grad_norm": 4.445528984069824,
+      "grad_norm": 4.576138973236084,
       "learning_rate": 2.4630186672044687e-06,
-      "loss": 0.7012,
+      "loss": 0.5247,
       "step": 2420
     },
     {
       "epoch": 2.336872586872587,
-      "grad_norm": 4.964884281158447,
+      "grad_norm": 4.736876487731934,
       "learning_rate": 2.4561723317553556e-06,
-      "loss": 0.7092,
+      "loss": 0.5292,
       "step": 2421
     },
     {
       "epoch": 2.3378378378378377,
-      "grad_norm": 4.606667995452881,
+      "grad_norm": 4.695431232452393,
       "learning_rate": 2.4493341923632195e-06,
-      "loss": 0.7299,
+      "loss": 0.5403,
       "step": 2422
     },
     {
       "epoch": 2.338803088803089,
-      "grad_norm": 5.229047775268555,
+      "grad_norm": 4.85709285736084,
       "learning_rate": 2.4425042564574186e-06,
-      "loss": 0.7875,
+      "loss": 0.6057,
       "step": 2423
     },
     {
       "epoch": 2.33976833976834,
-      "grad_norm": 5.045839786529541,
+      "grad_norm": 4.700776100158691,
       "learning_rate": 2.435682531458393e-06,
-      "loss": 0.7264,
+      "loss": 0.5477,
       "step": 2424
     },
     {
       "epoch": 2.3407335907335907,
-      "grad_norm": 5.008421421051025,
+      "grad_norm": 4.920938968658447,
       "learning_rate": 2.428869024777668e-06,
-      "loss": 0.7375,
+      "loss": 0.5659,
       "step": 2425
     },
     {
       "epoch": 2.341698841698842,
-      "grad_norm": 5.249884605407715,
+      "grad_norm": 4.95789098739624,
       "learning_rate": 2.422063743817832e-06,
-      "loss": 0.7812,
+      "loss": 0.5785,
       "step": 2426
     },
     {
       "epoch": 2.3426640926640925,
-      "grad_norm": 4.918494701385498,
+      "grad_norm": 4.656590938568115,
       "learning_rate": 2.4152666959725513e-06,
-      "loss": 0.7217,
+      "loss": 0.531,
       "step": 2427
     },
     {
       "epoch": 2.3436293436293436,
-      "grad_norm": 4.714098930358887,
+      "grad_norm": 4.613043308258057,
       "learning_rate": 2.408477888626527e-06,
-      "loss": 0.7027,
+      "loss": 0.5347,
       "step": 2428
     },
     {
       "epoch": 2.3445945945945947,
-      "grad_norm": 5.372392177581787,
+      "grad_norm": 5.030503273010254,
       "learning_rate": 2.4016973291555235e-06,
-      "loss": 0.7622,
+      "loss": 0.5769,
       "step": 2429
     },
     {
       "epoch": 2.3455598455598454,
-      "grad_norm": 4.705463409423828,
+      "grad_norm": 4.705520153045654,
       "learning_rate": 2.3949250249263333e-06,
-      "loss": 0.6922,
+      "loss": 0.5218,
       "step": 2430
     },
     {
       "epoch": 2.3465250965250966,
-      "grad_norm": 4.9439520835876465,
+      "grad_norm": 4.98171329498291,
       "learning_rate": 2.3881609832967934e-06,
-      "loss": 0.7497,
+      "loss": 0.5693,
       "step": 2431
     },
     {
       "epoch": 2.3474903474903477,
-      "grad_norm": 4.718496799468994,
+      "grad_norm": 4.874391555786133,
       "learning_rate": 2.3814052116157495e-06,
-      "loss": 0.7282,
+      "loss": 0.5637,
       "step": 2432
     },
     {
       "epoch": 2.3484555984555984,
-      "grad_norm": 4.703372955322266,
+      "grad_norm": 4.691925525665283,
       "learning_rate": 2.374657717223068e-06,
-      "loss": 0.7032,
+      "loss": 0.5246,
       "step": 2433
     },
     {
       "epoch": 2.3494208494208495,
-      "grad_norm": 4.857315540313721,
+      "grad_norm": 5.188472747802734,
       "learning_rate": 2.3679185074496247e-06,
-      "loss": 0.7654,
+      "loss": 0.5982,
       "step": 2434
     },
     {
       "epoch": 2.3503861003861006,
-      "grad_norm": 5.146064758300781,
+      "grad_norm": 5.233055114746094,
       "learning_rate": 2.361187589617292e-06,
-      "loss": 0.7491,
+      "loss": 0.5855,
       "step": 2435
     },
     {
       "epoch": 2.3513513513513513,
-      "grad_norm": 4.706638813018799,
+      "grad_norm": 4.739895343780518,
       "learning_rate": 2.354464971038933e-06,
-      "loss": 0.7369,
+      "loss": 0.5688,
       "step": 2436
     },
     {
       "epoch": 2.3523166023166024,
-      "grad_norm": 4.841254234313965,
+      "grad_norm": 4.980683326721191,
       "learning_rate": 2.347750659018397e-06,
-      "loss": 0.7545,
+      "loss": 0.5733,
       "step": 2437
     },
     {
       "epoch": 2.353281853281853,
-      "grad_norm": 4.904670715332031,
+      "grad_norm": 5.276948928833008,
       "learning_rate": 2.3410446608505044e-06,
-      "loss": 0.7085,
+      "loss": 0.5406,
       "step": 2438
     },
     {
       "epoch": 2.3542471042471043,
-      "grad_norm": 4.515750885009766,
+      "grad_norm": 4.721471309661865,
       "learning_rate": 2.3343469838210444e-06,
-      "loss": 0.6409,
+      "loss": 0.4915,
       "step": 2439
     },
     {
       "epoch": 2.3552123552123554,
-      "grad_norm": 4.757826805114746,
+      "grad_norm": 4.795527458190918,
       "learning_rate": 2.3276576352067725e-06,
-      "loss": 0.7561,
+      "loss": 0.5756,
       "step": 2440
     },
     {
       "epoch": 2.356177606177606,
-      "grad_norm": 4.912268161773682,
+      "grad_norm": 4.786769866943359,
       "learning_rate": 2.3209766222753814e-06,
-      "loss": 0.7219,
+      "loss": 0.5432,
       "step": 2441
     },
     {
       "epoch": 2.357142857142857,
-      "grad_norm": 4.534762382507324,
+      "grad_norm": 4.583132743835449,
       "learning_rate": 2.3143039522855204e-06,
-      "loss": 0.7202,
+      "loss": 0.5339,
       "step": 2442
     },
     {
       "epoch": 2.358108108108108,
-      "grad_norm": 4.946676254272461,
+      "grad_norm": 5.071953296661377,
       "learning_rate": 2.3076396324867643e-06,
-      "loss": 0.8139,
+      "loss": 0.6155,
       "step": 2443
     },
     {
       "epoch": 2.359073359073359,
-      "grad_norm": 4.679795742034912,
+      "grad_norm": 4.878026962280273,
       "learning_rate": 2.300983670119632e-06,
-      "loss": 0.6618,
+      "loss": 0.5089,
       "step": 2444
     },
     {
       "epoch": 2.36003861003861,
-      "grad_norm": 4.652212619781494,
+      "grad_norm": 4.868014335632324,
       "learning_rate": 2.2943360724155406e-06,
-      "loss": 0.7275,
+      "loss": 0.5368,
       "step": 2445
     },
     {
       "epoch": 2.361003861003861,
-      "grad_norm": 4.834445953369141,
+      "grad_norm": 4.96063756942749,
       "learning_rate": 2.2876968465968354e-06,
-      "loss": 0.7166,
+      "loss": 0.5238,
       "step": 2446
     },
     {
       "epoch": 2.361969111969112,
-      "grad_norm": 4.824753761291504,
+      "grad_norm": 4.64378547668457,
       "learning_rate": 2.281065999876756e-06,
-      "loss": 0.7317,
+      "loss": 0.5482,
       "step": 2447
     },
     {
       "epoch": 2.362934362934363,
-      "grad_norm": 5.067233085632324,
+      "grad_norm": 4.8883256912231445,
       "learning_rate": 2.27444353945945e-06,
-      "loss": 0.7301,
+      "loss": 0.5391,
       "step": 2448
     },
     {
       "epoch": 2.363899613899614,
-      "grad_norm": 5.02011775970459,
+      "grad_norm": 4.983039855957031,
       "learning_rate": 2.2678294725399465e-06,
-      "loss": 0.7139,
+      "loss": 0.5489,
       "step": 2449
     },
     {
       "epoch": 2.364864864864865,
-      "grad_norm": 4.794853687286377,
+      "grad_norm": 4.538534164428711,
       "learning_rate": 2.2612238063041515e-06,
-      "loss": 0.7047,
+      "loss": 0.5057,
       "step": 2450
     },
     {
       "epoch": 2.365830115830116,
-      "grad_norm": 5.3170599937438965,
+      "grad_norm": 5.266324520111084,
       "learning_rate": 2.2546265479288464e-06,
-      "loss": 0.7272,
+      "loss": 0.5445,
       "step": 2451
     },
     {
       "epoch": 2.3667953667953667,
-      "grad_norm": 4.975372791290283,
+      "grad_norm": 4.870588779449463,
       "learning_rate": 2.2480377045816858e-06,
-      "loss": 0.7372,
+      "loss": 0.5447,
       "step": 2452
     },
     {
       "epoch": 2.367760617760618,
-      "grad_norm": 5.004469394683838,
+      "grad_norm": 5.364177227020264,
       "learning_rate": 2.241457283421178e-06,
-      "loss": 0.7281,
+      "loss": 0.5393,
       "step": 2453
     },
     {
       "epoch": 2.3687258687258685,
-      "grad_norm": 5.1172380447387695,
+      "grad_norm": 4.956655025482178,
       "learning_rate": 2.23488529159667e-06,
-      "loss": 0.7877,
+      "loss": 0.5753,
       "step": 2454
     },
     {
       "epoch": 2.3696911196911197,
-      "grad_norm": 4.929131984710693,
+      "grad_norm": 4.822515964508057,
       "learning_rate": 2.22832173624836e-06,
-      "loss": 0.7335,
+      "loss": 0.5422,
       "step": 2455
     },
     {
       "epoch": 2.370656370656371,
-      "grad_norm": 4.840761184692383,
+      "grad_norm": 4.724470138549805,
       "learning_rate": 2.2217666245072864e-06,
-      "loss": 0.6732,
+      "loss": 0.4771,
       "step": 2456
     },
     {
       "epoch": 2.3716216216216215,
-      "grad_norm": 5.0260329246521,
+      "grad_norm": 4.903692722320557,
       "learning_rate": 2.2152199634953043e-06,
-      "loss": 0.7467,
+      "loss": 0.5387,
       "step": 2457
     },
     {
       "epoch": 2.3725868725868726,
-      "grad_norm": 4.6075439453125,
+      "grad_norm": 4.76098108291626,
       "learning_rate": 2.208681760325092e-06,
-      "loss": 0.6816,
+      "loss": 0.4984,
       "step": 2458
     },
     {
       "epoch": 2.3735521235521233,
-      "grad_norm": 5.479339599609375,
+      "grad_norm": 5.208647727966309,
       "learning_rate": 2.2021520221001304e-06,
-      "loss": 0.7389,
+      "loss": 0.5522,
       "step": 2459
     },
     {
       "epoch": 2.3745173745173744,
-      "grad_norm": 4.929660797119141,
+      "grad_norm": 4.945209503173828,
       "learning_rate": 2.195630755914716e-06,
-      "loss": 0.7207,
+      "loss": 0.5153,
       "step": 2460
     },
     {
       "epoch": 2.3754826254826256,
-      "grad_norm": 5.115887641906738,
+      "grad_norm": 4.85383939743042,
       "learning_rate": 2.189117968853933e-06,
-      "loss": 0.7422,
+      "loss": 0.5577,
       "step": 2461
     },
     {
       "epoch": 2.3764478764478767,
-      "grad_norm": 4.739163398742676,
+      "grad_norm": 4.616394519805908,
       "learning_rate": 2.1826136679936605e-06,
-      "loss": 0.7021,
+      "loss": 0.5265,
       "step": 2462
     },
     {
       "epoch": 2.3774131274131274,
-      "grad_norm": 4.605644226074219,
+      "grad_norm": 4.882351875305176,
       "learning_rate": 2.1761178604005418e-06,
-      "loss": 0.6962,
+      "loss": 0.5196,
       "step": 2463
     },
     {
       "epoch": 2.3783783783783785,
-      "grad_norm": 4.777822971343994,
+      "grad_norm": 4.6104912757873535,
       "learning_rate": 2.1696305531320115e-06,
-      "loss": 0.6872,
+      "loss": 0.5225,
       "step": 2464
     },
     {
       "epoch": 2.379343629343629,
-      "grad_norm": 5.216665744781494,
+      "grad_norm": 4.999615669250488,
       "learning_rate": 2.16315175323626e-06,
-      "loss": 0.733,
+      "loss": 0.5504,
       "step": 2465
     },
     {
       "epoch": 2.3803088803088803,
-      "grad_norm": 4.957310676574707,
+      "grad_norm": 5.243192195892334,
       "learning_rate": 2.156681467752234e-06,
-      "loss": 0.7233,
+      "loss": 0.5477,
       "step": 2466
     },
     {
       "epoch": 2.3812741312741315,
-      "grad_norm": 4.596826553344727,
+      "grad_norm": 4.59197473526001,
       "learning_rate": 2.1502197037096317e-06,
-      "loss": 0.7304,
+      "loss": 0.5554,
       "step": 2467
     },
     {
       "epoch": 2.382239382239382,
-      "grad_norm": 4.776923656463623,
+      "grad_norm": 4.822226524353027,
       "learning_rate": 2.1437664681288926e-06,
-      "loss": 0.7066,
+      "loss": 0.5361,
       "step": 2468
     },
     {
       "epoch": 2.3832046332046333,
-      "grad_norm": 4.835448741912842,
+      "grad_norm": 5.079570293426514,
       "learning_rate": 2.1373217680211924e-06,
-      "loss": 0.7082,
+      "loss": 0.5453,
       "step": 2469
     },
     {
       "epoch": 2.384169884169884,
-      "grad_norm": 4.912600040435791,
+      "grad_norm": 5.263392448425293,
       "learning_rate": 2.130885610388428e-06,
-      "loss": 0.7089,
+      "loss": 0.5254,
       "step": 2470
     },
     {
       "epoch": 2.385135135135135,
-      "grad_norm": 5.509889125823975,
+      "grad_norm": 4.964388370513916,
       "learning_rate": 2.12445800222322e-06,
-      "loss": 0.7146,
+      "loss": 0.5146,
       "step": 2471
     },
     {
       "epoch": 2.386100386100386,
-      "grad_norm": 4.585874080657959,
+      "grad_norm": 4.747281074523926,
       "learning_rate": 2.1180389505089007e-06,
-      "loss": 0.7554,
+      "loss": 0.5652,
       "step": 2472
     },
     {
       "epoch": 2.387065637065637,
-      "grad_norm": 4.63887357711792,
+      "grad_norm": 4.432153701782227,
       "learning_rate": 2.1116284622195018e-06,
-      "loss": 0.7053,
+      "loss": 0.543,
       "step": 2473
     },
     {
       "epoch": 2.388030888030888,
-      "grad_norm": 5.154671669006348,
+      "grad_norm": 4.785538196563721,
       "learning_rate": 2.105226544319756e-06,
-      "loss": 0.7142,
+      "loss": 0.5232,
       "step": 2474
     },
     {
       "epoch": 2.388996138996139,
-      "grad_norm": 4.6797566413879395,
+      "grad_norm": 4.856302261352539,
       "learning_rate": 2.0988332037650816e-06,
-      "loss": 0.685,
+      "loss": 0.5285,
       "step": 2475
     },
     {
       "epoch": 2.38996138996139,
-      "grad_norm": 4.851925373077393,
+      "grad_norm": 4.766124248504639,
       "learning_rate": 2.0924484475015783e-06,
-      "loss": 0.7357,
+      "loss": 0.5617,
       "step": 2476
     },
     {
       "epoch": 2.390926640926641,
-      "grad_norm": 5.000600337982178,
+      "grad_norm": 5.159885883331299,
       "learning_rate": 2.0860722824660217e-06,
-      "loss": 0.7511,
+      "loss": 0.5654,
       "step": 2477
     },
     {
       "epoch": 2.391891891891892,
-      "grad_norm": 4.8918352127075195,
+      "grad_norm": 4.937465667724609,
       "learning_rate": 2.0797047155858495e-06,
-      "loss": 0.7263,
+      "loss": 0.5571,
       "step": 2478
     },
     {
       "epoch": 2.392857142857143,
-      "grad_norm": 4.8164591789245605,
+      "grad_norm": 4.661976337432861,
       "learning_rate": 2.0733457537791614e-06,
-      "loss": 0.7303,
+      "loss": 0.5429,
       "step": 2479
     },
     {
       "epoch": 2.393822393822394,
-      "grad_norm": 5.150565147399902,
+      "grad_norm": 4.966793060302734,
       "learning_rate": 2.066995403954706e-06,
-      "loss": 0.7473,
+      "loss": 0.5658,
       "step": 2480
     },
     {
       "epoch": 2.3947876447876446,
-      "grad_norm": 4.9704742431640625,
+      "grad_norm": 4.89058256149292,
       "learning_rate": 2.0606536730118767e-06,
-      "loss": 0.7989,
+      "loss": 0.6114,
       "step": 2481
     },
     {
       "epoch": 2.3957528957528957,
-      "grad_norm": 4.806553363800049,
+      "grad_norm": 4.81455135345459,
       "learning_rate": 2.0543205678407e-06,
-      "loss": 0.7698,
+      "loss": 0.5816,
       "step": 2482
     },
     {
       "epoch": 2.396718146718147,
-      "grad_norm": 4.680107593536377,
+      "grad_norm": 4.495682239532471,
       "learning_rate": 2.047996095321835e-06,
-      "loss": 0.7222,
+      "loss": 0.5411,
       "step": 2483
     },
     {
       "epoch": 2.3976833976833976,
-      "grad_norm": 4.774284362792969,
+      "grad_norm": 4.705482482910156,
       "learning_rate": 2.0416802623265596e-06,
-      "loss": 0.7484,
+      "loss": 0.5562,
       "step": 2484
     },
     {
       "epoch": 2.3986486486486487,
-      "grad_norm": 5.096017360687256,
+      "grad_norm": 4.689180850982666,
       "learning_rate": 2.0353730757167634e-06,
-      "loss": 0.7417,
+      "loss": 0.557,
       "step": 2485
     },
     {
       "epoch": 2.3996138996138994,
-      "grad_norm": 4.331223011016846,
+      "grad_norm": 4.329068183898926,
       "learning_rate": 2.0290745423449465e-06,
-      "loss": 0.6919,
+      "loss": 0.5151,
       "step": 2486
     },
     {
       "epoch": 2.4005791505791505,
-      "grad_norm": 4.920547008514404,
+      "grad_norm": 5.295011043548584,
       "learning_rate": 2.0227846690542043e-06,
-      "loss": 0.7301,
+      "loss": 0.5564,
       "step": 2487
     },
     {
       "epoch": 2.4015444015444016,
-      "grad_norm": 5.212554454803467,
+      "grad_norm": 5.0405473709106445,
       "learning_rate": 2.016503462678222e-06,
-      "loss": 0.7452,
+      "loss": 0.5337,
       "step": 2488
     },
     {
       "epoch": 2.4025096525096523,
-      "grad_norm": 5.136991500854492,
+      "grad_norm": 5.489158630371094,
       "learning_rate": 2.010230930041274e-06,
-      "loss": 0.7219,
+      "loss": 0.5585,
       "step": 2489
     },
     {
       "epoch": 2.4034749034749034,
-      "grad_norm": 4.706001281738281,
+      "grad_norm": 4.395766258239746,
       "learning_rate": 2.0039670779582066e-06,
-      "loss": 0.6659,
+      "loss": 0.4763,
       "step": 2490
     },
     {
       "epoch": 2.4044401544401546,
-      "grad_norm": 5.0232133865356445,
+      "grad_norm": 5.271701812744141,
       "learning_rate": 1.997711913234435e-06,
-      "loss": 0.7514,
+      "loss": 0.5694,
       "step": 2491
     },
     {
       "epoch": 2.4054054054054053,
-      "grad_norm": 5.077817440032959,
+      "grad_norm": 4.626297473907471,
       "learning_rate": 1.9914654426659374e-06,
-      "loss": 0.7229,
+      "loss": 0.5148,
       "step": 2492
     },
     {
       "epoch": 2.4063706563706564,
-      "grad_norm": 5.236948490142822,
+      "grad_norm": 5.056756019592285,
       "learning_rate": 1.985227673039245e-06,
-      "loss": 0.7808,
+      "loss": 0.5954,
       "step": 2493
     },
     {
       "epoch": 2.4073359073359075,
-      "grad_norm": 4.87125825881958,
+      "grad_norm": 4.903386116027832,
       "learning_rate": 1.978998611131443e-06,
-      "loss": 0.7886,
+      "loss": 0.5818,
       "step": 2494
     },
     {
       "epoch": 2.408301158301158,
-      "grad_norm": 5.276411533355713,
+      "grad_norm": 5.086071968078613,
       "learning_rate": 1.9727782637101435e-06,
-      "loss": 0.7445,
+      "loss": 0.5391,
       "step": 2495
     },
     {
       "epoch": 2.4092664092664093,
-      "grad_norm": 5.497075080871582,
+      "grad_norm": 5.549103260040283,
       "learning_rate": 1.966566637533497e-06,
-      "loss": 0.7403,
+      "loss": 0.5664,
       "step": 2496
     },
     {
       "epoch": 2.41023166023166,
-      "grad_norm": 5.233363151550293,
+      "grad_norm": 5.048281192779541,
       "learning_rate": 1.960363739350177e-06,
-      "loss": 0.714,
+      "loss": 0.5107,
       "step": 2497
     },
     {
       "epoch": 2.411196911196911,
-      "grad_norm": 5.141330718994141,
+      "grad_norm": 5.3985466957092285,
       "learning_rate": 1.954169575899386e-06,
-      "loss": 0.7169,
+      "loss": 0.5328,
       "step": 2498
     },
     {
       "epoch": 2.4121621621621623,
-      "grad_norm": 4.863707542419434,
+      "grad_norm": 4.683662414550781,
       "learning_rate": 1.9479841539108156e-06,
-      "loss": 0.7577,
+      "loss": 0.5536,
       "step": 2499
     },
     {
       "epoch": 2.413127413127413,
-      "grad_norm": 5.165623188018799,
+      "grad_norm": 5.321349143981934,
       "learning_rate": 1.941807480104676e-06,
-      "loss": 0.7102,
+      "loss": 0.5276,
       "step": 2500
     },
     {
       "epoch": 2.414092664092664,
-      "grad_norm": 4.734787464141846,
+      "grad_norm": 4.7699432373046875,
       "learning_rate": 1.935639561191663e-06,
-      "loss": 0.7166,
+      "loss": 0.5324,
       "step": 2501
     },
     {
       "epoch": 2.4150579150579152,
-      "grad_norm": 5.163553714752197,
+      "grad_norm": 4.978326320648193,
       "learning_rate": 1.9294804038729743e-06,
-      "loss": 0.7574,
+      "loss": 0.5586,
       "step": 2502
     },
     {
       "epoch": 2.416023166023166,
-      "grad_norm": 5.120594024658203,
+      "grad_norm": 4.980433464050293,
       "learning_rate": 1.9233300148402767e-06,
-      "loss": 0.6955,
+      "loss": 0.5237,
       "step": 2503
     },
     {
       "epoch": 2.416988416988417,
-      "grad_norm": 5.021568298339844,
+      "grad_norm": 5.136814117431641,
       "learning_rate": 1.9171884007757093e-06,
-      "loss": 0.7281,
+      "loss": 0.5366,
       "step": 2504
     },
     {
       "epoch": 2.417953667953668,
-      "grad_norm": 4.861301422119141,
+      "grad_norm": 4.977025032043457,
       "learning_rate": 1.911055568351883e-06,
-      "loss": 0.7986,
+      "loss": 0.5867,
       "step": 2505
     },
     {
       "epoch": 2.418918918918919,
-      "grad_norm": 5.154460430145264,
+      "grad_norm": 5.0297651290893555,
       "learning_rate": 1.90493152423187e-06,
-      "loss": 0.6855,
+      "loss": 0.5025,
       "step": 2506
     },
     {
       "epoch": 2.41988416988417,
-      "grad_norm": 4.927275657653809,
+      "grad_norm": 4.725804805755615,
       "learning_rate": 1.898816275069193e-06,
-      "loss": 0.7509,
+      "loss": 0.5529,
       "step": 2507
     },
     {
       "epoch": 2.4208494208494207,
-      "grad_norm": 5.266952991485596,
+      "grad_norm": 4.853487014770508,
       "learning_rate": 1.8927098275078116e-06,
-      "loss": 0.7192,
+      "loss": 0.5117,
       "step": 2508
     },
     {
       "epoch": 2.421814671814672,
-      "grad_norm": 4.80417537689209,
+      "grad_norm": 4.7407612800598145,
       "learning_rate": 1.886612188182132e-06,
-      "loss": 0.7568,
+      "loss": 0.5515,
       "step": 2509
     },
     {
       "epoch": 2.422779922779923,
-      "grad_norm": 5.437583923339844,
+      "grad_norm": 4.966141700744629,
       "learning_rate": 1.8805233637169829e-06,
-      "loss": 0.6822,
+      "loss": 0.497,
       "step": 2510
     },
     {
       "epoch": 2.4237451737451736,
-      "grad_norm": 4.913822650909424,
+      "grad_norm": 4.955407619476318,
       "learning_rate": 1.8744433607276269e-06,
-      "loss": 0.7758,
+      "loss": 0.589,
       "step": 2511
     },
     {
       "epoch": 2.4247104247104247,
-      "grad_norm": 5.271089553833008,
+      "grad_norm": 5.178981781005859,
       "learning_rate": 1.8683721858197369e-06,
-      "loss": 0.7822,
+      "loss": 0.58,
       "step": 2512
     },
     {
       "epoch": 2.4256756756756754,
-      "grad_norm": 4.6216139793396,
+      "grad_norm": 4.665440082550049,
       "learning_rate": 1.8623098455893862e-06,
-      "loss": 0.7506,
+      "loss": 0.5554,
       "step": 2513
     },
     {
       "epoch": 2.4266409266409266,
-      "grad_norm": 4.506705284118652,
+      "grad_norm": 4.465180397033691,
       "learning_rate": 1.8562563466230577e-06,
-      "loss": 0.7118,
+      "loss": 0.5191,
       "step": 2514
     },
     {
       "epoch": 2.4276061776061777,
-      "grad_norm": 4.784862041473389,
+      "grad_norm": 4.388522148132324,
       "learning_rate": 1.8502116954976334e-06,
-      "loss": 0.7221,
+      "loss": 0.5283,
       "step": 2515
     },
     {
       "epoch": 2.4285714285714284,
-      "grad_norm": 5.663732528686523,
+      "grad_norm": 5.072386264801025,
       "learning_rate": 1.844175898780377e-06,
-      "loss": 0.7711,
+      "loss": 0.5616,
       "step": 2516
     },
     {
       "epoch": 2.4295366795366795,
-      "grad_norm": 4.902246952056885,
+      "grad_norm": 4.71058988571167,
       "learning_rate": 1.8381489630289273e-06,
-      "loss": 0.73,
+      "loss": 0.5353,
       "step": 2517
     },
     {
       "epoch": 2.4305019305019306,
-      "grad_norm": 4.988832473754883,
+      "grad_norm": 4.704596519470215,
       "learning_rate": 1.8321308947912997e-06,
-      "loss": 0.6607,
+      "loss": 0.4515,
       "step": 2518
     },
     {
       "epoch": 2.4314671814671813,
-      "grad_norm": 4.855826377868652,
+      "grad_norm": 4.8011603355407715,
       "learning_rate": 1.8261217006058828e-06,
-      "loss": 0.7027,
+      "loss": 0.5074,
       "step": 2519
     },
     {
       "epoch": 2.4324324324324325,
-      "grad_norm": 4.841279029846191,
+      "grad_norm": 4.964382171630859,
       "learning_rate": 1.8201213870014168e-06,
-      "loss": 0.7623,
+      "loss": 0.5755,
       "step": 2520
     },
     {
       "epoch": 2.4333976833976836,
-      "grad_norm": 5.101052761077881,
+      "grad_norm": 4.7918314933776855,
       "learning_rate": 1.8141299604969908e-06,
-      "loss": 0.7258,
+      "loss": 0.5234,
       "step": 2521
     },
     {
       "epoch": 2.4343629343629343,
-      "grad_norm": 5.110799312591553,
+      "grad_norm": 4.950772762298584,
       "learning_rate": 1.8081474276020406e-06,
-      "loss": 0.6804,
+      "loss": 0.496,
       "step": 2522
     },
     {
       "epoch": 2.4353281853281854,
-      "grad_norm": 5.3934125900268555,
+      "grad_norm": 4.8687520027160645,
       "learning_rate": 1.8021737948163465e-06,
-      "loss": 0.7129,
+      "loss": 0.5056,
       "step": 2523
     },
     {
       "epoch": 2.436293436293436,
-      "grad_norm": 5.148820877075195,
+      "grad_norm": 5.125008583068848,
       "learning_rate": 1.7962090686300116e-06,
-      "loss": 0.7749,
+      "loss": 0.5581,
       "step": 2524
     },
     {
       "epoch": 2.437258687258687,
-      "grad_norm": 4.828898906707764,
+      "grad_norm": 4.7674970626831055,
       "learning_rate": 1.7902532555234653e-06,
-      "loss": 0.7241,
+      "loss": 0.5316,
       "step": 2525
     },
     {
       "epoch": 2.4382239382239383,
-      "grad_norm": 4.930122375488281,
+      "grad_norm": 5.063551425933838,
       "learning_rate": 1.7843063619674473e-06,
-      "loss": 0.7818,
+      "loss": 0.6009,
       "step": 2526
     },
     {
       "epoch": 2.439189189189189,
-      "grad_norm": 4.935842514038086,
+      "grad_norm": 5.20596170425415,
       "learning_rate": 1.778368394423018e-06,
-      "loss": 0.7692,
+      "loss": 0.5833,
       "step": 2527
     },
     {
       "epoch": 2.44015444015444,
-      "grad_norm": 5.2509074211120605,
+      "grad_norm": 5.126626968383789,
       "learning_rate": 1.7724393593415302e-06,
-      "loss": 0.7382,
+      "loss": 0.5497,
       "step": 2528
     },
     {
       "epoch": 2.4411196911196913,
-      "grad_norm": 4.885476112365723,
+      "grad_norm": 5.051726818084717,
       "learning_rate": 1.76651926316464e-06,
-      "loss": 0.7435,
+      "loss": 0.5366,
       "step": 2529
     },
     {
       "epoch": 2.442084942084942,
-      "grad_norm": 5.482947826385498,
+      "grad_norm": 5.063938617706299,
       "learning_rate": 1.760608112324278e-06,
-      "loss": 0.7631,
+      "loss": 0.5567,
       "step": 2530
     },
     {
       "epoch": 2.443050193050193,
-      "grad_norm": 5.086392879486084,
+      "grad_norm": 5.195236682891846,
       "learning_rate": 1.7547059132426735e-06,
-      "loss": 0.7685,
+      "loss": 0.5374,
       "step": 2531
     },
     {
       "epoch": 2.4440154440154442,
-      "grad_norm": 5.346570014953613,
+      "grad_norm": 4.944610118865967,
       "learning_rate": 1.7488126723323185e-06,
-      "loss": 0.7837,
+      "loss": 0.5862,
       "step": 2532
     },
     {
       "epoch": 2.444980694980695,
-      "grad_norm": 6.160499095916748,
+      "grad_norm": 5.61105489730835,
       "learning_rate": 1.7429283959959776e-06,
-      "loss": 0.7581,
+      "loss": 0.5483,
       "step": 2533
     },
     {
       "epoch": 2.445945945945946,
-      "grad_norm": 5.070865154266357,
+      "grad_norm": 5.157047271728516,
       "learning_rate": 1.7370530906266724e-06,
-      "loss": 0.7536,
+      "loss": 0.5403,
       "step": 2534
     },
     {
       "epoch": 2.4469111969111967,
-      "grad_norm": 5.467529296875,
+      "grad_norm": 4.74455451965332,
       "learning_rate": 1.7311867626076795e-06,
-      "loss": 0.6851,
+      "loss": 0.4803,
       "step": 2535
     },
     {
       "epoch": 2.447876447876448,
-      "grad_norm": 5.569533824920654,
+      "grad_norm": 5.031951427459717,
       "learning_rate": 1.7253294183125223e-06,
-      "loss": 0.7724,
+      "loss": 0.5608,
       "step": 2536
     },
     {
       "epoch": 2.448841698841699,
-      "grad_norm": 4.970886707305908,
+      "grad_norm": 5.138765811920166,
       "learning_rate": 1.719481064104962e-06,
-      "loss": 0.7247,
+      "loss": 0.523,
       "step": 2537
     },
     {
       "epoch": 2.4498069498069497,
-      "grad_norm": 5.251448631286621,
+      "grad_norm": 5.185892105102539,
       "learning_rate": 1.7136417063389964e-06,
-      "loss": 0.7191,
+      "loss": 0.5348,
       "step": 2538
     },
     {
       "epoch": 2.450772200772201,
-      "grad_norm": 5.369867324829102,
+      "grad_norm": 5.227701663970947,
       "learning_rate": 1.7078113513588434e-06,
-      "loss": 0.7434,
+      "loss": 0.5271,
       "step": 2539
     },
     {
       "epoch": 2.4517374517374515,
-      "grad_norm": 4.895583152770996,
+      "grad_norm": 4.948006629943848,
       "learning_rate": 1.7019900054989446e-06,
-      "loss": 0.7787,
+      "loss": 0.5873,
       "step": 2540
     },
     {
       "epoch": 2.4527027027027026,
-      "grad_norm": 5.491269588470459,
+      "grad_norm": 5.098097324371338,
       "learning_rate": 1.6961776750839509e-06,
-      "loss": 0.7176,
+      "loss": 0.5254,
       "step": 2541
     },
     {
       "epoch": 2.4536679536679538,
-      "grad_norm": 4.857673645019531,
+      "grad_norm": 5.095490455627441,
       "learning_rate": 1.6903743664287197e-06,
-      "loss": 0.7052,
+      "loss": 0.5237,
       "step": 2542
     },
     {
       "epoch": 2.4546332046332044,
-      "grad_norm": 5.4430623054504395,
+      "grad_norm": 5.12826681137085,
       "learning_rate": 1.6845800858383055e-06,
-      "loss": 0.7157,
+      "loss": 0.5263,
       "step": 2543
     },
     {
       "epoch": 2.4555984555984556,
-      "grad_norm": 5.276530742645264,
+      "grad_norm": 5.058716297149658,
       "learning_rate": 1.6787948396079544e-06,
-      "loss": 0.7353,
+      "loss": 0.5212,
       "step": 2544
     },
     {
       "epoch": 2.4565637065637067,
-      "grad_norm": 5.065103054046631,
+      "grad_norm": 4.896842956542969,
       "learning_rate": 1.6730186340230981e-06,
-      "loss": 0.6974,
+      "loss": 0.4966,
       "step": 2545
     },
     {
       "epoch": 2.4575289575289574,
-      "grad_norm": 5.144730567932129,
+      "grad_norm": 5.015899658203125,
       "learning_rate": 1.667251475359344e-06,
-      "loss": 0.7688,
+      "loss": 0.5638,
       "step": 2546
     },
     {
       "epoch": 2.4584942084942085,
-      "grad_norm": 5.180382251739502,
+      "grad_norm": 5.2632951736450195,
       "learning_rate": 1.661493369882473e-06,
-      "loss": 0.7028,
+      "loss": 0.5032,
       "step": 2547
     },
     {
       "epoch": 2.4594594594594597,
-      "grad_norm": 5.192044734954834,
+      "grad_norm": 4.960742473602295,
       "learning_rate": 1.6557443238484283e-06,
-      "loss": 0.7423,
+      "loss": 0.5297,
       "step": 2548
     },
     {
       "epoch": 2.4604247104247103,
-      "grad_norm": 5.120516300201416,
+      "grad_norm": 5.411482334136963,
       "learning_rate": 1.6500043435033109e-06,
-      "loss": 0.6733,
+      "loss": 0.4959,
       "step": 2549
     },
     {
       "epoch": 2.4613899613899615,
-      "grad_norm": 4.712008953094482,
+      "grad_norm": 4.962620258331299,
       "learning_rate": 1.6442734350833721e-06,
-      "loss": 0.7562,
+      "loss": 0.5451,
       "step": 2550
     },
     {
       "epoch": 2.462355212355212,
-      "grad_norm": 5.138809680938721,
+      "grad_norm": 5.040472984313965,
       "learning_rate": 1.6385516048150063e-06,
-      "loss": 0.7383,
+      "loss": 0.5352,
       "step": 2551
     },
     {
       "epoch": 2.4633204633204633,
-      "grad_norm": 5.04536247253418,
+      "grad_norm": 4.710428714752197,
       "learning_rate": 1.632838858914747e-06,
-      "loss": 0.7934,
+      "loss": 0.5716,
       "step": 2552
     },
     {
       "epoch": 2.4642857142857144,
-      "grad_norm": 5.438734531402588,
+      "grad_norm": 5.127421855926514,
       "learning_rate": 1.6271352035892552e-06,
-      "loss": 0.7686,
+      "loss": 0.5436,
       "step": 2553
     },
     {
       "epoch": 2.465250965250965,
-      "grad_norm": 4.890718460083008,
+      "grad_norm": 4.792416572570801,
       "learning_rate": 1.621440645035318e-06,
-      "loss": 0.7321,
+      "loss": 0.5258,
       "step": 2554
     },
     {
       "epoch": 2.4662162162162162,
-      "grad_norm": 5.157116413116455,
+      "grad_norm": 5.2342023849487305,
       "learning_rate": 1.6157551894398371e-06,
-      "loss": 0.7909,
+      "loss": 0.5835,
       "step": 2555
     },
     {
       "epoch": 2.4671814671814674,
-      "grad_norm": 4.768998146057129,
+      "grad_norm": 4.816235065460205,
       "learning_rate": 1.6100788429798242e-06,
-      "loss": 0.7467,
+      "loss": 0.5229,
       "step": 2556
     },
     {
       "epoch": 2.468146718146718,
-      "grad_norm": 5.029572486877441,
+      "grad_norm": 5.381363391876221,
       "learning_rate": 1.6044116118223973e-06,
-      "loss": 0.7762,
+      "loss": 0.596,
       "step": 2557
     },
     {
       "epoch": 2.469111969111969,
-      "grad_norm": 5.010144233703613,
+      "grad_norm": 4.885616779327393,
       "learning_rate": 1.5987535021247668e-06,
-      "loss": 0.77,
+      "loss": 0.546,
       "step": 2558
     },
     {
       "epoch": 2.4700772200772203,
-      "grad_norm": 4.539247989654541,
+      "grad_norm": 4.584182262420654,
       "learning_rate": 1.5931045200342355e-06,
-      "loss": 0.7277,
+      "loss": 0.5159,
       "step": 2559
     },
     {
       "epoch": 2.471042471042471,
-      "grad_norm": 4.728090286254883,
+      "grad_norm": 5.129901885986328,
       "learning_rate": 1.587464671688187e-06,
-      "loss": 0.7132,
+      "loss": 0.534,
       "step": 2560
     },
     {
       "epoch": 2.472007722007722,
-      "grad_norm": 4.758804798126221,
+      "grad_norm": 4.932203769683838,
       "learning_rate": 1.581833963214091e-06,
-      "loss": 0.7277,
+      "loss": 0.5149,
       "step": 2561
     },
     {
       "epoch": 2.472972972972973,
-      "grad_norm": 5.113906383514404,
+      "grad_norm": 5.150115013122559,
       "learning_rate": 1.5762124007294721e-06,
-      "loss": 0.7661,
+      "loss": 0.578,
       "step": 2562
     },
     {
       "epoch": 2.473938223938224,
-      "grad_norm": 5.134305953979492,
+      "grad_norm": 5.037455081939697,
       "learning_rate": 1.5705999903419267e-06,
-      "loss": 0.7827,
+      "loss": 0.5504,
       "step": 2563
     },
     {
       "epoch": 2.474903474903475,
-      "grad_norm": 5.465189456939697,
+      "grad_norm": 5.31726598739624,
       "learning_rate": 1.564996738149106e-06,
-      "loss": 0.7497,
+      "loss": 0.5449,
       "step": 2564
     },
     {
       "epoch": 2.4758687258687258,
-      "grad_norm": 5.321998119354248,
+      "grad_norm": 4.975057125091553,
       "learning_rate": 1.5594026502387184e-06,
-      "loss": 0.759,
+      "loss": 0.5472,
       "step": 2565
     },
     {
       "epoch": 2.476833976833977,
-      "grad_norm": 5.111255168914795,
+      "grad_norm": 4.776544094085693,
       "learning_rate": 1.553817732688503e-06,
-      "loss": 0.7343,
+      "loss": 0.5245,
       "step": 2566
     },
     {
       "epoch": 2.4777992277992276,
-      "grad_norm": 5.24220085144043,
+      "grad_norm": 4.868185520172119,
       "learning_rate": 1.5482419915662418e-06,
-      "loss": 0.7317,
+      "loss": 0.5241,
       "step": 2567
     },
     {
       "epoch": 2.4787644787644787,
-      "grad_norm": 4.93293571472168,
+      "grad_norm": 5.049710273742676,
       "learning_rate": 1.5426754329297477e-06,
-      "loss": 0.7427,
+      "loss": 0.5523,
       "step": 2568
     },
     {
       "epoch": 2.47972972972973,
-      "grad_norm": 4.531349182128906,
+      "grad_norm": 4.725720405578613,
       "learning_rate": 1.537118062826859e-06,
-      "loss": 0.7641,
+      "loss": 0.5591,
       "step": 2569
     },
     {
       "epoch": 2.4806949806949805,
-      "grad_norm": 4.671472549438477,
+      "grad_norm": 4.379458427429199,
       "learning_rate": 1.5315698872954299e-06,
-      "loss": 0.7524,
+      "loss": 0.5451,
       "step": 2570
     },
     {
       "epoch": 2.4816602316602316,
-      "grad_norm": 5.036346435546875,
+      "grad_norm": 5.342967510223389,
       "learning_rate": 1.5260309123633189e-06,
-      "loss": 0.7977,
+      "loss": 0.6081,
       "step": 2571
     },
     {
       "epoch": 2.4826254826254828,
-      "grad_norm": 5.153173446655273,
+      "grad_norm": 5.54488468170166,
       "learning_rate": 1.520501144048393e-06,
-      "loss": 0.7819,
+      "loss": 0.5776,
       "step": 2572
     },
     {
       "epoch": 2.4835907335907335,
-      "grad_norm": 4.5762858390808105,
+      "grad_norm": 4.853941917419434,
       "learning_rate": 1.5149805883585221e-06,
-      "loss": 0.6937,
+      "loss": 0.4967,
       "step": 2573
     },
     {
       "epoch": 2.4845559845559846,
-      "grad_norm": 4.898032188415527,
+      "grad_norm": 5.128244876861572,
       "learning_rate": 1.5094692512915621e-06,
-      "loss": 0.7118,
+      "loss": 0.5048,
       "step": 2574
     },
     {
       "epoch": 2.4855212355212357,
-      "grad_norm": 4.975152969360352,
+      "grad_norm": 4.976890563964844,
       "learning_rate": 1.5039671388353472e-06,
-      "loss": 0.7288,
+      "loss": 0.534,
       "step": 2575
     },
     {
       "epoch": 2.4864864864864864,
-      "grad_norm": 4.596181869506836,
+      "grad_norm": 4.490545272827148,
       "learning_rate": 1.4984742569676957e-06,
-      "loss": 0.7118,
+      "loss": 0.5121,
       "step": 2576
     },
     {
       "epoch": 2.4874517374517375,
-      "grad_norm": 5.141880512237549,
+      "grad_norm": 4.944030284881592,
       "learning_rate": 1.4929906116564008e-06,
-      "loss": 0.793,
+      "loss": 0.5681,
       "step": 2577
     },
     {
       "epoch": 2.488416988416988,
-      "grad_norm": 4.9118194580078125,
+      "grad_norm": 4.702122688293457,
       "learning_rate": 1.4875162088592155e-06,
-      "loss": 0.7374,
+      "loss": 0.5242,
       "step": 2578
     },
     {
       "epoch": 2.4893822393822393,
-      "grad_norm": 5.6777849197387695,
+      "grad_norm": 5.538225173950195,
       "learning_rate": 1.4820510545238542e-06,
-      "loss": 0.7466,
+      "loss": 0.5376,
       "step": 2579
     },
     {
       "epoch": 2.4903474903474905,
-      "grad_norm": 5.995486259460449,
+      "grad_norm": 4.954733371734619,
       "learning_rate": 1.4765951545879732e-06,
-      "loss": 0.7335,
+      "loss": 0.4935,
       "step": 2580
     },
     {
       "epoch": 2.491312741312741,
-      "grad_norm": 4.940185546875,
+      "grad_norm": 4.844915866851807,
       "learning_rate": 1.4711485149791894e-06,
-      "loss": 0.7946,
+      "loss": 0.5857,
       "step": 2581
     },
     {
       "epoch": 2.4922779922779923,
-      "grad_norm": 4.679786682128906,
+      "grad_norm": 4.477333068847656,
       "learning_rate": 1.4657111416150504e-06,
-      "loss": 0.7744,
+      "loss": 0.5499,
       "step": 2582
     },
     {
       "epoch": 2.4932432432432434,
-      "grad_norm": 5.902797222137451,
+      "grad_norm": 4.836726665496826,
       "learning_rate": 1.4602830404030387e-06,
-      "loss": 0.7534,
+      "loss": 0.5439,
       "step": 2583
     },
     {
       "epoch": 2.494208494208494,
-      "grad_norm": 5.293987274169922,
+      "grad_norm": 5.163521766662598,
       "learning_rate": 1.4548642172405569e-06,
-      "loss": 0.7632,
+      "loss": 0.5595,
       "step": 2584
     },
     {
       "epoch": 2.4951737451737452,
-      "grad_norm": 5.381636142730713,
+      "grad_norm": 5.177822589874268,
       "learning_rate": 1.4494546780149366e-06,
-      "loss": 0.7585,
+      "loss": 0.5657,
       "step": 2585
     },
     {
       "epoch": 2.4961389961389964,
-      "grad_norm": 5.651233673095703,
+      "grad_norm": 4.960970878601074,
       "learning_rate": 1.4440544286034186e-06,
-      "loss": 0.7293,
+      "loss": 0.5173,
       "step": 2586
     },
     {
       "epoch": 2.497104247104247,
-      "grad_norm": 5.446207046508789,
+      "grad_norm": 5.376863956451416,
       "learning_rate": 1.4386634748731499e-06,
-      "loss": 0.7867,
+      "loss": 0.5762,
       "step": 2587
     },
     {
       "epoch": 2.498069498069498,
-      "grad_norm": 5.752668857574463,
+      "grad_norm": 5.361690044403076,
       "learning_rate": 1.4332818226811807e-06,
-      "loss": 0.692,
+      "loss": 0.5109,
       "step": 2588
     },
     {
       "epoch": 2.499034749034749,
-      "grad_norm": 5.912139892578125,
+      "grad_norm": 5.333489418029785,
       "learning_rate": 1.4279094778744473e-06,
-      "loss": 0.8358,
+      "loss": 0.593,
       "step": 2589
     },
     {
       "epoch": 2.5,
-      "grad_norm": 5.182679176330566,
+      "grad_norm": 5.0431060791015625,
       "learning_rate": 1.4225464462897876e-06,
-      "loss": 0.7419,
+      "loss": 0.5214,
       "step": 2590
     },
     {
       "epoch": 2.500965250965251,
-      "grad_norm": 5.262983798980713,
+      "grad_norm": 4.874368667602539,
       "learning_rate": 1.4171927337539103e-06,
-      "loss": 0.76,
+      "loss": 0.55,
       "step": 2591
     },
     {
       "epoch": 2.501930501930502,
-      "grad_norm": 4.978514671325684,
+      "grad_norm": 4.766120910644531,
       "learning_rate": 1.4118483460834065e-06,
-      "loss": 0.7344,
+      "loss": 0.5369,
       "step": 2592
     },
     {
       "epoch": 2.502895752895753,
-      "grad_norm": 5.735084533691406,
+      "grad_norm": 5.093378067016602,
       "learning_rate": 1.4065132890847245e-06,
-      "loss": 0.7396,
+      "loss": 0.5145,
       "step": 2593
     },
     {
       "epoch": 2.5038610038610036,
-      "grad_norm": 5.626955986022949,
+      "grad_norm": 5.536632537841797,
       "learning_rate": 1.4011875685541898e-06,
-      "loss": 0.7295,
+      "loss": 0.5065,
       "step": 2594
     },
     {
       "epoch": 2.5048262548262548,
-      "grad_norm": 4.752853870391846,
+      "grad_norm": 4.396800518035889,
       "learning_rate": 1.3958711902779775e-06,
-      "loss": 0.6756,
+      "loss": 0.4691,
       "step": 2595
     },
     {
       "epoch": 2.505791505791506,
-      "grad_norm": 5.7477569580078125,
+      "grad_norm": 4.920122146606445,
       "learning_rate": 1.390564160032113e-06,
-      "loss": 0.7603,
+      "loss": 0.5331,
       "step": 2596
     },
     {
       "epoch": 2.506756756756757,
-      "grad_norm": 4.698903560638428,
+      "grad_norm": 4.640273094177246,
       "learning_rate": 1.3852664835824592e-06,
-      "loss": 0.7626,
+      "loss": 0.5553,
       "step": 2597
     },
     {
       "epoch": 2.5077220077220077,
-      "grad_norm": 5.006353378295898,
+      "grad_norm": 4.775614261627197,
       "learning_rate": 1.3799781666847301e-06,
-      "loss": 0.7236,
+      "loss": 0.5103,
       "step": 2598
     },
     {
       "epoch": 2.508687258687259,
-      "grad_norm": 6.56751823425293,
+      "grad_norm": 5.377407550811768,
       "learning_rate": 1.3746992150844595e-06,
-      "loss": 0.7565,
+      "loss": 0.5553,
       "step": 2599
     },
     {
       "epoch": 2.5096525096525095,
-      "grad_norm": 4.768529415130615,
+      "grad_norm": 4.803976535797119,
       "learning_rate": 1.369429634517012e-06,
-      "loss": 0.7437,
+      "loss": 0.5258,
       "step": 2600
     },
     {
       "epoch": 2.5106177606177607,
-      "grad_norm": 5.13110876083374,
+      "grad_norm": 4.5876359939575195,
       "learning_rate": 1.3641694307075682e-06,
-      "loss": 0.7305,
+      "loss": 0.4968,
       "step": 2601
     },
     {
       "epoch": 2.511583011583012,
-      "grad_norm": 4.623649597167969,
+      "grad_norm": 4.6867194175720215,
       "learning_rate": 1.3589186093711227e-06,
-      "loss": 0.7282,
+      "loss": 0.542,
       "step": 2602
     },
     {
       "epoch": 2.5125482625482625,
-      "grad_norm": 4.7761759757995605,
+      "grad_norm": 4.6872735023498535,
       "learning_rate": 1.3536771762124745e-06,
-      "loss": 0.7194,
+      "loss": 0.5069,
       "step": 2603
     },
     {
       "epoch": 2.5135135135135136,
-      "grad_norm": 5.18038272857666,
+      "grad_norm": 5.25548791885376,
       "learning_rate": 1.3484451369262251e-06,
-      "loss": 0.7949,
+      "loss": 0.5776,
       "step": 2604
     },
     {
       "epoch": 2.5144787644787643,
-      "grad_norm": 4.617712020874023,
+      "grad_norm": 4.618780136108398,
       "learning_rate": 1.34322249719677e-06,
-      "loss": 0.7398,
+      "loss": 0.5197,
       "step": 2605
     },
     {
       "epoch": 2.5154440154440154,
-      "grad_norm": 5.1880784034729,
+      "grad_norm": 5.138356685638428,
       "learning_rate": 1.3380092626982887e-06,
-      "loss": 0.6584,
+      "loss": 0.4896,
       "step": 2606
     },
     {
       "epoch": 2.5164092664092665,
-      "grad_norm": 5.24415922164917,
+      "grad_norm": 4.957482814788818,
       "learning_rate": 1.3328054390947475e-06,
-      "loss": 0.735,
+      "loss": 0.5488,
       "step": 2607
     },
     {
       "epoch": 2.5173745173745172,
-      "grad_norm": 4.836585521697998,
+      "grad_norm": 5.0967116355896,
       "learning_rate": 1.3276110320398849e-06,
-      "loss": 0.7223,
+      "loss": 0.5193,
       "step": 2608
     },
     {
       "epoch": 2.5183397683397684,
-      "grad_norm": 5.203280925750732,
+      "grad_norm": 5.083591938018799,
       "learning_rate": 1.3224260471772088e-06,
-      "loss": 0.7055,
+      "loss": 0.487,
       "step": 2609
     },
     {
       "epoch": 2.519305019305019,
-      "grad_norm": 5.076012134552002,
+      "grad_norm": 4.914400100708008,
       "learning_rate": 1.317250490139992e-06,
-      "loss": 0.745,
+      "loss": 0.5087,
       "step": 2610
     },
     {
       "epoch": 2.52027027027027,
-      "grad_norm": 5.236003398895264,
+      "grad_norm": 4.9238481521606445,
       "learning_rate": 1.3120843665512617e-06,
-      "loss": 0.7623,
+      "loss": 0.5417,
       "step": 2611
     },
     {
       "epoch": 2.5212355212355213,
-      "grad_norm": 5.240046977996826,
+      "grad_norm": 5.124131202697754,
       "learning_rate": 1.3069276820237997e-06,
-      "loss": 0.7261,
+      "loss": 0.5209,
       "step": 2612
     },
     {
       "epoch": 2.5222007722007724,
-      "grad_norm": 5.299560546875,
+      "grad_norm": 5.464865684509277,
       "learning_rate": 1.3017804421601298e-06,
-      "loss": 0.8015,
+      "loss": 0.585,
       "step": 2613
     },
     {
       "epoch": 2.523166023166023,
-      "grad_norm": 5.6360015869140625,
+      "grad_norm": 4.946651458740234,
       "learning_rate": 1.2966426525525144e-06,
-      "loss": 0.7365,
+      "loss": 0.5276,
       "step": 2614
     },
     {
       "epoch": 2.5241312741312742,
-      "grad_norm": 5.0282158851623535,
+      "grad_norm": 4.8147454261779785,
       "learning_rate": 1.2915143187829516e-06,
-      "loss": 0.728,
+      "loss": 0.495,
       "step": 2615
     },
     {
       "epoch": 2.525096525096525,
-      "grad_norm": 4.797794342041016,
+      "grad_norm": 4.769637584686279,
       "learning_rate": 1.286395446423162e-06,
-      "loss": 0.7516,
+      "loss": 0.5263,
       "step": 2616
     },
     {
       "epoch": 2.526061776061776,
-      "grad_norm": 4.75595235824585,
+      "grad_norm": 5.0410051345825195,
       "learning_rate": 1.2812860410345894e-06,
-      "loss": 0.7432,
+      "loss": 0.5383,
       "step": 2617
     },
     {
       "epoch": 2.527027027027027,
-      "grad_norm": 4.534605503082275,
+      "grad_norm": 4.58768367767334,
       "learning_rate": 1.2761861081683914e-06,
-      "loss": 0.8098,
+      "loss": 0.5882,
       "step": 2618
     },
     {
       "epoch": 2.527992277992278,
-      "grad_norm": 4.790217399597168,
+      "grad_norm": 5.219866752624512,
       "learning_rate": 1.2710956533654384e-06,
-      "loss": 0.7833,
+      "loss": 0.5591,
       "step": 2619
     },
     {
       "epoch": 2.528957528957529,
-      "grad_norm": 4.9109721183776855,
+      "grad_norm": 5.143247127532959,
       "learning_rate": 1.2660146821562936e-06,
-      "loss": 0.7315,
+      "loss": 0.5189,
       "step": 2620
     },
     {
       "epoch": 2.5299227799227797,
-      "grad_norm": 4.9708781242370605,
+      "grad_norm": 4.844534873962402,
       "learning_rate": 1.2609432000612232e-06,
-      "loss": 0.7869,
+      "loss": 0.5468,
       "step": 2621
     },
     {
       "epoch": 2.530888030888031,
-      "grad_norm": 5.2403082847595215,
+      "grad_norm": 5.229526042938232,
       "learning_rate": 1.255881212590183e-06,
-      "loss": 0.742,
+      "loss": 0.5375,
       "step": 2622
     },
     {
       "epoch": 2.531853281853282,
-      "grad_norm": 5.221992015838623,
+      "grad_norm": 5.4178853034973145,
       "learning_rate": 1.2508287252428175e-06,
-      "loss": 0.7807,
+      "loss": 0.5647,
       "step": 2623
     },
     {
       "epoch": 2.532818532818533,
-      "grad_norm": 4.320801734924316,
+      "grad_norm": 4.532337665557861,
       "learning_rate": 1.245785743508441e-06,
-      "loss": 0.6937,
+      "loss": 0.517,
       "step": 2624
     },
     {
       "epoch": 2.5337837837837838,
-      "grad_norm": 5.410978317260742,
+      "grad_norm": 5.332575798034668,
       "learning_rate": 1.2407522728660447e-06,
-      "loss": 0.7456,
+      "loss": 0.5544,
       "step": 2625
     },
     {
       "epoch": 2.534749034749035,
-      "grad_norm": 5.150344371795654,
+      "grad_norm": 5.031994819641113,
       "learning_rate": 1.2357283187842872e-06,
-      "loss": 0.7478,
+      "loss": 0.5512,
       "step": 2626
     },
     {
       "epoch": 2.5357142857142856,
-      "grad_norm": 5.0697174072265625,
+      "grad_norm": 4.665736198425293,
       "learning_rate": 1.2307138867214863e-06,
-      "loss": 0.7477,
+      "loss": 0.5399,
       "step": 2627
     },
     {
       "epoch": 2.5366795366795367,
-      "grad_norm": 5.012232780456543,
+      "grad_norm": 4.951563835144043,
       "learning_rate": 1.2257089821256196e-06,
-      "loss": 0.764,
+      "loss": 0.5285,
       "step": 2628
     },
     {
       "epoch": 2.537644787644788,
-      "grad_norm": 4.838174819946289,
+      "grad_norm": 4.878172397613525,
       "learning_rate": 1.220713610434303e-06,
-      "loss": 0.7387,
+      "loss": 0.5234,
       "step": 2629
     },
     {
       "epoch": 2.5386100386100385,
-      "grad_norm": 5.2238311767578125,
+      "grad_norm": 4.664141654968262,
       "learning_rate": 1.2157277770748044e-06,
-      "loss": 0.7306,
+      "loss": 0.5228,
       "step": 2630
     },
     {
       "epoch": 2.5395752895752897,
-      "grad_norm": 4.850396633148193,
+      "grad_norm": 4.9251532554626465,
       "learning_rate": 1.2107514874640235e-06,
-      "loss": 0.7291,
+      "loss": 0.5118,
       "step": 2631
     },
     {
       "epoch": 2.5405405405405403,
-      "grad_norm": 4.7762322425842285,
+      "grad_norm": 4.885021209716797,
       "learning_rate": 1.2057847470084993e-06,
-      "loss": 0.6686,
+      "loss": 0.4737,
       "step": 2632
     },
     {
       "epoch": 2.5415057915057915,
-      "grad_norm": 5.246326446533203,
+      "grad_norm": 5.117430686950684,
       "learning_rate": 1.200827561104383e-06,
-      "loss": 0.7399,
+      "loss": 0.5478,
       "step": 2633
     },
     {
       "epoch": 2.5424710424710426,
-      "grad_norm": 5.028595924377441,
+      "grad_norm": 4.5831732749938965,
       "learning_rate": 1.195879935137455e-06,
-      "loss": 0.7027,
+      "loss": 0.4841,
       "step": 2634
     },
     {
       "epoch": 2.5434362934362933,
-      "grad_norm": 4.831367492675781,
+      "grad_norm": 4.682000160217285,
       "learning_rate": 1.1909418744831048e-06,
-      "loss": 0.7693,
+      "loss": 0.5426,
       "step": 2635
     },
     {
       "epoch": 2.5444015444015444,
-      "grad_norm": 5.539283275604248,
+      "grad_norm": 4.87588357925415,
       "learning_rate": 1.186013384506335e-06,
-      "loss": 0.7085,
+      "loss": 0.4986,
       "step": 2636
     },
     {
       "epoch": 2.545366795366795,
-      "grad_norm": 5.044233798980713,
+      "grad_norm": 4.957118988037109,
       "learning_rate": 1.1810944705617466e-06,
-      "loss": 0.7303,
+      "loss": 0.5369,
       "step": 2637
     },
     {
       "epoch": 2.5463320463320462,
-      "grad_norm": 4.916104316711426,
+      "grad_norm": 4.897390842437744,
       "learning_rate": 1.1761851379935307e-06,
-      "loss": 0.7268,
+      "loss": 0.5258,
       "step": 2638
     },
     {
       "epoch": 2.5472972972972974,
-      "grad_norm": 5.304819583892822,
+      "grad_norm": 5.071802139282227,
       "learning_rate": 1.171285392135476e-06,
-      "loss": 0.8003,
+      "loss": 0.5668,
       "step": 2639
     },
     {
       "epoch": 2.5482625482625485,
-      "grad_norm": 4.751123428344727,
+      "grad_norm": 4.867398262023926,
       "learning_rate": 1.1663952383109566e-06,
-      "loss": 0.7035,
+      "loss": 0.4985,
       "step": 2640
     },
     {
       "epoch": 2.549227799227799,
-      "grad_norm": 4.719592094421387,
+      "grad_norm": 4.763899326324463,
       "learning_rate": 1.1615146818329237e-06,
-      "loss": 0.738,
+      "loss": 0.5096,
       "step": 2641
     },
     {
       "epoch": 2.5501930501930503,
-      "grad_norm": 4.5217180252075195,
+      "grad_norm": 4.5873517990112305,
       "learning_rate": 1.1566437280038956e-06,
-      "loss": 0.7048,
+      "loss": 0.4995,
       "step": 2642
     },
     {
       "epoch": 2.551158301158301,
-      "grad_norm": 4.832311630249023,
+      "grad_norm": 4.499822616577148,
       "learning_rate": 1.1517823821159624e-06,
-      "loss": 0.7116,
+      "loss": 0.4775,
       "step": 2643
     },
     {
       "epoch": 2.552123552123552,
-      "grad_norm": 5.188864231109619,
+      "grad_norm": 5.088689804077148,
       "learning_rate": 1.1469306494507792e-06,
-      "loss": 0.6957,
+      "loss": 0.4986,
       "step": 2644
     },
     {
       "epoch": 2.5530888030888033,
-      "grad_norm": 4.863763809204102,
+      "grad_norm": 4.933341979980469,
       "learning_rate": 1.1420885352795519e-06,
-      "loss": 0.7794,
+      "loss": 0.5556,
       "step": 2645
     },
     {
       "epoch": 2.554054054054054,
-      "grad_norm": 4.8022003173828125,
+      "grad_norm": 4.7844133377075195,
       "learning_rate": 1.1372560448630377e-06,
-      "loss": 0.7648,
+      "loss": 0.5543,
       "step": 2646
     },
     {
       "epoch": 2.555019305019305,
-      "grad_norm": 4.957478046417236,
+      "grad_norm": 4.794405937194824,
       "learning_rate": 1.1324331834515335e-06,
-      "loss": 0.7656,
+      "loss": 0.553,
       "step": 2647
     },
     {
       "epoch": 2.5559845559845558,
-      "grad_norm": 4.873451232910156,
+      "grad_norm": 4.613048076629639,
       "learning_rate": 1.1276199562848845e-06,
-      "loss": 0.752,
+      "loss": 0.5269,
       "step": 2648
     },
     {
       "epoch": 2.556949806949807,
-      "grad_norm": 5.428621768951416,
+      "grad_norm": 5.2739739418029785,
       "learning_rate": 1.1228163685924599e-06,
-      "loss": 0.7545,
+      "loss": 0.5291,
       "step": 2649
     },
     {
       "epoch": 2.557915057915058,
-      "grad_norm": 4.873611927032471,
+      "grad_norm": 4.597942352294922,
       "learning_rate": 1.1180224255931626e-06,
-      "loss": 0.7147,
+      "loss": 0.5136,
       "step": 2650
     },
     {
       "epoch": 2.558880308880309,
-      "grad_norm": 4.914524555206299,
+      "grad_norm": 4.725526332855225,
       "learning_rate": 1.1132381324954067e-06,
-      "loss": 0.7488,
+      "loss": 0.5282,
       "step": 2651
     },
     {
       "epoch": 2.55984555984556,
-      "grad_norm": 4.954052448272705,
+      "grad_norm": 4.896808624267578,
       "learning_rate": 1.108463494497135e-06,
-      "loss": 0.79,
+      "loss": 0.5483,
       "step": 2652
     },
     {
       "epoch": 2.560810810810811,
-      "grad_norm": 4.874828815460205,
+      "grad_norm": 4.725536823272705,
       "learning_rate": 1.1036985167857938e-06,
-      "loss": 0.7259,
+      "loss": 0.5166,
       "step": 2653
     },
     {
       "epoch": 2.5617760617760617,
-      "grad_norm": 5.06655740737915,
+      "grad_norm": 4.951056957244873,
       "learning_rate": 1.098943204538332e-06,
-      "loss": 0.7827,
+      "loss": 0.5575,
       "step": 2654
     },
     {
       "epoch": 2.562741312741313,
-      "grad_norm": 4.803916931152344,
+      "grad_norm": 4.902041435241699,
       "learning_rate": 1.094197562921202e-06,
-      "loss": 0.7496,
+      "loss": 0.5393,
       "step": 2655
     },
     {
       "epoch": 2.563706563706564,
-      "grad_norm": 5.09578800201416,
+      "grad_norm": 5.162410736083984,
       "learning_rate": 1.0894615970903454e-06,
-      "loss": 0.7658,
+      "loss": 0.5653,
       "step": 2656
     },
     {
       "epoch": 2.5646718146718146,
-      "grad_norm": 4.7430219650268555,
+      "grad_norm": 4.854796886444092,
       "learning_rate": 1.0847353121911952e-06,
-      "loss": 0.7573,
+      "loss": 0.5554,
       "step": 2657
     },
     {
       "epoch": 2.5656370656370657,
-      "grad_norm": 5.272439479827881,
+      "grad_norm": 5.427006244659424,
       "learning_rate": 1.0800187133586626e-06,
-      "loss": 0.7421,
+      "loss": 0.5228,
       "step": 2658
     },
     {
       "epoch": 2.5666023166023164,
-      "grad_norm": 4.683495998382568,
+      "grad_norm": 4.887862205505371,
       "learning_rate": 1.0753118057171396e-06,
-      "loss": 0.7881,
+      "loss": 0.6028,
       "step": 2659
     },
     {
       "epoch": 2.5675675675675675,
-      "grad_norm": 5.072673797607422,
+      "grad_norm": 5.142760276794434,
       "learning_rate": 1.0706145943804846e-06,
-      "loss": 0.7413,
+      "loss": 0.5343,
       "step": 2660
     },
     {
       "epoch": 2.5685328185328187,
-      "grad_norm": 4.405806541442871,
+      "grad_norm": 4.502059459686279,
       "learning_rate": 1.0659270844520253e-06,
-      "loss": 0.7472,
+      "loss": 0.5427,
       "step": 2661
     },
     {
       "epoch": 2.5694980694980694,
-      "grad_norm": 4.982200622558594,
+      "grad_norm": 4.744590759277344,
       "learning_rate": 1.0612492810245466e-06,
-      "loss": 0.7096,
+      "loss": 0.4981,
       "step": 2662
     },
     {
       "epoch": 2.5704633204633205,
-      "grad_norm": 4.694039821624756,
+      "grad_norm": 4.498823642730713,
       "learning_rate": 1.056581189180288e-06,
-      "loss": 0.7413,
+      "loss": 0.5092,
       "step": 2663
     },
     {
       "epoch": 2.571428571428571,
-      "grad_norm": 4.818814754486084,
+      "grad_norm": 4.577084064483643,
       "learning_rate": 1.05192281399094e-06,
-      "loss": 0.7524,
+      "loss": 0.5331,
       "step": 2664
     },
     {
       "epoch": 2.5723938223938223,
-      "grad_norm": 4.986637592315674,
+      "grad_norm": 4.900193691253662,
       "learning_rate": 1.0472741605176339e-06,
-      "loss": 0.788,
+      "loss": 0.5681,
       "step": 2665
     },
     {
       "epoch": 2.5733590733590734,
-      "grad_norm": 4.786144733428955,
+      "grad_norm": 4.39017915725708,
       "learning_rate": 1.0426352338109392e-06,
-      "loss": 0.7132,
+      "loss": 0.4973,
       "step": 2666
     },
     {
       "epoch": 2.5743243243243246,
-      "grad_norm": 5.189314365386963,
+      "grad_norm": 4.927492141723633,
       "learning_rate": 1.038006038910858e-06,
-      "loss": 0.7619,
+      "loss": 0.5471,
       "step": 2667
     },
     {
       "epoch": 2.5752895752895753,
-      "grad_norm": 4.994247913360596,
+      "grad_norm": 4.808524131774902,
       "learning_rate": 1.0333865808468203e-06,
-      "loss": 0.7311,
+      "loss": 0.525,
       "step": 2668
     },
     {
       "epoch": 2.5762548262548264,
-      "grad_norm": 5.909048557281494,
+      "grad_norm": 4.852983474731445,
       "learning_rate": 1.0287768646376762e-06,
-      "loss": 0.7497,
+      "loss": 0.529,
       "step": 2669
     },
     {
       "epoch": 2.577220077220077,
-      "grad_norm": 4.981646537780762,
+      "grad_norm": 5.177182197570801,
       "learning_rate": 1.0241768952916898e-06,
-      "loss": 0.778,
+      "loss": 0.5766,
       "step": 2670
     },
     {
       "epoch": 2.578185328185328,
-      "grad_norm": 5.011918067932129,
+      "grad_norm": 4.912561416625977,
       "learning_rate": 1.01958667780654e-06,
-      "loss": 0.7496,
+      "loss": 0.5447,
       "step": 2671
     },
     {
       "epoch": 2.5791505791505793,
-      "grad_norm": 4.762938976287842,
+      "grad_norm": 4.673938751220703,
       "learning_rate": 1.0150062171693075e-06,
-      "loss": 0.7419,
+      "loss": 0.5195,
       "step": 2672
     },
     {
       "epoch": 2.58011583011583,
-      "grad_norm": 5.066151142120361,
+      "grad_norm": 5.0424981117248535,
       "learning_rate": 1.010435518356474e-06,
-      "loss": 0.736,
+      "loss": 0.5244,
       "step": 2673
     },
     {
       "epoch": 2.581081081081081,
-      "grad_norm": 5.22730827331543,
+      "grad_norm": 4.975863933563232,
       "learning_rate": 1.0058745863339136e-06,
-      "loss": 0.7731,
+      "loss": 0.577,
       "step": 2674
     },
     {
       "epoch": 2.582046332046332,
-      "grad_norm": 5.501034736633301,
+      "grad_norm": 4.993093013763428,
       "learning_rate": 1.0013234260568915e-06,
-      "loss": 0.7585,
+      "loss": 0.5319,
       "step": 2675
     },
     {
       "epoch": 2.583011583011583,
-      "grad_norm": 5.197367191314697,
+      "grad_norm": 5.188309192657471,
       "learning_rate": 9.967820424700548e-07,
-      "loss": 0.7524,
+      "loss": 0.535,
       "step": 2676
     },
     {
       "epoch": 2.583976833976834,
-      "grad_norm": 5.409716606140137,
+      "grad_norm": 5.098466396331787,
       "learning_rate": 9.922504405074295e-07,
-      "loss": 0.7613,
+      "loss": 0.5507,
       "step": 2677
     },
     {
       "epoch": 2.584942084942085,
-      "grad_norm": 5.282143592834473,
+      "grad_norm": 5.406103610992432,
       "learning_rate": 9.877286250924145e-07,
-      "loss": 0.7369,
+      "loss": 0.5378,
       "step": 2678
     },
     {
       "epoch": 2.585907335907336,
-      "grad_norm": 5.7618842124938965,
+      "grad_norm": 5.13191556930542,
       "learning_rate": 9.83216601137773e-07,
-      "loss": 0.7047,
+      "loss": 0.4863,
       "step": 2679
     },
     {
       "epoch": 2.586872586872587,
-      "grad_norm": 4.768263339996338,
+      "grad_norm": 4.861323833465576,
       "learning_rate": 9.787143735456351e-07,
-      "loss": 0.8062,
+      "loss": 0.5926,
       "step": 2680
     },
     {
       "epoch": 2.5878378378378377,
-      "grad_norm": 4.665050029754639,
+      "grad_norm": 4.965173721313477,
       "learning_rate": 9.742219472074833e-07,
-      "loss": 0.7731,
+      "loss": 0.5582,
       "step": 2681
     },
     {
       "epoch": 2.588803088803089,
-      "grad_norm": 4.817323207855225,
+      "grad_norm": 5.006649494171143,
       "learning_rate": 9.697393270041533e-07,
-      "loss": 0.7083,
+      "loss": 0.5101,
       "step": 2682
     },
     {
       "epoch": 2.58976833976834,
-      "grad_norm": 4.751098155975342,
+      "grad_norm": 4.78377628326416,
       "learning_rate": 9.652665178058285e-07,
-      "loss": 0.8147,
+      "loss": 0.5886,
       "step": 2683
     },
     {
       "epoch": 2.5907335907335907,
-      "grad_norm": 4.651120662689209,
+      "grad_norm": 4.934759616851807,
       "learning_rate": 9.608035244720293e-07,
-      "loss": 0.7698,
+      "loss": 0.5637,
       "step": 2684
     },
     {
       "epoch": 2.591698841698842,
-      "grad_norm": 5.18508768081665,
+      "grad_norm": 5.294219017028809,
       "learning_rate": 9.563503518516114e-07,
-      "loss": 0.7625,
+      "loss": 0.54,
       "step": 2685
     },
     {
       "epoch": 2.5926640926640925,
-      "grad_norm": 4.685789108276367,
+      "grad_norm": 4.688608646392822,
       "learning_rate": 9.519070047827705e-07,
-      "loss": 0.7353,
+      "loss": 0.5301,
       "step": 2686
     },
     {
       "epoch": 2.5936293436293436,
-      "grad_norm": 4.945529937744141,
+      "grad_norm": 5.03803014755249,
       "learning_rate": 9.474734880930114e-07,
-      "loss": 0.7487,
+      "loss": 0.5124,
       "step": 2687
     },
     {
       "epoch": 2.5945945945945947,
-      "grad_norm": 4.744202136993408,
+      "grad_norm": 4.730742454528809,
       "learning_rate": 9.430498065991712e-07,
-      "loss": 0.7357,
+      "loss": 0.5321,
       "step": 2688
     },
     {
       "epoch": 2.5955598455598454,
-      "grad_norm": 5.035425186157227,
+      "grad_norm": 5.0808844566345215,
       "learning_rate": 9.386359651073929e-07,
-      "loss": 0.7739,
+      "loss": 0.5583,
       "step": 2689
     },
     {
       "epoch": 2.5965250965250966,
-      "grad_norm": 4.681142330169678,
+      "grad_norm": 4.648793697357178,
       "learning_rate": 9.342319684131396e-07,
-      "loss": 0.7957,
+      "loss": 0.5714,
       "step": 2690
     },
     {
       "epoch": 2.5974903474903472,
-      "grad_norm": 5.089695930480957,
+      "grad_norm": 5.011692523956299,
       "learning_rate": 9.298378213011661e-07,
-      "loss": 0.717,
+      "loss": 0.5098,
       "step": 2691
     },
     {
       "epoch": 2.5984555984555984,
-      "grad_norm": 4.988779544830322,
+      "grad_norm": 4.8456549644470215,
       "learning_rate": 9.254535285455335e-07,
-      "loss": 0.6986,
+      "loss": 0.4814,
       "step": 2692
     },
     {
       "epoch": 2.5994208494208495,
-      "grad_norm": 4.7568254470825195,
+      "grad_norm": 4.961777687072754,
       "learning_rate": 9.210790949095916e-07,
-      "loss": 0.7577,
+      "loss": 0.5469,
       "step": 2693
     },
     {
       "epoch": 2.6003861003861006,
-      "grad_norm": 4.567595481872559,
+      "grad_norm": 5.076168060302734,
       "learning_rate": 9.167145251459875e-07,
-      "loss": 0.7213,
+      "loss": 0.528,
       "step": 2694
     },
     {
       "epoch": 2.6013513513513513,
-      "grad_norm": 4.910531044006348,
+      "grad_norm": 4.8708038330078125,
       "learning_rate": 9.123598239966447e-07,
-      "loss": 0.7957,
+      "loss": 0.5814,
       "step": 2695
     },
     {
       "epoch": 2.6023166023166024,
-      "grad_norm": 5.033100605010986,
+      "grad_norm": 5.015824317932129,
       "learning_rate": 9.08014996192762e-07,
-      "loss": 0.7237,
+      "loss": 0.518,
       "step": 2696
     },
     {
       "epoch": 2.603281853281853,
-      "grad_norm": 4.96120023727417,
+      "grad_norm": 5.016436576843262,
       "learning_rate": 9.036800464548157e-07,
-      "loss": 0.6612,
+      "loss": 0.4676,
       "step": 2697
     },
     {
       "epoch": 2.6042471042471043,
-      "grad_norm": 4.651401519775391,
+      "grad_norm": 4.84147310256958,
       "learning_rate": 8.993549794925538e-07,
-      "loss": 0.7619,
+      "loss": 0.5548,
       "step": 2698
     },
     {
       "epoch": 2.6052123552123554,
-      "grad_norm": 5.121455192565918,
+      "grad_norm": 4.757695198059082,
       "learning_rate": 8.95039800004982e-07,
-      "loss": 0.7651,
+      "loss": 0.527,
       "step": 2699
     },
     {
       "epoch": 2.606177606177606,
-      "grad_norm": 4.698480606079102,
+      "grad_norm": 4.609194755554199,
       "learning_rate": 8.907345126803602e-07,
-      "loss": 0.7207,
+      "loss": 0.4956,
       "step": 2700
     },
     {
       "epoch": 2.607142857142857,
-      "grad_norm": 4.910830020904541,
+      "grad_norm": 4.868758678436279,
       "learning_rate": 8.864391221962065e-07,
-      "loss": 0.7565,
+      "loss": 0.5491,
       "step": 2701
     },
     {
       "epoch": 2.608108108108108,
-      "grad_norm": 4.684881687164307,
+      "grad_norm": 4.633437633514404,
       "learning_rate": 8.821536332192882e-07,
-      "loss": 0.7329,
+      "loss": 0.524,
       "step": 2702
     },
     {
       "epoch": 2.609073359073359,
-      "grad_norm": 4.938525676727295,
+      "grad_norm": 4.7568039894104,
       "learning_rate": 8.778780504056106e-07,
-      "loss": 0.7575,
+      "loss": 0.5006,
       "step": 2703
     },
     {
       "epoch": 2.61003861003861,
-      "grad_norm": 5.000374794006348,
+      "grad_norm": 4.724079132080078,
       "learning_rate": 8.73612378400418e-07,
-      "loss": 0.7125,
+      "loss": 0.4897,
       "step": 2704
     },
     {
       "epoch": 2.611003861003861,
-      "grad_norm": 5.096011638641357,
+      "grad_norm": 5.096314430236816,
       "learning_rate": 8.693566218381854e-07,
-      "loss": 0.7542,
+      "loss": 0.5224,
       "step": 2705
     },
     {
       "epoch": 2.611969111969112,
-      "grad_norm": 4.76540470123291,
+      "grad_norm": 4.7161641120910645,
       "learning_rate": 8.651107853426144e-07,
-      "loss": 0.7275,
+      "loss": 0.5207,
       "step": 2706
     },
     {
       "epoch": 2.612934362934363,
-      "grad_norm": 5.00642204284668,
+      "grad_norm": 4.984276294708252,
       "learning_rate": 8.608748735266359e-07,
-      "loss": 0.7828,
+      "loss": 0.5598,
       "step": 2707
     },
     {
       "epoch": 2.613899613899614,
-      "grad_norm": 5.640402793884277,
+      "grad_norm": 5.071376800537109,
       "learning_rate": 8.56648890992392e-07,
-      "loss": 0.7276,
+      "loss": 0.5031,
       "step": 2708
     },
     {
       "epoch": 2.614864864864865,
-      "grad_norm": 4.735208034515381,
+      "grad_norm": 4.666943073272705,
       "learning_rate": 8.524328423312356e-07,
-      "loss": 0.7951,
+      "loss": 0.5527,
       "step": 2709
     },
     {
       "epoch": 2.615830115830116,
-      "grad_norm": 5.167140960693359,
+      "grad_norm": 4.890416622161865,
       "learning_rate": 8.482267321237281e-07,
-      "loss": 0.8058,
+      "loss": 0.5635,
       "step": 2710
     },
     {
       "epoch": 2.6167953667953667,
-      "grad_norm": 4.77217960357666,
+      "grad_norm": 4.743790626525879,
       "learning_rate": 8.440305649396385e-07,
-      "loss": 0.7804,
+      "loss": 0.5589,
       "step": 2711
     },
     {
       "epoch": 2.617760617760618,
-      "grad_norm": 4.561986446380615,
+      "grad_norm": 4.881480693817139,
       "learning_rate": 8.398443453379268e-07,
-      "loss": 0.7066,
+      "loss": 0.5105,
       "step": 2712
     },
     {
       "epoch": 2.6187258687258685,
-      "grad_norm": 4.737310886383057,
+      "grad_norm": 4.7660017013549805,
       "learning_rate": 8.356680778667503e-07,
-      "loss": 0.7665,
+      "loss": 0.5434,
       "step": 2713
     },
     {
       "epoch": 2.6196911196911197,
-      "grad_norm": 4.607104778289795,
+      "grad_norm": 4.853164196014404,
       "learning_rate": 8.31501767063445e-07,
-      "loss": 0.7518,
+      "loss": 0.5483,
       "step": 2714
     },
     {
       "epoch": 2.620656370656371,
-      "grad_norm": 4.817365646362305,
+      "grad_norm": 4.690718650817871,
       "learning_rate": 8.273454174545415e-07,
-      "loss": 0.7109,
+      "loss": 0.499,
       "step": 2715
     },
     {
       "epoch": 2.6216216216216215,
-      "grad_norm": 4.927873611450195,
+      "grad_norm": 4.9637250900268555,
       "learning_rate": 8.231990335557383e-07,
-      "loss": 0.7881,
+      "loss": 0.5388,
       "step": 2716
     },
     {
       "epoch": 2.6225868725868726,
-      "grad_norm": 4.71023416519165,
+      "grad_norm": 4.850188732147217,
       "learning_rate": 8.190626198719143e-07,
-      "loss": 0.7334,
+      "loss": 0.5263,
       "step": 2717
     },
     {
       "epoch": 2.6235521235521233,
-      "grad_norm": 4.634425163269043,
+      "grad_norm": 4.883005619049072,
       "learning_rate": 8.149361808971045e-07,
-      "loss": 0.7472,
+      "loss": 0.5341,
       "step": 2718
     },
     {
       "epoch": 2.6245173745173744,
-      "grad_norm": 4.826690196990967,
+      "grad_norm": 4.711854457855225,
       "learning_rate": 8.108197211145219e-07,
-      "loss": 0.7295,
+      "loss": 0.5241,
       "step": 2719
     },
     {
       "epoch": 2.6254826254826256,
-      "grad_norm": 4.9166483879089355,
+      "grad_norm": 4.918217658996582,
       "learning_rate": 8.067132449965254e-07,
-      "loss": 0.7745,
+      "loss": 0.5449,
       "step": 2720
     },
     {
       "epoch": 2.6264478764478767,
-      "grad_norm": 4.875312805175781,
+      "grad_norm": 4.926039695739746,
       "learning_rate": 8.02616757004634e-07,
-      "loss": 0.7593,
+      "loss": 0.5474,
       "step": 2721
     },
     {
       "epoch": 2.6274131274131274,
-      "grad_norm": 5.300418853759766,
+      "grad_norm": 4.7104692459106445,
       "learning_rate": 7.985302615895119e-07,
-      "loss": 0.7508,
+      "loss": 0.518,
       "step": 2722
     },
     {
       "epoch": 2.6283783783783785,
-      "grad_norm": 5.059196949005127,
+      "grad_norm": 4.662424087524414,
       "learning_rate": 7.944537631909666e-07,
-      "loss": 0.7335,
+      "loss": 0.5005,
       "step": 2723
     },
     {
       "epoch": 2.629343629343629,
-      "grad_norm": 5.13384485244751,
+      "grad_norm": 5.005817413330078,
       "learning_rate": 7.903872662379453e-07,
-      "loss": 0.7502,
+      "loss": 0.5208,
       "step": 2724
     },
     {
       "epoch": 2.6303088803088803,
-      "grad_norm": 7.005805492401123,
+      "grad_norm": 5.4079508781433105,
       "learning_rate": 7.863307751485294e-07,
-      "loss": 0.7461,
+      "loss": 0.5265,
       "step": 2725
     },
     {
       "epoch": 2.6312741312741315,
-      "grad_norm": 4.686102867126465,
+      "grad_norm": 4.554139137268066,
       "learning_rate": 7.822842943299292e-07,
-      "loss": 0.6944,
+      "loss": 0.4799,
       "step": 2726
     },
     {
       "epoch": 2.632239382239382,
-      "grad_norm": 4.866771697998047,
+      "grad_norm": 4.755030155181885,
       "learning_rate": 7.782478281784789e-07,
-      "loss": 0.7448,
+      "loss": 0.5265,
       "step": 2727
     },
     {
       "epoch": 2.6332046332046333,
-      "grad_norm": 5.382166385650635,
+      "grad_norm": 4.804332256317139,
       "learning_rate": 7.742213810796306e-07,
-      "loss": 0.7336,
+      "loss": 0.5272,
       "step": 2728
     },
     {
       "epoch": 2.634169884169884,
-      "grad_norm": 4.498585224151611,
+      "grad_norm": 4.2574028968811035,
       "learning_rate": 7.70204957407954e-07,
-      "loss": 0.7431,
+      "loss": 0.5141,
       "step": 2729
     },
     {
       "epoch": 2.635135135135135,
-      "grad_norm": 4.91465950012207,
+      "grad_norm": 4.7146711349487305,
       "learning_rate": 7.661985615271283e-07,
-      "loss": 0.7126,
+      "loss": 0.5029,
       "step": 2730
     },
     {
       "epoch": 2.636100386100386,
-      "grad_norm": 4.96571683883667,
+      "grad_norm": 4.7528276443481445,
       "learning_rate": 7.62202197789934e-07,
-      "loss": 0.7496,
+      "loss": 0.524,
       "step": 2731
     },
     {
       "epoch": 2.637065637065637,
-      "grad_norm": 4.990151405334473,
+      "grad_norm": 4.855458736419678,
       "learning_rate": 7.582158705382581e-07,
-      "loss": 0.7053,
+      "loss": 0.4719,
       "step": 2732
     },
     {
       "epoch": 2.638030888030888,
-      "grad_norm": 5.347865581512451,
+      "grad_norm": 5.035953044891357,
       "learning_rate": 7.542395841030792e-07,
-      "loss": 0.7827,
+      "loss": 0.5443,
       "step": 2733
     },
     {
       "epoch": 2.638996138996139,
-      "grad_norm": 5.1041646003723145,
+      "grad_norm": 4.722480773925781,
       "learning_rate": 7.502733428044684e-07,
-      "loss": 0.7695,
+      "loss": 0.5342,
       "step": 2734
     },
     {
       "epoch": 2.63996138996139,
-      "grad_norm": 4.740970611572266,
+      "grad_norm": 4.642751693725586,
       "learning_rate": 7.463171509515832e-07,
-      "loss": 0.7504,
+      "loss": 0.5263,
       "step": 2735
     },
     {
       "epoch": 2.640926640926641,
-      "grad_norm": 5.272965908050537,
+      "grad_norm": 5.1721625328063965,
       "learning_rate": 7.423710128426608e-07,
-      "loss": 0.7869,
+      "loss": 0.5367,
       "step": 2736
     },
     {
       "epoch": 2.641891891891892,
-      "grad_norm": 5.517708778381348,
+      "grad_norm": 5.225581645965576,
       "learning_rate": 7.384349327650198e-07,
-      "loss": 0.7672,
+      "loss": 0.5389,
       "step": 2737
     },
     {
       "epoch": 2.642857142857143,
-      "grad_norm": 4.674668312072754,
+      "grad_norm": 4.7233567237854,
       "learning_rate": 7.345089149950468e-07,
-      "loss": 0.677,
+      "loss": 0.4775,
       "step": 2738
     },
     {
       "epoch": 2.643822393822394,
-      "grad_norm": 5.323797225952148,
+      "grad_norm": 5.208678245544434,
       "learning_rate": 7.305929637981968e-07,
-      "loss": 0.7509,
+      "loss": 0.5385,
       "step": 2739
     },
     {
       "epoch": 2.6447876447876446,
-      "grad_norm": 4.841979026794434,
+      "grad_norm": 4.656902313232422,
       "learning_rate": 7.26687083428993e-07,
-      "loss": 0.7427,
+      "loss": 0.5195,
       "step": 2740
     },
     {
       "epoch": 2.6457528957528957,
-      "grad_norm": 4.6486639976501465,
+      "grad_norm": 4.821692943572998,
       "learning_rate": 7.227912781310099e-07,
-      "loss": 0.7341,
+      "loss": 0.5183,
       "step": 2741
     },
     {
       "epoch": 2.646718146718147,
-      "grad_norm": 5.009174823760986,
+      "grad_norm": 5.12415885925293,
       "learning_rate": 7.189055521368793e-07,
-      "loss": 0.7128,
+      "loss": 0.4966,
       "step": 2742
     },
     {
       "epoch": 2.6476833976833976,
-      "grad_norm": 4.864299297332764,
+      "grad_norm": 4.975660800933838,
       "learning_rate": 7.150299096682812e-07,
-      "loss": 0.6911,
+      "loss": 0.4755,
       "step": 2743
     },
     {
       "epoch": 2.6486486486486487,
-      "grad_norm": 4.875516414642334,
+      "grad_norm": 4.825127124786377,
       "learning_rate": 7.111643549359426e-07,
-      "loss": 0.7303,
+      "loss": 0.4924,
       "step": 2744
     },
     {
       "epoch": 2.6496138996138994,
-      "grad_norm": 4.866146564483643,
+      "grad_norm": 4.76726770401001,
       "learning_rate": 7.073088921396287e-07,
-      "loss": 0.7547,
+      "loss": 0.5219,
       "step": 2745
     },
     {
       "epoch": 2.6505791505791505,
-      "grad_norm": 5.089334011077881,
+      "grad_norm": 4.654300689697266,
       "learning_rate": 7.034635254681388e-07,
-      "loss": 0.7418,
+      "loss": 0.4997,
       "step": 2746
     },
     {
       "epoch": 2.6515444015444016,
-      "grad_norm": 4.971090793609619,
+      "grad_norm": 5.472558498382568,
       "learning_rate": 6.996282590993064e-07,
-      "loss": 0.7802,
+      "loss": 0.5557,
       "step": 2747
     },
     {
       "epoch": 2.6525096525096528,
-      "grad_norm": 4.696529865264893,
+      "grad_norm": 4.755984306335449,
       "learning_rate": 6.958030971999885e-07,
-      "loss": 0.7193,
+      "loss": 0.4935,
       "step": 2748
     },
     {
       "epoch": 2.6534749034749034,
-      "grad_norm": 4.897533416748047,
+      "grad_norm": 4.6794915199279785,
       "learning_rate": 6.919880439260685e-07,
-      "loss": 0.7503,
+      "loss": 0.5289,
       "step": 2749
     },
     {
       "epoch": 2.6544401544401546,
-      "grad_norm": 4.853023052215576,
+      "grad_norm": 4.992674350738525,
       "learning_rate": 6.881831034224417e-07,
-      "loss": 0.7827,
+      "loss": 0.5744,
       "step": 2750
     },
     {
       "epoch": 2.6554054054054053,
-      "grad_norm": 5.431102275848389,
+      "grad_norm": 5.203206539154053,
       "learning_rate": 6.843882798230184e-07,
-      "loss": 0.7609,
+      "loss": 0.5499,
       "step": 2751
     },
     {
       "epoch": 2.6563706563706564,
-      "grad_norm": 5.013240337371826,
+      "grad_norm": 5.38931131362915,
       "learning_rate": 6.806035772507169e-07,
-      "loss": 0.7242,
+      "loss": 0.526,
       "step": 2752
     },
     {
       "epoch": 2.6573359073359075,
-      "grad_norm": 6.189661026000977,
+      "grad_norm": 4.988227367401123,
       "learning_rate": 6.768289998174649e-07,
-      "loss": 0.7656,
+      "loss": 0.5342,
       "step": 2753
     },
     {
       "epoch": 2.658301158301158,
-      "grad_norm": 5.869439125061035,
+      "grad_norm": 5.64236307144165,
       "learning_rate": 6.730645516241785e-07,
-      "loss": 0.7486,
+      "loss": 0.5406,
       "step": 2754
     },
     {
       "epoch": 2.6592664092664093,
-      "grad_norm": 5.316051006317139,
+      "grad_norm": 5.056393623352051,
       "learning_rate": 6.693102367607795e-07,
-      "loss": 0.736,
+      "loss": 0.5184,
       "step": 2755
     },
     {
       "epoch": 2.66023166023166,
-      "grad_norm": 5.744836330413818,
+      "grad_norm": 4.964641571044922,
       "learning_rate": 6.655660593061719e-07,
-      "loss": 0.7973,
+      "loss": 0.5724,
       "step": 2756
     },
     {
       "epoch": 2.661196911196911,
-      "grad_norm": 6.7306809425354,
+      "grad_norm": 5.1389479637146,
       "learning_rate": 6.618320233282549e-07,
-      "loss": 0.7991,
+      "loss": 0.5597,
       "step": 2757
     },
     {
       "epoch": 2.6621621621621623,
-      "grad_norm": 5.0219268798828125,
+      "grad_norm": 5.0884504318237305,
       "learning_rate": 6.581081328839012e-07,
-      "loss": 0.7495,
+      "loss": 0.5289,
       "step": 2758
     },
     {
       "epoch": 2.663127413127413,
-      "grad_norm": 6.37299108505249,
+      "grad_norm": 4.794864177703857,
       "learning_rate": 6.543943920189633e-07,
-      "loss": 0.8041,
+      "loss": 0.5746,
       "step": 2759
     },
     {
       "epoch": 2.664092664092664,
-      "grad_norm": 5.728517055511475,
+      "grad_norm": 4.880391597747803,
       "learning_rate": 6.50690804768267e-07,
-      "loss": 0.794,
+      "loss": 0.5426,
       "step": 2760
     },
     {
       "epoch": 2.665057915057915,
-      "grad_norm": 5.198686122894287,
+      "grad_norm": 5.269046306610107,
       "learning_rate": 6.46997375155608e-07,
-      "loss": 0.7419,
+      "loss": 0.5001,
       "step": 2761
     },
     {
       "epoch": 2.666023166023166,
-      "grad_norm": 6.464615821838379,
+      "grad_norm": 5.2381415367126465,
       "learning_rate": 6.433141071937476e-07,
-      "loss": 0.7804,
+      "loss": 0.535,
       "step": 2762
     },
     {
       "epoch": 2.666988416988417,
-      "grad_norm": 5.360623836517334,
+      "grad_norm": 5.110682487487793,
       "learning_rate": 6.396410048843992e-07,
-      "loss": 0.7121,
+      "loss": 0.4976,
       "step": 2763
     },
     {
       "epoch": 2.667953667953668,
-      "grad_norm": 5.197407245635986,
+      "grad_norm": 5.019687175750732,
       "learning_rate": 6.359780722182374e-07,
-      "loss": 0.7632,
+      "loss": 0.5227,
       "step": 2764
     },
     {
       "epoch": 2.668918918918919,
-      "grad_norm": 5.302798748016357,
+      "grad_norm": 4.705749988555908,
       "learning_rate": 6.323253131748908e-07,
-      "loss": 0.7396,
+      "loss": 0.498,
       "step": 2765
     },
     {
       "epoch": 2.66988416988417,
-      "grad_norm": 4.82261323928833,
+      "grad_norm": 4.832378387451172,
       "learning_rate": 6.286827317229293e-07,
-      "loss": 0.7525,
+      "loss": 0.5085,
       "step": 2766
     },
     {
       "epoch": 2.6708494208494207,
-      "grad_norm": 5.102636814117432,
+      "grad_norm": 5.103469371795654,
       "learning_rate": 6.250503318198664e-07,
-      "loss": 0.7265,
+      "loss": 0.5051,
       "step": 2767
     },
     {
       "epoch": 2.671814671814672,
-      "grad_norm": 5.377608299255371,
+      "grad_norm": 4.912531852722168,
       "learning_rate": 6.214281174121518e-07,
-      "loss": 0.7501,
+      "loss": 0.5181,
       "step": 2768
     },
     {
       "epoch": 2.672779922779923,
-      "grad_norm": 5.08730411529541,
+      "grad_norm": 4.841454029083252,
       "learning_rate": 6.178160924351772e-07,
-      "loss": 0.7851,
+      "loss": 0.5533,
       "step": 2769
     },
     {
       "epoch": 2.6737451737451736,
-      "grad_norm": 5.115046501159668,
+      "grad_norm": 4.948421001434326,
       "learning_rate": 6.142142608132551e-07,
-      "loss": 0.7517,
+      "loss": 0.5223,
       "step": 2770
     },
     {
       "epoch": 2.6747104247104247,
-      "grad_norm": 4.9304022789001465,
+      "grad_norm": 4.87682580947876,
       "learning_rate": 6.106226264596293e-07,
-      "loss": 0.796,
+      "loss": 0.5428,
       "step": 2771
     },
     {
       "epoch": 2.6756756756756754,
-      "grad_norm": 5.703690052032471,
+      "grad_norm": 5.059298038482666,
       "learning_rate": 6.070411932764586e-07,
-      "loss": 0.8081,
+      "loss": 0.5672,
       "step": 2772
     },
     {
       "epoch": 2.6766409266409266,
-      "grad_norm": 4.732430934906006,
+      "grad_norm": 4.800141334533691,
       "learning_rate": 6.034699651548237e-07,
-      "loss": 0.7497,
+      "loss": 0.5173,
       "step": 2773
     },
     {
       "epoch": 2.6776061776061777,
-      "grad_norm": 5.347548484802246,
+      "grad_norm": 4.64623498916626,
       "learning_rate": 5.999089459747176e-07,
-      "loss": 0.7851,
+      "loss": 0.5372,
       "step": 2774
     },
     {
       "epoch": 2.678571428571429,
-      "grad_norm": 5.401774883270264,
+      "grad_norm": 5.102084636688232,
       "learning_rate": 5.9635813960504e-07,
-      "loss": 0.7221,
+      "loss": 0.4831,
       "step": 2775
     },
     {
       "epoch": 2.6795366795366795,
-      "grad_norm": 4.879080295562744,
+      "grad_norm": 4.993703365325928,
       "learning_rate": 5.928175499035926e-07,
-      "loss": 0.7224,
+      "loss": 0.5069,
       "step": 2776
     },
     {
       "epoch": 2.6805019305019306,
-      "grad_norm": 4.759188175201416,
+      "grad_norm": 4.624363899230957,
       "learning_rate": 5.89287180717083e-07,
-      "loss": 0.7287,
+      "loss": 0.5077,
       "step": 2777
     },
     {
       "epoch": 2.6814671814671813,
-      "grad_norm": 5.15285587310791,
+      "grad_norm": 4.758932590484619,
       "learning_rate": 5.857670358811096e-07,
-      "loss": 0.7374,
+      "loss": 0.5216,
       "step": 2778
     },
     {
       "epoch": 2.6824324324324325,
-      "grad_norm": 5.157477378845215,
+      "grad_norm": 5.287724494934082,
       "learning_rate": 5.822571192201654e-07,
-      "loss": 0.7771,
+      "loss": 0.5453,
       "step": 2779
     },
     {
       "epoch": 2.6833976833976836,
-      "grad_norm": 5.011092662811279,
+      "grad_norm": 5.294573783874512,
       "learning_rate": 5.787574345476299e-07,
-      "loss": 0.733,
+      "loss": 0.5024,
       "step": 2780
     },
     {
       "epoch": 2.6843629343629343,
-      "grad_norm": 5.20009708404541,
+      "grad_norm": 5.406548976898193,
       "learning_rate": 5.752679856657639e-07,
-      "loss": 0.7647,
+      "loss": 0.539,
       "step": 2781
     },
     {
       "epoch": 2.6853281853281854,
-      "grad_norm": 5.075290679931641,
+      "grad_norm": 5.139636516571045,
       "learning_rate": 5.717887763657115e-07,
-      "loss": 0.7759,
+      "loss": 0.5553,
       "step": 2782
     },
     {
       "epoch": 2.686293436293436,
-      "grad_norm": 4.610193252563477,
+      "grad_norm": 4.546214580535889,
       "learning_rate": 5.68319810427489e-07,
-      "loss": 0.7193,
+      "loss": 0.4936,
       "step": 2783
     },
     {
       "epoch": 2.687258687258687,
-      "grad_norm": 5.327815055847168,
+      "grad_norm": 5.247474193572998,
       "learning_rate": 5.64861091619987e-07,
-      "loss": 0.7594,
+      "loss": 0.5508,
       "step": 2784
     },
     {
       "epoch": 2.6882239382239383,
-      "grad_norm": 5.171527862548828,
+      "grad_norm": 4.687647342681885,
       "learning_rate": 5.614126237009543e-07,
-      "loss": 0.7818,
+      "loss": 0.5476,
       "step": 2785
     },
     {
       "epoch": 2.689189189189189,
-      "grad_norm": 5.051294326782227,
+      "grad_norm": 5.077347278594971,
       "learning_rate": 5.579744104170137e-07,
-      "loss": 0.753,
+      "loss": 0.5206,
       "step": 2786
     },
     {
       "epoch": 2.69015444015444,
-      "grad_norm": 5.165798664093018,
+      "grad_norm": 4.948495388031006,
       "learning_rate": 5.545464555036418e-07,
-      "loss": 0.7747,
+      "loss": 0.5108,
       "step": 2787
     },
     {
       "epoch": 2.691119691119691,
-      "grad_norm": 5.248556137084961,
+      "grad_norm": 5.382138729095459,
       "learning_rate": 5.511287626851681e-07,
-      "loss": 0.7073,
+      "loss": 0.5043,
       "step": 2788
     },
     {
       "epoch": 2.692084942084942,
-      "grad_norm": 4.877704620361328,
+      "grad_norm": 4.686574935913086,
       "learning_rate": 5.477213356747746e-07,
-      "loss": 0.7038,
+      "loss": 0.4846,
       "step": 2789
     },
     {
       "epoch": 2.693050193050193,
-      "grad_norm": 5.552155017852783,
+      "grad_norm": 4.853424072265625,
       "learning_rate": 5.443241781744923e-07,
-      "loss": 0.7444,
+      "loss": 0.5086,
       "step": 2790
     },
     {
       "epoch": 2.6940154440154442,
-      "grad_norm": 5.087784767150879,
+      "grad_norm": 4.842031955718994,
       "learning_rate": 5.4093729387519e-07,
-      "loss": 0.7423,
+      "loss": 0.514,
       "step": 2791
     },
     {
       "epoch": 2.694980694980695,
-      "grad_norm": 4.976923942565918,
+      "grad_norm": 4.946706771850586,
       "learning_rate": 5.375606864565786e-07,
-      "loss": 0.7266,
+      "loss": 0.5079,
       "step": 2792
     },
     {
       "epoch": 2.695945945945946,
-      "grad_norm": 5.3527302742004395,
+      "grad_norm": 4.790452003479004,
       "learning_rate": 5.341943595872012e-07,
-      "loss": 0.7541,
+      "loss": 0.5238,
       "step": 2793
     },
     {
       "epoch": 2.6969111969111967,
-      "grad_norm": 4.737863063812256,
+      "grad_norm": 4.842235088348389,
       "learning_rate": 5.308383169244346e-07,
-      "loss": 0.7284,
+      "loss": 0.5108,
       "step": 2794
     },
     {
       "epoch": 2.697876447876448,
-      "grad_norm": 4.960092067718506,
+      "grad_norm": 4.8339667320251465,
       "learning_rate": 5.274925621144799e-07,
-      "loss": 0.7621,
+      "loss": 0.5149,
       "step": 2795
     },
     {
       "epoch": 2.698841698841699,
-      "grad_norm": 5.724166393280029,
+      "grad_norm": 5.027866840362549,
       "learning_rate": 5.241570987923616e-07,
-      "loss": 0.7272,
+      "loss": 0.496,
       "step": 2796
     },
     {
       "epoch": 2.6998069498069497,
-      "grad_norm": 4.756486892700195,
+      "grad_norm": 5.168914318084717,
       "learning_rate": 5.208319305819209e-07,
-      "loss": 0.7545,
+      "loss": 0.5425,
       "step": 2797
     },
     {
       "epoch": 2.700772200772201,
-      "grad_norm": 4.879261016845703,
+      "grad_norm": 4.743222713470459,
       "learning_rate": 5.17517061095818e-07,
-      "loss": 0.77,
+      "loss": 0.5368,
       "step": 2798
     },
     {
       "epoch": 2.7017374517374515,
-      "grad_norm": 5.662070274353027,
+      "grad_norm": 5.201723575592041,
       "learning_rate": 5.142124939355209e-07,
-      "loss": 0.7919,
+      "loss": 0.5618,
       "step": 2799
     },
     {
       "epoch": 2.7027027027027026,
-      "grad_norm": 4.820952892303467,
+      "grad_norm": 4.524865627288818,
       "learning_rate": 5.109182326913053e-07,
-      "loss": 0.7558,
+      "loss": 0.5187,
       "step": 2800
     },
     {
       "epoch": 2.7036679536679538,
-      "grad_norm": 4.854535102844238,
+      "grad_norm": 4.588018417358398,
       "learning_rate": 5.076342809422485e-07,
-      "loss": 0.7414,
+      "loss": 0.5141,
       "step": 2801
     },
     {
       "epoch": 2.704633204633205,
-      "grad_norm": 5.061697483062744,
+      "grad_norm": 4.917291641235352,
       "learning_rate": 5.043606422562308e-07,
-      "loss": 0.7599,
+      "loss": 0.5147,
       "step": 2802
     },
     {
       "epoch": 2.7055984555984556,
-      "grad_norm": 4.653299331665039,
+      "grad_norm": 4.6834917068481445,
       "learning_rate": 5.010973201899228e-07,
-      "loss": 0.7372,
+      "loss": 0.5104,
       "step": 2803
     },
     {
       "epoch": 2.7065637065637067,
-      "grad_norm": 4.892110824584961,
+      "grad_norm": 4.701769828796387,
       "learning_rate": 4.978443182887904e-07,
-      "loss": 0.745,
+      "loss": 0.5255,
       "step": 2804
     },
     {
       "epoch": 2.7075289575289574,
-      "grad_norm": 4.729368686676025,
+      "grad_norm": 4.813390731811523,
       "learning_rate": 4.946016400870857e-07,
-      "loss": 0.6932,
+      "loss": 0.4905,
       "step": 2805
     },
     {
       "epoch": 2.7084942084942085,
-      "grad_norm": 5.155167579650879,
+      "grad_norm": 5.0029168128967285,
       "learning_rate": 4.913692891078437e-07,
-      "loss": 0.7692,
+      "loss": 0.5225,
       "step": 2806
     },
     {
       "epoch": 2.7094594594594597,
-      "grad_norm": 4.721940040588379,
+      "grad_norm": 4.860783576965332,
       "learning_rate": 4.88147268862883e-07,
-      "loss": 0.7415,
+      "loss": 0.5233,
       "step": 2807
     },
     {
       "epoch": 2.7104247104247103,
-      "grad_norm": 5.228612422943115,
+      "grad_norm": 5.056766986846924,
       "learning_rate": 4.849355828527924e-07,
-      "loss": 0.7987,
+      "loss": 0.557,
       "step": 2808
     },
     {
       "epoch": 2.7113899613899615,
-      "grad_norm": 5.587022304534912,
+      "grad_norm": 5.188930511474609,
       "learning_rate": 4.817342345669363e-07,
-      "loss": 0.7307,
+      "loss": 0.5081,
       "step": 2809
     },
     {
       "epoch": 2.712355212355212,
-      "grad_norm": 5.123222351074219,
+      "grad_norm": 4.9118757247924805,
       "learning_rate": 4.785432274834468e-07,
-      "loss": 0.7939,
+      "loss": 0.5432,
       "step": 2810
     },
     {
       "epoch": 2.7133204633204633,
-      "grad_norm": 4.989898204803467,
+      "grad_norm": 4.907232761383057,
       "learning_rate": 4.7536256506922507e-07,
-      "loss": 0.7548,
+      "loss": 0.5338,
       "step": 2811
     },
     {
       "epoch": 2.7142857142857144,
-      "grad_norm": 5.328431606292725,
+      "grad_norm": 4.8400702476501465,
       "learning_rate": 4.721922507799248e-07,
-      "loss": 0.7577,
+      "loss": 0.5221,
       "step": 2812
     },
     {
       "epoch": 2.715250965250965,
-      "grad_norm": 4.798675060272217,
+      "grad_norm": 4.920558452606201,
       "learning_rate": 4.6903228805996406e-07,
-      "loss": 0.757,
+      "loss": 0.5291,
       "step": 2813
     },
     {
       "epoch": 2.7162162162162162,
-      "grad_norm": 4.8693952560424805,
+      "grad_norm": 4.828951358795166,
       "learning_rate": 4.658826803425087e-07,
-      "loss": 0.7315,
+      "loss": 0.5272,
       "step": 2814
     },
     {
       "epoch": 2.717181467181467,
-      "grad_norm": 5.654747486114502,
+      "grad_norm": 5.254495143890381,
       "learning_rate": 4.6274343104948163e-07,
-      "loss": 0.7477,
+      "loss": 0.5129,
       "step": 2815
     },
     {
       "epoch": 2.718146718146718,
-      "grad_norm": 4.620950222015381,
+      "grad_norm": 4.74537467956543,
       "learning_rate": 4.596145435915467e-07,
-      "loss": 0.7357,
+      "loss": 0.5008,
       "step": 2816
     },
     {
       "epoch": 2.719111969111969,
-      "grad_norm": 4.7026567459106445,
+      "grad_norm": 4.618557929992676,
       "learning_rate": 4.56496021368108e-07,
-      "loss": 0.7102,
+      "loss": 0.4914,
       "step": 2817
     },
     {
       "epoch": 2.7200772200772203,
-      "grad_norm": 5.4267730712890625,
+      "grad_norm": 4.587148666381836,
       "learning_rate": 4.533878677673131e-07,
-      "loss": 0.7338,
+      "loss": 0.5077,
       "step": 2818
     },
     {
       "epoch": 2.721042471042471,
-      "grad_norm": 4.934582233428955,
+      "grad_norm": 4.922990798950195,
       "learning_rate": 4.502900861660398e-07,
-      "loss": 0.7352,
+      "loss": 0.5095,
       "step": 2819
     },
     {
       "epoch": 2.722007722007722,
-      "grad_norm": 5.04288911819458,
+      "grad_norm": 5.105484485626221,
       "learning_rate": 4.472026799299045e-07,
-      "loss": 0.7491,
+      "loss": 0.5398,
       "step": 2820
     },
     {
       "epoch": 2.722972972972973,
-      "grad_norm": 5.7374467849731445,
+      "grad_norm": 5.164768695831299,
       "learning_rate": 4.441256524132431e-07,
-      "loss": 0.7443,
+      "loss": 0.5334,
       "step": 2821
     },
     {
       "epoch": 2.723938223938224,
-      "grad_norm": 5.140861988067627,
+      "grad_norm": 5.0256524085998535,
       "learning_rate": 4.410590069591192e-07,
-      "loss": 0.7479,
+      "loss": 0.5193,
       "step": 2822
     },
     {
       "epoch": 2.724903474903475,
-      "grad_norm": 5.024938583374023,
+      "grad_norm": 4.9391655921936035,
       "learning_rate": 4.380027468993153e-07,
-      "loss": 0.7909,
+      "loss": 0.5445,
       "step": 2823
     },
     {
       "epoch": 2.7258687258687258,
-      "grad_norm": 6.299558639526367,
+      "grad_norm": 5.219255447387695,
       "learning_rate": 4.3495687555433544e-07,
-      "loss": 0.7606,
+      "loss": 0.5313,
       "step": 2824
     },
     {
       "epoch": 2.726833976833977,
-      "grad_norm": 5.034975051879883,
+      "grad_norm": 5.054938316345215,
       "learning_rate": 4.3192139623339125e-07,
-      "loss": 0.7507,
+      "loss": 0.5332,
       "step": 2825
     },
     {
       "epoch": 2.7277992277992276,
-      "grad_norm": 4.692763328552246,
+      "grad_norm": 4.7713189125061035,
       "learning_rate": 4.2889631223440255e-07,
-      "loss": 0.7735,
+      "loss": 0.5423,
       "step": 2826
     },
     {
       "epoch": 2.7287644787644787,
-      "grad_norm": 5.218514919281006,
+      "grad_norm": 4.929327487945557,
       "learning_rate": 4.2588162684400027e-07,
-      "loss": 0.7663,
+      "loss": 0.541,
       "step": 2827
     },
     {
       "epoch": 2.72972972972973,
-      "grad_norm": 5.270998954772949,
+      "grad_norm": 4.945908546447754,
       "learning_rate": 4.228773433375166e-07,
-      "loss": 0.795,
+      "loss": 0.5551,
       "step": 2828
     },
     {
       "epoch": 2.730694980694981,
-      "grad_norm": 4.963054656982422,
+      "grad_norm": 5.095919609069824,
       "learning_rate": 4.198834649789818e-07,
-      "loss": 0.7967,
+      "loss": 0.5727,
       "step": 2829
     },
     {
       "epoch": 2.7316602316602316,
-      "grad_norm": 4.898987293243408,
+      "grad_norm": 4.697289943695068,
       "learning_rate": 4.1689999502111835e-07,
-      "loss": 0.7473,
+      "loss": 0.5341,
       "step": 2830
     },
     {
       "epoch": 2.7326254826254828,
-      "grad_norm": 6.245449066162109,
+      "grad_norm": 5.276443004608154,
       "learning_rate": 4.1392693670534223e-07,
-      "loss": 0.7795,
+      "loss": 0.5536,
       "step": 2831
     },
     {
       "epoch": 2.7335907335907335,
-      "grad_norm": 5.4050421714782715,
+      "grad_norm": 5.237566947937012,
       "learning_rate": 4.10964293261763e-07,
-      "loss": 0.7471,
+      "loss": 0.5265,
       "step": 2832
     },
     {
       "epoch": 2.7345559845559846,
-      "grad_norm": 5.049526691436768,
+      "grad_norm": 5.058318138122559,
       "learning_rate": 4.0801206790916815e-07,
-      "loss": 0.7621,
+      "loss": 0.539,
       "step": 2833
     },
     {
       "epoch": 2.7355212355212357,
-      "grad_norm": 4.990163326263428,
+      "grad_norm": 4.92175817489624,
       "learning_rate": 4.0507026385502747e-07,
-      "loss": 0.7178,
+      "loss": 0.5188,
       "step": 2834
     },
     {
       "epoch": 2.7364864864864864,
-      "grad_norm": 5.667870998382568,
+      "grad_norm": 5.164303302764893,
       "learning_rate": 4.0213888429548877e-07,
-      "loss": 0.8263,
+      "loss": 0.5686,
       "step": 2835
     },
     {
       "epoch": 2.7374517374517375,
-      "grad_norm": 5.045688152313232,
+      "grad_norm": 4.613215923309326,
       "learning_rate": 3.9921793241537666e-07,
-      "loss": 0.7756,
+      "loss": 0.5301,
       "step": 2836
     },
     {
       "epoch": 2.738416988416988,
-      "grad_norm": 5.432678699493408,
+      "grad_norm": 5.151228427886963,
       "learning_rate": 3.9630741138818484e-07,
-      "loss": 0.756,
+      "loss": 0.5266,
       "step": 2837
     },
     {
       "epoch": 2.7393822393822393,
-      "grad_norm": 5.021127700805664,
+      "grad_norm": 4.799189567565918,
       "learning_rate": 3.9340732437607496e-07,
-      "loss": 0.7377,
+      "loss": 0.5198,
       "step": 2838
     },
     {
       "epoch": 2.7403474903474905,
-      "grad_norm": 4.582510948181152,
+      "grad_norm": 4.604424476623535,
       "learning_rate": 3.9051767452986777e-07,
-      "loss": 0.7567,
+      "loss": 0.5222,
       "step": 2839
     },
     {
       "epoch": 2.741312741312741,
-      "grad_norm": 6.305225849151611,
+      "grad_norm": 5.120781898498535,
       "learning_rate": 3.8763846498905187e-07,
-      "loss": 0.7504,
+      "loss": 0.5251,
       "step": 2840
     },
     {
       "epoch": 2.7422779922779923,
-      "grad_norm": 5.327244758605957,
+      "grad_norm": 5.069425106048584,
       "learning_rate": 3.8476969888176843e-07,
-      "loss": 0.7837,
+      "loss": 0.5652,
       "step": 2841
     },
     {
       "epoch": 2.743243243243243,
-      "grad_norm": 5.5716233253479,
+      "grad_norm": 5.333296775817871,
       "learning_rate": 3.8191137932481414e-07,
-      "loss": 0.7646,
+      "loss": 0.5842,
       "step": 2842
     },
     {
       "epoch": 2.744208494208494,
-      "grad_norm": 6.495911598205566,
+      "grad_norm": 4.8124213218688965,
       "learning_rate": 3.790635094236317e-07,
-      "loss": 0.7561,
+      "loss": 0.5264,
       "step": 2843
     },
     {
       "epoch": 2.7451737451737452,
-      "grad_norm": 4.862516403198242,
+      "grad_norm": 4.8003692626953125,
       "learning_rate": 3.762260922723182e-07,
-      "loss": 0.7373,
+      "loss": 0.5223,
       "step": 2844
     },
     {
       "epoch": 2.7461389961389964,
-      "grad_norm": 5.6016998291015625,
+      "grad_norm": 4.855422496795654,
       "learning_rate": 3.733991309536078e-07,
-      "loss": 0.7469,
+      "loss": 0.501,
       "step": 2845
     },
     {
       "epoch": 2.747104247104247,
-      "grad_norm": 5.131885051727295,
+      "grad_norm": 4.553038120269775,
       "learning_rate": 3.7058262853887805e-07,
-      "loss": 0.7512,
+      "loss": 0.5198,
       "step": 2846
     },
     {
       "epoch": 2.748069498069498,
-      "grad_norm": 4.686842441558838,
+      "grad_norm": 4.722749710083008,
       "learning_rate": 3.6777658808814234e-07,
-      "loss": 0.7179,
+      "loss": 0.4946,
       "step": 2847
     },
     {
       "epoch": 2.749034749034749,
-      "grad_norm": 5.378399848937988,
+      "grad_norm": 5.077066898345947,
       "learning_rate": 3.649810126500475e-07,
-      "loss": 0.8191,
+      "loss": 0.5736,
       "step": 2848
     },
     {
       "epoch": 2.75,
-      "grad_norm": 5.172610282897949,
+      "grad_norm": 4.869683265686035,
       "learning_rate": 3.6219590526187063e-07,
-      "loss": 0.7824,
+      "loss": 0.5491,
       "step": 2849
     },
     {
       "epoch": 2.750965250965251,
-      "grad_norm": 4.6634321212768555,
+      "grad_norm": 4.449004650115967,
       "learning_rate": 3.594212689495147e-07,
-      "loss": 0.7388,
+      "loss": 0.5103,
       "step": 2850
     },
     {
       "epoch": 2.751930501930502,
-      "grad_norm": 5.370039463043213,
+      "grad_norm": 4.713835716247559,
       "learning_rate": 3.566571067275093e-07,
-      "loss": 0.6907,
+      "loss": 0.4531,
       "step": 2851
     },
     {
       "epoch": 2.752895752895753,
-      "grad_norm": 4.233461380004883,
+      "grad_norm": 4.121732711791992,
       "learning_rate": 3.5390342159900225e-07,
-      "loss": 0.7345,
+      "loss": 0.5111,
       "step": 2852
     },
     {
       "epoch": 2.7538610038610036,
-      "grad_norm": 5.249642372131348,
+      "grad_norm": 4.982783317565918,
       "learning_rate": 3.511602165557571e-07,
-      "loss": 0.7928,
+      "loss": 0.5387,
       "step": 2853
     },
     {
       "epoch": 2.7548262548262548,
-      "grad_norm": 4.909470081329346,
+      "grad_norm": 5.174259662628174,
       "learning_rate": 3.4842749457815406e-07,
-      "loss": 0.7511,
+      "loss": 0.5529,
       "step": 2854
     },
     {
       "epoch": 2.755791505791506,
-      "grad_norm": 4.895900726318359,
+      "grad_norm": 4.703104019165039,
       "learning_rate": 3.457052586351817e-07,
-      "loss": 0.7404,
+      "loss": 0.5184,
       "step": 2855
     },
     {
       "epoch": 2.756756756756757,
-      "grad_norm": 5.3179216384887695,
+      "grad_norm": 5.381026268005371,
       "learning_rate": 3.429935116844374e-07,
-      "loss": 0.806,
+      "loss": 0.5704,
       "step": 2856
     },
     {
       "epoch": 2.7577220077220077,
-      "grad_norm": 4.825327396392822,
+      "grad_norm": 4.818737030029297,
       "learning_rate": 3.402922566721223e-07,
-      "loss": 0.7589,
+      "loss": 0.5134,
       "step": 2857
     },
     {
       "epoch": 2.758687258687259,
-      "grad_norm": 5.171874523162842,
+      "grad_norm": 5.417940139770508,
       "learning_rate": 3.3760149653303877e-07,
-      "loss": 0.7502,
+      "loss": 0.5287,
       "step": 2858
     },
     {
       "epoch": 2.7596525096525095,
-      "grad_norm": 4.987893104553223,
+      "grad_norm": 5.112573623657227,
       "learning_rate": 3.3492123419058497e-07,
-      "loss": 0.7794,
+      "loss": 0.5422,
       "step": 2859
     },
     {
       "epoch": 2.7606177606177607,
-      "grad_norm": 5.187206745147705,
+      "grad_norm": 5.20232629776001,
       "learning_rate": 3.3225147255675716e-07,
-      "loss": 0.8164,
+      "loss": 0.5829,
       "step": 2860
     },
     {
       "epoch": 2.761583011583012,
-      "grad_norm": 4.622416973114014,
+      "grad_norm": 4.654224395751953,
       "learning_rate": 3.295922145321384e-07,
-      "loss": 0.728,
+      "loss": 0.5194,
       "step": 2861
     },
     {
       "epoch": 2.7625482625482625,
-      "grad_norm": 5.248575687408447,
+      "grad_norm": 4.957277774810791,
       "learning_rate": 3.269434630059054e-07,
-      "loss": 0.7108,
+      "loss": 0.5051,
       "step": 2862
     },
     {
       "epoch": 2.7635135135135136,
-      "grad_norm": 5.089781761169434,
+      "grad_norm": 5.176142692565918,
       "learning_rate": 3.243052208558151e-07,
-      "loss": 0.7604,
+      "loss": 0.5467,
       "step": 2863
     },
     {
       "epoch": 2.7644787644787643,
-      "grad_norm": 5.344825744628906,
+      "grad_norm": 5.377825736999512,
       "learning_rate": 3.216774909482101e-07,
-      "loss": 0.8224,
+      "loss": 0.5762,
       "step": 2864
     },
     {
       "epoch": 2.7654440154440154,
-      "grad_norm": 4.867672920227051,
+      "grad_norm": 4.523346900939941,
       "learning_rate": 3.1906027613800903e-07,
-      "loss": 0.7378,
+      "loss": 0.4963,
       "step": 2865
     },
     {
       "epoch": 2.7664092664092665,
-      "grad_norm": 5.2442216873168945,
+      "grad_norm": 5.246423244476318,
       "learning_rate": 3.1645357926870957e-07,
-      "loss": 0.8109,
+      "loss": 0.573,
       "step": 2866
     },
     {
       "epoch": 2.7673745173745172,
-      "grad_norm": 4.97896146774292,
+      "grad_norm": 4.898642063140869,
       "learning_rate": 3.1385740317237733e-07,
-      "loss": 0.7306,
+      "loss": 0.4977,
       "step": 2867
     },
     {
       "epoch": 2.7683397683397684,
-      "grad_norm": 4.8300089836120605,
+      "grad_norm": 4.9068145751953125,
       "learning_rate": 3.1127175066965276e-07,
-      "loss": 0.7886,
+      "loss": 0.5573,
       "step": 2868
     },
     {
       "epoch": 2.769305019305019,
-      "grad_norm": 4.99078369140625,
+      "grad_norm": 5.002114295959473,
       "learning_rate": 3.086966245697387e-07,
-      "loss": 0.7539,
+      "loss": 0.4971,
       "step": 2869
     },
     {
       "epoch": 2.77027027027027,
-      "grad_norm": 5.40231466293335,
+      "grad_norm": 5.3776326179504395,
       "learning_rate": 3.06132027670405e-07,
-      "loss": 0.763,
+      "loss": 0.5347,
       "step": 2870
     },
     {
       "epoch": 2.7712355212355213,
-      "grad_norm": 5.105173587799072,
+      "grad_norm": 4.914074897766113,
       "learning_rate": 3.035779627579782e-07,
-      "loss": 0.7722,
+      "loss": 0.5404,
       "step": 2871
     },
     {
       "epoch": 2.7722007722007724,
-      "grad_norm": 5.617315292358398,
+      "grad_norm": 4.961867809295654,
       "learning_rate": 3.0103443260734557e-07,
-      "loss": 0.8002,
+      "loss": 0.5614,
       "step": 2872
     },
     {
       "epoch": 2.773166023166023,
-      "grad_norm": 5.030704021453857,
+      "grad_norm": 4.8710832595825195,
       "learning_rate": 2.9850143998194635e-07,
-      "loss": 0.7475,
+      "loss": 0.5434,
       "step": 2873
     },
     {
       "epoch": 2.7741312741312742,
-      "grad_norm": 5.375137805938721,
+      "grad_norm": 5.236742973327637,
       "learning_rate": 2.9597898763377595e-07,
-      "loss": 0.7464,
+      "loss": 0.5142,
       "step": 2874
     },
     {
       "epoch": 2.775096525096525,
-      "grad_norm": 4.688440799713135,
+      "grad_norm": 4.659385681152344,
       "learning_rate": 2.9346707830336997e-07,
-      "loss": 0.7595,
+      "loss": 0.5304,
       "step": 2875
     },
     {
       "epoch": 2.776061776061776,
-      "grad_norm": 4.799991130828857,
+      "grad_norm": 4.7530293464660645,
       "learning_rate": 2.9096571471981636e-07,
-      "loss": 0.7795,
+      "loss": 0.5468,
       "step": 2876
     },
     {
       "epoch": 2.777027027027027,
-      "grad_norm": 5.409911632537842,
+      "grad_norm": 5.276031494140625,
       "learning_rate": 2.8847489960074136e-07,
-      "loss": 0.7786,
+      "loss": 0.5575,
       "step": 2877
     },
     {
       "epoch": 2.777992277992278,
-      "grad_norm": 5.098364353179932,
+      "grad_norm": 4.971229553222656,
       "learning_rate": 2.8599463565231687e-07,
-      "loss": 0.8011,
+      "loss": 0.5413,
       "step": 2878
     },
     {
       "epoch": 2.778957528957529,
-      "grad_norm": 5.285067081451416,
+      "grad_norm": 5.172731876373291,
       "learning_rate": 2.8352492556924295e-07,
-      "loss": 0.7708,
+      "loss": 0.5527,
       "step": 2879
     },
     {
       "epoch": 2.7799227799227797,
-      "grad_norm": 4.846736431121826,
+      "grad_norm": 4.953556060791016,
       "learning_rate": 2.810657720347587e-07,
-      "loss": 0.7992,
+      "loss": 0.5669,
       "step": 2880
     },
     {
       "epoch": 2.780888030888031,
-      "grad_norm": 4.905336380004883,
+      "grad_norm": 4.756893157958984,
       "learning_rate": 2.786171777206337e-07,
-      "loss": 0.7135,
+      "loss": 0.4913,
       "step": 2881
     },
     {
       "epoch": 2.781853281853282,
-      "grad_norm": 4.571996212005615,
+      "grad_norm": 4.732292652130127,
       "learning_rate": 2.7617914528716317e-07,
-      "loss": 0.79,
+      "loss": 0.5542,
       "step": 2882
     },
     {
       "epoch": 2.782818532818533,
-      "grad_norm": 5.4335527420043945,
+      "grad_norm": 5.016957759857178,
       "learning_rate": 2.737516773831717e-07,
-      "loss": 0.7503,
+      "loss": 0.4999,
       "step": 2883
     },
     {
       "epoch": 2.7837837837837838,
-      "grad_norm": 4.6285223960876465,
+      "grad_norm": 4.592807292938232,
       "learning_rate": 2.713347766459984e-07,
-      "loss": 0.7467,
+      "loss": 0.5061,
       "step": 2884
     },
     {
       "epoch": 2.784749034749035,
-      "grad_norm": 4.802557945251465,
+      "grad_norm": 5.09492301940918,
       "learning_rate": 2.689284457015073e-07,
-      "loss": 0.7849,
+      "loss": 0.5639,
       "step": 2885
     },
     {
       "epoch": 2.7857142857142856,
-      "grad_norm": 5.64417028427124,
+      "grad_norm": 5.29189395904541,
       "learning_rate": 2.6653268716407944e-07,
-      "loss": 0.7614,
+      "loss": 0.5231,
       "step": 2886
     },
     {
       "epoch": 2.7866795366795367,
-      "grad_norm": 5.090156555175781,
+      "grad_norm": 4.804929256439209,
       "learning_rate": 2.6414750363660814e-07,
-      "loss": 0.7759,
+      "loss": 0.5328,
       "step": 2887
     },
     {
       "epoch": 2.787644787644788,
-      "grad_norm": 5.5157880783081055,
+      "grad_norm": 5.041073799133301,
       "learning_rate": 2.6177289771049274e-07,
-      "loss": 0.7346,
+      "loss": 0.5179,
       "step": 2888
     },
     {
       "epoch": 2.7886100386100385,
-      "grad_norm": 4.951792240142822,
+      "grad_norm": 4.780815601348877,
       "learning_rate": 2.5940887196564625e-07,
-      "loss": 0.7408,
+      "loss": 0.52,
       "step": 2889
     },
     {
       "epoch": 2.7895752895752897,
-      "grad_norm": 4.753869533538818,
+      "grad_norm": 4.883152008056641,
       "learning_rate": 2.570554289704863e-07,
-      "loss": 0.7802,
+      "loss": 0.5596,
       "step": 2890
     },
     {
       "epoch": 2.7905405405405403,
-      "grad_norm": 5.767086982727051,
+      "grad_norm": 5.0725626945495605,
       "learning_rate": 2.5471257128193094e-07,
-      "loss": 0.7615,
+      "loss": 0.5175,
       "step": 2891
     },
     {
       "epoch": 2.7915057915057915,
-      "grad_norm": 5.374652862548828,
+      "grad_norm": 4.442497253417969,
       "learning_rate": 2.5238030144539737e-07,
-      "loss": 0.6966,
+      "loss": 0.4589,
       "step": 2892
     },
     {
       "epoch": 2.7924710424710426,
-      "grad_norm": 5.251793384552002,
+      "grad_norm": 5.348597049713135,
       "learning_rate": 2.500586219947998e-07,
-      "loss": 0.7848,
+      "loss": 0.5183,
       "step": 2893
     },
     {
       "epoch": 2.7934362934362933,
-      "grad_norm": 5.737508296966553,
+      "grad_norm": 5.256956100463867,
       "learning_rate": 2.4774753545254936e-07,
-      "loss": 0.7652,
+      "loss": 0.5191,
       "step": 2894
     },
     {
       "epoch": 2.7944015444015444,
-      "grad_norm": 5.297785758972168,
+      "grad_norm": 5.04073429107666,
       "learning_rate": 2.4544704432954313e-07,
-      "loss": 0.7817,
+      "loss": 0.5186,
       "step": 2895
     },
     {
       "epoch": 2.795366795366795,
-      "grad_norm": 5.0009636878967285,
+      "grad_norm": 4.935091495513916,
       "learning_rate": 2.431571511251729e-07,
-      "loss": 0.8392,
+      "loss": 0.5618,
       "step": 2896
     },
     {
       "epoch": 2.7963320463320462,
-      "grad_norm": 4.5383100509643555,
+      "grad_norm": 4.528614521026611,
       "learning_rate": 2.408778583273108e-07,
-      "loss": 0.7715,
+      "loss": 0.5397,
       "step": 2897
     },
     {
       "epoch": 2.7972972972972974,
-      "grad_norm": 4.934019565582275,
+      "grad_norm": 4.703637599945068,
       "learning_rate": 2.3860916841231265e-07,
-      "loss": 0.7585,
+      "loss": 0.5119,
       "step": 2898
     },
     {
       "epoch": 2.7982625482625485,
-      "grad_norm": 4.724292755126953,
+      "grad_norm": 4.759828567504883,
       "learning_rate": 2.3635108384502003e-07,
-      "loss": 0.7789,
+      "loss": 0.5492,
       "step": 2899
     },
     {
       "epoch": 2.799227799227799,
-      "grad_norm": 5.758883476257324,
+      "grad_norm": 5.0521979331970215,
       "learning_rate": 2.3410360707874724e-07,
-      "loss": 0.7811,
+      "loss": 0.5246,
       "step": 2900
     },
     {
       "epoch": 2.8001930501930503,
-      "grad_norm": 4.7594475746154785,
+      "grad_norm": 4.533651351928711,
       "learning_rate": 2.3186674055528436e-07,
-      "loss": 0.7249,
+      "loss": 0.4833,
       "step": 2901
     },
     {
       "epoch": 2.801158301158301,
-      "grad_norm": 4.8631815910339355,
+      "grad_norm": 4.7712883949279785,
       "learning_rate": 2.2964048670489293e-07,
-      "loss": 0.7789,
+      "loss": 0.5419,
       "step": 2902
     },
     {
       "epoch": 2.802123552123552,
-      "grad_norm": 5.352580547332764,
+      "grad_norm": 5.2773756980896,
       "learning_rate": 2.2742484794630813e-07,
-      "loss": 0.7284,
+      "loss": 0.4973,
       "step": 2903
     },
     {
       "epoch": 2.8030888030888033,
-      "grad_norm": 5.275538921356201,
+      "grad_norm": 4.940196990966797,
       "learning_rate": 2.2521982668673003e-07,
-      "loss": 0.6938,
+      "loss": 0.4674,
       "step": 2904
     },
     {
       "epoch": 2.804054054054054,
-      "grad_norm": 5.052992820739746,
+      "grad_norm": 5.0227766036987305,
       "learning_rate": 2.2302542532182226e-07,
-      "loss": 0.7923,
+      "loss": 0.5493,
       "step": 2905
     },
     {
       "epoch": 2.805019305019305,
-      "grad_norm": 5.165040969848633,
+      "grad_norm": 4.966673374176025,
       "learning_rate": 2.208416462357088e-07,
-      "loss": 0.7193,
+      "loss": 0.4774,
       "step": 2906
     },
     {
       "epoch": 2.8059845559845558,
-      "grad_norm": 4.947180271148682,
+      "grad_norm": 4.703286647796631,
       "learning_rate": 2.186684918009796e-07,
-      "loss": 0.7712,
+      "loss": 0.5484,
       "step": 2907
     },
     {
       "epoch": 2.806949806949807,
-      "grad_norm": 5.16188907623291,
+      "grad_norm": 5.246286869049072,
       "learning_rate": 2.165059643786749e-07,
-      "loss": 0.7505,
+      "loss": 0.5228,
       "step": 2908
     },
     {
       "epoch": 2.807915057915058,
-      "grad_norm": 4.7499680519104,
+      "grad_norm": 4.638299942016602,
       "learning_rate": 2.1435406631829192e-07,
-      "loss": 0.7404,
+      "loss": 0.5238,
       "step": 2909
     },
     {
       "epoch": 2.808880308880309,
-      "grad_norm": 5.207340717315674,
+      "grad_norm": 4.797356128692627,
       "learning_rate": 2.1221279995777833e-07,
-      "loss": 0.7567,
+      "loss": 0.5242,
       "step": 2910
     },
     {
       "epoch": 2.80984555984556,
-      "grad_norm": 4.701102256774902,
+      "grad_norm": 4.508692264556885,
       "learning_rate": 2.1008216762353207e-07,
-      "loss": 0.7376,
+      "loss": 0.5089,
       "step": 2911
     },
     {
       "epoch": 2.810810810810811,
-      "grad_norm": 4.868410110473633,
+      "grad_norm": 4.9089179039001465,
       "learning_rate": 2.0796217163039588e-07,
-      "loss": 0.7915,
+      "loss": 0.5572,
       "step": 2912
     },
     {
       "epoch": 2.8117760617760617,
-      "grad_norm": 5.2054667472839355,
+      "grad_norm": 5.02157735824585,
       "learning_rate": 2.0585281428166071e-07,
-      "loss": 0.8091,
+      "loss": 0.5762,
       "step": 2913
     },
     {
       "epoch": 2.812741312741313,
-      "grad_norm": 4.966453552246094,
+      "grad_norm": 5.1007208824157715,
       "learning_rate": 2.037540978690533e-07,
-      "loss": 0.7822,
+      "loss": 0.5426,
       "step": 2914
     },
     {
       "epoch": 2.813706563706564,
-      "grad_norm": 4.5922532081604,
+      "grad_norm": 4.4503326416015625,
       "learning_rate": 2.016660246727431e-07,
-      "loss": 0.7637,
+      "loss": 0.5067,
       "step": 2915
     },
     {
       "epoch": 2.8146718146718146,
-      "grad_norm": 4.752979755401611,
+      "grad_norm": 4.977926254272461,
       "learning_rate": 1.995885969613365e-07,
-      "loss": 0.7835,
+      "loss": 0.5501,
       "step": 2916
     },
     {
       "epoch": 2.8156370656370657,
-      "grad_norm": 4.5410027503967285,
+      "grad_norm": 4.974874973297119,
       "learning_rate": 1.9752181699187133e-07,
-      "loss": 0.7396,
+      "loss": 0.5142,
       "step": 2917
     },
     {
       "epoch": 2.8166023166023164,
-      "grad_norm": 5.296461582183838,
+      "grad_norm": 5.1573076248168945,
       "learning_rate": 1.9546568700981927e-07,
-      "loss": 0.7387,
+      "loss": 0.5189,
       "step": 2918
     },
     {
       "epoch": 2.8175675675675675,
-      "grad_norm": 5.087576866149902,
+      "grad_norm": 5.160093784332275,
       "learning_rate": 1.9342020924908e-07,
-      "loss": 0.7756,
+      "loss": 0.5307,
       "step": 2919
     },
     {
       "epoch": 2.8185328185328187,
-      "grad_norm": 5.443266868591309,
+      "grad_norm": 5.529516220092773,
       "learning_rate": 1.9138538593198141e-07,
-      "loss": 0.7282,
+      "loss": 0.5053,
       "step": 2920
     },
     {
       "epoch": 2.8194980694980694,
-      "grad_norm": 5.033198356628418,
+      "grad_norm": 5.158353805541992,
       "learning_rate": 1.8936121926927508e-07,
-      "loss": 0.7677,
+      "loss": 0.5341,
       "step": 2921
     },
     {
       "epoch": 2.8204633204633205,
-      "grad_norm": 4.999009132385254,
+      "grad_norm": 4.881100654602051,
       "learning_rate": 1.8734771146013297e-07,
-      "loss": 0.7393,
+      "loss": 0.5111,
       "step": 2922
     },
     {
       "epoch": 2.821428571428571,
-      "grad_norm": 5.431258201599121,
+      "grad_norm": 5.3278279304504395,
       "learning_rate": 1.8534486469214962e-07,
-      "loss": 0.7643,
+      "loss": 0.5444,
       "step": 2923
     },
     {
       "epoch": 2.8223938223938223,
-      "grad_norm": 5.312672138214111,
+      "grad_norm": 5.186180591583252,
       "learning_rate": 1.8335268114133443e-07,
-      "loss": 0.7246,
+      "loss": 0.5101,
       "step": 2924
     },
     {
       "epoch": 2.8233590733590734,
-      "grad_norm": 4.68110466003418,
+      "grad_norm": 4.60059118270874,
       "learning_rate": 1.8137116297211266e-07,
-      "loss": 0.7418,
+      "loss": 0.5046,
       "step": 2925
     },
     {
       "epoch": 2.8243243243243246,
-      "grad_norm": 5.30441951751709,
+      "grad_norm": 4.951251029968262,
       "learning_rate": 1.7940031233732224e-07,
-      "loss": 0.7541,
+      "loss": 0.528,
       "step": 2926
     },
     {
       "epoch": 2.8252895752895753,
-      "grad_norm": 5.100651264190674,
+      "grad_norm": 4.834125518798828,
       "learning_rate": 1.7744013137821038e-07,
-      "loss": 0.7718,
+      "loss": 0.5406,
       "step": 2927
     },
     {
       "epoch": 2.8262548262548264,
-      "grad_norm": 4.791323184967041,
+      "grad_norm": 4.861802577972412,
       "learning_rate": 1.7549062222443347e-07,
-      "loss": 0.79,
+      "loss": 0.5703,
       "step": 2928
     },
     {
       "epoch": 2.827220077220077,
-      "grad_norm": 4.730557441711426,
+      "grad_norm": 4.645813465118408,
       "learning_rate": 1.7355178699405172e-07,
-      "loss": 0.7372,
+      "loss": 0.5079,
       "step": 2929
     },
     {
       "epoch": 2.828185328185328,
-      "grad_norm": 4.413017749786377,
+      "grad_norm": 4.52653169631958,
       "learning_rate": 1.7162362779353014e-07,
-      "loss": 0.7552,
+      "loss": 0.5234,
       "step": 2930
     },
     {
       "epoch": 2.8291505791505793,
-      "grad_norm": 4.423736572265625,
+      "grad_norm": 4.552811622619629,
       "learning_rate": 1.69706146717733e-07,
-      "loss": 0.768,
+      "loss": 0.507,
       "step": 2931
     },
     {
       "epoch": 2.83011583011583,
-      "grad_norm": 4.955407619476318,
+      "grad_norm": 4.709280967712402,
       "learning_rate": 1.677993458499272e-07,
-      "loss": 0.7212,
+      "loss": 0.4907,
       "step": 2932
     },
     {
       "epoch": 2.831081081081081,
-      "grad_norm": 5.887596130371094,
+      "grad_norm": 5.274842262268066,
       "learning_rate": 1.659032272617689e-07,
-      "loss": 0.7751,
+      "loss": 0.5311,
       "step": 2933
     },
     {
       "epoch": 2.832046332046332,
-      "grad_norm": 5.196951866149902,
+      "grad_norm": 4.888405799865723,
       "learning_rate": 1.6401779301331467e-07,
-      "loss": 0.7516,
+      "loss": 0.5063,
       "step": 2934
     },
     {
       "epoch": 2.833011583011583,
-      "grad_norm": 5.311980247497559,
+      "grad_norm": 5.198635101318359,
       "learning_rate": 1.621430451530115e-07,
-      "loss": 0.749,
+      "loss": 0.5162,
       "step": 2935
     },
     {
       "epoch": 2.833976833976834,
-      "grad_norm": 5.174939155578613,
+      "grad_norm": 4.855655193328857,
       "learning_rate": 1.602789857176945e-07,
-      "loss": 0.7625,
+      "loss": 0.4939,
       "step": 2936
     },
     {
       "epoch": 2.834942084942085,
-      "grad_norm": 4.715925216674805,
+      "grad_norm": 4.655534267425537,
       "learning_rate": 1.5842561673258705e-07,
-      "loss": 0.8064,
+      "loss": 0.5484,
       "step": 2937
     },
     {
       "epoch": 2.835907335907336,
-      "grad_norm": 4.72916841506958,
+      "grad_norm": 4.629755020141602,
       "learning_rate": 1.565829402112995e-07,
-      "loss": 0.7617,
+      "loss": 0.5345,
       "step": 2938
     },
     {
       "epoch": 2.836872586872587,
-      "grad_norm": 5.083279132843018,
+      "grad_norm": 5.064570426940918,
       "learning_rate": 1.5475095815582152e-07,
-      "loss": 0.8057,
+      "loss": 0.5761,
       "step": 2939
     },
     {
       "epoch": 2.8378378378378377,
-      "grad_norm": 4.998275279998779,
+      "grad_norm": 4.776557922363281,
       "learning_rate": 1.5292967255652657e-07,
-      "loss": 0.7013,
+      "loss": 0.4737,
       "step": 2940
     },
     {
       "epoch": 2.838803088803089,
-      "grad_norm": 4.9365553855896,
+      "grad_norm": 4.641363143920898,
       "learning_rate": 1.5111908539216845e-07,
-      "loss": 0.7415,
+      "loss": 0.5019,
       "step": 2941
     },
     {
       "epoch": 2.83976833976834,
-      "grad_norm": 4.951539039611816,
+      "grad_norm": 4.976378440856934,
       "learning_rate": 1.4931919862987143e-07,
-      "loss": 0.761,
+      "loss": 0.5325,
       "step": 2942
     },
     {
       "epoch": 2.8407335907335907,
-      "grad_norm": 5.639662265777588,
+      "grad_norm": 5.464055061340332,
       "learning_rate": 1.4753001422514125e-07,
-      "loss": 0.7355,
+      "loss": 0.5117,
       "step": 2943
     },
     {
       "epoch": 2.841698841698842,
-      "grad_norm": 5.438138961791992,
+      "grad_norm": 5.116255283355713,
       "learning_rate": 1.4575153412185072e-07,
-      "loss": 0.7831,
+      "loss": 0.5397,
       "step": 2944
     },
     {
       "epoch": 2.8426640926640925,
-      "grad_norm": 5.091670513153076,
+      "grad_norm": 4.943606853485107,
       "learning_rate": 1.4398376025224758e-07,
-      "loss": 0.7536,
+      "loss": 0.5075,
       "step": 2945
     },
     {
       "epoch": 2.8436293436293436,
-      "grad_norm": 5.158112525939941,
+      "grad_norm": 4.961364269256592,
       "learning_rate": 1.4222669453694215e-07,
-      "loss": 0.7313,
+      "loss": 0.492,
       "step": 2946
     },
     {
       "epoch": 2.8445945945945947,
-      "grad_norm": 5.176087379455566,
+      "grad_norm": 5.134489059448242,
       "learning_rate": 1.4048033888491518e-07,
-      "loss": 0.7136,
+      "loss": 0.4938,
       "step": 2947
     },
     {
       "epoch": 2.8455598455598454,
-      "grad_norm": 4.935799598693848,
+      "grad_norm": 5.117803573608398,
       "learning_rate": 1.3874469519350896e-07,
-      "loss": 0.8053,
+      "loss": 0.5615,
       "step": 2948
     },
     {
       "epoch": 2.8465250965250966,
-      "grad_norm": 4.985904216766357,
+      "grad_norm": 4.942277908325195,
       "learning_rate": 1.3701976534842953e-07,
-      "loss": 0.7894,
+      "loss": 0.5437,
       "step": 2949
     },
     {
       "epoch": 2.8474903474903472,
-      "grad_norm": 4.679719924926758,
+      "grad_norm": 4.37016487121582,
       "learning_rate": 1.3530555122374335e-07,
-      "loss": 0.7374,
+      "loss": 0.4907,
       "step": 2950
     },
     {
       "epoch": 2.8484555984555984,
-      "grad_norm": 5.193964004516602,
+      "grad_norm": 4.954007625579834,
       "learning_rate": 1.336020546818706e-07,
-      "loss": 0.704,
+      "loss": 0.4873,
       "step": 2951
     },
     {
       "epoch": 2.8494208494208495,
-      "grad_norm": 4.97507905960083,
+      "grad_norm": 4.970139980316162,
       "learning_rate": 1.3190927757358974e-07,
-      "loss": 0.7801,
+      "loss": 0.5373,
       "step": 2952
     },
     {
       "epoch": 2.8503861003861006,
-      "grad_norm": 4.854886531829834,
+      "grad_norm": 4.965250492095947,
       "learning_rate": 1.3022722173803516e-07,
-      "loss": 0.7937,
+      "loss": 0.5532,
       "step": 2953
     },
     {
       "epoch": 2.8513513513513513,
-      "grad_norm": 5.049645900726318,
+      "grad_norm": 5.065099239349365,
       "learning_rate": 1.2855588900269057e-07,
-      "loss": 0.8131,
+      "loss": 0.5525,
       "step": 2954
     },
     {
       "epoch": 2.8523166023166024,
-      "grad_norm": 4.761787414550781,
+      "grad_norm": 4.821765899658203,
       "learning_rate": 1.26895281183389e-07,
-      "loss": 0.7051,
+      "loss": 0.4907,
       "step": 2955
     },
     {
       "epoch": 2.853281853281853,
-      "grad_norm": 4.782556533813477,
+      "grad_norm": 4.620850563049316,
       "learning_rate": 1.2524540008431173e-07,
-      "loss": 0.7114,
+      "loss": 0.4862,
       "step": 2956
     },
     {
       "epoch": 2.8542471042471043,
-      "grad_norm": 5.348335266113281,
+      "grad_norm": 5.3231120109558105,
       "learning_rate": 1.2360624749798822e-07,
-      "loss": 0.7948,
+      "loss": 0.5543,
       "step": 2957
     },
     {
       "epoch": 2.8552123552123554,
-      "grad_norm": 5.1443657875061035,
+      "grad_norm": 4.985795974731445,
       "learning_rate": 1.2197782520528834e-07,
-      "loss": 0.7879,
+      "loss": 0.5321,
       "step": 2958
     },
     {
       "epoch": 2.856177606177606,
-      "grad_norm": 5.073531627655029,
+      "grad_norm": 5.238483905792236,
       "learning_rate": 1.2036013497542798e-07,
-      "loss": 0.7857,
+      "loss": 0.5714,
       "step": 2959
     },
     {
       "epoch": 2.857142857142857,
-      "grad_norm": 4.599514961242676,
+      "grad_norm": 4.567643642425537,
       "learning_rate": 1.187531785659568e-07,
-      "loss": 0.7515,
+      "loss": 0.5137,
       "step": 2960
     },
     {
       "epoch": 2.858108108108108,
-      "grad_norm": 4.990761756896973,
+      "grad_norm": 5.1691789627075195,
       "learning_rate": 1.1715695772277047e-07,
-      "loss": 0.7379,
+      "loss": 0.5098,
       "step": 2961
     },
     {
       "epoch": 2.859073359073359,
-      "grad_norm": 4.7035346031188965,
+      "grad_norm": 4.490543842315674,
       "learning_rate": 1.1557147418009396e-07,
-      "loss": 0.7611,
+      "loss": 0.5268,
       "step": 2962
     },
     {
       "epoch": 2.86003861003861,
-      "grad_norm": 5.617299556732178,
+      "grad_norm": 5.627073764801025,
       "learning_rate": 1.1399672966049269e-07,
-      "loss": 0.7622,
+      "loss": 0.5322,
       "step": 2963
     },
     {
       "epoch": 2.861003861003861,
-      "grad_norm": 4.790339946746826,
+      "grad_norm": 4.845139503479004,
       "learning_rate": 1.1243272587485809e-07,
-      "loss": 0.7545,
+      "loss": 0.5411,
       "step": 2964
     },
     {
       "epoch": 2.861969111969112,
-      "grad_norm": 5.162209510803223,
+      "grad_norm": 5.226559162139893,
       "learning_rate": 1.1087946452241871e-07,
-      "loss": 0.7503,
+      "loss": 0.5185,
       "step": 2965
     },
     {
       "epoch": 2.862934362934363,
-      "grad_norm": 4.988022327423096,
+      "grad_norm": 5.115107536315918,
       "learning_rate": 1.0933694729072686e-07,
-      "loss": 0.7562,
+      "loss": 0.5243,
       "step": 2966
     },
     {
       "epoch": 2.863899613899614,
-      "grad_norm": 4.976837635040283,
+      "grad_norm": 4.881818771362305,
       "learning_rate": 1.0780517585566419e-07,
-      "loss": 0.7398,
+      "loss": 0.4978,
       "step": 2967
     },
     {
       "epoch": 2.864864864864865,
-      "grad_norm": 4.8899922370910645,
+      "grad_norm": 4.9509172439575195,
       "learning_rate": 1.0628415188143726e-07,
-      "loss": 0.7984,
+      "loss": 0.5443,
       "step": 2968
     },
     {
       "epoch": 2.865830115830116,
-      "grad_norm": 4.969272613525391,
+      "grad_norm": 4.939781188964844,
       "learning_rate": 1.0477387702057529e-07,
-      "loss": 0.7793,
+      "loss": 0.5351,
       "step": 2969
     },
     {
       "epoch": 2.8667953667953667,
-      "grad_norm": 5.026739597320557,
+      "grad_norm": 5.007850170135498,
       "learning_rate": 1.0327435291393018e-07,
-      "loss": 0.76,
+      "loss": 0.5242,
       "step": 2970
     },
     {
       "epoch": 2.867760617760618,
-      "grad_norm": 4.789086818695068,
+      "grad_norm": 5.053526401519775,
       "learning_rate": 1.0178558119067316e-07,
-      "loss": 0.7797,
+      "loss": 0.5608,
       "step": 2971
     },
     {
       "epoch": 2.8687258687258685,
-      "grad_norm": 4.7164225578308105,
+      "grad_norm": 4.78934907913208,
       "learning_rate": 1.0030756346829151e-07,
-      "loss": 0.7619,
+      "loss": 0.5232,
       "step": 2972
     },
     {
       "epoch": 2.8696911196911197,
-      "grad_norm": 4.880717754364014,
+      "grad_norm": 4.513342380523682,
       "learning_rate": 9.884030135259182e-08,
-      "loss": 0.736,
+      "loss": 0.5065,
       "step": 2973
     },
     {
       "epoch": 2.870656370656371,
-      "grad_norm": 4.9621171951293945,
+      "grad_norm": 4.656712055206299,
       "learning_rate": 9.738379643769446e-08,
-      "loss": 0.7874,
+      "loss": 0.5505,
       "step": 2974
     },
     {
       "epoch": 2.8716216216216215,
-      "grad_norm": 4.993303298950195,
+      "grad_norm": 4.887301445007324,
       "learning_rate": 9.593805030602921e-08,
-      "loss": 0.7836,
+      "loss": 0.5464,
       "step": 2975
     },
     {
       "epoch": 2.8725868725868726,
-      "grad_norm": 5.17083740234375,
+      "grad_norm": 5.140177249908447,
       "learning_rate": 9.45030645283418e-08,
-      "loss": 0.7683,
+      "loss": 0.5334,
       "step": 2976
     },
     {
       "epoch": 2.8735521235521233,
-      "grad_norm": 5.310078144073486,
+      "grad_norm": 5.473716735839844,
       "learning_rate": 9.3078840663684e-08,
-      "loss": 0.7087,
+      "loss": 0.5133,
       "step": 2977
     },
     {
       "epoch": 2.8745173745173744,
-      "grad_norm": 4.97788667678833,
+      "grad_norm": 4.9361186027526855,
       "learning_rate": 9.166538025941695e-08,
-      "loss": 0.791,
+      "loss": 0.5424,
       "step": 2978
     },
     {
       "epoch": 2.8754826254826256,
-      "grad_norm": 4.916007041931152,
+      "grad_norm": 4.956519603729248,
       "learning_rate": 9.02626848512067e-08,
-      "loss": 0.7509,
+      "loss": 0.5201,
       "step": 2979
     },
     {
       "epoch": 2.8764478764478767,
-      "grad_norm": 4.999144554138184,
+      "grad_norm": 5.262781143188477,
       "learning_rate": 8.887075596302197e-08,
-      "loss": 0.7848,
+      "loss": 0.5357,
       "step": 2980
     },
     {
       "epoch": 2.8774131274131274,
-      "grad_norm": 5.314653396606445,
+      "grad_norm": 5.275564193725586,
       "learning_rate": 8.748959510713861e-08,
-      "loss": 0.7556,
+      "loss": 0.5292,
       "step": 2981
     },
     {
       "epoch": 2.8783783783783785,
-      "grad_norm": 5.080662727355957,
+      "grad_norm": 4.906208038330078,
       "learning_rate": 8.611920378412963e-08,
-      "loss": 0.793,
+      "loss": 0.5362,
       "step": 2982
     },
     {
       "epoch": 2.879343629343629,
-      "grad_norm": 5.135159015655518,
+      "grad_norm": 4.9854254722595215,
       "learning_rate": 8.475958348286962e-08,
-      "loss": 0.7788,
+      "loss": 0.5462,
       "step": 2983
     },
     {
       "epoch": 2.8803088803088803,
-      "grad_norm": 4.982119083404541,
+      "grad_norm": 4.777050971984863,
       "learning_rate": 8.341073568052804e-08,
-      "loss": 0.7507,
+      "loss": 0.4829,
       "step": 2984
     },
     {
       "epoch": 2.8812741312741315,
-      "grad_norm": 4.8198018074035645,
+      "grad_norm": 4.587100505828857,
       "learning_rate": 8.207266184257268e-08,
-      "loss": 0.7585,
+      "loss": 0.5246,
       "step": 2985
     },
     {
       "epoch": 2.882239382239382,
-      "grad_norm": 4.909982681274414,
+      "grad_norm": 4.932624340057373,
       "learning_rate": 8.074536342276618e-08,
-      "loss": 0.7613,
+      "loss": 0.5314,
       "step": 2986
     },
     {
       "epoch": 2.8832046332046333,
-      "grad_norm": 4.988618850708008,
+      "grad_norm": 5.164953708648682,
       "learning_rate": 7.94288418631639e-08,
-      "loss": 0.8453,
+      "loss": 0.5913,
       "step": 2987
     },
     {
       "epoch": 2.884169884169884,
-      "grad_norm": 4.8787689208984375,
+      "grad_norm": 4.875906944274902,
       "learning_rate": 7.81230985941106e-08,
-      "loss": 0.7528,
+      "loss": 0.5201,
       "step": 2988
     },
     {
       "epoch": 2.885135135135135,
-      "grad_norm": 4.880171775817871,
+      "grad_norm": 4.863702774047852,
       "learning_rate": 7.68281350342437e-08,
-      "loss": 0.7982,
+      "loss": 0.5673,
       "step": 2989
     },
     {
       "epoch": 2.886100386100386,
-      "grad_norm": 4.924539566040039,
+      "grad_norm": 5.056474208831787,
       "learning_rate": 7.554395259048664e-08,
-      "loss": 0.7348,
+      "loss": 0.4953,
       "step": 2990
     },
     {
       "epoch": 2.887065637065637,
-      "grad_norm": 4.793140888214111,
+      "grad_norm": 4.7832746505737305,
       "learning_rate": 7.427055265805006e-08,
-      "loss": 0.7723,
+      "loss": 0.5234,
       "step": 2991
     },
     {
       "epoch": 2.888030888030888,
-      "grad_norm": 4.872824668884277,
+      "grad_norm": 4.696780204772949,
       "learning_rate": 7.300793662043282e-08,
-      "loss": 0.746,
+      "loss": 0.5176,
       "step": 2992
     },
     {
       "epoch": 2.888996138996139,
-      "grad_norm": 4.631551742553711,
+      "grad_norm": 4.492559432983398,
       "learning_rate": 7.175610584941317e-08,
-      "loss": 0.7264,
+      "loss": 0.5104,
       "step": 2993
     },
     {
       "epoch": 2.88996138996139,
-      "grad_norm": 4.891885757446289,
+      "grad_norm": 5.119144916534424,
       "learning_rate": 7.051506170505318e-08,
-      "loss": 0.8092,
+      "loss": 0.5611,
       "step": 2994
     },
     {
       "epoch": 2.890926640926641,
-      "grad_norm": 5.292384624481201,
+      "grad_norm": 5.008601665496826,
       "learning_rate": 6.92848055356976e-08,
-      "loss": 0.7981,
+      "loss": 0.5426,
       "step": 2995
     },
     {
       "epoch": 2.891891891891892,
-      "grad_norm": 4.880120754241943,
+      "grad_norm": 4.926332950592041,
       "learning_rate": 6.806533867796728e-08,
-      "loss": 0.788,
+      "loss": 0.5467,
       "step": 2996
     },
     {
       "epoch": 2.892857142857143,
-      "grad_norm": 4.910423278808594,
+      "grad_norm": 4.861738204956055,
       "learning_rate": 6.685666245676348e-08,
-      "loss": 0.7714,
+      "loss": 0.5351,
       "step": 2997
     },
     {
       "epoch": 2.893822393822394,
-      "grad_norm": 4.717833042144775,
+      "grad_norm": 4.890074729919434,
       "learning_rate": 6.565877818526245e-08,
-      "loss": 0.7899,
+      "loss": 0.5544,
       "step": 2998
     },
     {
       "epoch": 2.8947876447876446,
-      "grad_norm": 5.327867031097412,
+      "grad_norm": 5.253537654876709,
       "learning_rate": 6.447168716491648e-08,
-      "loss": 0.8035,
+      "loss": 0.5691,
       "step": 2999
     },
     {
       "epoch": 2.8957528957528957,
-      "grad_norm": 4.775325298309326,
+      "grad_norm": 4.624936103820801,
       "learning_rate": 6.329539068545054e-08,
-      "loss": 0.728,
+      "loss": 0.4892,
       "step": 3000
     },
     {
       "epoch": 2.896718146718147,
-      "grad_norm": 5.513738632202148,
+      "grad_norm": 5.1580657958984375,
       "learning_rate": 6.212989002486236e-08,
-      "loss": 0.7387,
+      "loss": 0.4961,
       "step": 3001
     },
     {
       "epoch": 2.8976833976833976,
-      "grad_norm": 4.775712490081787,
+      "grad_norm": 4.736073970794678,
       "learning_rate": 6.097518644942013e-08,
-      "loss": 0.7119,
+      "loss": 0.4873,
       "step": 3002
     },
     {
       "epoch": 2.8986486486486487,
-      "grad_norm": 4.993870258331299,
+      "grad_norm": 5.123494625091553,
       "learning_rate": 5.983128121366255e-08,
-      "loss": 0.7261,
+      "loss": 0.4967,
       "step": 3003
     },
     {
       "epoch": 2.8996138996138994,
-      "grad_norm": 4.865166187286377,
+      "grad_norm": 4.917851448059082,
       "learning_rate": 5.869817556039325e-08,
-      "loss": 0.7597,
+      "loss": 0.5182,
       "step": 3004
     },
     {
       "epoch": 2.9005791505791505,
-      "grad_norm": 5.616902828216553,
+      "grad_norm": 5.349608898162842,
       "learning_rate": 5.757587072068638e-08,
-      "loss": 0.8111,
+      "loss": 0.5628,
       "step": 3005
     },
     {
       "epoch": 2.9015444015444016,
-      "grad_norm": 5.1008429527282715,
+      "grad_norm": 4.981143474578857,
       "learning_rate": 5.646436791387766e-08,
-      "loss": 0.734,
+      "loss": 0.5124,
       "step": 3006
     },
     {
       "epoch": 2.9025096525096528,
-      "grad_norm": 4.946638107299805,
+      "grad_norm": 5.081432342529297,
       "learning_rate": 5.536366834757223e-08,
-      "loss": 0.7565,
+      "loss": 0.5233,
       "step": 3007
     },
     {
       "epoch": 2.9034749034749034,
-      "grad_norm": 4.6390061378479,
+      "grad_norm": 4.7435784339904785,
       "learning_rate": 5.427377321763239e-08,
-      "loss": 0.7207,
+      "loss": 0.5041,
       "step": 3008
     },
     {
       "epoch": 2.9044401544401546,
-      "grad_norm": 4.830086708068848,
+      "grad_norm": 4.702599048614502,
       "learning_rate": 5.319468370818537e-08,
-      "loss": 0.7857,
+      "loss": 0.5352,
       "step": 3009
     },
     {
       "epoch": 2.9054054054054053,
-      "grad_norm": 4.865915298461914,
+      "grad_norm": 4.6741814613342285,
       "learning_rate": 5.212640099161559e-08,
-      "loss": 0.8014,
+      "loss": 0.5499,
       "step": 3010
     },
     {
       "epoch": 2.9063706563706564,
-      "grad_norm": 4.121964454650879,
+      "grad_norm": 4.191152572631836,
       "learning_rate": 5.106892622857018e-08,
-      "loss": 0.7236,
+      "loss": 0.4999,
       "step": 3011
     },
     {
       "epoch": 2.9073359073359075,
-      "grad_norm": 5.154491901397705,
+      "grad_norm": 5.19315767288208,
       "learning_rate": 5.002226056795123e-08,
-      "loss": 0.7661,
+      "loss": 0.5279,
       "step": 3012
     },
     {
       "epoch": 2.908301158301158,
-      "grad_norm": 5.089937686920166,
+      "grad_norm": 4.906758785247803,
       "learning_rate": 4.898640514691577e-08,
-      "loss": 0.7765,
+      "loss": 0.5402,
       "step": 3013
     },
     {
       "epoch": 2.9092664092664093,
-      "grad_norm": 4.757026195526123,
+      "grad_norm": 4.853614330291748,
       "learning_rate": 4.796136109088023e-08,
-      "loss": 0.7503,
+      "loss": 0.5181,
       "step": 3014
     },
     {
       "epoch": 2.91023166023166,
-      "grad_norm": 4.677746295928955,
+      "grad_norm": 4.901845932006836,
       "learning_rate": 4.694712951351044e-08,
-      "loss": 0.7914,
+      "loss": 0.5507,
       "step": 3015
     },
     {
       "epoch": 2.911196911196911,
-      "grad_norm": 4.440855026245117,
+      "grad_norm": 4.614941120147705,
       "learning_rate": 4.594371151673049e-08,
-      "loss": 0.7548,
+      "loss": 0.5179,
       "step": 3016
     },
     {
       "epoch": 2.9121621621621623,
-      "grad_norm": 5.054029941558838,
+      "grad_norm": 4.903169631958008,
       "learning_rate": 4.4951108190709466e-08,
-      "loss": 0.7766,
+      "loss": 0.5166,
       "step": 3017
     },
     {
       "epoch": 2.913127413127413,
-      "grad_norm": 4.783627986907959,
+      "grad_norm": 4.7379326820373535,
       "learning_rate": 4.396932061387138e-08,
-      "loss": 0.7003,
+      "loss": 0.4655,
       "step": 3018
     },
     {
       "epoch": 2.914092664092664,
-      "grad_norm": 4.913122653961182,
+      "grad_norm": 4.942343711853027,
       "learning_rate": 4.299834985288631e-08,
-      "loss": 0.7566,
+      "loss": 0.5187,
       "step": 3019
     },
     {
       "epoch": 2.915057915057915,
-      "grad_norm": 5.131570339202881,
+      "grad_norm": 4.926549911499023,
       "learning_rate": 4.203819696267486e-08,
-      "loss": 0.7938,
+      "loss": 0.5599,
       "step": 3020
     },
     {
       "epoch": 2.916023166023166,
-      "grad_norm": 4.725548267364502,
+      "grad_norm": 4.897482395172119,
       "learning_rate": 4.10888629864048e-08,
-      "loss": 0.7956,
+      "loss": 0.5643,
       "step": 3021
     },
     {
       "epoch": 2.916988416988417,
-      "grad_norm": 4.535856246948242,
+      "grad_norm": 4.643405914306641,
       "learning_rate": 4.015034895548664e-08,
-      "loss": 0.7068,
+      "loss": 0.5105,
       "step": 3022
     },
     {
       "epoch": 2.917953667953668,
-      "grad_norm": 4.978786945343018,
+      "grad_norm": 5.248547554016113,
       "learning_rate": 3.922265588957586e-08,
-      "loss": 0.7321,
+      "loss": 0.4953,
       "step": 3023
     },
     {
       "epoch": 2.918918918918919,
-      "grad_norm": 4.783323764801025,
+      "grad_norm": 5.124320030212402,
       "learning_rate": 3.8305784796576203e-08,
-      "loss": 0.7729,
+      "loss": 0.5625,
       "step": 3024
     },
     {
       "epoch": 2.91988416988417,
-      "grad_norm": 4.754459381103516,
+      "grad_norm": 4.723893165588379,
       "learning_rate": 3.739973667262642e-08,
-      "loss": 0.7401,
+      "loss": 0.5038,
       "step": 3025
     },
     {
       "epoch": 2.9208494208494207,
-      "grad_norm": 4.781511306762695,
+      "grad_norm": 4.5857954025268555,
       "learning_rate": 3.6504512502113506e-08,
-      "loss": 0.7532,
+      "loss": 0.5027,
       "step": 3026
     },
     {
       "epoch": 2.921814671814672,
-      "grad_norm": 4.7534565925598145,
+      "grad_norm": 4.543224811553955,
       "learning_rate": 3.562011325765946e-08,
-      "loss": 0.7489,
+      "loss": 0.5162,
       "step": 3027
     },
     {
       "epoch": 2.922779922779923,
-      "grad_norm": 4.938658714294434,
+      "grad_norm": 4.846933841705322,
       "learning_rate": 3.4746539900129e-08,
-      "loss": 0.7472,
+      "loss": 0.5192,
       "step": 3028
     },
     {
       "epoch": 2.9237451737451736,
-      "grad_norm": 4.837621688842773,
+      "grad_norm": 4.467364311218262,
       "learning_rate": 3.38837933786218e-08,
-      "loss": 0.7241,
+      "loss": 0.4829,
       "step": 3029
     },
     {
       "epoch": 2.9247104247104247,
-      "grad_norm": 4.8044939041137695,
+      "grad_norm": 4.420899868011475,
       "learning_rate": 3.3031874630478076e-08,
-      "loss": 0.7251,
+      "loss": 0.4921,
       "step": 3030
     },
     {
       "epoch": 2.9256756756756754,
-      "grad_norm": 4.589970588684082,
+      "grad_norm": 4.69775915145874,
       "learning_rate": 3.2190784581270786e-08,
-      "loss": 0.7609,
+      "loss": 0.5262,
       "step": 3031
     },
     {
       "epoch": 2.9266409266409266,
-      "grad_norm": 5.225532531738281,
+      "grad_norm": 5.100448131561279,
       "learning_rate": 3.1360524144810055e-08,
-      "loss": 0.8266,
+      "loss": 0.571,
       "step": 3032
     },
     {
       "epoch": 2.9276061776061777,
-      "grad_norm": 4.62082576751709,
+      "grad_norm": 4.896040916442871,
       "learning_rate": 3.054109422313989e-08,
-      "loss": 0.7959,
+      "loss": 0.5567,
       "step": 3033
     },
     {
       "epoch": 2.928571428571429,
-      "grad_norm": 4.918736457824707,
+      "grad_norm": 5.331302642822266,
       "learning_rate": 2.9732495706535913e-08,
-      "loss": 0.7869,
+      "loss": 0.5642,
       "step": 3034
     },
     {
       "epoch": 2.9295366795366795,
-      "grad_norm": 4.885461330413818,
+      "grad_norm": 4.965645790100098,
       "learning_rate": 2.8934729473508726e-08,
-      "loss": 0.7288,
+      "loss": 0.5136,
       "step": 3035
     },
     {
       "epoch": 2.9305019305019306,
-      "grad_norm": 4.932406425476074,
+      "grad_norm": 4.881459712982178,
       "learning_rate": 2.814779639079723e-08,
-      "loss": 0.7362,
+      "loss": 0.4848,
       "step": 3036
     },
     {
       "epoch": 2.9314671814671813,
-      "grad_norm": 5.010663032531738,
+      "grad_norm": 4.871085166931152,
       "learning_rate": 2.737169731337197e-08,
-      "loss": 0.7421,
+      "loss": 0.4915,
       "step": 3037
     },
     {
       "epoch": 2.9324324324324325,
-      "grad_norm": 5.282849311828613,
+      "grad_norm": 4.781154632568359,
       "learning_rate": 2.6606433084432892e-08,
-      "loss": 0.7304,
+      "loss": 0.4827,
       "step": 3038
     },
     {
       "epoch": 2.9333976833976836,
-      "grad_norm": 4.926002025604248,
+      "grad_norm": 5.2007222175598145,
       "learning_rate": 2.585200453540715e-08,
-      "loss": 0.7978,
+      "loss": 0.561,
       "step": 3039
     },
     {
       "epoch": 2.9343629343629343,
-      "grad_norm": 4.964679718017578,
+      "grad_norm": 5.135045528411865,
       "learning_rate": 2.5108412485951306e-08,
-      "loss": 0.7656,
+      "loss": 0.5337,
       "step": 3040
     },
     {
       "epoch": 2.9353281853281854,
-      "grad_norm": 4.717856407165527,
+      "grad_norm": 5.009424209594727,
       "learning_rate": 2.437565774394579e-08,
-      "loss": 0.8208,
+      "loss": 0.5717,
       "step": 3041
     },
     {
       "epoch": 2.936293436293436,
-      "grad_norm": 4.755115032196045,
+      "grad_norm": 5.051687240600586,
       "learning_rate": 2.3653741105499338e-08,
-      "loss": 0.7739,
+      "loss": 0.535,
       "step": 3042
     },
     {
       "epoch": 2.937258687258687,
-      "grad_norm": 4.922940731048584,
+      "grad_norm": 5.135486602783203,
       "learning_rate": 2.294266335494566e-08,
-      "loss": 0.7458,
+      "loss": 0.5096,
       "step": 3043
     },
     {
       "epoch": 2.9382239382239383,
-      "grad_norm": 4.712069988250732,
+      "grad_norm": 4.8160719871521,
       "learning_rate": 2.2242425264838997e-08,
-      "loss": 0.7617,
+      "loss": 0.5157,
       "step": 3044
     },
     {
       "epoch": 2.939189189189189,
-      "grad_norm": 4.8516106605529785,
+      "grad_norm": 5.10272741317749,
       "learning_rate": 2.1553027595959676e-08,
-      "loss": 0.8564,
+      "loss": 0.5826,
       "step": 3045
     },
     {
       "epoch": 2.94015444015444,
-      "grad_norm": 5.042926788330078,
+      "grad_norm": 4.727243900299072,
       "learning_rate": 2.0874471097311886e-08,
-      "loss": 0.7718,
+      "loss": 0.5118,
       "step": 3046
     },
     {
       "epoch": 2.941119691119691,
-      "grad_norm": 4.81359338760376,
+      "grad_norm": 4.879641056060791,
       "learning_rate": 2.020675650611703e-08,
-      "loss": 0.6968,
+      "loss": 0.4693,
       "step": 3047
     },
     {
       "epoch": 2.942084942084942,
-      "grad_norm": 5.007816314697266,
+      "grad_norm": 4.865490913391113,
       "learning_rate": 1.9549884547819253e-08,
-      "loss": 0.7924,
+      "loss": 0.5472,
       "step": 3048
     },
     {
       "epoch": 2.943050193050193,
-      "grad_norm": 4.917684078216553,
+      "grad_norm": 5.047770977020264,
       "learning_rate": 1.890385593608435e-08,
-      "loss": 0.7483,
+      "loss": 0.5222,
       "step": 3049
     },
     {
       "epoch": 2.9440154440154442,
-      "grad_norm": 4.8487114906311035,
+      "grad_norm": 4.634869575500488,
       "learning_rate": 1.82686713727942e-08,
-      "loss": 0.7795,
+      "loss": 0.5428,
       "step": 3050
     },
     {
       "epoch": 2.944980694980695,
-      "grad_norm": 4.602555274963379,
+      "grad_norm": 4.592961311340332,
       "learning_rate": 1.7644331548052342e-08,
-      "loss": 0.7561,
+      "loss": 0.51,
       "step": 3051
     },
     {
       "epoch": 2.945945945945946,
-      "grad_norm": 5.175610065460205,
+      "grad_norm": 4.75779914855957,
       "learning_rate": 1.703083714017617e-08,
-      "loss": 0.783,
+      "loss": 0.5292,
       "step": 3052
     },
     {
       "epoch": 2.9469111969111967,
-      "grad_norm": 5.09755277633667,
+      "grad_norm": 4.959355354309082,
       "learning_rate": 1.6428188815703627e-08,
-      "loss": 0.7955,
+      "loss": 0.5405,
       "step": 3053
     },
     {
       "epoch": 2.947876447876448,
-      "grad_norm": 4.747951030731201,
+      "grad_norm": 4.939132213592529,
       "learning_rate": 1.583638722938652e-08,
-      "loss": 0.7569,
+      "loss": 0.5175,
       "step": 3054
     },
     {
       "epoch": 2.948841698841699,
-      "grad_norm": 4.651163578033447,
+      "grad_norm": 4.663990497589111,
       "learning_rate": 1.5255433024192746e-08,
-      "loss": 0.756,
+      "loss": 0.5092,
       "step": 3055
     },
     {
       "epoch": 2.9498069498069497,
-      "grad_norm": 4.646449565887451,
+      "grad_norm": 4.954626560211182,
       "learning_rate": 1.4685326831305192e-08,
-      "loss": 0.7379,
+      "loss": 0.5195,
       "step": 3056
     },
     {
       "epoch": 2.950772200772201,
-      "grad_norm": 4.513665199279785,
+      "grad_norm": 4.627216339111328,
       "learning_rate": 1.4126069270120613e-08,
-      "loss": 0.7685,
+      "loss": 0.5284,
       "step": 3057
     },
     {
       "epoch": 2.9517374517374515,
-      "grad_norm": 4.776824474334717,
+      "grad_norm": 4.8601789474487305,
       "learning_rate": 1.3577660948249637e-08,
-      "loss": 0.7659,
+      "loss": 0.5246,
       "step": 3058
     },
     {
       "epoch": 2.9527027027027026,
-      "grad_norm": 4.89701509475708,
+      "grad_norm": 4.986196041107178,
       "learning_rate": 1.3040102461514548e-08,
-      "loss": 0.7627,
+      "loss": 0.535,
       "step": 3059
     },
     {
       "epoch": 2.9536679536679538,
-      "grad_norm": 4.525649070739746,
+      "grad_norm": 4.6612653732299805,
       "learning_rate": 1.2513394393950384e-08,
-      "loss": 0.7146,
+      "loss": 0.502,
       "step": 3060
     },
     {
       "epoch": 2.954633204633205,
-      "grad_norm": 4.8755316734313965,
+      "grad_norm": 5.21066951751709,
       "learning_rate": 1.1997537317804953e-08,
-      "loss": 0.7616,
+      "loss": 0.5279,
       "step": 3061
     },
     {
       "epoch": 2.9555984555984556,
-      "grad_norm": 4.713533878326416,
+      "grad_norm": 4.365143299102783,
       "learning_rate": 1.1492531793534378e-08,
-      "loss": 0.6995,
+      "loss": 0.4601,
       "step": 3062
     },
     {
       "epoch": 2.9565637065637067,
-      "grad_norm": 4.9669342041015625,
+      "grad_norm": 4.665617942810059,
       "learning_rate": 1.0998378369806439e-08,
-      "loss": 0.7321,
+      "loss": 0.5052,
       "step": 3063
     },
     {
       "epoch": 2.9575289575289574,
-      "grad_norm": 4.8890180587768555,
+      "grad_norm": 4.648453712463379,
       "learning_rate": 1.0515077583498346e-08,
-      "loss": 0.7737,
+      "loss": 0.5266,
       "step": 3064
     },
     {
       "epoch": 2.9584942084942085,
-      "grad_norm": 5.029232501983643,
+      "grad_norm": 5.01582145690918,
       "learning_rate": 1.0042629959696737e-08,
-      "loss": 0.7617,
+      "loss": 0.5109,
       "step": 3065
     },
     {
       "epoch": 2.9594594594594597,
-      "grad_norm": 4.720171928405762,
+      "grad_norm": 4.600343704223633,
       "learning_rate": 9.581036011696577e-09,
-      "loss": 0.7643,
+      "loss": 0.5279,
       "step": 3066
     },
     {
       "epoch": 2.9604247104247103,
-      "grad_norm": 4.735867977142334,
+      "grad_norm": 4.735622882843018,
       "learning_rate": 9.130296241001146e-09,
-      "loss": 0.7611,
+      "loss": 0.5251,
       "step": 3067
     },
     {
       "epoch": 2.9613899613899615,
-      "grad_norm": 4.768474578857422,
+      "grad_norm": 4.944203853607178,
       "learning_rate": 8.690411137318722e-09,
-      "loss": 0.7638,
+      "loss": 0.5349,
       "step": 3068
     },
     {
       "epoch": 2.962355212355212,
-      "grad_norm": 5.289783954620361,
+      "grad_norm": 5.110785961151123,
       "learning_rate": 8.261381178568117e-09,
-      "loss": 0.7812,
+      "loss": 0.5415,
       "step": 3069
     },
     {
       "epoch": 2.9633204633204633,
-      "grad_norm": 4.914121150970459,
+      "grad_norm": 4.999091625213623,
       "learning_rate": 7.84320683087203e-09,
-      "loss": 0.8026,
+      "loss": 0.5667,
       "step": 3070
     },
     {
       "epoch": 2.9642857142857144,
-      "grad_norm": 5.514883995056152,
+      "grad_norm": 5.431160926818848,
       "learning_rate": 7.435888548559256e-09,
-      "loss": 0.7871,
+      "loss": 0.5404,
       "step": 3071
     },
     {
       "epoch": 2.965250965250965,
-      "grad_norm": 5.027594089508057,
+      "grad_norm": 5.32690954208374,
       "learning_rate": 7.0394267741646925e-09,
-      "loss": 0.8206,
+      "loss": 0.5807,
       "step": 3072
     },
     {
       "epoch": 2.9662162162162162,
-      "grad_norm": 5.023658752441406,
+      "grad_norm": 5.04693078994751,
       "learning_rate": 6.653821938427119e-09,
-      "loss": 0.7488,
+      "loss": 0.5301,
       "step": 3073
     },
     {
       "epoch": 2.967181467181467,
-      "grad_norm": 5.08175802230835,
+      "grad_norm": 5.046180248260498,
       "learning_rate": 6.279074460291412e-09,
-      "loss": 0.7689,
+      "loss": 0.5222,
       "step": 3074
     },
     {
       "epoch": 2.968146718146718,
-      "grad_norm": 5.168396472930908,
+      "grad_norm": 5.3946685791015625,
       "learning_rate": 5.9151847469041125e-09,
-      "loss": 0.7105,
+      "loss": 0.5065,
       "step": 3075
     },
     {
       "epoch": 2.969111969111969,
-      "grad_norm": 4.691485404968262,
+      "grad_norm": 4.633338451385498,
       "learning_rate": 5.562153193616749e-09,
-      "loss": 0.732,
+      "loss": 0.5042,
       "step": 3076
     },
     {
       "epoch": 2.9700772200772203,
-      "grad_norm": 4.773744106292725,
+      "grad_norm": 5.00629186630249,
       "learning_rate": 5.2199801839836195e-09,
-      "loss": 0.7813,
+      "loss": 0.5605,
       "step": 3077
     },
     {
       "epoch": 2.971042471042471,
-      "grad_norm": 4.5500617027282715,
+      "grad_norm": 5.058832168579102,
       "learning_rate": 4.8886660897617955e-09,
-      "loss": 0.8096,
+      "loss": 0.5728,
       "step": 3078
     },
     {
       "epoch": 2.972007722007722,
-      "grad_norm": 4.949380874633789,
+      "grad_norm": 4.761467456817627,
       "learning_rate": 4.568211270910005e-09,
-      "loss": 0.7535,
+      "loss": 0.5301,
       "step": 3079
     },
     {
       "epoch": 2.972972972972973,
-      "grad_norm": 4.675527572631836,
+      "grad_norm": 4.579498291015625,
       "learning_rate": 4.25861607558864e-09,
-      "loss": 0.7593,
+      "loss": 0.5156,
       "step": 3080
     },
     {
       "epoch": 2.973938223938224,
-      "grad_norm": 5.3298773765563965,
+      "grad_norm": 5.203444004058838,
       "learning_rate": 3.9598808401597466e-09,
-      "loss": 0.808,
+      "loss": 0.5658,
       "step": 3081
     },
     {
       "epoch": 2.974903474903475,
-      "grad_norm": 4.755771160125732,
+      "grad_norm": 4.986360549926758,
       "learning_rate": 3.6720058891892564e-09,
-      "loss": 0.7844,
+      "loss": 0.5575,
       "step": 3082
     },
     {
       "epoch": 2.9758687258687258,
-      "grad_norm": 4.750619411468506,
+      "grad_norm": 4.718313694000244,
       "learning_rate": 3.3949915354380968e-09,
-      "loss": 0.754,
+      "loss": 0.5172,
       "step": 3083
     },
     {
       "epoch": 2.976833976833977,
-      "grad_norm": 4.936907768249512,
+      "grad_norm": 4.813381671905518,
       "learning_rate": 3.1288380798744078e-09,
-      "loss": 0.7575,
+      "loss": 0.5313,
       "step": 3084
     },
     {
       "epoch": 2.9777992277992276,
-      "grad_norm": 4.750420093536377,
+      "grad_norm": 4.974298000335693,
       "learning_rate": 2.8735458116602164e-09,
-      "loss": 0.7027,
+      "loss": 0.5037,
       "step": 3085
     },
     {
       "epoch": 2.9787644787644787,
-      "grad_norm": 5.266950607299805,
+      "grad_norm": 5.3536057472229,
       "learning_rate": 2.629115008160321e-09,
-      "loss": 0.7765,
+      "loss": 0.5084,
       "step": 3086
     },
     {
       "epoch": 2.97972972972973,
-      "grad_norm": 5.024481773376465,
+      "grad_norm": 5.087066650390625,
       "learning_rate": 2.395545934940069e-09,
-      "loss": 0.7972,
+      "loss": 0.5351,
       "step": 3087
     },
     {
       "epoch": 2.980694980694981,
-      "grad_norm": 5.134521007537842,
+      "grad_norm": 5.070369720458984,
       "learning_rate": 2.1728388457620263e-09,
-      "loss": 0.76,
+      "loss": 0.5252,
       "step": 3088
     },
     {
       "epoch": 2.9816602316602316,
-      "grad_norm": 5.2135138511657715,
+      "grad_norm": 4.965612888336182,
       "learning_rate": 1.9609939825881996e-09,
-      "loss": 0.731,
+      "loss": 0.512,
       "step": 3089
     },
     {
       "epoch": 2.9826254826254828,
-      "grad_norm": 4.94840669631958,
+      "grad_norm": 5.179541110992432,
       "learning_rate": 1.7600115755789237e-09,
-      "loss": 0.757,
+      "loss": 0.5336,
       "step": 3090
     },
     {
       "epoch": 2.9835907335907335,
-      "grad_norm": 5.121521472930908,
+      "grad_norm": 4.8364973068237305,
       "learning_rate": 1.5698918430928634e-09,
-      "loss": 0.7751,
+      "loss": 0.5093,
       "step": 3091
     },
     {
       "epoch": 2.9845559845559846,
-      "grad_norm": 5.261178016662598,
+      "grad_norm": 5.100462913513184,
       "learning_rate": 1.3906349916881223e-09,
-      "loss": 0.7607,
+      "loss": 0.5167,
       "step": 3092
     },
     {
       "epoch": 2.9855212355212357,
-      "grad_norm": 4.680968284606934,
+      "grad_norm": 4.575159072875977,
       "learning_rate": 1.2222412161200237e-09,
-      "loss": 0.7714,
+      "loss": 0.5213,
       "step": 3093
     },
     {
       "epoch": 2.9864864864864864,
-      "grad_norm": 5.011970520019531,
+      "grad_norm": 4.8422322273254395,
       "learning_rate": 1.0647106993411094e-09,
-      "loss": 0.7493,
+      "loss": 0.5032,
       "step": 3094
     },
     {
       "epoch": 2.9874517374517375,
-      "grad_norm": 4.68505334854126,
+      "grad_norm": 4.758364677429199,
       "learning_rate": 9.180436125011405e-10,
-      "loss": 0.7664,
+      "loss": 0.5292,
       "step": 3095
     },
     {
       "epoch": 2.988416988416988,
-      "grad_norm": 5.225539684295654,
+      "grad_norm": 4.876880168914795,
       "learning_rate": 7.822401149482073e-10,
-      "loss": 0.7783,
+      "loss": 0.5292,
       "step": 3096
     },
     {
       "epoch": 2.9893822393822393,
-      "grad_norm": 4.609559535980225,
+      "grad_norm": 4.759739398956299,
       "learning_rate": 6.573003542276191e-10,
-      "loss": 0.7553,
+      "loss": 0.5189,
       "step": 3097
     },
     {
       "epoch": 2.9903474903474905,
-      "grad_norm": 4.871151924133301,
+      "grad_norm": 4.953545093536377,
       "learning_rate": 5.432244660819041e-10,
-      "loss": 0.7406,
+      "loss": 0.5122,
       "step": 3098
     },
     {
       "epoch": 2.991312741312741,
-      "grad_norm": 4.843747615814209,
+      "grad_norm": 4.62905740737915,
       "learning_rate": 4.4001257444747924e-10,
-      "loss": 0.823,
+      "loss": 0.5702,
       "step": 3099
     },
     {
       "epoch": 2.9922779922779923,
-      "grad_norm": 4.668362140655518,
+      "grad_norm": 4.5414934158325195,
       "learning_rate": 3.4766479146242095e-10,
-      "loss": 0.7192,
+      "loss": 0.4941,
       "step": 3100
     },
     {
       "epoch": 2.993243243243243,
-      "grad_norm": 5.2617316246032715,
+      "grad_norm": 5.267070770263672,
       "learning_rate": 2.66181217457584e-10,
-      "loss": 0.7802,
+      "loss": 0.5238,
       "step": 3101
     },
     {
       "epoch": 2.994208494208494,
-      "grad_norm": 5.22498083114624,
+      "grad_norm": 5.189320087432861,
       "learning_rate": 1.955619409610421e-10,
-      "loss": 0.7905,
+      "loss": 0.5513,
       "step": 3102
     },
     {
       "epoch": 2.9951737451737452,
-      "grad_norm": 5.287078857421875,
+      "grad_norm": 5.146084785461426,
       "learning_rate": 1.358070386980881e-10,
-      "loss": 0.7469,
+      "loss": 0.5179,
       "step": 3103
     },
     {
       "epoch": 2.9961389961389964,
-      "grad_norm": 4.943191051483154,
+      "grad_norm": 5.037194728851318,
       "learning_rate": 8.691657559012356e-11,
-      "loss": 0.8264,
+      "loss": 0.5697,
       "step": 3104
     },
     {
       "epoch": 2.997104247104247,
-      "grad_norm": 5.016753673553467,
+      "grad_norm": 5.0243401527404785,
       "learning_rate": 4.88906047546589e-11,
-      "loss": 0.7621,
+      "loss": 0.5293,
       "step": 3105
     },
     {
       "epoch": 2.998069498069498,
-      "grad_norm": 4.927329063415527,
+      "grad_norm": 4.676773548126221,
       "learning_rate": 2.1729167505313288e-11,
-      "loss": 0.702,
+      "loss": 0.4941,
       "step": 3106
     },
     {
       "epoch": 2.999034749034749,
-      "grad_norm": 4.855465412139893,
+      "grad_norm": 4.810445785522461,
       "learning_rate": 5.432293351814721e-12,
-      "loss": 0.7606,
+      "loss": 0.5043,
       "step": 3107
     },
     {
       "epoch": 3.0,
-      "grad_norm": 4.297211170196533,
+      "grad_norm": 4.053826332092285,
       "learning_rate": 0.0,
-      "loss": 0.6057,
+      "loss": 0.3939,
       "step": 3108
     },
     {
       "epoch": 3.0,
       "step": 3108,
-      "total_flos": 8.955621288433418e+17,
-      "train_loss": 0.6617098891558642,
-      "train_runtime": 8230.5312,
-      "train_samples_per_second": 36.226,
-      "train_steps_per_second": 0.378
+      "total_flos": 8.955962624080609e+17,
+      "train_loss": 1.1321901073115979,
+      "train_runtime": 12345.8235,
+      "train_samples_per_second": 24.151,
+      "train_steps_per_second": 0.252
     }
   ],
   "logging_steps": 1.0,
   "max_steps": 3108,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
-  "save_steps": 1000,
-  "total_flos": 8.955621288433418e+17,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.955962624080609e+17,
   "train_batch_size": 24,
   "trial_name": null,
   "trial_params": null