diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -3,4910 +3,5141 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 3488,
+  "global_step": 3653,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0014334862385321102,
-      "grad_norm": 10.930955258982435,
-      "learning_rate": 2.865329512893983e-07,
-      "loss": 1.1122,
+      "epoch": 0.001368738023542294,
+      "grad_norm": 1.1110930744661907,
+      "learning_rate": 2.73224043715847e-07,
+      "loss": 1.71,
       "step": 5
     },
     {
-      "epoch": 0.0028669724770642203,
-      "grad_norm": 7.957126318923027,
-      "learning_rate": 5.730659025787966e-07,
-      "loss": 1.1033,
+      "epoch": 0.002737476047084588,
+      "grad_norm": 1.0450218593816867,
+      "learning_rate": 5.46448087431694e-07,
+      "loss": 1.7206,
       "step": 10
     },
     {
-      "epoch": 0.0043004587155963305,
-      "grad_norm": 6.252488126522345,
-      "learning_rate": 8.595988538681949e-07,
-      "loss": 1.1448,
+      "epoch": 0.004106214070626882,
+      "grad_norm": 1.2104539142499893,
+      "learning_rate": 8.196721311475409e-07,
+      "loss": 1.8131,
       "step": 15
     },
     {
-      "epoch": 0.005733944954128441,
-      "grad_norm": 6.589440813762702,
-      "learning_rate": 1.1461318051575932e-06,
-      "loss": 1.0284,
+      "epoch": 0.005474952094169176,
+      "grad_norm": 1.1014228621000113,
+      "learning_rate": 1.092896174863388e-06,
+      "loss": 1.7571,
       "step": 20
     },
     {
-      "epoch": 0.007167431192660551,
-      "grad_norm": 4.917441825534674,
-      "learning_rate": 1.4326647564469915e-06,
-      "loss": 1.0131,
+      "epoch": 0.00684369011771147,
+      "grad_norm": 1.0213348170440788,
+      "learning_rate": 1.3661202185792352e-06,
+      "loss": 1.7504,
       "step": 25
     },
     {
-      "epoch": 0.008600917431192661,
-      "grad_norm": 4.611821748972413,
-      "learning_rate": 1.7191977077363897e-06,
-      "loss": 1.0492,
+      "epoch": 0.008212428141253765,
+      "grad_norm": 0.907763844914787,
+      "learning_rate": 1.6393442622950819e-06,
+      "loss": 1.726,
       "step": 30
     },
     {
-      "epoch": 0.010034403669724771,
-      "grad_norm": 4.602326061969326,
-      "learning_rate": 2.005730659025788e-06,
-      "loss": 1.0249,
+      "epoch": 0.009581166164796057,
+      "grad_norm": 0.9051570945453739,
+      "learning_rate": 1.912568306010929e-06,
+      "loss": 1.7158,
       "step": 35
     },
     {
-      "epoch": 0.011467889908256881,
-      "grad_norm": 4.795975437768488,
-      "learning_rate": 2.2922636103151864e-06,
-      "loss": 0.9889,
+      "epoch": 0.010949904188338352,
+      "grad_norm": 0.7396508509648158,
+      "learning_rate": 2.185792349726776e-06,
+      "loss": 1.7692,
       "step": 40
     },
     {
-      "epoch": 0.012901376146788992,
-      "grad_norm": 4.881452695659322,
-      "learning_rate": 2.5787965616045845e-06,
-      "loss": 0.9654,
+      "epoch": 0.012318642211880646,
+      "grad_norm": 0.6316750205627214,
+      "learning_rate": 2.459016393442623e-06,
+      "loss": 1.7118,
       "step": 45
     },
     {
-      "epoch": 0.014334862385321102,
-      "grad_norm": 4.816747748013977,
-      "learning_rate": 2.865329512893983e-06,
-      "loss": 1.0479,
+      "epoch": 0.01368738023542294,
+      "grad_norm": 0.6071126407750962,
+      "learning_rate": 2.7322404371584705e-06,
+      "loss": 1.6533,
       "step": 50
     },
     {
-      "epoch": 0.01576834862385321,
-      "grad_norm": 4.672562758642394,
-      "learning_rate": 3.151862464183381e-06,
-      "loss": 0.9692,
+      "epoch": 0.015056118258965233,
+      "grad_norm": 0.6129917793535419,
+      "learning_rate": 3.0054644808743173e-06,
+      "loss": 1.6471,
       "step": 55
     },
     {
-      "epoch": 0.017201834862385322,
-      "grad_norm": 4.665632444621185,
-      "learning_rate": 3.4383954154727795e-06,
-      "loss": 1.0033,
+      "epoch": 0.01642485628250753,
+      "grad_norm": 0.6342687394025227,
+      "learning_rate": 3.2786885245901638e-06,
+      "loss": 1.6505,
       "step": 60
     },
     {
-      "epoch": 0.01863532110091743,
-      "grad_norm": 4.5487339829077795,
-      "learning_rate": 3.724928366762178e-06,
-      "loss": 1.0206,
+      "epoch": 0.017793594306049824,
+      "grad_norm": 0.5732154266113143,
+      "learning_rate": 3.551912568306011e-06,
+      "loss": 1.6261,
       "step": 65
     },
     {
-      "epoch": 0.020068807339449542,
-      "grad_norm": 4.72120821418095,
-      "learning_rate": 4.011461318051576e-06,
-      "loss": 1.0065,
+      "epoch": 0.019162332329592115,
+      "grad_norm": 0.5200300967595174,
+      "learning_rate": 3.825136612021858e-06,
+      "loss": 1.5861,
       "step": 70
     },
     {
-      "epoch": 0.02150229357798165,
-      "grad_norm": 4.488698654642803,
-      "learning_rate": 4.2979942693409744e-06,
-      "loss": 0.9634,
+      "epoch": 0.02053107035313441,
+      "grad_norm": 0.5356030837166187,
+      "learning_rate": 4.098360655737705e-06,
+      "loss": 1.6002,
       "step": 75
     },
     {
-      "epoch": 0.022935779816513763,
-      "grad_norm": 4.502540106177661,
-      "learning_rate": 4.584527220630373e-06,
-      "loss": 0.9633,
+      "epoch": 0.021899808376676703,
+      "grad_norm": 0.4893081141758957,
+      "learning_rate": 4.371584699453552e-06,
+      "loss": 1.5388,
       "step": 80
     },
     {
-      "epoch": 0.02436926605504587,
-      "grad_norm": 4.538891592130377,
-      "learning_rate": 4.871060171919771e-06,
-      "loss": 0.9148,
+      "epoch": 0.023268546400218998,
+      "grad_norm": 0.5042585034665032,
+      "learning_rate": 4.6448087431694e-06,
+      "loss": 1.5788,
       "step": 85
     },
     {
-      "epoch": 0.025802752293577983,
-      "grad_norm": 4.89489963864345,
-      "learning_rate": 5.157593123209169e-06,
-      "loss": 1.0303,
+      "epoch": 0.024637284423761292,
+      "grad_norm": 0.4756627065397568,
+      "learning_rate": 4.918032786885246e-06,
+      "loss": 1.5832,
       "step": 90
     },
     {
-      "epoch": 0.02723623853211009,
-      "grad_norm": 4.546009685259081,
-      "learning_rate": 5.444126074498568e-06,
-      "loss": 1.0174,
+      "epoch": 0.026006022447303587,
+      "grad_norm": 0.4712228645477175,
+      "learning_rate": 5.191256830601094e-06,
+      "loss": 1.5221,
       "step": 95
     },
     {
-      "epoch": 0.028669724770642203,
-      "grad_norm": 4.542941293141969,
-      "learning_rate": 5.730659025787966e-06,
-      "loss": 1.0251,
+      "epoch": 0.02737476047084588,
+      "grad_norm": 0.47712827241082845,
+      "learning_rate": 5.464480874316941e-06,
+      "loss": 1.5554,
       "step": 100
     },
     {
-      "epoch": 0.030103211009174312,
-      "grad_norm": 4.571406282018777,
-      "learning_rate": 6.017191977077364e-06,
-      "loss": 1.0376,
+      "epoch": 0.028743498494388176,
+      "grad_norm": 0.4747245775681728,
+      "learning_rate": 5.737704918032787e-06,
+      "loss": 1.5051,
       "step": 105
     },
     {
-      "epoch": 0.03153669724770642,
-      "grad_norm": 4.75986045778734,
-      "learning_rate": 6.303724928366762e-06,
-      "loss": 0.9827,
+      "epoch": 0.030112236517930466,
+      "grad_norm": 0.4619152125132191,
+      "learning_rate": 6.010928961748635e-06,
+      "loss": 1.5731,
       "step": 110
     },
     {
-      "epoch": 0.03297018348623853,
-      "grad_norm": 4.393838501543807,
-      "learning_rate": 6.590257879656161e-06,
-      "loss": 0.9533,
+      "epoch": 0.031480974541472764,
+      "grad_norm": 0.46656513563983254,
+      "learning_rate": 6.284153005464482e-06,
+      "loss": 1.5719,
       "step": 115
     },
     {
-      "epoch": 0.034403669724770644,
-      "grad_norm": 4.780066323607276,
-      "learning_rate": 6.876790830945559e-06,
-      "loss": 1.0018,
+      "epoch": 0.03284971256501506,
+      "grad_norm": 0.45441304688271805,
+      "learning_rate": 6.5573770491803276e-06,
+      "loss": 1.5359,
       "step": 120
     },
     {
-      "epoch": 0.03583715596330275,
-      "grad_norm": 4.414441537089264,
-      "learning_rate": 7.163323782234957e-06,
-      "loss": 1.0135,
+      "epoch": 0.03421845058855735,
+      "grad_norm": 0.4492251581811985,
+      "learning_rate": 6.830601092896175e-06,
+      "loss": 1.5254,
       "step": 125
     },
     {
-      "epoch": 0.03727064220183486,
-      "grad_norm": 4.301618750664585,
-      "learning_rate": 7.449856733524356e-06,
-      "loss": 1.0201,
+      "epoch": 0.03558718861209965,
+      "grad_norm": 0.4901256909260675,
+      "learning_rate": 7.103825136612022e-06,
+      "loss": 1.5531,
       "step": 130
     },
     {
-      "epoch": 0.03870412844036697,
-      "grad_norm": 4.3543481672910245,
-      "learning_rate": 7.736389684813753e-06,
-      "loss": 1.0236,
+      "epoch": 0.036955926635641935,
+      "grad_norm": 0.4521871797566739,
+      "learning_rate": 7.3770491803278695e-06,
+      "loss": 1.5203,
       "step": 135
     },
     {
-      "epoch": 0.040137614678899085,
-      "grad_norm": 4.31887452196869,
-      "learning_rate": 8.022922636103152e-06,
-      "loss": 1.0413,
+      "epoch": 0.03832466465918423,
+      "grad_norm": 0.46301007561132373,
+      "learning_rate": 7.650273224043716e-06,
+      "loss": 1.5597,
       "step": 140
     },
     {
-      "epoch": 0.04157110091743119,
-      "grad_norm": 4.19580797580846,
-      "learning_rate": 8.30945558739255e-06,
-      "loss": 0.9999,
+      "epoch": 0.039693402682726524,
+      "grad_norm": 0.4443455697452246,
+      "learning_rate": 7.923497267759564e-06,
+      "loss": 1.4751,
       "step": 145
     },
     {
-      "epoch": 0.0430045871559633,
-      "grad_norm": 4.2941554889344635,
-      "learning_rate": 8.595988538681949e-06,
-      "loss": 1.0824,
+      "epoch": 0.04106214070626882,
+      "grad_norm": 0.4529255518299648,
+      "learning_rate": 8.19672131147541e-06,
+      "loss": 1.5645,
       "step": 150
     },
     {
-      "epoch": 0.04443807339449541,
-      "grad_norm": 4.430699539319586,
-      "learning_rate": 8.882521489971347e-06,
-      "loss": 1.0545,
+      "epoch": 0.04243087872981111,
+      "grad_norm": 0.45575204993786755,
+      "learning_rate": 8.469945355191259e-06,
+      "loss": 1.5297,
       "step": 155
     },
     {
-      "epoch": 0.045871559633027525,
-      "grad_norm": 4.357465541547218,
-      "learning_rate": 9.169054441260746e-06,
-      "loss": 1.0662,
+      "epoch": 0.04379961675335341,
+      "grad_norm": 0.4619195990067295,
+      "learning_rate": 8.743169398907103e-06,
+      "loss": 1.4945,
       "step": 160
     },
     {
-      "epoch": 0.047305045871559634,
-      "grad_norm": 5.722224893724134,
-      "learning_rate": 9.455587392550144e-06,
-      "loss": 1.0197,
+      "epoch": 0.0451683547768957,
+      "grad_norm": 0.43620516601045234,
+      "learning_rate": 9.016393442622952e-06,
+      "loss": 1.5064,
       "step": 165
     },
     {
-      "epoch": 0.04873853211009174,
-      "grad_norm": 4.374513298911362,
-      "learning_rate": 9.742120343839543e-06,
-      "loss": 1.046,
+      "epoch": 0.046537092800437996,
+      "grad_norm": 0.48507057683395305,
+      "learning_rate": 9.2896174863388e-06,
+      "loss": 1.4873,
       "step": 170
     },
     {
-      "epoch": 0.05017201834862385,
-      "grad_norm": 4.593743802840811,
-      "learning_rate": 1.0028653295128941e-05,
-      "loss": 1.0739,
+      "epoch": 0.04790583082398029,
+      "grad_norm": 0.4591779126577447,
+      "learning_rate": 9.562841530054644e-06,
+      "loss": 1.4989,
       "step": 175
     },
     {
-      "epoch": 0.051605504587155966,
-      "grad_norm": 4.333415176606838,
-      "learning_rate": 1.0315186246418338e-05,
-      "loss": 1.0833,
+      "epoch": 0.049274568847522585,
+      "grad_norm": 0.4477837953889451,
+      "learning_rate": 9.836065573770493e-06,
+      "loss": 1.4755,
       "step": 180
     },
     {
-      "epoch": 0.053038990825688075,
-      "grad_norm": 4.426601251103325,
-      "learning_rate": 1.0601719197707738e-05,
-      "loss": 1.0799,
+      "epoch": 0.05064330687106488,
+      "grad_norm": 0.4580627295319435,
+      "learning_rate": 1.0109289617486339e-05,
+      "loss": 1.4545,
       "step": 185
     },
     {
-      "epoch": 0.05447247706422018,
-      "grad_norm": 4.556659147888879,
-      "learning_rate": 1.0888252148997137e-05,
-      "loss": 1.0405,
+      "epoch": 0.05201204489460717,
+      "grad_norm": 0.45313033315292356,
+      "learning_rate": 1.0382513661202187e-05,
+      "loss": 1.4985,
       "step": 190
     },
     {
-      "epoch": 0.05590596330275229,
-      "grad_norm": 4.3383547238785605,
-      "learning_rate": 1.1174785100286533e-05,
-      "loss": 1.0189,
+      "epoch": 0.05338078291814947,
+      "grad_norm": 0.42708720087605334,
+      "learning_rate": 1.0655737704918034e-05,
+      "loss": 1.4229,
       "step": 195
     },
     {
-      "epoch": 0.05733944954128441,
-      "grad_norm": 4.125983340413087,
-      "learning_rate": 1.1461318051575932e-05,
-      "loss": 1.0592,
+      "epoch": 0.05474952094169176,
+      "grad_norm": 0.4211311153329928,
+      "learning_rate": 1.0928961748633882e-05,
+      "loss": 1.4669,
       "step": 200
     },
     {
-      "epoch": 0.058772935779816515,
-      "grad_norm": 4.479422603616749,
-      "learning_rate": 1.1747851002865332e-05,
-      "loss": 1.0626,
+      "epoch": 0.05611825896523406,
+      "grad_norm": 0.4553582543296074,
+      "learning_rate": 1.1202185792349727e-05,
+      "loss": 1.4444,
       "step": 205
     },
     {
-      "epoch": 0.060206422018348624,
-      "grad_norm": 4.506035260293105,
-      "learning_rate": 1.2034383954154729e-05,
-      "loss": 1.0506,
+      "epoch": 0.05748699698877635,
+      "grad_norm": 0.43888054055275466,
+      "learning_rate": 1.1475409836065575e-05,
+      "loss": 1.4622,
       "step": 210
     },
     {
-      "epoch": 0.06163990825688073,
-      "grad_norm": 5.22525942718489,
-      "learning_rate": 1.2320916905444127e-05,
-      "loss": 1.0855,
+      "epoch": 0.058855735012318645,
+      "grad_norm": 0.43839040680573754,
+      "learning_rate": 1.1748633879781421e-05,
+      "loss": 1.3921,
       "step": 215
     },
     {
-      "epoch": 0.06307339449541284,
-      "grad_norm": 4.779500493076456,
-      "learning_rate": 1.2607449856733524e-05,
-      "loss": 1.0872,
+      "epoch": 0.06022447303586093,
+      "grad_norm": 0.442929256824184,
+      "learning_rate": 1.202185792349727e-05,
+      "loss": 1.4695,
       "step": 220
     },
     {
-      "epoch": 0.06450688073394495,
-      "grad_norm": 4.1419371246845555,
-      "learning_rate": 1.2893982808022924e-05,
-      "loss": 1.0158,
+      "epoch": 0.06159321105940323,
+      "grad_norm": 0.42979267539703714,
+      "learning_rate": 1.2295081967213116e-05,
+      "loss": 1.4444,
       "step": 225
     },
     {
-      "epoch": 0.06594036697247706,
-      "grad_norm": 4.55193247523381,
-      "learning_rate": 1.3180515759312323e-05,
-      "loss": 1.1341,
+      "epoch": 0.06296194908294553,
+      "grad_norm": 0.43942113048062365,
+      "learning_rate": 1.2568306010928964e-05,
+      "loss": 1.4581,
       "step": 230
     },
     {
-      "epoch": 0.06737385321100918,
-      "grad_norm": 4.821922591302919,
-      "learning_rate": 1.346704871060172e-05,
-      "loss": 1.0317,
+      "epoch": 0.06433068710648782,
+      "grad_norm": 0.43105155511527676,
+      "learning_rate": 1.284153005464481e-05,
+      "loss": 1.4335,
       "step": 235
     },
     {
-      "epoch": 0.06880733944954129,
-      "grad_norm": 5.7218678278363075,
-      "learning_rate": 1.3753581661891118e-05,
-      "loss": 1.0929,
+      "epoch": 0.06569942513003012,
+      "grad_norm": 0.4347334195115935,
+      "learning_rate": 1.3114754098360655e-05,
+      "loss": 1.5437,
       "step": 240
     },
     {
-      "epoch": 0.0702408256880734,
-      "grad_norm": 4.5897462960681885,
-      "learning_rate": 1.4040114613180518e-05,
-      "loss": 1.0707,
+      "epoch": 0.0670681631535724,
+      "grad_norm": 0.41558733251657665,
+      "learning_rate": 1.3387978142076503e-05,
+      "loss": 1.4246,
       "step": 245
     },
     {
-      "epoch": 0.0716743119266055,
-      "grad_norm": 4.9044704729529265,
-      "learning_rate": 1.4326647564469915e-05,
-      "loss": 1.0589,
+      "epoch": 0.0684369011771147,
+      "grad_norm": 0.4166574660442,
+      "learning_rate": 1.366120218579235e-05,
+      "loss": 1.4672,
       "step": 250
     },
     {
-      "epoch": 0.07310779816513761,
-      "grad_norm": 4.317942267526155,
-      "learning_rate": 1.4613180515759313e-05,
-      "loss": 1.0944,
+      "epoch": 0.069805639200657,
+      "grad_norm": 0.40343261677226544,
+      "learning_rate": 1.3934426229508198e-05,
+      "loss": 1.4695,
       "step": 255
     },
     {
-      "epoch": 0.07454128440366972,
-      "grad_norm": 4.019443011772793,
-      "learning_rate": 1.4899713467048712e-05,
-      "loss": 1.1014,
+      "epoch": 0.0711743772241993,
+      "grad_norm": 0.41945935612278784,
+      "learning_rate": 1.4207650273224044e-05,
+      "loss": 1.4173,
       "step": 260
     },
     {
-      "epoch": 0.07597477064220183,
-      "grad_norm": 5.05645456273739,
-      "learning_rate": 1.518624641833811e-05,
-      "loss": 1.0855,
+      "epoch": 0.07254311524774158,
+      "grad_norm": 0.4091410585881606,
+      "learning_rate": 1.4480874316939892e-05,
+      "loss": 1.3669,
       "step": 265
     },
     {
-      "epoch": 0.07740825688073394,
-      "grad_norm": 4.631991611940915,
-      "learning_rate": 1.5472779369627507e-05,
-      "loss": 1.1797,
+      "epoch": 0.07391185327128387,
+      "grad_norm": 0.40428264801464686,
+      "learning_rate": 1.4754098360655739e-05,
+      "loss": 1.4712,
       "step": 270
     },
     {
-      "epoch": 0.07884174311926606,
-      "grad_norm": 4.863391448983006,
-      "learning_rate": 1.5759312320916907e-05,
-      "loss": 1.1189,
+      "epoch": 0.07528059129482617,
+      "grad_norm": 0.4074017153449847,
+      "learning_rate": 1.5027322404371585e-05,
+      "loss": 1.4477,
       "step": 275
     },
     {
-      "epoch": 0.08027522935779817,
-      "grad_norm": 5.67171228826149,
-      "learning_rate": 1.6045845272206304e-05,
-      "loss": 1.0687,
+      "epoch": 0.07664932931836846,
+      "grad_norm": 0.41299926058270664,
+      "learning_rate": 1.5300546448087432e-05,
+      "loss": 1.4015,
       "step": 280
     },
     {
-      "epoch": 0.08170871559633028,
-      "grad_norm": 4.863402157719064,
-      "learning_rate": 1.6332378223495704e-05,
-      "loss": 1.1033,
+      "epoch": 0.07801806734191076,
+      "grad_norm": 0.39622800283246956,
+      "learning_rate": 1.5573770491803278e-05,
+      "loss": 1.443,
       "step": 285
     },
     {
-      "epoch": 0.08314220183486239,
-      "grad_norm": 4.660992837053409,
-      "learning_rate": 1.66189111747851e-05,
-      "loss": 1.1543,
+      "epoch": 0.07938680536545305,
+      "grad_norm": 0.4050370810855124,
+      "learning_rate": 1.5846994535519128e-05,
+      "loss": 1.4454,
       "step": 290
     },
     {
-      "epoch": 0.0845756880733945,
-      "grad_norm": 4.552069083735871,
-      "learning_rate": 1.69054441260745e-05,
-      "loss": 1.1464,
+      "epoch": 0.08075554338899535,
+      "grad_norm": 0.405591352468174,
+      "learning_rate": 1.6120218579234975e-05,
+      "loss": 1.4077,
       "step": 295
     },
     {
-      "epoch": 0.0860091743119266,
-      "grad_norm": 4.61726981167071,
-      "learning_rate": 1.7191977077363898e-05,
-      "loss": 1.1332,
+      "epoch": 0.08212428141253764,
+      "grad_norm": 0.4188674346008248,
+      "learning_rate": 1.639344262295082e-05,
+      "loss": 1.4503,
       "step": 300
     },
     {
-      "epoch": 0.08744266055045871,
-      "grad_norm": 4.9734394128477755,
-      "learning_rate": 1.7478510028653298e-05,
-      "loss": 1.0608,
+      "epoch": 0.08349301943607994,
+      "grad_norm": 0.4015518431467613,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.4027,
       "step": 305
     },
     {
-      "epoch": 0.08887614678899082,
-      "grad_norm": 4.080774915213187,
-      "learning_rate": 1.7765042979942695e-05,
-      "loss": 1.1737,
+      "epoch": 0.08486175745962223,
+      "grad_norm": 0.40825550305371383,
+      "learning_rate": 1.6939890710382517e-05,
+      "loss": 1.4021,
       "step": 310
     },
     {
-      "epoch": 0.09030963302752294,
-      "grad_norm": 4.9010757003007805,
-      "learning_rate": 1.805157593123209e-05,
-      "loss": 1.152,
+      "epoch": 0.08623049548316453,
+      "grad_norm": 0.4103687720629476,
+      "learning_rate": 1.721311475409836e-05,
+      "loss": 1.4304,
       "step": 315
     },
     {
-      "epoch": 0.09174311926605505,
-      "grad_norm": 4.348268075612745,
-      "learning_rate": 1.833810888252149e-05,
-      "loss": 1.1493,
+      "epoch": 0.08759923350670681,
+      "grad_norm": 0.4048205442390437,
+      "learning_rate": 1.7486338797814207e-05,
+      "loss": 1.4,
       "step": 320
     },
     {
-      "epoch": 0.09317660550458716,
-      "grad_norm": 5.525129625035676,
-      "learning_rate": 1.8624641833810892e-05,
-      "loss": 1.109,
+      "epoch": 0.08896797153024912,
+      "grad_norm": 0.41661892344943224,
+      "learning_rate": 1.7759562841530057e-05,
+      "loss": 1.4035,
       "step": 325
     },
     {
-      "epoch": 0.09461009174311927,
-      "grad_norm": 6.128793939259254,
-      "learning_rate": 1.891117478510029e-05,
-      "loss": 1.1883,
+      "epoch": 0.0903367095537914,
+      "grad_norm": 0.4308332237873975,
+      "learning_rate": 1.8032786885245903e-05,
+      "loss": 1.4302,
       "step": 330
     },
     {
-      "epoch": 0.09604357798165138,
-      "grad_norm": 4.553408862504849,
-      "learning_rate": 1.9197707736389685e-05,
-      "loss": 1.1294,
+      "epoch": 0.0917054475773337,
+      "grad_norm": 0.41012677210314796,
+      "learning_rate": 1.830601092896175e-05,
+      "loss": 1.3969,
       "step": 335
     },
     {
-      "epoch": 0.09747706422018348,
-      "grad_norm": 4.512129266800345,
-      "learning_rate": 1.9484240687679085e-05,
-      "loss": 1.2076,
+      "epoch": 0.09307418560087599,
+      "grad_norm": 0.40743633096989823,
+      "learning_rate": 1.85792349726776e-05,
+      "loss": 1.4425,
       "step": 340
     },
     {
-      "epoch": 0.0989105504587156,
-      "grad_norm": 4.294698339999475,
-      "learning_rate": 1.9770773638968482e-05,
-      "loss": 1.1455,
+      "epoch": 0.09444292362441829,
+      "grad_norm": 0.39899819661715447,
+      "learning_rate": 1.8852459016393446e-05,
+      "loss": 1.4303,
       "step": 345
     },
     {
-      "epoch": 0.1003440366972477,
-      "grad_norm": 4.649384755228939,
-      "learning_rate": 1.999999499173752e-05,
-      "loss": 1.142,
+      "epoch": 0.09581166164796058,
+      "grad_norm": 0.41870768132807307,
+      "learning_rate": 1.912568306010929e-05,
+      "loss": 1.4002,
       "step": 350
     },
     {
-      "epoch": 0.10177752293577981,
-      "grad_norm": 4.639825467109486,
-      "learning_rate": 1.999981970307739e-05,
-      "loss": 1.19,
+      "epoch": 0.09718039967150287,
+      "grad_norm": 0.40412869941796903,
+      "learning_rate": 1.939890710382514e-05,
+      "loss": 1.403,
       "step": 355
     },
     {
-      "epoch": 0.10321100917431193,
-      "grad_norm": 4.447236882707046,
-      "learning_rate": 1.999939400630968e-05,
-      "loss": 1.1753,
+      "epoch": 0.09854913769504517,
+      "grad_norm": 0.40478167238457397,
+      "learning_rate": 1.9672131147540985e-05,
+      "loss": 1.4227,
       "step": 360
     },
     {
-      "epoch": 0.10464449541284404,
-      "grad_norm": 4.092068716884874,
-      "learning_rate": 1.999871791209438e-05,
-      "loss": 1.1893,
+      "epoch": 0.09991787571858746,
+      "grad_norm": 0.41341003482817107,
+      "learning_rate": 1.994535519125683e-05,
+      "loss": 1.4302,
       "step": 365
     },
     {
-      "epoch": 0.10607798165137615,
-      "grad_norm": 4.398948638445328,
-      "learning_rate": 1.9997791437361734e-05,
-      "loss": 1.1555,
+      "epoch": 0.10128661374212976,
+      "grad_norm": 0.41658269655258945,
+      "learning_rate": 1.999992692147127e-05,
+      "loss": 1.445,
       "step": 370
     },
     {
-      "epoch": 0.10751146788990826,
-      "grad_norm": 4.701129620714216,
-      "learning_rate": 1.9996614605311848e-05,
-      "loss": 1.2011,
+      "epoch": 0.10265535176567205,
+      "grad_norm": 0.40746899077621546,
+      "learning_rate": 1.999963004177886e-05,
+      "loss": 1.458,
       "step": 375
     },
     {
-      "epoch": 0.10894495412844037,
-      "grad_norm": 4.821236633722439,
-      "learning_rate": 1.999518744541407e-05,
-      "loss": 1.1776,
+      "epoch": 0.10402408978921435,
+      "grad_norm": 0.3960893049170255,
+      "learning_rate": 1.9999104800289367e-05,
+      "loss": 1.4382,
       "step": 380
     },
     {
-      "epoch": 0.11037844036697247,
-      "grad_norm": 4.335100849608752,
-      "learning_rate": 1.9993509993406297e-05,
-      "loss": 1.1902,
+      "epoch": 0.10539282781275663,
+      "grad_norm": 0.3950983233516754,
+      "learning_rate": 1.9998351208997734e-05,
+      "loss": 1.4365,
       "step": 385
     },
     {
-      "epoch": 0.11181192660550458,
-      "grad_norm": 4.068825093466267,
-      "learning_rate": 1.9991582291294042e-05,
-      "loss": 1.1836,
+      "epoch": 0.10676156583629894,
+      "grad_norm": 0.38845890924227394,
+      "learning_rate": 1.9997369285113754e-05,
+      "loss": 1.3731,
       "step": 390
     },
     {
-      "epoch": 0.11324541284403669,
-      "grad_norm": 3.9520872546047143,
-      "learning_rate": 1.9989404387349393e-05,
-      "loss": 1.1439,
+      "epoch": 0.10813030385984122,
+      "grad_norm": 0.39526066204949717,
+      "learning_rate": 1.9996159051061638e-05,
+      "loss": 1.4111,
       "step": 395
     },
     {
-      "epoch": 0.11467889908256881,
-      "grad_norm": 4.582077126321336,
-      "learning_rate": 1.998697633610982e-05,
-      "loss": 1.2094,
+      "epoch": 0.10949904188338352,
+      "grad_norm": 0.4133763360642962,
+      "learning_rate": 1.9994720534479543e-05,
+      "loss": 1.4294,
       "step": 400
     },
     {
-      "epoch": 0.11611238532110092,
-      "grad_norm": 4.38687782745242,
-      "learning_rate": 1.998429819837679e-05,
-      "loss": 1.1866,
+      "epoch": 0.11086777990692581,
+      "grad_norm": 0.39671784657884285,
+      "learning_rate": 1.999305376821889e-05,
+      "loss": 1.3708,
       "step": 405
     },
     {
-      "epoch": 0.11754587155963303,
-      "grad_norm": 4.6771014427247275,
-      "learning_rate": 1.998137004121425e-05,
-      "loss": 1.1587,
+      "epoch": 0.11223651793046811,
+      "grad_norm": 0.3847825062400942,
+      "learning_rate": 1.999115879034368e-05,
+      "loss": 1.4638,
       "step": 410
     },
     {
-      "epoch": 0.11897935779816514,
-      "grad_norm": 4.218200613872475,
-      "learning_rate": 1.9978191937946955e-05,
-      "loss": 1.1411,
+      "epoch": 0.1136052559540104,
+      "grad_norm": 0.41459555705408424,
+      "learning_rate": 1.9989035644129553e-05,
+      "loss": 1.4105,
       "step": 415
     },
     {
-      "epoch": 0.12041284403669725,
-      "grad_norm": 4.509609936000089,
-      "learning_rate": 1.9974763968158614e-05,
-      "loss": 1.1772,
+      "epoch": 0.1149739939775527,
+      "grad_norm": 0.4270268157501032,
+      "learning_rate": 1.998668437806286e-05,
+      "loss": 1.4243,
       "step": 420
     },
     {
-      "epoch": 0.12184633027522936,
-      "grad_norm": 5.925174301278395,
-      "learning_rate": 1.9971086217689928e-05,
-      "loss": 1.2411,
+      "epoch": 0.11634273200109499,
+      "grad_norm": 0.42083549100480927,
+      "learning_rate": 1.998410504583952e-05,
+      "loss": 1.4285,
       "step": 425
     },
     {
-      "epoch": 0.12327981651376146,
-      "grad_norm": 4.022377669658578,
-      "learning_rate": 1.9967158778636405e-05,
-      "loss": 1.1887,
+      "epoch": 0.11771147002463729,
+      "grad_norm": 0.3929543512101535,
+      "learning_rate": 1.998129770636381e-05,
+      "loss": 1.3692,
       "step": 430
     },
     {
-      "epoch": 0.12471330275229357,
-      "grad_norm": 4.539584147379229,
-      "learning_rate": 1.996298174934608e-05,
-      "loss": 1.1717,
+      "epoch": 0.11908020804817958,
+      "grad_norm": 0.40163841769109054,
+      "learning_rate": 1.9978262423747003e-05,
+      "loss": 1.3842,
       "step": 435
     },
     {
-      "epoch": 0.12614678899082568,
-      "grad_norm": 9.545718854729863,
-      "learning_rate": 1.9958555234417035e-05,
-      "loss": 1.1879,
+      "epoch": 0.12044894607172187,
+      "grad_norm": 0.40750648380479343,
+      "learning_rate": 1.997499926730593e-05,
+      "loss": 1.4412,
       "step": 440
     },
     {
-      "epoch": 0.1275802752293578,
-      "grad_norm": 22.053598752620175,
-      "learning_rate": 1.995387934469479e-05,
-      "loss": 1.172,
+      "epoch": 0.12181768409526417,
+      "grad_norm": 0.4029670313800167,
+      "learning_rate": 1.9971508311561373e-05,
+      "loss": 1.4481,
       "step": 445
     },
     {
-      "epoch": 0.1290137614678899,
-      "grad_norm": 12.911881630750296,
-      "learning_rate": 1.994895419726953e-05,
-      "loss": 1.2034,
+      "epoch": 0.12318642211880645,
+      "grad_norm": 0.3934576991168861,
+      "learning_rate": 1.996778963623637e-05,
+      "loss": 1.3706,
       "step": 450
     },
     {
-      "epoch": 0.13044724770642202,
-      "grad_norm": 4.856805114224933,
-      "learning_rate": 1.9943779915473165e-05,
-      "loss": 1.1795,
+      "epoch": 0.12455516014234876,
+      "grad_norm": 0.4126991514447529,
+      "learning_rate": 1.9963843326254406e-05,
+      "loss": 1.4096,
       "step": 455
     },
     {
-      "epoch": 0.13188073394495411,
-      "grad_norm": 4.286105662410557,
-      "learning_rate": 1.9938356628876235e-05,
-      "loss": 1.1866,
+      "epoch": 0.12592389816589106,
+      "grad_norm": 0.400896785065692,
+      "learning_rate": 1.9959669471737456e-05,
+      "loss": 1.3284,
       "step": 460
     },
     {
-      "epoch": 0.13331422018348624,
-      "grad_norm": 4.2624287742900995,
-      "learning_rate": 1.9932684473284687e-05,
-      "loss": 1.1894,
+      "epoch": 0.12729263618943334,
+      "grad_norm": 0.4084884002554967,
+      "learning_rate": 1.9955268168003938e-05,
+      "loss": 1.4346,
       "step": 465
     },
     {
-      "epoch": 0.13474770642201836,
-      "grad_norm": 4.716264114725749,
-      "learning_rate": 1.9926763590736457e-05,
-      "loss": 1.2835,
+      "epoch": 0.12866137421297563,
+      "grad_norm": 0.4031311283565323,
+      "learning_rate": 1.9950639515566537e-05,
+      "loss": 1.3632,
       "step": 470
     },
     {
-      "epoch": 0.13618119266055045,
-      "grad_norm": 4.2462330499210115,
-      "learning_rate": 1.992059412949791e-05,
-      "loss": 1.2412,
+      "epoch": 0.13003011223651792,
+      "grad_norm": 0.4017916786924823,
+      "learning_rate": 1.99457836201299e-05,
+      "loss": 1.4083,
       "step": 475
     },
     {
-      "epoch": 0.13761467889908258,
-      "grad_norm": 4.391420347453553,
-      "learning_rate": 1.9914176244060158e-05,
-      "loss": 1.1516,
+      "epoch": 0.13139885026006023,
+      "grad_norm": 0.38801834482390074,
+      "learning_rate": 1.9940700592588228e-05,
+      "loss": 1.3823,
       "step": 480
     },
     {
-      "epoch": 0.13904816513761467,
-      "grad_norm": 75.66535297260735,
-      "learning_rate": 1.9907510095135142e-05,
-      "loss": 1.2722,
+      "epoch": 0.13276758828360252,
+      "grad_norm": 0.40725954819087723,
+      "learning_rate": 1.993539054902275e-05,
+      "loss": 1.369,
       "step": 485
     },
     {
-      "epoch": 0.1404816513761468,
-      "grad_norm": 5.310902667535963,
-      "learning_rate": 1.9900595849651645e-05,
-      "loss": 1.2421,
+      "epoch": 0.1341363263071448,
+      "grad_norm": 0.42191731863948084,
+      "learning_rate": 1.992985361069906e-05,
+      "loss": 1.3907,
       "step": 490
     },
     {
-      "epoch": 0.1419151376146789,
-      "grad_norm": 5.025394485398511,
-      "learning_rate": 1.9893433680751105e-05,
-      "loss": 1.2633,
+      "epoch": 0.1355050643306871,
+      "grad_norm": 0.3981980013765845,
+      "learning_rate": 1.9924089904064354e-05,
+      "loss": 1.4253,
       "step": 495
     },
     {
-      "epoch": 0.143348623853211,
-      "grad_norm": 5.520994924491881,
-      "learning_rate": 1.988602376778327e-05,
-      "loss": 1.189,
+      "epoch": 0.1368738023542294,
+      "grad_norm": 0.3968358847514907,
+      "learning_rate": 1.9918099560744545e-05,
+      "loss": 1.4478,
       "step": 500
     },
     {
-      "epoch": 0.14478211009174313,
-      "grad_norm": 4.515316969142533,
-      "learning_rate": 1.9878366296301713e-05,
-      "loss": 1.1607,
+      "epoch": 0.1382425403777717,
+      "grad_norm": 0.4020407861005804,
+      "learning_rate": 1.991188271754125e-05,
+      "loss": 1.4063,
       "step": 505
     },
     {
-      "epoch": 0.14621559633027523,
-      "grad_norm": 4.532857835025907,
-      "learning_rate": 1.9870461458059188e-05,
-      "loss": 1.2315,
+      "epoch": 0.139611278401314,
+      "grad_norm": 0.3983008409093381,
+      "learning_rate": 1.990543951642866e-05,
+      "loss": 1.3838,
       "step": 510
     },
     {
-      "epoch": 0.14764908256880735,
-      "grad_norm": 4.001206919662932,
-      "learning_rate": 1.9862309451002827e-05,
-      "loss": 1.1659,
+      "epoch": 0.14098001642485627,
+      "grad_norm": 0.38772247745474425,
+      "learning_rate": 1.9898770104550335e-05,
+      "loss": 1.4193,
       "step": 515
     },
     {
-      "epoch": 0.14908256880733944,
-      "grad_norm": 4.137717431705149,
-      "learning_rate": 1.9853910479269165e-05,
-      "loss": 1.1732,
+      "epoch": 0.1423487544483986,
+      "grad_norm": 0.4172192092889838,
+      "learning_rate": 1.9891874634215784e-05,
+      "loss": 1.3831,
       "step": 520
     },
     {
-      "epoch": 0.15051605504587157,
-      "grad_norm": 4.367905503345653,
-      "learning_rate": 1.9845264753179064e-05,
-      "loss": 1.2219,
+      "epoch": 0.14371749247194088,
+      "grad_norm": 0.39316534732966774,
+      "learning_rate": 1.9884753262897042e-05,
+      "loss": 1.3813,
       "step": 525
     },
     {
-      "epoch": 0.15194954128440366,
-      "grad_norm": 3.942691834220351,
-      "learning_rate": 1.9836372489232416e-05,
-      "loss": 1.1208,
+      "epoch": 0.14508623049548316,
+      "grad_norm": 0.40825756476362335,
+      "learning_rate": 1.9877406153225028e-05,
+      "loss": 1.4233,
       "step": 530
     },
     {
-      "epoch": 0.15338302752293578,
-      "grad_norm": 4.028647286087348,
-      "learning_rate": 1.982723391010273e-05,
-      "loss": 1.1499,
+      "epoch": 0.14645496851902545,
+      "grad_norm": 0.4303738354233706,
+      "learning_rate": 1.9869833472985882e-05,
+      "loss": 1.4267,
       "step": 535
     },
     {
-      "epoch": 0.15481651376146788,
-      "grad_norm": 4.262774154684462,
-      "learning_rate": 1.9817849244631575e-05,
-      "loss": 1.2052,
+      "epoch": 0.14782370654256774,
+      "grad_norm": 0.39536263390575654,
+      "learning_rate": 1.9862035395117075e-05,
+      "loss": 1.3688,
       "step": 540
     },
     {
-      "epoch": 0.15625,
-      "grad_norm": 4.0350996506196175,
-      "learning_rate": 1.9808218727822808e-05,
-      "loss": 1.1899,
+      "epoch": 0.14919244456611006,
+      "grad_norm": 0.39303998157704906,
+      "learning_rate": 1.9854012097703515e-05,
+      "loss": 1.4259,
       "step": 545
     },
     {
-      "epoch": 0.15768348623853212,
-      "grad_norm": 3.806300412200292,
-      "learning_rate": 1.979834260083673e-05,
-      "loss": 1.2099,
+      "epoch": 0.15056118258965234,
+      "grad_norm": 0.3939411519174778,
+      "learning_rate": 1.9845763763973433e-05,
+      "loss": 1.354,
       "step": 550
     },
     {
-      "epoch": 0.15911697247706422,
-      "grad_norm": 4.447162495747635,
-      "learning_rate": 1.9788221110984026e-05,
-      "loss": 1.2123,
+      "epoch": 0.15192992061319463,
+      "grad_norm": 0.39277763892363776,
+      "learning_rate": 1.9837290582294233e-05,
+      "loss": 1.3468,
       "step": 555
     },
     {
-      "epoch": 0.16055045871559634,
-      "grad_norm": 4.094827599966712,
-      "learning_rate": 1.977785451171958e-05,
-      "loss": 1.1411,
+      "epoch": 0.15329865863673692,
+      "grad_norm": 0.39121990838351217,
+      "learning_rate": 1.9828592746168172e-05,
+      "loss": 1.3508,
       "step": 560
     },
     {
-      "epoch": 0.16198394495412843,
-      "grad_norm": 4.6429132125696535,
-      "learning_rate": 1.9767243062636122e-05,
-      "loss": 1.1914,
+      "epoch": 0.15466739666027923,
+      "grad_norm": 0.4035101254769364,
+      "learning_rate": 1.981967045422795e-05,
+      "loss": 1.369,
       "step": 565
     },
     {
-      "epoch": 0.16341743119266056,
-      "grad_norm": 3.9726685572488227,
-      "learning_rate": 1.975638702945773e-05,
-      "loss": 1.1758,
+      "epoch": 0.15603613468382152,
+      "grad_norm": 0.38407623227911974,
+      "learning_rate": 1.9810523910232165e-05,
+      "loss": 1.4039,
       "step": 570
     },
     {
-      "epoch": 0.16485091743119265,
-      "grad_norm": 3.9582277412235367,
-      "learning_rate": 1.974528668403318e-05,
-      "loss": 1.1839,
+      "epoch": 0.1574048727073638,
+      "grad_norm": 0.3915340445821982,
+      "learning_rate": 1.9801153323060667e-05,
+      "loss": 1.3634,
       "step": 575
     },
     {
-      "epoch": 0.16628440366972477,
-      "grad_norm": 4.611199474059946,
-      "learning_rate": 1.973394230432913e-05,
-      "loss": 1.2155,
+      "epoch": 0.1587736107309061,
+      "grad_norm": 0.39241210575593644,
+      "learning_rate": 1.9791558906709787e-05,
+      "loss": 1.4257,
       "step": 580
     },
     {
-      "epoch": 0.16771788990825687,
-      "grad_norm": 3.964842824608755,
-      "learning_rate": 1.972235417442317e-05,
-      "loss": 1.168,
+      "epoch": 0.1601423487544484,
+      "grad_norm": 0.3946963973614525,
+      "learning_rate": 1.9781740880287444e-05,
+      "loss": 1.3136,
       "step": 585
     },
     {
-      "epoch": 0.169151376146789,
-      "grad_norm": 4.083964710145448,
-      "learning_rate": 1.9710522584496695e-05,
-      "loss": 1.2135,
+      "epoch": 0.1615110867779907,
+      "grad_norm": 0.39148042578465725,
+      "learning_rate": 1.9771699468008156e-05,
+      "loss": 1.3654,
       "step": 590
     },
     {
-      "epoch": 0.1705848623853211,
-      "grad_norm": 3.8905730852414155,
-      "learning_rate": 1.9698447830827655e-05,
-      "loss": 1.1893,
+      "epoch": 0.16287982480153299,
+      "grad_norm": 0.3895106088217061,
+      "learning_rate": 1.9761434899187893e-05,
+      "loss": 1.3762,
       "step": 595
     },
     {
-      "epoch": 0.1720183486238532,
-      "grad_norm": 4.399950123881084,
-      "learning_rate": 1.9686130215783124e-05,
-      "loss": 1.2244,
+      "epoch": 0.16424856282507527,
+      "grad_norm": 0.3997937267692177,
+      "learning_rate": 1.9750947408238872e-05,
+      "loss": 1.4064,
       "step": 600
     },
     {
-      "epoch": 0.17345183486238533,
-      "grad_norm": 4.3066835429404495,
-      "learning_rate": 1.967357004781173e-05,
-      "loss": 1.1532,
+      "epoch": 0.1656173008486176,
+      "grad_norm": 0.4002294154735745,
+      "learning_rate": 1.974023723466418e-05,
+      "loss": 1.3771,
       "step": 605
     },
     {
-      "epoch": 0.17488532110091742,
-      "grad_norm": 5.071390390632784,
-      "learning_rate": 1.9660767641435926e-05,
-      "loss": 1.2137,
+      "epoch": 0.16698603887215988,
+      "grad_norm": 0.40550620529076886,
+      "learning_rate": 1.9729304623052315e-05,
+      "loss": 1.4551,
       "step": 610
     },
     {
-      "epoch": 0.17631880733944955,
-      "grad_norm": 3.738266843064544,
-      "learning_rate": 1.964772331724414e-05,
-      "loss": 1.165,
+      "epoch": 0.16835477689570216,
+      "grad_norm": 0.40410760524836825,
+      "learning_rate": 1.9718149823071592e-05,
+      "loss": 1.3394,
       "step": 615
     },
     {
-      "epoch": 0.17775229357798164,
-      "grad_norm": 3.885352429025116,
-      "learning_rate": 1.9634437401882707e-05,
-      "loss": 1.1899,
+      "epoch": 0.16972351491924445,
+      "grad_norm": 0.40142195594131513,
+      "learning_rate": 1.970677308946446e-05,
+      "loss": 1.3872,
       "step": 620
     },
     {
-      "epoch": 0.17918577981651376,
-      "grad_norm": 3.653935761439161,
-      "learning_rate": 1.962091022804772e-05,
-      "loss": 1.1506,
+      "epoch": 0.17109225294278674,
+      "grad_norm": 0.4041136169142662,
+      "learning_rate": 1.9695174682041652e-05,
+      "loss": 1.3644,
       "step": 625
     },
     {
-      "epoch": 0.18061926605504589,
-      "grad_norm": 4.559227033021131,
-      "learning_rate": 1.960714213447668e-05,
-      "loss": 1.1645,
+      "epoch": 0.17246099096632905,
+      "grad_norm": 0.39436310577761385,
+      "learning_rate": 1.9683354865676298e-05,
+      "loss": 1.434,
       "step": 630
     },
     {
-      "epoch": 0.18205275229357798,
-      "grad_norm": 4.618925411728112,
-      "learning_rate": 1.959313346594004e-05,
-      "loss": 1.1802,
+      "epoch": 0.17382972898987134,
+      "grad_norm": 0.400517293758074,
+      "learning_rate": 1.9671313910297826e-05,
+      "loss": 1.3941,
       "step": 635
     },
     {
-      "epoch": 0.1834862385321101,
-      "grad_norm": 4.308837236054774,
-      "learning_rate": 1.9578884573232538e-05,
-      "loss": 1.2023,
+      "epoch": 0.17519846701341363,
+      "grad_norm": 0.4059408812570113,
+      "learning_rate": 1.9659052090885834e-05,
+      "loss": 1.3907,
       "step": 640
     },
     {
-      "epoch": 0.1849197247706422,
-      "grad_norm": 4.304427359268019,
-      "learning_rate": 1.9564395813164428e-05,
-      "loss": 1.2244,
+      "epoch": 0.17656720503695592,
+      "grad_norm": 0.3929001302562383,
+      "learning_rate": 1.9646569687463796e-05,
+      "loss": 1.351,
       "step": 645
     },
     {
-      "epoch": 0.18635321100917432,
-      "grad_norm": 3.486413458583046,
-      "learning_rate": 1.9549667548552557e-05,
-      "loss": 1.2111,
+      "epoch": 0.17793594306049823,
+      "grad_norm": 0.4090718300758474,
+      "learning_rate": 1.9633866985092655e-05,
+      "loss": 1.4102,
       "step": 650
     },
     {
-      "epoch": 0.1877866972477064,
-      "grad_norm": 3.7640427311235403,
-      "learning_rate": 1.9534700148211255e-05,
-      "loss": 1.1956,
+      "epoch": 0.17930468108404052,
+      "grad_norm": 0.40995200812054255,
+      "learning_rate": 1.9620944273864343e-05,
+      "loss": 1.3575,
       "step": 655
     },
     {
-      "epoch": 0.18922018348623854,
-      "grad_norm": 4.222766727318349,
-      "learning_rate": 1.9519493986943125e-05,
-      "loss": 1.1996,
+      "epoch": 0.1806734191075828,
+      "grad_norm": 0.40276642674265933,
+      "learning_rate": 1.960780184889514e-05,
+      "loss": 1.3941,
       "step": 660
     },
     {
-      "epoch": 0.19065366972477063,
-      "grad_norm": 3.7868143111703256,
-      "learning_rate": 1.9504049445529632e-05,
-      "loss": 1.2093,
+      "epoch": 0.1820421571311251,
+      "grad_norm": 0.400136397383414,
+      "learning_rate": 1.9594440010318924e-05,
+      "loss": 1.4171,
       "step": 665
     },
     {
-      "epoch": 0.19208715596330275,
-      "grad_norm": 3.579431474524795,
-      "learning_rate": 1.94883669107216e-05,
-      "loss": 1.1785,
+      "epoch": 0.1834108951546674,
+      "grad_norm": 0.40872717576906,
+      "learning_rate": 1.9580859063280326e-05,
+      "loss": 1.391,
       "step": 670
     },
     {
-      "epoch": 0.19352064220183487,
-      "grad_norm": 4.14899759337832,
-      "learning_rate": 1.9472446775229486e-05,
-      "loss": 1.2402,
+      "epoch": 0.1847796331782097,
+      "grad_norm": 0.39405384932557985,
+      "learning_rate": 1.956705931792777e-05,
+      "loss": 1.4042,
       "step": 675
     },
     {
-      "epoch": 0.19495412844036697,
-      "grad_norm": 3.7259985492864804,
-      "learning_rate": 1.9456289437713578e-05,
-      "loss": 1.1713,
+      "epoch": 0.18614837120175198,
+      "grad_norm": 0.396632571946212,
+      "learning_rate": 1.9553041089406387e-05,
+      "loss": 1.3598,
       "step": 680
     },
     {
-      "epoch": 0.1963876146788991,
-      "grad_norm": 4.525451494816849,
-      "learning_rate": 1.9439895302774007e-05,
-      "loss": 1.2179,
+      "epoch": 0.18751710922529427,
+      "grad_norm": 0.39535225715229844,
+      "learning_rate": 1.95388046978508e-05,
+      "loss": 1.3652,
       "step": 685
     },
     {
-      "epoch": 0.1978211009174312,
-      "grad_norm": 3.9639891914703185,
-      "learning_rate": 1.9423264780940602e-05,
-      "loss": 1.2163,
+      "epoch": 0.18888584724883659,
+      "grad_norm": 0.39747240968211434,
+      "learning_rate": 1.9524350468377828e-05,
+      "loss": 1.3857,
       "step": 690
     },
     {
-      "epoch": 0.1992545871559633,
-      "grad_norm": 3.8210369748297612,
-      "learning_rate": 1.940639828866262e-05,
-      "loss": 1.2462,
+      "epoch": 0.19025458527237887,
+      "grad_norm": 0.4045442697980633,
+      "learning_rate": 1.9509678731079074e-05,
+      "loss": 1.3724,
       "step": 695
     },
     {
-      "epoch": 0.2006880733944954,
-      "grad_norm": 4.499363334347953,
-      "learning_rate": 1.938929624829832e-05,
-      "loss": 1.2572,
+      "epoch": 0.19162332329592116,
+      "grad_norm": 0.39648112930634666,
+      "learning_rate": 1.949478982101336e-05,
+      "loss": 1.3642,
       "step": 700
     },
     {
-      "epoch": 0.20212155963302753,
-      "grad_norm": 32.95893389851377,
-      "learning_rate": 1.937195908810438e-05,
-      "loss": 1.2332,
+      "epoch": 0.19299206131946345,
+      "grad_norm": 0.3966559822195587,
+      "learning_rate": 1.947968407819909e-05,
+      "loss": 1.3704,
       "step": 705
     },
     {
-      "epoch": 0.20355504587155962,
-      "grad_norm": 4.056487188350767,
-      "learning_rate": 1.935438724222517e-05,
-      "loss": 1.1797,
+      "epoch": 0.19436079934300574,
+      "grad_norm": 0.3803799928500003,
+      "learning_rate": 1.9464361847606486e-05,
+      "loss": 1.3718,
       "step": 710
     },
     {
-      "epoch": 0.20498853211009174,
-      "grad_norm": 3.940310991182425,
-      "learning_rate": 1.93365811506819e-05,
-      "loss": 1.2315,
+      "epoch": 0.19572953736654805,
+      "grad_norm": 0.3920580621341347,
+      "learning_rate": 1.9448823479149705e-05,
+      "loss": 1.3994,
       "step": 715
     },
     {
-      "epoch": 0.20642201834862386,
-      "grad_norm": 3.92517292280482,
-      "learning_rate": 1.9318541259361573e-05,
-      "loss": 1.1824,
+      "epoch": 0.19709827539009034,
+      "grad_norm": 0.38159551727148144,
+      "learning_rate": 1.9433069327678847e-05,
+      "loss": 1.3539,
       "step": 720
     },
     {
-      "epoch": 0.20785550458715596,
-      "grad_norm": 4.109451546798357,
-      "learning_rate": 1.9300268020005832e-05,
-      "loss": 1.2375,
+      "epoch": 0.19846701341363263,
+      "grad_norm": 0.3979848267470623,
+      "learning_rate": 1.9417099752971858e-05,
+      "loss": 1.3824,
       "step": 725
     },
     {
-      "epoch": 0.20928899082568808,
-      "grad_norm": 4.168329616635684,
-      "learning_rate": 1.9281761890199666e-05,
-      "loss": 1.1812,
+      "epoch": 0.1998357514371749,
+      "grad_norm": 0.36228811525683013,
+      "learning_rate": 1.9400915119726305e-05,
+      "loss": 1.2942,
       "step": 730
     },
     {
-      "epoch": 0.21072247706422018,
-      "grad_norm": 3.8618338414153435,
-      "learning_rate": 1.9263023333359918e-05,
-      "loss": 1.1903,
+      "epoch": 0.20120448946071723,
+      "grad_norm": 0.40336839547536973,
+      "learning_rate": 1.938451579755106e-05,
+      "loss": 1.3305,
       "step": 735
     },
     {
-      "epoch": 0.2121559633027523,
-      "grad_norm": 4.070643005805358,
-      "learning_rate": 1.9244052818723706e-05,
-      "loss": 1.2539,
+      "epoch": 0.20257322748425952,
+      "grad_norm": 0.4093706132368206,
+      "learning_rate": 1.9367902160957843e-05,
+      "loss": 1.4047,
       "step": 740
     },
     {
-      "epoch": 0.2135894495412844,
-      "grad_norm": 3.9207051720216004,
-      "learning_rate": 1.9224850821336664e-05,
-      "loss": 1.2375,
+      "epoch": 0.2039419655078018,
+      "grad_norm": 0.4012678664539993,
+      "learning_rate": 1.9351074589352684e-05,
+      "loss": 1.34,
       "step": 745
     },
     {
-      "epoch": 0.21502293577981652,
-      "grad_norm": 4.14495928553583,
-      "learning_rate": 1.920541782204104e-05,
-      "loss": 1.253,
+      "epoch": 0.2053107035313441,
+      "grad_norm": 0.3982341789573293,
+      "learning_rate": 1.933403346702725e-05,
+      "loss": 1.3518,
       "step": 750
     },
     {
-      "epoch": 0.21645642201834864,
-      "grad_norm": 4.210904171053472,
-      "learning_rate": 1.918575430746367e-05,
-      "loss": 1.1987,
+      "epoch": 0.2066794415548864,
+      "grad_norm": 0.3895851509088621,
+      "learning_rate": 1.931677918315007e-05,
+      "loss": 1.3698,
       "step": 755
     },
     {
-      "epoch": 0.21788990825688073,
-      "grad_norm": 3.6674431622452883,
-      "learning_rate": 1.9165860770003774e-05,
-      "loss": 1.2427,
+      "epoch": 0.2080481795784287,
+      "grad_norm": 0.40873783451123424,
+      "learning_rate": 1.9299312131757645e-05,
+      "loss": 1.3768,
       "step": 760
     },
     {
-      "epoch": 0.21932339449541285,
-      "grad_norm": 3.7503035449224127,
-      "learning_rate": 1.914573770782065e-05,
-      "loss": 1.1564,
+      "epoch": 0.20941691760197098,
+      "grad_norm": 0.38863227885235524,
+      "learning_rate": 1.928163271174546e-05,
+      "loss": 1.368,
       "step": 765
     },
     {
-      "epoch": 0.22075688073394495,
-      "grad_norm": 4.206820549333469,
-      "learning_rate": 1.9125385624821162e-05,
-      "loss": 1.2026,
+      "epoch": 0.21078565562551327,
+      "grad_norm": 0.40218624042286666,
+      "learning_rate": 1.9263741326858866e-05,
+      "loss": 1.4002,
       "step": 770
     },
     {
-      "epoch": 0.22219036697247707,
-      "grad_norm": 4.079050524248544,
-      "learning_rate": 1.9104805030647164e-05,
-      "loss": 1.2521,
+      "epoch": 0.21215439364905558,
+      "grad_norm": 0.38750285278794777,
+      "learning_rate": 1.9245638385683857e-05,
+      "loss": 1.3808,
       "step": 775
     },
     {
-      "epoch": 0.22362385321100917,
-      "grad_norm": 3.8439532493920585,
-      "learning_rate": 1.908399644066272e-05,
-      "loss": 1.2247,
+      "epoch": 0.21352313167259787,
+      "grad_norm": 0.39332958062090156,
+      "learning_rate": 1.9227324301637747e-05,
+      "loss": 1.3991,
       "step": 780
     },
     {
-      "epoch": 0.2250573394495413,
-      "grad_norm": 4.419153034660572,
-      "learning_rate": 1.906296037594117e-05,
-      "loss": 1.2827,
+      "epoch": 0.21489186969614016,
+      "grad_norm": 0.39759945858946194,
+      "learning_rate": 1.9208799492959723e-05,
+      "loss": 1.3765,
       "step": 785
     },
     {
-      "epoch": 0.22649082568807338,
-      "grad_norm": 4.802038277224569,
-      "learning_rate": 1.904169736325215e-05,
-      "loss": 1.262,
+      "epoch": 0.21626060771968245,
+      "grad_norm": 0.3779639266645249,
+      "learning_rate": 1.9190064382701296e-05,
+      "loss": 1.378,
       "step": 790
     },
     {
-      "epoch": 0.2279243119266055,
-      "grad_norm": 4.284188903637777,
-      "learning_rate": 1.9020207935048317e-05,
-      "loss": 1.1532,
+      "epoch": 0.21762934574322473,
+      "grad_norm": 0.3980685187213535,
+      "learning_rate": 1.917111939871664e-05,
+      "loss": 1.3559,
       "step": 795
     },
     {
-      "epoch": 0.22935779816513763,
-      "grad_norm": 3.639715866595117,
-      "learning_rate": 1.8998492629452087e-05,
-      "loss": 1.2381,
+      "epoch": 0.21899808376676705,
+      "grad_norm": 0.3955990129038042,
+      "learning_rate": 1.915196497365282e-05,
+      "loss": 1.3688,
       "step": 800
     },
     {
-      "epoch": 0.23079128440366972,
-      "grad_norm": 3.741005239883841,
-      "learning_rate": 1.8976551990242122e-05,
-      "loss": 1.216,
+      "epoch": 0.22036682179030934,
+      "grad_norm": 0.3814772041353902,
+      "learning_rate": 1.9132601544939914e-05,
+      "loss": 1.3106,
       "step": 805
     },
     {
-      "epoch": 0.23222477064220184,
-      "grad_norm": 3.6139511602383276,
-      "learning_rate": 1.895438656683972e-05,
-      "loss": 1.2473,
+      "epoch": 0.22173555981385162,
+      "grad_norm": 0.4033095650227253,
+      "learning_rate": 1.9113029554781014e-05,
+      "loss": 1.3389,
       "step": 810
     },
     {
-      "epoch": 0.23365825688073394,
-      "grad_norm": 3.8300619382359846,
-      "learning_rate": 1.8931996914295065e-05,
-      "loss": 1.1882,
+      "epoch": 0.2231042978373939,
+      "grad_norm": 0.3934860150762494,
+      "learning_rate": 1.9093249450142144e-05,
+      "loss": 1.3343,
       "step": 815
     },
     {
-      "epoch": 0.23509174311926606,
-      "grad_norm": 3.943889343109655,
-      "learning_rate": 1.8909383593273317e-05,
-      "loss": 1.2444,
+      "epoch": 0.22447303586093623,
+      "grad_norm": 0.4008166941226203,
+      "learning_rate": 1.907326168274204e-05,
+      "loss": 1.3627,
       "step": 820
     },
     {
-      "epoch": 0.23652522935779816,
-      "grad_norm": 4.115538048586424,
-      "learning_rate": 1.8886547170040575e-05,
-      "loss": 1.233,
+      "epoch": 0.2258417738844785,
+      "grad_norm": 0.38863913805386574,
+      "learning_rate": 1.905306670904184e-05,
+      "loss": 1.3978,
       "step": 825
     },
     {
-      "epoch": 0.23795871559633028,
-      "grad_norm": 5.787154468595948,
-      "learning_rate": 1.8863488216449702e-05,
-      "loss": 1.2236,
+      "epoch": 0.2272105119080208,
+      "grad_norm": 0.39887884072448215,
+      "learning_rate": 1.9032664990234648e-05,
+      "loss": 1.3548,
       "step": 830
     },
     {
-      "epoch": 0.23939220183486237,
-      "grad_norm": 3.778052893153268,
-      "learning_rate": 1.8840207309926003e-05,
-      "loss": 1.2286,
+      "epoch": 0.2285792499315631,
+      "grad_norm": 0.3942528539065832,
+      "learning_rate": 1.9012056992235025e-05,
+      "loss": 1.3589,
       "step": 835
     },
     {
-      "epoch": 0.2408256880733945,
-      "grad_norm": 3.435004162216508,
-      "learning_rate": 1.881670503345277e-05,
-      "loss": 1.1868,
+      "epoch": 0.2299479879551054,
+      "grad_norm": 0.4003093485695292,
+      "learning_rate": 1.899124318566832e-05,
+      "loss": 1.4008,
       "step": 840
     },
     {
-      "epoch": 0.24225917431192662,
-      "grad_norm": 5.261326157501783,
-      "learning_rate": 1.879298197555666e-05,
-      "loss": 1.2486,
+      "epoch": 0.2313167259786477,
+      "grad_norm": 0.38851656838431264,
+      "learning_rate": 1.897022404585996e-05,
+      "loss": 1.3396,
       "step": 845
     },
     {
-      "epoch": 0.2436926605504587,
-      "grad_norm": 3.651324240819765,
-      "learning_rate": 1.8769038730292993e-05,
-      "loss": 1.2383,
+      "epoch": 0.23268546400218998,
+      "grad_norm": 0.39282411404685663,
+      "learning_rate": 1.894900005282454e-05,
+      "loss": 1.3118,
       "step": 850
     },
     {
-      "epoch": 0.24512614678899083,
-      "grad_norm": 4.834151575695704,
-      "learning_rate": 1.8744875897230853e-05,
-      "loss": 1.2592,
+      "epoch": 0.23405420202573227,
+      "grad_norm": 0.3995600743628729,
+      "learning_rate": 1.892757169125492e-05,
+      "loss": 1.3261,
       "step": 855
     },
     {
-      "epoch": 0.24655963302752293,
-      "grad_norm": 3.9673796319727606,
-      "learning_rate": 1.872049408143808e-05,
-      "loss": 1.2135,
+      "epoch": 0.23542294004927458,
+      "grad_norm": 0.40585943733175783,
+      "learning_rate": 1.8905939450511117e-05,
+      "loss": 1.2896,
       "step": 860
     },
     {
-      "epoch": 0.24799311926605505,
-      "grad_norm": 3.9495356227550293,
-      "learning_rate": 1.869589389346611e-05,
-      "loss": 1.2443,
+      "epoch": 0.23679167807281687,
+      "grad_norm": 0.38676778157652897,
+      "learning_rate": 1.888410382460915e-05,
+      "loss": 1.3769,
       "step": 865
     },
     {
-      "epoch": 0.24942660550458715,
-      "grad_norm": 3.8171283870857917,
-      "learning_rate": 1.8671075949334713e-05,
-      "loss": 1.1502,
+      "epoch": 0.23816041609635916,
+      "grad_norm": 0.4006135132967053,
+      "learning_rate": 1.8862065312209735e-05,
+      "loss": 1.3744,
       "step": 870
     },
     {
-      "epoch": 0.25086009174311924,
-      "grad_norm": 3.594654406802359,
-      "learning_rate": 1.8646040870516526e-05,
-      "loss": 1.1831,
+      "epoch": 0.23952915411990144,
+      "grad_norm": 0.38183538010624907,
+      "learning_rate": 1.8839824416606932e-05,
+      "loss": 1.3994,
       "step": 875
     },
     {
-      "epoch": 0.25229357798165136,
-      "grad_norm": 3.5772784131514337,
-      "learning_rate": 1.862078928392153e-05,
-      "loss": 1.2116,
+      "epoch": 0.24089789214344373,
+      "grad_norm": 0.40012172515979694,
+      "learning_rate": 1.8817381645716613e-05,
+      "loss": 1.3736,
       "step": 880
     },
     {
-      "epoch": 0.2537270642201835,
-      "grad_norm": 4.016497035159711,
-      "learning_rate": 1.8595321821881322e-05,
-      "loss": 1.2156,
+      "epoch": 0.24226663016698605,
+      "grad_norm": 0.4036320450863606,
+      "learning_rate": 1.879473751206489e-05,
+      "loss": 1.3679,
       "step": 885
     },
     {
-      "epoch": 0.2551605504587156,
-      "grad_norm": 3.6584048014823622,
-      "learning_rate": 1.8569639122133304e-05,
-      "loss": 1.1687,
+      "epoch": 0.24363536819052833,
+      "grad_norm": 0.39250548017399456,
+      "learning_rate": 1.8771892532776406e-05,
+      "loss": 1.352,
       "step": 890
     },
     {
-      "epoch": 0.25659403669724773,
-      "grad_norm": 4.092581816708313,
-      "learning_rate": 1.8543741827804685e-05,
-      "loss": 1.2433,
+      "epoch": 0.24500410621407062,
+      "grad_norm": 0.384618237208868,
+      "learning_rate": 1.8748847229562504e-05,
+      "loss": 1.3415,
       "step": 895
     },
     {
-      "epoch": 0.2580275229357798,
-      "grad_norm": 10.965320915487338,
-      "learning_rate": 1.8517630587396413e-05,
-      "loss": 1.2365,
+      "epoch": 0.2463728442376129,
+      "grad_norm": 0.3847231065307316,
+      "learning_rate": 1.8725602128709348e-05,
+      "loss": 1.436,
       "step": 900
     },
     {
-      "epoch": 0.2594610091743119,
-      "grad_norm": 3.744425931996878,
-      "learning_rate": 1.8491306054766907e-05,
-      "loss": 1.2315,
+      "epoch": 0.24774158226115522,
+      "grad_norm": 0.39329176880661687,
+      "learning_rate": 1.8702157761065877e-05,
+      "loss": 1.344,
       "step": 905
     },
     {
-      "epoch": 0.26089449541284404,
-      "grad_norm": 4.446159019549649,
-      "learning_rate": 1.8464768889115684e-05,
-      "loss": 1.165,
+      "epoch": 0.2491103202846975,
+      "grad_norm": 0.4202514383981607,
+      "learning_rate": 1.8678514662031688e-05,
+      "loss": 1.3384,
       "step": 910
     },
     {
-      "epoch": 0.26232798165137616,
-      "grad_norm": 3.4122914752905227,
-      "learning_rate": 1.8438019754966877e-05,
-      "loss": 1.1715,
+      "epoch": 0.2504790583082398,
+      "grad_norm": 0.3866724615350219,
+      "learning_rate": 1.8654673371544815e-05,
+      "loss": 1.3374,
       "step": 915
     },
     {
-      "epoch": 0.26376146788990823,
-      "grad_norm": 3.681090000606899,
-      "learning_rate": 1.841105932215256e-05,
-      "loss": 1.2197,
+      "epoch": 0.2518477963317821,
+      "grad_norm": 0.4109333525553728,
+      "learning_rate": 1.8630634434069397e-05,
+      "loss": 1.3949,
       "step": 920
     },
     {
-      "epoch": 0.26519495412844035,
-      "grad_norm": 3.723633399780894,
-      "learning_rate": 1.838388826579601e-05,
-      "loss": 1.2308,
+      "epoch": 0.2532165343553244,
+      "grad_norm": 0.3782421450027903,
+      "learning_rate": 1.860639839858324e-05,
+      "loss": 1.3162,
       "step": 925
     },
     {
-      "epoch": 0.2666284403669725,
-      "grad_norm": 3.79765489569063,
-      "learning_rate": 1.835650726629477e-05,
-      "loss": 1.2388,
+      "epoch": 0.2545852723788667,
+      "grad_norm": 0.3909532902784873,
+      "learning_rate": 1.8581965818565278e-05,
+      "loss": 1.3829,
       "step": 930
     },
     {
-      "epoch": 0.2680619266055046,
-      "grad_norm": 3.636387647388923,
-      "learning_rate": 1.8328917009303634e-05,
-      "loss": 1.2296,
+      "epoch": 0.255954010402409,
+      "grad_norm": 0.4051170609748319,
+      "learning_rate": 1.855733725198295e-05,
+      "loss": 1.3462,
       "step": 935
     },
     {
-      "epoch": 0.2694954128440367,
-      "grad_norm": 3.696962989830932,
-      "learning_rate": 1.830111818571745e-05,
-      "loss": 1.2048,
+      "epoch": 0.25732274842595126,
+      "grad_norm": 0.39349552836368845,
+      "learning_rate": 1.8532513261279433e-05,
+      "loss": 1.4015,
       "step": 940
     },
     {
-      "epoch": 0.2709288990825688,
-      "grad_norm": 3.896582872078181,
-      "learning_rate": 1.8273111491653867e-05,
-      "loss": 1.2522,
+      "epoch": 0.2586914864494936,
+      "grad_norm": 0.39347934180028793,
+      "learning_rate": 1.8507494413360808e-05,
+      "loss": 1.3367,
       "step": 945
     },
     {
-      "epoch": 0.2723623853211009,
-      "grad_norm": 3.7244583842983454,
-      "learning_rate": 1.824489762843584e-05,
-      "loss": 1.1896,
+      "epoch": 0.26006022447303584,
+      "grad_norm": 0.4158855876544813,
+      "learning_rate": 1.848228127958312e-05,
+      "loss": 1.353,
       "step": 950
     },
     {
-      "epoch": 0.27379587155963303,
-      "grad_norm": 3.66702244239725,
-      "learning_rate": 1.821647730257413e-05,
-      "loss": 1.1777,
+      "epoch": 0.26142896249657815,
+      "grad_norm": 0.3793682443730692,
+      "learning_rate": 1.8456874435739337e-05,
+      "loss": 1.3398,
       "step": 955
     },
     {
-      "epoch": 0.27522935779816515,
-      "grad_norm": 6.778664388382065,
-      "learning_rate": 1.818785122574956e-05,
-      "loss": 1.2239,
+      "epoch": 0.26279770052012047,
+      "grad_norm": 0.38047885709566037,
+      "learning_rate": 1.843127446204616e-05,
+      "loss": 1.329,
       "step": 960
     },
     {
-      "epoch": 0.2766628440366973,
-      "grad_norm": 3.5798872160414095,
-      "learning_rate": 1.8159020114795226e-05,
-      "loss": 1.1532,
+      "epoch": 0.26416643854366273,
+      "grad_norm": 0.3991275596683987,
+      "learning_rate": 1.8405481943130827e-05,
+      "loss": 1.3115,
       "step": 965
     },
     {
-      "epoch": 0.27809633027522934,
-      "grad_norm": 4.080407231530825,
-      "learning_rate": 1.8129984691678547e-05,
-      "loss": 1.2046,
+      "epoch": 0.26553517656720504,
+      "grad_norm": 0.389649193690539,
+      "learning_rate": 1.8379497468017726e-05,
+      "loss": 1.3746,
       "step": 970
     },
     {
-      "epoch": 0.27952981651376146,
-      "grad_norm": 3.9179230686635234,
-      "learning_rate": 1.8100745683483168e-05,
-      "loss": 1.254,
+      "epoch": 0.2669039145907473,
+      "grad_norm": 0.40470694231468296,
+      "learning_rate": 1.8353321630114952e-05,
+      "loss": 1.3354,
       "step": 975
     },
     {
-      "epoch": 0.2809633027522936,
-      "grad_norm": 4.97899730667325,
-      "learning_rate": 1.807130382239075e-05,
-      "loss": 1.2084,
+      "epoch": 0.2682726526142896,
+      "grad_norm": 0.40734933283960667,
+      "learning_rate": 1.832695502720076e-05,
+      "loss": 1.3295,
       "step": 980
     },
     {
-      "epoch": 0.2823967889908257,
-      "grad_norm": 3.8642747207026242,
-      "learning_rate": 1.8041659845662663e-05,
-      "loss": 1.2014,
+      "epoch": 0.26964139063783193,
+      "grad_norm": 0.39323434740751423,
+      "learning_rate": 1.8300398261409912e-05,
+      "loss": 1.3069,
       "step": 985
     },
     {
-      "epoch": 0.2838302752293578,
-      "grad_norm": 3.4483181116099964,
-      "learning_rate": 1.8011814495621506e-05,
-      "loss": 1.1435,
+      "epoch": 0.2710101286613742,
+      "grad_norm": 0.3837841492160018,
+      "learning_rate": 1.8273651939219914e-05,
+      "loss": 1.3543,
       "step": 990
     },
     {
-      "epoch": 0.2852637614678899,
-      "grad_norm": 3.8042720182477434,
-      "learning_rate": 1.798176851963251e-05,
-      "loss": 1.2047,
+      "epoch": 0.2723788666849165,
+      "grad_norm": 0.38795485386053835,
+      "learning_rate": 1.8246716671437186e-05,
+      "loss": 1.3798,
       "step": 995
     },
     {
-      "epoch": 0.286697247706422,
-      "grad_norm": 3.483685408627929,
-      "learning_rate": 1.7951522670084847e-05,
-      "loss": 1.171,
+      "epoch": 0.2737476047084588,
+      "grad_norm": 0.38509661706339016,
+      "learning_rate": 1.8219593073183106e-05,
+      "loss": 1.3604,
       "step": 1000
     },
     {
-      "epoch": 0.28813073394495414,
-      "grad_norm": 3.487052716444198,
-      "learning_rate": 1.792107770437276e-05,
-      "loss": 1.2396,
+      "epoch": 0.2751163427320011,
+      "grad_norm": 0.3882331224756399,
+      "learning_rate": 1.8192281763879946e-05,
+      "loss": 1.3417,
       "step": 1005
     },
     {
-      "epoch": 0.28956422018348627,
-      "grad_norm": 3.553005650055509,
-      "learning_rate": 1.789043438487662e-05,
-      "loss": 1.1149,
+      "epoch": 0.2764850807555434,
+      "grad_norm": 0.3842715356979672,
+      "learning_rate": 1.816478336723675e-05,
+      "loss": 1.4061,
       "step": 1010
     },
     {
-      "epoch": 0.29099770642201833,
-      "grad_norm": 3.5693421017081306,
-      "learning_rate": 1.785959347894383e-05,
-      "loss": 1.2387,
+      "epoch": 0.27785381877908566,
+      "grad_norm": 0.4013303335538485,
+      "learning_rate": 1.8137098511235084e-05,
+      "loss": 1.3509,
       "step": 1015
     },
     {
-      "epoch": 0.29243119266055045,
-      "grad_norm": 3.535948224388696,
-      "learning_rate": 1.7828555758869602e-05,
-      "loss": 1.232,
+      "epoch": 0.279222556802628,
+      "grad_norm": 0.389250409772092,
+      "learning_rate": 1.810922782811468e-05,
+      "loss": 1.3371,
       "step": 1020
     },
     {
-      "epoch": 0.2938646788990826,
-      "grad_norm": 3.522215138188648,
-      "learning_rate": 1.7797322001877625e-05,
-      "loss": 1.2004,
+      "epoch": 0.2805912948261703,
+      "grad_norm": 0.3891214912082097,
+      "learning_rate": 1.808117195435901e-05,
+      "loss": 1.3756,
       "step": 1025
     },
     {
-      "epoch": 0.2952981651376147,
-      "grad_norm": 3.9401114869933362,
-      "learning_rate": 1.7765892990100593e-05,
-      "loss": 1.1954,
+      "epoch": 0.28196003284971255,
+      "grad_norm": 0.39295322317665116,
+      "learning_rate": 1.805293153068076e-05,
+      "loss": 1.3525,
       "step": 1030
     },
     {
-      "epoch": 0.29673165137614677,
-      "grad_norm": 3.9452874322873424,
-      "learning_rate": 1.773426951056064e-05,
-      "loss": 1.1752,
+      "epoch": 0.28332877087325486,
+      "grad_norm": 0.38744159952014173,
+      "learning_rate": 1.802450720200718e-05,
+      "loss": 1.3508,
       "step": 1035
     },
     {
-      "epoch": 0.2981651376146789,
-      "grad_norm": 3.77900474922427,
-      "learning_rate": 1.7702452355149606e-05,
-      "loss": 1.2023,
+      "epoch": 0.2846975088967972,
+      "grad_norm": 0.3971694811346686,
+      "learning_rate": 1.7995899617465357e-05,
+      "loss": 1.3242,
       "step": 1040
     },
     {
-      "epoch": 0.299598623853211,
-      "grad_norm": 4.194810233115317,
-      "learning_rate": 1.7670442320609226e-05,
-      "loss": 1.1762,
+      "epoch": 0.28606624692033944,
+      "grad_norm": 0.3892979964335795,
+      "learning_rate": 1.7967109430367406e-05,
+      "loss": 1.2919,
       "step": 1045
     },
     {
-      "epoch": 0.30103211009174313,
-      "grad_norm": 3.6317516711250524,
-      "learning_rate": 1.7638240208511162e-05,
-      "loss": 1.2036,
+      "epoch": 0.28743498494388176,
+      "grad_norm": 0.40651116778504365,
+      "learning_rate": 1.793813729819553e-05,
+      "loss": 1.4047,
       "step": 1050
     },
     {
-      "epoch": 0.30246559633027525,
-      "grad_norm": 3.613998915069475,
-      "learning_rate": 1.760584682523696e-05,
-      "loss": 1.1794,
+      "epoch": 0.288803722967424,
+      "grad_norm": 0.39365601735181455,
+      "learning_rate": 1.7908983882587038e-05,
+      "loss": 1.3622,
       "step": 1055
     },
     {
-      "epoch": 0.3038990825688073,
-      "grad_norm": 3.168795755449482,
-      "learning_rate": 1.7573262981957814e-05,
-      "loss": 1.1812,
+      "epoch": 0.29017246099096633,
+      "grad_norm": 0.38324138023204,
+      "learning_rate": 1.787964984931919e-05,
+      "loss": 1.3773,
       "step": 1060
     },
     {
-      "epoch": 0.30533256880733944,
-      "grad_norm": 3.883990610964808,
-      "learning_rate": 1.7540489494614294e-05,
-      "loss": 1.2273,
+      "epoch": 0.29154119901450865,
+      "grad_norm": 0.3887104304366927,
+      "learning_rate": 1.7850135868294023e-05,
+      "loss": 1.3973,
       "step": 1065
     },
     {
-      "epoch": 0.30676605504587157,
-      "grad_norm": 4.068560981758666,
-      "learning_rate": 1.7507527183895893e-05,
-      "loss": 1.2297,
+      "epoch": 0.2929099370380509,
+      "grad_norm": 0.4087532681641359,
+      "learning_rate": 1.782044261352305e-05,
+      "loss": 1.3246,
       "step": 1070
     },
     {
-      "epoch": 0.3081995412844037,
-      "grad_norm": 3.812795569574441,
-      "learning_rate": 1.747437687522047e-05,
-      "loss": 1.2037,
+      "epoch": 0.2942786750615932,
+      "grad_norm": 0.39381567774906023,
+      "learning_rate": 1.7790570763111864e-05,
+      "loss": 1.3683,
       "step": 1075
     },
     {
-      "epoch": 0.30963302752293576,
-      "grad_norm": 3.7886870743963077,
-      "learning_rate": 1.744103939871361e-05,
-      "loss": 1.2235,
+      "epoch": 0.2956474130851355,
+      "grad_norm": 0.38720936830053604,
+      "learning_rate": 1.7760520999244638e-05,
+      "loss": 1.375,
       "step": 1080
     },
     {
-      "epoch": 0.3110665137614679,
-      "grad_norm": 3.3064001495116555,
-      "learning_rate": 1.7407515589187793e-05,
-      "loss": 1.213,
+      "epoch": 0.2970161511086778,
+      "grad_norm": 0.38699812271583683,
+      "learning_rate": 1.7730294008168578e-05,
+      "loss": 1.403,
       "step": 1085
     },
     {
-      "epoch": 0.3125,
-      "grad_norm": 3.999062074927365,
-      "learning_rate": 1.7373806286121532e-05,
-      "loss": 1.2586,
+      "epoch": 0.2983848891322201,
+      "grad_norm": 0.3938155567988771,
+      "learning_rate": 1.7699890480178216e-05,
+      "loss": 1.3567,
       "step": 1090
     },
     {
-      "epoch": 0.3139334862385321,
-      "grad_norm": 4.371331122546795,
-      "learning_rate": 1.7339912333638322e-05,
-      "loss": 1.1731,
+      "epoch": 0.29975362715576237,
+      "grad_norm": 0.4051470823072413,
+      "learning_rate": 1.766931110959967e-05,
+      "loss": 1.4228,
       "step": 1095
     },
     {
-      "epoch": 0.31536697247706424,
-      "grad_norm": 3.8167405663507,
-      "learning_rate": 1.730583458048552e-05,
-      "loss": 1.2098,
+      "epoch": 0.3011223651793047,
+      "grad_norm": 0.4013404711196811,
+      "learning_rate": 1.763855659477478e-05,
+      "loss": 1.3689,
       "step": 1100
     },
     {
-      "epoch": 0.3168004587155963,
-      "grad_norm": 3.398825843337014,
-      "learning_rate": 1.727157388001307e-05,
-      "loss": 1.187,
+      "epoch": 0.302491103202847,
+      "grad_norm": 0.4239789885244797,
+      "learning_rate": 1.7607627638045156e-05,
+      "loss": 1.3988,
       "step": 1105
     },
     {
-      "epoch": 0.31823394495412843,
-      "grad_norm": 3.323984794217734,
-      "learning_rate": 1.723713109015217e-05,
-      "loss": 1.229,
+      "epoch": 0.30385984122638926,
+      "grad_norm": 0.4005302271519552,
+      "learning_rate": 1.7576524945736137e-05,
+      "loss": 1.3368,
       "step": 1110
     },
     {
-      "epoch": 0.31966743119266056,
-      "grad_norm": 3.388696780319304,
-      "learning_rate": 1.720250707339374e-05,
-      "loss": 1.1616,
+      "epoch": 0.3052285792499316,
+      "grad_norm": 0.3900827670562639,
+      "learning_rate": 1.754524922814068e-05,
+      "loss": 1.3633,
       "step": 1115
     },
     {
-      "epoch": 0.3211009174311927,
-      "grad_norm": 3.7284978608483934,
-      "learning_rate": 1.7167702696766877e-05,
-      "loss": 1.173,
+      "epoch": 0.30659731727347384,
+      "grad_norm": 0.39004513777273,
+      "learning_rate": 1.751380119950311e-05,
+      "loss": 1.4024,
       "step": 1120
     },
     {
-      "epoch": 0.32253440366972475,
-      "grad_norm": 4.037999124478957,
-      "learning_rate": 1.7132718831817093e-05,
-      "loss": 1.1695,
+      "epoch": 0.30796605529701615,
+      "grad_norm": 0.3955440868581025,
+      "learning_rate": 1.7482181578002837e-05,
+      "loss": 1.3667,
       "step": 1125
     },
     {
-      "epoch": 0.32396788990825687,
-      "grad_norm": 3.2060752585491086,
-      "learning_rate": 1.7097556354584526e-05,
-      "loss": 1.1464,
+      "epoch": 0.30933479332055847,
+      "grad_norm": 0.41261816602720663,
+      "learning_rate": 1.745039108573793e-05,
+      "loss": 1.357,
       "step": 1130
     },
     {
-      "epoch": 0.325401376146789,
-      "grad_norm": 3.72775795176933,
-      "learning_rate": 1.7062216145581997e-05,
-      "loss": 1.1237,
+      "epoch": 0.3107035313441007,
+      "grad_norm": 0.37732770567847945,
+      "learning_rate": 1.7418430448708644e-05,
+      "loss": 1.3337,
       "step": 1135
     },
     {
-      "epoch": 0.3268348623853211,
-      "grad_norm": 3.783961044208216,
-      "learning_rate": 1.7026699089772937e-05,
-      "loss": 1.1899,
+      "epoch": 0.31207226936764304,
+      "grad_norm": 0.3764972884033891,
+      "learning_rate": 1.738630039680083e-05,
+      "loss": 1.3556,
       "step": 1140
     },
     {
-      "epoch": 0.32826834862385323,
-      "grad_norm": 3.5936905948051323,
-      "learning_rate": 1.699100607654926e-05,
-      "loss": 1.176,
+      "epoch": 0.3134410073911853,
+      "grad_norm": 0.40764805847336166,
+      "learning_rate": 1.7354001663769278e-05,
+      "loss": 1.3679,
       "step": 1145
     },
     {
-      "epoch": 0.3297018348623853,
-      "grad_norm": 3.5296084880895156,
-      "learning_rate": 1.6955137999709075e-05,
-      "loss": 1.1445,
+      "epoch": 0.3148097454147276,
+      "grad_norm": 0.3794101029629702,
+      "learning_rate": 1.7321534987220942e-05,
+      "loss": 1.3176,
       "step": 1150
     },
     {
-      "epoch": 0.3311353211009174,
-      "grad_norm": 3.278850491610551,
-      "learning_rate": 1.6919095757434288e-05,
-      "loss": 1.2269,
+      "epoch": 0.31617848343826993,
+      "grad_norm": 0.40248164169870704,
+      "learning_rate": 1.728890110859812e-05,
+      "loss": 1.3458,
       "step": 1155
     },
     {
-      "epoch": 0.33256880733944955,
-      "grad_norm": 3.6404322855144247,
-      "learning_rate": 1.6882880252268156e-05,
-      "loss": 1.1836,
+      "epoch": 0.3175472214618122,
+      "grad_norm": 0.39012890957455026,
+      "learning_rate": 1.7256100773161492e-05,
+      "loss": 1.296,
       "step": 1160
     },
     {
-      "epoch": 0.33400229357798167,
-      "grad_norm": 3.437933045156408,
-      "learning_rate": 1.6846492391092625e-05,
-      "loss": 1.2295,
+      "epoch": 0.3189159594853545,
+      "grad_norm": 0.39856381680118136,
+      "learning_rate": 1.7223134729973134e-05,
+      "loss": 1.3614,
       "step": 1165
     },
     {
-      "epoch": 0.33543577981651373,
-      "grad_norm": 3.5611595737623167,
-      "learning_rate": 1.680993308510568e-05,
-      "loss": 1.1862,
+      "epoch": 0.3202846975088968,
+      "grad_norm": 0.40493526767721577,
+      "learning_rate": 1.7190003731879375e-05,
+      "loss": 1.3533,
       "step": 1170
     },
     {
-      "epoch": 0.33686926605504586,
-      "grad_norm": 4.312054532006292,
-      "learning_rate": 1.6773203249798482e-05,
-      "loss": 1.1602,
+      "epoch": 0.3216534355324391,
+      "grad_norm": 0.38868285184435786,
+      "learning_rate": 1.715670853549364e-05,
+      "loss": 1.3377,
       "step": 1175
     },
     {
-      "epoch": 0.338302752293578,
-      "grad_norm": 3.598611274354031,
-      "learning_rate": 1.6736303804932475e-05,
-      "loss": 1.1825,
+      "epoch": 0.3230221735559814,
+      "grad_norm": 0.3840977803535061,
+      "learning_rate": 1.7123249901179142e-05,
+      "loss": 1.3753,
       "step": 1180
     },
     {
-      "epoch": 0.3397362385321101,
-      "grad_norm": 3.779933724951351,
-      "learning_rate": 1.6699235674516334e-05,
-      "loss": 1.1847,
+      "epoch": 0.32439091157952366,
+      "grad_norm": 0.37510550805682613,
+      "learning_rate": 1.708962859303154e-05,
+      "loss": 1.3557,
       "step": 1185
     },
     {
-      "epoch": 0.3411697247706422,
-      "grad_norm": 3.590038112136874,
-      "learning_rate": 1.666199978678283e-05,
-      "loss": 1.1984,
+      "epoch": 0.32575964960306597,
+      "grad_norm": 0.39070855119805253,
+      "learning_rate": 1.7055845378861476e-05,
+      "loss": 1.3584,
       "step": 1190
     },
     {
-      "epoch": 0.3426032110091743,
-      "grad_norm": 4.110223116861788,
-      "learning_rate": 1.6624597074165597e-05,
-      "loss": 1.2468,
+      "epoch": 0.3271283876266083,
+      "grad_norm": 0.39021160718672115,
+      "learning_rate": 1.7021901030177036e-05,
+      "loss": 1.3399,
       "step": 1195
     },
     {
-      "epoch": 0.3440366972477064,
-      "grad_norm": 3.603557931052651,
-      "learning_rate": 1.6587028473275772e-05,
-      "loss": 1.1962,
+      "epoch": 0.32849712565015055,
+      "grad_norm": 0.3983342506003189,
+      "learning_rate": 1.698779632216615e-05,
+      "loss": 1.3965,
       "step": 1200
     },
     {
-      "epoch": 0.34547018348623854,
-      "grad_norm": 3.209591671498054,
-      "learning_rate": 1.6549294924878532e-05,
-      "loss": 1.1714,
+      "epoch": 0.32986586367369286,
+      "grad_norm": 0.38690562360273995,
+      "learning_rate": 1.6953532033678874e-05,
+      "loss": 1.4221,
       "step": 1205
     },
     {
-      "epoch": 0.34690366972477066,
-      "grad_norm": 3.6243010245754337,
-      "learning_rate": 1.651139737386957e-05,
-      "loss": 1.1733,
+      "epoch": 0.3312346016972352,
+      "grad_norm": 0.39029540911089444,
+      "learning_rate": 1.69191089472096e-05,
+      "loss": 1.3201,
       "step": 1210
     },
     {
-      "epoch": 0.3483371559633027,
-      "grad_norm": 3.9078886802263697,
-      "learning_rate": 1.6473336769251388e-05,
-      "loss": 1.1715,
+      "epoch": 0.33260333972077744,
+      "grad_norm": 0.40512131600219015,
+      "learning_rate": 1.688452784887921e-05,
+      "loss": 1.3495,
       "step": 1215
     },
     {
-      "epoch": 0.34977064220183485,
-      "grad_norm": 3.6344742284976306,
-      "learning_rate": 1.6435114064109575e-05,
-      "loss": 1.1244,
+      "epoch": 0.33397207774431975,
+      "grad_norm": 0.3804586676740124,
+      "learning_rate": 1.684978952841709e-05,
+      "loss": 1.3471,
       "step": 1220
     },
     {
-      "epoch": 0.35120412844036697,
-      "grad_norm": 3.6952536271050964,
-      "learning_rate": 1.6396730215588913e-05,
-      "loss": 1.2089,
+      "epoch": 0.335340815767862,
+      "grad_norm": 0.38428741373037245,
+      "learning_rate": 1.681489477914312e-05,
+      "loss": 1.3196,
       "step": 1225
     },
     {
-      "epoch": 0.3526376146788991,
-      "grad_norm": 3.429886933846167,
-      "learning_rate": 1.6358186184869417e-05,
-      "loss": 1.1892,
+      "epoch": 0.3367095537914043,
+      "grad_norm": 0.384047691017285,
+      "learning_rate": 1.677984439794954e-05,
+      "loss": 1.3303,
       "step": 1230
     },
     {
-      "epoch": 0.3540711009174312,
-      "grad_norm": 3.093455767598654,
-      "learning_rate": 1.631948293714227e-05,
-      "loss": 1.2106,
+      "epoch": 0.33807829181494664,
+      "grad_norm": 0.3839900646299871,
+      "learning_rate": 1.6744639185282784e-05,
+      "loss": 1.2792,
       "step": 1235
     },
     {
-      "epoch": 0.3555045871559633,
-      "grad_norm": 4.038152481713509,
-      "learning_rate": 1.6280621441585647e-05,
-      "loss": 1.1697,
+      "epoch": 0.3394470298384889,
+      "grad_norm": 0.3884042208952324,
+      "learning_rate": 1.670927994512514e-05,
+      "loss": 1.3223,
       "step": 1240
     },
     {
-      "epoch": 0.3569380733944954,
-      "grad_norm": 3.6999557746923095,
-      "learning_rate": 1.6241602671340448e-05,
-      "loss": 1.2376,
+      "epoch": 0.3408157678620312,
+      "grad_norm": 0.3911385295369844,
+      "learning_rate": 1.667376748497646e-05,
+      "loss": 1.3546,
       "step": 1245
     },
     {
-      "epoch": 0.3583715596330275,
-      "grad_norm": 3.6041781324392868,
-      "learning_rate": 1.6202427603485933e-05,
-      "loss": 1.1327,
+      "epoch": 0.3421845058855735,
+      "grad_norm": 0.3807410219808285,
+      "learning_rate": 1.6638102615835658e-05,
+      "loss": 1.3148,
       "step": 1250
     },
     {
-      "epoch": 0.35980504587155965,
-      "grad_norm": 3.4145472213007473,
-      "learning_rate": 1.6163097219015245e-05,
-      "loss": 1.1372,
+      "epoch": 0.3435532439091158,
+      "grad_norm": 0.39423887853753,
+      "learning_rate": 1.6602286152182236e-05,
+      "loss": 1.361,
       "step": 1255
     },
     {
-      "epoch": 0.36123853211009177,
-      "grad_norm": 3.429021637188865,
-      "learning_rate": 1.6123612502810865e-05,
-      "loss": 1.1646,
+      "epoch": 0.3449219819326581,
+      "grad_norm": 0.39187255483162625,
+      "learning_rate": 1.6566318911957647e-05,
+      "loss": 1.339,
       "step": 1260
     },
     {
-      "epoch": 0.36267201834862384,
-      "grad_norm": 3.6122171856293708,
-      "learning_rate": 1.6083974443619922e-05,
-      "loss": 1.2031,
+      "epoch": 0.34629071995620037,
+      "grad_norm": 0.38859511955741677,
+      "learning_rate": 1.6530201716546647e-05,
+      "loss": 1.3556,
       "step": 1265
     },
     {
-      "epoch": 0.36410550458715596,
-      "grad_norm": 3.2252592769247617,
-      "learning_rate": 1.6044184034029445e-05,
-      "loss": 1.1669,
+      "epoch": 0.3476594579797427,
+      "grad_norm": 0.3993643770992966,
+      "learning_rate": 1.649393539075851e-05,
+      "loss": 1.3476,
       "step": 1270
     },
     {
-      "epoch": 0.3655389908256881,
-      "grad_norm": 3.513524941132868,
-      "learning_rate": 1.6004242270441523e-05,
-      "loss": 1.1871,
+      "epoch": 0.349028196003285,
+      "grad_norm": 0.39035176617866785,
+      "learning_rate": 1.6457520762808217e-05,
+      "loss": 1.3228,
       "step": 1275
     },
     {
-      "epoch": 0.3669724770642202,
-      "grad_norm": 3.5159805187675115,
-      "learning_rate": 1.596415015304833e-05,
-      "loss": 1.1851,
+      "epoch": 0.35039693402682726,
+      "grad_norm": 0.3830707286890397,
+      "learning_rate": 1.6420958664297514e-05,
+      "loss": 1.3094,
       "step": 1280
     },
     {
-      "epoch": 0.36840596330275227,
-      "grad_norm": 3.3054440445349837,
-      "learning_rate": 1.5923908685807087e-05,
-      "loss": 1.1074,
+      "epoch": 0.35176567205036957,
+      "grad_norm": 0.36321444690348953,
+      "learning_rate": 1.638424993019595e-05,
+      "loss": 1.3853,
       "step": 1285
     },
     {
-      "epoch": 0.3698394495412844,
-      "grad_norm": 3.6946632308744136,
-      "learning_rate": 1.588351887641494e-05,
-      "loss": 1.1596,
+      "epoch": 0.35313441007391183,
+      "grad_norm": 0.371835147945527,
+      "learning_rate": 1.634739539882178e-05,
+      "loss": 1.3477,
       "step": 1290
     },
     {
-      "epoch": 0.3712729357798165,
-      "grad_norm": 3.3924561467755785,
-      "learning_rate": 1.5842981736283686e-05,
-      "loss": 1.2262,
+      "epoch": 0.35450314809745415,
+      "grad_norm": 0.3906744323143432,
+      "learning_rate": 1.6310395911822848e-05,
+      "loss": 1.3149,
       "step": 1295
     },
     {
-      "epoch": 0.37270642201834864,
-      "grad_norm": 3.623083608828368,
-      "learning_rate": 1.5802298280514487e-05,
-      "loss": 1.1823,
+      "epoch": 0.35587188612099646,
+      "grad_norm": 0.37408867437468385,
+      "learning_rate": 1.6273252314157352e-05,
+      "loss": 1.3135,
       "step": 1300
     },
     {
-      "epoch": 0.37413990825688076,
-      "grad_norm": 3.0148932024272437,
-      "learning_rate": 1.5761469527872427e-05,
-      "loss": 1.1513,
+      "epoch": 0.3572406241445387,
+      "grad_norm": 0.37403440281560435,
+      "learning_rate": 1.6235965454074535e-05,
+      "loss": 1.2301,
       "step": 1305
     },
     {
-      "epoch": 0.3755733944954128,
-      "grad_norm": 10.198288104380973,
-      "learning_rate": 1.572049650076101e-05,
-      "loss": 1.1498,
+      "epoch": 0.35860936216808104,
+      "grad_norm": 0.3867708656040568,
+      "learning_rate": 1.619853618309535e-05,
+      "loss": 1.3716,
       "step": 1310
     },
     {
-      "epoch": 0.37700688073394495,
-      "grad_norm": 3.623884808969302,
-      "learning_rate": 1.5679380225196546e-05,
-      "loss": 1.1857,
+      "epoch": 0.3599781001916233,
+      "grad_norm": 0.3885328796144267,
+      "learning_rate": 1.6160965355992966e-05,
+      "loss": 1.366,
       "step": 1315
     },
     {
-      "epoch": 0.37844036697247707,
-      "grad_norm": 21.814389727799938,
-      "learning_rate": 1.5638121730782486e-05,
-      "loss": 1.1748,
+      "epoch": 0.3613468382151656,
+      "grad_norm": 0.403719948973195,
+      "learning_rate": 1.6123253830773293e-05,
+      "loss": 1.3661,
       "step": 1320
     },
     {
-      "epoch": 0.3798738532110092,
-      "grad_norm": 3.7586731538464684,
-      "learning_rate": 1.5596722050683598e-05,
-      "loss": 1.2024,
+      "epoch": 0.3627155762387079,
+      "grad_norm": 0.3827196192323514,
+      "learning_rate": 1.6085402468655356e-05,
+      "loss": 1.3567,
       "step": 1325
     },
     {
-      "epoch": 0.38130733944954126,
-      "grad_norm": 3.914930197959616,
-      "learning_rate": 1.555518222160013e-05,
-      "loss": 1.1873,
+      "epoch": 0.3640843142622502,
+      "grad_norm": 0.3968500270407259,
+      "learning_rate": 1.6047412134051645e-05,
+      "loss": 1.4044,
       "step": 1330
     },
     {
-      "epoch": 0.3827408256880734,
-      "grad_norm": 3.645421580899915,
-      "learning_rate": 1.551350328374184e-05,
-      "loss": 1.1252,
+      "epoch": 0.3654530522857925,
+      "grad_norm": 0.38847711656237954,
+      "learning_rate": 1.6009283694548365e-05,
+      "loss": 1.3591,
       "step": 1335
     },
     {
-      "epoch": 0.3841743119266055,
-      "grad_norm": 3.1973889011870376,
-      "learning_rate": 1.5471686280801933e-05,
-      "loss": 1.1839,
+      "epoch": 0.3668217903093348,
+      "grad_norm": 0.4036707276972636,
+      "learning_rate": 1.5971018020885623e-05,
+      "loss": 1.3916,
       "step": 1340
     },
     {
-      "epoch": 0.3856077981651376,
-      "grad_norm": 3.5501087765790915,
-      "learning_rate": 1.5429732259930955e-05,
-      "loss": 1.1232,
+      "epoch": 0.3681905283328771,
+      "grad_norm": 0.39526777473833846,
+      "learning_rate": 1.593261598693755e-05,
+      "loss": 1.3478,
       "step": 1345
     },
     {
-      "epoch": 0.38704128440366975,
-      "grad_norm": 3.3946063622500016,
-      "learning_rate": 1.538764227171054e-05,
-      "loss": 1.1768,
+      "epoch": 0.3695592663564194,
+      "grad_norm": 0.3837685918142903,
+      "learning_rate": 1.5894078469692343e-05,
+      "loss": 1.3604,
       "step": 1350
     },
     {
-      "epoch": 0.3884747706422018,
-      "grad_norm": 3.3089407247339975,
-      "learning_rate": 1.5345417370127123e-05,
-      "loss": 1.1697,
+      "epoch": 0.37092800437996165,
+      "grad_norm": 0.4184389090754904,
+      "learning_rate": 1.585540634923223e-05,
+      "loss": 1.4042,
       "step": 1355
     },
     {
-      "epoch": 0.38990825688073394,
-      "grad_norm": 4.435494305269373,
-      "learning_rate": 1.5303058612545534e-05,
-      "loss": 1.1301,
+      "epoch": 0.37229674240350397,
+      "grad_norm": 0.4132734637825316,
+      "learning_rate": 1.5816600508713372e-05,
+      "loss": 1.3901,
       "step": 1360
     },
     {
-      "epoch": 0.39134174311926606,
-      "grad_norm": 3.1427532195841343,
-      "learning_rate": 1.5260567059682535e-05,
-      "loss": 1.1857,
+      "epoch": 0.3736654804270463,
+      "grad_norm": 0.3876929955451876,
+      "learning_rate": 1.5777661834345708e-05,
+      "loss": 1.3296,
       "step": 1365
     },
     {
-      "epoch": 0.3927752293577982,
-      "grad_norm": 3.5070367063445684,
-      "learning_rate": 1.521794377558024e-05,
-      "loss": 1.1294,
+      "epoch": 0.37503421845058854,
+      "grad_norm": 0.39294627819783745,
+      "learning_rate": 1.57385912153727e-05,
+      "loss": 1.3581,
       "step": 1370
     },
     {
-      "epoch": 0.39420871559633025,
-      "grad_norm": 3.6077040833104896,
-      "learning_rate": 1.5175189827579489e-05,
-      "loss": 1.1469,
+      "epoch": 0.37640295647413086,
+      "grad_norm": 0.3875865415567979,
+      "learning_rate": 1.5699389544051028e-05,
+      "loss": 1.3167,
       "step": 1375
     },
     {
-      "epoch": 0.3956422018348624,
-      "grad_norm": 3.401651617393664,
-      "learning_rate": 1.5132306286293096e-05,
-      "loss": 1.1383,
+      "epoch": 0.37777169449767317,
+      "grad_norm": 0.3863898154888909,
+      "learning_rate": 1.566005771563023e-05,
+      "loss": 1.3694,
       "step": 1380
     },
     {
-      "epoch": 0.3970756880733945,
-      "grad_norm": 3.3826428424139925,
-      "learning_rate": 1.5089294225579077e-05,
-      "loss": 1.1887,
+      "epoch": 0.37914043252121543,
+      "grad_norm": 0.38577432820095847,
+      "learning_rate": 1.5620596628332242e-05,
+      "loss": 1.323,
       "step": 1385
     },
     {
-      "epoch": 0.3985091743119266,
-      "grad_norm": 3.2830968866458345,
-      "learning_rate": 1.5046154722513718e-05,
-      "loss": 1.15,
+      "epoch": 0.38050917054475775,
+      "grad_norm": 0.3971392209603939,
+      "learning_rate": 1.5581007183330877e-05,
+      "loss": 1.3432,
       "step": 1390
     },
     {
-      "epoch": 0.39994266055045874,
-      "grad_norm": 3.2832815737637113,
-      "learning_rate": 1.5002888857364624e-05,
-      "loss": 1.1811,
+      "epoch": 0.3818779085683,
+      "grad_norm": 0.40311344444707337,
+      "learning_rate": 1.554129028473127e-05,
+      "loss": 1.3802,
       "step": 1395
     },
     {
-      "epoch": 0.4013761467889908,
-      "grad_norm": 3.222829419424945,
-      "learning_rate": 1.4959497713563677e-05,
-      "loss": 1.1696,
+      "epoch": 0.3832466465918423,
+      "grad_norm": 0.40527858159368374,
+      "learning_rate": 1.5501446839549207e-05,
+      "loss": 1.3445,
       "step": 1400
     },
     {
-      "epoch": 0.40280963302752293,
-      "grad_norm": 3.739181624579295,
-      "learning_rate": 1.4915982377679885e-05,
-      "loss": 1.1978,
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.4025758901452214,
+      "learning_rate": 1.5461477757690424e-05,
+      "loss": 1.3321,
       "step": 1405
     },
     {
-      "epoch": 0.40424311926605505,
-      "grad_norm": 3.383948957829039,
-      "learning_rate": 1.4872343939392189e-05,
-      "loss": 1.1724,
+      "epoch": 0.3859841226389269,
+      "grad_norm": 0.3855393398790419,
+      "learning_rate": 1.542138395192983e-05,
+      "loss": 1.3405,
       "step": 1410
     },
     {
-      "epoch": 0.4056766055045872,
-      "grad_norm": 3.7382791698251427,
-      "learning_rate": 1.482858349146216e-05,
-      "loss": 1.1471,
+      "epoch": 0.3873528606624692,
+      "grad_norm": 0.38944762738859545,
+      "learning_rate": 1.538116633789065e-05,
+      "loss": 1.3289,
       "step": 1415
     },
     {
-      "epoch": 0.40711009174311924,
-      "grad_norm": 3.510878161789182,
-      "learning_rate": 1.4784702129706655e-05,
-      "loss": 1.1337,
+      "epoch": 0.38872159868601147,
+      "grad_norm": 0.4028684971128393,
+      "learning_rate": 1.5340825834023526e-05,
+      "loss": 1.3798,
       "step": 1420
     },
     {
-      "epoch": 0.40854357798165136,
-      "grad_norm": 3.505598802015759,
-      "learning_rate": 1.474070095297036e-05,
-      "loss": 1.187,
+      "epoch": 0.3900903367095538,
+      "grad_norm": 0.37654892707807947,
+      "learning_rate": 1.530036336158553e-05,
+      "loss": 1.3432,
       "step": 1425
     },
     {
-      "epoch": 0.4099770642201835,
-      "grad_norm": 3.2069715268799204,
-      "learning_rate": 1.469658106309828e-05,
-      "loss": 1.1136,
+      "epoch": 0.3914590747330961,
+      "grad_norm": 0.3986727790830818,
+      "learning_rate": 1.5259779844619152e-05,
+      "loss": 1.3422,
       "step": 1430
     },
     {
-      "epoch": 0.4114105504587156,
-      "grad_norm": 4.1812266615604345,
-      "learning_rate": 1.465234356490815e-05,
-      "loss": 1.1718,
+      "epoch": 0.39282781275663836,
+      "grad_norm": 0.4029735444716236,
+      "learning_rate": 1.5219076209931159e-05,
+      "loss": 1.3136,
       "step": 1435
     },
     {
-      "epoch": 0.41284403669724773,
-      "grad_norm": 3.3985858466934338,
-      "learning_rate": 1.4607989566162761e-05,
-      "loss": 1.1414,
+      "epoch": 0.3941965507801807,
+      "grad_norm": 0.41012078556001347,
+      "learning_rate": 1.5178253387071458e-05,
+      "loss": 1.4002,
       "step": 1440
     },
     {
-      "epoch": 0.4142775229357798,
-      "grad_norm": 3.6381472550431093,
-      "learning_rate": 1.4563520177542226e-05,
-      "loss": 1.2166,
+      "epoch": 0.395565288803723,
+      "grad_norm": 0.40771281882367433,
+      "learning_rate": 1.5137312308311857e-05,
+      "loss": 1.3684,
       "step": 1445
     },
     {
-      "epoch": 0.4157110091743119,
-      "grad_norm": 3.268767608874384,
-      "learning_rate": 1.451893651261617e-05,
-      "loss": 1.1695,
+      "epoch": 0.39693402682726525,
+      "grad_norm": 0.38445172076552614,
+      "learning_rate": 1.5096253908624778e-05,
+      "loss": 1.3137,
       "step": 1450
     },
     {
-      "epoch": 0.41714449541284404,
-      "grad_norm": 4.519674348990799,
-      "learning_rate": 1.4474239687815838e-05,
-      "loss": 1.189,
+      "epoch": 0.39830276485080757,
+      "grad_norm": 0.3869608798791883,
+      "learning_rate": 1.5055079125661908e-05,
+      "loss": 1.2812,
       "step": 1455
     },
     {
-      "epoch": 0.41857798165137616,
-      "grad_norm": 3.594106437257434,
-      "learning_rate": 1.4429430822406138e-05,
-      "loss": 1.1956,
+      "epoch": 0.3996715028743498,
+      "grad_norm": 0.3987871366347409,
+      "learning_rate": 1.5013788899732775e-05,
+      "loss": 1.3394,
       "step": 1460
     },
     {
-      "epoch": 0.42001146788990823,
-      "grad_norm": 3.838114634468203,
-      "learning_rate": 1.4384511038457624e-05,
-      "loss": 1.1348,
+      "epoch": 0.40104024089789214,
+      "grad_norm": 0.3970490183809672,
+      "learning_rate": 1.4972384173783284e-05,
+      "loss": 1.3544,
       "step": 1465
     },
     {
-      "epoch": 0.42144495412844035,
-      "grad_norm": 3.3721934098468083,
-      "learning_rate": 1.4339481460818385e-05,
-      "loss": 1.135,
+      "epoch": 0.40240897892143446,
+      "grad_norm": 0.3925074770480503,
+      "learning_rate": 1.493086589337418e-05,
+      "loss": 1.3294,
       "step": 1470
     },
     {
-      "epoch": 0.4228784403669725,
-      "grad_norm": 3.3513743473211512,
-      "learning_rate": 1.429434321708588e-05,
-      "loss": 1.1443,
+      "epoch": 0.4037777169449767,
+      "grad_norm": 0.40973551864547075,
+      "learning_rate": 1.4889235006659448e-05,
+      "loss": 1.3675,
       "step": 1475
     },
     {
-      "epoch": 0.4243119266055046,
-      "grad_norm": 3.3279635877417304,
-      "learning_rate": 1.4249097437578712e-05,
-      "loss": 1.1574,
+      "epoch": 0.40514645496851903,
+      "grad_norm": 0.3812235133361238,
+      "learning_rate": 1.484749246436468e-05,
+      "loss": 1.3288,
       "step": 1480
     },
     {
-      "epoch": 0.4257454128440367,
-      "grad_norm": 3.5340814143703896,
-      "learning_rate": 1.4203745255308306e-05,
-      "loss": 1.1607,
+      "epoch": 0.40651519299206135,
+      "grad_norm": 0.3827389411340829,
+      "learning_rate": 1.4805639219765337e-05,
+      "loss": 1.4128,
       "step": 1485
     },
     {
-      "epoch": 0.4271788990825688,
-      "grad_norm": 3.3296679892148995,
-      "learning_rate": 1.4158287805950557e-05,
-      "loss": 1.199,
+      "epoch": 0.4078839310156036,
+      "grad_norm": 0.37343264715636293,
+      "learning_rate": 1.476367622866499e-05,
+      "loss": 1.3419,
       "step": 1490
     },
     {
-      "epoch": 0.4286123853211009,
-      "grad_norm": 3.401879333323161,
-      "learning_rate": 1.411272622781737e-05,
-      "loss": 1.1364,
+      "epoch": 0.4092526690391459,
+      "grad_norm": 0.4116326320287375,
+      "learning_rate": 1.4721604449373505e-05,
+      "loss": 1.297,
       "step": 1495
     },
     {
-      "epoch": 0.43004587155963303,
-      "grad_norm": 3.131916232702695,
-      "learning_rate": 1.4067061661828176e-05,
-      "loss": 1.1472,
+      "epoch": 0.4106214070626882,
+      "grad_norm": 0.38955441577160277,
+      "learning_rate": 1.4679424842685137e-05,
+      "loss": 1.3138,
       "step": 1500
     },
     {
-      "epoch": 0.43147935779816515,
-      "grad_norm": 3.543349954734912,
-      "learning_rate": 1.4021295251481347e-05,
-      "loss": 1.1813,
+      "epoch": 0.4119901450862305,
+      "grad_norm": 0.38444568811838686,
+      "learning_rate": 1.4637138371856601e-05,
+      "loss": 1.3284,
       "step": 1505
     },
     {
-      "epoch": 0.4329128440366973,
-      "grad_norm": 3.309506087624542,
-      "learning_rate": 1.3975428142825562e-05,
-      "loss": 1.1581,
+      "epoch": 0.4133588831097728,
+      "grad_norm": 0.3938144538178601,
+      "learning_rate": 1.4594746002585072e-05,
+      "loss": 1.3498,
       "step": 1510
     },
     {
-      "epoch": 0.43434633027522934,
-      "grad_norm": 3.9281940551258097,
-      "learning_rate": 1.392946148443112e-05,
-      "loss": 1.1808,
+      "epoch": 0.41472762113331507,
+      "grad_norm": 0.39696555787686244,
+      "learning_rate": 1.4552248702986127e-05,
+      "loss": 1.3524,
       "step": 1515
     },
     {
-      "epoch": 0.43577981651376146,
-      "grad_norm": 3.8214812468244346,
-      "learning_rate": 1.3883396427361169e-05,
-      "loss": 1.1616,
+      "epoch": 0.4160963591568574,
+      "grad_norm": 0.3802714110333375,
+      "learning_rate": 1.4509647443571643e-05,
+      "loss": 1.288,
       "step": 1520
     },
     {
-      "epoch": 0.4372133027522936,
-      "grad_norm": 3.997270424750147,
-      "learning_rate": 1.383723412514288e-05,
-      "loss": 1.1113,
+      "epoch": 0.41746509718039965,
+      "grad_norm": 0.38532238774246486,
+      "learning_rate": 1.446694319722763e-05,
+      "loss": 1.3656,
       "step": 1525
     },
     {
-      "epoch": 0.4386467889908257,
-      "grad_norm": 3.437008178425516,
-      "learning_rate": 1.3790975733738576e-05,
-      "loss": 1.1713,
+      "epoch": 0.41883383520394196,
+      "grad_norm": 0.3866193343216763,
+      "learning_rate": 1.4424136939192009e-05,
+      "loss": 1.3696,
       "step": 1530
     },
     {
-      "epoch": 0.4400802752293578,
-      "grad_norm": 3.3811402901395984,
-      "learning_rate": 1.3744622411516758e-05,
-      "loss": 1.1347,
+      "epoch": 0.4202025732274843,
+      "grad_norm": 0.37501111401637777,
+      "learning_rate": 1.4381229647032346e-05,
+      "loss": 1.3512,
       "step": 1535
     },
     {
-      "epoch": 0.4415137614678899,
-      "grad_norm": 3.3962664452333993,
-      "learning_rate": 1.3698175319223133e-05,
-      "loss": 1.1347,
+      "epoch": 0.42157131125102654,
+      "grad_norm": 0.3700012749063157,
+      "learning_rate": 1.4338222300623533e-05,
+      "loss": 1.3092,
       "step": 1540
     },
     {
-      "epoch": 0.442947247706422,
-      "grad_norm": 3.6162635705687127,
-      "learning_rate": 1.3651635619951509e-05,
-      "loss": 1.1574,
+      "epoch": 0.42294004927456885,
+      "grad_norm": 0.38806617391675347,
+      "learning_rate": 1.4295115882125393e-05,
+      "loss": 1.3471,
       "step": 1545
     },
     {
-      "epoch": 0.44438073394495414,
-      "grad_norm": 3.502705575568886,
-      "learning_rate": 1.360500447911471e-05,
-      "loss": 1.0691,
+      "epoch": 0.42430878729811117,
+      "grad_norm": 0.39116631927262135,
+      "learning_rate": 1.4251911375960261e-05,
+      "loss": 1.4043,
       "step": 1550
     },
     {
-      "epoch": 0.44581422018348627,
-      "grad_norm": 3.473102068128584,
-      "learning_rate": 1.3558283064415357e-05,
-      "loss": 1.1018,
+      "epoch": 0.4256775253216534,
+      "grad_norm": 0.3907489868883888,
+      "learning_rate": 1.4208609768790513e-05,
+      "loss": 1.3476,
       "step": 1555
     },
     {
-      "epoch": 0.44724770642201833,
-      "grad_norm": 3.5587976176438656,
-      "learning_rate": 1.3511472545816648e-05,
-      "loss": 1.1293,
+      "epoch": 0.42704626334519574,
+      "grad_norm": 0.40188011896979126,
+      "learning_rate": 1.4165212049496013e-05,
+      "loss": 1.3398,
       "step": 1560
     },
     {
-      "epoch": 0.44868119266055045,
-      "grad_norm": 3.613085860404909,
-      "learning_rate": 1.3464574095513058e-05,
-      "loss": 1.1685,
+      "epoch": 0.428415001368738,
+      "grad_norm": 0.39583117671628015,
+      "learning_rate": 1.4121719209151545e-05,
+      "loss": 1.3827,
       "step": 1565
     },
     {
-      "epoch": 0.4501146788990826,
-      "grad_norm": 4.06622018524846,
-      "learning_rate": 1.3417588887900978e-05,
-      "loss": 1.1067,
+      "epoch": 0.4297837393922803,
+      "grad_norm": 0.3948036088686505,
+      "learning_rate": 1.4078132241004174e-05,
+      "loss": 1.3509,
       "step": 1570
     },
     {
-      "epoch": 0.4515481651376147,
-      "grad_norm": 3.4649254770088698,
-      "learning_rate": 1.3370518099549315e-05,
-      "loss": 1.1026,
+      "epoch": 0.43115247741582263,
+      "grad_norm": 0.3844018510427473,
+      "learning_rate": 1.4034452140450561e-05,
+      "loss": 1.3619,
       "step": 1575
     },
     {
-      "epoch": 0.45298165137614677,
-      "grad_norm": 2.952840725725241,
-      "learning_rate": 1.3323362909170018e-05,
-      "loss": 1.1283,
+      "epoch": 0.4325212154393649,
+      "grad_norm": 0.40654356409937986,
+      "learning_rate": 1.3990679905014235e-05,
+      "loss": 1.2983,
       "step": 1580
     },
     {
-      "epoch": 0.4544151376146789,
-      "grad_norm": 3.253136251920175,
-      "learning_rate": 1.3276124497588585e-05,
-      "loss": 1.2414,
+      "epoch": 0.4338899534629072,
+      "grad_norm": 0.401731234750035,
+      "learning_rate": 1.3946816534322815e-05,
+      "loss": 1.3573,
       "step": 1585
     },
     {
-      "epoch": 0.455848623853211,
-      "grad_norm": 3.162846986211554,
-      "learning_rate": 1.3228804047714462e-05,
-      "loss": 1.1169,
+      "epoch": 0.43525869148644947,
+      "grad_norm": 0.40604771069832035,
+      "learning_rate": 1.3902863030085176e-05,
+      "loss": 1.333,
       "step": 1590
     },
     {
-      "epoch": 0.45728211009174313,
-      "grad_norm": 3.3272179350821505,
-      "learning_rate": 1.3181402744511446e-05,
-      "loss": 1.1756,
+      "epoch": 0.4366274295099918,
+      "grad_norm": 0.3821888564061559,
+      "learning_rate": 1.3858820396068572e-05,
+      "loss": 1.3062,
       "step": 1595
     },
     {
-      "epoch": 0.45871559633027525,
-      "grad_norm": 3.1731644889698183,
-      "learning_rate": 1.3133921774968001e-05,
-      "loss": 1.1305,
+      "epoch": 0.4379961675335341,
+      "grad_norm": 0.3992586017031061,
+      "learning_rate": 1.3814689638075725e-05,
+      "loss": 1.3671,
       "step": 1600
     },
     {
-      "epoch": 0.4601490825688073,
-      "grad_norm": 3.1249681327250904,
-      "learning_rate": 1.3086362328067536e-05,
-      "loss": 1.1603,
+      "epoch": 0.43936490555707636,
+      "grad_norm": 0.4018110425948897,
+      "learning_rate": 1.3770471763921833e-05,
+      "loss": 1.3709,
       "step": 1605
     },
     {
-      "epoch": 0.46158256880733944,
-      "grad_norm": 3.376924777276381,
-      "learning_rate": 1.3038725594758632e-05,
-      "loss": 1.118,
+      "epoch": 0.4407336435806187,
+      "grad_norm": 0.387994622633552,
+      "learning_rate": 1.372616778341158e-05,
+      "loss": 1.3138,
       "step": 1610
     },
     {
-      "epoch": 0.46301605504587157,
-      "grad_norm": 3.193633263846461,
-      "learning_rate": 1.2991012767925224e-05,
-      "loss": 1.1337,
+      "epoch": 0.442102381604161,
+      "grad_norm": 0.3974604574879118,
+      "learning_rate": 1.3681778708316054e-05,
+      "loss": 1.4087,
       "step": 1615
     },
     {
-      "epoch": 0.4644495412844037,
-      "grad_norm": 3.0745547065753915,
-      "learning_rate": 1.2943225042356714e-05,
-      "loss": 1.1252,
+      "epoch": 0.44347111962770325,
+      "grad_norm": 0.39844593969419934,
+      "learning_rate": 1.3637305552349656e-05,
+      "loss": 1.3599,
       "step": 1620
     },
     {
-      "epoch": 0.46588302752293576,
-      "grad_norm": 7.556185912909887,
-      "learning_rate": 1.2895363614718082e-05,
-      "loss": 1.1517,
+      "epoch": 0.44483985765124556,
+      "grad_norm": 0.3857194594528982,
+      "learning_rate": 1.3592749331146941e-05,
+      "loss": 1.2897,
       "step": 1625
     },
     {
-      "epoch": 0.4673165137614679,
-      "grad_norm": 3.3618721338615787,
-      "learning_rate": 1.2847429683519879e-05,
-      "loss": 1.1321,
+      "epoch": 0.4462085956747878,
+      "grad_norm": 0.3750305160936592,
+      "learning_rate": 1.3548111062239432e-05,
+      "loss": 1.2819,
       "step": 1630
     },
     {
-      "epoch": 0.46875,
-      "grad_norm": 3.1654091942580904,
-      "learning_rate": 1.2799424449088246e-05,
-      "loss": 1.1475,
+      "epoch": 0.44757733369833014,
+      "grad_norm": 0.4007647993356723,
+      "learning_rate": 1.350339176503237e-05,
+      "loss": 1.3194,
       "step": 1635
     },
     {
-      "epoch": 0.4701834862385321,
-      "grad_norm": 3.3736619152907386,
-      "learning_rate": 1.2751349113534856e-05,
-      "loss": 1.1297,
+      "epoch": 0.44894607172187245,
+      "grad_norm": 0.4013887994126227,
+      "learning_rate": 1.3458592460781446e-05,
+      "loss": 1.3199,
       "step": 1640
     },
     {
-      "epoch": 0.47161697247706424,
-      "grad_norm": 3.711148861718645,
-      "learning_rate": 1.2703204880726788e-05,
-      "loss": 1.1654,
+      "epoch": 0.4503148097454147,
+      "grad_norm": 0.3981164766269796,
+      "learning_rate": 1.341371417256947e-05,
+      "loss": 1.3464,
       "step": 1645
     },
     {
-      "epoch": 0.4730504587155963,
-      "grad_norm": 3.0399296608693946,
-      "learning_rate": 1.2654992956256397e-05,
-      "loss": 1.1072,
+      "epoch": 0.451683547768957,
+      "grad_norm": 0.39477092552678616,
+      "learning_rate": 1.3368757925283015e-05,
+      "loss": 1.3529,
       "step": 1650
     },
     {
-      "epoch": 0.47448394495412843,
-      "grad_norm": 3.299482801946242,
-      "learning_rate": 1.2606714547411138e-05,
-      "loss": 1.1338,
+      "epoch": 0.45305228579249934,
+      "grad_norm": 0.3896314211042654,
+      "learning_rate": 1.3323724745589007e-05,
+      "loss": 1.3881,
       "step": 1655
     },
     {
-      "epoch": 0.47591743119266056,
-      "grad_norm": 4.109998693762478,
-      "learning_rate": 1.2558370863143298e-05,
-      "loss": 1.1517,
+      "epoch": 0.4544210238160416,
+      "grad_norm": 0.3954835469300208,
+      "learning_rate": 1.3278615661911274e-05,
+      "loss": 1.3124,
       "step": 1660
     },
     {
-      "epoch": 0.4773509174311927,
-      "grad_norm": 3.6028647178766167,
-      "learning_rate": 1.250996311403976e-05,
-      "loss": 1.2193,
+      "epoch": 0.4557897618395839,
+      "grad_norm": 0.3708379387210574,
+      "learning_rate": 1.3233431704407072e-05,
+      "loss": 1.2866,
       "step": 1665
     },
     {
-      "epoch": 0.47878440366972475,
-      "grad_norm": 3.088463522622337,
-      "learning_rate": 1.246149251229166e-05,
-      "loss": 1.1481,
+      "epoch": 0.4571584998631262,
+      "grad_norm": 0.38510327105983166,
+      "learning_rate": 1.318817390494355e-05,
+      "loss": 1.2968,
       "step": 1670
     },
     {
-      "epoch": 0.48021788990825687,
-      "grad_norm": 3.1966754732870033,
-      "learning_rate": 1.2412960271664046e-05,
-      "loss": 1.0923,
+      "epoch": 0.4585272378866685,
+      "grad_norm": 0.38362958034859684,
+      "learning_rate": 1.3142843297074182e-05,
+      "loss": 1.3155,
       "step": 1675
     },
     {
-      "epoch": 0.481651376146789,
-      "grad_norm": 3.5520419027791044,
-      "learning_rate": 1.2364367607465483e-05,
-      "loss": 1.131,
+      "epoch": 0.4598959759102108,
+      "grad_norm": 0.4082500339960413,
+      "learning_rate": 1.3097440916015179e-05,
+      "loss": 1.3646,
       "step": 1680
     },
     {
-      "epoch": 0.4830848623853211,
-      "grad_norm": 3.1911366000417103,
-      "learning_rate": 1.2315715736517624e-05,
-      "loss": 1.1524,
+      "epoch": 0.46126471393375307,
+      "grad_norm": 0.3927808793669975,
+      "learning_rate": 1.3051967798621834e-05,
+      "loss": 1.3165,
       "step": 1685
     },
     {
-      "epoch": 0.48451834862385323,
-      "grad_norm": 3.1232221206590847,
-      "learning_rate": 1.2267005877124721e-05,
-      "loss": 1.1336,
+      "epoch": 0.4626334519572954,
+      "grad_norm": 0.4002328011893471,
+      "learning_rate": 1.300642498336484e-05,
+      "loss": 1.3287,
       "step": 1690
     },
     {
-      "epoch": 0.4859518348623853,
-      "grad_norm": 3.1919229672837606,
-      "learning_rate": 1.2218239249043143e-05,
-      "loss": 1.1163,
+      "epoch": 0.46400218998083764,
+      "grad_norm": 0.38096569725907414,
+      "learning_rate": 1.2960813510306599e-05,
+      "loss": 1.319,
       "step": 1695
     },
     {
-      "epoch": 0.4873853211009174,
-      "grad_norm": 3.6466393865332223,
-      "learning_rate": 1.2169417073450805e-05,
-      "loss": 1.1113,
+      "epoch": 0.46537092800437996,
+      "grad_norm": 0.38869356480860534,
+      "learning_rate": 1.2915134421077433e-05,
+      "loss": 1.3763,
       "step": 1700
     },
     {
-      "epoch": 0.48881880733944955,
-      "grad_norm": 3.275240485720321,
-      "learning_rate": 1.2120540572916617e-05,
-      "loss": 1.1516,
+      "epoch": 0.4667396660279223,
+      "grad_norm": 0.38290589898614796,
+      "learning_rate": 1.2869388758851828e-05,
+      "loss": 1.3204,
       "step": 1705
     },
     {
-      "epoch": 0.49025229357798167,
-      "grad_norm": 3.350383352712665,
-      "learning_rate": 1.2071610971369842e-05,
-      "loss": 1.1564,
+      "epoch": 0.46810840405146453,
+      "grad_norm": 0.41880465304512116,
+      "learning_rate": 1.2823577568324604e-05,
+      "loss": 1.3858,
       "step": 1710
     },
     {
-      "epoch": 0.49168577981651373,
-      "grad_norm": 3.5074589174762236,
-      "learning_rate": 1.2022629494069466e-05,
-      "loss": 1.1004,
+      "epoch": 0.46947714207500685,
+      "grad_norm": 0.4031243358335159,
+      "learning_rate": 1.2777701895687034e-05,
+      "loss": 1.3346,
       "step": 1715
     },
     {
-      "epoch": 0.49311926605504586,
-      "grad_norm": 3.4164962172597035,
-      "learning_rate": 1.1973597367573509e-05,
-      "loss": 1.0966,
+      "epoch": 0.47084588009854916,
+      "grad_norm": 0.3919138760842289,
+      "learning_rate": 1.2731762788602988e-05,
+      "loss": 1.4046,
       "step": 1720
     },
     {
-      "epoch": 0.494552752293578,
-      "grad_norm": 3.3927103725177825,
-      "learning_rate": 1.19245158197083e-05,
-      "loss": 1.1635,
+      "epoch": 0.4722146181220914,
+      "grad_norm": 0.40045145219664335,
+      "learning_rate": 1.2685761296184987e-05,
+      "loss": 1.3808,
       "step": 1725
     },
     {
-      "epoch": 0.4959862385321101,
-      "grad_norm": 2.9590270815185193,
-      "learning_rate": 1.1875386079537762e-05,
-      "loss": 1.1415,
+      "epoch": 0.47358335614563374,
+      "grad_norm": 0.38631426280799075,
+      "learning_rate": 1.2639698468970237e-05,
+      "loss": 1.3563,
       "step": 1730
     },
     {
-      "epoch": 0.4974197247706422,
-      "grad_norm": 3.257588158306314,
-      "learning_rate": 1.1826209377332593e-05,
-      "loss": 1.1326,
+      "epoch": 0.474952094169176,
+      "grad_norm": 0.3891224496732836,
+      "learning_rate": 1.259357535889666e-05,
+      "loss": 1.368,
       "step": 1735
     },
     {
-      "epoch": 0.4988532110091743,
-      "grad_norm": 3.3051541372419737,
-      "learning_rate": 1.1776986944539498e-05,
-      "loss": 1.1278,
+      "epoch": 0.4763208321927183,
+      "grad_norm": 0.40044455029632475,
+      "learning_rate": 1.2547393019278853e-05,
+      "loss": 1.3548,
       "step": 1740
     },
     {
-      "epoch": 0.5002866972477065,
-      "grad_norm": 3.523346784327648,
-      "learning_rate": 1.1727720013750319e-05,
-      "loss": 1.1099,
+      "epoch": 0.47768957021626063,
+      "grad_norm": 0.3899915581292829,
+      "learning_rate": 1.2501152504784044e-05,
+      "loss": 1.3291,
       "step": 1745
     },
     {
-      "epoch": 0.5017201834862385,
-      "grad_norm": 3.3893012078265015,
-      "learning_rate": 1.1678409818671192e-05,
-      "loss": 1.1546,
+      "epoch": 0.4790583082398029,
+      "grad_norm": 0.3869143385442828,
+      "learning_rate": 1.2454854871407993e-05,
+      "loss": 1.3473,
       "step": 1750
     },
     {
-      "epoch": 0.5031536697247706,
-      "grad_norm": 3.3226857594304713,
-      "learning_rate": 1.1629057594091639e-05,
-      "loss": 1.1061,
+      "epoch": 0.4804270462633452,
+      "grad_norm": 0.3824047072867315,
+      "learning_rate": 1.2408501176450898e-05,
+      "loss": 1.3196,
       "step": 1755
     },
     {
-      "epoch": 0.5045871559633027,
-      "grad_norm": 3.507323841567208,
-      "learning_rate": 1.1579664575853667e-05,
-      "loss": 1.1551,
+      "epoch": 0.48179578428688746,
+      "grad_norm": 0.4057108906200233,
+      "learning_rate": 1.2362092478493226e-05,
+      "loss": 1.3554,
       "step": 1760
     },
     {
-      "epoch": 0.5060206422018348,
-      "grad_norm": 3.1700645014865856,
-      "learning_rate": 1.1530232000820791e-05,
-      "loss": 1.1151,
+      "epoch": 0.4831645223104298,
+      "grad_norm": 0.3776742288321122,
+      "learning_rate": 1.2315629837371556e-05,
+      "loss": 1.3014,
       "step": 1765
     },
     {
-      "epoch": 0.507454128440367,
-      "grad_norm": 3.397843452612661,
-      "learning_rate": 1.1480761106847088e-05,
-      "loss": 1.0967,
+      "epoch": 0.4845332603339721,
+      "grad_norm": 0.3977510874849243,
+      "learning_rate": 1.2269114314154365e-05,
+      "loss": 1.3967,
       "step": 1770
     },
     {
-      "epoch": 0.5088876146788991,
-      "grad_norm": 3.296593763189619,
-      "learning_rate": 1.1431253132746187e-05,
-      "loss": 1.1229,
+      "epoch": 0.48590199835751435,
+      "grad_norm": 0.3697505648941807,
+      "learning_rate": 1.2222546971117797e-05,
+      "loss": 1.3171,
       "step": 1775
     },
     {
-      "epoch": 0.5103211009174312,
-      "grad_norm": 3.1860797359537574,
-      "learning_rate": 1.138170931826025e-05,
-      "loss": 1.1305,
+      "epoch": 0.48727073638105667,
+      "grad_norm": 0.40788370625821474,
+      "learning_rate": 1.2175928871721411e-05,
+      "loss": 1.3222,
       "step": 1780
     },
     {
-      "epoch": 0.5117545871559633,
-      "grad_norm": 3.2917554463603333,
-      "learning_rate": 1.133213090402893e-05,
-      "loss": 1.0911,
+      "epoch": 0.488639474404599,
+      "grad_norm": 0.39090111193092664,
+      "learning_rate": 1.2129261080583897e-05,
+      "loss": 1.3676,
       "step": 1785
     },
     {
-      "epoch": 0.5131880733944955,
-      "grad_norm": 3.158139008213761,
-      "learning_rate": 1.1282519131558302e-05,
-      "loss": 1.1095,
+      "epoch": 0.49000821242814124,
+      "grad_norm": 0.3983751803246069,
+      "learning_rate": 1.2082544663458736e-05,
+      "loss": 1.3364,
       "step": 1790
     },
     {
-      "epoch": 0.5146215596330275,
-      "grad_norm": 3.22490301605298,
-      "learning_rate": 1.1232875243189765e-05,
-      "loss": 1.1695,
+      "epoch": 0.49137695045168356,
+      "grad_norm": 0.3848833412781912,
+      "learning_rate": 1.20357806872099e-05,
+      "loss": 1.2978,
       "step": 1795
     },
     {
-      "epoch": 0.5160550458715596,
-      "grad_norm": 3.586519697086409,
-      "learning_rate": 1.1183200482068949e-05,
-      "loss": 1.0989,
+      "epoch": 0.4927456884752258,
+      "grad_norm": 0.37897812252926655,
+      "learning_rate": 1.1988970219787467e-05,
+      "loss": 1.3333,
       "step": 1800
     },
     {
-      "epoch": 0.5174885321100917,
-      "grad_norm": 3.265891742392178,
-      "learning_rate": 1.1133496092114576e-05,
-      "loss": 1.0851,
+      "epoch": 0.49411442649876813,
+      "grad_norm": 0.3948100327011319,
+      "learning_rate": 1.1942114330203227e-05,
+      "loss": 1.3665,
       "step": 1805
     },
     {
-      "epoch": 0.5189220183486238,
-      "grad_norm": 3.1099004425125334,
-      "learning_rate": 1.1083763317987304e-05,
-      "loss": 1.1119,
+      "epoch": 0.49548316452231045,
+      "grad_norm": 0.3844309877664426,
+      "learning_rate": 1.1895214088506284e-05,
+      "loss": 1.3383,
       "step": 1810
     },
     {
-      "epoch": 0.520355504587156,
-      "grad_norm": 4.5404304007913305,
-      "learning_rate": 1.103400340505858e-05,
-      "loss": 1.1143,
+      "epoch": 0.4968519025458527,
+      "grad_norm": 0.3785571762413702,
+      "learning_rate": 1.1848270565758616e-05,
+      "loss": 1.2997,
       "step": 1815
     },
     {
-      "epoch": 0.5217889908256881,
-      "grad_norm": 3.4230865219558764,
-      "learning_rate": 1.0984217599379425e-05,
-      "loss": 1.1182,
+      "epoch": 0.498220640569395,
+      "grad_norm": 0.3873700740980167,
+      "learning_rate": 1.1801284834010596e-05,
+      "loss": 1.3332,
       "step": 1820
     },
     {
-      "epoch": 0.5232224770642202,
-      "grad_norm": 3.1427663664508456,
-      "learning_rate": 1.093440714764926e-05,
-      "loss": 1.1317,
+      "epoch": 0.49958937859293734,
+      "grad_norm": 0.3971195600650439,
+      "learning_rate": 1.1754257966276544e-05,
+      "loss": 1.3468,
       "step": 1825
     },
     {
-      "epoch": 0.5246559633027523,
-      "grad_norm": 2.9784943507834476,
-      "learning_rate": 1.088457329718467e-05,
-      "loss": 1.0364,
+      "epoch": 0.5009581166164796,
+      "grad_norm": 0.37924897770225724,
+      "learning_rate": 1.1707191036510189e-05,
+      "loss": 1.3152,
       "step": 1830
     },
     {
-      "epoch": 0.5260894495412844,
-      "grad_norm": 3.3656116327984806,
-      "learning_rate": 1.0834717295888168e-05,
-      "loss": 1.1432,
+      "epoch": 0.5023268546400219,
+      "grad_norm": 0.37889309262468557,
+      "learning_rate": 1.1660085119580165e-05,
+      "loss": 1.3074,
       "step": 1835
     },
     {
-      "epoch": 0.5275229357798165,
-      "grad_norm": 2.8639878500754774,
-      "learning_rate": 1.0784840392216961e-05,
-      "loss": 1.1421,
+      "epoch": 0.5036955926635642,
+      "grad_norm": 0.38526267448234497,
+      "learning_rate": 1.1612941291245456e-05,
+      "loss": 1.2863,
       "step": 1840
     },
     {
-      "epoch": 0.5289564220183486,
-      "grad_norm": 3.340151227567289,
-      "learning_rate": 1.0734943835151674e-05,
-      "loss": 1.0931,
+      "epoch": 0.5050643306871064,
+      "grad_norm": 0.4007327343298376,
+      "learning_rate": 1.1565760628130824e-05,
+      "loss": 1.348,
       "step": 1845
     },
     {
-      "epoch": 0.5303899082568807,
-      "grad_norm": 3.8035180968079993,
-      "learning_rate": 1.0685028874165075e-05,
-      "loss": 1.1179,
+      "epoch": 0.5064330687106487,
+      "grad_norm": 0.3917232580410847,
+      "learning_rate": 1.1518544207702238e-05,
+      "loss": 1.3193,
       "step": 1850
     },
     {
-      "epoch": 0.5318233944954128,
-      "grad_norm": 3.070748628370571,
-      "learning_rate": 1.0635096759190792e-05,
-      "loss": 1.123,
+      "epoch": 0.5078018067341911,
+      "grad_norm": 0.41154960813532804,
+      "learning_rate": 1.1471293108242251e-05,
+      "loss": 1.3921,
       "step": 1855
     },
     {
-      "epoch": 0.533256880733945,
-      "grad_norm": 3.226775333715127,
-      "learning_rate": 1.0585148740592013e-05,
-      "loss": 1.1159,
+      "epoch": 0.5091705447577334,
+      "grad_norm": 0.3808027664016428,
+      "learning_rate": 1.1424008408825383e-05,
+      "loss": 1.337,
       "step": 1860
     },
     {
-      "epoch": 0.5346903669724771,
-      "grad_norm": 3.236504164334084,
-      "learning_rate": 1.053518606913017e-05,
-      "loss": 1.1494,
+      "epoch": 0.5105392827812757,
+      "grad_norm": 0.38767901381709907,
+      "learning_rate": 1.1376691189293474e-05,
+      "loss": 1.3142,
       "step": 1865
     },
     {
-      "epoch": 0.5361238532110092,
-      "grad_norm": 3.2302791306950702,
-      "learning_rate": 1.048520999593362e-05,
-      "loss": 1.1444,
+      "epoch": 0.511908020804818,
+      "grad_norm": 0.39402951617560983,
+      "learning_rate": 1.1329342530231036e-05,
+      "loss": 1.358,
       "step": 1870
     },
     {
-      "epoch": 0.5375573394495413,
-      "grad_norm": 3.2832052841745916,
-      "learning_rate": 1.0435221772466318e-05,
-      "loss": 1.1469,
+      "epoch": 0.5132767588283602,
+      "grad_norm": 0.38538496638703307,
+      "learning_rate": 1.128196351294055e-05,
+      "loss": 1.2994,
       "step": 1875
     },
     {
-      "epoch": 0.5389908256880734,
-      "grad_norm": 3.2524360195076496,
-      "learning_rate": 1.0385222650496479e-05,
-      "loss": 1.1313,
+      "epoch": 0.5146454968519025,
+      "grad_norm": 0.38082093170114384,
+      "learning_rate": 1.1234555219417804e-05,
+      "loss": 1.3208,
       "step": 1880
     },
     {
-      "epoch": 0.5404243119266054,
-      "grad_norm": 2.919300243143409,
-      "learning_rate": 1.0335213882065225e-05,
-      "loss": 1.1112,
+      "epoch": 0.5160142348754448,
+      "grad_norm": 0.3748307863634792,
+      "learning_rate": 1.1187118732327167e-05,
+      "loss": 1.3506,
       "step": 1885
     },
     {
-      "epoch": 0.5418577981651376,
-      "grad_norm": 3.7356105416504546,
-      "learning_rate": 1.0285196719455242e-05,
-      "loss": 1.1113,
+      "epoch": 0.5173829728989872,
+      "grad_norm": 0.38705593276141625,
+      "learning_rate": 1.1139655134976855e-05,
+      "loss": 1.2906,
       "step": 1890
     },
     {
-      "epoch": 0.5432912844036697,
-      "grad_norm": 3.1163300953894306,
-      "learning_rate": 1.0235172415159418e-05,
-      "loss": 1.0671,
+      "epoch": 0.5187517109225295,
+      "grad_norm": 0.39888481423655353,
+      "learning_rate": 1.1092165511294206e-05,
+      "loss": 1.367,
       "step": 1895
     },
     {
-      "epoch": 0.5447247706422018,
-      "grad_norm": 3.6338243642002976,
-      "learning_rate": 1.0185142221849469e-05,
-      "loss": 1.1405,
+      "epoch": 0.5201204489460717,
+      "grad_norm": 0.3971141048112466,
+      "learning_rate": 1.104465094580093e-05,
+      "loss": 1.3405,
       "step": 1900
     },
     {
-      "epoch": 0.5461582568807339,
-      "grad_norm": 3.4265846748133613,
-      "learning_rate": 1.0135107392344594e-05,
-      "loss": 1.1052,
+      "epoch": 0.521489186969614,
+      "grad_norm": 0.3837313790185174,
+      "learning_rate": 1.0997112523588322e-05,
+      "loss": 1.3351,
       "step": 1905
     },
     {
-      "epoch": 0.5475917431192661,
-      "grad_norm": 3.6433669703963774,
-      "learning_rate": 1.0085069179580076e-05,
-      "loss": 1.0958,
+      "epoch": 0.5228579249931563,
+      "grad_norm": 0.39019433157562,
+      "learning_rate": 1.0949551330292502e-05,
+      "loss": 1.2482,
       "step": 1910
     },
     {
-      "epoch": 0.5490252293577982,
-      "grad_norm": 3.070116617818365,
-      "learning_rate": 1.0035028836575922e-05,
-      "loss": 1.0856,
+      "epoch": 0.5242266630166986,
+      "grad_norm": 0.39062012523465395,
+      "learning_rate": 1.090196845206961e-05,
+      "loss": 1.343,
       "step": 1915
     },
     {
-      "epoch": 0.5504587155963303,
-      "grad_norm": 3.3094892848191373,
-      "learning_rate": 9.984987616405486e-06,
-      "loss": 1.1775,
+      "epoch": 0.5255954010402409,
+      "grad_norm": 0.3948259084399907,
+      "learning_rate": 1.0854364975571004e-05,
+      "loss": 1.3001,
       "step": 1920
     },
     {
-      "epoch": 0.5518922018348624,
-      "grad_norm": 3.1253255381505465,
-      "learning_rate": 9.934946772164082e-06,
-      "loss": 1.0962,
+      "epoch": 0.5269641390637831,
+      "grad_norm": 0.3845620013681871,
+      "learning_rate": 1.0806741987918448e-05,
+      "loss": 1.3449,
       "step": 1925
     },
     {
-      "epoch": 0.5533256880733946,
-      "grad_norm": 3.2006466765888377,
-      "learning_rate": 9.884907556937619e-06,
-      "loss": 1.0732,
+      "epoch": 0.5283328770873255,
+      "grad_norm": 0.3976520791295208,
+      "learning_rate": 1.075910057667928e-05,
+      "loss": 1.3007,
       "step": 1930
     },
     {
-      "epoch": 0.5547591743119266,
-      "grad_norm": 3.1292256633487647,
-      "learning_rate": 9.834871223771204e-06,
-      "loss": 1.0887,
+      "epoch": 0.5297016151108678,
+      "grad_norm": 0.3817709747936168,
+      "learning_rate": 1.071144182984158e-05,
+      "loss": 1.3453,
       "step": 1935
     },
     {
-      "epoch": 0.5561926605504587,
-      "grad_norm": 3.1916225285491886,
-      "learning_rate": 9.78483902563778e-06,
-      "loss": 1.1239,
+      "epoch": 0.5310703531344101,
+      "grad_norm": 0.38542609340341144,
+      "learning_rate": 1.0663766835789327e-05,
+      "loss": 1.3187,
       "step": 1940
     },
     {
-      "epoch": 0.5576261467889908,
-      "grad_norm": 3.0332428019262623,
-      "learning_rate": 9.73481221540674e-06,
-      "loss": 1.1231,
+      "epoch": 0.5324390911579524,
+      "grad_norm": 0.3775171570445251,
+      "learning_rate": 1.0616076683277524e-05,
+      "loss": 1.3315,
       "step": 1945
     },
     {
-      "epoch": 0.5590596330275229,
-      "grad_norm": 3.045815826748736,
-      "learning_rate": 9.684792045812555e-06,
-      "loss": 1.1098,
+      "epoch": 0.5338078291814946,
+      "grad_norm": 0.40024298975169065,
+      "learning_rate": 1.056837246140736e-05,
+      "loss": 1.3772,
       "step": 1950
     },
     {
-      "epoch": 0.560493119266055,
-      "grad_norm": 3.0391892370593334,
-      "learning_rate": 9.634779769423412e-06,
-      "loss": 1.1044,
+      "epoch": 0.5351765672050369,
+      "grad_norm": 0.3882397104456436,
+      "learning_rate": 1.0520655259601325e-05,
+      "loss": 1.3284,
       "step": 1955
     },
     {
-      "epoch": 0.5619266055045872,
-      "grad_norm": 3.055569668736528,
-      "learning_rate": 9.584776638609841e-06,
-      "loss": 1.0894,
+      "epoch": 0.5365453052285792,
+      "grad_norm": 0.3892143818918897,
+      "learning_rate": 1.0472926167578323e-05,
+      "loss": 1.302,
       "step": 1960
     },
     {
-      "epoch": 0.5633600917431193,
-      "grad_norm": 3.5406439314630167,
-      "learning_rate": 9.534783905513355e-06,
-      "loss": 1.1109,
+      "epoch": 0.5379140432521216,
+      "grad_norm": 0.41282669532669847,
+      "learning_rate": 1.042518627532881e-05,
+      "loss": 1.3261,
       "step": 1965
     },
     {
-      "epoch": 0.5647935779816514,
-      "grad_norm": 3.345905555913882,
-      "learning_rate": 9.484802822015087e-06,
-      "loss": 1.1138,
+      "epoch": 0.5392827812756639,
+      "grad_norm": 0.37154889529358315,
+      "learning_rate": 1.0377436673089873e-05,
+      "loss": 1.3142,
       "step": 1970
     },
     {
-      "epoch": 0.5662270642201835,
-      "grad_norm": 3.296141105361695,
-      "learning_rate": 9.434834639704464e-06,
-      "loss": 1.1059,
+      "epoch": 0.5406515192992062,
+      "grad_norm": 0.4001478843028637,
+      "learning_rate": 1.0329678451320352e-05,
+      "loss": 1.3894,
       "step": 1975
     },
     {
-      "epoch": 0.5676605504587156,
-      "grad_norm": 2.8031036854561697,
-      "learning_rate": 9.384880609847838e-06,
-      "loss": 1.0806,
+      "epoch": 0.5420202573227484,
+      "grad_norm": 0.4082737912884333,
+      "learning_rate": 1.0281912700675937e-05,
+      "loss": 1.38,
       "step": 1980
     },
     {
-      "epoch": 0.5690940366972477,
-      "grad_norm": 3.0977315843393165,
-      "learning_rate": 9.33494198335717e-06,
-      "loss": 1.1364,
+      "epoch": 0.5433889953462907,
+      "grad_norm": 0.3976694377900842,
+      "learning_rate": 1.0234140511984246e-05,
+      "loss": 1.337,
       "step": 1985
     },
     {
-      "epoch": 0.5705275229357798,
-      "grad_norm": 3.3929218329827604,
-      "learning_rate": 9.285020010758706e-06,
-      "loss": 1.1644,
+      "epoch": 0.544757733369833,
+      "grad_norm": 0.4090490835246688,
+      "learning_rate": 1.0186362976219926e-05,
+      "loss": 1.3389,
       "step": 1990
     },
     {
-      "epoch": 0.5719610091743119,
-      "grad_norm": 3.2746802917151396,
-      "learning_rate": 9.235115942161656e-06,
-      "loss": 1.0867,
+      "epoch": 0.5461264713933753,
+      "grad_norm": 0.39123091240448754,
+      "learning_rate": 1.0138581184479737e-05,
+      "loss": 1.2726,
       "step": 1995
     },
     {
-      "epoch": 0.573394495412844,
-      "grad_norm": 3.4111204084118594,
-      "learning_rate": 9.18523102722688e-06,
-      "loss": 1.1372,
+      "epoch": 0.5474952094169176,
+      "grad_norm": 0.383808685203924,
+      "learning_rate": 1.0090796227957633e-05,
+      "loss": 1.3344,
       "step": 2000
     },
     {
-      "epoch": 0.5748279816513762,
-      "grad_norm": 3.0339025073036883,
-      "learning_rate": 9.135366515135617e-06,
-      "loss": 1.1156,
+      "epoch": 0.5488639474404599,
+      "grad_norm": 0.39282920042398406,
+      "learning_rate": 1.0043009197919836e-05,
+      "loss": 1.3188,
       "step": 2005
     },
     {
-      "epoch": 0.5762614678899083,
-      "grad_norm": 3.152200571534001,
-      "learning_rate": 9.0855236545582e-06,
-      "loss": 1.2007,
+      "epoch": 0.5502326854640022,
+      "grad_norm": 0.3877735838132022,
+      "learning_rate": 9.99522118567993e-06,
+      "loss": 1.3412,
       "step": 2010
     },
     {
-      "epoch": 0.5776949541284404,
-      "grad_norm": 5.672098040763039,
-      "learning_rate": 9.035703693622762e-06,
-      "loss": 1.1011,
+      "epoch": 0.5516014234875445,
+      "grad_norm": 0.3855977732261345,
+      "learning_rate": 9.947433282573926e-06,
+      "loss": 1.31,
       "step": 2015
     },
     {
-      "epoch": 0.5791284403669725,
-      "grad_norm": 3.0337007868840207,
-      "learning_rate": 8.985907879884011e-06,
-      "loss": 1.0502,
+      "epoch": 0.5529701615110868,
+      "grad_norm": 0.39182359203627265,
+      "learning_rate": 9.899646579935336e-06,
+      "loss": 1.3288,
       "step": 2020
     },
     {
-      "epoch": 0.5805619266055045,
-      "grad_norm": 3.397456497600358,
-      "learning_rate": 8.936137460291985e-06,
-      "loss": 1.0686,
+      "epoch": 0.5543388995346291,
+      "grad_norm": 0.37896193597955086,
+      "learning_rate": 9.85186216907027e-06,
+      "loss": 1.3222,
       "step": 2025
     },
     {
-      "epoch": 0.5819954128440367,
-      "grad_norm": 3.1557825825649983,
-      "learning_rate": 8.886393681160804e-06,
-      "loss": 1.1242,
+      "epoch": 0.5557076375581713,
+      "grad_norm": 0.39627324276578396,
+      "learning_rate": 9.80408114123249e-06,
+      "loss": 1.3659,
       "step": 2030
     },
     {
-      "epoch": 0.5834288990825688,
-      "grad_norm": 3.014868609795805,
-      "learning_rate": 8.836677788137488e-06,
-      "loss": 1.0823,
+      "epoch": 0.5570763755817136,
+      "grad_norm": 0.39331176053433153,
+      "learning_rate": 9.756304587598503e-06,
+      "loss": 1.4234,
       "step": 2035
     },
     {
-      "epoch": 0.5848623853211009,
-      "grad_norm": 3.1697976935253256,
-      "learning_rate": 8.78699102617076e-06,
-      "loss": 1.1477,
+      "epoch": 0.558445113605256,
+      "grad_norm": 0.38413569943245757,
+      "learning_rate": 9.708533599242643e-06,
+      "loss": 1.2809,
       "step": 2040
     },
     {
-      "epoch": 0.586295871559633,
-      "grad_norm": 3.060910090014108,
-      "learning_rate": 8.737334639479843e-06,
-      "loss": 1.1047,
+      "epoch": 0.5598138516287983,
+      "grad_norm": 0.378050582923272,
+      "learning_rate": 9.660769267112137e-06,
+      "loss": 1.3,
       "step": 2045
     },
     {
-      "epoch": 0.5877293577981652,
-      "grad_norm": 3.3975873769480525,
-      "learning_rate": 8.687709871523346e-06,
-      "loss": 1.1472,
+      "epoch": 0.5611825896523406,
+      "grad_norm": 0.3810287903511996,
+      "learning_rate": 9.61301268200222e-06,
+      "loss": 1.328,
       "step": 2050
     },
     {
-      "epoch": 0.5891628440366973,
-      "grad_norm": 3.103294993064734,
-      "learning_rate": 8.638117964968098e-06,
-      "loss": 1.1147,
+      "epoch": 0.5625513276758828,
+      "grad_norm": 0.3968883040797777,
+      "learning_rate": 9.5652649345312e-06,
+      "loss": 1.3861,
       "step": 2055
     },
     {
-      "epoch": 0.5905963302752294,
-      "grad_norm": 3.1038713521837176,
-      "learning_rate": 8.588560161658039e-06,
-      "loss": 1.1201,
+      "epoch": 0.5639200656994251,
+      "grad_norm": 0.381754630880566,
+      "learning_rate": 9.517527115115554e-06,
+      "loss": 1.2931,
       "step": 2060
     },
     {
-      "epoch": 0.5920298165137615,
-      "grad_norm": 3.579612587244007,
-      "learning_rate": 8.539037702583108e-06,
-      "loss": 1.2046,
+      "epoch": 0.5652888037229674,
+      "grad_norm": 0.3934475062551264,
+      "learning_rate": 9.46980031394504e-06,
+      "loss": 1.357,
       "step": 2065
     },
     {
-      "epoch": 0.5934633027522935,
-      "grad_norm": 3.1302516904022766,
-      "learning_rate": 8.489551827848197e-06,
-      "loss": 1.0237,
+      "epoch": 0.5666575417465097,
+      "grad_norm": 0.3764960791780743,
+      "learning_rate": 9.422085620957795e-06,
+      "loss": 1.3112,
       "step": 2070
     },
     {
-      "epoch": 0.5948967889908257,
-      "grad_norm": 3.24638481758472,
-      "learning_rate": 8.440103776642074e-06,
-      "loss": 1.0678,
+      "epoch": 0.568026279770052,
+      "grad_norm": 0.3819475902039454,
+      "learning_rate": 9.374384125815427e-06,
+      "loss": 1.317,
       "step": 2075
     },
     {
-      "epoch": 0.5963302752293578,
-      "grad_norm": 3.4237594408353083,
-      "learning_rate": 8.390694787206349e-06,
-      "loss": 1.1734,
+      "epoch": 0.5693950177935944,
+      "grad_norm": 0.39075935188345634,
+      "learning_rate": 9.326696917878159e-06,
+      "loss": 1.329,
       "step": 2080
     },
     {
-      "epoch": 0.5977637614678899,
-      "grad_norm": 3.0719379679166625,
-      "learning_rate": 8.341326096804489e-06,
-      "loss": 1.0969,
+      "epoch": 0.5707637558171366,
+      "grad_norm": 0.4115410325839856,
+      "learning_rate": 9.27902508617993e-06,
+      "loss": 1.3405,
       "step": 2085
     },
     {
-      "epoch": 0.599197247706422,
-      "grad_norm": 3.114117068959359,
-      "learning_rate": 8.291998941690821e-06,
-      "loss": 1.1036,
+      "epoch": 0.5721324938406789,
+      "grad_norm": 0.3929327529237431,
+      "learning_rate": 9.23136971940353e-06,
+      "loss": 1.2811,
       "step": 2090
     },
     {
-      "epoch": 0.6006307339449541,
-      "grad_norm": 3.058349370894333,
-      "learning_rate": 8.242714557079563e-06,
-      "loss": 1.0749,
+      "epoch": 0.5735012318642212,
+      "grad_norm": 0.38257057379328274,
+      "learning_rate": 9.183731905855746e-06,
+      "loss": 1.3234,
       "step": 2095
     },
     {
-      "epoch": 0.6020642201834863,
-      "grad_norm": 3.036934322182637,
-      "learning_rate": 8.193474177113918e-06,
-      "loss": 1.0583,
+      "epoch": 0.5748699698877635,
+      "grad_norm": 0.38674820150451145,
+      "learning_rate": 9.136112733442493e-06,
+      "loss": 1.3466,
       "step": 2100
     },
     {
-      "epoch": 0.6034977064220184,
-      "grad_norm": 3.13799906852602,
-      "learning_rate": 8.144279034835157e-06,
-      "loss": 1.1374,
+      "epoch": 0.5762387079113058,
+      "grad_norm": 0.39368922489348585,
+      "learning_rate": 9.088513289643982e-06,
+      "loss": 1.3149,
       "step": 2105
     },
     {
-      "epoch": 0.6049311926605505,
-      "grad_norm": 3.303135805293391,
-      "learning_rate": 8.095130362151737e-06,
-      "loss": 1.0465,
+      "epoch": 0.577607445934848,
+      "grad_norm": 0.38199335563674713,
+      "learning_rate": 9.040934661489885e-06,
+      "loss": 1.3576,
       "step": 2110
     },
     {
-      "epoch": 0.6063646788990825,
-      "grad_norm": 3.3978912865766686,
-      "learning_rate": 8.046029389808457e-06,
-      "loss": 1.055,
+      "epoch": 0.5789761839583903,
+      "grad_norm": 0.39323361965983067,
+      "learning_rate": 8.9933779355345e-06,
+      "loss": 1.3159,
       "step": 2115
     },
     {
-      "epoch": 0.6077981651376146,
-      "grad_norm": 3.47121454527395,
-      "learning_rate": 7.996977347355647e-06,
-      "loss": 1.1155,
+      "epoch": 0.5803449219819327,
+      "grad_norm": 0.39256654497554794,
+      "learning_rate": 8.945844197831945e-06,
+      "loss": 1.3623,
       "step": 2120
     },
     {
-      "epoch": 0.6092316513761468,
-      "grad_norm": 3.156947339229311,
-      "learning_rate": 7.947975463118361e-06,
-      "loss": 1.0363,
+      "epoch": 0.581713660005475,
+      "grad_norm": 0.3905410660434604,
+      "learning_rate": 8.898334533911362e-06,
+      "loss": 1.3074,
       "step": 2125
     },
     {
-      "epoch": 0.6106651376146789,
-      "grad_norm": 3.2797688368791955,
-      "learning_rate": 7.899024964165634e-06,
-      "loss": 1.1381,
+      "epoch": 0.5830823980290173,
+      "grad_norm": 0.3700729231212675,
+      "learning_rate": 8.850850028752108e-06,
+      "loss": 1.3436,
       "step": 2130
     },
     {
-      "epoch": 0.612098623853211,
-      "grad_norm": 2.9662302688805275,
-      "learning_rate": 7.850127076279747e-06,
-      "loss": 1.1032,
+      "epoch": 0.5844511360525595,
+      "grad_norm": 0.3701852614302805,
+      "learning_rate": 8.803391766758998e-06,
+      "loss": 1.2672,
       "step": 2135
     },
     {
-      "epoch": 0.6135321100917431,
-      "grad_norm": 3.203748291163447,
-      "learning_rate": 7.801283023925536e-06,
-      "loss": 1.0517,
+      "epoch": 0.5858198740761018,
+      "grad_norm": 0.3809658344783075,
+      "learning_rate": 8.755960831737529e-06,
+      "loss": 1.3638,
       "step": 2140
     },
     {
-      "epoch": 0.6149655963302753,
-      "grad_norm": 3.319287430110112,
-      "learning_rate": 7.752494030219724e-06,
-      "loss": 1.115,
+      "epoch": 0.5871886120996441,
+      "grad_norm": 0.3852501406589481,
+      "learning_rate": 8.708558306869125e-06,
+      "loss": 1.3389,
       "step": 2145
     },
     {
-      "epoch": 0.6163990825688074,
-      "grad_norm": 3.7703557404425823,
-      "learning_rate": 7.703761316900293e-06,
-      "loss": 1.0813,
+      "epoch": 0.5885573501231864,
+      "grad_norm": 0.4003501899309122,
+      "learning_rate": 8.661185274686418e-06,
+      "loss": 1.3228,
       "step": 2150
     },
     {
-      "epoch": 0.6178325688073395,
-      "grad_norm": 3.2303522091544727,
-      "learning_rate": 7.655086104295904e-06,
-      "loss": 1.0668,
+      "epoch": 0.5899260881467288,
+      "grad_norm": 0.38933438340786924,
+      "learning_rate": 8.613842817048503e-06,
+      "loss": 1.3627,
       "step": 2155
     },
     {
-      "epoch": 0.6192660550458715,
-      "grad_norm": 3.1818989111713676,
-      "learning_rate": 7.606469611295315e-06,
-      "loss": 1.1368,
+      "epoch": 0.591294826170271,
+      "grad_norm": 0.39608781513600605,
+      "learning_rate": 8.566532015116251e-06,
+      "loss": 1.3626,
       "step": 2160
     },
     {
-      "epoch": 0.6206995412844036,
-      "grad_norm": 3.0516455298261347,
-      "learning_rate": 7.5579130553168815e-06,
-      "loss": 1.1132,
+      "epoch": 0.5926635641938133,
+      "grad_norm": 0.37582004076749886,
+      "learning_rate": 8.51925394932761e-06,
+      "loss": 1.2766,
       "step": 2165
     },
     {
-      "epoch": 0.6221330275229358,
-      "grad_norm": 3.3307988467764087,
-      "learning_rate": 7.50941765227805e-06,
-      "loss": 1.0518,
+      "epoch": 0.5940323022173556,
+      "grad_norm": 0.39046439420533485,
+      "learning_rate": 8.47200969937293e-06,
+      "loss": 1.3342,
       "step": 2170
     },
     {
-      "epoch": 0.6235665137614679,
-      "grad_norm": 3.3783423662774363,
-      "learning_rate": 7.460984616564929e-06,
-      "loss": 1.1132,
+      "epoch": 0.5954010402408979,
+      "grad_norm": 0.3846853640345239,
+      "learning_rate": 8.424800344170308e-06,
+      "loss": 1.2973,
       "step": 2175
     },
     {
-      "epoch": 0.625,
-      "grad_norm": 3.634134930894941,
-      "learning_rate": 7.412615161001866e-06,
-      "loss": 1.0798,
+      "epoch": 0.5967697782644402,
+      "grad_norm": 0.3794031254673053,
+      "learning_rate": 8.37762696184095e-06,
+      "loss": 1.3288,
       "step": 2180
     },
     {
-      "epoch": 0.6264334862385321,
-      "grad_norm": 3.2417955659870796,
-      "learning_rate": 7.364310496821086e-06,
-      "loss": 1.0864,
+      "epoch": 0.5981385162879824,
+      "grad_norm": 0.38487683969290326,
+      "learning_rate": 8.330490629684547e-06,
+      "loss": 1.3374,
       "step": 2185
     },
     {
-      "epoch": 0.6278669724770642,
-      "grad_norm": 3.1150031782926395,
-      "learning_rate": 7.316071833632346e-06,
-      "loss": 1.1044,
+      "epoch": 0.5995072543115247,
+      "grad_norm": 0.38666671699682115,
+      "learning_rate": 8.28339242415468e-06,
+      "loss": 1.3202,
       "step": 2190
     },
     {
-      "epoch": 0.6293004587155964,
-      "grad_norm": 3.058907033640181,
-      "learning_rate": 7.2679003793926626e-06,
-      "loss": 1.1097,
+      "epoch": 0.600875992335067,
+      "grad_norm": 0.3788541652969095,
+      "learning_rate": 8.236333420834216e-06,
+      "loss": 1.2778,
       "step": 2195
     },
     {
-      "epoch": 0.6307339449541285,
-      "grad_norm": 3.2942538258713348,
-      "learning_rate": 7.2197973403760614e-06,
-      "loss": 1.1005,
+      "epoch": 0.6022447303586094,
+      "grad_norm": 0.4003559394681289,
+      "learning_rate": 8.189314694410781e-06,
+      "loss": 1.3349,
       "step": 2200
     },
     {
-      "epoch": 0.6321674311926605,
-      "grad_norm": 3.225308816069398,
-      "learning_rate": 7.171763921143346e-06,
-      "loss": 1.0627,
+      "epoch": 0.6036134683821517,
+      "grad_norm": 0.40672129713209243,
+      "learning_rate": 8.14233731865218e-06,
+      "loss": 1.3343,
       "step": 2205
     },
     {
-      "epoch": 0.6336009174311926,
-      "grad_norm": 3.0409935206984566,
-      "learning_rate": 7.123801324511972e-06,
-      "loss": 1.0937,
+      "epoch": 0.604982206405694,
+      "grad_norm": 0.392344465653996,
+      "learning_rate": 8.0954023663819e-06,
+      "loss": 1.2804,
       "step": 2210
     },
     {
-      "epoch": 0.6350344036697247,
-      "grad_norm": 9.664703329802007,
-      "learning_rate": 7.075910751525895e-06,
-      "loss": 1.1024,
+      "epoch": 0.6063509444292362,
+      "grad_norm": 0.396931681160349,
+      "learning_rate": 8.048510909454601e-06,
+      "loss": 1.3712,
       "step": 2215
     },
     {
-      "epoch": 0.6364678899082569,
-      "grad_norm": 3.057262418668561,
-      "learning_rate": 7.0280934014255195e-06,
-      "loss": 1.0857,
+      "epoch": 0.6077196824527785,
+      "grad_norm": 0.3882127247657791,
+      "learning_rate": 8.001664018731637e-06,
+      "loss": 1.3313,
       "step": 2220
     },
     {
-      "epoch": 0.637901376146789,
-      "grad_norm": 3.0509477926802395,
-      "learning_rate": 6.980350471617638e-06,
-      "loss": 1.0727,
+      "epoch": 0.6090884204763208,
+      "grad_norm": 0.3727086035483801,
+      "learning_rate": 7.954862764056602e-06,
+      "loss": 1.2715,
       "step": 2225
     },
     {
-      "epoch": 0.6393348623853211,
-      "grad_norm": 3.4647112333044743,
-      "learning_rate": 6.9326831576454835e-06,
-      "loss": 1.1128,
+      "epoch": 0.6104571584998632,
+      "grad_norm": 0.39519191088539024,
+      "learning_rate": 7.908108214230902e-06,
+      "loss": 1.3018,
       "step": 2230
     },
     {
-      "epoch": 0.6407683486238532,
-      "grad_norm": 8.615283468155432,
-      "learning_rate": 6.885092653158768e-06,
-      "loss": 1.1205,
+      "epoch": 0.6118258965234055,
+      "grad_norm": 0.39954515244878724,
+      "learning_rate": 7.861401436989337e-06,
+      "loss": 1.3697,
       "step": 2235
     },
     {
-      "epoch": 0.6422018348623854,
-      "grad_norm": 3.6518697308480927,
-      "learning_rate": 6.837580149883787e-06,
-      "loss": 1.0998,
+      "epoch": 0.6131946345469477,
+      "grad_norm": 0.3961862899698687,
+      "learning_rate": 7.814743498975731e-06,
+      "loss": 1.3651,
       "step": 2240
     },
     {
-      "epoch": 0.6436353211009175,
-      "grad_norm": 3.2292043015948493,
-      "learning_rate": 6.790146837593599e-06,
-      "loss": 1.0467,
+      "epoch": 0.61456337257049,
+      "grad_norm": 0.38565731038310297,
+      "learning_rate": 7.768135465718559e-06,
+      "loss": 1.3462,
       "step": 2245
     },
     {
-      "epoch": 0.6450688073394495,
-      "grad_norm": 3.060662379650319,
-      "learning_rate": 6.7427939040782175e-06,
-      "loss": 1.0426,
+      "epoch": 0.6159321105940323,
+      "grad_norm": 0.38796255525719986,
+      "learning_rate": 7.72157840160662e-06,
+      "loss": 1.3087,
       "step": 2250
     },
     {
-      "epoch": 0.6465022935779816,
-      "grad_norm": 3.3641577098688087,
-      "learning_rate": 6.695522535114866e-06,
-      "loss": 1.0968,
+      "epoch": 0.6173008486175746,
+      "grad_norm": 0.3947254708668997,
+      "learning_rate": 7.67507336986474e-06,
+      "loss": 1.262,
       "step": 2255
     },
     {
-      "epoch": 0.6479357798165137,
-      "grad_norm": 3.025310634435396,
-      "learning_rate": 6.64833391443829e-06,
-      "loss": 1.1066,
+      "epoch": 0.6186695866411169,
+      "grad_norm": 0.39167259958488637,
+      "learning_rate": 7.628621432529461e-06,
+      "loss": 1.3117,
       "step": 2260
     },
     {
-      "epoch": 0.6493692660550459,
-      "grad_norm": 3.3638409022750477,
-      "learning_rate": 6.601229223711123e-06,
-      "loss": 1.061,
+      "epoch": 0.6200383246646591,
+      "grad_norm": 0.39458361491505556,
+      "learning_rate": 7.582223650424825e-06,
+      "loss": 1.3067,
       "step": 2265
     },
     {
-      "epoch": 0.650802752293578,
-      "grad_norm": 3.0591413772292415,
-      "learning_rate": 6.554209642494267e-06,
-      "loss": 1.1305,
+      "epoch": 0.6214070626882015,
+      "grad_norm": 0.3948240515553033,
+      "learning_rate": 7.5358810831381225e-06,
+      "loss": 1.3177,
       "step": 2270
     },
     {
-      "epoch": 0.6522362385321101,
-      "grad_norm": 3.048895188924846,
-      "learning_rate": 6.507276348217393e-06,
-      "loss": 1.0676,
+      "epoch": 0.6227758007117438,
+      "grad_norm": 0.3814626352546649,
+      "learning_rate": 7.489594788995698e-06,
+      "loss": 1.3259,
       "step": 2275
     },
     {
-      "epoch": 0.6536697247706422,
-      "grad_norm": 3.1625182877065,
-      "learning_rate": 6.460430516149433e-06,
-      "loss": 0.9783,
+      "epoch": 0.6241445387352861,
+      "grad_norm": 0.39070667312634827,
+      "learning_rate": 7.443365825038793e-06,
+      "loss": 1.3212,
       "step": 2280
     },
     {
-      "epoch": 0.6551032110091743,
-      "grad_norm": 3.0971219975300093,
-      "learning_rate": 6.413673319369145e-06,
-      "loss": 1.0947,
+      "epoch": 0.6255132767588284,
+      "grad_norm": 0.38913877038236866,
+      "learning_rate": 7.397195246999391e-06,
+      "loss": 1.346,
       "step": 2285
     },
     {
-      "epoch": 0.6565366972477065,
-      "grad_norm": 3.0087113841047706,
-      "learning_rate": 6.36700592873576e-06,
-      "loss": 1.1115,
+      "epoch": 0.6268820147823706,
+      "grad_norm": 0.4070153012541291,
+      "learning_rate": 7.351084109276119e-06,
+      "loss": 1.3826,
       "step": 2290
     },
     {
-      "epoch": 0.6579701834862385,
-      "grad_norm": 3.240656660427129,
-      "learning_rate": 6.320429512859645e-06,
-      "loss": 1.0604,
+      "epoch": 0.6282507528059129,
+      "grad_norm": 0.36428744429147014,
+      "learning_rate": 7.305033464910164e-06,
+      "loss": 1.2659,
       "step": 2295
     },
     {
-      "epoch": 0.6594036697247706,
-      "grad_norm": 2.8631951161021423,
-      "learning_rate": 6.273945238073047e-06,
-      "loss": 1.0521,
+      "epoch": 0.6296194908294552,
+      "grad_norm": 0.3877290507395246,
+      "learning_rate": 7.25904436556122e-06,
+      "loss": 1.304,
       "step": 2300
     },
     {
-      "epoch": 0.6608371559633027,
-      "grad_norm": 3.372695675980184,
-      "learning_rate": 6.227554268400875e-06,
-      "loss": 1.1259,
+      "epoch": 0.6309882288529975,
+      "grad_norm": 0.3803920075270922,
+      "learning_rate": 7.21311786148348e-06,
+      "loss": 1.3333,
       "step": 2305
     },
     {
-      "epoch": 0.6622706422018348,
-      "grad_norm": 3.0645684579682353,
-      "learning_rate": 6.1812577655315695e-06,
-      "loss": 1.0146,
+      "epoch": 0.6323569668765399,
+      "grad_norm": 0.3822541348646928,
+      "learning_rate": 7.167255001501651e-06,
+      "loss": 1.3307,
       "step": 2310
     },
     {
-      "epoch": 0.663704128440367,
-      "grad_norm": 3.2992667272685456,
-      "learning_rate": 6.135056888788004e-06,
-      "loss": 1.0908,
+      "epoch": 0.6337257049000822,
+      "grad_norm": 0.3977624877197784,
+      "learning_rate": 7.121456832986988e-06,
+      "loss": 1.3329,
       "step": 2315
     },
     {
-      "epoch": 0.6651376146788991,
-      "grad_norm": 3.134342552266428,
-      "learning_rate": 6.088952795098442e-06,
-      "loss": 1.0408,
+      "epoch": 0.6350944429236244,
+      "grad_norm": 0.38990040940646387,
+      "learning_rate": 7.075724401833395e-06,
+      "loss": 1.3647,
       "step": 2320
     },
     {
-      "epoch": 0.6665711009174312,
-      "grad_norm": 3.195978033002455,
-      "learning_rate": 6.042946638967586e-06,
-      "loss": 1.0734,
+      "epoch": 0.6364631809471667,
+      "grad_norm": 0.39550975070062716,
+      "learning_rate": 7.030058752433526e-06,
+      "loss": 1.2715,
       "step": 2325
     },
     {
-      "epoch": 0.6680045871559633,
-      "grad_norm": 3.244082712886425,
-      "learning_rate": 5.997039572447658e-06,
-      "loss": 1.0466,
+      "epoch": 0.637831918970709,
+      "grad_norm": 0.3857148917491234,
+      "learning_rate": 6.984460927654937e-06,
+      "loss": 1.3191,
       "step": 2330
     },
     {
-      "epoch": 0.6694380733944955,
-      "grad_norm": 2.9373127993862105,
-      "learning_rate": 5.951232745109552e-06,
-      "loss": 1.0658,
+      "epoch": 0.6392006569942513,
+      "grad_norm": 0.38855941306596253,
+      "learning_rate": 6.938931968816275e-06,
+      "loss": 1.2979,
       "step": 2335
     },
     {
-      "epoch": 0.6708715596330275,
-      "grad_norm": 3.1038594855264967,
-      "learning_rate": 5.9055273040140374e-06,
-      "loss": 1.0916,
+      "epoch": 0.6405693950177936,
+      "grad_norm": 0.3892088605550479,
+      "learning_rate": 6.893472915663493e-06,
+      "loss": 1.3826,
       "step": 2340
     },
     {
-      "epoch": 0.6723050458715596,
-      "grad_norm": 3.1281256278361256,
-      "learning_rate": 5.859924393683056e-06,
-      "loss": 1.1042,
+      "epoch": 0.6419381330413358,
+      "grad_norm": 0.4024108620394257,
+      "learning_rate": 6.8480848063461035e-06,
+      "loss": 1.2943,
       "step": 2345
     },
     {
-      "epoch": 0.6737385321100917,
-      "grad_norm": 2.9165714439580825,
-      "learning_rate": 5.8144251560710415e-06,
-      "loss": 1.0735,
+      "epoch": 0.6433068710648782,
+      "grad_norm": 0.3976600247063458,
+      "learning_rate": 6.8027686773934765e-06,
+      "loss": 1.3241,
       "step": 2350
     },
     {
-      "epoch": 0.6751720183486238,
-      "grad_norm": 3.3690225312568955,
-      "learning_rate": 5.769030730536336e-06,
-      "loss": 1.0574,
+      "epoch": 0.6446756090884205,
+      "grad_norm": 0.38538373822690386,
+      "learning_rate": 6.7575255636911626e-06,
+      "loss": 1.3316,
       "step": 2355
     },
     {
-      "epoch": 0.676605504587156,
-      "grad_norm": 3.298580191163876,
-      "learning_rate": 5.723742253812658e-06,
-      "loss": 1.1132,
+      "epoch": 0.6460443471119628,
+      "grad_norm": 0.3961847974725629,
+      "learning_rate": 6.7123564984572596e-06,
+      "loss": 1.3481,
       "step": 2360
     },
     {
-      "epoch": 0.6780389908256881,
-      "grad_norm": 3.004930197106605,
-      "learning_rate": 5.678560859980621e-06,
-      "loss": 1.0691,
+      "epoch": 0.6474130851355051,
+      "grad_norm": 0.3739018408786073,
+      "learning_rate": 6.667262513218824e-06,
+      "loss": 1.2857,
       "step": 2365
     },
     {
-      "epoch": 0.6794724770642202,
-      "grad_norm": 3.1318112002832197,
-      "learning_rate": 5.633487680439362e-06,
-      "loss": 1.0305,
+      "epoch": 0.6487818231590473,
+      "grad_norm": 0.4058027362822848,
+      "learning_rate": 6.622244637788302e-06,
+      "loss": 1.3623,
       "step": 2370
     },
     {
-      "epoch": 0.6809059633027523,
-      "grad_norm": 3.139769183129407,
-      "learning_rate": 5.588523843878189e-06,
-      "loss": 1.0547,
+      "epoch": 0.6501505611825896,
+      "grad_norm": 0.3887614780397367,
+      "learning_rate": 6.577303900240023e-06,
+      "loss": 1.3466,
       "step": 2375
     },
     {
-      "epoch": 0.6823394495412844,
-      "grad_norm": 3.060531894281262,
-      "learning_rate": 5.543670476248327e-06,
-      "loss": 1.0354,
+      "epoch": 0.6515192992061319,
+      "grad_norm": 0.39583219449776136,
+      "learning_rate": 6.532441326886716e-06,
+      "loss": 1.3826,
       "step": 2380
     },
     {
-      "epoch": 0.6837729357798165,
-      "grad_norm": 3.1023960334424556,
-      "learning_rate": 5.498928700734713e-06,
-      "loss": 1.0565,
+      "epoch": 0.6528880372296743,
+      "grad_norm": 0.38600694323078566,
+      "learning_rate": 6.487657942256069e-06,
+      "loss": 1.3124,
       "step": 2385
     },
     {
-      "epoch": 0.6852064220183486,
-      "grad_norm": 3.1211535834048396,
-      "learning_rate": 5.454299637727885e-06,
-      "loss": 1.0662,
+      "epoch": 0.6542567752532166,
+      "grad_norm": 0.35581732986644143,
+      "learning_rate": 6.442954769067341e-06,
+      "loss": 1.3352,
       "step": 2390
     },
     {
-      "epoch": 0.6866399082568807,
-      "grad_norm": 2.99851195359665,
-      "learning_rate": 5.409784404795913e-06,
-      "loss": 1.0467,
+      "epoch": 0.6556255132767588,
+      "grad_norm": 0.3977530272775962,
+      "learning_rate": 6.398332828207996e-06,
+      "loss": 1.3226,
       "step": 2395
     },
     {
-      "epoch": 0.6880733944954128,
-      "grad_norm": 3.0166642308548606,
-      "learning_rate": 5.365384116656415e-06,
-      "loss": 1.1033,
+      "epoch": 0.6569942513003011,
+      "grad_norm": 0.3928175408666658,
+      "learning_rate": 6.3537931387103925e-06,
+      "loss": 1.3262,
       "step": 2400
     },
     {
-      "epoch": 0.689506880733945,
-      "grad_norm": 3.1825150095020724,
-      "learning_rate": 5.321099885148652e-06,
-      "loss": 1.0541,
+      "epoch": 0.6583629893238434,
+      "grad_norm": 0.3795840583774513,
+      "learning_rate": 6.309336717728516e-06,
+      "loss": 1.3203,
       "step": 2405
     },
     {
-      "epoch": 0.6909403669724771,
-      "grad_norm": 3.3490852800179174,
-      "learning_rate": 5.2769328192056824e-06,
-      "loss": 1.0624,
+      "epoch": 0.6597317273473857,
+      "grad_norm": 0.396679060826067,
+      "learning_rate": 6.264964580514745e-06,
+      "loss": 1.2842,
       "step": 2410
     },
     {
-      "epoch": 0.6923738532110092,
-      "grad_norm": 3.1423133524388698,
-      "learning_rate": 5.23288402482658e-06,
-      "loss": 1.0582,
+      "epoch": 0.661100465370928,
+      "grad_norm": 0.3699325537006795,
+      "learning_rate": 6.220677740396668e-06,
+      "loss": 1.2936,
       "step": 2415
     },
     {
-      "epoch": 0.6938073394495413,
-      "grad_norm": 3.1573558030480235,
-      "learning_rate": 5.18895460504876e-06,
-      "loss": 1.0325,
+      "epoch": 0.6624692033944704,
+      "grad_norm": 0.38098356916961845,
+      "learning_rate": 6.176477208753944e-06,
+      "loss": 1.276,
       "step": 2420
     },
     {
-      "epoch": 0.6952408256880734,
-      "grad_norm": 3.107763639677989,
-      "learning_rate": 5.145145659920348e-06,
-      "loss": 1.0017,
+      "epoch": 0.6638379414180126,
+      "grad_norm": 0.38873579028440075,
+      "learning_rate": 6.132363994995194e-06,
+      "loss": 1.2348,
       "step": 2425
     },
     {
-      "epoch": 0.6966743119266054,
-      "grad_norm": 3.3802618734734793,
-      "learning_rate": 5.101458286472618e-06,
-      "loss": 1.0718,
+      "epoch": 0.6652066794415549,
+      "grad_norm": 0.39405073870487267,
+      "learning_rate": 6.088339106534971e-06,
+      "loss": 1.2659,
       "step": 2430
     },
     {
-      "epoch": 0.6981077981651376,
-      "grad_norm": 2.918248752013308,
-      "learning_rate": 5.05789357869255e-06,
-      "loss": 1.0316,
+      "epoch": 0.6665754174650972,
+      "grad_norm": 0.3833620202872981,
+      "learning_rate": 6.044403548770735e-06,
+      "loss": 1.3103,
       "step": 2435
     },
     {
-      "epoch": 0.6995412844036697,
-      "grad_norm": 3.189799059727885,
-      "learning_rate": 5.01445262749542e-06,
-      "loss": 1.0851,
+      "epoch": 0.6679441554886395,
+      "grad_norm": 0.39527725366864974,
+      "learning_rate": 6.000558325059894e-06,
+      "loss": 1.3423,
       "step": 2440
     },
     {
-      "epoch": 0.7009747706422018,
-      "grad_norm": 3.2475620018827676,
-      "learning_rate": 4.9711365206974716e-06,
-      "loss": 1.0537,
+      "epoch": 0.6693128935121818,
+      "grad_norm": 0.38996615233504806,
+      "learning_rate": 5.956804436696904e-06,
+      "loss": 1.303,
       "step": 2445
     },
     {
-      "epoch": 0.7024082568807339,
-      "grad_norm": 3.0707000997716047,
-      "learning_rate": 4.927946342988699e-06,
-      "loss": 1.0271,
+      "epoch": 0.670681631535724,
+      "grad_norm": 0.3943489977366666,
+      "learning_rate": 5.9131428828903905e-06,
+      "loss": 1.3547,
       "step": 2450
     },
     {
-      "epoch": 0.7038417431192661,
-      "grad_norm": 2.9176475930747543,
-      "learning_rate": 4.884883175905671e-06,
-      "loss": 1.0535,
+      "epoch": 0.6720503695592663,
+      "grad_norm": 0.3899208460332646,
+      "learning_rate": 5.8695746607403285e-06,
+      "loss": 1.2969,
       "step": 2455
     },
     {
-      "epoch": 0.7052752293577982,
-      "grad_norm": 3.0577673743440923,
-      "learning_rate": 4.8419480978044395e-06,
-      "loss": 1.0848,
+      "epoch": 0.6734191075828087,
+      "grad_norm": 0.3940611248671062,
+      "learning_rate": 5.826100765215273e-06,
+      "loss": 1.318,
       "step": 2460
     },
     {
-      "epoch": 0.7067087155963303,
-      "grad_norm": 3.3735135945939136,
-      "learning_rate": 4.799142183833561e-06,
-      "loss": 1.0651,
+      "epoch": 0.674787845606351,
+      "grad_norm": 0.3930083338553073,
+      "learning_rate": 5.782722189129655e-06,
+      "loss": 1.2906,
       "step": 2465
     },
     {
-      "epoch": 0.7081422018348624,
-      "grad_norm": 3.170994136975781,
-      "learning_rate": 4.75646650590715e-06,
-      "loss": 1.0736,
+      "epoch": 0.6761565836298933,
+      "grad_norm": 0.39311675749157776,
+      "learning_rate": 5.739439923121077e-06,
+      "loss": 1.3569,
       "step": 2470
     },
     {
-      "epoch": 0.7095756880733946,
-      "grad_norm": 2.993299399999142,
-      "learning_rate": 4.713922132678055e-06,
-      "loss": 1.0415,
+      "epoch": 0.6775253216534355,
+      "grad_norm": 0.4114954929599597,
+      "learning_rate": 5.6962549556277134e-06,
+      "loss": 1.3773,
       "step": 2475
     },
     {
-      "epoch": 0.7110091743119266,
-      "grad_norm": 3.4225173298945117,
-      "learning_rate": 4.671510129511074e-06,
-      "loss": 1.0363,
+      "epoch": 0.6788940596769778,
+      "grad_norm": 0.38652001787654966,
+      "learning_rate": 5.653168272865724e-06,
+      "loss": 1.3008,
       "step": 2480
     },
     {
-      "epoch": 0.7124426605504587,
-      "grad_norm": 3.0344762068899365,
-      "learning_rate": 4.629231558456306e-06,
-      "loss": 1.0414,
+      "epoch": 0.6802627977005201,
+      "grad_norm": 0.3878760723325068,
+      "learning_rate": 5.6101808588067505e-06,
+      "loss": 1.2992,
       "step": 2485
     },
     {
-      "epoch": 0.7138761467889908,
-      "grad_norm": 3.1844264445491297,
-      "learning_rate": 4.587087478222539e-06,
-      "loss": 1.0714,
+      "epoch": 0.6816315357240624,
+      "grad_norm": 0.39884745029373364,
+      "learning_rate": 5.56729369515542e-06,
+      "loss": 1.3088,
       "step": 2490
     },
     {
-      "epoch": 0.7153096330275229,
-      "grad_norm": 3.09219275586818,
-      "learning_rate": 4.545078944150728e-06,
-      "loss": 1.017,
+      "epoch": 0.6830002737476047,
+      "grad_norm": 0.38961302860625757,
+      "learning_rate": 5.52450776132694e-06,
+      "loss": 1.2904,
       "step": 2495
     },
     {
-      "epoch": 0.716743119266055,
-      "grad_norm": 3.4728979799373976,
-      "learning_rate": 4.5032070081876e-06,
-      "loss": 1.0612,
+      "epoch": 0.684369011771147,
+      "grad_norm": 0.3970778873479824,
+      "learning_rate": 5.481824034424741e-06,
+      "loss": 1.3464,
       "step": 2500
     },
     {
-      "epoch": 0.7181766055045872,
-      "grad_norm": 3.063801560649391,
-      "learning_rate": 4.4614727188592835e-06,
-      "loss": 1.0798,
+      "epoch": 0.6857377497946893,
+      "grad_norm": 0.4014161218820373,
+      "learning_rate": 5.439243489218138e-06,
+      "loss": 1.3376,
       "step": 2505
     },
     {
-      "epoch": 0.7196100917431193,
-      "grad_norm": 3.823900363914515,
-      "learning_rate": 4.419877121245058e-06,
-      "loss": 1.0435,
+      "epoch": 0.6871064878182316,
+      "grad_norm": 0.3874962119825426,
+      "learning_rate": 5.396767098120087e-06,
+      "loss": 1.3078,
       "step": 2510
     },
     {
-      "epoch": 0.7210435779816514,
-      "grad_norm": 2.8618913314337107,
-      "learning_rate": 4.378421256951192e-06,
-      "loss": 1.0102,
+      "epoch": 0.6884752258417739,
+      "grad_norm": 0.37396226223807555,
+      "learning_rate": 5.354395831164982e-06,
+      "loss": 1.3097,
       "step": 2515
     },
     {
-      "epoch": 0.7224770642201835,
-      "grad_norm": 2.9378204976736466,
-      "learning_rate": 4.337106164084861e-06,
-      "loss": 1.0824,
+      "epoch": 0.6898439638653162,
+      "grad_norm": 0.3792883560955228,
+      "learning_rate": 5.312130655986485e-06,
+      "loss": 1.3158,
       "step": 2520
     },
     {
-      "epoch": 0.7239105504587156,
-      "grad_norm": 2.945681604419468,
-      "learning_rate": 4.295932877228132e-06,
-      "loss": 0.9985,
+      "epoch": 0.6912127018888585,
+      "grad_norm": 0.3798203691318429,
+      "learning_rate": 5.269972537795434e-06,
+      "loss": 1.3087,
       "step": 2525
     },
     {
-      "epoch": 0.7253440366972477,
-      "grad_norm": 3.106814303596456,
-      "learning_rate": 4.254902427412082e-06,
-      "loss": 1.0317,
+      "epoch": 0.6925814399124007,
+      "grad_norm": 0.38253235315284295,
+      "learning_rate": 5.227922439357823e-06,
+      "loss": 1.3198,
       "step": 2530
     },
     {
-      "epoch": 0.7267775229357798,
-      "grad_norm": 3.1281573814995363,
-      "learning_rate": 4.214015842090969e-06,
-      "loss": 1.0491,
+      "epoch": 0.693950177935943,
+      "grad_norm": 0.3817014501814054,
+      "learning_rate": 5.1859813209727775e-06,
+      "loss": 1.321,
       "step": 2535
     },
     {
-      "epoch": 0.7282110091743119,
-      "grad_norm": 2.795954987429199,
-      "learning_rate": 4.173274145116491e-06,
-      "loss": 0.9885,
+      "epoch": 0.6953189159594854,
+      "grad_norm": 0.3950376035714224,
+      "learning_rate": 5.144150140450643e-06,
+      "loss": 1.3024,
       "step": 2540
     },
     {
-      "epoch": 0.729644495412844,
-      "grad_norm": 3.3191476249085565,
-      "learning_rate": 4.1326783567121685e-06,
-      "loss": 1.0517,
+      "epoch": 0.6966876539830277,
+      "grad_norm": 0.4084629435066235,
+      "learning_rate": 5.102429853091128e-06,
+      "loss": 1.3437,
       "step": 2545
     },
     {
-      "epoch": 0.7310779816513762,
-      "grad_norm": 3.07441160825076,
-      "learning_rate": 4.092229493447788e-06,
-      "loss": 1.0512,
+      "epoch": 0.69805639200657,
+      "grad_norm": 0.3808185831587325,
+      "learning_rate": 5.060821411661459e-06,
+      "loss": 1.3264,
       "step": 2550
     },
     {
-      "epoch": 0.7325114678899083,
-      "grad_norm": 3.3162975946352824,
-      "learning_rate": 4.051928568213942e-06,
-      "loss": 1.0476,
+      "epoch": 0.6994251300301122,
+      "grad_norm": 0.3937638284493571,
+      "learning_rate": 5.019325766374625e-06,
+      "loss": 1.314,
       "step": 2555
     },
     {
-      "epoch": 0.7339449541284404,
-      "grad_norm": 3.222715903509193,
-      "learning_rate": 4.0117765901966635e-06,
-      "loss": 1.0401,
+      "epoch": 0.7007938680536545,
+      "grad_norm": 0.3920386340433558,
+      "learning_rate": 4.977943864867712e-06,
+      "loss": 1.3271,
       "step": 2560
     },
     {
-      "epoch": 0.7353784403669725,
-      "grad_norm": 3.1385505256509014,
-      "learning_rate": 3.9717745648521646e-06,
-      "loss": 1.0243,
+      "epoch": 0.7021626060771968,
+      "grad_norm": 0.3898729624487372,
+      "learning_rate": 4.936676652180215e-06,
+      "loss": 1.3467,
       "step": 2565
     },
     {
-      "epoch": 0.7368119266055045,
-      "grad_norm": 3.5703129773472506,
-      "learning_rate": 3.931923493881659e-06,
-      "loss": 1.0319,
+      "epoch": 0.7035313441007391,
+      "grad_norm": 0.39902316286587364,
+      "learning_rate": 4.89552507073248e-06,
+      "loss": 1.3102,
       "step": 2570
     },
     {
-      "epoch": 0.7382454128440367,
-      "grad_norm": 3.0801290385363393,
-      "learning_rate": 3.892224375206256e-06,
-      "loss": 1.0382,
+      "epoch": 0.7049000821242815,
+      "grad_norm": 0.3673934948431135,
+      "learning_rate": 4.854490060304192e-06,
+      "loss": 1.2612,
       "step": 2575
     },
     {
-      "epoch": 0.7396788990825688,
-      "grad_norm": 3.2917177730741676,
-      "learning_rate": 3.8526782029420005e-06,
-      "loss": 1.1191,
+      "epoch": 0.7062688201478237,
+      "grad_norm": 0.3922497888421968,
+      "learning_rate": 4.813572558012892e-06,
+      "loss": 1.329,
       "step": 2580
     },
     {
-      "epoch": 0.7411123853211009,
-      "grad_norm": 3.1425801332889827,
-      "learning_rate": 3.8132859673749688e-06,
-      "loss": 1.0481,
+      "epoch": 0.707637558171366,
+      "grad_norm": 0.3860080742782407,
+      "learning_rate": 4.772773498292579e-06,
+      "loss": 1.3332,
       "step": 2585
     },
     {
-      "epoch": 0.742545871559633,
-      "grad_norm": 3.1741701966057447,
-      "learning_rate": 3.774048654936454e-06,
-      "loss": 1.0072,
+      "epoch": 0.7090062961949083,
+      "grad_norm": 0.3695087849805836,
+      "learning_rate": 4.732093812872391e-06,
+      "loss": 1.2697,
       "step": 2590
     },
     {
-      "epoch": 0.7439793577981652,
-      "grad_norm": 3.4923814958157466,
-      "learning_rate": 3.7349672481782894e-06,
-      "loss": 1.01,
+      "epoch": 0.7103750342184506,
+      "grad_norm": 0.4012769440751328,
+      "learning_rate": 4.691534430755302e-06,
+      "loss": 1.3213,
       "step": 2595
     },
     {
-      "epoch": 0.7454128440366973,
-      "grad_norm": 3.121310877940095,
-      "learning_rate": 3.6960427257482343e-06,
-      "loss": 1.0804,
+      "epoch": 0.7117437722419929,
+      "grad_norm": 0.38657730746893026,
+      "learning_rate": 4.651096278196916e-06,
+      "loss": 1.3665,
       "step": 2600
     },
     {
-      "epoch": 0.7468463302752294,
-      "grad_norm": 3.1172405011336908,
-      "learning_rate": 3.657276062365457e-06,
-      "loss": 1.0313,
+      "epoch": 0.7131125102655351,
+      "grad_norm": 0.38560469814966725,
+      "learning_rate": 4.610780278684315e-06,
+      "loss": 1.3369,
       "step": 2605
     },
     {
-      "epoch": 0.7482798165137615,
-      "grad_norm": 3.0906434983571223,
-      "learning_rate": 3.618668228796143e-06,
-      "loss": 1.0378,
+      "epoch": 0.7144812482890774,
+      "grad_norm": 0.39542221048440374,
+      "learning_rate": 4.570587352914977e-06,
+      "loss": 1.3048,
       "step": 2610
     },
     {
-      "epoch": 0.7497133027522935,
-      "grad_norm": 3.1863927645425725,
-      "learning_rate": 3.580220191829178e-06,
-      "loss": 0.9929,
+      "epoch": 0.7158499863126198,
+      "grad_norm": 0.3910002548641959,
+      "learning_rate": 4.530518418775734e-06,
+      "loss": 1.3247,
       "step": 2615
     },
     {
-      "epoch": 0.7511467889908257,
-      "grad_norm": 3.0251252557702437,
-      "learning_rate": 3.5419329142519433e-06,
-      "loss": 1.0814,
+      "epoch": 0.7172187243361621,
+      "grad_norm": 0.3825608407076146,
+      "learning_rate": 4.490574391321814e-06,
+      "loss": 1.2967,
       "step": 2620
     },
     {
-      "epoch": 0.7525802752293578,
-      "grad_norm": 4.943706664249104,
-      "learning_rate": 3.5038073548261888e-06,
-      "loss": 1.0629,
+      "epoch": 0.7185874623597044,
+      "grad_norm": 0.4028460389272689,
+      "learning_rate": 4.450756182755963e-06,
+      "loss": 1.3671,
       "step": 2625
     },
     {
-      "epoch": 0.7540137614678899,
-      "grad_norm": 3.16322658334476,
-      "learning_rate": 3.46584446826405e-06,
-      "loss": 0.9881,
+      "epoch": 0.7199562003832466,
+      "grad_norm": 0.40485919090431155,
+      "learning_rate": 4.411064702407585e-06,
+      "loss": 1.3556,
       "step": 2630
     },
     {
-      "epoch": 0.755447247706422,
-      "grad_norm": 3.125095249964922,
-      "learning_rate": 3.428045205204125e-06,
-      "loss": 0.9923,
+      "epoch": 0.7213249384067889,
+      "grad_norm": 0.39021669719654245,
+      "learning_rate": 4.371500856711988e-06,
+      "loss": 1.3774,
       "step": 2635
     },
     {
-      "epoch": 0.7568807339449541,
-      "grad_norm": 2.973224447074909,
-      "learning_rate": 3.3904105121876764e-06,
-      "loss": 1.1052,
+      "epoch": 0.7226936764303312,
+      "grad_norm": 0.3958279714364836,
+      "learning_rate": 4.332065549189697e-06,
+      "loss": 1.2982,
       "step": 2640
     },
     {
-      "epoch": 0.7583142201834863,
-      "grad_norm": 2.996637163056179,
-      "learning_rate": 3.3529413316349145e-06,
-      "loss": 1.1049,
+      "epoch": 0.7240624144538735,
+      "grad_norm": 0.4060690524084766,
+      "learning_rate": 4.292759680425794e-06,
+      "loss": 1.3767,
       "step": 2645
     },
     {
-      "epoch": 0.7597477064220184,
-      "grad_norm": 2.935944096948497,
-      "learning_rate": 3.3156386018214193e-06,
-      "loss": 1.0712,
+      "epoch": 0.7254311524774159,
+      "grad_norm": 0.40926421108177974,
+      "learning_rate": 4.253584148049369e-06,
+      "loss": 1.339,
       "step": 2650
     },
     {
-      "epoch": 0.7611811926605505,
-      "grad_norm": 3.038494065615691,
-      "learning_rate": 3.2785032568546304e-06,
-      "loss": 1.0472,
+      "epoch": 0.7267998905009582,
+      "grad_norm": 0.37375698755072967,
+      "learning_rate": 4.214539846713024e-06,
+      "loss": 1.2961,
       "step": 2655
     },
     {
-      "epoch": 0.7626146788990825,
-      "grad_norm": 3.2871929054536397,
-      "learning_rate": 3.2415362266504626e-06,
-      "loss": 1.0342,
+      "epoch": 0.7281686285245004,
+      "grad_norm": 0.3968852233360893,
+      "learning_rate": 4.175627668072425e-06,
+      "loss": 1.2962,
       "step": 2660
     },
     {
-      "epoch": 0.7640481651376146,
-      "grad_norm": 3.098314519979228,
-      "learning_rate": 3.2047384369100065e-06,
-      "loss": 1.0903,
+      "epoch": 0.7295373665480427,
+      "grad_norm": 0.4073445114415598,
+      "learning_rate": 4.136848500765948e-06,
+      "loss": 1.4056,
       "step": 2665
     },
     {
-      "epoch": 0.7654816513761468,
-      "grad_norm": 2.9569161427598787,
-      "learning_rate": 3.16811080909637e-06,
-      "loss": 1.0007,
+      "epoch": 0.730906104571585,
+      "grad_norm": 0.3950383171619063,
+      "learning_rate": 4.098203230394399e-06,
+      "loss": 1.3192,
       "step": 2670
     },
     {
-      "epoch": 0.7669151376146789,
-      "grad_norm": 3.355325001454353,
-      "learning_rate": 3.1316542604115853e-06,
-      "loss": 1.0525,
+      "epoch": 0.7322748425951273,
+      "grad_norm": 0.3986322017273241,
+      "learning_rate": 4.059692739500761e-06,
+      "loss": 1.3485,
       "step": 2675
     },
     {
-      "epoch": 0.768348623853211,
-      "grad_norm": 3.7866942452127113,
-      "learning_rate": 3.095369703773652e-06,
-      "loss": 1.0607,
+      "epoch": 0.7336435806186696,
+      "grad_norm": 0.3848930065792978,
+      "learning_rate": 4.02131790755006e-06,
+      "loss": 1.378,
       "step": 2680
     },
     {
-      "epoch": 0.7697821100917431,
-      "grad_norm": 2.8549205403217024,
-      "learning_rate": 3.0592580477936606e-06,
-      "loss": 1.0612,
+      "epoch": 0.7350123186422118,
+      "grad_norm": 0.39139666434515563,
+      "learning_rate": 3.983079610909283e-06,
+      "loss": 1.3334,
       "step": 2685
     },
     {
-      "epoch": 0.7712155963302753,
-      "grad_norm": 3.1911721995721547,
-      "learning_rate": 3.0233201967530647e-06,
-      "loss": 1.0756,
+      "epoch": 0.7363810566657542,
+      "grad_norm": 0.38858454704969564,
+      "learning_rate": 3.944978722827347e-06,
+      "loss": 1.3235,
       "step": 2690
     },
     {
-      "epoch": 0.7726490825688074,
-      "grad_norm": 3.254594926845702,
-      "learning_rate": 2.987557050581017e-06,
-      "loss": 1.0928,
+      "epoch": 0.7377497946892965,
+      "grad_norm": 0.390122324459421,
+      "learning_rate": 3.907016113415166e-06,
+      "loss": 1.3583,
       "step": 2695
     },
     {
-      "epoch": 0.7740825688073395,
-      "grad_norm": 3.029915420885045,
-      "learning_rate": 2.9519695048318353e-06,
-      "loss": 0.9834,
+      "epoch": 0.7391185327128388,
+      "grad_norm": 0.3919502066927978,
+      "learning_rate": 3.869192649625792e-06,
+      "loss": 1.3092,
       "step": 2700
     },
     {
-      "epoch": 0.7755160550458715,
-      "grad_norm": 3.7010803603107614,
-      "learning_rate": 2.9165584506625864e-06,
-      "loss": 1.0616,
+      "epoch": 0.7404872707363811,
+      "grad_norm": 0.39777490757009926,
+      "learning_rate": 3.831509195234598e-06,
+      "loss": 1.3358,
       "step": 2705
     },
     {
-      "epoch": 0.7769495412844036,
-      "grad_norm": 3.0973140862278155,
-      "learning_rate": 2.8813247748107665e-06,
-      "loss": 1.0493,
+      "epoch": 0.7418560087599233,
+      "grad_norm": 0.39353868348346743,
+      "learning_rate": 3.793966610819545e-06,
+      "loss": 1.3515,
       "step": 2710
     },
     {
-      "epoch": 0.7783830275229358,
-      "grad_norm": 3.3576787529696217,
-      "learning_rate": 2.8462693595720938e-06,
-      "loss": 1.0743,
+      "epoch": 0.7432247467834656,
+      "grad_norm": 0.39342447998988617,
+      "learning_rate": 3.756565753741569e-06,
+      "loss": 1.2885,
       "step": 2715
     },
     {
-      "epoch": 0.7798165137614679,
-      "grad_norm": 3.074976481415647,
-      "learning_rate": 2.8113930827784076e-06,
-      "loss": 1.0157,
+      "epoch": 0.7445934848070079,
+      "grad_norm": 0.37781062124467246,
+      "learning_rate": 3.7193074781249585e-06,
+      "loss": 1.276,
       "step": 2720
     },
     {
-      "epoch": 0.78125,
-      "grad_norm": 3.0753397191274945,
-      "learning_rate": 2.776696817775707e-06,
-      "loss": 1.0266,
+      "epoch": 0.7459622228305502,
+      "grad_norm": 0.38397833073923665,
+      "learning_rate": 3.6821926348378666e-06,
+      "loss": 1.3031,
       "step": 2725
     },
     {
-      "epoch": 0.7826834862385321,
-      "grad_norm": 3.09366006809429,
-      "learning_rate": 2.7421814334022624e-06,
-      "loss": 1.0243,
+      "epoch": 0.7473309608540926,
+      "grad_norm": 0.3993776914034358,
+      "learning_rate": 3.6452220714728883e-06,
+      "loss": 1.3719,
       "step": 2730
     },
     {
-      "epoch": 0.7841169724770642,
-      "grad_norm": 3.6168913804023326,
-      "learning_rate": 2.7078477939668625e-06,
-      "loss": 1.0652,
+      "epoch": 0.7486996988776348,
+      "grad_norm": 0.387591612397714,
+      "learning_rate": 3.608396632327684e-06,
+      "loss": 1.3006,
       "step": 2735
     },
     {
-      "epoch": 0.7855504587155964,
-      "grad_norm": 4.050361789957049,
-      "learning_rate": 2.673696759227177e-06,
-      "loss": 1.0636,
+      "epoch": 0.7500684369011771,
+      "grad_norm": 0.4059102325008591,
+      "learning_rate": 3.5717171583857115e-06,
+      "loss": 1.3689,
       "step": 2740
     },
     {
-      "epoch": 0.7869839449541285,
-      "grad_norm": 3.249151536222312,
-      "learning_rate": 2.639729184368226e-06,
-      "loss": 1.0643,
+      "epoch": 0.7514371749247194,
+      "grad_norm": 0.3772375625162727,
+      "learning_rate": 3.5351844872970233e-06,
+      "loss": 1.3091,
       "step": 2745
     },
     {
-      "epoch": 0.7884174311926605,
-      "grad_norm": 2.9287260819204692,
-      "learning_rate": 2.6059459199809545e-06,
-      "loss": 1.0028,
+      "epoch": 0.7528059129482617,
+      "grad_norm": 0.3853940006094772,
+      "learning_rate": 3.498799453359124e-06,
+      "loss": 1.3454,
       "step": 2750
     },
     {
-      "epoch": 0.7898509174311926,
-      "grad_norm": 3.0465964857735663,
-      "learning_rate": 2.5723478120409474e-06,
-      "loss": 1.0421,
+      "epoch": 0.754174650971804,
+      "grad_norm": 0.3771144405512011,
+      "learning_rate": 3.462562887497927e-06,
+      "loss": 1.2998,
       "step": 2755
     },
     {
-      "epoch": 0.7912844036697247,
-      "grad_norm": 3.153016800252537,
-      "learning_rate": 2.5389357018872405e-06,
-      "loss": 1.0201,
+      "epoch": 0.7555433889953463,
+      "grad_norm": 0.39604150906533137,
+      "learning_rate": 3.4264756172487813e-06,
+      "loss": 1.3129,
       "step": 2760
     },
     {
-      "epoch": 0.7927178899082569,
-      "grad_norm": 3.115248902197113,
-      "learning_rate": 2.505710426201239e-06,
-      "loss": 1.1122,
+      "epoch": 0.7569121270188885,
+      "grad_norm": 0.3843657920576998,
+      "learning_rate": 3.390538466737564e-06,
+      "loss": 1.2981,
       "step": 2765
     },
     {
-      "epoch": 0.794151376146789,
-      "grad_norm": 3.456526732392788,
-      "learning_rate": 2.4726728169857885e-06,
-      "loss": 0.9736,
+      "epoch": 0.7582808650424309,
+      "grad_norm": 0.39238834291334906,
+      "learning_rate": 3.3547522566618593e-06,
+      "loss": 1.2697,
       "step": 2770
     },
     {
-      "epoch": 0.7955848623853211,
-      "grad_norm": 3.078706890825068,
-      "learning_rate": 2.439823701544328e-06,
-      "loss": 1.0208,
+      "epoch": 0.7596496030659732,
+      "grad_norm": 0.4030855832137369,
+      "learning_rate": 3.319117804272236e-06,
+      "loss": 1.3152,
       "step": 2775
     },
     {
-      "epoch": 0.7970183486238532,
-      "grad_norm": 3.0852747181382485,
-      "learning_rate": 2.407163902460167e-06,
-      "loss": 1.0112,
+      "epoch": 0.7610183410895155,
+      "grad_norm": 0.38853914068753576,
+      "learning_rate": 3.283635923353553e-06,
+      "loss": 1.3125,
       "step": 2780
     },
     {
-      "epoch": 0.7984518348623854,
-      "grad_norm": 2.9795351626224202,
-      "learning_rate": 2.3746942375758986e-06,
-      "loss": 1.0495,
+      "epoch": 0.7623870791130578,
+      "grad_norm": 0.39575307616938343,
+      "learning_rate": 3.248307424206395e-06,
+      "loss": 1.4215,
       "step": 2785
     },
     {
-      "epoch": 0.7998853211009175,
-      "grad_norm": 3.662965002906388,
-      "learning_rate": 2.3424155199729206e-06,
-      "loss": 1.0088,
+      "epoch": 0.7637558171366,
+      "grad_norm": 0.3903008557955737,
+      "learning_rate": 3.2131331136285717e-06,
+      "loss": 1.3034,
       "step": 2790
     },
     {
-      "epoch": 0.8013188073394495,
-      "grad_norm": 2.9979503759811146,
-      "learning_rate": 2.310328557951065e-06,
-      "loss": 0.9978,
+      "epoch": 0.7651245551601423,
+      "grad_norm": 0.38760587663102863,
+      "learning_rate": 3.1781137948966754e-06,
+      "loss": 1.291,
       "step": 2795
     },
     {
-      "epoch": 0.8027522935779816,
-      "grad_norm": 3.1993479502657856,
-      "learning_rate": 2.2784341550083577e-06,
-      "loss": 1.08,
+      "epoch": 0.7664932931836846,
+      "grad_norm": 0.3917779069091076,
+      "learning_rate": 3.1432502677477494e-06,
+      "loss": 1.3318,
       "step": 2800
     },
     {
-      "epoch": 0.8041857798165137,
-      "grad_norm": 2.9645752085086716,
-      "learning_rate": 2.2467331098209098e-06,
-      "loss": 1.0019,
+      "epoch": 0.767862031207227,
+      "grad_norm": 0.3978532288818969,
+      "learning_rate": 3.108543328361017e-06,
+      "loss": 1.3074,
       "step": 2805
     },
     {
-      "epoch": 0.8056192660550459,
-      "grad_norm": 3.0604845129074856,
-      "learning_rate": 2.215226216222911e-06,
-      "loss": 0.9847,
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.3931552217963087,
+      "learning_rate": 3.0739937693397113e-06,
+      "loss": 1.3181,
       "step": 2810
     },
     {
-      "epoch": 0.807052752293578,
-      "grad_norm": 3.1249395082814235,
-      "learning_rate": 2.1839142631867396e-06,
-      "loss": 1.003,
+      "epoch": 0.7705995072543115,
+      "grad_norm": 0.38132685696899987,
+      "learning_rate": 3.0396023796929597e-06,
+      "loss": 1.3202,
       "step": 2815
     },
     {
-      "epoch": 0.8084862385321101,
-      "grad_norm": 3.5887794377789746,
-      "learning_rate": 2.1527980348032263e-06,
-      "loss": 1.0915,
+      "epoch": 0.7719682452778538,
+      "grad_norm": 0.3895674318092009,
+      "learning_rate": 3.0053699448177687e-06,
+      "loss": 1.2912,
       "step": 2820
     },
     {
-      "epoch": 0.8099197247706422,
-      "grad_norm": 2.9913673167068056,
-      "learning_rate": 2.121878310262008e-06,
-      "loss": 1.0346,
+      "epoch": 0.7733369833013961,
+      "grad_norm": 0.3923145835781509,
+      "learning_rate": 2.971297246481101e-06,
+      "loss": 1.2874,
       "step": 2825
     },
     {
-      "epoch": 0.8113532110091743,
-      "grad_norm": 3.076522728603463,
-      "learning_rate": 2.0911558638320117e-06,
-      "loss": 1.0254,
+      "epoch": 0.7747057213249384,
+      "grad_norm": 0.3964424700594031,
+      "learning_rate": 2.937385062802004e-06,
+      "loss": 1.3723,
       "step": 2830
     },
     {
-      "epoch": 0.8127866972477065,
-      "grad_norm": 3.4952190302417625,
-      "learning_rate": 2.0606314648420757e-06,
-      "loss": 1.0164,
+      "epoch": 0.7760744593484807,
+      "grad_norm": 0.39752354179970156,
+      "learning_rate": 2.9036341682338466e-06,
+      "loss": 1.3329,
       "step": 2835
     },
     {
-      "epoch": 0.8142201834862385,
-      "grad_norm": 2.987547482751348,
-      "learning_rate": 2.0303058776616847e-06,
-      "loss": 1.0241,
+      "epoch": 0.7774431973720229,
+      "grad_norm": 0.39323110376722525,
+      "learning_rate": 2.870045333546644e-06,
+      "loss": 1.2905,
       "step": 2840
     },
     {
-      "epoch": 0.8156536697247706,
-      "grad_norm": 3.0295895740792265,
-      "learning_rate": 2.0001798616818137e-06,
-      "loss": 0.9397,
+      "epoch": 0.7788119353955653,
+      "grad_norm": 0.3924911670687984,
+      "learning_rate": 2.8366193258094355e-06,
+      "loss": 1.292,
       "step": 2845
     },
     {
-      "epoch": 0.8170871559633027,
-      "grad_norm": 3.2124562930166376,
-      "learning_rate": 1.970254171295931e-06,
-      "loss": 1.0349,
+      "epoch": 0.7801806734191076,
+      "grad_norm": 0.372337059578975,
+      "learning_rate": 2.8033569083727797e-06,
+      "loss": 1.3368,
       "step": 2850
     },
     {
-      "epoch": 0.8185206422018348,
-      "grad_norm": 2.7811469673449403,
-      "learning_rate": 1.940529555881101e-06,
-      "loss": 1.0372,
+      "epoch": 0.7815494114426499,
+      "grad_norm": 0.386256983183401,
+      "learning_rate": 2.7702588408513276e-06,
+      "loss": 1.3236,
       "step": 2855
     },
     {
-      "epoch": 0.819954128440367,
-      "grad_norm": 3.0716049369726854,
-      "learning_rate": 1.9110067597792094e-06,
-      "loss": 1.0258,
+      "epoch": 0.7829181494661922,
+      "grad_norm": 0.4151085195929508,
+      "learning_rate": 2.7373258791064572e-06,
+      "loss": 1.3342,
       "step": 2860
     },
     {
-      "epoch": 0.8213876146788991,
-      "grad_norm": 3.31583577307639,
-      "learning_rate": 1.8816865222783354e-06,
-      "loss": 1.0659,
+      "epoch": 0.7842868874897345,
+      "grad_norm": 0.39410387314037304,
+      "learning_rate": 2.7045587752290224e-06,
+      "loss": 1.28,
       "step": 2865
     },
     {
-      "epoch": 0.8228211009174312,
-      "grad_norm": 3.3545193733702616,
-      "learning_rate": 1.8525695775942376e-06,
-      "loss": 0.997,
+      "epoch": 0.7856556255132767,
+      "grad_norm": 0.40070665770653285,
+      "learning_rate": 2.6719582775221862e-06,
+      "loss": 1.3336,
       "step": 2870
     },
     {
-      "epoch": 0.8242545871559633,
-      "grad_norm": 3.082738506583357,
-      "learning_rate": 1.8236566548519664e-06,
-      "loss": 1.0109,
+      "epoch": 0.787024363536819,
+      "grad_norm": 0.3842777541641571,
+      "learning_rate": 2.6395251304843137e-06,
+      "loss": 1.2757,
       "step": 2875
     },
     {
-      "epoch": 0.8256880733944955,
-      "grad_norm": 2.9272236463346326,
-      "learning_rate": 1.7949484780675941e-06,
-      "loss": 1.0096,
+      "epoch": 0.7883931015603614,
+      "grad_norm": 0.40580643578053277,
+      "learning_rate": 2.6072600747919773e-06,
+      "loss": 1.3196,
       "step": 2880
     },
     {
-      "epoch": 0.8271215596330275,
-      "grad_norm": 2.9288594245035937,
-      "learning_rate": 1.7664457661301103e-06,
-      "loss": 1.0126,
+      "epoch": 0.7897618395839037,
+      "grad_norm": 0.388179506322075,
+      "learning_rate": 2.575163847283053e-06,
+      "loss": 1.293,
       "step": 2885
     },
     {
-      "epoch": 0.8285550458715596,
-      "grad_norm": 2.9288774104923077,
-      "learning_rate": 1.7381492327834004e-06,
-      "loss": 1.0499,
+      "epoch": 0.791130577607446,
+      "grad_norm": 0.3732450173203748,
+      "learning_rate": 2.543237180939875e-06,
+      "loss": 1.2751,
       "step": 2890
     },
     {
-      "epoch": 0.8299885321100917,
-      "grad_norm": 2.7877742435190065,
-      "learning_rate": 1.7100595866083713e-06,
-      "loss": 1.0041,
+      "epoch": 0.7924993156309882,
+      "grad_norm": 0.38506878802774175,
+      "learning_rate": 2.5114808048725035e-06,
+      "loss": 1.3083,
       "step": 2895
     },
     {
-      "epoch": 0.8314220183486238,
-      "grad_norm": 3.0286412164914656,
-      "learning_rate": 1.6821775310052212e-06,
-      "loss": 0.9811,
+      "epoch": 0.7938680536545305,
+      "grad_norm": 0.4075860384902042,
+      "learning_rate": 2.479895444302086e-06,
+      "loss": 1.307,
       "step": 2900
     },
     {
-      "epoch": 0.832855504587156,
-      "grad_norm": 2.7532642945829537,
-      "learning_rate": 1.65450376417582e-06,
-      "loss": 1.0096,
+      "epoch": 0.7952367916780728,
+      "grad_norm": 0.39287683770161946,
+      "learning_rate": 2.4484818205442763e-06,
+      "loss": 1.3645,
       "step": 2905
     },
     {
-      "epoch": 0.8342889908256881,
-      "grad_norm": 2.9265157315568375,
-      "learning_rate": 1.6270389791062146e-06,
-      "loss": 1.0491,
+      "epoch": 0.7966055297016151,
+      "grad_norm": 0.3882005380231924,
+      "learning_rate": 2.417240650992767e-06,
+      "loss": 1.3512,
       "step": 2910
     },
     {
-      "epoch": 0.8357224770642202,
-      "grad_norm": 3.216652891176282,
-      "learning_rate": 1.5997838635492936e-06,
-      "loss": 1.0956,
+      "epoch": 0.7979742677251574,
+      "grad_norm": 0.3866431713970654,
+      "learning_rate": 2.3861726491029237e-06,
+      "loss": 1.2793,
       "step": 2915
     },
     {
-      "epoch": 0.8371559633027523,
-      "grad_norm": 3.242345149185627,
-      "learning_rate": 1.5727391000075542e-06,
-      "loss": 0.9864,
+      "epoch": 0.7993430057486997,
+      "grad_norm": 0.3798968085022952,
+      "learning_rate": 2.355278524375465e-06,
+      "loss": 1.2865,
       "step": 2920
     },
     {
-      "epoch": 0.8385894495412844,
-      "grad_norm": 3.2259601086687204,
-      "learning_rate": 1.5459053657160084e-06,
-      "loss": 1.1009,
+      "epoch": 0.800711743772242,
+      "grad_norm": 0.39453328832547074,
+      "learning_rate": 2.324558982340275e-06,
+      "loss": 1.2576,
       "step": 2925
     },
     {
-      "epoch": 0.8400229357798165,
-      "grad_norm": 2.744523780896053,
-      "learning_rate": 1.5192833326252377e-06,
-      "loss": 1.0163,
+      "epoch": 0.8020804817957843,
+      "grad_norm": 0.3764906658011197,
+      "learning_rate": 2.2940147245402944e-06,
+      "loss": 1.3167,
       "step": 2930
     },
     {
-      "epoch": 0.8414564220183486,
-      "grad_norm": 2.828910888875833,
-      "learning_rate": 1.4928736673845534e-06,
-      "loss": 1.0037,
+      "epoch": 0.8034492198193266,
+      "grad_norm": 0.3823354264503374,
+      "learning_rate": 2.2636464485154875e-06,
+      "loss": 1.334,
       "step": 2935
     },
     {
-      "epoch": 0.8428899082568807,
-      "grad_norm": 3.062048982707707,
-      "learning_rate": 1.4666770313253054e-06,
-      "loss": 1.0248,
+      "epoch": 0.8048179578428689,
+      "grad_norm": 0.3911098766218853,
+      "learning_rate": 2.23345484778692e-06,
+      "loss": 1.3856,
       "step": 2940
     },
     {
-      "epoch": 0.8443233944954128,
-      "grad_norm": 3.1169358681462174,
-      "learning_rate": 1.4406940804443303e-06,
-      "loss": 1.0094,
+      "epoch": 0.8061866958664111,
+      "grad_norm": 0.4024307217930752,
+      "learning_rate": 2.2034406118409178e-06,
+      "loss": 1.3704,
       "step": 2945
     },
     {
-      "epoch": 0.845756880733945,
-      "grad_norm": 3.045993870767794,
-      "learning_rate": 1.4149254653875167e-06,
-      "loss": 1.0523,
+      "epoch": 0.8075554338899534,
+      "grad_norm": 0.38312668558103635,
+      "learning_rate": 2.1736044261133305e-06,
+      "loss": 1.32,
       "step": 2950
     },
     {
-      "epoch": 0.8471903669724771,
-      "grad_norm": 3.0564501410586637,
-      "learning_rate": 1.389371831433507e-06,
-      "loss": 1.0095,
+      "epoch": 0.8089241719134957,
+      "grad_norm": 0.39944235808095113,
+      "learning_rate": 2.1439469719738615e-06,
+      "loss": 1.3348,
       "step": 2955
     },
     {
-      "epoch": 0.8486238532110092,
-      "grad_norm": 3.4108278621499917,
-      "learning_rate": 1.3640338184775526e-06,
-      "loss": 1.0366,
+      "epoch": 0.8102929099370381,
+      "grad_norm": 0.37362368240209753,
+      "learning_rate": 2.1144689267105213e-06,
+      "loss": 1.3058,
       "step": 2960
     },
     {
-      "epoch": 0.8500573394495413,
-      "grad_norm": 3.309723042860519,
-      "learning_rate": 1.3389120610154804e-06,
-      "loss": 1.015,
+      "epoch": 0.8116616479605804,
+      "grad_norm": 0.3977426993128578,
+      "learning_rate": 2.0851709635141526e-06,
+      "loss": 1.329,
       "step": 2965
     },
     {
-      "epoch": 0.8514908256880734,
-      "grad_norm": 3.40140275029476,
-      "learning_rate": 1.3140071881278106e-06,
-      "loss": 1.1159,
+      "epoch": 0.8130303859841227,
+      "grad_norm": 0.390250388707541,
+      "learning_rate": 2.0560537514630595e-06,
+      "loss": 1.3435,
       "step": 2970
     },
     {
-      "epoch": 0.8529243119266054,
-      "grad_norm": 3.0642718892804344,
-      "learning_rate": 1.2893198234639904e-06,
-      "loss": 0.983,
+      "epoch": 0.8143991240076649,
+      "grad_norm": 0.41070469591969644,
+      "learning_rate": 2.0271179555077357e-06,
+      "loss": 1.3172,
       "step": 2975
     },
     {
-      "epoch": 0.8543577981651376,
-      "grad_norm": 3.5196505954999315,
-      "learning_rate": 1.2648505852267956e-06,
-      "loss": 1.0422,
+      "epoch": 0.8157678620312072,
+      "grad_norm": 0.378348117809942,
+      "learning_rate": 1.998364236455661e-06,
+      "loss": 1.3027,
       "step": 2980
     },
     {
-      "epoch": 0.8557912844036697,
-      "grad_norm": 3.012388196153649,
-      "learning_rate": 1.240600086156839e-06,
-      "loss": 1.0161,
+      "epoch": 0.8171366000547495,
+      "grad_norm": 0.380190609171277,
+      "learning_rate": 1.969793250956221e-06,
+      "loss": 1.2577,
       "step": 2985
     },
     {
-      "epoch": 0.8572247706422018,
-      "grad_norm": 4.224630882517724,
-      "learning_rate": 1.2165689335172248e-06,
-      "loss": 1.0435,
+      "epoch": 0.8185053380782918,
+      "grad_norm": 0.38655656053220244,
+      "learning_rate": 1.9414056514857205e-06,
+      "loss": 1.3137,
       "step": 2990
     },
     {
-      "epoch": 0.8586582568807339,
-      "grad_norm": 3.3927019439454558,
-      "learning_rate": 1.1927577290783488e-06,
-      "loss": 1.0549,
+      "epoch": 0.8198740761018342,
+      "grad_norm": 0.38867016212920535,
+      "learning_rate": 1.913202086332463e-06,
+      "loss": 1.3597,
       "step": 2995
     },
     {
-      "epoch": 0.8600917431192661,
-      "grad_norm": 2.8680598337924965,
-      "learning_rate": 1.169167069102828e-06,
-      "loss": 1.062,
+      "epoch": 0.8212428141253764,
+      "grad_norm": 0.38641157633799067,
+      "learning_rate": 1.8851831995819569e-06,
+      "loss": 1.3184,
       "step": 3000
     },
     {
-      "epoch": 0.8615252293577982,
-      "grad_norm": 3.2446852778299453,
-      "learning_rate": 1.1457975443305625e-06,
-      "loss": 1.0439,
+      "epoch": 0.8226115521489187,
+      "grad_norm": 0.39081114681314566,
+      "learning_rate": 1.8573496311022133e-06,
+      "loss": 1.3219,
       "step": 3005
     },
     {
-      "epoch": 0.8629587155963303,
-      "grad_norm": 3.074025983059374,
-      "learning_rate": 1.1226497399639501e-06,
-      "loss": 1.0202,
+      "epoch": 0.823980290172461,
+      "grad_norm": 0.37733544335013924,
+      "learning_rate": 1.8297020165291158e-06,
+      "loss": 1.2927,
       "step": 3010
     },
     {
-      "epoch": 0.8643922018348624,
-      "grad_norm": 2.9805357814293423,
-      "learning_rate": 1.0997242356532335e-06,
-      "loss": 0.9824,
+      "epoch": 0.8253490281960033,
+      "grad_norm": 0.38748075262103254,
+      "learning_rate": 1.8022409872519197e-06,
+      "loss": 1.3184,
       "step": 3015
     },
     {
-      "epoch": 0.8658256880733946,
-      "grad_norm": 3.377450943710277,
-      "learning_rate": 1.0770216054819782e-06,
-      "loss": 1.0292,
+      "epoch": 0.8267177662195456,
+      "grad_norm": 0.40138106139790847,
+      "learning_rate": 1.7749671703988226e-06,
+      "loss": 1.3312,
       "step": 3020
     },
     {
-      "epoch": 0.8672591743119266,
-      "grad_norm": 3.203588510488362,
-      "learning_rate": 1.0545424179526963e-06,
-      "loss": 1.0008,
+      "epoch": 0.8280865042430878,
+      "grad_norm": 0.3876362883781664,
+      "learning_rate": 1.7478811888226555e-06,
+      "loss": 1.3101,
       "step": 3025
     },
     {
-      "epoch": 0.8686926605504587,
-      "grad_norm": 3.3838873878755438,
-      "learning_rate": 1.03228723597262e-06,
-      "loss": 1.0276,
+      "epoch": 0.8294552422666301,
+      "grad_norm": 0.3936459314036851,
+      "learning_rate": 1.7209836610866426e-06,
+      "loss": 1.3431,
       "step": 3030
     },
     {
-      "epoch": 0.8701261467889908,
-      "grad_norm": 3.0207979512808247,
-      "learning_rate": 1.0102566168395977e-06,
-      "loss": 1.0448,
+      "epoch": 0.8308239802901725,
+      "grad_norm": 0.38764243587502206,
+      "learning_rate": 1.694275201450284e-06,
+      "loss": 1.3386,
       "step": 3035
     },
     {
-      "epoch": 0.8715596330275229,
-      "grad_norm": 3.294936816848852,
-      "learning_rate": 9.884511122281427e-07,
-      "loss": 1.0123,
+      "epoch": 0.8321927183137148,
+      "grad_norm": 0.3908568598595201,
+      "learning_rate": 1.6677564198553332e-06,
+      "loss": 1.3342,
       "step": 3040
     },
     {
-      "epoch": 0.872993119266055,
-      "grad_norm": 3.0263492251087802,
-      "learning_rate": 9.668712681756087e-07,
-      "loss": 0.9943,
+      "epoch": 0.8335614563372571,
+      "grad_norm": 0.38588972696708185,
+      "learning_rate": 1.6414279219118568e-06,
+      "loss": 1.3527,
       "step": 3045
     },
     {
-      "epoch": 0.8744266055045872,
-      "grad_norm": 2.987557644263372,
-      "learning_rate": 9.455176250685338e-07,
-      "loss": 1.0392,
+      "epoch": 0.8349301943607993,
+      "grad_norm": 0.3907806499233958,
+      "learning_rate": 1.6152903088844051e-06,
+      "loss": 1.3104,
       "step": 3050
     },
     {
-      "epoch": 0.8758600917431193,
-      "grad_norm": 2.827866834943389,
-      "learning_rate": 9.243907176290945e-07,
-      "loss": 1.0009,
+      "epoch": 0.8362989323843416,
+      "grad_norm": 0.3825824287507951,
+      "learning_rate": 1.5893441776782947e-06,
+      "loss": 1.3062,
       "step": 3055
     },
     {
-      "epoch": 0.8772935779816514,
-      "grad_norm": 3.0479507652012106,
-      "learning_rate": 9.034910749017211e-07,
-      "loss": 1.0362,
+      "epoch": 0.8376676704078839,
+      "grad_norm": 0.3842728055099776,
+      "learning_rate": 1.5635901208259608e-06,
+      "loss": 1.3581,
       "step": 3060
     },
     {
-      "epoch": 0.8787270642201835,
-      "grad_norm": 3.166615479232945,
-      "learning_rate": 8.828192202398455e-07,
-      "loss": 1.0649,
+      "epoch": 0.8390364084314262,
+      "grad_norm": 0.3905623616859072,
+      "learning_rate": 1.5380287264734285e-06,
+      "loss": 1.3148,
       "step": 3065
     },
     {
-      "epoch": 0.8801605504587156,
-      "grad_norm": 3.086530733524286,
-      "learning_rate": 8.623756712928022e-07,
-      "loss": 1.0161,
+      "epoch": 0.8404051464549686,
+      "grad_norm": 0.39844460827488454,
+      "learning_rate": 1.5126605783668945e-06,
+      "loss": 1.3074,
       "step": 3070
     },
     {
-      "epoch": 0.8815940366972477,
-      "grad_norm": 3.2339942958362995,
-      "learning_rate": 8.421609399928621e-07,
-      "loss": 1.0274,
+      "epoch": 0.8417738844785108,
+      "grad_norm": 0.39217288885390583,
+      "learning_rate": 1.4874862558393787e-06,
+      "loss": 1.3171,
       "step": 3075
     },
     {
-      "epoch": 0.8830275229357798,
-      "grad_norm": 3.1624079222447845,
-      "learning_rate": 8.221755325424152e-07,
-      "loss": 0.963,
+      "epoch": 0.8431426225020531,
+      "grad_norm": 0.38303308206260517,
+      "learning_rate": 1.462506333797501e-06,
+      "loss": 1.2985,
       "step": 3080
     },
     {
-      "epoch": 0.8844610091743119,
-      "grad_norm": 6.67242001014344,
-      "learning_rate": 8.024199494012863e-07,
-      "loss": 0.994,
+      "epoch": 0.8445113605255954,
+      "grad_norm": 0.37178133903556354,
+      "learning_rate": 1.4377213827083602e-06,
+      "loss": 1.278,
       "step": 3085
     },
     {
-      "epoch": 0.885894495412844,
-      "grad_norm": 3.0370868865276424,
-      "learning_rate": 7.828946852742148e-07,
-      "loss": 1.0551,
+      "epoch": 0.8458800985491377,
+      "grad_norm": 0.383873936282195,
+      "learning_rate": 1.413131968586491e-06,
+      "loss": 1.2989,
       "step": 3090
     },
     {
-      "epoch": 0.8873279816513762,
-      "grad_norm": 3.1653796046419185,
-      "learning_rate": 7.636002290984634e-07,
-      "loss": 1.0085,
+      "epoch": 0.84724883657268,
+      "grad_norm": 0.3877611566994199,
+      "learning_rate": 1.3887386529809454e-06,
+      "loss": 1.3543,
       "step": 3095
     },
     {
-      "epoch": 0.8887614678899083,
-      "grad_norm": 3.018418033888261,
-      "learning_rate": 7.445370640315642e-07,
-      "loss": 1.0017,
+      "epoch": 0.8486175745962223,
+      "grad_norm": 0.3844140889489012,
+      "learning_rate": 1.364541992962476e-06,
+      "loss": 1.345,
       "step": 3100
     },
     {
-      "epoch": 0.8901949541284404,
-      "grad_norm": 3.543389563066827,
-      "learning_rate": 7.257056674392359e-07,
-      "loss": 1.002,
+      "epoch": 0.8499863126197645,
+      "grad_norm": 0.3791779816963577,
+      "learning_rate": 1.3405425411108008e-06,
+      "loss": 1.3202,
       "step": 3105
     },
     {
-      "epoch": 0.8916284403669725,
-      "grad_norm": 3.1380320671176194,
-      "learning_rate": 7.071065108834197e-07,
-      "loss": 1.0311,
+      "epoch": 0.8513550506433069,
+      "grad_norm": 0.37802806800903216,
+      "learning_rate": 1.3167408455019903e-06,
+      "loss": 1.3,
       "step": 3110
     },
     {
-      "epoch": 0.8930619266055045,
-      "grad_norm": 3.107368067479966,
-      "learning_rate": 6.887400601104688e-07,
-      "loss": 0.9969,
+      "epoch": 0.8527237886668492,
+      "grad_norm": 0.38421255363489804,
+      "learning_rate": 1.2931374496959548e-06,
+      "loss": 1.3033,
       "step": 3115
     },
     {
-      "epoch": 0.8944954128440367,
-      "grad_norm": 2.86265063408354,
-      "learning_rate": 6.706067750394951e-07,
-      "loss": 1.0004,
+      "epoch": 0.8540925266903915,
+      "grad_norm": 0.3890245703911996,
+      "learning_rate": 1.2697328927240238e-06,
+      "loss": 1.3155,
       "step": 3120
     },
     {
-      "epoch": 0.8959288990825688,
-      "grad_norm": 2.921057543794976,
-      "learning_rate": 6.527071097508475e-07,
-      "loss": 1.0568,
+      "epoch": 0.8554612647139338,
+      "grad_norm": 0.38965532355017113,
+      "learning_rate": 1.2465277090766381e-06,
+      "loss": 1.3408,
       "step": 3125
     },
     {
-      "epoch": 0.8973623853211009,
-      "grad_norm": 3.413457128089623,
-      "learning_rate": 6.350415124747378e-07,
-      "loss": 1.0966,
+      "epoch": 0.856830002737476,
+      "grad_norm": 0.38847581936098935,
+      "learning_rate": 1.2235224286911495e-06,
+      "loss": 1.3619,
       "step": 3130
     },
     {
-      "epoch": 0.898795871559633,
-      "grad_norm": 3.2754484907337913,
-      "learning_rate": 6.176104255800175e-07,
-      "loss": 1.0012,
+      "epoch": 0.8581987407610183,
+      "grad_norm": 0.3901578768967818,
+      "learning_rate": 1.2007175769397117e-06,
+      "loss": 1.3714,
       "step": 3135
     },
     {
-      "epoch": 0.9002293577981652,
-      "grad_norm": 3.096817656277008,
-      "learning_rate": 6.004142855631068e-07,
-      "loss": 1.0409,
+      "epoch": 0.8595674787845606,
+      "grad_norm": 0.38170228292380465,
+      "learning_rate": 1.178113674617285e-06,
+      "loss": 1.3144,
       "step": 3140
     },
     {
-      "epoch": 0.9016628440366973,
-      "grad_norm": 3.2107227855247737,
-      "learning_rate": 5.834535230370586e-07,
-      "loss": 1.022,
+      "epoch": 0.860936216808103,
+      "grad_norm": 0.3852269233134472,
+      "learning_rate": 1.1557112379297385e-06,
+      "loss": 1.3542,
       "step": 3145
     },
     {
-      "epoch": 0.9030963302752294,
-      "grad_norm": 3.087915467643749,
-      "learning_rate": 5.66728562720772e-07,
-      "loss": 1.1023,
+      "epoch": 0.8623049548316453,
+      "grad_norm": 0.385512160462019,
+      "learning_rate": 1.1335107784820741e-06,
+      "loss": 1.3556,
       "step": 3150
     },
     {
-      "epoch": 0.9045298165137615,
-      "grad_norm": 3.1987341877696034,
-      "learning_rate": 5.502398234283657e-07,
-      "loss": 1.004,
+      "epoch": 0.8636736928551875,
+      "grad_norm": 0.39758307605071247,
+      "learning_rate": 1.1115128032667288e-06,
+      "loss": 1.2992,
       "step": 3155
     },
     {
-      "epoch": 0.9059633027522935,
-      "grad_norm": 2.8182259878196527,
-      "learning_rate": 5.339877180586872e-07,
-      "loss": 1.0388,
+      "epoch": 0.8650424308787298,
+      "grad_norm": 0.3702170793773961,
+      "learning_rate": 1.0897178146520014e-06,
+      "loss": 1.3861,
       "step": 3160
     },
     {
-      "epoch": 0.9073967889908257,
-      "grad_norm": 3.060289512500466,
-      "learning_rate": 5.179726535849649e-07,
-      "loss": 1.0461,
+      "epoch": 0.8664111689022721,
+      "grad_norm": 0.3862887179246134,
+      "learning_rate": 1.0681263103705853e-06,
+      "loss": 1.3317,
       "step": 3165
     },
     {
-      "epoch": 0.9088302752293578,
-      "grad_norm": 3.1698304516587794,
-      "learning_rate": 5.0219503104463e-07,
-      "loss": 1.0247,
+      "epoch": 0.8677799069258144,
+      "grad_norm": 0.3979382501364706,
+      "learning_rate": 1.0467387835081944e-06,
+      "loss": 1.351,
       "step": 3170
     },
     {
-      "epoch": 0.9102637614678899,
-      "grad_norm": 3.130302466725309,
-      "learning_rate": 4.866552455292673e-07,
-      "loss": 1.0121,
+      "epoch": 0.8691486449493567,
+      "grad_norm": 0.3853783562962658,
+      "learning_rate": 1.0255557224923018e-06,
+      "loss": 1.3474,
       "step": 3175
     },
     {
-      "epoch": 0.911697247706422,
-      "grad_norm": 3.42566263140553,
-      "learning_rate": 4.713536861747181e-07,
-      "loss": 1.0241,
+      "epoch": 0.8705173829728989,
+      "grad_norm": 0.3921947417156507,
+      "learning_rate": 1.004577611080998e-06,
+      "loss": 1.3162,
       "step": 3180
     },
     {
-      "epoch": 0.9131307339449541,
-      "grad_norm": 2.9858737991061974,
-      "learning_rate": 4.5629073615134466e-07,
-      "loss": 1.0616,
+      "epoch": 0.8718861209964412,
+      "grad_norm": 0.38229199053031476,
+      "learning_rate": 9.838049283519258e-07,
+      "loss": 1.3265,
       "step": 3185
     },
     {
-      "epoch": 0.9145642201834863,
-      "grad_norm": 3.285076013359202,
-      "learning_rate": 4.414667726544308e-07,
-      "loss": 0.9933,
+      "epoch": 0.8732548590199836,
+      "grad_norm": 0.38057187484097027,
+      "learning_rate": 9.63238148691351e-07,
+      "loss": 1.3087,
       "step": 3190
     },
     {
-      "epoch": 0.9159977064220184,
-      "grad_norm": 3.2338711561101006,
-      "learning_rate": 4.2688216689472984e-07,
-      "loss": 1.0764,
+      "epoch": 0.8746235970435259,
+      "grad_norm": 0.3942222628186198,
+      "learning_rate": 9.42877741783328e-07,
+      "loss": 1.366,
       "step": 3195
     },
     {
-      "epoch": 0.9174311926605505,
-      "grad_norm": 3.1747891589130766,
-      "learning_rate": 4.125372840891817e-07,
-      "loss": 1.0288,
+      "epoch": 0.8759923350670682,
+      "grad_norm": 0.38514993886291565,
+      "learning_rate": 9.227241725989699e-07,
+      "loss": 1.3212,
       "step": 3200
     },
     {
-      "epoch": 0.9188646788990825,
-      "grad_norm": 3.4369303388998604,
-      "learning_rate": 3.984324834517583e-07,
-      "loss": 1.0151,
+      "epoch": 0.8773610730906105,
+      "grad_norm": 0.3836381492014198,
+      "learning_rate": 9.027779013858284e-07,
+      "loss": 1.2787,
       "step": 3205
     },
     {
-      "epoch": 0.9202981651376146,
-      "grad_norm": 2.9486157404109927,
-      "learning_rate": 3.845681181844718e-07,
-      "loss": 1.0339,
+      "epoch": 0.8787298111141527,
+      "grad_norm": 0.38734970197899504,
+      "learning_rate": 8.830393836573947e-07,
+      "loss": 1.3387,
       "step": 3210
     },
     {
-      "epoch": 0.9217316513761468,
-      "grad_norm": 3.026266781773246,
-      "learning_rate": 3.7094453546852706e-07,
-      "loss": 1.0372,
+      "epoch": 0.880098549137695,
+      "grad_norm": 0.3817865137800505,
+      "learning_rate": 8.635090701826799e-07,
+      "loss": 1.3753,
       "step": 3215
     },
     {
-      "epoch": 0.9231651376146789,
-      "grad_norm": 3.6005808302403715,
-      "learning_rate": 3.575620764556331e-07,
-      "loss": 1.0675,
+      "epoch": 0.8814672871612373,
+      "grad_norm": 0.38200788300985267,
+      "learning_rate": 8.441874069759337e-07,
+      "loss": 1.2776,
       "step": 3220
     },
     {
-      "epoch": 0.924598623853211,
-      "grad_norm": 3.065531017189199,
-      "learning_rate": 3.4442107625945577e-07,
-      "loss": 0.9969,
+      "epoch": 0.8828360251847797,
+      "grad_norm": 0.37300363369723033,
+      "learning_rate": 8.250748352864546e-07,
+      "loss": 1.317,
       "step": 3225
     },
     {
-      "epoch": 0.9260321100917431,
-      "grad_norm": 3.101619885450492,
-      "learning_rate": 3.3152186394722506e-07,
-      "loss": 1.004,
+      "epoch": 0.884204763208322,
+      "grad_norm": 0.39737199946658025,
+      "learning_rate": 8.061717915885103e-07,
+      "loss": 1.3048,
       "step": 3230
     },
     {
-      "epoch": 0.9274655963302753,
-      "grad_norm": 3.069142098455305,
-      "learning_rate": 3.188647625315011e-07,
-      "loss": 1.0277,
+      "epoch": 0.8855735012318642,
+      "grad_norm": 0.3771135467998678,
+      "learning_rate": 7.874787075713742e-07,
+      "loss": 1.2507,
       "step": 3235
     },
     {
-      "epoch": 0.9288990825688074,
-      "grad_norm": 2.981358314660114,
-      "learning_rate": 3.064500889620792e-07,
-      "loss": 1.012,
+      "epoch": 0.8869422392554065,
+      "grad_norm": 0.37723109683472905,
+      "learning_rate": 7.689960101294691e-07,
+      "loss": 1.3081,
       "step": 3240
     },
     {
-      "epoch": 0.9303325688073395,
-      "grad_norm": 3.146209185290999,
-      "learning_rate": 2.9427815411805616e-07,
-      "loss": 1.0409,
+      "epoch": 0.8883109772789488,
+      "grad_norm": 0.3871833961992763,
+      "learning_rate": 7.507241213526073e-07,
+      "loss": 1.3122,
       "step": 3245
     },
     {
-      "epoch": 0.9317660550458715,
-      "grad_norm": 2.9139523212945,
-      "learning_rate": 2.823492628000435e-07,
-      "loss": 1.0108,
+      "epoch": 0.8896797153024911,
+      "grad_norm": 0.3774090621861952,
+      "learning_rate": 7.326634585163617e-07,
+      "loss": 1.3243,
       "step": 3250
     },
     {
-      "epoch": 0.9331995412844036,
-      "grad_norm": 3.2427768967167068,
-      "learning_rate": 2.7066371372253873e-07,
-      "loss": 1.0033,
+      "epoch": 0.8910484533260334,
+      "grad_norm": 0.39061075194420053,
+      "learning_rate": 7.148144340725371e-07,
+      "loss": 1.3123,
       "step": 3255
     },
     {
-      "epoch": 0.9346330275229358,
-      "grad_norm": 3.2929773628001073,
-      "learning_rate": 2.5922179950643833e-07,
-      "loss": 1.053,
+      "epoch": 0.8924171913495756,
+      "grad_norm": 0.3844913895203542,
+      "learning_rate": 6.971774556397415e-07,
+      "loss": 1.3238,
       "step": 3260
     },
     {
-      "epoch": 0.9360665137614679,
-      "grad_norm": 3.0857731418567473,
-      "learning_rate": 2.480238066717178e-07,
-      "loss": 1.0159,
+      "epoch": 0.893785929373118,
+      "grad_norm": 0.3954658749141559,
+      "learning_rate": 6.797529259940827e-07,
+      "loss": 1.3421,
       "step": 3265
     },
     {
-      "epoch": 0.9375,
-      "grad_norm": 2.913155618655732,
-      "learning_rate": 2.370700156302541e-07,
-      "loss": 0.9997,
+      "epoch": 0.8951546673966603,
+      "grad_norm": 0.38848733821104914,
+      "learning_rate": 6.625412430599765e-07,
+      "loss": 1.3485,
       "step": 3270
     },
     {
-      "epoch": 0.9389334862385321,
-      "grad_norm": 2.9236470377703134,
-      "learning_rate": 2.2636070067879933e-07,
-      "loss": 1.0315,
+      "epoch": 0.8965234054202026,
+      "grad_norm": 0.38275187348582623,
+      "learning_rate": 6.455427999010466e-07,
+      "loss": 1.3343,
       "step": 3275
     },
     {
-      "epoch": 0.9403669724770642,
-      "grad_norm": 3.1616081864710113,
-      "learning_rate": 2.1589612999211697e-07,
-      "loss": 1.0573,
+      "epoch": 0.8978921434437449,
+      "grad_norm": 0.3946581641011151,
+      "learning_rate": 6.287579847111569e-07,
+      "loss": 1.3362,
       "step": 3280
     },
     {
-      "epoch": 0.9418004587155964,
-      "grad_norm": 3.2061935710058305,
-      "learning_rate": 2.056765656162685e-07,
-      "loss": 1.0344,
+      "epoch": 0.8992608814672871,
+      "grad_norm": 0.3931302904815353,
+      "learning_rate": 6.121871808055479e-07,
+      "loss": 1.3095,
       "step": 3285
     },
     {
-      "epoch": 0.9432339449541285,
-      "grad_norm": 3.1445007223648087,
-      "learning_rate": 1.9570226346204312e-07,
-      "loss": 1.0502,
+      "epoch": 0.9006296194908294,
+      "grad_norm": 0.39874206377284055,
+      "learning_rate": 5.958307666120733e-07,
+      "loss": 1.2925,
       "step": 3290
     },
     {
-      "epoch": 0.9446674311926605,
-      "grad_norm": 3.0523955190740226,
-      "learning_rate": 1.8597347329855742e-07,
-      "loss": 1.0277,
+      "epoch": 0.9019983575143717,
+      "grad_norm": 0.38856468686205425,
+      "learning_rate": 5.796891156625639e-07,
+      "loss": 1.2878,
       "step": 3295
     },
     {
-      "epoch": 0.9461009174311926,
-      "grad_norm": 3.10853151235872,
-      "learning_rate": 1.764904387469979e-07,
-      "loss": 0.9951,
+      "epoch": 0.903367095537914,
+      "grad_norm": 0.3830311422743617,
+      "learning_rate": 5.637625965843041e-07,
+      "loss": 1.3247,
       "step": 3300
     },
     {
-      "epoch": 0.9475344036697247,
-      "grad_norm": 2.8610073341168327,
-      "learning_rate": 1.672533972745194e-07,
-      "loss": 1.0133,
+      "epoch": 0.9047358335614564,
+      "grad_norm": 0.39286283244773124,
+      "learning_rate": 5.480515730915992e-07,
+      "loss": 1.2902,
       "step": 3305
     },
     {
-      "epoch": 0.9489678899082569,
-      "grad_norm": 2.9659296802362154,
-      "learning_rate": 1.5826258018829866e-07,
-      "loss": 1.0146,
+      "epoch": 0.9061045715849987,
+      "grad_norm": 0.37864973007608094,
+      "learning_rate": 5.325564039774777e-07,
+      "loss": 1.308,
       "step": 3310
     },
     {
-      "epoch": 0.950401376146789,
-      "grad_norm": 3.0101090249896187,
-      "learning_rate": 1.4951821262974563e-07,
-      "loss": 1.0064,
+      "epoch": 0.9074733096085409,
+      "grad_norm": 0.3932055048918823,
+      "learning_rate": 5.172774431054995e-07,
+      "loss": 1.3245,
       "step": 3315
     },
     {
-      "epoch": 0.9518348623853211,
-      "grad_norm": 2.873431144530414,
-      "learning_rate": 1.4102051356886027e-07,
-      "loss": 0.9988,
+      "epoch": 0.9088420476320832,
+      "grad_norm": 0.3861464572512248,
+      "learning_rate": 5.022150394016701e-07,
+      "loss": 1.3345,
       "step": 3320
     },
     {
-      "epoch": 0.9532683486238532,
-      "grad_norm": 2.861730520518137,
-      "learning_rate": 1.3276969579875453e-07,
-      "loss": 1.007,
+      "epoch": 0.9102107856556255,
+      "grad_norm": 0.3849211658260205,
+      "learning_rate": 4.873695368464693e-07,
+      "loss": 1.3522,
       "step": 3325
     },
     {
-      "epoch": 0.9547018348623854,
-      "grad_norm": 3.1539277667342964,
-      "learning_rate": 1.24765965930318e-07,
-      "loss": 1.017,
+      "epoch": 0.9115795236791678,
+      "grad_norm": 0.3850036124623499,
+      "learning_rate": 4.72741274467009e-07,
+      "loss": 1.3205,
       "step": 3330
     },
     {
-      "epoch": 0.9561353211009175,
-      "grad_norm": 3.182662519707628,
-      "learning_rate": 1.1700952438705171e-07,
-      "loss": 1.0008,
+      "epoch": 0.9129482617027102,
+      "grad_norm": 0.3858389082786252,
+      "learning_rate": 4.5833058632927417e-07,
+      "loss": 1.2984,
       "step": 3335
     },
     {
-      "epoch": 0.9575688073394495,
-      "grad_norm": 3.078518777565793,
-      "learning_rate": 1.0950056540004029e-07,
-      "loss": 1.0649,
+      "epoch": 0.9143169997262524,
+      "grad_norm": 0.3974200110164499,
+      "learning_rate": 4.441378015305031e-07,
+      "loss": 1.2907,
       "step": 3340
     },
     {
-      "epoch": 0.9590022935779816,
-      "grad_norm": 2.8736168807619507,
-      "learning_rate": 1.0223927700309667e-07,
-      "loss": 1.0177,
+      "epoch": 0.9156857377497947,
+      "grad_norm": 0.3885026941845055,
+      "learning_rate": 4.3016324419167365e-07,
+      "loss": 1.3571,
       "step": 3345
     },
     {
-      "epoch": 0.9604357798165137,
-      "grad_norm": 3.380358687506272,
-      "learning_rate": 9.522584102804599e-08,
-      "loss": 0.9955,
+      "epoch": 0.917054475773337,
+      "grad_norm": 0.391701334822345,
+      "learning_rate": 4.164072334500935e-07,
+      "loss": 1.2946,
       "step": 3350
     },
     {
-      "epoch": 0.9618692660550459,
-      "grad_norm": 3.209744698526145,
-      "learning_rate": 8.846043310017927e-08,
-      "loss": 1.0051,
+      "epoch": 0.9184232137968793,
+      "grad_norm": 0.371298260058105,
+      "learning_rate": 4.028700834521193e-07,
+      "loss": 1.2734,
       "step": 3355
     },
     {
-      "epoch": 0.963302752293578,
-      "grad_norm": 3.135993664577884,
-      "learning_rate": 8.194322263385013e-08,
-      "loss": 1.0096,
+      "epoch": 0.9197919518204216,
+      "grad_norm": 0.39257123679148415,
+      "learning_rate": 3.8955210334597595e-07,
+      "loss": 1.3792,
       "step": 3360
     },
     {
-      "epoch": 0.9647362385321101,
-      "grad_norm": 2.9702419226658128,
-      "learning_rate": 7.567437282823386e-08,
-      "loss": 0.9581,
+      "epoch": 0.9211606898439638,
+      "grad_norm": 0.37739382436504915,
+      "learning_rate": 3.764535972747052e-07,
+      "loss": 1.3182,
       "step": 3365
     },
     {
-      "epoch": 0.9661697247706422,
-      "grad_norm": 2.9902064692936365,
-      "learning_rate": 6.96540406632451e-08,
-      "loss": 1.0254,
+      "epoch": 0.9225294278675061,
+      "grad_norm": 0.4040354445639437,
+      "learning_rate": 3.6357486436921164e-07,
+      "loss": 1.3149,
       "step": 3370
     },
     {
-      "epoch": 0.9676032110091743,
-      "grad_norm": 3.136047137814621,
-      "learning_rate": 6.388237689559762e-08,
-      "loss": 1.0386,
+      "epoch": 0.9238981658910485,
+      "grad_norm": 0.3885892463097669,
+      "learning_rate": 3.5091619874143446e-07,
+      "loss": 1.3612,
       "step": 3375
     },
     {
-      "epoch": 0.9690366972477065,
-      "grad_norm": 3.209855131586944,
-      "learning_rate": 5.8359526055038476e-08,
-      "loss": 1.0493,
+      "epoch": 0.9252669039145908,
+      "grad_norm": 0.39478973344601664,
+      "learning_rate": 3.3847788947763194e-07,
+      "loss": 1.3338,
       "step": 3380
     },
     {
-      "epoch": 0.9704701834862385,
-      "grad_norm": 3.2935877965995966,
-      "learning_rate": 5.3085626440724235e-08,
-      "loss": 1.0391,
+      "epoch": 0.9266356419381331,
+      "grad_norm": 0.4002415191432085,
+      "learning_rate": 3.2626022063177997e-07,
+      "loss": 1.3854,
       "step": 3385
     },
     {
-      "epoch": 0.9719036697247706,
-      "grad_norm": 3.406107022140581,
-      "learning_rate": 4.8060810117757096e-08,
-      "loss": 1.0244,
+      "epoch": 0.9280043799616753,
+      "grad_norm": 0.3871640152525417,
+      "learning_rate": 3.142634712190795e-07,
+      "loss": 1.2663,
       "step": 3390
     },
     {
-      "epoch": 0.9733371559633027,
-      "grad_norm": 3.3748108058363377,
-      "learning_rate": 4.3285202913881944e-08,
-      "loss": 0.9787,
+      "epoch": 0.9293731179852176,
+      "grad_norm": 0.3844217189319342,
+      "learning_rate": 3.0248791520959387e-07,
+      "loss": 1.3304,
       "step": 3395
     },
     {
-      "epoch": 0.9747706422018348,
-      "grad_norm": 3.0832449198208978,
-      "learning_rate": 3.875892441633e-08,
-      "loss": 0.9951,
+      "epoch": 0.9307418560087599,
+      "grad_norm": 0.40010784379733844,
+      "learning_rate": 2.909338215219859e-07,
+      "loss": 1.3458,
       "step": 3400
     },
     {
-      "epoch": 0.976204128440367,
-      "grad_norm": 3.186455532049428,
-      "learning_rate": 3.4482087968829014e-08,
-      "loss": 1.0315,
+      "epoch": 0.9321105940323022,
+      "grad_norm": 0.38274676057072776,
+      "learning_rate": 2.7960145401737415e-07,
+      "loss": 1.2606,
       "step": 3405
     },
     {
-      "epoch": 0.9776376146788991,
-      "grad_norm": 2.9104798470989883,
-      "learning_rate": 3.045480066876105e-08,
-      "loss": 1.0092,
+      "epoch": 0.9334793320558445,
+      "grad_norm": 0.39275920250834023,
+      "learning_rate": 2.6849107149331756e-07,
+      "loss": 1.2825,
       "step": 3410
     },
     {
-      "epoch": 0.9790711009174312,
-      "grad_norm": 3.037035090176394,
-      "learning_rate": 2.667716336448356e-08,
-      "loss": 0.9587,
+      "epoch": 0.9348480700793868,
+      "grad_norm": 0.3947633255379109,
+      "learning_rate": 2.576029276778924e-07,
+      "loss": 1.3441,
       "step": 3415
     },
     {
-      "epoch": 0.9805045871559633,
-      "grad_norm": 2.9918210850993785,
-      "learning_rate": 2.3149270652803592e-08,
-      "loss": 0.9973,
+      "epoch": 0.9362168081029291,
+      "grad_norm": 0.41407416506322803,
+      "learning_rate": 2.4693727122390597e-07,
+      "loss": 1.371,
       "step": 3420
     },
     {
-      "epoch": 0.9819380733944955,
-      "grad_norm": 3.183820306323636,
-      "learning_rate": 1.9871210876607484e-08,
-      "loss": 1.011,
+      "epoch": 0.9375855461264714,
+      "grad_norm": 0.37625661449174036,
+      "learning_rate": 2.3649434570321984e-07,
+      "loss": 1.2862,
       "step": 3425
     },
     {
-      "epoch": 0.9833715596330275,
-      "grad_norm": 3.214411554963273,
-      "learning_rate": 1.6843066122649297e-08,
-      "loss": 0.9913,
+      "epoch": 0.9389542841500137,
+      "grad_norm": 0.37104631834710733,
+      "learning_rate": 2.2627438960117876e-07,
+      "loss": 1.2833,
       "step": 3430
     },
     {
-      "epoch": 0.9848050458715596,
-      "grad_norm": 3.5216487010248088,
-      "learning_rate": 1.4064912219496907e-08,
-      "loss": 1.0442,
+      "epoch": 0.940323022173556,
+      "grad_norm": 0.3868409112001441,
+      "learning_rate": 2.1627763631117182e-07,
+      "loss": 1.3551,
       "step": 3435
     },
     {
-      "epoch": 0.9862385321100917,
-      "grad_norm": 3.1877265670501598,
-      "learning_rate": 1.1536818735630172e-08,
-      "loss": 1.0318,
+      "epoch": 0.9416917601970983,
+      "grad_norm": 0.4061569745095073,
+      "learning_rate": 2.0650431412930104e-07,
+      "loss": 1.3273,
       "step": 3440
     },
     {
-      "epoch": 0.9876720183486238,
-      "grad_norm": 3.087482942801695,
-      "learning_rate": 9.25884897770013e-09,
-      "loss": 1.0607,
+      "epoch": 0.9430604982206405,
+      "grad_norm": 0.3923199404039606,
+      "learning_rate": 1.969546462491634e-07,
+      "loss": 1.3093,
       "step": 3445
     },
     {
-      "epoch": 0.989105504587156,
-      "grad_norm": 3.059255678756277,
-      "learning_rate": 7.231059988945799e-09,
-      "loss": 1.0148,
+      "epoch": 0.9444292362441828,
+      "grad_norm": 0.40035298460473323,
+      "learning_rate": 1.876288507567592e-07,
+      "loss": 1.2859,
       "step": 3450
     },
     {
-      "epoch": 0.9905389908256881,
-      "grad_norm": 3.239412598410617,
-      "learning_rate": 5.4535025477642224e-09,
-      "loss": 1.0422,
+      "epoch": 0.9457979742677252,
+      "grad_norm": 0.39270377743419116,
+      "learning_rate": 1.785271406255107e-07,
+      "loss": 1.3086,
       "step": 3455
     },
     {
-      "epoch": 0.9919724770642202,
-      "grad_norm": 3.144154368562322,
-      "learning_rate": 3.926221166434818e-09,
-      "loss": 1.0226,
+      "epoch": 0.9471667122912675,
+      "grad_norm": 0.391272234134139,
+      "learning_rate": 1.6964972371139588e-07,
+      "loss": 1.3324,
       "step": 3460
     },
     {
-      "epoch": 0.9934059633027523,
-      "grad_norm": 3.4686686147707877,
-      "learning_rate": 2.6492540900135976e-09,
-      "loss": 1.0359,
+      "epoch": 0.9485354503148098,
+      "grad_norm": 0.3971408419550116,
+      "learning_rate": 1.609968027482012e-07,
+      "loss": 1.3241,
       "step": 3465
     },
     {
-      "epoch": 0.9948394495412844,
-      "grad_norm": 2.9818281052270432,
-      "learning_rate": 1.6226332953661605e-09,
-      "loss": 0.988,
+      "epoch": 0.949904188338352,
+      "grad_norm": 0.397310674294107,
+      "learning_rate": 1.5256857534289626e-07,
+      "loss": 1.344,
       "step": 3470
     },
     {
-      "epoch": 0.9962729357798165,
-      "grad_norm": 2.940056641516156,
-      "learning_rate": 8.46384490373886e-10,
-      "loss": 1.0299,
+      "epoch": 0.9512729263618943,
+      "grad_norm": 0.37737586107823284,
+      "learning_rate": 1.443652339711199e-07,
+      "loss": 1.3227,
       "step": 3475
     },
     {
-      "epoch": 0.9977064220183486,
-      "grad_norm": 3.0373605695085883,
-      "learning_rate": 3.2052711328778297e-10,
-      "loss": 0.9876,
+      "epoch": 0.9526416643854366,
+      "grad_norm": 0.3953334287452581,
+      "learning_rate": 1.3638696597277678e-07,
+      "loss": 1.3323,
       "step": 3480
     },
     {
-      "epoch": 0.9991399082568807,
-      "grad_norm": 3.0988136193131335,
-      "learning_rate": 4.5074332237771134e-11,
-      "loss": 1.0197,
+      "epoch": 0.9540104024089789,
+      "grad_norm": 0.39727863290285664,
+      "learning_rate": 1.2863395354777097e-07,
+      "loss": 1.2965,
       "step": 3485
     },
+    {
+      "epoch": 0.9553791404325213,
+      "grad_norm": 0.3915026545665889,
+      "learning_rate": 1.211063737518392e-07,
+      "loss": 1.2945,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9567478784560635,
+      "grad_norm": 0.3963237143375255,
+      "learning_rate": 1.1380439849250414e-07,
+      "loss": 1.3079,
+      "step": 3495
+    },
+    {
+      "epoch": 0.9581166164796058,
+      "grad_norm": 0.38628531323230814,
+      "learning_rate": 1.0672819452515526e-07,
+      "loss": 1.347,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9594853545031481,
+      "grad_norm": 0.39250735459020125,
+      "learning_rate": 9.987792344923753e-08,
+      "loss": 1.3292,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9608540925266904,
+      "grad_norm": 0.39214485753203543,
+      "learning_rate": 9.32537417045576e-08,
+      "loss": 1.2703,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9622228305502327,
+      "grad_norm": 0.38156832143297204,
+      "learning_rate": 8.685580056771781e-08,
+      "loss": 1.3404,
+      "step": 3515
+    },
+    {
+      "epoch": 0.9635915685737749,
+      "grad_norm": 0.3907257773875925,
+      "learning_rate": 8.0684246148659e-08,
+      "loss": 1.2681,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9649603065973172,
+      "grad_norm": 0.4194193748219591,
+      "learning_rate": 7.473921938731865e-08,
+      "loss": 1.382,
+      "step": 3525
+    },
+    {
+      "epoch": 0.9663290446208596,
+      "grad_norm": 0.3858101635490962,
+      "learning_rate": 6.902085605042019e-08,
+      "loss": 1.3671,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9676977826444019,
+      "grad_norm": 0.3813102160301827,
+      "learning_rate": 6.352928672836767e-08,
+      "loss": 1.3013,
+      "step": 3535
+    },
+    {
+      "epoch": 0.9690665206679442,
+      "grad_norm": 0.38117598807083153,
+      "learning_rate": 5.82646368322648e-08,
+      "loss": 1.3406,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9704352586914865,
+      "grad_norm": 0.3869856138402919,
+      "learning_rate": 5.3227026591049505e-08,
+      "loss": 1.3311,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9718039967150287,
+      "grad_norm": 0.3866858707608126,
+      "learning_rate": 4.841657104875275e-08,
+      "loss": 1.3593,
+      "step": 3550
+    },
+    {
+      "epoch": 0.973172734738571,
+      "grad_norm": 0.3866568363141188,
+      "learning_rate": 4.3833380061865104e-08,
+      "loss": 1.3318,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9745414727621133,
+      "grad_norm": 0.39553603732869785,
+      "learning_rate": 3.947755829683097e-08,
+      "loss": 1.3403,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9759102107856557,
+      "grad_norm": 0.3990993940103786,
+      "learning_rate": 3.5349205227660496e-08,
+      "loss": 1.3812,
+      "step": 3565
+    },
+    {
+      "epoch": 0.977278948809198,
+      "grad_norm": 0.3925614431869845,
+      "learning_rate": 3.144841513365249e-08,
+      "loss": 1.3025,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9786476868327402,
+      "grad_norm": 0.3883373657835178,
+      "learning_rate": 2.7775277097247255e-08,
+      "loss": 1.313,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9800164248562825,
+      "grad_norm": 0.38572501548529475,
+      "learning_rate": 2.4329875001989356e-08,
+      "loss": 1.3058,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9813851628798248,
+      "grad_norm": 0.3838975342711645,
+      "learning_rate": 2.1112287530609122e-08,
+      "loss": 1.3165,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9827539009033671,
+      "grad_norm": 0.3918735272830008,
+      "learning_rate": 1.812258816323187e-08,
+      "loss": 1.3388,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9841226389269094,
+      "grad_norm": 0.38644635827545804,
+      "learning_rate": 1.5360845175695916e-08,
+      "loss": 1.3378,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9854913769504516,
+      "grad_norm": 0.4046028492421605,
+      "learning_rate": 1.2827121637992712e-08,
+      "loss": 1.3104,
+      "step": 3600
+    },
+    {
+      "epoch": 0.986860114973994,
+      "grad_norm": 0.37588594630122674,
+      "learning_rate": 1.0521475412830218e-08,
+      "loss": 1.3345,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9882288529975363,
+      "grad_norm": 0.39556552792879707,
+      "learning_rate": 8.44395915430729e-09,
+      "loss": 1.3184,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9895975910210786,
+      "grad_norm": 0.3745912876489376,
+      "learning_rate": 6.5946203067135395e-09,
+      "loss": 1.3036,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9909663290446209,
+      "grad_norm": 0.39112472329549625,
+      "learning_rate": 4.9735011034457434e-09,
+      "loss": 1.2682,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9923350670681631,
+      "grad_norm": 0.380307545731869,
+      "learning_rate": 3.580638566043071e-09,
+      "loss": 1.2837,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9937038050917054,
+      "grad_norm": 0.3955743424810463,
+      "learning_rate": 2.416064503342197e-09,
+      "loss": 1.3092,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9950725431152477,
+      "grad_norm": 0.3853245568354318,
+      "learning_rate": 1.4798055107489996e-09,
+      "loss": 1.3025,
+      "step": 3635
+    },
+    {
+      "epoch": 0.99644128113879,
+      "grad_norm": 0.39071516823719027,
+      "learning_rate": 7.718829696334862e-10,
+      "loss": 1.3332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9978100191623324,
+      "grad_norm": 0.38346889130554324,
+      "learning_rate": 2.9231304683907667e-10,
+      "loss": 1.3369,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9991787571858747,
+      "grad_norm": 0.4139215786504167,
+      "learning_rate": 4.1106694317338826e-11,
+      "loss": 1.3518,
+      "step": 3650
+    },
     {
       "epoch": 1.0,
-      "eval_loss": 1.0320301055908203,
-      "eval_runtime": 4528.3515,
-      "eval_samples_per_second": 18.469,
-      "eval_steps_per_second": 1.155,
-      "step": 3488
+      "eval_loss": 1.320330262184143,
+      "eval_runtime": 951.5767,
+      "eval_samples_per_second": 92.065,
+      "eval_steps_per_second": 5.755,
+      "step": 3653
     },
     {
       "epoch": 1.0,
-      "step": 3488,
-      "total_flos": 182579059752960.0,
-      "train_loss": 0.025541797006895784,
-      "train_runtime": 4827.4873,
-      "train_samples_per_second": 11.557,
-      "train_steps_per_second": 0.723
+      "step": 3653,
+      "total_flos": 66190143651840.0,
+      "train_loss": 1.3573657579367173,
+      "train_runtime": 8062.8425,
+      "train_samples_per_second": 7.249,
+      "train_steps_per_second": 0.453
     }
   ],
   "logging_steps": 5,
-  "max_steps": 3488,
+  "max_steps": 3653,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 100,
@@ -4922,7 +5153,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 182579059752960.0,
+  "total_flos": 66190143651840.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null