diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -10,6330 +10,6330 @@
   "log_history": [
     {
       "epoch": 0.001105460977227504,
-      "grad_norm": 2.856322765350342,
+      "grad_norm": 3.39754056930542,
       "learning_rate": 5.000000000000001e-07,
-      "loss": 4.5165,
+      "loss": 4.4162,
       "step": 5
     },
     {
       "epoch": 0.002210921954455008,
-      "grad_norm": 2.6451292037963867,
+      "grad_norm": 3.1306285858154297,
       "learning_rate": 1.0000000000000002e-06,
-      "loss": 4.3082,
+      "loss": 4.1884,
       "step": 10
     },
     {
       "epoch": 0.0033163829316825116,
-      "grad_norm": 3.0033833980560303,
+      "grad_norm": 3.5547330379486084,
       "learning_rate": 1.5e-06,
-      "loss": 4.3804,
+      "loss": 4.2556,
       "step": 15
     },
     {
       "epoch": 0.004421843908910016,
-      "grad_norm": 2.5815796852111816,
+      "grad_norm": 3.078263521194458,
       "learning_rate": 2.0000000000000003e-06,
-      "loss": 4.3296,
+      "loss": 4.2079,
       "step": 20
     },
     {
       "epoch": 0.0055273048861375195,
-      "grad_norm": 2.6665921211242676,
+      "grad_norm": 3.162038803100586,
       "learning_rate": 2.5e-06,
-      "loss": 4.274,
+      "loss": 4.1394,
       "step": 25
     },
     {
       "epoch": 0.006632765863365023,
-      "grad_norm": 2.6247975826263428,
+      "grad_norm": 3.037843704223633,
       "learning_rate": 3e-06,
-      "loss": 4.3979,
+      "loss": 4.2716,
       "step": 30
     },
     {
       "epoch": 0.007738226840592527,
-      "grad_norm": 2.822925329208374,
+      "grad_norm": 3.3284013271331787,
       "learning_rate": 3.5000000000000004e-06,
-      "loss": 4.3966,
+      "loss": 4.2629,
       "step": 35
     },
     {
       "epoch": 0.008843687817820032,
-      "grad_norm": 2.9031052589416504,
+      "grad_norm": 3.4477779865264893,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 4.2606,
+      "loss": 4.127,
       "step": 40
     },
     {
       "epoch": 0.009949148795047534,
-      "grad_norm": 2.43031907081604,
+      "grad_norm": 2.864109516143799,
       "learning_rate": 4.5e-06,
-      "loss": 4.3432,
+      "loss": 4.2129,
       "step": 45
     },
     {
       "epoch": 0.011054609772275039,
-      "grad_norm": 2.6758840084075928,
+      "grad_norm": 3.1708180904388428,
       "learning_rate": 5e-06,
-      "loss": 4.3543,
+      "loss": 4.2202,
       "step": 50
     },
     {
       "epoch": 0.012160070749502542,
-      "grad_norm": 2.5539205074310303,
+      "grad_norm": 3.034501075744629,
       "learning_rate": 5.500000000000001e-06,
-      "loss": 4.2527,
+      "loss": 4.1114,
       "step": 55
     },
     {
       "epoch": 0.013265531726730046,
-      "grad_norm": 2.552877187728882,
+      "grad_norm": 3.006770133972168,
       "learning_rate": 6e-06,
-      "loss": 4.2603,
+      "loss": 4.1129,
       "step": 60
     },
     {
       "epoch": 0.014370992703957551,
-      "grad_norm": 2.7067909240722656,
+      "grad_norm": 3.2049851417541504,
       "learning_rate": 6.5000000000000004e-06,
-      "loss": 4.3525,
+      "loss": 4.2145,
       "step": 65
     },
     {
       "epoch": 0.015476453681185054,
-      "grad_norm": 2.633598566055298,
+      "grad_norm": 3.1332285404205322,
       "learning_rate": 7.000000000000001e-06,
-      "loss": 4.2775,
+      "loss": 4.1246,
       "step": 70
     },
     {
       "epoch": 0.016581914658412557,
-      "grad_norm": 2.582083225250244,
+      "grad_norm": 3.0644774436950684,
       "learning_rate": 7.5e-06,
-      "loss": 4.4059,
+      "loss": 4.254,
       "step": 75
     },
     {
       "epoch": 0.017687375635640063,
-      "grad_norm": 2.7137420177459717,
+      "grad_norm": 3.2055327892303467,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 4.2763,
+      "loss": 4.1074,
       "step": 80
     },
     {
       "epoch": 0.018792836612867566,
-      "grad_norm": 2.743177652359009,
+      "grad_norm": 3.2684853076934814,
       "learning_rate": 8.500000000000002e-06,
-      "loss": 4.3027,
+      "loss": 4.1476,
       "step": 85
     },
     {
       "epoch": 0.01989829759009507,
-      "grad_norm": 2.9156761169433594,
+      "grad_norm": 3.4426097869873047,
       "learning_rate": 9e-06,
-      "loss": 4.3686,
+      "loss": 4.1999,
       "step": 90
     },
     {
       "epoch": 0.021003758567322575,
-      "grad_norm": 2.936218738555908,
+      "grad_norm": 3.511049509048462,
       "learning_rate": 9.5e-06,
-      "loss": 4.3308,
+      "loss": 4.171,
       "step": 95
     },
     {
       "epoch": 0.022109219544550078,
-      "grad_norm": 2.5734968185424805,
+      "grad_norm": 3.096892833709717,
       "learning_rate": 1e-05,
-      "loss": 4.144,
+      "loss": 3.9748,
       "step": 100
     },
     {
       "epoch": 0.02321468052177758,
-      "grad_norm": 3.0580193996429443,
+      "grad_norm": 3.6330976486206055,
       "learning_rate": 1.05e-05,
-      "loss": 4.2334,
+      "loss": 4.0609,
       "step": 105
     },
     {
       "epoch": 0.024320141499005084,
-      "grad_norm": 2.8130428791046143,
+      "grad_norm": 3.3570828437805176,
       "learning_rate": 1.1000000000000001e-05,
-      "loss": 4.0793,
+      "loss": 3.8997,
       "step": 110
     },
     {
       "epoch": 0.02542560247623259,
-      "grad_norm": 2.9316952228546143,
+      "grad_norm": 3.4774484634399414,
       "learning_rate": 1.1500000000000002e-05,
-      "loss": 4.2116,
+      "loss": 4.0389,
       "step": 115
     },
     {
       "epoch": 0.026531063453460093,
-      "grad_norm": 2.8418164253234863,
+      "grad_norm": 3.3079240322113037,
       "learning_rate": 1.2e-05,
-      "loss": 4.1086,
+      "loss": 3.9314,
       "step": 120
     },
     {
       "epoch": 0.027636524430687596,
-      "grad_norm": 2.6649138927459717,
+      "grad_norm": 3.205742120742798,
       "learning_rate": 1.25e-05,
-      "loss": 4.1225,
+      "loss": 3.9449,
       "step": 125
     },
     {
       "epoch": 0.028741985407915102,
-      "grad_norm": 2.6316049098968506,
+      "grad_norm": 3.1243762969970703,
       "learning_rate": 1.3000000000000001e-05,
-      "loss": 4.3653,
+      "loss": 4.1831,
       "step": 130
     },
     {
       "epoch": 0.029847446385142605,
-      "grad_norm": 2.9526383876800537,
+      "grad_norm": 3.541386365890503,
       "learning_rate": 1.3500000000000001e-05,
-      "loss": 4.2687,
+      "loss": 4.0843,
       "step": 135
     },
     {
       "epoch": 0.030952907362370108,
-      "grad_norm": 2.761291980743408,
+      "grad_norm": 3.3051278591156006,
       "learning_rate": 1.4000000000000001e-05,
-      "loss": 4.3326,
+      "loss": 4.1357,
       "step": 140
     },
     {
       "epoch": 0.03205836833959761,
-      "grad_norm": 2.6319758892059326,
+      "grad_norm": 3.1360301971435547,
       "learning_rate": 1.45e-05,
-      "loss": 4.3805,
+      "loss": 4.1699,
       "step": 145
     },
     {
       "epoch": 0.033163829316825114,
-      "grad_norm": 2.6983299255371094,
+      "grad_norm": 3.2172937393188477,
       "learning_rate": 1.5e-05,
-      "loss": 4.2701,
+      "loss": 4.0711,
       "step": 150
     },
     {
       "epoch": 0.03426929029405262,
-      "grad_norm": 2.749418020248413,
+      "grad_norm": 3.301384687423706,
       "learning_rate": 1.55e-05,
-      "loss": 4.3861,
+      "loss": 4.1846,
       "step": 155
     },
     {
       "epoch": 0.035374751271280126,
-      "grad_norm": 2.784226179122925,
+      "grad_norm": 3.359994411468506,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 4.1366,
+      "loss": 3.9317,
       "step": 160
     },
     {
       "epoch": 0.03648021224850763,
-      "grad_norm": 2.6632113456726074,
+      "grad_norm": 3.1677932739257812,
       "learning_rate": 1.65e-05,
-      "loss": 4.2244,
+      "loss": 4.0274,
       "step": 165
     },
     {
       "epoch": 0.03758567322573513,
-      "grad_norm": 2.804885149002075,
+      "grad_norm": 3.3650851249694824,
       "learning_rate": 1.7000000000000003e-05,
-      "loss": 4.1976,
+      "loss": 3.9811,
       "step": 170
     },
     {
       "epoch": 0.038691134202962635,
-      "grad_norm": 2.806664228439331,
+      "grad_norm": 3.3627660274505615,
       "learning_rate": 1.75e-05,
-      "loss": 4.0235,
+      "loss": 3.8106,
       "step": 175
     },
     {
       "epoch": 0.03979659518019014,
-      "grad_norm": 2.6123688220977783,
+      "grad_norm": 3.13092041015625,
       "learning_rate": 1.8e-05,
-      "loss": 4.1966,
+      "loss": 3.987,
       "step": 180
     },
     {
       "epoch": 0.04090205615741764,
-      "grad_norm": 2.80129075050354,
+      "grad_norm": 3.3445138931274414,
       "learning_rate": 1.85e-05,
-      "loss": 4.1403,
+      "loss": 3.9161,
       "step": 185
     },
     {
       "epoch": 0.04200751713464515,
-      "grad_norm": 2.7253201007843018,
+      "grad_norm": 3.246854305267334,
       "learning_rate": 1.9e-05,
-      "loss": 4.1317,
+      "loss": 3.909,
       "step": 190
     },
     {
       "epoch": 0.04311297811187265,
-      "grad_norm": 2.852238178253174,
+      "grad_norm": 3.4233202934265137,
       "learning_rate": 1.9500000000000003e-05,
-      "loss": 4.243,
+      "loss": 4.0135,
       "step": 195
     },
     {
       "epoch": 0.044218439089100156,
-      "grad_norm": 2.968660831451416,
+      "grad_norm": 3.4935128688812256,
       "learning_rate": 2e-05,
-      "loss": 4.1083,
+      "loss": 3.8746,
       "step": 200
     },
     {
       "epoch": 0.04532390006632766,
-      "grad_norm": 2.7089550495147705,
+      "grad_norm": 3.2535643577575684,
       "learning_rate": 2.05e-05,
-      "loss": 4.3324,
+      "loss": 4.106,
       "step": 205
     },
     {
       "epoch": 0.04642936104355516,
-      "grad_norm": 2.6991310119628906,
+      "grad_norm": 3.2571001052856445,
       "learning_rate": 2.1e-05,
-      "loss": 4.3528,
+      "loss": 4.1086,
       "step": 210
     },
     {
       "epoch": 0.047534822020782665,
-      "grad_norm": 2.5547115802764893,
+      "grad_norm": 3.044970989227295,
       "learning_rate": 2.15e-05,
-      "loss": 4.0371,
+      "loss": 3.8028,
       "step": 215
     },
     {
       "epoch": 0.04864028299801017,
-      "grad_norm": 2.708559989929199,
+      "grad_norm": 3.2755327224731445,
       "learning_rate": 2.2000000000000003e-05,
-      "loss": 4.1409,
+      "loss": 3.9024,
       "step": 220
     },
     {
       "epoch": 0.04974574397523768,
-      "grad_norm": 2.709721565246582,
+      "grad_norm": 3.273817777633667,
       "learning_rate": 2.25e-05,
-      "loss": 4.0498,
+      "loss": 3.8103,
       "step": 225
     },
     {
       "epoch": 0.05085120495246518,
-      "grad_norm": 2.7421419620513916,
+      "grad_norm": 3.3369789123535156,
       "learning_rate": 2.3000000000000003e-05,
-      "loss": 4.1863,
+      "loss": 3.9451,
       "step": 230
     },
     {
       "epoch": 0.05195666592969268,
-      "grad_norm": 2.776456832885742,
+      "grad_norm": 3.309906244277954,
       "learning_rate": 2.35e-05,
-      "loss": 4.1545,
+      "loss": 3.8952,
       "step": 235
     },
     {
       "epoch": 0.053062126906920186,
-      "grad_norm": 2.8448917865753174,
+      "grad_norm": 3.415665864944458,
       "learning_rate": 2.4e-05,
-      "loss": 4.0625,
+      "loss": 3.8063,
       "step": 240
     },
     {
       "epoch": 0.05416758788414769,
-      "grad_norm": 2.933760404586792,
+      "grad_norm": 3.543761968612671,
       "learning_rate": 2.45e-05,
-      "loss": 4.1578,
+      "loss": 3.9076,
       "step": 245
     },
     {
       "epoch": 0.05527304886137519,
-      "grad_norm": 3.026527166366577,
+      "grad_norm": 3.6634442806243896,
       "learning_rate": 2.5e-05,
-      "loss": 4.1339,
+      "loss": 3.8706,
       "step": 250
     },
     {
       "epoch": 0.056378509838602694,
-      "grad_norm": 2.5931596755981445,
+      "grad_norm": 3.18397855758667,
       "learning_rate": 2.5500000000000003e-05,
-      "loss": 4.0458,
+      "loss": 3.7844,
       "step": 255
     },
     {
       "epoch": 0.057483970815830204,
-      "grad_norm": 2.9681997299194336,
+      "grad_norm": 3.5998668670654297,
       "learning_rate": 2.6000000000000002e-05,
-      "loss": 4.2949,
+      "loss": 4.0349,
       "step": 260
     },
     {
       "epoch": 0.05858943179305771,
-      "grad_norm": 2.822819232940674,
+      "grad_norm": 3.3828001022338867,
       "learning_rate": 2.6500000000000004e-05,
-      "loss": 4.1542,
+      "loss": 3.8898,
       "step": 265
     },
     {
       "epoch": 0.05969489277028521,
-      "grad_norm": 2.794525623321533,
+      "grad_norm": 3.3752715587615967,
       "learning_rate": 2.7000000000000002e-05,
-      "loss": 4.1644,
+      "loss": 3.8881,
       "step": 270
     },
     {
       "epoch": 0.06080035374751271,
-      "grad_norm": 2.6282451152801514,
+      "grad_norm": 3.2270023822784424,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 4.0558,
+      "loss": 3.7873,
       "step": 275
     },
     {
       "epoch": 0.061905814724740216,
-      "grad_norm": 2.87127947807312,
+      "grad_norm": 3.488190174102783,
       "learning_rate": 2.8000000000000003e-05,
-      "loss": 4.2151,
+      "loss": 3.943,
       "step": 280
     },
     {
       "epoch": 0.06301127570196773,
-      "grad_norm": 2.7771425247192383,
+      "grad_norm": 3.401060104370117,
       "learning_rate": 2.8499999999999998e-05,
-      "loss": 4.1902,
+      "loss": 3.9121,
       "step": 285
     },
     {
       "epoch": 0.06411673667919522,
-      "grad_norm": 2.7243714332580566,
+      "grad_norm": 3.3256752490997314,
       "learning_rate": 2.9e-05,
-      "loss": 4.1633,
+      "loss": 3.8748,
       "step": 290
     },
     {
       "epoch": 0.06522219765642273,
-      "grad_norm": 2.733858108520508,
+      "grad_norm": 3.3622047901153564,
       "learning_rate": 2.95e-05,
-      "loss": 4.2343,
+      "loss": 3.9412,
       "step": 295
     },
     {
       "epoch": 0.06632765863365023,
-      "grad_norm": 3.054060935974121,
+      "grad_norm": 3.694291353225708,
       "learning_rate": 3e-05,
-      "loss": 4.0605,
+      "loss": 3.7536,
       "step": 300
     },
     {
       "epoch": 0.06743311961087774,
-      "grad_norm": 2.681039333343506,
+      "grad_norm": 3.3076677322387695,
       "learning_rate": 3.05e-05,
-      "loss": 4.0797,
+      "loss": 3.7886,
       "step": 305
     },
     {
       "epoch": 0.06853858058810525,
-      "grad_norm": 2.594285011291504,
+      "grad_norm": 3.1887643337249756,
       "learning_rate": 3.1e-05,
-      "loss": 4.0443,
+      "loss": 3.7669,
       "step": 310
     },
     {
       "epoch": 0.06964404156533274,
-      "grad_norm": 2.9265353679656982,
+      "grad_norm": 3.589118003845215,
       "learning_rate": 3.15e-05,
-      "loss": 4.2976,
+      "loss": 4.0068,
       "step": 315
     },
     {
       "epoch": 0.07074950254256025,
-      "grad_norm": 3.3384079933166504,
+      "grad_norm": 4.053341388702393,
       "learning_rate": 3.2000000000000005e-05,
-      "loss": 4.2831,
+      "loss": 3.9767,
       "step": 320
     },
     {
       "epoch": 0.07185496351978775,
-      "grad_norm": 2.9113404750823975,
+      "grad_norm": 3.607832431793213,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 4.0696,
+      "loss": 3.7606,
       "step": 325
     },
     {
       "epoch": 0.07296042449701526,
-      "grad_norm": 2.746483087539673,
+      "grad_norm": 3.382657766342163,
       "learning_rate": 3.3e-05,
-      "loss": 3.9145,
+      "loss": 3.6169,
       "step": 330
     },
     {
       "epoch": 0.07406588547424275,
-      "grad_norm": 2.892920970916748,
+      "grad_norm": 3.5504579544067383,
       "learning_rate": 3.35e-05,
-      "loss": 4.0664,
+      "loss": 3.7617,
       "step": 335
     },
     {
       "epoch": 0.07517134645147026,
-      "grad_norm": 3.030963659286499,
+      "grad_norm": 3.64894700050354,
       "learning_rate": 3.4000000000000007e-05,
-      "loss": 4.14,
+      "loss": 3.8215,
       "step": 340
     },
     {
       "epoch": 0.07627680742869777,
-      "grad_norm": 3.1139981746673584,
+      "grad_norm": 3.817647695541382,
       "learning_rate": 3.45e-05,
-      "loss": 4.0361,
+      "loss": 3.732,
       "step": 345
     },
     {
       "epoch": 0.07738226840592527,
-      "grad_norm": 2.646188497543335,
+      "grad_norm": 3.2936594486236572,
       "learning_rate": 3.5e-05,
-      "loss": 4.0998,
+      "loss": 3.7829,
       "step": 350
     },
     {
       "epoch": 0.07848772938315278,
-      "grad_norm": 2.9719629287719727,
+      "grad_norm": 3.607374906539917,
       "learning_rate": 3.55e-05,
-      "loss": 3.9815,
+      "loss": 3.6527,
       "step": 355
     },
     {
       "epoch": 0.07959319036038028,
-      "grad_norm": 2.6908960342407227,
+      "grad_norm": 3.3296027183532715,
       "learning_rate": 3.6e-05,
-      "loss": 4.016,
+      "loss": 3.6855,
       "step": 360
     },
     {
       "epoch": 0.08069865133760779,
-      "grad_norm": 3.2028872966766357,
+      "grad_norm": 3.9575257301330566,
       "learning_rate": 3.65e-05,
-      "loss": 4.0359,
+      "loss": 3.6992,
       "step": 365
     },
     {
       "epoch": 0.08180411231483528,
-      "grad_norm": 2.9519758224487305,
+      "grad_norm": 3.71858811378479,
       "learning_rate": 3.7e-05,
-      "loss": 4.0871,
+      "loss": 3.7658,
       "step": 370
     },
     {
       "epoch": 0.08290957329206279,
-      "grad_norm": 2.844874143600464,
+      "grad_norm": 3.562725067138672,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 4.1217,
+      "loss": 3.7899,
       "step": 375
     },
     {
       "epoch": 0.0840150342692903,
-      "grad_norm": 2.73949933052063,
+      "grad_norm": 3.426848888397217,
       "learning_rate": 3.8e-05,
-      "loss": 4.0189,
+      "loss": 3.6863,
       "step": 380
     },
     {
       "epoch": 0.0851204952465178,
-      "grad_norm": 2.928393840789795,
+      "grad_norm": 3.6758880615234375,
       "learning_rate": 3.85e-05,
-      "loss": 3.9084,
+      "loss": 3.5705,
       "step": 385
     },
     {
       "epoch": 0.0862259562237453,
-      "grad_norm": 2.726449966430664,
+      "grad_norm": 3.432365655899048,
       "learning_rate": 3.9000000000000006e-05,
-      "loss": 4.2775,
+      "loss": 3.9429,
       "step": 390
     },
     {
       "epoch": 0.0873314172009728,
-      "grad_norm": 2.5583412647247314,
+      "grad_norm": 3.2312800884246826,
       "learning_rate": 3.9500000000000005e-05,
-      "loss": 4.0654,
+      "loss": 3.732,
       "step": 395
     },
     {
       "epoch": 0.08843687817820031,
-      "grad_norm": 2.8123371601104736,
+      "grad_norm": 3.485888957977295,
       "learning_rate": 4e-05,
-      "loss": 4.0601,
+      "loss": 3.7195,
       "step": 400
     },
     {
       "epoch": 0.08954233915542781,
-      "grad_norm": 3.2048697471618652,
+      "grad_norm": 3.9309921264648438,
       "learning_rate": 4.05e-05,
-      "loss": 4.1568,
+      "loss": 3.8056,
       "step": 405
     },
     {
       "epoch": 0.09064780013265532,
-      "grad_norm": 2.8617966175079346,
+      "grad_norm": 3.572824716567993,
       "learning_rate": 4.1e-05,
-      "loss": 4.209,
+      "loss": 3.847,
       "step": 410
     },
     {
       "epoch": 0.09175326110988283,
-      "grad_norm": 3.07211971282959,
+      "grad_norm": 3.7535312175750732,
       "learning_rate": 4.15e-05,
-      "loss": 3.9661,
+      "loss": 3.6189,
       "step": 415
     },
     {
       "epoch": 0.09285872208711032,
-      "grad_norm": 2.84535813331604,
+      "grad_norm": 3.558069944381714,
       "learning_rate": 4.2e-05,
-      "loss": 4.0976,
+      "loss": 3.7393,
       "step": 420
     },
     {
       "epoch": 0.09396418306433783,
-      "grad_norm": 2.6337199211120605,
+      "grad_norm": 3.339397668838501,
       "learning_rate": 4.25e-05,
-      "loss": 3.9352,
+      "loss": 3.5769,
       "step": 425
     },
     {
       "epoch": 0.09506964404156533,
-      "grad_norm": 3.1465373039245605,
+      "grad_norm": 3.8922882080078125,
       "learning_rate": 4.3e-05,
-      "loss": 4.0732,
+      "loss": 3.7001,
       "step": 430
     },
     {
       "epoch": 0.09617510501879284,
-      "grad_norm": 2.9059720039367676,
+      "grad_norm": 3.590777635574341,
       "learning_rate": 4.35e-05,
-      "loss": 4.0377,
+      "loss": 3.6682,
       "step": 435
     },
     {
       "epoch": 0.09728056599602034,
-      "grad_norm": 3.147087812423706,
+      "grad_norm": 3.932199001312256,
       "learning_rate": 4.4000000000000006e-05,
-      "loss": 4.0003,
+      "loss": 3.6135,
       "step": 440
     },
     {
       "epoch": 0.09838602697324784,
-      "grad_norm": 2.878849983215332,
+      "grad_norm": 3.55391263961792,
       "learning_rate": 4.4500000000000004e-05,
-      "loss": 4.1393,
+      "loss": 3.754,
       "step": 445
     },
     {
       "epoch": 0.09949148795047535,
-      "grad_norm": 2.9624218940734863,
+      "grad_norm": 3.6753768920898438,
       "learning_rate": 4.5e-05,
-      "loss": 3.9989,
+      "loss": 3.6095,
       "step": 450
     },
     {
       "epoch": 0.10059694892770285,
-      "grad_norm": 3.047313690185547,
+      "grad_norm": 3.8276047706604004,
       "learning_rate": 4.55e-05,
-      "loss": 3.9942,
+      "loss": 3.6001,
       "step": 455
     },
     {
       "epoch": 0.10170240990493036,
-      "grad_norm": 3.069126605987549,
+      "grad_norm": 3.8610167503356934,
       "learning_rate": 4.600000000000001e-05,
-      "loss": 4.0184,
+      "loss": 3.6077,
       "step": 460
     },
     {
       "epoch": 0.10280787088215786,
-      "grad_norm": 3.046513319015503,
+      "grad_norm": 3.790121078491211,
       "learning_rate": 4.6500000000000005e-05,
-      "loss": 4.0441,
+      "loss": 3.6594,
       "step": 465
     },
     {
       "epoch": 0.10391333185938537,
-      "grad_norm": 2.829324722290039,
+      "grad_norm": 3.6155998706817627,
       "learning_rate": 4.7e-05,
-      "loss": 3.9685,
+      "loss": 3.5735,
       "step": 470
     },
     {
       "epoch": 0.10501879283661286,
-      "grad_norm": 3.0912318229675293,
+      "grad_norm": 3.8067572116851807,
       "learning_rate": 4.75e-05,
-      "loss": 3.9195,
+      "loss": 3.4999,
       "step": 475
     },
     {
       "epoch": 0.10612425381384037,
-      "grad_norm": 3.3232522010803223,
+      "grad_norm": 4.257463455200195,
       "learning_rate": 4.8e-05,
-      "loss": 4.092,
+      "loss": 3.6716,
       "step": 480
     },
     {
       "epoch": 0.10722971479106788,
-      "grad_norm": 3.12263822555542,
+      "grad_norm": 3.9768128395080566,
       "learning_rate": 4.85e-05,
-      "loss": 4.1614,
+      "loss": 3.7557,
       "step": 485
     },
     {
       "epoch": 0.10833517576829538,
-      "grad_norm": 3.245594024658203,
+      "grad_norm": 4.108499526977539,
       "learning_rate": 4.9e-05,
-      "loss": 3.9858,
+      "loss": 3.5771,
       "step": 490
     },
     {
       "epoch": 0.10944063674552289,
-      "grad_norm": 3.0725033283233643,
+      "grad_norm": 3.813082218170166,
       "learning_rate": 4.9500000000000004e-05,
-      "loss": 4.0024,
+      "loss": 3.5872,
       "step": 495
     },
     {
       "epoch": 0.11054609772275038,
-      "grad_norm": 3.056286334991455,
+      "grad_norm": 3.9025115966796875,
       "learning_rate": 5e-05,
-      "loss": 3.9464,
+      "loss": 3.5307,
       "step": 500
     },
     {
       "epoch": 0.1116515586999779,
-      "grad_norm": 3.068084478378296,
+      "grad_norm": 3.9042763710021973,
       "learning_rate": 4.993785732040766e-05,
-      "loss": 4.1743,
+      "loss": 3.7688,
       "step": 505
     },
     {
       "epoch": 0.11275701967720539,
-      "grad_norm": 3.189666509628296,
+      "grad_norm": 4.018752098083496,
       "learning_rate": 4.9875714640815315e-05,
-      "loss": 3.9721,
+      "loss": 3.5532,
       "step": 510
     },
     {
       "epoch": 0.1138624806544329,
-      "grad_norm": 3.0129644870758057,
+      "grad_norm": 3.9098243713378906,
       "learning_rate": 4.981357196122297e-05,
-      "loss": 4.0626,
+      "loss": 3.6392,
       "step": 515
     },
     {
       "epoch": 0.11496794163166041,
-      "grad_norm": 2.962771415710449,
+      "grad_norm": 3.8272063732147217,
       "learning_rate": 4.975142928163063e-05,
-      "loss": 4.0752,
+      "loss": 3.6571,
       "step": 520
     },
     {
       "epoch": 0.1160734026088879,
-      "grad_norm": 3.028667688369751,
+      "grad_norm": 3.8089969158172607,
       "learning_rate": 4.968928660203828e-05,
-      "loss": 4.0011,
+      "loss": 3.5747,
       "step": 525
     },
     {
       "epoch": 0.11717886358611541,
-      "grad_norm": 2.98563551902771,
+      "grad_norm": 3.8167307376861572,
       "learning_rate": 4.962714392244594e-05,
-      "loss": 4.0049,
+      "loss": 3.596,
       "step": 530
     },
     {
       "epoch": 0.11828432456334291,
-      "grad_norm": 3.0009968280792236,
+      "grad_norm": 3.860056161880493,
       "learning_rate": 4.9565001242853596e-05,
-      "loss": 3.8586,
+      "loss": 3.4414,
       "step": 535
     },
     {
       "epoch": 0.11938978554057042,
-      "grad_norm": 3.038587808609009,
+      "grad_norm": 3.8806769847869873,
       "learning_rate": 4.950285856326125e-05,
-      "loss": 4.2185,
+      "loss": 3.8098,
       "step": 540
     },
     {
       "epoch": 0.12049524651779792,
-      "grad_norm": 2.9189321994781494,
+      "grad_norm": 3.712733745574951,
       "learning_rate": 4.944071588366891e-05,
-      "loss": 4.0958,
+      "loss": 3.6793,
       "step": 545
     },
     {
       "epoch": 0.12160070749502543,
-      "grad_norm": 2.9720592498779297,
+      "grad_norm": 3.6758031845092773,
       "learning_rate": 4.9378573204076564e-05,
-      "loss": 4.0651,
+      "loss": 3.6438,
       "step": 550
     },
     {
       "epoch": 0.12270616847225294,
-      "grad_norm": 3.094660520553589,
+      "grad_norm": 3.8948066234588623,
       "learning_rate": 4.931643052448422e-05,
-      "loss": 4.0115,
+      "loss": 3.6079,
       "step": 555
     },
     {
       "epoch": 0.12381162944948043,
-      "grad_norm": 3.197223663330078,
+      "grad_norm": 4.064329624176025,
       "learning_rate": 4.925428784489187e-05,
-      "loss": 3.9921,
+      "loss": 3.5837,
       "step": 560
     },
     {
       "epoch": 0.12491709042670794,
-      "grad_norm": 3.033642053604126,
+      "grad_norm": 3.877742052078247,
       "learning_rate": 4.919214516529953e-05,
-      "loss": 3.9471,
+      "loss": 3.5395,
       "step": 565
     },
     {
       "epoch": 0.12602255140393545,
-      "grad_norm": 3.1595492362976074,
+      "grad_norm": 3.997528553009033,
       "learning_rate": 4.913000248570719e-05,
-      "loss": 3.9529,
+      "loss": 3.5445,
       "step": 570
     },
     {
       "epoch": 0.12712801238116295,
-      "grad_norm": 2.948946714401245,
+      "grad_norm": 3.845088481903076,
       "learning_rate": 4.906785980611484e-05,
-      "loss": 4.0634,
+      "loss": 3.6454,
       "step": 575
     },
     {
       "epoch": 0.12823347335839044,
-      "grad_norm": 3.366753339767456,
+      "grad_norm": 4.189253807067871,
       "learning_rate": 4.90057171265225e-05,
-      "loss": 3.8098,
+      "loss": 3.3949,
       "step": 580
     },
     {
       "epoch": 0.12933893433561794,
-      "grad_norm": 3.2447152137756348,
+      "grad_norm": 4.1266021728515625,
       "learning_rate": 4.894357444693015e-05,
-      "loss": 3.9081,
+      "loss": 3.4912,
       "step": 585
     },
     {
       "epoch": 0.13044439531284546,
-      "grad_norm": 3.0394585132598877,
+      "grad_norm": 3.861191749572754,
       "learning_rate": 4.888143176733781e-05,
-      "loss": 3.8164,
+      "loss": 3.4098,
       "step": 590
     },
     {
       "epoch": 0.13154985629007296,
-      "grad_norm": 2.983616828918457,
+      "grad_norm": 3.8402774333953857,
       "learning_rate": 4.881928908774547e-05,
-      "loss": 3.9558,
+      "loss": 3.5374,
       "step": 595
     },
     {
       "epoch": 0.13265531726730045,
-      "grad_norm": 3.1075408458709717,
+      "grad_norm": 4.020076751708984,
       "learning_rate": 4.875714640815312e-05,
-      "loss": 3.9563,
+      "loss": 3.558,
       "step": 600
     },
     {
       "epoch": 0.13376077824452798,
-      "grad_norm": 3.068930149078369,
+      "grad_norm": 4.0010809898376465,
       "learning_rate": 4.8695003728560775e-05,
-      "loss": 3.9616,
+      "loss": 3.535,
       "step": 605
     },
     {
       "epoch": 0.13486623922175547,
-      "grad_norm": 3.2127275466918945,
+      "grad_norm": 4.180361747741699,
       "learning_rate": 4.863286104896843e-05,
-      "loss": 4.0268,
+      "loss": 3.6221,
       "step": 610
     },
     {
       "epoch": 0.13597170019898297,
-      "grad_norm": 3.0738019943237305,
+      "grad_norm": 3.923491954803467,
       "learning_rate": 4.857071836937609e-05,
-      "loss": 4.0659,
+      "loss": 3.671,
       "step": 615
     },
     {
       "epoch": 0.1370771611762105,
-      "grad_norm": 3.2203280925750732,
+      "grad_norm": 4.088878631591797,
       "learning_rate": 4.850857568978375e-05,
-      "loss": 4.052,
+      "loss": 3.6537,
       "step": 620
     },
     {
       "epoch": 0.138182622153438,
-      "grad_norm": 3.0868825912475586,
+      "grad_norm": 3.926711320877075,
       "learning_rate": 4.84464330101914e-05,
-      "loss": 4.1941,
+      "loss": 3.7874,
       "step": 625
     },
     {
       "epoch": 0.13928808313066549,
-      "grad_norm": 2.9370384216308594,
+      "grad_norm": 3.8330297470092773,
       "learning_rate": 4.8384290330599056e-05,
-      "loss": 3.9984,
+      "loss": 3.5945,
       "step": 630
     },
     {
       "epoch": 0.14039354410789298,
-      "grad_norm": 3.230595111846924,
+      "grad_norm": 4.105088710784912,
       "learning_rate": 4.832214765100672e-05,
-      "loss": 4.0905,
+      "loss": 3.6789,
       "step": 635
     },
     {
       "epoch": 0.1414990050851205,
-      "grad_norm": 3.1805593967437744,
+      "grad_norm": 4.167165279388428,
       "learning_rate": 4.826000497141437e-05,
-      "loss": 4.0938,
+      "loss": 3.6994,
       "step": 640
     },
     {
       "epoch": 0.142604466062348,
-      "grad_norm": 2.952800989151001,
+      "grad_norm": 3.771599769592285,
       "learning_rate": 4.8197862291822025e-05,
-      "loss": 3.9694,
+      "loss": 3.5726,
       "step": 645
     },
     {
       "epoch": 0.1437099270395755,
-      "grad_norm": 2.96767520904541,
+      "grad_norm": 3.9164323806762695,
       "learning_rate": 4.813571961222968e-05,
-      "loss": 4.089,
+      "loss": 3.6797,
       "step": 650
     },
     {
       "epoch": 0.14481538801680302,
-      "grad_norm": 3.2061245441436768,
+      "grad_norm": 4.078146934509277,
       "learning_rate": 4.807357693263734e-05,
-      "loss": 3.925,
+      "loss": 3.5247,
       "step": 655
     },
     {
       "epoch": 0.14592084899403052,
-      "grad_norm": 3.4966869354248047,
+      "grad_norm": 4.486090660095215,
       "learning_rate": 4.801143425304499e-05,
-      "loss": 4.0303,
+      "loss": 3.6171,
       "step": 660
     },
     {
       "epoch": 0.147026309971258,
-      "grad_norm": 3.0343263149261475,
+      "grad_norm": 3.9272730350494385,
       "learning_rate": 4.794929157345265e-05,
-      "loss": 4.1302,
+      "loss": 3.7362,
       "step": 665
     },
     {
       "epoch": 0.1481317709484855,
-      "grad_norm": 3.1001501083374023,
+      "grad_norm": 4.070497035980225,
       "learning_rate": 4.7887148893860305e-05,
-      "loss": 4.136,
+      "loss": 3.7417,
       "step": 670
     },
     {
       "epoch": 0.14923723192571303,
-      "grad_norm": 3.0706558227539062,
+      "grad_norm": 3.9726619720458984,
       "learning_rate": 4.782500621426796e-05,
-      "loss": 3.97,
+      "loss": 3.5877,
       "step": 675
     },
     {
       "epoch": 0.15034269290294053,
-      "grad_norm": 3.4160215854644775,
+      "grad_norm": 4.311509609222412,
       "learning_rate": 4.776286353467562e-05,
-      "loss": 4.1283,
+      "loss": 3.7235,
       "step": 680
     },
     {
       "epoch": 0.15144815388016802,
-      "grad_norm": 3.6512129306793213,
+      "grad_norm": 4.604835033416748,
       "learning_rate": 4.7700720855083274e-05,
-      "loss": 4.1619,
+      "loss": 3.7623,
       "step": 685
     },
     {
       "epoch": 0.15255361485739555,
-      "grad_norm": 2.8638243675231934,
+      "grad_norm": 3.8830602169036865,
       "learning_rate": 4.763857817549093e-05,
-      "loss": 4.0563,
+      "loss": 3.681,
       "step": 690
     },
     {
       "epoch": 0.15365907583462304,
-      "grad_norm": 2.87731671333313,
+      "grad_norm": 3.7675650119781494,
       "learning_rate": 4.7576435495898586e-05,
-      "loss": 4.0609,
+      "loss": 3.6754,
       "step": 695
     },
     {
       "epoch": 0.15476453681185054,
-      "grad_norm": 3.2787325382232666,
+      "grad_norm": 4.328310489654541,
       "learning_rate": 4.751429281630624e-05,
-      "loss": 4.031,
+      "loss": 3.6266,
       "step": 700
     },
     {
       "epoch": 0.15586999778907804,
-      "grad_norm": 2.9089596271514893,
+      "grad_norm": 3.7267017364501953,
       "learning_rate": 4.74521501367139e-05,
-      "loss": 3.8997,
+      "loss": 3.5039,
       "step": 705
     },
     {
       "epoch": 0.15697545876630556,
-      "grad_norm": 3.02470326423645,
+      "grad_norm": 3.9472036361694336,
       "learning_rate": 4.7390007457121555e-05,
-      "loss": 4.1458,
+      "loss": 3.7627,
       "step": 710
     },
     {
       "epoch": 0.15808091974353305,
-      "grad_norm": 3.1005873680114746,
+      "grad_norm": 3.9889473915100098,
       "learning_rate": 4.7327864777529204e-05,
-      "loss": 3.8473,
+      "loss": 3.4427,
       "step": 715
     },
     {
       "epoch": 0.15918638072076055,
-      "grad_norm": 3.2032277584075928,
+      "grad_norm": 4.245467185974121,
       "learning_rate": 4.726572209793687e-05,
-      "loss": 4.0817,
+      "loss": 3.683,
       "step": 720
     },
     {
       "epoch": 0.16029184169798807,
-      "grad_norm": 3.1510956287384033,
+      "grad_norm": 4.044693470001221,
       "learning_rate": 4.720357941834452e-05,
-      "loss": 4.0441,
+      "loss": 3.6367,
       "step": 725
     },
     {
       "epoch": 0.16139730267521557,
-      "grad_norm": 3.088815689086914,
+      "grad_norm": 3.9949588775634766,
       "learning_rate": 4.714143673875217e-05,
-      "loss": 3.8953,
+      "loss": 3.512,
       "step": 730
     },
     {
       "epoch": 0.16250276365244307,
-      "grad_norm": 3.099492073059082,
+      "grad_norm": 4.007084369659424,
       "learning_rate": 4.7079294059159836e-05,
-      "loss": 3.8765,
+      "loss": 3.4813,
       "step": 735
     },
     {
       "epoch": 0.16360822462967056,
-      "grad_norm": 2.95200252532959,
+      "grad_norm": 3.873394012451172,
       "learning_rate": 4.7017151379567485e-05,
-      "loss": 4.0126,
+      "loss": 3.6249,
       "step": 740
     },
     {
       "epoch": 0.16471368560689809,
-      "grad_norm": 3.2879955768585205,
+      "grad_norm": 4.209449768066406,
       "learning_rate": 4.695500869997515e-05,
-      "loss": 4.0581,
+      "loss": 3.6696,
       "step": 745
     },
     {
       "epoch": 0.16581914658412558,
-      "grad_norm": 3.344324827194214,
+      "grad_norm": 4.280659198760986,
       "learning_rate": 4.6892866020382804e-05,
-      "loss": 4.056,
+      "loss": 3.6461,
       "step": 750
     },
     {
       "epoch": 0.16692460756135308,
-      "grad_norm": 3.3089466094970703,
+      "grad_norm": 4.279236793518066,
       "learning_rate": 4.6830723340790454e-05,
-      "loss": 3.9941,
+      "loss": 3.5907,
       "step": 755
     },
     {
       "epoch": 0.1680300685385806,
-      "grad_norm": 3.3503427505493164,
+      "grad_norm": 4.385336875915527,
       "learning_rate": 4.6768580661198117e-05,
-      "loss": 3.9987,
+      "loss": 3.5952,
       "step": 760
     },
     {
       "epoch": 0.1691355295158081,
-      "grad_norm": 3.3430700302124023,
+      "grad_norm": 4.472828388214111,
       "learning_rate": 4.670643798160577e-05,
-      "loss": 3.8631,
+      "loss": 3.4825,
       "step": 765
     },
     {
       "epoch": 0.1702409904930356,
-      "grad_norm": 3.0984108448028564,
+      "grad_norm": 3.9934606552124023,
       "learning_rate": 4.664429530201342e-05,
-      "loss": 4.0144,
+      "loss": 3.6254,
       "step": 770
     },
     {
       "epoch": 0.1713464514702631,
-      "grad_norm": 3.1141326427459717,
+      "grad_norm": 4.048763275146484,
       "learning_rate": 4.6582152622421085e-05,
-      "loss": 3.9256,
+      "loss": 3.528,
       "step": 775
     },
     {
       "epoch": 0.1724519124474906,
-      "grad_norm": 3.1998496055603027,
+      "grad_norm": 4.2004170417785645,
       "learning_rate": 4.6520009942828734e-05,
-      "loss": 3.9675,
+      "loss": 3.5779,
       "step": 780
     },
     {
       "epoch": 0.1735573734247181,
-      "grad_norm": 3.034891128540039,
+      "grad_norm": 3.957515239715576,
       "learning_rate": 4.645786726323639e-05,
-      "loss": 4.0099,
+      "loss": 3.6286,
       "step": 785
     },
     {
       "epoch": 0.1746628344019456,
-      "grad_norm": 3.2506675720214844,
+      "grad_norm": 4.240476131439209,
       "learning_rate": 4.6395724583644054e-05,
-      "loss": 3.9246,
+      "loss": 3.5195,
       "step": 790
     },
     {
       "epoch": 0.17576829537917313,
-      "grad_norm": 3.485947608947754,
+      "grad_norm": 4.494534492492676,
       "learning_rate": 4.63335819040517e-05,
-      "loss": 3.9919,
+      "loss": 3.5998,
       "step": 795
     },
     {
       "epoch": 0.17687375635640062,
-      "grad_norm": 3.2420520782470703,
+      "grad_norm": 4.247699737548828,
       "learning_rate": 4.627143922445936e-05,
-      "loss": 4.021,
+      "loss": 3.6292,
       "step": 800
     },
     {
       "epoch": 0.17797921733362812,
-      "grad_norm": 2.989863872528076,
+      "grad_norm": 3.8181638717651367,
       "learning_rate": 4.6209296544867015e-05,
-      "loss": 3.9956,
+      "loss": 3.6066,
       "step": 805
     },
     {
       "epoch": 0.17908467831085562,
-      "grad_norm": 2.9505488872528076,
+      "grad_norm": 3.8398866653442383,
       "learning_rate": 4.614715386527467e-05,
-      "loss": 4.1098,
+      "loss": 3.7303,
       "step": 810
     },
     {
       "epoch": 0.18019013928808314,
-      "grad_norm": 3.1943299770355225,
+      "grad_norm": 4.112417221069336,
       "learning_rate": 4.608501118568233e-05,
-      "loss": 3.962,
+      "loss": 3.5656,
       "step": 815
     },
     {
       "epoch": 0.18129560026531064,
-      "grad_norm": 3.1761474609375,
+      "grad_norm": 4.167758464813232,
       "learning_rate": 4.6022868506089984e-05,
-      "loss": 3.8666,
+      "loss": 3.4776,
       "step": 820
     },
     {
       "epoch": 0.18240106124253813,
-      "grad_norm": 3.454538345336914,
+      "grad_norm": 4.43293571472168,
       "learning_rate": 4.596072582649764e-05,
-      "loss": 4.1169,
+      "loss": 3.7155,
       "step": 825
     },
     {
       "epoch": 0.18350652221976566,
-      "grad_norm": 3.3881819248199463,
+      "grad_norm": 4.3459153175354,
       "learning_rate": 4.5898583146905296e-05,
-      "loss": 4.0902,
+      "loss": 3.6927,
       "step": 830
     },
     {
       "epoch": 0.18461198319699315,
-      "grad_norm": 3.0427277088165283,
+      "grad_norm": 3.971358060836792,
       "learning_rate": 4.583644046731295e-05,
-      "loss": 3.9533,
+      "loss": 3.567,
       "step": 835
     },
     {
       "epoch": 0.18571744417422065,
-      "grad_norm": 3.062037944793701,
+      "grad_norm": 4.023475646972656,
       "learning_rate": 4.577429778772061e-05,
-      "loss": 4.0955,
+      "loss": 3.7008,
       "step": 840
     },
     {
       "epoch": 0.18682290515144814,
-      "grad_norm": 3.1821091175079346,
+      "grad_norm": 4.0871052742004395,
       "learning_rate": 4.5712155108128265e-05,
-      "loss": 4.1441,
+      "loss": 3.7617,
       "step": 845
     },
     {
       "epoch": 0.18792836612867567,
-      "grad_norm": 3.1128711700439453,
+      "grad_norm": 4.028955936431885,
       "learning_rate": 4.565001242853592e-05,
-      "loss": 4.1418,
+      "loss": 3.7648,
       "step": 850
     },
     {
       "epoch": 0.18903382710590316,
-      "grad_norm": 3.0755162239074707,
+      "grad_norm": 4.022873401641846,
       "learning_rate": 4.558786974894358e-05,
-      "loss": 4.0246,
+      "loss": 3.6412,
       "step": 855
     },
     {
       "epoch": 0.19013928808313066,
-      "grad_norm": 3.2559144496917725,
+      "grad_norm": 4.154995918273926,
       "learning_rate": 4.552572706935123e-05,
-      "loss": 4.0333,
+      "loss": 3.6437,
       "step": 860
     },
     {
       "epoch": 0.19124474906035818,
-      "grad_norm": 2.929656744003296,
+      "grad_norm": 3.7861580848693848,
       "learning_rate": 4.546358438975889e-05,
-      "loss": 4.102,
+      "loss": 3.7265,
       "step": 865
     },
     {
       "epoch": 0.19235021003758568,
-      "grad_norm": 3.1212410926818848,
+      "grad_norm": 4.051858425140381,
       "learning_rate": 4.5401441710166546e-05,
-      "loss": 3.8648,
+      "loss": 3.4807,
       "step": 870
     },
     {
       "epoch": 0.19345567101481317,
-      "grad_norm": 3.0112760066986084,
+      "grad_norm": 3.948338508605957,
       "learning_rate": 4.53392990305742e-05,
-      "loss": 4.0,
+      "loss": 3.6138,
       "step": 875
     },
     {
       "epoch": 0.19456113199204067,
-      "grad_norm": 3.1704013347625732,
+      "grad_norm": 4.054975986480713,
       "learning_rate": 4.527715635098186e-05,
-      "loss": 4.0259,
+      "loss": 3.6342,
       "step": 880
     },
     {
       "epoch": 0.1956665929692682,
-      "grad_norm": 2.999876022338867,
+      "grad_norm": 3.926056146621704,
       "learning_rate": 4.5215013671389514e-05,
-      "loss": 3.8822,
+      "loss": 3.5105,
       "step": 885
     },
     {
       "epoch": 0.1967720539464957,
-      "grad_norm": 3.1141977310180664,
+      "grad_norm": 4.02896785736084,
       "learning_rate": 4.515287099179717e-05,
-      "loss": 4.1104,
+      "loss": 3.7238,
       "step": 890
     },
     {
       "epoch": 0.19787751492372319,
-      "grad_norm": 3.2327237129211426,
+      "grad_norm": 4.17403507232666,
       "learning_rate": 4.509072831220482e-05,
-      "loss": 3.8755,
+      "loss": 3.4809,
       "step": 895
     },
     {
       "epoch": 0.1989829759009507,
-      "grad_norm": 3.019273519515991,
+      "grad_norm": 3.9228296279907227,
       "learning_rate": 4.502858563261248e-05,
-      "loss": 3.9992,
+      "loss": 3.6422,
       "step": 900
     },
     {
       "epoch": 0.2000884368781782,
-      "grad_norm": 3.203974962234497,
+      "grad_norm": 4.167940139770508,
       "learning_rate": 4.496644295302014e-05,
-      "loss": 3.978,
+      "loss": 3.6114,
       "step": 905
     },
     {
       "epoch": 0.2011938978554057,
-      "grad_norm": 3.0810108184814453,
+      "grad_norm": 4.080869674682617,
       "learning_rate": 4.490430027342779e-05,
-      "loss": 4.0461,
+      "loss": 3.6734,
       "step": 910
     },
     {
       "epoch": 0.2022993588326332,
-      "grad_norm": 3.004460096359253,
+      "grad_norm": 3.9201180934906006,
       "learning_rate": 4.484215759383545e-05,
-      "loss": 3.9562,
+      "loss": 3.5669,
       "step": 915
     },
     {
       "epoch": 0.20340481980986072,
-      "grad_norm": 3.146409034729004,
+      "grad_norm": 4.063509941101074,
       "learning_rate": 4.478001491424311e-05,
-      "loss": 4.0321,
+      "loss": 3.6556,
       "step": 920
     },
     {
       "epoch": 0.20451028078708822,
-      "grad_norm": 3.180551528930664,
+      "grad_norm": 4.066250324249268,
       "learning_rate": 4.471787223465076e-05,
-      "loss": 4.0203,
+      "loss": 3.661,
       "step": 925
     },
     {
       "epoch": 0.2056157417643157,
-      "grad_norm": 3.2521543502807617,
+      "grad_norm": 4.142704010009766,
       "learning_rate": 4.465572955505842e-05,
-      "loss": 3.9592,
+      "loss": 3.583,
       "step": 930
     },
     {
       "epoch": 0.20672120274154324,
-      "grad_norm": 3.3072097301483154,
+      "grad_norm": 4.202151298522949,
       "learning_rate": 4.459358687546607e-05,
-      "loss": 3.9383,
+      "loss": 3.5656,
       "step": 935
     },
     {
       "epoch": 0.20782666371877073,
-      "grad_norm": 3.152592182159424,
+      "grad_norm": 4.072716236114502,
       "learning_rate": 4.4531444195873725e-05,
-      "loss": 3.9695,
+      "loss": 3.6085,
       "step": 940
     },
     {
       "epoch": 0.20893212469599823,
-      "grad_norm": 3.3956856727600098,
+      "grad_norm": 4.414867877960205,
       "learning_rate": 4.446930151628139e-05,
-      "loss": 4.1435,
+      "loss": 3.765,
       "step": 945
     },
     {
       "epoch": 0.21003758567322572,
-      "grad_norm": 3.2591230869293213,
+      "grad_norm": 4.334674835205078,
       "learning_rate": 4.440715883668904e-05,
-      "loss": 3.9847,
+      "loss": 3.6208,
       "step": 950
     },
     {
       "epoch": 0.21114304665045325,
-      "grad_norm": 3.197763204574585,
+      "grad_norm": 4.102890968322754,
       "learning_rate": 4.4345016157096694e-05,
-      "loss": 4.0096,
+      "loss": 3.6259,
       "step": 955
     },
     {
       "epoch": 0.21224850762768074,
-      "grad_norm": 3.1687469482421875,
+      "grad_norm": 4.096250534057617,
       "learning_rate": 4.428287347750435e-05,
-      "loss": 3.9947,
+      "loss": 3.6078,
       "step": 960
     },
     {
       "epoch": 0.21335396860490824,
-      "grad_norm": 3.01877498626709,
+      "grad_norm": 3.8845198154449463,
       "learning_rate": 4.4220730797912006e-05,
-      "loss": 3.9609,
+      "loss": 3.6027,
       "step": 965
     },
     {
       "epoch": 0.21445942958213576,
-      "grad_norm": 3.0294318199157715,
+      "grad_norm": 3.9897491931915283,
       "learning_rate": 4.415858811831967e-05,
-      "loss": 4.1849,
+      "loss": 3.8051,
       "step": 970
     },
     {
       "epoch": 0.21556489055936326,
-      "grad_norm": 3.6619277000427246,
+      "grad_norm": 4.8230791091918945,
       "learning_rate": 4.409644543872732e-05,
-      "loss": 4.0503,
+      "loss": 3.683,
       "step": 975
     },
     {
       "epoch": 0.21667035153659076,
-      "grad_norm": 3.24751353263855,
+      "grad_norm": 4.1993560791015625,
       "learning_rate": 4.4034302759134975e-05,
-      "loss": 4.1227,
+      "loss": 3.7456,
       "step": 980
     },
     {
       "epoch": 0.21777581251381825,
-      "grad_norm": 3.2298481464385986,
+      "grad_norm": 4.264660358428955,
       "learning_rate": 4.397216007954264e-05,
-      "loss": 4.0815,
+      "loss": 3.7117,
       "step": 985
     },
     {
       "epoch": 0.21888127349104577,
-      "grad_norm": 3.2555155754089355,
+      "grad_norm": 4.276547908782959,
       "learning_rate": 4.391001739995029e-05,
-      "loss": 4.1461,
+      "loss": 3.7765,
       "step": 990
     },
     {
       "epoch": 0.21998673446827327,
-      "grad_norm": 3.141761064529419,
+      "grad_norm": 4.0318284034729,
       "learning_rate": 4.384787472035794e-05,
-      "loss": 4.021,
+      "loss": 3.6558,
       "step": 995
     },
     {
       "epoch": 0.22109219544550077,
-      "grad_norm": 3.0659165382385254,
+      "grad_norm": 3.9856860637664795,
       "learning_rate": 4.37857320407656e-05,
-      "loss": 3.8781,
+      "loss": 3.515,
       "step": 1000
     },
     {
       "epoch": 0.2221976564227283,
-      "grad_norm": 3.1628031730651855,
+      "grad_norm": 4.203017711639404,
       "learning_rate": 4.3723589361173255e-05,
-      "loss": 4.0618,
+      "loss": 3.6813,
       "step": 1005
     },
     {
       "epoch": 0.2233031173999558,
-      "grad_norm": 3.143479347229004,
+      "grad_norm": 4.113800525665283,
       "learning_rate": 4.366144668158091e-05,
-      "loss": 4.0251,
+      "loss": 3.6484,
       "step": 1010
     },
     {
       "epoch": 0.22440857837718328,
-      "grad_norm": 3.302840232849121,
+      "grad_norm": 4.346760272979736,
       "learning_rate": 4.359930400198857e-05,
-      "loss": 3.8384,
+      "loss": 3.4778,
       "step": 1015
     },
     {
       "epoch": 0.22551403935441078,
-      "grad_norm": 2.7286899089813232,
+      "grad_norm": 3.6219234466552734,
       "learning_rate": 4.3537161322396224e-05,
-      "loss": 4.0165,
+      "loss": 3.6632,
       "step": 1020
     },
     {
       "epoch": 0.2266195003316383,
-      "grad_norm": 3.0600860118865967,
+      "grad_norm": 4.0094380378723145,
       "learning_rate": 4.347501864280388e-05,
-      "loss": 4.0364,
+      "loss": 3.6682,
       "step": 1025
     },
     {
       "epoch": 0.2277249613088658,
-      "grad_norm": 2.9517204761505127,
+      "grad_norm": 3.9358749389648438,
       "learning_rate": 4.3412875963211536e-05,
-      "loss": 4.0458,
+      "loss": 3.6907,
       "step": 1030
     },
     {
       "epoch": 0.2288304222860933,
-      "grad_norm": 3.2530035972595215,
+      "grad_norm": 4.201631546020508,
       "learning_rate": 4.335073328361919e-05,
-      "loss": 4.1022,
+      "loss": 3.7309,
       "step": 1035
     },
     {
       "epoch": 0.22993588326332082,
-      "grad_norm": 3.277559280395508,
+      "grad_norm": 4.2483696937561035,
       "learning_rate": 4.328859060402685e-05,
-      "loss": 3.9183,
+      "loss": 3.5525,
       "step": 1040
     },
     {
       "epoch": 0.2310413442405483,
-      "grad_norm": 3.286675453186035,
+      "grad_norm": 4.221437454223633,
       "learning_rate": 4.3226447924434505e-05,
-      "loss": 4.1264,
+      "loss": 3.7521,
       "step": 1045
     },
     {
       "epoch": 0.2321468052177758,
-      "grad_norm": 3.010737180709839,
+      "grad_norm": 3.9680721759796143,
       "learning_rate": 4.3164305244842154e-05,
-      "loss": 4.0477,
+      "loss": 3.6752,
       "step": 1050
     },
     {
       "epoch": 0.2332522661950033,
-      "grad_norm": 3.050497055053711,
+      "grad_norm": 3.949784517288208,
       "learning_rate": 4.310216256524982e-05,
-      "loss": 4.0464,
+      "loss": 3.6751,
       "step": 1055
     },
     {
       "epoch": 0.23435772717223083,
-      "grad_norm": 3.201765537261963,
+      "grad_norm": 4.090748310089111,
       "learning_rate": 4.304001988565747e-05,
-      "loss": 4.0519,
+      "loss": 3.696,
       "step": 1060
     },
     {
       "epoch": 0.23546318814945832,
-      "grad_norm": 3.3649299144744873,
+      "grad_norm": 4.347312927246094,
       "learning_rate": 4.297787720606512e-05,
-      "loss": 3.81,
+      "loss": 3.4433,
       "step": 1065
     },
     {
       "epoch": 0.23656864912668582,
-      "grad_norm": 3.5535190105438232,
+      "grad_norm": 4.687427997589111,
       "learning_rate": 4.2915734526472786e-05,
-      "loss": 4.1328,
+      "loss": 3.7585,
       "step": 1070
     },
     {
       "epoch": 0.23767411010391334,
-      "grad_norm": 3.1812844276428223,
+      "grad_norm": 4.179584980010986,
       "learning_rate": 4.285359184688044e-05,
-      "loss": 4.052,
+      "loss": 3.6715,
       "step": 1075
     },
     {
       "epoch": 0.23877957108114084,
-      "grad_norm": 3.303905725479126,
+      "grad_norm": 4.280942916870117,
       "learning_rate": 4.279144916728809e-05,
-      "loss": 3.8988,
+      "loss": 3.5565,
       "step": 1080
     },
     {
       "epoch": 0.23988503205836834,
-      "grad_norm": 3.1050772666931152,
+      "grad_norm": 4.036306381225586,
       "learning_rate": 4.2729306487695754e-05,
-      "loss": 3.9013,
+      "loss": 3.5392,
       "step": 1085
     },
     {
       "epoch": 0.24099049303559583,
-      "grad_norm": 3.2585289478302,
+      "grad_norm": 4.2910332679748535,
       "learning_rate": 4.2667163808103404e-05,
-      "loss": 4.1435,
+      "loss": 3.7715,
       "step": 1090
     },
     {
       "epoch": 0.24209595401282336,
-      "grad_norm": 3.3238561153411865,
+      "grad_norm": 4.373847484588623,
       "learning_rate": 4.2605021128511067e-05,
-      "loss": 3.9212,
+      "loss": 3.5579,
       "step": 1095
     },
     {
       "epoch": 0.24320141499005085,
-      "grad_norm": 3.151242971420288,
+      "grad_norm": 4.097213268280029,
       "learning_rate": 4.254287844891872e-05,
-      "loss": 4.0018,
+      "loss": 3.6472,
       "step": 1100
     },
     {
       "epoch": 0.24430687596727835,
-      "grad_norm": 2.9132590293884277,
+      "grad_norm": 3.8386526107788086,
       "learning_rate": 4.248073576932637e-05,
-      "loss": 3.9902,
+      "loss": 3.6468,
       "step": 1105
     },
     {
       "epoch": 0.24541233694450587,
-      "grad_norm": 3.318678140640259,
+      "grad_norm": 4.21539831161499,
       "learning_rate": 4.2418593089734035e-05,
-      "loss": 3.9546,
+      "loss": 3.5911,
       "step": 1110
     },
     {
       "epoch": 0.24651779792173337,
-      "grad_norm": 3.3934099674224854,
+      "grad_norm": 4.4001312255859375,
       "learning_rate": 4.2356450410141684e-05,
-      "loss": 3.9115,
+      "loss": 3.5429,
       "step": 1115
     },
     {
       "epoch": 0.24762325889896086,
-      "grad_norm": 3.0218331813812256,
+      "grad_norm": 3.9963278770446777,
       "learning_rate": 4.229430773054934e-05,
-      "loss": 4.0945,
+      "loss": 3.7329,
       "step": 1120
     },
     {
       "epoch": 0.24872871987618836,
-      "grad_norm": 3.152254581451416,
+      "grad_norm": 4.068180084228516,
       "learning_rate": 4.2232165050957004e-05,
-      "loss": 4.1443,
+      "loss": 3.7902,
       "step": 1125
     },
     {
       "epoch": 0.24983418085341588,
-      "grad_norm": 3.2911226749420166,
+      "grad_norm": 4.273662567138672,
       "learning_rate": 4.217002237136465e-05,
-      "loss": 4.0634,
+      "loss": 3.6911,
       "step": 1130
     },
     {
       "epoch": 0.2509396418306434,
-      "grad_norm": 3.0462334156036377,
+      "grad_norm": 3.9411628246307373,
       "learning_rate": 4.210787969177231e-05,
-      "loss": 4.0296,
+      "loss": 3.6796,
       "step": 1135
     },
     {
       "epoch": 0.2520451028078709,
-      "grad_norm": 3.0708699226379395,
+      "grad_norm": 3.9177472591400146,
       "learning_rate": 4.204573701217997e-05,
-      "loss": 4.1341,
+      "loss": 3.7818,
       "step": 1140
     },
     {
       "epoch": 0.25315056378509837,
-      "grad_norm": 3.381535053253174,
+      "grad_norm": 4.356219291687012,
       "learning_rate": 4.198359433258762e-05,
-      "loss": 3.9333,
+      "loss": 3.5686,
       "step": 1145
     },
     {
       "epoch": 0.2542560247623259,
-      "grad_norm": 3.021491050720215,
+      "grad_norm": 3.958259105682373,
       "learning_rate": 4.192145165299528e-05,
-      "loss": 4.0599,
+      "loss": 3.7071,
       "step": 1150
     },
     {
       "epoch": 0.2553614857395534,
-      "grad_norm": 3.339264154434204,
+      "grad_norm": 4.27962589263916,
       "learning_rate": 4.1859308973402934e-05,
-      "loss": 4.0867,
+      "loss": 3.7341,
       "step": 1155
     },
     {
       "epoch": 0.2564669467167809,
-      "grad_norm": 2.9898245334625244,
+      "grad_norm": 4.052038669586182,
       "learning_rate": 4.179716629381059e-05,
-      "loss": 4.0395,
+      "loss": 3.6849,
       "step": 1160
     },
     {
       "epoch": 0.2575724076940084,
-      "grad_norm": 3.3147876262664795,
+      "grad_norm": 4.264370441436768,
       "learning_rate": 4.1735023614218246e-05,
-      "loss": 3.9406,
+      "loss": 3.5877,
       "step": 1165
     },
     {
       "epoch": 0.2586778686712359,
-      "grad_norm": 3.3725435733795166,
+      "grad_norm": 4.356472969055176,
       "learning_rate": 4.16728809346259e-05,
-      "loss": 3.9498,
+      "loss": 3.5851,
       "step": 1170
     },
     {
       "epoch": 0.2597833296484634,
-      "grad_norm": 3.2875232696533203,
+      "grad_norm": 4.2516045570373535,
       "learning_rate": 4.161073825503356e-05,
-      "loss": 3.9765,
+      "loss": 3.6137,
       "step": 1175
     },
     {
       "epoch": 0.2608887906256909,
-      "grad_norm": 3.117985248565674,
+      "grad_norm": 4.144331932067871,
       "learning_rate": 4.1548595575441215e-05,
-      "loss": 4.2161,
+      "loss": 3.8551,
       "step": 1180
     },
     {
       "epoch": 0.2619942516029184,
-      "grad_norm": 3.326371669769287,
+      "grad_norm": 4.3178324699401855,
       "learning_rate": 4.148645289584887e-05,
-      "loss": 3.8891,
+      "loss": 3.5357,
       "step": 1185
     },
     {
       "epoch": 0.2630997125801459,
-      "grad_norm": 3.4053702354431152,
+      "grad_norm": 4.498015403747559,
       "learning_rate": 4.142431021625653e-05,
-      "loss": 4.1167,
+      "loss": 3.7317,
       "step": 1190
     },
     {
       "epoch": 0.26420517355737344,
-      "grad_norm": 2.9902451038360596,
+      "grad_norm": 3.982158660888672,
       "learning_rate": 4.136216753666418e-05,
-      "loss": 4.1837,
+      "loss": 3.8332,
       "step": 1195
     },
     {
       "epoch": 0.2653106345346009,
-      "grad_norm": 3.04341721534729,
+      "grad_norm": 3.9977242946624756,
       "learning_rate": 4.130002485707184e-05,
-      "loss": 3.9783,
+      "loss": 3.6242,
       "step": 1200
     },
     {
       "epoch": 0.26641609551182843,
-      "grad_norm": 3.1881587505340576,
+      "grad_norm": 4.109318256378174,
       "learning_rate": 4.123788217747949e-05,
-      "loss": 4.0327,
+      "loss": 3.6697,
       "step": 1205
     },
     {
       "epoch": 0.26752155648905596,
-      "grad_norm": 3.1782286167144775,
+      "grad_norm": 4.06707763671875,
       "learning_rate": 4.117573949788715e-05,
-      "loss": 3.9614,
+      "loss": 3.6063,
       "step": 1210
     },
     {
       "epoch": 0.2686270174662834,
-      "grad_norm": 3.0777156352996826,
+      "grad_norm": 4.068865776062012,
       "learning_rate": 4.111359681829481e-05,
-      "loss": 3.9946,
+      "loss": 3.6385,
       "step": 1215
     },
     {
       "epoch": 0.26973247844351095,
-      "grad_norm": 3.0450563430786133,
+      "grad_norm": 4.006341457366943,
       "learning_rate": 4.1051454138702464e-05,
-      "loss": 4.0267,
+      "loss": 3.6703,
       "step": 1220
     },
     {
       "epoch": 0.27083793942073847,
-      "grad_norm": 3.516542673110962,
+      "grad_norm": 4.60530424118042,
       "learning_rate": 4.098931145911012e-05,
-      "loss": 4.0077,
+      "loss": 3.6532,
       "step": 1225
     },
     {
       "epoch": 0.27194340039796594,
-      "grad_norm": 3.6443097591400146,
+      "grad_norm": 4.59467077255249,
       "learning_rate": 4.0927168779517776e-05,
-      "loss": 3.9799,
+      "loss": 3.6235,
       "step": 1230
     },
     {
       "epoch": 0.27304886137519346,
-      "grad_norm": 3.004601240158081,
+      "grad_norm": 3.8726587295532227,
       "learning_rate": 4.086502609992543e-05,
-      "loss": 3.997,
+      "loss": 3.6436,
       "step": 1235
     },
     {
       "epoch": 0.274154322352421,
-      "grad_norm": 2.9626457691192627,
+      "grad_norm": 3.865142822265625,
       "learning_rate": 4.080288342033309e-05,
-      "loss": 3.9609,
+      "loss": 3.5955,
       "step": 1240
     },
     {
       "epoch": 0.27525978332964846,
-      "grad_norm": 3.267373561859131,
+      "grad_norm": 4.22714900970459,
       "learning_rate": 4.074074074074074e-05,
-      "loss": 4.0279,
+      "loss": 3.6649,
       "step": 1245
     },
     {
       "epoch": 0.276365244306876,
-      "grad_norm": 3.2012808322906494,
+      "grad_norm": 4.154287338256836,
       "learning_rate": 4.06785980611484e-05,
-      "loss": 4.0551,
+      "loss": 3.7048,
       "step": 1250
     },
     {
       "epoch": 0.27747070528410345,
-      "grad_norm": 3.1443517208099365,
+      "grad_norm": 4.046567916870117,
       "learning_rate": 4.061645538155606e-05,
-      "loss": 3.9241,
+      "loss": 3.5729,
       "step": 1255
     },
     {
       "epoch": 0.27857616626133097,
-      "grad_norm": 3.201756238937378,
+      "grad_norm": 4.118610382080078,
       "learning_rate": 4.055431270196371e-05,
-      "loss": 4.0168,
+      "loss": 3.6564,
       "step": 1260
     },
     {
       "epoch": 0.2796816272385585,
-      "grad_norm": 3.381840229034424,
+      "grad_norm": 4.433300495147705,
       "learning_rate": 4.049217002237137e-05,
-      "loss": 4.0506,
+      "loss": 3.7124,
       "step": 1265
     },
     {
       "epoch": 0.28078708821578596,
-      "grad_norm": 3.3655803203582764,
+      "grad_norm": 4.387540340423584,
       "learning_rate": 4.043002734277902e-05,
-      "loss": 4.0166,
+      "loss": 3.6662,
       "step": 1270
     },
     {
       "epoch": 0.2818925491930135,
-      "grad_norm": 3.1821653842926025,
+      "grad_norm": 4.078158378601074,
       "learning_rate": 4.0367884663186675e-05,
-      "loss": 4.0161,
+      "loss": 3.6688,
       "step": 1275
     },
     {
       "epoch": 0.282998010170241,
-      "grad_norm": 3.2986061573028564,
+      "grad_norm": 4.215695381164551,
       "learning_rate": 4.030574198359434e-05,
-      "loss": 3.8855,
+      "loss": 3.5357,
       "step": 1280
     },
     {
       "epoch": 0.2841034711474685,
-      "grad_norm": 3.3557889461517334,
+      "grad_norm": 4.387265682220459,
       "learning_rate": 4.024359930400199e-05,
-      "loss": 4.0151,
+      "loss": 3.6528,
       "step": 1285
     },
     {
       "epoch": 0.285208932124696,
-      "grad_norm": 3.358522891998291,
+      "grad_norm": 4.288345813751221,
       "learning_rate": 4.0181456624409644e-05,
-      "loss": 3.9199,
+      "loss": 3.5628,
       "step": 1290
     },
     {
       "epoch": 0.2863143931019235,
-      "grad_norm": 3.4547970294952393,
+      "grad_norm": 4.415268898010254,
       "learning_rate": 4.011931394481731e-05,
-      "loss": 4.0687,
+      "loss": 3.7136,
       "step": 1295
     },
     {
       "epoch": 0.287419854079151,
-      "grad_norm": 3.0661280155181885,
+      "grad_norm": 4.021255970001221,
       "learning_rate": 4.0057171265224956e-05,
-      "loss": 4.0077,
+      "loss": 3.6683,
       "step": 1300
     },
     {
       "epoch": 0.2885253150563785,
-      "grad_norm": 3.2720112800598145,
+      "grad_norm": 4.272186756134033,
       "learning_rate": 3.999502858563262e-05,
-      "loss": 3.8876,
+      "loss": 3.5302,
       "step": 1305
     },
     {
       "epoch": 0.28963077603360604,
-      "grad_norm": 3.0981643199920654,
+      "grad_norm": 4.09537935256958,
       "learning_rate": 3.993288590604027e-05,
-      "loss": 3.9393,
+      "loss": 3.5839,
       "step": 1310
     },
     {
       "epoch": 0.2907362370108335,
-      "grad_norm": 3.2599971294403076,
+      "grad_norm": 4.1642165184021,
       "learning_rate": 3.9870743226447925e-05,
-      "loss": 3.8995,
+      "loss": 3.5569,
       "step": 1315
     },
     {
       "epoch": 0.29184169798806103,
-      "grad_norm": 3.6165876388549805,
+      "grad_norm": 4.633291721343994,
       "learning_rate": 3.980860054685559e-05,
-      "loss": 4.0319,
+      "loss": 3.6757,
       "step": 1320
     },
     {
       "epoch": 0.2929471589652885,
-      "grad_norm": 3.432969331741333,
+      "grad_norm": 4.420826435089111,
       "learning_rate": 3.974645786726324e-05,
-      "loss": 4.0085,
+      "loss": 3.6516,
       "step": 1325
     },
     {
       "epoch": 0.294052619942516,
-      "grad_norm": 3.2116641998291016,
+      "grad_norm": 4.221107482910156,
       "learning_rate": 3.968431518767089e-05,
-      "loss": 3.9819,
+      "loss": 3.6325,
       "step": 1330
     },
     {
       "epoch": 0.29515808091974355,
-      "grad_norm": 3.476435661315918,
+      "grad_norm": 4.3750176429748535,
       "learning_rate": 3.962217250807855e-05,
-      "loss": 4.028,
+      "loss": 3.6964,
       "step": 1335
     },
     {
       "epoch": 0.296263541896971,
-      "grad_norm": 3.428138017654419,
+      "grad_norm": 4.3734259605407715,
       "learning_rate": 3.9560029828486205e-05,
-      "loss": 3.9686,
+      "loss": 3.6174,
       "step": 1340
     },
     {
       "epoch": 0.29736900287419854,
-      "grad_norm": 3.2953410148620605,
+      "grad_norm": 4.195644378662109,
       "learning_rate": 3.949788714889386e-05,
-      "loss": 3.9535,
+      "loss": 3.6082,
       "step": 1345
     },
     {
       "epoch": 0.29847446385142606,
-      "grad_norm": 3.800462245941162,
+      "grad_norm": 4.865828514099121,
       "learning_rate": 3.943574446930152e-05,
-      "loss": 4.02,
+      "loss": 3.6746,
       "step": 1350
     },
     {
       "epoch": 0.29957992482865353,
-      "grad_norm": 3.0902063846588135,
+      "grad_norm": 4.039443016052246,
       "learning_rate": 3.9373601789709174e-05,
-      "loss": 4.0621,
+      "loss": 3.7412,
       "step": 1355
     },
     {
       "epoch": 0.30068538580588106,
-      "grad_norm": 3.0530946254730225,
+      "grad_norm": 4.0772905349731445,
       "learning_rate": 3.931145911011683e-05,
-      "loss": 4.0547,
+      "loss": 3.7229,
       "step": 1360
     },
     {
       "epoch": 0.3017908467831086,
-      "grad_norm": 3.3780524730682373,
+      "grad_norm": 4.307041168212891,
       "learning_rate": 3.9249316430524486e-05,
-      "loss": 3.8966,
+      "loss": 3.5379,
       "step": 1365
     },
     {
       "epoch": 0.30289630776033605,
-      "grad_norm": 3.302295207977295,
+      "grad_norm": 4.247729301452637,
       "learning_rate": 3.918717375093214e-05,
-      "loss": 4.1423,
+      "loss": 3.8061,
       "step": 1370
     },
     {
       "epoch": 0.30400176873756357,
-      "grad_norm": 3.452106237411499,
+      "grad_norm": 4.3970184326171875,
       "learning_rate": 3.91250310713398e-05,
-      "loss": 3.95,
+      "loss": 3.5913,
       "step": 1375
     },
     {
       "epoch": 0.3051072297147911,
-      "grad_norm": 3.3365650177001953,
+      "grad_norm": 4.335420608520508,
       "learning_rate": 3.9062888391747455e-05,
-      "loss": 4.0451,
+      "loss": 3.7037,
       "step": 1380
     },
     {
       "epoch": 0.30621269069201856,
-      "grad_norm": 3.3903305530548096,
+      "grad_norm": 4.382383823394775,
       "learning_rate": 3.900074571215511e-05,
-      "loss": 3.8807,
+      "loss": 3.5466,
       "step": 1385
     },
     {
       "epoch": 0.3073181516692461,
-      "grad_norm": 3.6150190830230713,
+      "grad_norm": 4.490661144256592,
       "learning_rate": 3.893860303256277e-05,
-      "loss": 4.0183,
+      "loss": 3.6931,
       "step": 1390
     },
     {
       "epoch": 0.30842361264647356,
-      "grad_norm": 3.298021078109741,
+      "grad_norm": 4.237279891967773,
       "learning_rate": 3.887646035297042e-05,
-      "loss": 4.0159,
+      "loss": 3.6754,
       "step": 1395
     },
     {
       "epoch": 0.3095290736237011,
-      "grad_norm": 3.3884518146514893,
+      "grad_norm": 4.381764888763428,
       "learning_rate": 3.881431767337807e-05,
-      "loss": 4.0274,
+      "loss": 3.6724,
       "step": 1400
     },
     {
       "epoch": 0.3106345346009286,
-      "grad_norm": 3.0882458686828613,
+      "grad_norm": 4.063532829284668,
       "learning_rate": 3.8752174993785736e-05,
-      "loss": 4.0236,
+      "loss": 3.6731,
       "step": 1405
     },
     {
       "epoch": 0.31173999557815607,
-      "grad_norm": 3.4634859561920166,
+      "grad_norm": 4.436192512512207,
       "learning_rate": 3.869003231419339e-05,
-      "loss": 4.106,
+      "loss": 3.7704,
       "step": 1410
     },
     {
       "epoch": 0.3128454565553836,
-      "grad_norm": 3.3966925144195557,
+      "grad_norm": 4.449820518493652,
       "learning_rate": 3.862788963460104e-05,
-      "loss": 4.1579,
+      "loss": 3.8232,
       "step": 1415
     },
     {
       "epoch": 0.3139509175326111,
-      "grad_norm": 3.643110990524292,
+      "grad_norm": 4.727301120758057,
       "learning_rate": 3.8565746955008704e-05,
-      "loss": 3.8821,
+      "loss": 3.5416,
       "step": 1420
     },
     {
       "epoch": 0.3150563785098386,
-      "grad_norm": 3.37382435798645,
+      "grad_norm": 4.49858283996582,
       "learning_rate": 3.8503604275416354e-05,
-      "loss": 4.1456,
+      "loss": 3.8056,
       "step": 1425
     },
     {
       "epoch": 0.3161618394870661,
-      "grad_norm": 3.523825168609619,
+      "grad_norm": 4.507175445556641,
       "learning_rate": 3.8441461595824017e-05,
-      "loss": 4.0356,
+      "loss": 3.6909,
       "step": 1430
     },
     {
       "epoch": 0.31726730046429363,
-      "grad_norm": 3.146383762359619,
+      "grad_norm": 4.115423202514648,
       "learning_rate": 3.837931891623167e-05,
-      "loss": 4.1187,
+      "loss": 3.7772,
       "step": 1435
     },
     {
       "epoch": 0.3183727614415211,
-      "grad_norm": 3.3049044609069824,
+      "grad_norm": 4.257082939147949,
       "learning_rate": 3.831717623663932e-05,
-      "loss": 3.9896,
+      "loss": 3.6642,
       "step": 1440
     },
     {
       "epoch": 0.3194782224187486,
-      "grad_norm": 3.3387224674224854,
+      "grad_norm": 4.415299415588379,
       "learning_rate": 3.8255033557046985e-05,
-      "loss": 4.0838,
+      "loss": 3.7436,
       "step": 1445
     },
     {
       "epoch": 0.32058368339597615,
-      "grad_norm": 3.432584047317505,
+      "grad_norm": 4.489107131958008,
       "learning_rate": 3.819289087745464e-05,
-      "loss": 4.2188,
+      "loss": 3.8687,
       "step": 1450
     },
     {
       "epoch": 0.3216891443732036,
-      "grad_norm": 3.689253568649292,
+      "grad_norm": 4.746647357940674,
       "learning_rate": 3.813074819786229e-05,
-      "loss": 4.0942,
+      "loss": 3.7641,
       "step": 1455
     },
     {
       "epoch": 0.32279460535043114,
-      "grad_norm": 3.4148080348968506,
+      "grad_norm": 4.403326511383057,
       "learning_rate": 3.8068605518269954e-05,
-      "loss": 4.0352,
+      "loss": 3.7055,
       "step": 1460
     },
     {
       "epoch": 0.3239000663276586,
-      "grad_norm": 3.3507676124572754,
+      "grad_norm": 4.340537071228027,
       "learning_rate": 3.80064628386776e-05,
-      "loss": 4.0372,
+      "loss": 3.7061,
       "step": 1465
     },
     {
       "epoch": 0.32500552730488613,
-      "grad_norm": 3.4236788749694824,
+      "grad_norm": 4.400414943695068,
       "learning_rate": 3.794432015908526e-05,
-      "loss": 4.0303,
+      "loss": 3.685,
       "step": 1470
     },
     {
       "epoch": 0.32611098828211366,
-      "grad_norm": 3.2741448879241943,
+      "grad_norm": 4.250907897949219,
       "learning_rate": 3.788217747949292e-05,
-      "loss": 3.9362,
+      "loss": 3.5962,
       "step": 1475
     },
     {
       "epoch": 0.3272164492593411,
-      "grad_norm": 3.177788734436035,
+      "grad_norm": 4.332138538360596,
       "learning_rate": 3.782003479990057e-05,
-      "loss": 4.0183,
+      "loss": 3.6838,
       "step": 1480
     },
     {
       "epoch": 0.32832191023656865,
-      "grad_norm": 3.6237776279449463,
+      "grad_norm": 4.640033721923828,
       "learning_rate": 3.775789212030823e-05,
-      "loss": 4.0285,
+      "loss": 3.6904,
       "step": 1485
     },
     {
       "epoch": 0.32942737121379617,
-      "grad_norm": 3.418241024017334,
+      "grad_norm": 4.409541130065918,
       "learning_rate": 3.7695749440715884e-05,
-      "loss": 4.1458,
+      "loss": 3.817,
       "step": 1490
     },
     {
       "epoch": 0.33053283219102364,
-      "grad_norm": 3.0317554473876953,
+      "grad_norm": 3.9391915798187256,
       "learning_rate": 3.763360676112354e-05,
-      "loss": 3.9586,
+      "loss": 3.6441,
       "step": 1495
     },
     {
       "epoch": 0.33163829316825116,
-      "grad_norm": 3.402616024017334,
+      "grad_norm": 4.386765956878662,
       "learning_rate": 3.7571464081531196e-05,
-      "loss": 4.1311,
+      "loss": 3.809,
       "step": 1500
     },
     {
       "epoch": 0.3327437541454787,
-      "grad_norm": 3.386590003967285,
+      "grad_norm": 4.390594005584717,
       "learning_rate": 3.750932140193885e-05,
-      "loss": 4.189,
+      "loss": 3.8462,
       "step": 1505
     },
     {
       "epoch": 0.33384921512270616,
-      "grad_norm": 3.329336404800415,
+      "grad_norm": 4.307095527648926,
       "learning_rate": 3.744717872234651e-05,
-      "loss": 3.9931,
+      "loss": 3.6736,
       "step": 1510
     },
     {
       "epoch": 0.3349546760999337,
-      "grad_norm": 3.281658411026001,
+      "grad_norm": 4.342155456542969,
       "learning_rate": 3.7385036042754165e-05,
-      "loss": 4.0458,
+      "loss": 3.7116,
       "step": 1515
     },
     {
       "epoch": 0.3360601370771612,
-      "grad_norm": 3.196786880493164,
+      "grad_norm": 4.138860702514648,
       "learning_rate": 3.732289336316182e-05,
-      "loss": 3.9526,
+      "loss": 3.6249,
       "step": 1520
     },
     {
       "epoch": 0.33716559805438867,
-      "grad_norm": 3.386678695678711,
+      "grad_norm": 4.34747314453125,
       "learning_rate": 3.726075068356948e-05,
-      "loss": 4.1347,
+      "loss": 3.8029,
       "step": 1525
     },
     {
       "epoch": 0.3382710590316162,
-      "grad_norm": 2.9931721687316895,
+      "grad_norm": 3.868788480758667,
       "learning_rate": 3.719860800397713e-05,
-      "loss": 3.9369,
+      "loss": 3.6167,
       "step": 1530
     },
     {
       "epoch": 0.33937652000884366,
-      "grad_norm": 3.7105250358581543,
+      "grad_norm": 4.68723201751709,
       "learning_rate": 3.713646532438479e-05,
-      "loss": 3.8733,
+      "loss": 3.5402,
       "step": 1535
     },
     {
       "epoch": 0.3404819809860712,
-      "grad_norm": 3.0669617652893066,
+      "grad_norm": 3.953012704849243,
       "learning_rate": 3.7074322644792446e-05,
-      "loss": 3.8466,
+      "loss": 3.5284,
       "step": 1540
     },
     {
       "epoch": 0.3415874419632987,
-      "grad_norm": 3.449889898300171,
+      "grad_norm": 4.5136027336120605,
       "learning_rate": 3.70121799652001e-05,
-      "loss": 4.0733,
+      "loss": 3.7272,
       "step": 1545
     },
     {
       "epoch": 0.3426929029405262,
-      "grad_norm": 3.4569785594940186,
+      "grad_norm": 4.450701713562012,
       "learning_rate": 3.695003728560776e-05,
-      "loss": 3.9711,
+      "loss": 3.6296,
       "step": 1550
     },
     {
       "epoch": 0.3437983639177537,
-      "grad_norm": 3.4246673583984375,
+      "grad_norm": 4.464328289031982,
       "learning_rate": 3.6887894606015414e-05,
-      "loss": 4.0172,
+      "loss": 3.6706,
       "step": 1555
     },
     {
       "epoch": 0.3449038248949812,
-      "grad_norm": 3.5262482166290283,
+      "grad_norm": 4.6134209632873535,
       "learning_rate": 3.682575192642307e-05,
-      "loss": 4.1475,
+      "loss": 3.8335,
       "step": 1560
     },
     {
       "epoch": 0.3460092858722087,
-      "grad_norm": 3.057406425476074,
+      "grad_norm": 3.961596965789795,
       "learning_rate": 3.6763609246830726e-05,
-      "loss": 4.0023,
+      "loss": 3.6837,
       "step": 1565
     },
     {
       "epoch": 0.3471147468494362,
-      "grad_norm": 3.6714344024658203,
+      "grad_norm": 4.72334623336792,
       "learning_rate": 3.670146656723838e-05,
-      "loss": 3.9847,
+      "loss": 3.6439,
       "step": 1570
     },
     {
       "epoch": 0.34822020782666374,
-      "grad_norm": 3.396587371826172,
+      "grad_norm": 4.351592540740967,
       "learning_rate": 3.663932388764604e-05,
-      "loss": 4.1175,
+      "loss": 3.8006,
       "step": 1575
     },
     {
       "epoch": 0.3493256688038912,
-      "grad_norm": 3.11995530128479,
+      "grad_norm": 4.027406692504883,
       "learning_rate": 3.6577181208053695e-05,
-      "loss": 4.0171,
+      "loss": 3.6907,
       "step": 1580
     },
     {
       "epoch": 0.35043112978111873,
-      "grad_norm": 3.4781930446624756,
+      "grad_norm": 4.581493377685547,
       "learning_rate": 3.651503852846135e-05,
-      "loss": 3.9343,
+      "loss": 3.6095,
       "step": 1585
     },
     {
       "epoch": 0.35153659075834626,
-      "grad_norm": 3.264204263687134,
+      "grad_norm": 4.25584077835083,
       "learning_rate": 3.645289584886901e-05,
-      "loss": 4.1221,
+      "loss": 3.7931,
       "step": 1590
     },
     {
       "epoch": 0.3526420517355737,
-      "grad_norm": 3.2987558841705322,
+      "grad_norm": 4.197746276855469,
       "learning_rate": 3.639075316927666e-05,
-      "loss": 4.1632,
+      "loss": 3.8354,
       "step": 1595
     },
     {
       "epoch": 0.35374751271280125,
-      "grad_norm": 3.6787593364715576,
+      "grad_norm": 4.751527309417725,
       "learning_rate": 3.632861048968432e-05,
-      "loss": 3.9153,
+      "loss": 3.5876,
       "step": 1600
     },
     {
       "epoch": 0.3548529736900287,
-      "grad_norm": 3.2717323303222656,
+      "grad_norm": 4.196956634521484,
       "learning_rate": 3.6266467810091976e-05,
-      "loss": 3.9902,
+      "loss": 3.6597,
       "step": 1605
     },
     {
       "epoch": 0.35595843466725624,
-      "grad_norm": 3.1607632637023926,
+      "grad_norm": 4.098910808563232,
       "learning_rate": 3.6204325130499625e-05,
-      "loss": 4.0374,
+      "loss": 3.7147,
       "step": 1610
     },
     {
       "epoch": 0.35706389564448376,
-      "grad_norm": 3.187629461288452,
+      "grad_norm": 4.055737018585205,
       "learning_rate": 3.614218245090729e-05,
-      "loss": 4.2059,
+      "loss": 3.8782,
       "step": 1615
     },
     {
       "epoch": 0.35816935662171123,
-      "grad_norm": 3.6148953437805176,
+      "grad_norm": 4.633082389831543,
       "learning_rate": 3.608003977131494e-05,
-      "loss": 4.0533,
+      "loss": 3.7296,
       "step": 1620
     },
     {
       "epoch": 0.35927481759893876,
-      "grad_norm": 3.3978331089019775,
+      "grad_norm": 4.442163467407227,
       "learning_rate": 3.6017897091722594e-05,
-      "loss": 4.0545,
+      "loss": 3.7361,
       "step": 1625
     },
     {
       "epoch": 0.3603802785761663,
-      "grad_norm": 3.5654563903808594,
+      "grad_norm": 4.672637939453125,
       "learning_rate": 3.595575441213026e-05,
-      "loss": 4.0747,
+      "loss": 3.7541,
       "step": 1630
     },
     {
       "epoch": 0.36148573955339375,
-      "grad_norm": 3.0887868404388428,
+      "grad_norm": 3.967427968978882,
       "learning_rate": 3.5893611732537906e-05,
-      "loss": 4.0406,
+      "loss": 3.7434,
       "step": 1635
     },
     {
       "epoch": 0.36259120053062127,
-      "grad_norm": 2.8452141284942627,
+      "grad_norm": 3.670264959335327,
       "learning_rate": 3.583146905294556e-05,
-      "loss": 4.0105,
+      "loss": 3.7046,
       "step": 1640
     },
     {
       "epoch": 0.3636966615078488,
-      "grad_norm": 3.3485066890716553,
+      "grad_norm": 4.298936367034912,
       "learning_rate": 3.576932637335322e-05,
-      "loss": 4.1587,
+      "loss": 3.8335,
       "step": 1645
     },
     {
       "epoch": 0.36480212248507626,
-      "grad_norm": 3.476148843765259,
+      "grad_norm": 4.5899810791015625,
       "learning_rate": 3.5707183693760875e-05,
-      "loss": 3.9972,
+      "loss": 3.6835,
       "step": 1650
     },
     {
       "epoch": 0.3659075834623038,
-      "grad_norm": 3.3700621128082275,
+      "grad_norm": 4.329667091369629,
       "learning_rate": 3.564504101416854e-05,
-      "loss": 3.8875,
+      "loss": 3.557,
       "step": 1655
     },
     {
       "epoch": 0.3670130444395313,
-      "grad_norm": 3.48191237449646,
+      "grad_norm": 4.458032131195068,
       "learning_rate": 3.558289833457619e-05,
-      "loss": 4.0436,
+      "loss": 3.6997,
       "step": 1660
     },
     {
       "epoch": 0.3681185054167588,
-      "grad_norm": 2.992255926132202,
+      "grad_norm": 3.9379196166992188,
       "learning_rate": 3.552075565498384e-05,
-      "loss": 4.143,
+      "loss": 3.8344,
       "step": 1665
     },
     {
       "epoch": 0.3692239663939863,
-      "grad_norm": 3.511962413787842,
+      "grad_norm": 4.534510612487793,
       "learning_rate": 3.5458612975391506e-05,
-      "loss": 4.0267,
+      "loss": 3.6883,
       "step": 1670
     },
     {
       "epoch": 0.37032942737121377,
-      "grad_norm": 3.1641499996185303,
+      "grad_norm": 4.058669090270996,
       "learning_rate": 3.5396470295799155e-05,
-      "loss": 3.9213,
+      "loss": 3.6038,
       "step": 1675
     },
     {
       "epoch": 0.3714348883484413,
-      "grad_norm": 3.7594759464263916,
+      "grad_norm": 4.719683647155762,
       "learning_rate": 3.533432761620681e-05,
-      "loss": 4.1522,
+      "loss": 3.8287,
       "step": 1680
     },
     {
       "epoch": 0.3725403493256688,
-      "grad_norm": 3.7265207767486572,
+      "grad_norm": 4.753529071807861,
       "learning_rate": 3.527218493661447e-05,
-      "loss": 3.9366,
+      "loss": 3.612,
       "step": 1685
     },
     {
       "epoch": 0.3736458103028963,
-      "grad_norm": 3.301990270614624,
+      "grad_norm": 4.247401714324951,
       "learning_rate": 3.5210042257022124e-05,
-      "loss": 3.9142,
+      "loss": 3.5975,
       "step": 1690
     },
     {
       "epoch": 0.3747512712801238,
-      "grad_norm": 3.2270445823669434,
+      "grad_norm": 4.180871963500977,
       "learning_rate": 3.514789957742978e-05,
-      "loss": 4.0301,
+      "loss": 3.7198,
       "step": 1695
     },
     {
       "epoch": 0.37585673225735133,
-      "grad_norm": 3.4519598484039307,
+      "grad_norm": 4.501943588256836,
       "learning_rate": 3.5085756897837436e-05,
-      "loss": 3.9566,
+      "loss": 3.6294,
       "step": 1700
     },
     {
       "epoch": 0.3769621932345788,
-      "grad_norm": 3.3497774600982666,
+      "grad_norm": 4.247684478759766,
       "learning_rate": 3.502361421824509e-05,
-      "loss": 3.9327,
+      "loss": 3.6145,
       "step": 1705
     },
     {
       "epoch": 0.3780676542118063,
-      "grad_norm": 3.5343832969665527,
+      "grad_norm": 4.6985626220703125,
       "learning_rate": 3.496147153865275e-05,
-      "loss": 4.0446,
+      "loss": 3.7075,
       "step": 1710
     },
     {
       "epoch": 0.37917311518903385,
-      "grad_norm": 3.369101047515869,
+      "grad_norm": 4.337338447570801,
       "learning_rate": 3.4899328859060405e-05,
-      "loss": 3.9529,
+      "loss": 3.6306,
       "step": 1715
     },
     {
       "epoch": 0.3802785761662613,
-      "grad_norm": 3.0477051734924316,
+      "grad_norm": 3.948371171951294,
       "learning_rate": 3.483718617946806e-05,
-      "loss": 3.8606,
+      "loss": 3.5587,
       "step": 1720
     },
     {
       "epoch": 0.38138403714348884,
-      "grad_norm": 3.516953468322754,
+      "grad_norm": 4.552607536315918,
       "learning_rate": 3.477504349987572e-05,
-      "loss": 3.9936,
+      "loss": 3.6804,
       "step": 1725
     },
     {
       "epoch": 0.38248949812071636,
-      "grad_norm": 3.628263235092163,
+      "grad_norm": 4.62523889541626,
       "learning_rate": 3.471290082028337e-05,
-      "loss": 3.9455,
+      "loss": 3.6199,
       "step": 1730
     },
     {
       "epoch": 0.38359495909794383,
-      "grad_norm": 3.476489305496216,
+      "grad_norm": 4.478798866271973,
       "learning_rate": 3.465075814069103e-05,
-      "loss": 3.972,
+      "loss": 3.6626,
       "step": 1735
     },
     {
       "epoch": 0.38470042007517136,
-      "grad_norm": 3.296743154525757,
+      "grad_norm": 4.268756866455078,
       "learning_rate": 3.4588615461098686e-05,
-      "loss": 4.093,
+      "loss": 3.7864,
       "step": 1740
     },
     {
       "epoch": 0.3858058810523988,
-      "grad_norm": 3.523559331893921,
+      "grad_norm": 4.523069381713867,
       "learning_rate": 3.452647278150634e-05,
-      "loss": 3.9767,
+      "loss": 3.6488,
       "step": 1745
     },
     {
       "epoch": 0.38691134202962635,
-      "grad_norm": 3.2359955310821533,
+      "grad_norm": 4.231037139892578,
       "learning_rate": 3.446433010191399e-05,
-      "loss": 3.9597,
+      "loss": 3.6529,
       "step": 1750
     },
     {
       "epoch": 0.38801680300685387,
-      "grad_norm": 3.318793296813965,
+      "grad_norm": 4.2496185302734375,
       "learning_rate": 3.4402187422321654e-05,
-      "loss": 4.1788,
+      "loss": 3.8646,
       "step": 1755
     },
     {
       "epoch": 0.38912226398408134,
-      "grad_norm": 3.055785655975342,
+      "grad_norm": 3.9765138626098633,
       "learning_rate": 3.434004474272931e-05,
-      "loss": 3.873,
+      "loss": 3.5648,
       "step": 1760
     },
     {
       "epoch": 0.39022772496130886,
-      "grad_norm": 3.787897825241089,
+      "grad_norm": 4.81071138381958,
       "learning_rate": 3.427790206313696e-05,
-      "loss": 4.0092,
+      "loss": 3.7035,
       "step": 1765
     },
     {
       "epoch": 0.3913331859385364,
-      "grad_norm": 3.6127915382385254,
+      "grad_norm": 4.6892409324646,
       "learning_rate": 3.421575938354462e-05,
-      "loss": 3.9295,
+      "loss": 3.6122,
       "step": 1770
     },
     {
       "epoch": 0.39243864691576386,
-      "grad_norm": 3.254620313644409,
+      "grad_norm": 4.230915546417236,
       "learning_rate": 3.415361670395227e-05,
-      "loss": 3.965,
+      "loss": 3.662,
       "step": 1775
     },
     {
       "epoch": 0.3935441078929914,
-      "grad_norm": 3.480854034423828,
+      "grad_norm": 4.445470809936523,
       "learning_rate": 3.4091474024359935e-05,
-      "loss": 4.0151,
+      "loss": 3.7102,
       "step": 1780
     },
     {
       "epoch": 0.3946495688702189,
-      "grad_norm": 3.200242280960083,
+      "grad_norm": 4.06646728515625,
       "learning_rate": 3.402933134476759e-05,
-      "loss": 4.0929,
+      "loss": 3.7789,
       "step": 1785
     },
     {
       "epoch": 0.39575502984744637,
-      "grad_norm": 3.1364223957061768,
+      "grad_norm": 4.134965896606445,
       "learning_rate": 3.396718866517524e-05,
-      "loss": 3.9891,
+      "loss": 3.6949,
       "step": 1790
     },
     {
       "epoch": 0.3968604908246739,
-      "grad_norm": 3.4453999996185303,
+      "grad_norm": 4.332599639892578,
       "learning_rate": 3.3905045985582904e-05,
-      "loss": 4.1357,
+      "loss": 3.8316,
       "step": 1795
     },
     {
       "epoch": 0.3979659518019014,
-      "grad_norm": 3.265876531600952,
+      "grad_norm": 4.083302974700928,
       "learning_rate": 3.384290330599056e-05,
-      "loss": 3.8728,
+      "loss": 3.5698,
       "step": 1800
     },
     {
       "epoch": 0.3990714127791289,
-      "grad_norm": 3.2799103260040283,
+      "grad_norm": 4.3432440757751465,
       "learning_rate": 3.378076062639821e-05,
-      "loss": 4.1506,
+      "loss": 3.8478,
       "step": 1805
     },
     {
       "epoch": 0.4001768737563564,
-      "grad_norm": 3.2966063022613525,
+      "grad_norm": 4.224790573120117,
       "learning_rate": 3.371861794680587e-05,
-      "loss": 3.8413,
+      "loss": 3.5459,
       "step": 1810
     },
     {
       "epoch": 0.4012823347335839,
-      "grad_norm": 3.346560478210449,
+      "grad_norm": 4.279894828796387,
       "learning_rate": 3.365647526721352e-05,
-      "loss": 4.0029,
+      "loss": 3.7049,
       "step": 1815
     },
     {
       "epoch": 0.4023877957108114,
-      "grad_norm": 3.191598892211914,
+      "grad_norm": 4.1142659187316895,
       "learning_rate": 3.359433258762118e-05,
-      "loss": 4.0017,
+      "loss": 3.7022,
       "step": 1820
     },
     {
       "epoch": 0.4034932566880389,
-      "grad_norm": 3.689901113510132,
+      "grad_norm": 4.817364692687988,
       "learning_rate": 3.353218990802884e-05,
-      "loss": 4.1346,
+      "loss": 3.8185,
       "step": 1825
     },
     {
       "epoch": 0.4045987176652664,
-      "grad_norm": 3.4523544311523438,
+      "grad_norm": 4.536213397979736,
       "learning_rate": 3.347004722843649e-05,
-      "loss": 4.0593,
+      "loss": 3.7679,
       "step": 1830
     },
     {
       "epoch": 0.4057041786424939,
-      "grad_norm": 3.3706953525543213,
+      "grad_norm": 4.424204349517822,
       "learning_rate": 3.3407904548844146e-05,
-      "loss": 4.1312,
+      "loss": 3.8351,
       "step": 1835
     },
     {
       "epoch": 0.40680963961972144,
-      "grad_norm": 3.5654544830322266,
+      "grad_norm": 4.520392417907715,
       "learning_rate": 3.33457618692518e-05,
-      "loss": 4.0694,
+      "loss": 3.7669,
       "step": 1840
     },
     {
       "epoch": 0.4079151005969489,
-      "grad_norm": 3.540480136871338,
+      "grad_norm": 4.598485469818115,
       "learning_rate": 3.328361918965946e-05,
-      "loss": 4.123,
+      "loss": 3.8033,
       "step": 1845
     },
     {
       "epoch": 0.40902056157417643,
-      "grad_norm": 3.286994695663452,
+      "grad_norm": 4.315255165100098,
       "learning_rate": 3.3221476510067115e-05,
-      "loss": 4.1365,
+      "loss": 3.8388,
       "step": 1850
     },
     {
       "epoch": 0.41012602255140396,
-      "grad_norm": 3.0457570552825928,
+      "grad_norm": 3.9313549995422363,
       "learning_rate": 3.315933383047477e-05,
-      "loss": 3.9359,
+      "loss": 3.6465,
       "step": 1855
     },
     {
       "epoch": 0.4112314835286314,
-      "grad_norm": 3.2751758098602295,
+      "grad_norm": 4.2039475440979,
       "learning_rate": 3.309719115088243e-05,
-      "loss": 4.0003,
+      "loss": 3.7003,
       "step": 1860
     },
     {
       "epoch": 0.41233694450585895,
-      "grad_norm": 3.345170259475708,
+      "grad_norm": 4.258403301239014,
       "learning_rate": 3.303504847129008e-05,
-      "loss": 3.87,
+      "loss": 3.5857,
       "step": 1865
     },
     {
       "epoch": 0.4134424054830865,
-      "grad_norm": 3.398428440093994,
+      "grad_norm": 4.31795597076416,
       "learning_rate": 3.297290579169774e-05,
-      "loss": 3.9499,
+      "loss": 3.6514,
       "step": 1870
     },
     {
       "epoch": 0.41454786646031394,
-      "grad_norm": 3.3243329524993896,
+      "grad_norm": 4.225702285766602,
       "learning_rate": 3.2910763112105396e-05,
-      "loss": 4.0548,
+      "loss": 3.7454,
       "step": 1875
     },
     {
       "epoch": 0.41565332743754146,
-      "grad_norm": 3.449658155441284,
+      "grad_norm": 4.31908655166626,
       "learning_rate": 3.284862043251305e-05,
-      "loss": 3.8984,
+      "loss": 3.5983,
       "step": 1880
     },
     {
       "epoch": 0.41675878841476893,
-      "grad_norm": 3.741178035736084,
+      "grad_norm": 4.823517322540283,
       "learning_rate": 3.278647775292071e-05,
-      "loss": 4.1575,
+      "loss": 3.8533,
       "step": 1885
     },
     {
       "epoch": 0.41786424939199646,
-      "grad_norm": 3.4483730792999268,
+      "grad_norm": 4.4373273849487305,
       "learning_rate": 3.2724335073328364e-05,
-      "loss": 4.034,
+      "loss": 3.7375,
       "step": 1890
     },
     {
       "epoch": 0.418969710369224,
-      "grad_norm": 3.176455020904541,
+      "grad_norm": 4.050292491912842,
       "learning_rate": 3.266219239373602e-05,
-      "loss": 3.9522,
+      "loss": 3.6612,
       "step": 1895
     },
     {
       "epoch": 0.42007517134645145,
-      "grad_norm": 3.323781967163086,
+      "grad_norm": 4.318612575531006,
       "learning_rate": 3.2600049714143676e-05,
-      "loss": 4.04,
+      "loss": 3.7539,
       "step": 1900
     },
     {
       "epoch": 0.42118063232367897,
-      "grad_norm": 3.125051498413086,
+      "grad_norm": 4.0239362716674805,
       "learning_rate": 3.253790703455133e-05,
-      "loss": 3.9916,
+      "loss": 3.699,
       "step": 1905
     },
     {
       "epoch": 0.4222860933009065,
-      "grad_norm": 3.488311767578125,
+      "grad_norm": 4.380274772644043,
       "learning_rate": 3.247576435495899e-05,
-      "loss": 4.1544,
+      "loss": 3.8428,
       "step": 1910
     },
     {
       "epoch": 0.42339155427813396,
-      "grad_norm": 3.3193490505218506,
+      "grad_norm": 4.239674091339111,
       "learning_rate": 3.2413621675366645e-05,
-      "loss": 3.8267,
+      "loss": 3.523,
       "step": 1915
     },
     {
       "epoch": 0.4244970152553615,
-      "grad_norm": 3.118138313293457,
+      "grad_norm": 4.049400329589844,
       "learning_rate": 3.23514789957743e-05,
-      "loss": 4.0021,
+      "loss": 3.7156,
       "step": 1920
     },
     {
       "epoch": 0.425602476232589,
-      "grad_norm": 3.0843567848205566,
+      "grad_norm": 3.9598779678344727,
       "learning_rate": 3.228933631618196e-05,
-      "loss": 4.0595,
+      "loss": 3.7646,
       "step": 1925
     },
     {
       "epoch": 0.4267079372098165,
-      "grad_norm": 3.249384880065918,
+      "grad_norm": 4.1489667892456055,
       "learning_rate": 3.222719363658961e-05,
-      "loss": 4.0343,
+      "loss": 3.7338,
       "step": 1930
     },
     {
       "epoch": 0.427813398187044,
-      "grad_norm": 3.4635889530181885,
+      "grad_norm": 4.438506603240967,
       "learning_rate": 3.216505095699727e-05,
-      "loss": 4.1018,
+      "loss": 3.8092,
       "step": 1935
     },
     {
       "epoch": 0.4289188591642715,
-      "grad_norm": 3.705624580383301,
+      "grad_norm": 4.727476596832275,
       "learning_rate": 3.2102908277404926e-05,
-      "loss": 4.0875,
+      "loss": 3.7966,
       "step": 1940
     },
     {
       "epoch": 0.430024320141499,
-      "grad_norm": 3.6071228981018066,
+      "grad_norm": 4.53037166595459,
       "learning_rate": 3.2040765597812575e-05,
-      "loss": 4.0507,
+      "loss": 3.7643,
       "step": 1945
     },
     {
       "epoch": 0.4311297811187265,
-      "grad_norm": 3.513573169708252,
+      "grad_norm": 4.463996410369873,
       "learning_rate": 3.197862291822024e-05,
-      "loss": 3.9596,
+      "loss": 3.6561,
       "step": 1950
     },
     {
       "epoch": 0.432235242095954,
-      "grad_norm": 3.4200334548950195,
+      "grad_norm": 4.410264492034912,
       "learning_rate": 3.1916480238627894e-05,
-      "loss": 3.9723,
+      "loss": 3.671,
       "step": 1955
     },
     {
       "epoch": 0.4333407030731815,
-      "grad_norm": 3.472170114517212,
+      "grad_norm": 4.45218563079834,
       "learning_rate": 3.1854337559035544e-05,
-      "loss": 4.0224,
+      "loss": 3.7172,
       "step": 1960
     },
     {
       "epoch": 0.43444616405040903,
-      "grad_norm": 3.499969482421875,
+      "grad_norm": 4.5141072273254395,
       "learning_rate": 3.179219487944321e-05,
-      "loss": 3.935,
+      "loss": 3.6473,
       "step": 1965
     },
     {
       "epoch": 0.4355516250276365,
-      "grad_norm": 3.5393736362457275,
+      "grad_norm": 4.5253753662109375,
       "learning_rate": 3.1730052199850856e-05,
-      "loss": 4.1007,
+      "loss": 3.8088,
       "step": 1970
     },
     {
       "epoch": 0.436657086004864,
-      "grad_norm": 3.557710647583008,
+      "grad_norm": 4.525451183319092,
       "learning_rate": 3.166790952025851e-05,
-      "loss": 4.1932,
+      "loss": 3.8908,
       "step": 1975
     },
     {
       "epoch": 0.43776254698209155,
-      "grad_norm": 3.4602739810943604,
+      "grad_norm": 4.383676528930664,
       "learning_rate": 3.1605766840666175e-05,
-      "loss": 4.0741,
+      "loss": 3.7907,
       "step": 1980
     },
     {
       "epoch": 0.438868007959319,
-      "grad_norm": 3.578395366668701,
+      "grad_norm": 4.54648494720459,
       "learning_rate": 3.1543624161073825e-05,
-      "loss": 4.0468,
+      "loss": 3.751,
       "step": 1985
     },
     {
       "epoch": 0.43997346893654654,
-      "grad_norm": 3.3289973735809326,
+      "grad_norm": 4.296148777008057,
       "learning_rate": 3.148148148148148e-05,
-      "loss": 3.9823,
+      "loss": 3.6972,
       "step": 1990
     },
     {
       "epoch": 0.44107892991377406,
-      "grad_norm": 3.6602888107299805,
+      "grad_norm": 4.61207914352417,
       "learning_rate": 3.141933880188914e-05,
-      "loss": 4.0993,
+      "loss": 3.8076,
       "step": 1995
     },
     {
       "epoch": 0.44218439089100153,
-      "grad_norm": 3.5060999393463135,
+      "grad_norm": 4.51292610168457,
       "learning_rate": 3.135719612229679e-05,
-      "loss": 3.9399,
+      "loss": 3.6368,
       "step": 2000
     },
     {
       "epoch": 0.44328985186822906,
-      "grad_norm": 3.185040235519409,
+      "grad_norm": 4.101064205169678,
       "learning_rate": 3.1295053442704456e-05,
-      "loss": 4.0126,
+      "loss": 3.7129,
       "step": 2005
     },
     {
       "epoch": 0.4443953128454566,
-      "grad_norm": 3.3001205921173096,
+      "grad_norm": 4.213204383850098,
       "learning_rate": 3.1232910763112105e-05,
-      "loss": 4.1115,
+      "loss": 3.8288,
       "step": 2010
     },
     {
       "epoch": 0.44550077382268405,
-      "grad_norm": 3.4892706871032715,
+      "grad_norm": 4.428658485412598,
       "learning_rate": 3.117076808351976e-05,
-      "loss": 4.1656,
+      "loss": 3.8762,
       "step": 2015
     },
     {
       "epoch": 0.4466062347999116,
-      "grad_norm": 3.1955862045288086,
+      "grad_norm": 4.124545097351074,
       "learning_rate": 3.110862540392742e-05,
-      "loss": 4.0902,
+      "loss": 3.7953,
       "step": 2020
     },
     {
       "epoch": 0.44771169577713904,
-      "grad_norm": 3.3935418128967285,
+      "grad_norm": 4.37811803817749,
       "learning_rate": 3.1046482724335074e-05,
-      "loss": 3.9551,
+      "loss": 3.6624,
       "step": 2025
     },
     {
       "epoch": 0.44881715675436656,
-      "grad_norm": 3.6117637157440186,
+      "grad_norm": 4.66016960144043,
       "learning_rate": 3.098434004474273e-05,
-      "loss": 4.1856,
+      "loss": 3.8866,
       "step": 2030
     },
     {
       "epoch": 0.4499226177315941,
-      "grad_norm": 3.432446002960205,
+      "grad_norm": 4.389432907104492,
       "learning_rate": 3.0922197365150386e-05,
-      "loss": 4.1022,
+      "loss": 3.8125,
       "step": 2035
     },
     {
       "epoch": 0.45102807870882156,
-      "grad_norm": 3.3948235511779785,
+      "grad_norm": 4.264665126800537,
       "learning_rate": 3.086005468555804e-05,
-      "loss": 4.0777,
+      "loss": 3.7936,
       "step": 2040
     },
     {
       "epoch": 0.4521335396860491,
-      "grad_norm": 3.169699192047119,
+      "grad_norm": 4.068538665771484,
       "learning_rate": 3.07979120059657e-05,
-      "loss": 4.0572,
+      "loss": 3.7711,
       "step": 2045
     },
     {
       "epoch": 0.4532390006632766,
-      "grad_norm": 3.3817138671875,
+      "grad_norm": 4.360579013824463,
       "learning_rate": 3.0735769326373355e-05,
-      "loss": 4.0533,
+      "loss": 3.7848,
       "step": 2050
     },
     {
       "epoch": 0.45434446164050407,
-      "grad_norm": 3.4111692905426025,
+      "grad_norm": 4.351474285125732,
       "learning_rate": 3.067362664678101e-05,
-      "loss": 4.0991,
+      "loss": 3.8165,
       "step": 2055
     },
     {
       "epoch": 0.4554499226177316,
-      "grad_norm": 3.7082407474517822,
+      "grad_norm": 4.792845726013184,
       "learning_rate": 3.061148396718867e-05,
-      "loss": 4.0745,
+      "loss": 3.7943,
       "step": 2060
     },
     {
       "epoch": 0.4565553835949591,
-      "grad_norm": 3.393707036972046,
+      "grad_norm": 4.3848347663879395,
       "learning_rate": 3.054934128759632e-05,
-      "loss": 3.9929,
+      "loss": 3.7159,
       "step": 2065
     },
     {
       "epoch": 0.4576608445721866,
-      "grad_norm": 3.750239133834839,
+      "grad_norm": 4.771290302276611,
       "learning_rate": 3.048719860800398e-05,
-      "loss": 4.1835,
+      "loss": 3.909,
       "step": 2070
     },
     {
       "epoch": 0.4587663055494141,
-      "grad_norm": 3.366420030593872,
+      "grad_norm": 4.310284614562988,
       "learning_rate": 3.0425055928411632e-05,
-      "loss": 4.0436,
+      "loss": 3.7778,
       "step": 2075
     },
     {
       "epoch": 0.45987176652664163,
-      "grad_norm": 3.3570804595947266,
+      "grad_norm": 4.319281101226807,
       "learning_rate": 3.0362913248819292e-05,
-      "loss": 3.9977,
+      "loss": 3.7306,
       "step": 2080
     },
     {
       "epoch": 0.4609772275038691,
-      "grad_norm": 3.541613817214966,
+      "grad_norm": 4.402329444885254,
       "learning_rate": 3.0300770569226945e-05,
-      "loss": 4.0789,
+      "loss": 3.8035,
       "step": 2085
     },
     {
       "epoch": 0.4620826884810966,
-      "grad_norm": 3.697382926940918,
+      "grad_norm": 4.694875240325928,
       "learning_rate": 3.02386278896346e-05,
-      "loss": 4.1316,
+      "loss": 3.8439,
       "step": 2090
     },
     {
       "epoch": 0.4631881494583241,
-      "grad_norm": 3.375995397567749,
+      "grad_norm": 4.330312728881836,
       "learning_rate": 3.017648521004226e-05,
-      "loss": 3.935,
+      "loss": 3.6599,
       "step": 2095
     },
     {
       "epoch": 0.4642936104355516,
-      "grad_norm": 3.3144774436950684,
+      "grad_norm": 4.186635971069336,
       "learning_rate": 3.0114342530449913e-05,
-      "loss": 4.1222,
+      "loss": 3.846,
       "step": 2100
     },
     {
       "epoch": 0.46539907141277914,
-      "grad_norm": 3.600338935852051,
+      "grad_norm": 4.592949390411377,
       "learning_rate": 3.005219985085757e-05,
-      "loss": 4.1123,
+      "loss": 3.8302,
       "step": 2105
     },
     {
       "epoch": 0.4665045323900066,
-      "grad_norm": 3.3715898990631104,
+      "grad_norm": 4.36030387878418,
       "learning_rate": 2.999005717126523e-05,
-      "loss": 4.1952,
+      "loss": 3.9232,
       "step": 2110
     },
     {
       "epoch": 0.46760999336723413,
-      "grad_norm": 3.2076468467712402,
+      "grad_norm": 4.067847728729248,
       "learning_rate": 2.992791449167288e-05,
-      "loss": 3.9456,
+      "loss": 3.6619,
       "step": 2115
     },
     {
       "epoch": 0.46871545434446166,
-      "grad_norm": 3.7750439643859863,
+      "grad_norm": 4.751062393188477,
       "learning_rate": 2.986577181208054e-05,
-      "loss": 4.0785,
+      "loss": 3.8007,
       "step": 2120
     },
     {
       "epoch": 0.4698209153216891,
-      "grad_norm": 3.3552026748657227,
+      "grad_norm": 4.296187400817871,
       "learning_rate": 2.980362913248819e-05,
-      "loss": 4.3222,
+      "loss": 4.0479,
       "step": 2125
     },
     {
       "epoch": 0.47092637629891665,
-      "grad_norm": 3.4313700199127197,
+      "grad_norm": 4.393710613250732,
       "learning_rate": 2.974148645289585e-05,
-      "loss": 3.9145,
+      "loss": 3.6467,
       "step": 2130
     },
     {
       "epoch": 0.4720318372761442,
-      "grad_norm": 3.4928014278411865,
+      "grad_norm": 4.423858165740967,
       "learning_rate": 2.967934377330351e-05,
-      "loss": 3.8454,
+      "loss": 3.5571,
       "step": 2135
     },
     {
       "epoch": 0.47313729825337164,
-      "grad_norm": 3.6989784240722656,
+      "grad_norm": 4.678510665893555,
       "learning_rate": 2.9617201093711163e-05,
-      "loss": 3.9581,
+      "loss": 3.6886,
       "step": 2140
     },
     {
       "epoch": 0.47424275923059916,
-      "grad_norm": 3.152308702468872,
+      "grad_norm": 4.031360149383545,
       "learning_rate": 2.955505841411882e-05,
-      "loss": 3.9159,
+      "loss": 3.6449,
       "step": 2145
     },
     {
       "epoch": 0.4753482202078267,
-      "grad_norm": 3.2610297203063965,
+      "grad_norm": 4.206479549407959,
       "learning_rate": 2.949291573452647e-05,
-      "loss": 4.1646,
+      "loss": 3.8864,
       "step": 2150
     },
     {
       "epoch": 0.47645368118505416,
-      "grad_norm": 3.4919862747192383,
+      "grad_norm": 4.505343914031982,
       "learning_rate": 2.943077305493413e-05,
-      "loss": 3.9627,
+      "loss": 3.6903,
       "step": 2155
     },
     {
       "epoch": 0.4775591421622817,
-      "grad_norm": 3.323495388031006,
+      "grad_norm": 4.31165075302124,
       "learning_rate": 2.9368630375341787e-05,
-      "loss": 3.9826,
+      "loss": 3.709,
       "step": 2160
     },
     {
       "epoch": 0.47866460313950915,
-      "grad_norm": 3.4803435802459717,
+      "grad_norm": 4.356276988983154,
       "learning_rate": 2.930648769574944e-05,
-      "loss": 4.1256,
+      "loss": 3.8502,
       "step": 2165
     },
     {
       "epoch": 0.47977006411673667,
-      "grad_norm": 3.3792881965637207,
+      "grad_norm": 4.317960739135742,
       "learning_rate": 2.92443450161571e-05,
-      "loss": 3.9697,
+      "loss": 3.7002,
       "step": 2170
     },
     {
       "epoch": 0.4808755250939642,
-      "grad_norm": 3.5845255851745605,
+      "grad_norm": 4.5985541343688965,
       "learning_rate": 2.9182202336564756e-05,
-      "loss": 4.1054,
+      "loss": 3.8314,
       "step": 2175
     },
     {
       "epoch": 0.48198098607119166,
-      "grad_norm": 3.275973081588745,
+      "grad_norm": 4.173244476318359,
       "learning_rate": 2.912005965697241e-05,
-      "loss": 4.1417,
+      "loss": 3.8636,
       "step": 2180
     },
     {
       "epoch": 0.4830864470484192,
-      "grad_norm": 3.3241536617279053,
+      "grad_norm": 4.283194541931152,
       "learning_rate": 2.9057916977380068e-05,
-      "loss": 4.0629,
+      "loss": 3.793,
       "step": 2185
     },
     {
       "epoch": 0.4841919080256467,
-      "grad_norm": 3.298708200454712,
+      "grad_norm": 4.248963832855225,
       "learning_rate": 2.899577429778772e-05,
-      "loss": 3.9206,
+      "loss": 3.6518,
       "step": 2190
     },
     {
       "epoch": 0.4852973690028742,
-      "grad_norm": 3.18892502784729,
+      "grad_norm": 4.0418477058410645,
       "learning_rate": 2.8933631618195377e-05,
-      "loss": 4.0769,
+      "loss": 3.8178,
       "step": 2195
     },
     {
       "epoch": 0.4864028299801017,
-      "grad_norm": 3.206279993057251,
+      "grad_norm": 4.183856964111328,
       "learning_rate": 2.8871488938603037e-05,
-      "loss": 3.937,
+      "loss": 3.6664,
       "step": 2200
     },
     {
       "epoch": 0.4875082909573292,
-      "grad_norm": 3.4408323764801025,
+      "grad_norm": 4.381367206573486,
       "learning_rate": 2.880934625901069e-05,
-      "loss": 4.0496,
+      "loss": 3.7829,
       "step": 2205
     },
     {
       "epoch": 0.4886137519345567,
-      "grad_norm": 3.258359670639038,
+      "grad_norm": 4.142855167388916,
       "learning_rate": 2.8747203579418346e-05,
-      "loss": 4.038,
+      "loss": 3.7652,
       "step": 2210
     },
     {
       "epoch": 0.4897192129117842,
-      "grad_norm": 3.336268424987793,
+      "grad_norm": 4.279924392700195,
       "learning_rate": 2.8685060899826e-05,
-      "loss": 4.0437,
+      "loss": 3.7693,
       "step": 2215
     },
     {
       "epoch": 0.49082467388901174,
-      "grad_norm": 3.27437686920166,
+      "grad_norm": 4.1756134033203125,
       "learning_rate": 2.8622918220233658e-05,
-      "loss": 4.1238,
+      "loss": 3.8588,
       "step": 2220
     },
     {
       "epoch": 0.4919301348662392,
-      "grad_norm": 3.076141595840454,
+      "grad_norm": 3.9072275161743164,
       "learning_rate": 2.8560775540641317e-05,
-      "loss": 3.987,
+      "loss": 3.7163,
       "step": 2225
     },
     {
       "epoch": 0.49303559584346673,
-      "grad_norm": 3.2528483867645264,
+      "grad_norm": 4.12607479095459,
       "learning_rate": 2.8498632861048967e-05,
-      "loss": 3.9728,
+      "loss": 3.7101,
       "step": 2230
     },
     {
       "epoch": 0.4941410568206942,
-      "grad_norm": 3.397096872329712,
+      "grad_norm": 4.380521297454834,
       "learning_rate": 2.8436490181456626e-05,
-      "loss": 4.1707,
+      "loss": 3.905,
       "step": 2235
     },
     {
       "epoch": 0.4952465177979217,
-      "grad_norm": 3.2209689617156982,
+      "grad_norm": 4.143880844116211,
       "learning_rate": 2.837434750186428e-05,
-      "loss": 4.0548,
+      "loss": 3.7826,
       "step": 2240
     },
     {
       "epoch": 0.49635197877514925,
-      "grad_norm": 3.292736530303955,
+      "grad_norm": 4.162075996398926,
       "learning_rate": 2.831220482227194e-05,
-      "loss": 4.0244,
+      "loss": 3.7528,
       "step": 2245
     },
     {
       "epoch": 0.4974574397523767,
-      "grad_norm": 3.461022138595581,
+      "grad_norm": 4.480312824249268,
       "learning_rate": 2.8250062142679595e-05,
-      "loss": 4.0763,
+      "loss": 3.8186,
       "step": 2250
     },
     {
       "epoch": 0.49856290072960424,
-      "grad_norm": 3.4967451095581055,
+      "grad_norm": 4.4721503257751465,
       "learning_rate": 2.8187919463087248e-05,
-      "loss": 4.1721,
+      "loss": 3.9181,
       "step": 2255
     },
     {
       "epoch": 0.49966836170683177,
-      "grad_norm": 3.2440531253814697,
+      "grad_norm": 4.137672424316406,
       "learning_rate": 2.8125776783494907e-05,
-      "loss": 4.0546,
+      "loss": 3.7862,
       "step": 2260
     },
     {
       "epoch": 0.5007738226840592,
-      "grad_norm": 3.318380355834961,
+      "grad_norm": 4.334343433380127,
       "learning_rate": 2.8063634103902563e-05,
-      "loss": 4.0751,
+      "loss": 3.8156,
       "step": 2265
     },
     {
       "epoch": 0.5018792836612868,
-      "grad_norm": 3.1638567447662354,
+      "grad_norm": 3.9582431316375732,
       "learning_rate": 2.8001491424310216e-05,
-      "loss": 3.9274,
+      "loss": 3.6614,
       "step": 2270
     },
     {
       "epoch": 0.5029847446385143,
-      "grad_norm": 3.345717430114746,
+      "grad_norm": 4.25573205947876,
       "learning_rate": 2.7939348744717876e-05,
-      "loss": 4.1606,
+      "loss": 3.8853,
       "step": 2275
     },
     {
       "epoch": 0.5040902056157418,
-      "grad_norm": 3.5760574340820312,
+      "grad_norm": 4.529309272766113,
       "learning_rate": 2.787720606512553e-05,
-      "loss": 3.9832,
+      "loss": 3.7192,
       "step": 2280
     },
     {
       "epoch": 0.5051956665929692,
-      "grad_norm": 3.3899612426757812,
+      "grad_norm": 4.277974605560303,
       "learning_rate": 2.7815063385533185e-05,
-      "loss": 4.0456,
+      "loss": 3.7831,
       "step": 2285
     },
     {
       "epoch": 0.5063011275701967,
-      "grad_norm": 3.3774311542510986,
+      "grad_norm": 4.308351039886475,
       "learning_rate": 2.7752920705940844e-05,
-      "loss": 4.0956,
+      "loss": 3.8402,
       "step": 2290
     },
     {
       "epoch": 0.5074065885474243,
-      "grad_norm": 3.1358556747436523,
+      "grad_norm": 4.028602123260498,
       "learning_rate": 2.7690778026348497e-05,
-      "loss": 4.1611,
+      "loss": 3.8994,
       "step": 2295
     },
     {
       "epoch": 0.5085120495246518,
-      "grad_norm": 3.3426547050476074,
+      "grad_norm": 4.28765869140625,
       "learning_rate": 2.7628635346756153e-05,
-      "loss": 4.1298,
+      "loss": 3.873,
       "step": 2300
     },
     {
       "epoch": 0.5096175105018793,
-      "grad_norm": 3.252143383026123,
+      "grad_norm": 4.152209758758545,
       "learning_rate": 2.7566492667163806e-05,
-      "loss": 4.0572,
+      "loss": 3.803,
       "step": 2305
     },
     {
       "epoch": 0.5107229714791068,
-      "grad_norm": 3.4557764530181885,
+      "grad_norm": 4.325049877166748,
       "learning_rate": 2.7504349987571466e-05,
-      "loss": 4.0824,
+      "loss": 3.8222,
       "step": 2310
     },
     {
       "epoch": 0.5118284324563342,
-      "grad_norm": 3.2078895568847656,
+      "grad_norm": 4.071306228637695,
       "learning_rate": 2.7442207307979122e-05,
-      "loss": 3.9483,
+      "loss": 3.6962,
       "step": 2315
     },
     {
       "epoch": 0.5129338934335618,
-      "grad_norm": 3.4674055576324463,
+      "grad_norm": 4.453413963317871,
       "learning_rate": 2.7380064628386775e-05,
-      "loss": 4.0843,
+      "loss": 3.8324,
       "step": 2320
     },
     {
       "epoch": 0.5140393544107893,
-      "grad_norm": 3.7841782569885254,
+      "grad_norm": 4.732785224914551,
       "learning_rate": 2.7317921948794434e-05,
-      "loss": 4.2304,
+      "loss": 3.9613,
       "step": 2325
     },
     {
       "epoch": 0.5151448153880168,
-      "grad_norm": 3.267167091369629,
+      "grad_norm": 4.23225736618042,
       "learning_rate": 2.725577926920209e-05,
-      "loss": 4.0463,
+      "loss": 3.7984,
       "step": 2330
     },
     {
       "epoch": 0.5162502763652443,
-      "grad_norm": 3.782557725906372,
+      "grad_norm": 4.763914585113525,
       "learning_rate": 2.7193636589609743e-05,
-      "loss": 4.0149,
+      "loss": 3.7478,
       "step": 2335
     },
     {
       "epoch": 0.5173557373424718,
-      "grad_norm": 3.4802868366241455,
+      "grad_norm": 4.378221035003662,
       "learning_rate": 2.7131493910017403e-05,
-      "loss": 3.9961,
+      "loss": 3.7383,
       "step": 2340
     },
     {
       "epoch": 0.5184611983196993,
-      "grad_norm": 3.346196413040161,
+      "grad_norm": 4.221852779388428,
       "learning_rate": 2.7069351230425055e-05,
-      "loss": 3.9326,
+      "loss": 3.6738,
       "step": 2345
     },
     {
       "epoch": 0.5195666592969268,
-      "grad_norm": 3.166124105453491,
+      "grad_norm": 4.040189743041992,
       "learning_rate": 2.7007208550832715e-05,
-      "loss": 3.9516,
+      "loss": 3.6944,
       "step": 2350
     },
     {
       "epoch": 0.5206721202741543,
-      "grad_norm": 3.288295269012451,
+      "grad_norm": 4.224726676940918,
       "learning_rate": 2.694506587124037e-05,
-      "loss": 4.1143,
+      "loss": 3.8663,
       "step": 2355
     },
     {
       "epoch": 0.5217775812513819,
-      "grad_norm": 3.3296289443969727,
+      "grad_norm": 4.208111763000488,
       "learning_rate": 2.6882923191648024e-05,
-      "loss": 3.8863,
+      "loss": 3.6178,
       "step": 2360
     },
     {
       "epoch": 0.5228830422286094,
-      "grad_norm": 3.1221563816070557,
+      "grad_norm": 4.035094261169434,
       "learning_rate": 2.6820780512055683e-05,
-      "loss": 3.9889,
+      "loss": 3.7504,
       "step": 2365
     },
     {
       "epoch": 0.5239885032058368,
-      "grad_norm": 3.225713014602661,
+      "grad_norm": 4.1322855949401855,
       "learning_rate": 2.6758637832463336e-05,
-      "loss": 3.947,
+      "loss": 3.687,
       "step": 2370
     },
     {
       "epoch": 0.5250939641830643,
-      "grad_norm": 3.5291709899902344,
+      "grad_norm": 4.4745612144470215,
       "learning_rate": 2.6696495152870992e-05,
-      "loss": 4.1917,
+      "loss": 3.9329,
       "step": 2375
     },
     {
       "epoch": 0.5261994251602918,
-      "grad_norm": 3.4283344745635986,
+      "grad_norm": 4.301798343658447,
       "learning_rate": 2.6634352473278652e-05,
-      "loss": 4.0173,
+      "loss": 3.7627,
       "step": 2380
     },
     {
       "epoch": 0.5273048861375194,
-      "grad_norm": 3.4083287715911865,
+      "grad_norm": 4.328381538391113,
       "learning_rate": 2.6572209793686305e-05,
-      "loss": 4.1016,
+      "loss": 3.8483,
       "step": 2385
     },
     {
       "epoch": 0.5284103471147469,
-      "grad_norm": 3.3082547187805176,
+      "grad_norm": 4.222133159637451,
       "learning_rate": 2.651006711409396e-05,
-      "loss": 4.1025,
+      "loss": 3.8548,
       "step": 2390
     },
     {
       "epoch": 0.5295158080919744,
-      "grad_norm": 3.645259141921997,
+      "grad_norm": 4.592624187469482,
       "learning_rate": 2.644792443450162e-05,
-      "loss": 3.9657,
+      "loss": 3.7049,
       "step": 2395
     },
     {
       "epoch": 0.5306212690692018,
-      "grad_norm": 3.1570723056793213,
+      "grad_norm": 3.992602825164795,
       "learning_rate": 2.6385781754909273e-05,
-      "loss": 4.0965,
+      "loss": 3.8514,
       "step": 2400
     },
     {
       "epoch": 0.5317267300464293,
-      "grad_norm": 3.387300491333008,
+      "grad_norm": 4.265964031219482,
       "learning_rate": 2.632363907531693e-05,
-      "loss": 4.0099,
+      "loss": 3.766,
       "step": 2405
     },
     {
       "epoch": 0.5328321910236569,
-      "grad_norm": 3.4514920711517334,
+      "grad_norm": 4.394291877746582,
       "learning_rate": 2.6261496395724582e-05,
-      "loss": 3.9037,
+      "loss": 3.6601,
       "step": 2410
     },
     {
       "epoch": 0.5339376520008844,
-      "grad_norm": 3.7543208599090576,
+      "grad_norm": 4.654317378997803,
       "learning_rate": 2.6199353716132242e-05,
-      "loss": 4.0804,
+      "loss": 3.8274,
       "step": 2415
     },
     {
       "epoch": 0.5350431129781119,
-      "grad_norm": 3.4875600337982178,
+      "grad_norm": 4.4481635093688965,
       "learning_rate": 2.6137211036539898e-05,
-      "loss": 4.0105,
+      "loss": 3.7661,
       "step": 2420
     },
     {
       "epoch": 0.5361485739553393,
-      "grad_norm": 3.4124867916107178,
+      "grad_norm": 4.375431537628174,
       "learning_rate": 2.607506835694755e-05,
-      "loss": 4.1436,
+      "loss": 3.8995,
       "step": 2425
     },
     {
       "epoch": 0.5372540349325668,
-      "grad_norm": 3.392489194869995,
+      "grad_norm": 4.220302104949951,
       "learning_rate": 2.601292567735521e-05,
-      "loss": 4.087,
+      "loss": 3.841,
       "step": 2430
     },
     {
       "epoch": 0.5383594959097944,
-      "grad_norm": 3.3754377365112305,
+      "grad_norm": 4.340846538543701,
       "learning_rate": 2.5950782997762863e-05,
-      "loss": 4.0433,
+      "loss": 3.7965,
       "step": 2435
     },
     {
       "epoch": 0.5394649568870219,
-      "grad_norm": 3.23037052154541,
+      "grad_norm": 4.112565517425537,
       "learning_rate": 2.588864031817052e-05,
-      "loss": 3.9529,
+      "loss": 3.7081,
       "step": 2440
     },
     {
       "epoch": 0.5405704178642494,
-      "grad_norm": 3.4852147102355957,
+      "grad_norm": 4.324024677276611,
       "learning_rate": 2.582649763857818e-05,
-      "loss": 4.0165,
+      "loss": 3.7511,
       "step": 2445
     },
     {
       "epoch": 0.5416758788414769,
-      "grad_norm": 3.5113587379455566,
+      "grad_norm": 4.430271148681641,
       "learning_rate": 2.576435495898583e-05,
-      "loss": 4.1145,
+      "loss": 3.8567,
       "step": 2450
     },
     {
       "epoch": 0.5427813398187044,
-      "grad_norm": 3.569577693939209,
+      "grad_norm": 4.566709518432617,
       "learning_rate": 2.5702212279393488e-05,
-      "loss": 4.2112,
+      "loss": 3.9709,
       "step": 2455
     },
     {
       "epoch": 0.5438868007959319,
-      "grad_norm": 3.2119925022125244,
+      "grad_norm": 4.033237457275391,
       "learning_rate": 2.564006959980114e-05,
-      "loss": 3.9315,
+      "loss": 3.6954,
       "step": 2460
     },
     {
       "epoch": 0.5449922617731594,
-      "grad_norm": 3.502654790878296,
+      "grad_norm": 4.501426696777344,
       "learning_rate": 2.55779269202088e-05,
-      "loss": 4.0101,
+      "loss": 3.768,
       "step": 2465
     },
     {
       "epoch": 0.5460977227503869,
-      "grad_norm": 3.343017101287842,
+      "grad_norm": 4.246179580688477,
       "learning_rate": 2.551578424061646e-05,
-      "loss": 4.0618,
+      "loss": 3.8199,
       "step": 2470
     },
     {
       "epoch": 0.5472031837276145,
-      "grad_norm": 3.0435657501220703,
+      "grad_norm": 3.870412826538086,
       "learning_rate": 2.545364156102411e-05,
-      "loss": 4.2169,
+      "loss": 3.9862,
       "step": 2475
     },
     {
       "epoch": 0.548308644704842,
-      "grad_norm": 3.167151927947998,
+      "grad_norm": 3.9619927406311035,
       "learning_rate": 2.539149888143177e-05,
-      "loss": 4.043,
+      "loss": 3.7974,
       "step": 2480
     },
     {
       "epoch": 0.5494141056820694,
-      "grad_norm": 3.2351808547973633,
+      "grad_norm": 4.120594024658203,
       "learning_rate": 2.5329356201839428e-05,
-      "loss": 4.0381,
+      "loss": 3.8012,
       "step": 2485
     },
     {
       "epoch": 0.5505195666592969,
-      "grad_norm": 3.1816964149475098,
+      "grad_norm": 4.058072090148926,
       "learning_rate": 2.526721352224708e-05,
-      "loss": 4.1283,
+      "loss": 3.8845,
       "step": 2490
     },
     {
       "epoch": 0.5516250276365244,
-      "grad_norm": 3.2556283473968506,
+      "grad_norm": 4.13712739944458,
       "learning_rate": 2.5205070842654737e-05,
-      "loss": 4.1709,
+      "loss": 3.9284,
       "step": 2495
     },
     {
       "epoch": 0.552730488613752,
-      "grad_norm": 3.2887418270111084,
+      "grad_norm": 4.207889556884766,
       "learning_rate": 2.514292816306239e-05,
-      "loss": 4.1116,
+      "loss": 3.8678,
       "step": 2500
     },
     {
       "epoch": 0.5538359495909795,
-      "grad_norm": 3.559380531311035,
+      "grad_norm": 4.438493728637695,
       "learning_rate": 2.508078548347005e-05,
-      "loss": 4.0527,
+      "loss": 3.802,
       "step": 2505
     },
     {
       "epoch": 0.5549414105682069,
-      "grad_norm": 3.470162868499756,
+      "grad_norm": 4.422073841094971,
       "learning_rate": 2.5018642803877706e-05,
-      "loss": 4.0154,
+      "loss": 3.7657,
       "step": 2510
     },
     {
       "epoch": 0.5560468715454344,
-      "grad_norm": 3.294788122177124,
+      "grad_norm": 4.152473449707031,
       "learning_rate": 2.495650012428536e-05,
-      "loss": 4.0073,
+      "loss": 3.761,
       "step": 2515
     },
     {
       "epoch": 0.5571523325226619,
-      "grad_norm": 3.3408074378967285,
+      "grad_norm": 4.228184223175049,
       "learning_rate": 2.4894357444693018e-05,
-      "loss": 4.1111,
+      "loss": 3.8752,
       "step": 2520
     },
     {
       "epoch": 0.5582577934998895,
-      "grad_norm": 3.436032295227051,
+      "grad_norm": 4.336376667022705,
       "learning_rate": 2.4832214765100674e-05,
-      "loss": 4.0138,
+      "loss": 3.7806,
       "step": 2525
     },
     {
       "epoch": 0.559363254477117,
-      "grad_norm": 3.383261203765869,
+      "grad_norm": 4.211087703704834,
       "learning_rate": 2.4770072085508327e-05,
-      "loss": 4.1234,
+      "loss": 3.8946,
       "step": 2530
     },
     {
       "epoch": 0.5604687154543445,
-      "grad_norm": 3.479888916015625,
+      "grad_norm": 4.473411560058594,
       "learning_rate": 2.4707929405915983e-05,
-      "loss": 4.0519,
+      "loss": 3.8279,
       "step": 2535
     },
     {
       "epoch": 0.5615741764315719,
-      "grad_norm": 3.390536069869995,
+      "grad_norm": 4.341585636138916,
       "learning_rate": 2.4645786726323643e-05,
-      "loss": 4.1424,
+      "loss": 3.9098,
       "step": 2540
     },
     {
       "epoch": 0.5626796374087994,
-      "grad_norm": 3.320270538330078,
+      "grad_norm": 4.200295448303223,
       "learning_rate": 2.4583644046731296e-05,
-      "loss": 4.0054,
+      "loss": 3.7669,
       "step": 2545
     },
     {
       "epoch": 0.563785098386027,
-      "grad_norm": 3.477365016937256,
+      "grad_norm": 4.382364273071289,
       "learning_rate": 2.452150136713895e-05,
-      "loss": 4.0191,
+      "loss": 3.7919,
       "step": 2550
     },
     {
       "epoch": 0.5648905593632545,
-      "grad_norm": 3.547175884246826,
+      "grad_norm": 4.437411308288574,
       "learning_rate": 2.4459358687546608e-05,
-      "loss": 4.0718,
+      "loss": 3.849,
       "step": 2555
     },
     {
       "epoch": 0.565996020340482,
-      "grad_norm": 3.567544937133789,
+      "grad_norm": 4.420955657958984,
       "learning_rate": 2.4397216007954264e-05,
-      "loss": 4.1387,
+      "loss": 3.9096,
       "step": 2560
     },
     {
       "epoch": 0.5671014813177094,
-      "grad_norm": 3.351850748062134,
+      "grad_norm": 4.230448246002197,
       "learning_rate": 2.433507332836192e-05,
-      "loss": 4.0413,
+      "loss": 3.8093,
       "step": 2565
     },
     {
       "epoch": 0.568206942294937,
-      "grad_norm": 3.4294025897979736,
+      "grad_norm": 4.388647556304932,
       "learning_rate": 2.4272930648769576e-05,
-      "loss": 4.0404,
+      "loss": 3.8167,
       "step": 2570
     },
     {
       "epoch": 0.5693124032721645,
-      "grad_norm": 3.4079086780548096,
+      "grad_norm": 4.334132194519043,
       "learning_rate": 2.4210787969177233e-05,
-      "loss": 4.1669,
+      "loss": 3.9239,
       "step": 2575
     },
     {
       "epoch": 0.570417864249392,
-      "grad_norm": 3.6439168453216553,
+      "grad_norm": 4.52349853515625,
       "learning_rate": 2.4148645289584885e-05,
-      "loss": 4.109,
+      "loss": 3.8651,
       "step": 2580
     },
     {
       "epoch": 0.5715233252266195,
-      "grad_norm": 3.3144097328186035,
+      "grad_norm": 4.230077743530273,
       "learning_rate": 2.4086502609992545e-05,
-      "loss": 3.9591,
+      "loss": 3.7274,
       "step": 2585
     },
     {
       "epoch": 0.572628786203847,
-      "grad_norm": 3.3762526512145996,
+      "grad_norm": 4.300130367279053,
       "learning_rate": 2.40243599304002e-05,
-      "loss": 4.1867,
+      "loss": 3.9547,
       "step": 2590
     },
     {
       "epoch": 0.5737342471810745,
-      "grad_norm": 3.2939674854278564,
+      "grad_norm": 4.160315036773682,
       "learning_rate": 2.3962217250807857e-05,
-      "loss": 4.1226,
+      "loss": 3.8938,
       "step": 2595
     },
     {
       "epoch": 0.574839708158302,
-      "grad_norm": 3.094438314437866,
+      "grad_norm": 3.896183967590332,
       "learning_rate": 2.390007457121551e-05,
-      "loss": 3.9615,
+      "loss": 3.7268,
       "step": 2600
     },
     {
       "epoch": 0.5759451691355295,
-      "grad_norm": 3.3845763206481934,
+      "grad_norm": 4.2746663093566895,
       "learning_rate": 2.383793189162317e-05,
-      "loss": 3.9805,
+      "loss": 3.7383,
       "step": 2605
     },
     {
       "epoch": 0.577050630112757,
-      "grad_norm": 3.696262836456299,
+      "grad_norm": 4.682511329650879,
       "learning_rate": 2.3775789212030826e-05,
-      "loss": 3.8625,
+      "loss": 3.6358,
       "step": 2610
     },
     {
       "epoch": 0.5781560910899846,
-      "grad_norm": 3.3800036907196045,
+      "grad_norm": 4.271305561065674,
       "learning_rate": 2.371364653243848e-05,
-      "loss": 4.1462,
+      "loss": 3.9161,
       "step": 2615
     },
     {
       "epoch": 0.5792615520672121,
-      "grad_norm": 3.573200225830078,
+      "grad_norm": 4.486483573913574,
       "learning_rate": 2.3651503852846135e-05,
-      "loss": 4.1071,
+      "loss": 3.8655,
       "step": 2620
     },
     {
       "epoch": 0.5803670130444395,
-      "grad_norm": 3.651068925857544,
+      "grad_norm": 4.57070779800415,
       "learning_rate": 2.358936117325379e-05,
-      "loss": 4.0191,
+      "loss": 3.7881,
       "step": 2625
     },
     {
       "epoch": 0.581472474021667,
-      "grad_norm": 3.1807289123535156,
+      "grad_norm": 4.042338848114014,
       "learning_rate": 2.3527218493661447e-05,
-      "loss": 4.1579,
+      "loss": 3.9398,
       "step": 2630
     },
     {
       "epoch": 0.5825779349988945,
-      "grad_norm": 3.5472700595855713,
+      "grad_norm": 4.45573616027832,
       "learning_rate": 2.3465075814069103e-05,
-      "loss": 4.0699,
+      "loss": 3.8521,
       "step": 2635
     },
     {
       "epoch": 0.5836833959761221,
-      "grad_norm": 3.3236019611358643,
+      "grad_norm": 4.1683831214904785,
       "learning_rate": 2.340293313447676e-05,
-      "loss": 3.9927,
+      "loss": 3.762,
       "step": 2640
     },
     {
       "epoch": 0.5847888569533496,
-      "grad_norm": 3.5756359100341797,
+      "grad_norm": 4.501931667327881,
       "learning_rate": 2.3340790454884416e-05,
-      "loss": 4.2018,
+      "loss": 3.9792,
       "step": 2645
     },
     {
       "epoch": 0.585894317930577,
-      "grad_norm": 3.5606160163879395,
+      "grad_norm": 4.491320610046387,
       "learning_rate": 2.3278647775292072e-05,
-      "loss": 4.0626,
+      "loss": 3.8377,
       "step": 2650
     },
     {
       "epoch": 0.5869997789078045,
-      "grad_norm": 3.5119574069976807,
+      "grad_norm": 4.437068939208984,
       "learning_rate": 2.3216505095699728e-05,
-      "loss": 4.0997,
+      "loss": 3.8754,
       "step": 2655
     },
     {
       "epoch": 0.588105239885032,
-      "grad_norm": 3.373201847076416,
+      "grad_norm": 4.2230329513549805,
       "learning_rate": 2.3154362416107384e-05,
-      "loss": 3.9609,
+      "loss": 3.725,
       "step": 2660
     },
     {
       "epoch": 0.5892107008622596,
-      "grad_norm": 3.168120861053467,
+      "grad_norm": 3.982203960418701,
       "learning_rate": 2.309221973651504e-05,
-      "loss": 3.8898,
+      "loss": 3.671,
       "step": 2665
     },
     {
       "epoch": 0.5903161618394871,
-      "grad_norm": 3.260366678237915,
+      "grad_norm": 4.173470497131348,
       "learning_rate": 2.3030077056922693e-05,
-      "loss": 4.0445,
+      "loss": 3.8289,
       "step": 2670
     },
     {
       "epoch": 0.5914216228167146,
-      "grad_norm": 3.53143572807312,
+      "grad_norm": 4.4012837409973145,
       "learning_rate": 2.2967934377330353e-05,
-      "loss": 4.033,
+      "loss": 3.818,
       "step": 2675
     },
     {
       "epoch": 0.592527083793942,
-      "grad_norm": 3.4146888256073,
+      "grad_norm": 4.33328104019165,
       "learning_rate": 2.290579169773801e-05,
-      "loss": 3.9579,
+      "loss": 3.7354,
       "step": 2680
     },
     {
       "epoch": 0.5936325447711696,
-      "grad_norm": 3.554407835006714,
+      "grad_norm": 4.475124359130859,
       "learning_rate": 2.284364901814566e-05,
-      "loss": 4.0876,
+      "loss": 3.8638,
       "step": 2685
     },
     {
       "epoch": 0.5947380057483971,
-      "grad_norm": 3.302635431289673,
+      "grad_norm": 4.114334583282471,
       "learning_rate": 2.2781506338553318e-05,
-      "loss": 4.0015,
+      "loss": 3.7941,
       "step": 2690
     },
     {
       "epoch": 0.5958434667256246,
-      "grad_norm": 2.994694948196411,
+      "grad_norm": 3.7643494606018066,
       "learning_rate": 2.2719363658960977e-05,
-      "loss": 4.1925,
+      "loss": 3.9709,
       "step": 2695
     },
     {
       "epoch": 0.5969489277028521,
-      "grad_norm": 3.191727876663208,
+      "grad_norm": 3.9660909175872803,
       "learning_rate": 2.2657220979368633e-05,
-      "loss": 4.0834,
+      "loss": 3.873,
       "step": 2700
     },
     {
       "epoch": 0.5980543886800795,
-      "grad_norm": 3.187432050704956,
+      "grad_norm": 3.947577953338623,
       "learning_rate": 2.2595078299776286e-05,
-      "loss": 4.1476,
+      "loss": 3.9182,
       "step": 2705
     },
     {
       "epoch": 0.5991598496573071,
-      "grad_norm": 3.8028817176818848,
+      "grad_norm": 4.767286777496338,
       "learning_rate": 2.2532935620183942e-05,
-      "loss": 4.0108,
+      "loss": 3.7869,
       "step": 2710
     },
     {
       "epoch": 0.6002653106345346,
-      "grad_norm": 3.493286609649658,
+      "grad_norm": 4.299619197845459,
       "learning_rate": 2.2470792940591602e-05,
-      "loss": 4.1705,
+      "loss": 3.9475,
       "step": 2715
     },
     {
       "epoch": 0.6013707716117621,
-      "grad_norm": 3.4640684127807617,
+      "grad_norm": 4.320889949798584,
       "learning_rate": 2.2408650260999255e-05,
-      "loss": 4.1311,
+      "loss": 3.9141,
       "step": 2720
     },
     {
       "epoch": 0.6024762325889896,
-      "grad_norm": 3.8911242485046387,
+      "grad_norm": 4.803108215332031,
       "learning_rate": 2.234650758140691e-05,
-      "loss": 4.1535,
+      "loss": 3.9314,
       "step": 2725
     },
     {
       "epoch": 0.6035816935662172,
-      "grad_norm": 3.4392147064208984,
+      "grad_norm": 4.3592023849487305,
       "learning_rate": 2.2284364901814567e-05,
-      "loss": 4.1343,
+      "loss": 3.9067,
       "step": 2730
     },
     {
       "epoch": 0.6046871545434446,
-      "grad_norm": 3.2995851039886475,
+      "grad_norm": 4.13433837890625,
       "learning_rate": 2.2222222222222223e-05,
-      "loss": 4.0273,
+      "loss": 3.8067,
       "step": 2735
     },
     {
       "epoch": 0.6057926155206721,
-      "grad_norm": 3.1584272384643555,
+      "grad_norm": 3.916799783706665,
       "learning_rate": 2.216007954262988e-05,
-      "loss": 4.2191,
+      "loss": 4.0041,
       "step": 2740
     },
     {
       "epoch": 0.6068980764978996,
-      "grad_norm": 3.7929775714874268,
+      "grad_norm": 4.724668502807617,
       "learning_rate": 2.2097936863037536e-05,
-      "loss": 3.9746,
+      "loss": 3.7573,
       "step": 2745
     },
     {
       "epoch": 0.6080035374751271,
-      "grad_norm": 3.4396305084228516,
+      "grad_norm": 4.300846576690674,
       "learning_rate": 2.2035794183445192e-05,
-      "loss": 4.2164,
+      "loss": 4.003,
       "step": 2750
     },
     {
       "epoch": 0.6091089984523547,
-      "grad_norm": 3.2499279975891113,
+      "grad_norm": 4.073404788970947,
       "learning_rate": 2.1973651503852845e-05,
-      "loss": 3.9657,
+      "loss": 3.7454,
       "step": 2755
     },
     {
       "epoch": 0.6102144594295822,
-      "grad_norm": 3.682943105697632,
+      "grad_norm": 4.611419200897217,
       "learning_rate": 2.1911508824260504e-05,
-      "loss": 4.0552,
+      "loss": 3.843,
       "step": 2760
     },
     {
       "epoch": 0.6113199204068096,
-      "grad_norm": 3.217568874359131,
+      "grad_norm": 4.033169746398926,
       "learning_rate": 2.184936614466816e-05,
-      "loss": 4.1355,
+      "loss": 3.9166,
       "step": 2765
     },
     {
       "epoch": 0.6124253813840371,
-      "grad_norm": 3.696176528930664,
+      "grad_norm": 4.590342044830322,
       "learning_rate": 2.1787223465075816e-05,
-      "loss": 4.1971,
+      "loss": 3.9761,
       "step": 2770
     },
     {
       "epoch": 0.6135308423612647,
-      "grad_norm": 3.366211175918579,
+      "grad_norm": 4.150300979614258,
       "learning_rate": 2.172508078548347e-05,
-      "loss": 4.1779,
+      "loss": 3.9621,
       "step": 2775
     },
     {
       "epoch": 0.6146363033384922,
-      "grad_norm": 3.3090131282806396,
+      "grad_norm": 4.157660961151123,
       "learning_rate": 2.1662938105891125e-05,
-      "loss": 4.0138,
+      "loss": 3.7965,
       "step": 2780
     },
     {
       "epoch": 0.6157417643157197,
-      "grad_norm": 3.492255210876465,
+      "grad_norm": 4.348726749420166,
       "learning_rate": 2.1600795426298785e-05,
-      "loss": 4.113,
+      "loss": 3.9045,
       "step": 2785
     },
     {
       "epoch": 0.6168472252929471,
-      "grad_norm": 3.2298202514648438,
+      "grad_norm": 4.053340911865234,
       "learning_rate": 2.1538652746706438e-05,
-      "loss": 4.0822,
+      "loss": 3.8756,
       "step": 2790
     },
     {
       "epoch": 0.6179526862701746,
-      "grad_norm": 3.3362765312194824,
+      "grad_norm": 4.1914896965026855,
       "learning_rate": 2.1476510067114094e-05,
-      "loss": 4.1301,
+      "loss": 3.9241,
       "step": 2795
     },
     {
       "epoch": 0.6190581472474022,
-      "grad_norm": 3.1772379875183105,
+      "grad_norm": 3.9739878177642822,
       "learning_rate": 2.141436738752175e-05,
-      "loss": 4.0127,
+      "loss": 3.8039,
       "step": 2800
     },
     {
       "epoch": 0.6201636082246297,
-      "grad_norm": 3.5195131301879883,
+      "grad_norm": 4.364059925079346,
       "learning_rate": 2.1352224707929406e-05,
-      "loss": 4.0411,
+      "loss": 3.8287,
       "step": 2805
     },
     {
       "epoch": 0.6212690692018572,
-      "grad_norm": 3.1108715534210205,
+      "grad_norm": 3.8501362800598145,
       "learning_rate": 2.1290082028337062e-05,
-      "loss": 3.9511,
+      "loss": 3.7416,
       "step": 2810
     },
     {
       "epoch": 0.6223745301790847,
-      "grad_norm": 3.278776168823242,
+      "grad_norm": 4.111638069152832,
       "learning_rate": 2.122793934874472e-05,
-      "loss": 4.086,
+      "loss": 3.882,
       "step": 2815
     },
     {
       "epoch": 0.6234799911563121,
-      "grad_norm": 3.3844807147979736,
+      "grad_norm": 4.203191757202148,
       "learning_rate": 2.1165796669152375e-05,
-      "loss": 4.1389,
+      "loss": 3.936,
       "step": 2820
     },
     {
       "epoch": 0.6245854521335397,
-      "grad_norm": 3.547020673751831,
+      "grad_norm": 4.4386115074157715,
       "learning_rate": 2.110365398956003e-05,
-      "loss": 4.1154,
+      "loss": 3.9066,
       "step": 2825
     },
     {
       "epoch": 0.6256909131107672,
-      "grad_norm": 3.083136558532715,
+      "grad_norm": 3.8763813972473145,
       "learning_rate": 2.1041511309967687e-05,
-      "loss": 3.9761,
+      "loss": 3.7641,
       "step": 2830
     },
     {
       "epoch": 0.6267963740879947,
-      "grad_norm": 3.7824316024780273,
+      "grad_norm": 4.7053399085998535,
       "learning_rate": 2.0979368630375343e-05,
-      "loss": 4.04,
+      "loss": 3.8262,
       "step": 2835
     },
     {
       "epoch": 0.6279018350652222,
-      "grad_norm": 3.584540367126465,
+      "grad_norm": 4.445671081542969,
       "learning_rate": 2.0917225950783e-05,
-      "loss": 4.0237,
+      "loss": 3.8117,
       "step": 2840
     },
     {
       "epoch": 0.6290072960424496,
-      "grad_norm": 3.4071264266967773,
+      "grad_norm": 4.288915157318115,
       "learning_rate": 2.0855083271190652e-05,
-      "loss": 4.0866,
+      "loss": 3.8779,
       "step": 2845
     },
     {
       "epoch": 0.6301127570196772,
-      "grad_norm": 3.149873733520508,
+      "grad_norm": 3.978316307067871,
       "learning_rate": 2.0792940591598312e-05,
-      "loss": 4.0776,
+      "loss": 3.8757,
       "step": 2850
     },
     {
       "epoch": 0.6312182179969047,
-      "grad_norm": 3.3021628856658936,
+      "grad_norm": 4.085043907165527,
       "learning_rate": 2.0730797912005968e-05,
-      "loss": 4.1142,
+      "loss": 3.9212,
       "step": 2855
     },
     {
       "epoch": 0.6323236789741322,
-      "grad_norm": 3.379462957382202,
+      "grad_norm": 4.243721961975098,
       "learning_rate": 2.066865523241362e-05,
-      "loss": 4.197,
+      "loss": 4.0037,
       "step": 2860
     },
     {
       "epoch": 0.6334291399513597,
-      "grad_norm": 3.624547243118286,
+      "grad_norm": 4.492323875427246,
       "learning_rate": 2.0606512552821277e-05,
-      "loss": 4.1014,
+      "loss": 3.8815,
       "step": 2865
     },
     {
       "epoch": 0.6345346009285873,
-      "grad_norm": 3.391458511352539,
+      "grad_norm": 4.298487186431885,
       "learning_rate": 2.0544369873228937e-05,
-      "loss": 4.1709,
+      "loss": 3.9724,
       "step": 2870
     },
     {
       "epoch": 0.6356400619058147,
-      "grad_norm": 3.3703296184539795,
+      "grad_norm": 4.262723922729492,
       "learning_rate": 2.0482227193636593e-05,
-      "loss": 4.0528,
+      "loss": 3.8388,
       "step": 2875
     },
     {
       "epoch": 0.6367455228830422,
-      "grad_norm": 3.6773877143859863,
+      "grad_norm": 4.588946342468262,
       "learning_rate": 2.0420084514044246e-05,
-      "loss": 4.0063,
+      "loss": 3.7975,
       "step": 2880
     },
     {
       "epoch": 0.6378509838602697,
-      "grad_norm": 3.203677177429199,
+      "grad_norm": 4.005734443664551,
       "learning_rate": 2.03579418344519e-05,
-      "loss": 4.0733,
+      "loss": 3.8809,
       "step": 2885
     },
     {
       "epoch": 0.6389564448374973,
-      "grad_norm": 3.36698055267334,
+      "grad_norm": 4.173913955688477,
       "learning_rate": 2.0295799154859558e-05,
-      "loss": 4.0456,
+      "loss": 3.8513,
       "step": 2890
     },
     {
       "epoch": 0.6400619058147248,
-      "grad_norm": 3.412586212158203,
+      "grad_norm": 4.228811740875244,
       "learning_rate": 2.0233656475267214e-05,
-      "loss": 4.0807,
+      "loss": 3.8895,
       "step": 2895
     },
     {
       "epoch": 0.6411673667919523,
-      "grad_norm": 3.175722599029541,
+      "grad_norm": 3.9854164123535156,
       "learning_rate": 2.017151379567487e-05,
-      "loss": 4.1196,
+      "loss": 3.9249,
       "step": 2900
     },
     {
       "epoch": 0.6422728277691797,
-      "grad_norm": 3.315753936767578,
+      "grad_norm": 4.171172618865967,
       "learning_rate": 2.0109371116082526e-05,
-      "loss": 4.0987,
+      "loss": 3.9026,
       "step": 2905
     },
     {
       "epoch": 0.6433782887464072,
-      "grad_norm": 3.3233401775360107,
+      "grad_norm": 4.155889511108398,
       "learning_rate": 2.0047228436490183e-05,
-      "loss": 4.0974,
+      "loss": 3.893,
       "step": 2910
     },
     {
       "epoch": 0.6444837497236348,
-      "grad_norm": 3.648879051208496,
+      "grad_norm": 4.507026672363281,
       "learning_rate": 1.998508575689784e-05,
-      "loss": 4.0625,
+      "loss": 3.8608,
       "step": 2915
     },
     {
       "epoch": 0.6455892107008623,
-      "grad_norm": 3.3237850666046143,
+      "grad_norm": 4.152273654937744,
       "learning_rate": 1.9922943077305495e-05,
-      "loss": 4.138,
+      "loss": 3.9472,
       "step": 2920
     },
     {
       "epoch": 0.6466946716780898,
-      "grad_norm": 3.314603090286255,
+      "grad_norm": 4.080592632293701,
       "learning_rate": 1.986080039771315e-05,
-      "loss": 4.2303,
+      "loss": 4.0403,
       "step": 2925
     },
     {
       "epoch": 0.6478001326553172,
-      "grad_norm": 3.116244316101074,
+      "grad_norm": 3.855576515197754,
       "learning_rate": 1.9798657718120804e-05,
-      "loss": 4.1222,
+      "loss": 3.9324,
       "step": 2930
     },
     {
       "epoch": 0.6489055936325447,
-      "grad_norm": 3.232257127761841,
+      "grad_norm": 4.013497352600098,
       "learning_rate": 1.9736515038528463e-05,
-      "loss": 4.1056,
+      "loss": 3.9004,
       "step": 2935
     },
     {
       "epoch": 0.6500110546097723,
-      "grad_norm": 3.373582124710083,
+      "grad_norm": 4.193099498748779,
       "learning_rate": 1.967437235893612e-05,
-      "loss": 4.1316,
+      "loss": 3.9288,
       "step": 2940
     },
     {
       "epoch": 0.6511165155869998,
-      "grad_norm": 3.2493808269500732,
+      "grad_norm": 4.037634372711182,
       "learning_rate": 1.9612229679343776e-05,
-      "loss": 4.1764,
+      "loss": 3.9797,
       "step": 2945
     },
     {
       "epoch": 0.6522219765642273,
-      "grad_norm": 2.9851105213165283,
+      "grad_norm": 3.751389741897583,
       "learning_rate": 1.955008699975143e-05,
-      "loss": 4.074,
+      "loss": 3.8868,
       "step": 2950
     },
     {
       "epoch": 0.6533274375414548,
-      "grad_norm": 3.526233196258545,
+      "grad_norm": 4.334343910217285,
       "learning_rate": 1.9487944320159085e-05,
-      "loss": 4.0382,
+      "loss": 3.8372,
       "step": 2955
     },
     {
       "epoch": 0.6544328985186822,
-      "grad_norm": 3.4045310020446777,
+      "grad_norm": 4.235761642456055,
       "learning_rate": 1.9425801640566744e-05,
-      "loss": 4.012,
+      "loss": 3.8144,
       "step": 2960
     },
     {
       "epoch": 0.6555383594959098,
-      "grad_norm": 3.5040388107299805,
+      "grad_norm": 4.39980411529541,
       "learning_rate": 1.9363658960974397e-05,
-      "loss": 3.9922,
+      "loss": 3.7894,
       "step": 2965
     },
     {
       "epoch": 0.6566438204731373,
-      "grad_norm": 3.4251108169555664,
+      "grad_norm": 4.2445268630981445,
       "learning_rate": 1.9301516281382053e-05,
-      "loss": 4.0577,
+      "loss": 3.863,
       "step": 2970
     },
     {
       "epoch": 0.6577492814503648,
-      "grad_norm": 3.363278388977051,
+      "grad_norm": 4.2259650230407715,
       "learning_rate": 1.923937360178971e-05,
-      "loss": 4.1127,
+      "loss": 3.9236,
       "step": 2975
     },
     {
       "epoch": 0.6588547424275923,
-      "grad_norm": 3.2592687606811523,
+      "grad_norm": 4.067422389984131,
       "learning_rate": 1.917723092219737e-05,
-      "loss": 4.0898,
+      "loss": 3.8932,
       "step": 2980
     },
     {
       "epoch": 0.6599602034048198,
-      "grad_norm": 3.295732021331787,
+      "grad_norm": 4.111719608306885,
       "learning_rate": 1.9115088242605022e-05,
-      "loss": 4.0772,
+      "loss": 3.8891,
       "step": 2985
     },
     {
       "epoch": 0.6610656643820473,
-      "grad_norm": 3.302295684814453,
+      "grad_norm": 4.118727207183838,
       "learning_rate": 1.9052945563012678e-05,
-      "loss": 4.1688,
+      "loss": 3.9766,
       "step": 2990
     },
     {
       "epoch": 0.6621711253592748,
-      "grad_norm": 3.415590524673462,
+      "grad_norm": 4.237068176269531,
       "learning_rate": 1.8990802883420334e-05,
-      "loss": 4.0569,
+      "loss": 3.8585,
       "step": 2995
     },
     {
       "epoch": 0.6632765863365023,
-      "grad_norm": 3.4967286586761475,
+      "grad_norm": 4.3463568687438965,
       "learning_rate": 1.892866020382799e-05,
-      "loss": 4.0951,
+      "loss": 3.9038,
       "step": 3000
     },
     {
       "epoch": 0.6643820473137299,
-      "grad_norm": 3.3429524898529053,
+      "grad_norm": 4.205432891845703,
       "learning_rate": 1.8866517524235646e-05,
-      "loss": 4.0436,
+      "loss": 3.8587,
       "step": 3005
     },
     {
       "epoch": 0.6654875082909574,
-      "grad_norm": 3.2878565788269043,
+      "grad_norm": 4.115555763244629,
       "learning_rate": 1.8804374844643303e-05,
-      "loss": 4.0224,
+      "loss": 3.8332,
       "step": 3010
     },
     {
       "epoch": 0.6665929692681848,
-      "grad_norm": 3.4439568519592285,
+      "grad_norm": 4.279168128967285,
       "learning_rate": 1.874223216505096e-05,
-      "loss": 3.9529,
+      "loss": 3.7735,
       "step": 3015
     },
     {
       "epoch": 0.6676984302454123,
-      "grad_norm": 3.4221768379211426,
+      "grad_norm": 4.2163872718811035,
       "learning_rate": 1.868008948545861e-05,
-      "loss": 4.0604,
+      "loss": 3.8756,
       "step": 3020
     },
     {
       "epoch": 0.6688038912226398,
-      "grad_norm": 3.2308311462402344,
+      "grad_norm": 3.994424343109131,
       "learning_rate": 1.861794680586627e-05,
-      "loss": 4.0717,
+      "loss": 3.8942,
       "step": 3025
     },
     {
       "epoch": 0.6699093521998674,
-      "grad_norm": 3.7637572288513184,
+      "grad_norm": 4.58339786529541,
       "learning_rate": 1.8555804126273927e-05,
-      "loss": 4.3161,
+      "loss": 4.1267,
       "step": 3030
     },
     {
       "epoch": 0.6710148131770949,
-      "grad_norm": 3.2774343490600586,
+      "grad_norm": 4.057015419006348,
       "learning_rate": 1.849366144668158e-05,
-      "loss": 4.1447,
+      "loss": 3.9612,
       "step": 3035
     },
     {
       "epoch": 0.6721202741543224,
-      "grad_norm": 3.3979032039642334,
+      "grad_norm": 4.238908290863037,
       "learning_rate": 1.8431518767089236e-05,
-      "loss": 4.2971,
+      "loss": 4.1164,
       "step": 3040
     },
     {
       "epoch": 0.6732257351315498,
-      "grad_norm": 3.259497880935669,
+      "grad_norm": 4.066302299499512,
       "learning_rate": 1.8369376087496896e-05,
-      "loss": 4.2798,
+      "loss": 4.101,
       "step": 3045
     },
     {
       "epoch": 0.6743311961087773,
-      "grad_norm": 3.346216917037964,
+      "grad_norm": 4.119609355926514,
       "learning_rate": 1.8307233407904552e-05,
-      "loss": 4.2459,
+      "loss": 4.0514,
       "step": 3050
     },
     {
       "epoch": 0.6754366570860049,
-      "grad_norm": 3.195192813873291,
+      "grad_norm": 3.962416648864746,
       "learning_rate": 1.8245090728312205e-05,
-      "loss": 4.1107,
+      "loss": 3.9339,
       "step": 3055
     },
     {
       "epoch": 0.6765421180632324,
-      "grad_norm": 3.3949368000030518,
+      "grad_norm": 4.22977352142334,
       "learning_rate": 1.818294804871986e-05,
-      "loss": 4.1965,
+      "loss": 4.0103,
       "step": 3060
     },
     {
       "epoch": 0.6776475790404599,
-      "grad_norm": 3.1918063163757324,
+      "grad_norm": 3.9090285301208496,
       "learning_rate": 1.8120805369127517e-05,
-      "loss": 4.2232,
+      "loss": 4.0433,
       "step": 3065
     },
     {
       "epoch": 0.6787530400176873,
-      "grad_norm": 3.080773115158081,
+      "grad_norm": 3.774301290512085,
       "learning_rate": 1.8058662689535173e-05,
-      "loss": 4.1366,
+      "loss": 3.9531,
       "step": 3070
     },
     {
       "epoch": 0.6798585009949148,
-      "grad_norm": 3.573559284210205,
+      "grad_norm": 4.378792762756348,
       "learning_rate": 1.799652000994283e-05,
-      "loss": 4.0492,
+      "loss": 3.858,
       "step": 3075
     },
     {
       "epoch": 0.6809639619721424,
-      "grad_norm": 3.105289936065674,
+      "grad_norm": 3.832705020904541,
       "learning_rate": 1.7934377330350486e-05,
-      "loss": 4.019,
+      "loss": 3.8362,
       "step": 3080
     },
     {
       "epoch": 0.6820694229493699,
-      "grad_norm": 3.233858108520508,
+      "grad_norm": 3.976187229156494,
       "learning_rate": 1.7872234650758142e-05,
-      "loss": 4.2052,
+      "loss": 4.0236,
       "step": 3085
     },
     {
       "epoch": 0.6831748839265974,
-      "grad_norm": 3.489800214767456,
+      "grad_norm": 4.32907247543335,
       "learning_rate": 1.7810091971165798e-05,
-      "loss": 4.244,
+      "loss": 4.0618,
       "step": 3090
     },
     {
       "epoch": 0.6842803449038249,
-      "grad_norm": 3.919562339782715,
+      "grad_norm": 4.821102619171143,
       "learning_rate": 1.7747949291573454e-05,
-      "loss": 4.2778,
+      "loss": 4.0921,
       "step": 3095
     },
     {
       "epoch": 0.6853858058810524,
-      "grad_norm": 3.4953386783599854,
+      "grad_norm": 4.310387134552002,
       "learning_rate": 1.768580661198111e-05,
-      "loss": 4.0999,
+      "loss": 3.9262,
       "step": 3100
     },
     {
       "epoch": 0.6864912668582799,
-      "grad_norm": 3.0462942123413086,
+      "grad_norm": 3.790919542312622,
       "learning_rate": 1.7623663932388766e-05,
-      "loss": 4.1613,
+      "loss": 3.9805,
       "step": 3105
     },
     {
       "epoch": 0.6875967278355074,
-      "grad_norm": 3.604140520095825,
+      "grad_norm": 4.451929569244385,
       "learning_rate": 1.756152125279642e-05,
-      "loss": 4.037,
+      "loss": 3.8484,
       "step": 3110
     },
     {
       "epoch": 0.6887021888127349,
-      "grad_norm": 3.4862539768218994,
+      "grad_norm": 4.295001029968262,
       "learning_rate": 1.749937857320408e-05,
-      "loss": 4.136,
+      "loss": 3.9561,
       "step": 3115
     },
     {
       "epoch": 0.6898076497899625,
-      "grad_norm": 3.3312830924987793,
+      "grad_norm": 4.131453990936279,
       "learning_rate": 1.7437235893611735e-05,
-      "loss": 4.1436,
+      "loss": 3.9728,
       "step": 3120
     },
     {
       "epoch": 0.6909131107671899,
-      "grad_norm": 3.4092671871185303,
+      "grad_norm": 4.199705600738525,
       "learning_rate": 1.7375093214019388e-05,
-      "loss": 4.2998,
+      "loss": 4.1211,
       "step": 3125
     },
     {
       "epoch": 0.6920185717444174,
-      "grad_norm": 3.138869285583496,
+      "grad_norm": 3.898780345916748,
       "learning_rate": 1.7312950534427044e-05,
-      "loss": 3.9221,
+      "loss": 3.734,
       "step": 3130
     },
     {
       "epoch": 0.6931240327216449,
-      "grad_norm": 3.570099115371704,
+      "grad_norm": 4.383736610412598,
       "learning_rate": 1.7250807854834704e-05,
-      "loss": 4.1127,
+      "loss": 3.9447,
       "step": 3135
     },
     {
       "epoch": 0.6942294936988724,
-      "grad_norm": 3.4143168926239014,
+      "grad_norm": 4.196073532104492,
       "learning_rate": 1.7188665175242356e-05,
-      "loss": 4.1529,
+      "loss": 3.9812,
       "step": 3140
     },
     {
       "epoch": 0.6953349546761,
-      "grad_norm": 3.299022674560547,
+      "grad_norm": 4.070743083953857,
       "learning_rate": 1.7126522495650012e-05,
-      "loss": 4.094,
+      "loss": 3.9251,
       "step": 3145
     },
     {
       "epoch": 0.6964404156533275,
-      "grad_norm": 3.2752246856689453,
+      "grad_norm": 4.021485328674316,
       "learning_rate": 1.706437981605767e-05,
-      "loss": 4.0729,
+      "loss": 3.8985,
       "step": 3150
     },
     {
       "epoch": 0.6975458766305549,
-      "grad_norm": 3.453444004058838,
+      "grad_norm": 4.298070907592773,
       "learning_rate": 1.7002237136465328e-05,
-      "loss": 4.1417,
+      "loss": 3.9616,
       "step": 3155
     },
     {
       "epoch": 0.6986513376077824,
-      "grad_norm": 3.2120327949523926,
+      "grad_norm": 3.9657671451568604,
       "learning_rate": 1.694009445687298e-05,
-      "loss": 4.249,
+      "loss": 4.0888,
       "step": 3160
     },
     {
       "epoch": 0.6997567985850099,
-      "grad_norm": 3.4823880195617676,
+      "grad_norm": 4.285989761352539,
       "learning_rate": 1.6877951777280637e-05,
-      "loss": 4.197,
+      "loss": 4.0373,
       "step": 3165
     },
     {
       "epoch": 0.7008622595622375,
-      "grad_norm": 3.438119888305664,
+      "grad_norm": 4.256559371948242,
       "learning_rate": 1.6815809097688293e-05,
-      "loss": 4.1066,
+      "loss": 3.9443,
       "step": 3170
     },
     {
       "epoch": 0.701967720539465,
-      "grad_norm": 3.4621167182922363,
+      "grad_norm": 4.2700347900390625,
       "learning_rate": 1.675366641809595e-05,
-      "loss": 4.2766,
+      "loss": 4.1177,
       "step": 3175
     },
     {
       "epoch": 0.7030731815166925,
-      "grad_norm": 3.3527414798736572,
+      "grad_norm": 4.114231109619141,
       "learning_rate": 1.6691523738503606e-05,
-      "loss": 4.0086,
+      "loss": 3.8346,
       "step": 3180
     },
     {
       "epoch": 0.7041786424939199,
-      "grad_norm": 3.4415431022644043,
+      "grad_norm": 4.15029239654541,
       "learning_rate": 1.6629381058911262e-05,
-      "loss": 4.0336,
+      "loss": 3.8636,
       "step": 3185
     },
     {
       "epoch": 0.7052841034711474,
-      "grad_norm": 3.243367910385132,
+      "grad_norm": 4.081049919128418,
       "learning_rate": 1.6567238379318918e-05,
-      "loss": 4.1119,
+      "loss": 3.9365,
       "step": 3190
     },
     {
       "epoch": 0.706389564448375,
-      "grad_norm": 3.515403985977173,
+      "grad_norm": 4.327289581298828,
       "learning_rate": 1.650509569972657e-05,
-      "loss": 4.0391,
+      "loss": 3.8763,
       "step": 3195
     },
     {
       "epoch": 0.7074950254256025,
-      "grad_norm": 3.0629870891571045,
+      "grad_norm": 3.719695568084717,
       "learning_rate": 1.644295302013423e-05,
-      "loss": 4.2706,
+      "loss": 4.1082,
       "step": 3200
     },
     {
       "epoch": 0.70860048640283,
-      "grad_norm": 3.412379026412964,
+      "grad_norm": 4.164029121398926,
       "learning_rate": 1.6380810340541887e-05,
-      "loss": 4.3555,
+      "loss": 4.1869,
       "step": 3205
     },
     {
       "epoch": 0.7097059473800574,
-      "grad_norm": 3.250455141067505,
+      "grad_norm": 3.9678902626037598,
       "learning_rate": 1.631866766094954e-05,
-      "loss": 4.1877,
+      "loss": 4.023,
       "step": 3210
     },
     {
       "epoch": 0.710811408357285,
-      "grad_norm": 3.0698251724243164,
+      "grad_norm": 3.757974863052368,
       "learning_rate": 1.6256524981357195e-05,
-      "loss": 4.0128,
+      "loss": 3.8398,
       "step": 3215
     },
     {
       "epoch": 0.7119168693345125,
-      "grad_norm": 3.3195056915283203,
+      "grad_norm": 4.070690155029297,
       "learning_rate": 1.619438230176485e-05,
-      "loss": 4.2022,
+      "loss": 4.0439,
       "step": 3220
     },
     {
       "epoch": 0.71302233031174,
-      "grad_norm": 3.3622958660125732,
+      "grad_norm": 4.125089168548584,
       "learning_rate": 1.613223962217251e-05,
-      "loss": 4.187,
+      "loss": 4.0254,
       "step": 3225
     },
     {
       "epoch": 0.7141277912889675,
-      "grad_norm": 3.3840930461883545,
+      "grad_norm": 4.153202533721924,
       "learning_rate": 1.6070096942580164e-05,
-      "loss": 4.2928,
+      "loss": 4.1229,
       "step": 3230
     },
     {
       "epoch": 0.715233252266195,
-      "grad_norm": 3.4330742359161377,
+      "grad_norm": 4.196191310882568,
       "learning_rate": 1.600795426298782e-05,
-      "loss": 4.0841,
+      "loss": 3.9262,
       "step": 3235
     },
     {
       "epoch": 0.7163387132434225,
-      "grad_norm": 3.258180856704712,
+      "grad_norm": 3.993074417114258,
       "learning_rate": 1.5945811583395476e-05,
-      "loss": 4.1938,
+      "loss": 4.0286,
       "step": 3240
     },
     {
       "epoch": 0.71744417422065,
-      "grad_norm": 3.183001756668091,
+      "grad_norm": 3.895026206970215,
       "learning_rate": 1.5883668903803133e-05,
-      "loss": 4.078,
+      "loss": 3.9188,
       "step": 3245
     },
     {
       "epoch": 0.7185496351978775,
-      "grad_norm": 3.0564966201782227,
+      "grad_norm": 3.7398178577423096,
       "learning_rate": 1.582152622421079e-05,
-      "loss": 4.089,
+      "loss": 3.9278,
       "step": 3250
     },
     {
       "epoch": 0.719655096175105,
-      "grad_norm": 3.324143648147583,
+      "grad_norm": 4.117962837219238,
       "learning_rate": 1.5759383544618445e-05,
-      "loss": 4.2551,
+      "loss": 4.0958,
       "step": 3255
     },
     {
       "epoch": 0.7207605571523326,
-      "grad_norm": 3.4312210083007812,
+      "grad_norm": 4.1803083419799805,
       "learning_rate": 1.56972408650261e-05,
-      "loss": 4.1726,
+      "loss": 4.016,
       "step": 3260
     },
     {
       "epoch": 0.72186601812956,
-      "grad_norm": 3.168652057647705,
+      "grad_norm": 3.8384809494018555,
       "learning_rate": 1.5635098185433757e-05,
-      "loss": 4.0236,
+      "loss": 3.858,
       "step": 3265
     },
     {
       "epoch": 0.7229714791067875,
-      "grad_norm": 3.116694211959839,
+      "grad_norm": 3.79559326171875,
       "learning_rate": 1.5572955505841413e-05,
-      "loss": 4.2022,
+      "loss": 4.0283,
       "step": 3270
     },
     {
       "epoch": 0.724076940084015,
-      "grad_norm": 3.235372543334961,
+      "grad_norm": 3.969823122024536,
       "learning_rate": 1.551081282624907e-05,
-      "loss": 3.8518,
+      "loss": 3.6853,
       "step": 3275
     },
     {
       "epoch": 0.7251824010612425,
-      "grad_norm": 3.3609163761138916,
+      "grad_norm": 4.1237287521362305,
       "learning_rate": 1.5448670146656726e-05,
-      "loss": 3.968,
+      "loss": 3.8186,
       "step": 3280
     },
     {
       "epoch": 0.7262878620384701,
-      "grad_norm": 3.4579970836639404,
+      "grad_norm": 4.220582962036133,
       "learning_rate": 1.538652746706438e-05,
-      "loss": 4.212,
+      "loss": 4.043,
       "step": 3285
     },
     {
       "epoch": 0.7273933230156976,
-      "grad_norm": 3.582771062850952,
+      "grad_norm": 4.410422325134277,
       "learning_rate": 1.5324384787472038e-05,
-      "loss": 4.2005,
+      "loss": 4.044,
       "step": 3290
     },
     {
       "epoch": 0.728498783992925,
-      "grad_norm": 3.151522636413574,
+      "grad_norm": 3.880732536315918,
       "learning_rate": 1.5262242107879694e-05,
-      "loss": 4.0769,
+      "loss": 3.9236,
       "step": 3295
     },
     {
       "epoch": 0.7296042449701525,
-      "grad_norm": 3.194068193435669,
+      "grad_norm": 3.9242305755615234,
       "learning_rate": 1.5200099428287349e-05,
-      "loss": 4.2329,
+      "loss": 4.0774,
       "step": 3300
     },
     {
       "epoch": 0.73070970594738,
-      "grad_norm": 3.24617600440979,
+      "grad_norm": 3.967616081237793,
       "learning_rate": 1.5137956748695003e-05,
-      "loss": 4.0845,
+      "loss": 3.9249,
       "step": 3305
     },
     {
       "epoch": 0.7318151669246076,
-      "grad_norm": 3.347874641418457,
+      "grad_norm": 4.075644016265869,
       "learning_rate": 1.5075814069102661e-05,
-      "loss": 4.2557,
+      "loss": 4.1028,
       "step": 3310
     },
     {
       "epoch": 0.7329206279018351,
-      "grad_norm": 3.392652988433838,
+      "grad_norm": 4.196295261383057,
       "learning_rate": 1.5013671389510317e-05,
-      "loss": 4.1908,
+      "loss": 4.0395,
       "step": 3315
     },
     {
       "epoch": 0.7340260888790626,
-      "grad_norm": 3.364522933959961,
+      "grad_norm": 4.08765172958374,
       "learning_rate": 1.4951528709917972e-05,
-      "loss": 4.1181,
+      "loss": 3.9575,
       "step": 3320
     },
     {
       "epoch": 0.73513154985629,
-      "grad_norm": 3.217658042907715,
+      "grad_norm": 3.984494209289551,
       "learning_rate": 1.4889386030325628e-05,
-      "loss": 4.1129,
+      "loss": 3.961,
       "step": 3325
     },
     {
       "epoch": 0.7362370108335176,
-      "grad_norm": 3.741403102874756,
+      "grad_norm": 4.500490665435791,
       "learning_rate": 1.4827243350733282e-05,
-      "loss": 4.1941,
+      "loss": 4.0449,
       "step": 3330
     },
     {
       "epoch": 0.7373424718107451,
-      "grad_norm": 3.6244940757751465,
+      "grad_norm": 4.385171890258789,
       "learning_rate": 1.4765100671140942e-05,
-      "loss": 4.133,
+      "loss": 3.9644,
       "step": 3335
     },
     {
       "epoch": 0.7384479327879726,
-      "grad_norm": 3.455331563949585,
+      "grad_norm": 4.19306755065918,
       "learning_rate": 1.4702957991548596e-05,
-      "loss": 4.1993,
+      "loss": 4.055,
       "step": 3340
     },
     {
       "epoch": 0.7395533937652001,
-      "grad_norm": 3.3067119121551514,
+      "grad_norm": 4.029420375823975,
       "learning_rate": 1.4640815311956253e-05,
-      "loss": 4.1962,
+      "loss": 4.0404,
       "step": 3345
     },
     {
       "epoch": 0.7406588547424275,
-      "grad_norm": 3.3184375762939453,
+      "grad_norm": 4.044227600097656,
       "learning_rate": 1.4578672632363907e-05,
-      "loss": 4.0779,
+      "loss": 3.9228,
       "step": 3350
     },
     {
       "epoch": 0.7417643157196551,
-      "grad_norm": 3.617077350616455,
+      "grad_norm": 4.44707727432251,
       "learning_rate": 1.4516529952771565e-05,
-      "loss": 3.995,
+      "loss": 3.8386,
       "step": 3355
     },
     {
       "epoch": 0.7428697766968826,
-      "grad_norm": 3.471519947052002,
+      "grad_norm": 4.240484237670898,
       "learning_rate": 1.4454387273179221e-05,
-      "loss": 4.0302,
+      "loss": 3.8871,
       "step": 3360
     },
     {
       "epoch": 0.7439752376741101,
-      "grad_norm": 3.3337936401367188,
+      "grad_norm": 4.035573482513428,
       "learning_rate": 1.4392244593586876e-05,
-      "loss": 4.1125,
+      "loss": 3.9473,
       "step": 3365
     },
     {
       "epoch": 0.7450806986513376,
-      "grad_norm": 3.5475218296051025,
+      "grad_norm": 4.32297420501709,
       "learning_rate": 1.4330101913994532e-05,
-      "loss": 4.1158,
+      "loss": 3.9613,
       "step": 3370
     },
     {
       "epoch": 0.7461861596285652,
-      "grad_norm": 3.225281238555908,
+      "grad_norm": 3.930088996887207,
       "learning_rate": 1.4267959234402186e-05,
-      "loss": 4.1048,
+      "loss": 3.9554,
       "step": 3375
     },
     {
       "epoch": 0.7472916206057926,
-      "grad_norm": 2.9788243770599365,
+      "grad_norm": 3.607778549194336,
       "learning_rate": 1.4205816554809844e-05,
-      "loss": 4.1919,
+      "loss": 4.0278,
       "step": 3380
     },
     {
       "epoch": 0.7483970815830201,
-      "grad_norm": 2.9584922790527344,
+      "grad_norm": 3.6034600734710693,
       "learning_rate": 1.41436738752175e-05,
-      "loss": 3.9252,
+      "loss": 3.7776,
       "step": 3385
     },
     {
       "epoch": 0.7495025425602476,
-      "grad_norm": 3.4342474937438965,
+      "grad_norm": 4.144221782684326,
       "learning_rate": 1.4081531195625155e-05,
-      "loss": 4.2655,
+      "loss": 4.1134,
       "step": 3390
     },
     {
       "epoch": 0.7506080035374751,
-      "grad_norm": 3.157142400741577,
+      "grad_norm": 3.8719232082366943,
       "learning_rate": 1.4019388516032811e-05,
-      "loss": 4.0619,
+      "loss": 3.916,
       "step": 3395
     },
     {
       "epoch": 0.7517134645147027,
-      "grad_norm": 3.739959716796875,
+      "grad_norm": 4.523682117462158,
       "learning_rate": 1.3957245836440469e-05,
-      "loss": 4.1531,
+      "loss": 3.9996,
       "step": 3400
     },
     {
       "epoch": 0.7528189254919301,
-      "grad_norm": 3.4141812324523926,
+      "grad_norm": 4.147566795349121,
       "learning_rate": 1.3895103156848125e-05,
-      "loss": 4.0972,
+      "loss": 3.9515,
       "step": 3405
     },
     {
       "epoch": 0.7539243864691576,
-      "grad_norm": 3.140306234359741,
+      "grad_norm": 3.8327906131744385,
       "learning_rate": 1.383296047725578e-05,
-      "loss": 4.1615,
+      "loss": 4.0151,
       "step": 3410
     },
     {
       "epoch": 0.7550298474463851,
-      "grad_norm": 3.495731830596924,
+      "grad_norm": 4.266075134277344,
       "learning_rate": 1.3770817797663436e-05,
-      "loss": 4.322,
+      "loss": 4.1809,
       "step": 3415
     },
     {
       "epoch": 0.7561353084236127,
-      "grad_norm": 3.2486352920532227,
+      "grad_norm": 3.9193296432495117,
       "learning_rate": 1.3708675118071093e-05,
-      "loss": 4.1291,
+      "loss": 3.9779,
       "step": 3420
     },
     {
       "epoch": 0.7572407694008402,
-      "grad_norm": 3.405538320541382,
+      "grad_norm": 4.139795780181885,
       "learning_rate": 1.3646532438478748e-05,
-      "loss": 4.0567,
+      "loss": 3.9008,
       "step": 3425
     },
     {
       "epoch": 0.7583462303780677,
-      "grad_norm": 3.2491066455841064,
+      "grad_norm": 3.917721748352051,
       "learning_rate": 1.3584389758886404e-05,
-      "loss": 4.2248,
+      "loss": 4.086,
       "step": 3430
     },
     {
       "epoch": 0.7594516913552951,
-      "grad_norm": 3.415019989013672,
+      "grad_norm": 4.180150985717773,
       "learning_rate": 1.3522247079294059e-05,
-      "loss": 4.2429,
+      "loss": 4.0858,
       "step": 3435
     },
     {
       "epoch": 0.7605571523325226,
-      "grad_norm": 3.0789833068847656,
+      "grad_norm": 3.753077745437622,
       "learning_rate": 1.3460104399701715e-05,
-      "loss": 4.0823,
+      "loss": 3.9323,
       "step": 3440
     },
     {
       "epoch": 0.7616626133097502,
-      "grad_norm": 3.2663156986236572,
+      "grad_norm": 3.9792933464050293,
       "learning_rate": 1.3397961720109373e-05,
-      "loss": 4.1684,
+      "loss": 4.0261,
       "step": 3445
     },
     {
       "epoch": 0.7627680742869777,
-      "grad_norm": 3.3702750205993652,
+      "grad_norm": 4.080856800079346,
       "learning_rate": 1.3335819040517029e-05,
-      "loss": 4.1521,
+      "loss": 4.0075,
       "step": 3450
     },
     {
       "epoch": 0.7638735352642052,
-      "grad_norm": 3.318516731262207,
+      "grad_norm": 4.029576301574707,
       "learning_rate": 1.3273676360924683e-05,
-      "loss": 4.0572,
+      "loss": 3.9114,
       "step": 3455
     },
     {
       "epoch": 0.7649789962414327,
-      "grad_norm": 3.307229995727539,
+      "grad_norm": 4.0104756355285645,
       "learning_rate": 1.321153368133234e-05,
-      "loss": 4.2087,
+      "loss": 4.0611,
       "step": 3460
     },
     {
       "epoch": 0.7660844572186601,
-      "grad_norm": 3.141308546066284,
+      "grad_norm": 3.814962863922119,
       "learning_rate": 1.3149391001739997e-05,
-      "loss": 4.2045,
+      "loss": 4.0684,
       "step": 3465
     },
     {
       "epoch": 0.7671899181958877,
-      "grad_norm": 3.488524913787842,
+      "grad_norm": 4.239627838134766,
       "learning_rate": 1.3087248322147652e-05,
-      "loss": 4.1981,
+      "loss": 4.0581,
       "step": 3470
     },
     {
       "epoch": 0.7682953791731152,
-      "grad_norm": 3.333773612976074,
+      "grad_norm": 4.017524242401123,
       "learning_rate": 1.3025105642555308e-05,
-      "loss": 4.0546,
+      "loss": 3.9055,
       "step": 3475
     },
     {
       "epoch": 0.7694008401503427,
-      "grad_norm": 3.093600273132324,
+      "grad_norm": 3.7963919639587402,
       "learning_rate": 1.2962962962962962e-05,
-      "loss": 4.1297,
+      "loss": 3.9876,
       "step": 3480
     },
     {
       "epoch": 0.7705063011275702,
-      "grad_norm": 3.681091547012329,
+      "grad_norm": 4.4867424964904785,
       "learning_rate": 1.2900820283370619e-05,
-      "loss": 4.2743,
+      "loss": 4.1333,
       "step": 3485
     },
     {
       "epoch": 0.7716117621047976,
-      "grad_norm": 3.2113373279571533,
+      "grad_norm": 3.909165382385254,
       "learning_rate": 1.2838677603778276e-05,
-      "loss": 4.1716,
+      "loss": 4.0293,
       "step": 3490
     },
     {
       "epoch": 0.7727172230820252,
-      "grad_norm": 3.22847843170166,
+      "grad_norm": 3.8970353603363037,
       "learning_rate": 1.2776534924185931e-05,
-      "loss": 4.1038,
+      "loss": 3.9646,
       "step": 3495
     },
     {
       "epoch": 0.7738226840592527,
-      "grad_norm": 3.2960784435272217,
+      "grad_norm": 4.00996208190918,
       "learning_rate": 1.2714392244593587e-05,
-      "loss": 4.2599,
+      "loss": 4.1201,
       "step": 3500
     },
     {
       "epoch": 0.7749281450364802,
-      "grad_norm": 3.509111166000366,
+      "grad_norm": 4.233725070953369,
       "learning_rate": 1.2652249565001242e-05,
-      "loss": 4.2696,
+      "loss": 4.1307,
       "step": 3505
     },
     {
       "epoch": 0.7760336060137077,
-      "grad_norm": 3.4601404666900635,
+      "grad_norm": 4.174342632293701,
       "learning_rate": 1.2590106885408901e-05,
-      "loss": 4.0995,
+      "loss": 3.9678,
       "step": 3510
     },
     {
       "epoch": 0.7771390669909353,
-      "grad_norm": 3.166656017303467,
+      "grad_norm": 3.843369245529175,
       "learning_rate": 1.2527964205816556e-05,
-      "loss": 4.3323,
+      "loss": 4.2019,
       "step": 3515
     },
     {
       "epoch": 0.7782445279681627,
-      "grad_norm": 3.115483522415161,
+      "grad_norm": 3.7710518836975098,
       "learning_rate": 1.2465821526224212e-05,
-      "loss": 4.2784,
+      "loss": 4.1486,
       "step": 3520
     },
     {
       "epoch": 0.7793499889453902,
-      "grad_norm": 3.377978563308716,
+      "grad_norm": 4.084702491760254,
       "learning_rate": 1.2403678846631868e-05,
-      "loss": 4.1576,
+      "loss": 4.0051,
       "step": 3525
     },
     {
       "epoch": 0.7804554499226177,
-      "grad_norm": 3.291743278503418,
+      "grad_norm": 3.9926633834838867,
       "learning_rate": 1.2341536167039522e-05,
-      "loss": 4.3317,
+      "loss": 4.1913,
       "step": 3530
     },
     {
       "epoch": 0.7815609108998453,
-      "grad_norm": 3.091101884841919,
+      "grad_norm": 3.7536752223968506,
       "learning_rate": 1.227939348744718e-05,
-      "loss": 4.2178,
+      "loss": 4.0825,
       "step": 3535
     },
     {
       "epoch": 0.7826663718770728,
-      "grad_norm": 3.3874189853668213,
+      "grad_norm": 4.098823070526123,
       "learning_rate": 1.2217250807854835e-05,
-      "loss": 4.0266,
+      "loss": 3.8798,
       "step": 3540
     },
     {
       "epoch": 0.7837718328543002,
-      "grad_norm": 3.4406089782714844,
+      "grad_norm": 4.119952201843262,
       "learning_rate": 1.2155108128262491e-05,
-      "loss": 4.2243,
+      "loss": 4.0793,
       "step": 3545
     },
     {
       "epoch": 0.7848772938315277,
-      "grad_norm": 3.2707858085632324,
+      "grad_norm": 3.9421002864837646,
       "learning_rate": 1.2092965448670147e-05,
-      "loss": 4.1444,
+      "loss": 4.0086,
       "step": 3550
     },
     {
       "epoch": 0.7859827548087552,
-      "grad_norm": 3.2035396099090576,
+      "grad_norm": 3.87634015083313,
       "learning_rate": 1.2030822769077803e-05,
-      "loss": 3.981,
+      "loss": 3.8241,
       "step": 3555
     },
     {
       "epoch": 0.7870882157859828,
-      "grad_norm": 3.3851969242095947,
+      "grad_norm": 4.106171607971191,
       "learning_rate": 1.196868008948546e-05,
-      "loss": 4.219,
+      "loss": 4.0937,
       "step": 3560
     },
     {
       "epoch": 0.7881936767632103,
-      "grad_norm": 3.0952658653259277,
+      "grad_norm": 3.774790048599243,
       "learning_rate": 1.1906537409893114e-05,
-      "loss": 4.2355,
+      "loss": 4.11,
       "step": 3565
     },
     {
       "epoch": 0.7892991377404378,
-      "grad_norm": 3.3667149543762207,
+      "grad_norm": 4.0692458152771,
       "learning_rate": 1.1844394730300772e-05,
-      "loss": 4.2494,
+      "loss": 4.1116,
       "step": 3570
     },
     {
       "epoch": 0.7904045987176652,
-      "grad_norm": 3.6815719604492188,
+      "grad_norm": 4.495890140533447,
       "learning_rate": 1.1782252050708426e-05,
-      "loss": 4.15,
+      "loss": 4.0197,
       "step": 3575
     },
     {
       "epoch": 0.7915100596948927,
-      "grad_norm": 3.330397367477417,
+      "grad_norm": 4.008393287658691,
       "learning_rate": 1.1720109371116084e-05,
-      "loss": 4.0933,
+      "loss": 3.9503,
       "step": 3580
     },
     {
       "epoch": 0.7926155206721203,
-      "grad_norm": 3.213534355163574,
+      "grad_norm": 3.8918848037719727,
       "learning_rate": 1.1657966691523739e-05,
-      "loss": 4.0645,
+      "loss": 3.933,
       "step": 3585
     },
     {
       "epoch": 0.7937209816493478,
-      "grad_norm": 3.413196086883545,
+      "grad_norm": 4.125125408172607,
       "learning_rate": 1.1595824011931397e-05,
-      "loss": 4.2731,
+      "loss": 4.138,
       "step": 3590
     },
     {
       "epoch": 0.7948264426265753,
-      "grad_norm": 2.9504334926605225,
+      "grad_norm": 3.6134209632873535,
       "learning_rate": 1.1533681332339051e-05,
-      "loss": 4.0869,
+      "loss": 3.9578,
       "step": 3595
     },
     {
       "epoch": 0.7959319036038028,
-      "grad_norm": 3.48688006401062,
+      "grad_norm": 4.231503963470459,
       "learning_rate": 1.1471538652746707e-05,
-      "loss": 4.1732,
+      "loss": 4.0408,
       "step": 3600
     },
     {
       "epoch": 0.7970373645810302,
-      "grad_norm": 3.202857494354248,
+      "grad_norm": 3.8687455654144287,
       "learning_rate": 1.1409395973154363e-05,
-      "loss": 4.2084,
+      "loss": 4.0806,
       "step": 3605
     },
     {
       "epoch": 0.7981428255582578,
-      "grad_norm": 3.460794687271118,
+      "grad_norm": 4.175510883331299,
       "learning_rate": 1.1347253293562018e-05,
-      "loss": 4.2956,
+      "loss": 4.1758,
       "step": 3610
     },
     {
       "epoch": 0.7992482865354853,
-      "grad_norm": 3.3727447986602783,
+      "grad_norm": 4.044642448425293,
       "learning_rate": 1.1285110613969676e-05,
-      "loss": 4.1854,
+      "loss": 4.0496,
       "step": 3615
     },
     {
       "epoch": 0.8003537475127128,
-      "grad_norm": 3.3435420989990234,
+      "grad_norm": 4.010993957519531,
       "learning_rate": 1.122296793437733e-05,
-      "loss": 4.3749,
+      "loss": 4.2476,
       "step": 3620
     },
     {
       "epoch": 0.8014592084899403,
-      "grad_norm": 3.1651086807250977,
+      "grad_norm": 3.8528947830200195,
       "learning_rate": 1.1160825254784988e-05,
-      "loss": 4.132,
+      "loss": 4.0066,
       "step": 3625
     },
     {
       "epoch": 0.8025646694671678,
-      "grad_norm": 3.482461929321289,
+      "grad_norm": 4.169907569885254,
       "learning_rate": 1.1098682575192643e-05,
-      "loss": 4.3037,
+      "loss": 4.1723,
       "step": 3630
     },
     {
       "epoch": 0.8036701304443953,
-      "grad_norm": 3.5828919410705566,
+      "grad_norm": 4.301722526550293,
       "learning_rate": 1.1036539895600299e-05,
-      "loss": 4.1466,
+      "loss": 4.0186,
       "step": 3635
     },
     {
       "epoch": 0.8047755914216228,
-      "grad_norm": 3.344888687133789,
+      "grad_norm": 4.000545024871826,
       "learning_rate": 1.0974397216007955e-05,
-      "loss": 4.1947,
+      "loss": 4.0701,
       "step": 3640
     },
     {
       "epoch": 0.8058810523988503,
-      "grad_norm": 3.2426233291625977,
+      "grad_norm": 3.9018445014953613,
       "learning_rate": 1.091225453641561e-05,
-      "loss": 4.0683,
+      "loss": 3.9381,
       "step": 3645
     },
     {
       "epoch": 0.8069865133760779,
-      "grad_norm": 3.2281033992767334,
+      "grad_norm": 3.793440818786621,
       "learning_rate": 1.0850111856823267e-05,
-      "loss": 4.107,
+      "loss": 3.9743,
       "step": 3650
     },
     {
       "epoch": 0.8080919743533054,
-      "grad_norm": 3.1622958183288574,
+      "grad_norm": 3.7972564697265625,
       "learning_rate": 1.0787969177230922e-05,
-      "loss": 4.2085,
+      "loss": 4.0857,
       "step": 3655
     },
     {
       "epoch": 0.8091974353305328,
-      "grad_norm": 3.2309300899505615,
+      "grad_norm": 3.897294282913208,
       "learning_rate": 1.072582649763858e-05,
-      "loss": 4.3067,
+      "loss": 4.1872,
       "step": 3660
     },
     {
       "epoch": 0.8103028963077603,
-      "grad_norm": 3.1198458671569824,
+      "grad_norm": 3.738558769226074,
       "learning_rate": 1.0663683818046234e-05,
-      "loss": 4.0849,
+      "loss": 3.9514,
       "step": 3665
     },
     {
       "epoch": 0.8114083572849878,
-      "grad_norm": 3.5155203342437744,
+      "grad_norm": 4.222847938537598,
       "learning_rate": 1.060154113845389e-05,
-      "loss": 4.115,
+      "loss": 3.9862,
       "step": 3670
     },
     {
       "epoch": 0.8125138182622154,
-      "grad_norm": 3.102889060974121,
+      "grad_norm": 3.754302501678467,
       "learning_rate": 1.0539398458861546e-05,
-      "loss": 4.1175,
+      "loss": 3.9915,
       "step": 3675
     },
     {
       "epoch": 0.8136192792394429,
-      "grad_norm": 3.3019254207611084,
+      "grad_norm": 3.977128267288208,
       "learning_rate": 1.0477255779269203e-05,
-      "loss": 4.2803,
+      "loss": 4.1586,
       "step": 3680
     },
     {
       "epoch": 0.8147247402166704,
-      "grad_norm": 3.5849218368530273,
+      "grad_norm": 4.319126129150391,
       "learning_rate": 1.0415113099676859e-05,
-      "loss": 4.2005,
+      "loss": 4.0769,
       "step": 3685
     },
     {
       "epoch": 0.8158302011938978,
-      "grad_norm": 3.9152631759643555,
+      "grad_norm": 4.652560710906982,
       "learning_rate": 1.0352970420084515e-05,
-      "loss": 4.3163,
+      "loss": 4.1939,
       "step": 3690
     },
     {
       "epoch": 0.8169356621711253,
-      "grad_norm": 3.0798897743225098,
+      "grad_norm": 3.692349433898926,
       "learning_rate": 1.0290827740492171e-05,
-      "loss": 4.1536,
+      "loss": 4.0318,
       "step": 3695
     },
     {
       "epoch": 0.8180411231483529,
-      "grad_norm": 3.491821765899658,
+      "grad_norm": 4.215076923370361,
       "learning_rate": 1.0228685060899826e-05,
-      "loss": 4.3108,
+      "loss": 4.1897,
       "step": 3700
     },
     {
       "epoch": 0.8191465841255804,
-      "grad_norm": 3.093750238418579,
+      "grad_norm": 3.725860357284546,
       "learning_rate": 1.0166542381307482e-05,
-      "loss": 4.0467,
+      "loss": 3.9274,
       "step": 3705
     },
     {
       "epoch": 0.8202520451028079,
-      "grad_norm": 3.4779791831970215,
+      "grad_norm": 4.188419818878174,
       "learning_rate": 1.0104399701715138e-05,
-      "loss": 4.2484,
+      "loss": 4.1327,
       "step": 3710
     },
     {
       "epoch": 0.8213575060800353,
-      "grad_norm": 3.1915061473846436,
+      "grad_norm": 3.8532774448394775,
       "learning_rate": 1.0042257022122794e-05,
-      "loss": 4.3235,
+      "loss": 4.2108,
       "step": 3715
     },
     {
       "epoch": 0.8224629670572629,
-      "grad_norm": 3.1019785404205322,
+      "grad_norm": 3.7073981761932373,
       "learning_rate": 9.98011434253045e-06,
-      "loss": 4.1893,
+      "loss": 4.0719,
       "step": 3720
     },
     {
       "epoch": 0.8235684280344904,
-      "grad_norm": 3.3659591674804688,
+      "grad_norm": 4.034220218658447,
       "learning_rate": 9.917971662938106e-06,
-      "loss": 4.1759,
+      "loss": 4.0534,
       "step": 3725
     },
     {
       "epoch": 0.8246738890117179,
-      "grad_norm": 3.254364013671875,
+      "grad_norm": 3.895068645477295,
       "learning_rate": 9.855828983345763e-06,
-      "loss": 3.9382,
+      "loss": 3.8093,
       "step": 3730
     },
     {
       "epoch": 0.8257793499889454,
-      "grad_norm": 3.1901118755340576,
+      "grad_norm": 3.8226332664489746,
       "learning_rate": 9.793686303753419e-06,
-      "loss": 4.1601,
+      "loss": 4.0409,
       "step": 3735
     },
     {
       "epoch": 0.826884810966173,
-      "grad_norm": 3.040501832962036,
+      "grad_norm": 3.644611120223999,
       "learning_rate": 9.731543624161075e-06,
-      "loss": 4.0918,
+      "loss": 3.9602,
       "step": 3740
     },
     {
       "epoch": 0.8279902719434004,
-      "grad_norm": 3.3288450241088867,
+      "grad_norm": 3.9775216579437256,
       "learning_rate": 9.669400944568731e-06,
-      "loss": 4.1557,
+      "loss": 4.0428,
       "step": 3745
     },
     {
       "epoch": 0.8290957329206279,
-      "grad_norm": 3.145031213760376,
+      "grad_norm": 3.7982921600341797,
       "learning_rate": 9.607258264976386e-06,
-      "loss": 4.2639,
+      "loss": 4.1449,
       "step": 3750
     },
     {
       "epoch": 0.8302011938978554,
-      "grad_norm": 2.950425148010254,
+      "grad_norm": 3.5252296924591064,
       "learning_rate": 9.545115585384042e-06,
-      "loss": 4.0413,
+      "loss": 3.9183,
       "step": 3755
     },
     {
       "epoch": 0.8313066548750829,
-      "grad_norm": 3.336622714996338,
+      "grad_norm": 3.977123498916626,
       "learning_rate": 9.482972905791698e-06,
-      "loss": 4.2885,
+      "loss": 4.1698,
       "step": 3760
     },
     {
       "epoch": 0.8324121158523105,
-      "grad_norm": 3.403669834136963,
+      "grad_norm": 4.0719313621521,
       "learning_rate": 9.420830226199354e-06,
-      "loss": 4.224,
+      "loss": 4.1118,
       "step": 3765
     },
     {
       "epoch": 0.8335175768295379,
-      "grad_norm": 3.3747620582580566,
+      "grad_norm": 4.03733491897583,
       "learning_rate": 9.35868754660701e-06,
-      "loss": 4.1419,
+      "loss": 4.0254,
       "step": 3770
     },
     {
       "epoch": 0.8346230378067654,
-      "grad_norm": 3.3672516345977783,
+      "grad_norm": 4.0274786949157715,
       "learning_rate": 9.296544867014666e-06,
-      "loss": 4.2408,
+      "loss": 4.1203,
       "step": 3775
     },
     {
       "epoch": 0.8357284987839929,
-      "grad_norm": 3.1235463619232178,
+      "grad_norm": 3.746434450149536,
       "learning_rate": 9.234402187422323e-06,
-      "loss": 4.2304,
+      "loss": 4.1212,
       "step": 3780
     },
     {
       "epoch": 0.8368339597612204,
-      "grad_norm": 3.0135231018066406,
+      "grad_norm": 3.623635768890381,
       "learning_rate": 9.172259507829977e-06,
-      "loss": 4.3504,
+      "loss": 4.2391,
       "step": 3785
     },
     {
       "epoch": 0.837939420738448,
-      "grad_norm": 3.669422149658203,
+      "grad_norm": 4.408978462219238,
       "learning_rate": 9.110116828237635e-06,
-      "loss": 4.2286,
+      "loss": 4.1202,
       "step": 3790
     },
     {
       "epoch": 0.8390448817156755,
-      "grad_norm": 3.5061023235321045,
+      "grad_norm": 4.174715042114258,
       "learning_rate": 9.04797414864529e-06,
-      "loss": 4.043,
+      "loss": 3.9245,
       "step": 3795
     },
     {
       "epoch": 0.8401503426929029,
-      "grad_norm": 3.188978672027588,
+      "grad_norm": 3.8455190658569336,
       "learning_rate": 8.985831469052947e-06,
-      "loss": 4.2602,
+      "loss": 4.1507,
       "step": 3800
     },
     {
       "epoch": 0.8412558036701304,
-      "grad_norm": 3.4181642532348633,
+      "grad_norm": 4.084865093231201,
       "learning_rate": 8.923688789460602e-06,
-      "loss": 4.1946,
+      "loss": 4.0874,
       "step": 3805
     },
     {
       "epoch": 0.8423612646473579,
-      "grad_norm": 3.3051459789276123,
+      "grad_norm": 3.9332449436187744,
       "learning_rate": 8.861546109868258e-06,
-      "loss": 4.1812,
+      "loss": 4.0682,
       "step": 3810
     },
     {
       "epoch": 0.8434667256245855,
-      "grad_norm": 3.0405430793762207,
+      "grad_norm": 3.6512081623077393,
       "learning_rate": 8.799403430275914e-06,
-      "loss": 4.2455,
+      "loss": 4.1347,
       "step": 3815
     },
     {
       "epoch": 0.844572186601813,
-      "grad_norm": 3.1977388858795166,
+      "grad_norm": 3.837625741958618,
       "learning_rate": 8.737260750683569e-06,
-      "loss": 4.1665,
+      "loss": 4.0541,
       "step": 3820
     },
     {
       "epoch": 0.8456776475790405,
-      "grad_norm": 3.153214693069458,
+      "grad_norm": 3.768934488296509,
       "learning_rate": 8.675118071091226e-06,
-      "loss": 4.1227,
+      "loss": 4.0144,
       "step": 3825
     },
     {
       "epoch": 0.8467831085562679,
-      "grad_norm": 3.160295009613037,
+      "grad_norm": 3.7445712089538574,
       "learning_rate": 8.612975391498881e-06,
-      "loss": 4.1928,
+      "loss": 4.0745,
       "step": 3830
     },
     {
       "epoch": 0.8478885695334955,
-      "grad_norm": 3.522057294845581,
+      "grad_norm": 4.220143795013428,
       "learning_rate": 8.550832711906539e-06,
-      "loss": 4.3234,
+      "loss": 4.2218,
       "step": 3835
     },
     {
       "epoch": 0.848994030510723,
-      "grad_norm": 3.3850722312927246,
+      "grad_norm": 4.009212493896484,
       "learning_rate": 8.488690032314193e-06,
-      "loss": 4.2035,
+      "loss": 4.0958,
       "step": 3840
     },
     {
       "epoch": 0.8500994914879505,
-      "grad_norm": 3.237739324569702,
+      "grad_norm": 3.8734123706817627,
       "learning_rate": 8.42654735272185e-06,
-      "loss": 4.0377,
+      "loss": 3.9215,
       "step": 3845
     },
     {
       "epoch": 0.851204952465178,
-      "grad_norm": 3.3790619373321533,
+      "grad_norm": 4.034563064575195,
       "learning_rate": 8.364404673129506e-06,
-      "loss": 4.1112,
+      "loss": 4.0087,
       "step": 3850
     },
     {
       "epoch": 0.8523104134424054,
-      "grad_norm": 3.395925760269165,
+      "grad_norm": 4.0455474853515625,
       "learning_rate": 8.302261993537162e-06,
-      "loss": 4.3152,
+      "loss": 4.2065,
       "step": 3855
     },
     {
       "epoch": 0.853415874419633,
-      "grad_norm": 2.8968868255615234,
+      "grad_norm": 3.4714512825012207,
       "learning_rate": 8.240119313944818e-06,
-      "loss": 4.1642,
+      "loss": 4.0525,
       "step": 3860
     },
     {
       "epoch": 0.8545213353968605,
-      "grad_norm": 3.6181344985961914,
+      "grad_norm": 4.278322696685791,
       "learning_rate": 8.177976634352472e-06,
-      "loss": 4.27,
+      "loss": 4.161,
       "step": 3865
     },
     {
       "epoch": 0.855626796374088,
-      "grad_norm": 3.3780412673950195,
+      "grad_norm": 4.036164283752441,
       "learning_rate": 8.11583395476013e-06,
-      "loss": 4.2319,
+      "loss": 4.1256,
       "step": 3870
     },
     {
       "epoch": 0.8567322573513155,
-      "grad_norm": 3.0761659145355225,
+      "grad_norm": 3.6622910499572754,
       "learning_rate": 8.053691275167785e-06,
-      "loss": 4.2244,
+      "loss": 4.1116,
       "step": 3875
     },
     {
       "epoch": 0.857837718328543,
-      "grad_norm": 3.188369035720825,
+      "grad_norm": 3.780794620513916,
       "learning_rate": 7.991548595575441e-06,
-      "loss": 4.1855,
+      "loss": 4.0839,
       "step": 3880
     },
     {
       "epoch": 0.8589431793057705,
-      "grad_norm": 3.280965805053711,
+      "grad_norm": 3.8837084770202637,
       "learning_rate": 7.929405915983097e-06,
-      "loss": 4.2297,
+      "loss": 4.1146,
       "step": 3885
     },
     {
       "epoch": 0.860048640282998,
-      "grad_norm": 3.428769111633301,
+      "grad_norm": 4.092316150665283,
       "learning_rate": 7.867263236390753e-06,
-      "loss": 4.2635,
+      "loss": 4.1634,
       "step": 3890
     },
     {
       "epoch": 0.8611541012602255,
-      "grad_norm": 3.372145414352417,
+      "grad_norm": 4.024561882019043,
       "learning_rate": 7.80512055679841e-06,
-      "loss": 4.1799,
+      "loss": 4.077,
       "step": 3895
     },
     {
       "epoch": 0.862259562237453,
-      "grad_norm": 3.669572114944458,
+      "grad_norm": 4.362841606140137,
       "learning_rate": 7.742977877206066e-06,
-      "loss": 4.1279,
+      "loss": 4.0268,
       "step": 3900
     },
     {
       "epoch": 0.8633650232146806,
-      "grad_norm": 3.3069515228271484,
+      "grad_norm": 3.9448471069335938,
       "learning_rate": 7.680835197613722e-06,
-      "loss": 4.2423,
+      "loss": 4.1406,
       "step": 3905
     },
     {
       "epoch": 0.864470484191908,
-      "grad_norm": 3.4965929985046387,
+      "grad_norm": 4.135385036468506,
       "learning_rate": 7.618692518021378e-06,
-      "loss": 4.2445,
+      "loss": 4.1443,
       "step": 3910
     },
     {
       "epoch": 0.8655759451691355,
-      "grad_norm": 3.3007524013519287,
+      "grad_norm": 3.929286479949951,
       "learning_rate": 7.556549838429033e-06,
-      "loss": 4.3169,
+      "loss": 4.2124,
       "step": 3915
     },
     {
       "epoch": 0.866681406146363,
-      "grad_norm": 3.3031368255615234,
+      "grad_norm": 3.92202091217041,
       "learning_rate": 7.494407158836689e-06,
-      "loss": 4.2489,
+      "loss": 4.1477,
       "step": 3920
     },
     {
       "epoch": 0.8677868671235905,
-      "grad_norm": 3.3182923793792725,
+      "grad_norm": 3.9424068927764893,
       "learning_rate": 7.432264479244346e-06,
-      "loss": 4.1043,
+      "loss": 3.9998,
       "step": 3925
     },
     {
       "epoch": 0.8688923281008181,
-      "grad_norm": 3.1912918090820312,
+      "grad_norm": 3.804424524307251,
       "learning_rate": 7.370121799652001e-06,
-      "loss": 4.225,
+      "loss": 4.1245,
       "step": 3930
     },
     {
       "epoch": 0.8699977890780456,
-      "grad_norm": 3.4221689701080322,
+      "grad_norm": 4.067017078399658,
       "learning_rate": 7.307979120059657e-06,
-      "loss": 4.2911,
+      "loss": 4.1885,
       "step": 3935
     },
     {
       "epoch": 0.871103250055273,
-      "grad_norm": 3.3450770378112793,
+      "grad_norm": 3.9617257118225098,
       "learning_rate": 7.2458364404673125e-06,
-      "loss": 4.4661,
+      "loss": 4.3714,
       "step": 3940
     },
     {
       "epoch": 0.8722087110325005,
-      "grad_norm": 3.3857436180114746,
+      "grad_norm": 3.944721221923828,
       "learning_rate": 7.1836937608749695e-06,
-      "loss": 4.1062,
+      "loss": 4.0075,
       "step": 3945
     },
     {
       "epoch": 0.873314172009728,
-      "grad_norm": 3.2162883281707764,
+      "grad_norm": 3.816215753555298,
       "learning_rate": 7.121551081282625e-06,
-      "loss": 4.2926,
+      "loss": 4.1917,
       "step": 3950
     },
     {
       "epoch": 0.8744196329869556,
-      "grad_norm": 2.971797227859497,
+      "grad_norm": 3.535334587097168,
       "learning_rate": 7.059408401690282e-06,
-      "loss": 4.0731,
+      "loss": 3.9712,
       "step": 3955
     },
     {
       "epoch": 0.8755250939641831,
-      "grad_norm": 3.228489875793457,
+      "grad_norm": 3.8114192485809326,
       "learning_rate": 6.997265722097937e-06,
-      "loss": 4.1616,
+      "loss": 4.0627,
       "step": 3960
     },
     {
       "epoch": 0.8766305549414106,
-      "grad_norm": 3.2910053730010986,
+      "grad_norm": 3.934971570968628,
       "learning_rate": 6.935123042505594e-06,
-      "loss": 4.2075,
+      "loss": 4.1068,
       "step": 3965
     },
     {
       "epoch": 0.877736015918638,
-      "grad_norm": 3.1011228561401367,
+      "grad_norm": 3.6866393089294434,
       "learning_rate": 6.8729803629132495e-06,
-      "loss": 4.1851,
+      "loss": 4.0841,
       "step": 3970
     },
     {
       "epoch": 0.8788414768958656,
-      "grad_norm": 3.6701035499572754,
+      "grad_norm": 4.31212043762207,
       "learning_rate": 6.810837683320905e-06,
-      "loss": 4.1968,
+      "loss": 4.1102,
       "step": 3975
     },
     {
       "epoch": 0.8799469378730931,
-      "grad_norm": 3.310450315475464,
+      "grad_norm": 3.955538511276245,
       "learning_rate": 6.748695003728561e-06,
-      "loss": 4.3885,
+      "loss": 4.297,
       "step": 3980
     },
     {
       "epoch": 0.8810523988503206,
-      "grad_norm": 3.3232550621032715,
+      "grad_norm": 3.925504684448242,
       "learning_rate": 6.686552324136216e-06,
-      "loss": 4.202,
+      "loss": 4.1005,
       "step": 3985
     },
     {
       "epoch": 0.8821578598275481,
-      "grad_norm": 3.33705472946167,
+      "grad_norm": 3.9269564151763916,
       "learning_rate": 6.624409644543873e-06,
-      "loss": 4.2345,
+      "loss": 4.1324,
       "step": 3990
     },
     {
       "epoch": 0.8832633208047755,
-      "grad_norm": 3.648831605911255,
+      "grad_norm": 4.313398361206055,
       "learning_rate": 6.562266964951529e-06,
-      "loss": 4.2464,
+      "loss": 4.152,
       "step": 3995
     },
     {
       "epoch": 0.8843687817820031,
-      "grad_norm": 3.2218527793884277,
+      "grad_norm": 3.790170669555664,
       "learning_rate": 6.500124285359186e-06,
-      "loss": 4.0956,
+      "loss": 4.004,
       "step": 4000
     },
     {
       "epoch": 0.8854742427592306,
-      "grad_norm": 3.0550131797790527,
+      "grad_norm": 3.6186208724975586,
       "learning_rate": 6.437981605766841e-06,
-      "loss": 4.1712,
+      "loss": 4.0759,
       "step": 4005
     },
     {
       "epoch": 0.8865797037364581,
-      "grad_norm": 3.1984024047851562,
+      "grad_norm": 3.7495853900909424,
       "learning_rate": 6.375838926174497e-06,
-      "loss": 4.2718,
+      "loss": 4.1658,
       "step": 4010
     },
     {
       "epoch": 0.8876851647136856,
-      "grad_norm": 3.2509777545928955,
+      "grad_norm": 3.8539583683013916,
       "learning_rate": 6.3136962465821526e-06,
-      "loss": 4.0173,
+      "loss": 3.9178,
       "step": 4015
     },
     {
       "epoch": 0.8887906256909132,
-      "grad_norm": 3.146519899368286,
+      "grad_norm": 3.716893196105957,
       "learning_rate": 6.2515535669898096e-06,
-      "loss": 4.4115,
+      "loss": 4.3165,
       "step": 4020
     },
     {
       "epoch": 0.8898960866681406,
-      "grad_norm": 3.422335624694824,
+      "grad_norm": 4.035796165466309,
       "learning_rate": 6.189410887397465e-06,
-      "loss": 4.3307,
+      "loss": 4.24,
       "step": 4025
     },
     {
       "epoch": 0.8910015476453681,
-      "grad_norm": 3.50016188621521,
+      "grad_norm": 4.14146614074707,
       "learning_rate": 6.127268207805121e-06,
-      "loss": 4.0675,
+      "loss": 3.9737,
       "step": 4030
     },
     {
       "epoch": 0.8921070086225956,
-      "grad_norm": 3.059391975402832,
+      "grad_norm": 3.6457576751708984,
       "learning_rate": 6.065125528212777e-06,
-      "loss": 4.2215,
+      "loss": 4.1151,
       "step": 4035
     },
     {
       "epoch": 0.8932124695998231,
-      "grad_norm": 3.585162401199341,
+      "grad_norm": 4.226096153259277,
       "learning_rate": 6.002982848620433e-06,
-      "loss": 4.1206,
+      "loss": 4.0285,
       "step": 4040
     },
     {
       "epoch": 0.8943179305770507,
-      "grad_norm": 3.1658449172973633,
+      "grad_norm": 3.752408027648926,
       "learning_rate": 5.940840169028089e-06,
-      "loss": 4.1826,
+      "loss": 4.0855,
       "step": 4045
     },
     {
       "epoch": 0.8954233915542781,
-      "grad_norm": 3.30590558052063,
+      "grad_norm": 3.9440932273864746,
       "learning_rate": 5.878697489435745e-06,
-      "loss": 4.07,
+      "loss": 3.9758,
       "step": 4050
     },
     {
       "epoch": 0.8965288525315056,
-      "grad_norm": 3.5523128509521484,
+      "grad_norm": 4.19743537902832,
       "learning_rate": 5.8165548098434e-06,
-      "loss": 4.2302,
+      "loss": 4.1362,
       "step": 4055
     },
     {
       "epoch": 0.8976343135087331,
-      "grad_norm": 3.2362444400787354,
+      "grad_norm": 3.823868989944458,
       "learning_rate": 5.754412130251056e-06,
-      "loss": 4.1555,
+      "loss": 4.0663,
       "step": 4060
     },
     {
       "epoch": 0.8987397744859607,
-      "grad_norm": 2.9280905723571777,
+      "grad_norm": 3.4545068740844727,
       "learning_rate": 5.692269450658713e-06,
-      "loss": 4.1708,
+      "loss": 4.0821,
       "step": 4065
     },
     {
       "epoch": 0.8998452354631882,
-      "grad_norm": 3.277392625808716,
+      "grad_norm": 3.850423812866211,
       "learning_rate": 5.630126771066369e-06,
-      "loss": 4.1606,
+      "loss": 4.0663,
       "step": 4070
     },
     {
       "epoch": 0.9009506964404157,
-      "grad_norm": 2.9546451568603516,
+      "grad_norm": 3.5115787982940674,
       "learning_rate": 5.567984091474025e-06,
-      "loss": 4.1486,
+      "loss": 4.0567,
       "step": 4075
     },
     {
       "epoch": 0.9020561574176431,
-      "grad_norm": 3.33906888961792,
+      "grad_norm": 3.9357197284698486,
       "learning_rate": 5.50584141188168e-06,
-      "loss": 4.2423,
+      "loss": 4.1571,
       "step": 4080
     },
     {
       "epoch": 0.9031616183948706,
-      "grad_norm": 3.414642572402954,
+      "grad_norm": 3.9711387157440186,
       "learning_rate": 5.4436987322893364e-06,
-      "loss": 4.1806,
+      "loss": 4.0891,
       "step": 4085
     },
     {
       "epoch": 0.9042670793720982,
-      "grad_norm": 3.1724166870117188,
+      "grad_norm": 3.7287356853485107,
       "learning_rate": 5.381556052696993e-06,
-      "loss": 4.3395,
+      "loss": 4.2445,
       "step": 4090
     },
     {
       "epoch": 0.9053725403493257,
-      "grad_norm": 3.3159971237182617,
+      "grad_norm": 3.9396510124206543,
       "learning_rate": 5.319413373104649e-06,
-      "loss": 4.1692,
+      "loss": 4.0856,
       "step": 4095
     },
     {
       "epoch": 0.9064780013265532,
-      "grad_norm": 3.149585008621216,
+      "grad_norm": 3.733102321624756,
       "learning_rate": 5.257270693512305e-06,
-      "loss": 4.1873,
+      "loss": 4.0961,
       "step": 4100
     },
     {
       "epoch": 0.9075834623037807,
-      "grad_norm": 3.5617358684539795,
+      "grad_norm": 4.200262546539307,
       "learning_rate": 5.195128013919961e-06,
-      "loss": 4.2171,
+      "loss": 4.126,
       "step": 4105
     },
     {
       "epoch": 0.9086889232810081,
-      "grad_norm": 3.268549680709839,
+      "grad_norm": 3.8280789852142334,
       "learning_rate": 5.1329853343276164e-06,
-      "loss": 4.1768,
+      "loss": 4.0849,
       "step": 4110
     },
     {
       "epoch": 0.9097943842582357,
-      "grad_norm": 3.424433708190918,
+      "grad_norm": 4.028787136077881,
       "learning_rate": 5.070842654735273e-06,
-      "loss": 4.4327,
+      "loss": 4.3436,
       "step": 4115
     },
     {
       "epoch": 0.9108998452354632,
-      "grad_norm": 3.495929479598999,
+      "grad_norm": 4.13646936416626,
       "learning_rate": 5.008699975142928e-06,
-      "loss": 4.1886,
+      "loss": 4.1045,
       "step": 4120
     },
     {
       "epoch": 0.9120053062126907,
-      "grad_norm": 3.045023202896118,
+      "grad_norm": 3.6056816577911377,
       "learning_rate": 4.946557295550584e-06,
-      "loss": 4.3829,
+      "loss": 4.3006,
       "step": 4125
     },
     {
       "epoch": 0.9131107671899182,
-      "grad_norm": 3.1356985569000244,
+      "grad_norm": 3.693286895751953,
       "learning_rate": 4.88441461595824e-06,
-      "loss": 4.3304,
+      "loss": 4.2466,
       "step": 4130
     },
     {
       "epoch": 0.9142162281671457,
-      "grad_norm": 3.389559507369995,
+      "grad_norm": 3.950615167617798,
       "learning_rate": 4.8222719363658965e-06,
-      "loss": 4.1715,
+      "loss": 4.0856,
       "step": 4135
     },
     {
       "epoch": 0.9153216891443732,
-      "grad_norm": 3.1588001251220703,
+      "grad_norm": 3.727515935897827,
       "learning_rate": 4.760129256773553e-06,
-      "loss": 4.2491,
+      "loss": 4.1629,
       "step": 4140
     },
     {
       "epoch": 0.9164271501216007,
-      "grad_norm": 3.5233826637268066,
+      "grad_norm": 4.132688045501709,
       "learning_rate": 4.697986577181209e-06,
-      "loss": 4.409,
+      "loss": 4.3301,
       "step": 4145
     },
     {
       "epoch": 0.9175326110988282,
-      "grad_norm": 3.0876009464263916,
+      "grad_norm": 3.6707515716552734,
       "learning_rate": 4.635843897588864e-06,
-      "loss": 4.1037,
+      "loss": 4.0202,
       "step": 4150
     },
     {
       "epoch": 0.9186380720760557,
-      "grad_norm": 3.64609956741333,
+      "grad_norm": 4.277684211730957,
       "learning_rate": 4.57370121799652e-06,
-      "loss": 4.2202,
+      "loss": 4.1274,
       "step": 4155
     },
     {
       "epoch": 0.9197435330532833,
-      "grad_norm": 3.119335174560547,
+      "grad_norm": 3.684551477432251,
       "learning_rate": 4.511558538404176e-06,
-      "loss": 4.2293,
+      "loss": 4.1595,
       "step": 4160
     },
     {
       "epoch": 0.9208489940305107,
-      "grad_norm": 3.2007765769958496,
+      "grad_norm": 3.761082172393799,
       "learning_rate": 4.449415858811832e-06,
-      "loss": 4.2337,
+      "loss": 4.1573,
       "step": 4165
     },
     {
       "epoch": 0.9219544550077382,
-      "grad_norm": 2.860046625137329,
+      "grad_norm": 3.361445188522339,
       "learning_rate": 4.387273179219488e-06,
-      "loss": 4.2855,
+      "loss": 4.2017,
       "step": 4170
     },
     {
       "epoch": 0.9230599159849657,
-      "grad_norm": 3.472074270248413,
+      "grad_norm": 4.054385185241699,
       "learning_rate": 4.325130499627144e-06,
-      "loss": 4.2792,
+      "loss": 4.186,
       "step": 4175
     },
     {
       "epoch": 0.9241653769621933,
-      "grad_norm": 3.21456241607666,
+      "grad_norm": 3.7968826293945312,
       "learning_rate": 4.2629878200348e-06,
-      "loss": 4.2083,
+      "loss": 4.1265,
       "step": 4180
     },
     {
       "epoch": 0.9252708379394208,
-      "grad_norm": 3.0883960723876953,
+      "grad_norm": 3.6149940490722656,
       "learning_rate": 4.2008451404424565e-06,
-      "loss": 4.2125,
+      "loss": 4.1346,
       "step": 4185
     },
     {
       "epoch": 0.9263762989166482,
-      "grad_norm": 3.1821343898773193,
+      "grad_norm": 3.720431089401245,
       "learning_rate": 4.138702460850112e-06,
-      "loss": 4.3135,
+      "loss": 4.233,
       "step": 4190
     },
     {
       "epoch": 0.9274817598938757,
-      "grad_norm": 3.2891180515289307,
+      "grad_norm": 3.842650890350342,
       "learning_rate": 4.076559781257768e-06,
-      "loss": 4.2337,
+      "loss": 4.1558,
       "step": 4195
     },
     {
       "epoch": 0.9285872208711032,
-      "grad_norm": 3.036611557006836,
+      "grad_norm": 3.5617995262145996,
       "learning_rate": 4.014417101665424e-06,
-      "loss": 4.1799,
+      "loss": 4.1054,
       "step": 4200
     },
     {
       "epoch": 0.9296926818483308,
-      "grad_norm": 3.262669086456299,
+      "grad_norm": 3.804689884185791,
       "learning_rate": 3.95227442207308e-06,
-      "loss": 4.3257,
+      "loss": 4.2527,
       "step": 4205
     },
     {
       "epoch": 0.9307981428255583,
-      "grad_norm": 3.32913875579834,
+      "grad_norm": 3.900244951248169,
       "learning_rate": 3.8901317424807365e-06,
-      "loss": 4.2918,
+      "loss": 4.2147,
       "step": 4210
     },
     {
       "epoch": 0.9319036038027858,
-      "grad_norm": 3.221358299255371,
+      "grad_norm": 3.8041465282440186,
       "learning_rate": 3.827989062888392e-06,
-      "loss": 4.2922,
+      "loss": 4.2198,
       "step": 4215
     },
     {
       "epoch": 0.9330090647800132,
-      "grad_norm": 3.131178617477417,
+      "grad_norm": 3.6801540851593018,
       "learning_rate": 3.7658463832960476e-06,
-      "loss": 4.1484,
+      "loss": 4.0654,
       "step": 4220
     },
     {
       "epoch": 0.9341145257572407,
-      "grad_norm": 3.0813159942626953,
+      "grad_norm": 3.6228909492492676,
       "learning_rate": 3.7037037037037037e-06,
-      "loss": 4.2841,
+      "loss": 4.2119,
       "step": 4225
     },
     {
       "epoch": 0.9352199867344683,
-      "grad_norm": 2.8390700817108154,
+      "grad_norm": 3.339024305343628,
       "learning_rate": 3.64156102411136e-06,
-      "loss": 4.0598,
+      "loss": 3.9825,
       "step": 4230
     },
     {
       "epoch": 0.9363254477116958,
-      "grad_norm": 3.10927677154541,
+      "grad_norm": 3.6585726737976074,
       "learning_rate": 3.5794183445190157e-06,
-      "loss": 4.1328,
+      "loss": 4.0588,
       "step": 4235
     },
     {
       "epoch": 0.9374309086889233,
-      "grad_norm": 3.2241241931915283,
+      "grad_norm": 3.772191047668457,
       "learning_rate": 3.517275664926672e-06,
-      "loss": 4.2188,
+      "loss": 4.1361,
       "step": 4240
     },
     {
       "epoch": 0.9385363696661508,
-      "grad_norm": 2.9095420837402344,
+      "grad_norm": 3.402756690979004,
       "learning_rate": 3.455132985334328e-06,
-      "loss": 4.068,
+      "loss": 3.9858,
       "step": 4245
     },
     {
       "epoch": 0.9396418306433783,
-      "grad_norm": 3.1288955211639404,
+      "grad_norm": 3.6663706302642822,
       "learning_rate": 3.3929903057419838e-06,
-      "loss": 4.2663,
+      "loss": 4.1922,
       "step": 4250
     },
     {
       "epoch": 0.9407472916206058,
-      "grad_norm": 3.026554584503174,
+      "grad_norm": 3.5454792976379395,
       "learning_rate": 3.33084762614964e-06,
-      "loss": 4.1512,
+      "loss": 4.0807,
       "step": 4255
     },
     {
       "epoch": 0.9418527525978333,
-      "grad_norm": 3.222672462463379,
+      "grad_norm": 3.7887284755706787,
       "learning_rate": 3.268704946557296e-06,
-      "loss": 4.235,
+      "loss": 4.1674,
       "step": 4260
     },
     {
       "epoch": 0.9429582135750608,
-      "grad_norm": 3.381204605102539,
+      "grad_norm": 3.960259437561035,
       "learning_rate": 3.206562266964952e-06,
-      "loss": 4.1584,
+      "loss": 4.0886,
       "step": 4265
     },
     {
       "epoch": 0.9440636745522883,
-      "grad_norm": 3.3569135665893555,
+      "grad_norm": 3.9364819526672363,
       "learning_rate": 3.144419587372607e-06,
-      "loss": 4.2849,
+      "loss": 4.2069,
       "step": 4270
     },
     {
       "epoch": 0.9451691355295158,
-      "grad_norm": 3.2201907634735107,
+      "grad_norm": 3.784364700317383,
       "learning_rate": 3.0822769077802638e-06,
-      "loss": 4.1318,
+      "loss": 4.0632,
       "step": 4275
     },
     {
       "epoch": 0.9462745965067433,
-      "grad_norm": 3.078237771987915,
+      "grad_norm": 3.620039224624634,
       "learning_rate": 3.02013422818792e-06,
-      "loss": 4.2257,
+      "loss": 4.1531,
       "step": 4280
     },
     {
       "epoch": 0.9473800574839708,
-      "grad_norm": 2.9291415214538574,
+      "grad_norm": 3.4464962482452393,
       "learning_rate": 2.9579915485955753e-06,
-      "loss": 4.397,
+      "loss": 4.3285,
       "step": 4285
     },
     {
       "epoch": 0.9484855184611983,
-      "grad_norm": 3.3114891052246094,
+      "grad_norm": 3.883500576019287,
       "learning_rate": 2.8958488690032314e-06,
-      "loss": 4.1599,
+      "loss": 4.0893,
       "step": 4290
     },
     {
       "epoch": 0.9495909794384259,
-      "grad_norm": 3.3049850463867188,
+      "grad_norm": 3.8970224857330322,
       "learning_rate": 2.8337061894108876e-06,
-      "loss": 4.2123,
+      "loss": 4.1374,
       "step": 4295
     },
     {
       "epoch": 0.9506964404156534,
-      "grad_norm": 2.979609251022339,
+      "grad_norm": 3.481741428375244,
       "learning_rate": 2.7715635098185434e-06,
-      "loss": 4.1817,
+      "loss": 4.1142,
       "step": 4300
     },
     {
       "epoch": 0.9518019013928808,
-      "grad_norm": 3.1335394382476807,
+      "grad_norm": 3.664961814880371,
       "learning_rate": 2.7094208302261995e-06,
-      "loss": 4.2932,
+      "loss": 4.2209,
       "step": 4305
     },
     {
       "epoch": 0.9529073623701083,
-      "grad_norm": 3.3001952171325684,
+      "grad_norm": 3.868842840194702,
       "learning_rate": 2.6472781506338553e-06,
-      "loss": 4.4201,
+      "loss": 4.3574,
       "step": 4310
     },
     {
       "epoch": 0.9540128233473358,
-      "grad_norm": 3.1160495281219482,
+      "grad_norm": 3.64719557762146,
       "learning_rate": 2.5851354710415115e-06,
-      "loss": 4.1786,
+      "loss": 4.1112,
       "step": 4315
     },
     {
       "epoch": 0.9551182843245634,
-      "grad_norm": 2.8716208934783936,
+      "grad_norm": 3.3634984493255615,
       "learning_rate": 2.522992791449167e-06,
-      "loss": 3.9942,
+      "loss": 3.9221,
       "step": 4320
     },
     {
       "epoch": 0.9562237453017909,
-      "grad_norm": 3.0611040592193604,
+      "grad_norm": 3.578833818435669,
       "learning_rate": 2.4608501118568234e-06,
-      "loss": 4.4118,
+      "loss": 4.3452,
       "step": 4325
     },
     {
       "epoch": 0.9573292062790183,
-      "grad_norm": 2.9500648975372314,
+      "grad_norm": 3.468839168548584,
       "learning_rate": 2.3987074322644795e-06,
-      "loss": 4.305,
+      "loss": 4.243,
       "step": 4330
     },
     {
       "epoch": 0.9584346672562458,
-      "grad_norm": 3.5862972736358643,
+      "grad_norm": 4.166474342346191,
       "learning_rate": 2.3365647526721353e-06,
-      "loss": 4.3046,
+      "loss": 4.2432,
       "step": 4335
     },
     {
       "epoch": 0.9595401282334733,
-      "grad_norm": 3.304366111755371,
+      "grad_norm": 3.8754284381866455,
       "learning_rate": 2.274422073079791e-06,
-      "loss": 4.3483,
+      "loss": 4.2864,
       "step": 4340
     },
     {
       "epoch": 0.9606455892107009,
-      "grad_norm": 3.4040110111236572,
+      "grad_norm": 3.974168062210083,
       "learning_rate": 2.2122793934874472e-06,
-      "loss": 4.2975,
+      "loss": 4.2228,
       "step": 4345
     },
     {
       "epoch": 0.9617510501879284,
-      "grad_norm": 3.197815179824829,
+      "grad_norm": 3.7412900924682617,
       "learning_rate": 2.1501367138951034e-06,
-      "loss": 4.3031,
+      "loss": 4.2412,
       "step": 4350
     },
     {
       "epoch": 0.9628565111651559,
-      "grad_norm": 3.365293502807617,
+      "grad_norm": 3.9455902576446533,
       "learning_rate": 2.087994034302759e-06,
-      "loss": 4.2018,
+      "loss": 4.1292,
       "step": 4355
     },
     {
       "epoch": 0.9639619721423833,
-      "grad_norm": 3.179311990737915,
+      "grad_norm": 3.740370035171509,
       "learning_rate": 2.0258513547104153e-06,
-      "loss": 4.3385,
+      "loss": 4.2652,
       "step": 4360
     },
     {
       "epoch": 0.9650674331196109,
-      "grad_norm": 3.1740834712982178,
+      "grad_norm": 3.733259677886963,
       "learning_rate": 1.963708675118071e-06,
-      "loss": 4.4034,
+      "loss": 4.3391,
       "step": 4365
     },
     {
       "epoch": 0.9661728940968384,
-      "grad_norm": 3.0727176666259766,
+      "grad_norm": 3.6092658042907715,
       "learning_rate": 1.901565995525727e-06,
-      "loss": 4.2515,
+      "loss": 4.1877,
       "step": 4370
     },
     {
       "epoch": 0.9672783550740659,
-      "grad_norm": 2.9758899211883545,
+      "grad_norm": 3.488403797149658,
       "learning_rate": 1.8394233159333832e-06,
-      "loss": 4.1974,
+      "loss": 4.1456,
       "step": 4375
     },
     {
       "epoch": 0.9683838160512934,
-      "grad_norm": 3.014615774154663,
+      "grad_norm": 3.5342633724212646,
       "learning_rate": 1.7772806363410391e-06,
-      "loss": 4.3097,
+      "loss": 4.2478,
       "step": 4380
     },
     {
       "epoch": 0.969489277028521,
-      "grad_norm": 3.5511038303375244,
+      "grad_norm": 4.200704574584961,
       "learning_rate": 1.7151379567486951e-06,
-      "loss": 4.2784,
+      "loss": 4.2143,
       "step": 4385
     },
     {
       "epoch": 0.9705947380057484,
-      "grad_norm": 2.977102518081665,
+      "grad_norm": 3.4691877365112305,
       "learning_rate": 1.6529952771563513e-06,
-      "loss": 4.2234,
+      "loss": 4.1539,
       "step": 4390
     },
     {
       "epoch": 0.9717001989829759,
-      "grad_norm": 2.964914083480835,
+      "grad_norm": 3.464365243911743,
       "learning_rate": 1.5908525975640068e-06,
-      "loss": 4.1375,
+      "loss": 4.0741,
       "step": 4395
     },
     {
       "epoch": 0.9728056599602034,
-      "grad_norm": 2.916311025619507,
+      "grad_norm": 3.409198045730591,
       "learning_rate": 1.528709917971663e-06,
-      "loss": 4.1116,
+      "loss": 4.0496,
       "step": 4400
     },
     {
       "epoch": 0.9739111209374309,
-      "grad_norm": 3.3200995922088623,
+      "grad_norm": 3.876699686050415,
       "learning_rate": 1.466567238379319e-06,
-      "loss": 4.3596,
+      "loss": 4.2981,
       "step": 4405
     },
     {
       "epoch": 0.9750165819146585,
-      "grad_norm": 3.0481033325195312,
+      "grad_norm": 3.5573813915252686,
       "learning_rate": 1.4044245587869751e-06,
-      "loss": 4.303,
+      "loss": 4.2538,
       "step": 4410
     },
     {
       "epoch": 0.9761220428918859,
-      "grad_norm": 3.04089617729187,
+      "grad_norm": 3.57130765914917,
       "learning_rate": 1.3422818791946309e-06,
-      "loss": 4.3629,
+      "loss": 4.3062,
       "step": 4415
     },
     {
       "epoch": 0.9772275038691134,
-      "grad_norm": 3.03387713432312,
+      "grad_norm": 3.5500142574310303,
       "learning_rate": 1.280139199602287e-06,
-      "loss": 4.2679,
+      "loss": 4.212,
       "step": 4420
     },
     {
       "epoch": 0.9783329648463409,
-      "grad_norm": 3.1632862091064453,
+      "grad_norm": 3.7248690128326416,
       "learning_rate": 1.2179965200099428e-06,
-      "loss": 4.153,
+      "loss": 4.0904,
       "step": 4425
     },
     {
       "epoch": 0.9794384258235684,
-      "grad_norm": 3.382652759552002,
+      "grad_norm": 3.9716479778289795,
       "learning_rate": 1.1558538404175988e-06,
-      "loss": 4.1147,
+      "loss": 4.0558,
       "step": 4430
     },
     {
       "epoch": 0.980543886800796,
-      "grad_norm": 3.4399046897888184,
+      "grad_norm": 4.010632038116455,
       "learning_rate": 1.093711160825255e-06,
-      "loss": 4.2737,
+      "loss": 4.2186,
       "step": 4435
     },
     {
       "epoch": 0.9816493477780235,
-      "grad_norm": 3.3583288192749023,
+      "grad_norm": 3.9161455631256104,
       "learning_rate": 1.0315684812329107e-06,
-      "loss": 4.2274,
+      "loss": 4.1628,
       "step": 4440
     },
     {
       "epoch": 0.9827548087552509,
-      "grad_norm": 3.291776657104492,
+      "grad_norm": 3.856630325317383,
       "learning_rate": 9.694258016405668e-07,
-      "loss": 4.1284,
+      "loss": 4.0669,
       "step": 4445
     },
     {
       "epoch": 0.9838602697324784,
-      "grad_norm": 3.148688554763794,
+      "grad_norm": 3.680676221847534,
       "learning_rate": 9.072831220482228e-07,
-      "loss": 4.3734,
+      "loss": 4.3089,
       "step": 4450
     },
     {
       "epoch": 0.9849657307097059,
-      "grad_norm": 2.98494553565979,
+      "grad_norm": 3.4856209754943848,
       "learning_rate": 8.451404424558787e-07,
-      "loss": 4.2998,
+      "loss": 4.2488,
       "step": 4455
     },
     {
       "epoch": 0.9860711916869335,
-      "grad_norm": 3.550734043121338,
+      "grad_norm": 4.153370380401611,
       "learning_rate": 7.829977628635347e-07,
-      "loss": 4.131,
+      "loss": 4.0782,
       "step": 4460
     },
     {
       "epoch": 0.987176652664161,
-      "grad_norm": 3.148184299468994,
+      "grad_norm": 3.657689094543457,
       "learning_rate": 7.208550832711907e-07,
-      "loss": 4.211,
+      "loss": 4.1552,
       "step": 4465
     },
     {
       "epoch": 0.9882821136413884,
-      "grad_norm": 3.389477491378784,
+      "grad_norm": 3.9328858852386475,
       "learning_rate": 6.587124036788466e-07,
-      "loss": 4.3192,
+      "loss": 4.259,
       "step": 4470
     },
     {
       "epoch": 0.9893875746186159,
-      "grad_norm": 2.744230031967163,
+      "grad_norm": 3.2011306285858154,
       "learning_rate": 5.965697240865026e-07,
-      "loss": 4.3994,
+      "loss": 4.3455,
       "step": 4475
     },
     {
       "epoch": 0.9904930355958435,
-      "grad_norm": 3.189837694168091,
+      "grad_norm": 3.7026784420013428,
       "learning_rate": 5.344270444941587e-07,
-      "loss": 4.3435,
+      "loss": 4.2958,
       "step": 4480
     },
     {
       "epoch": 0.991598496573071,
-      "grad_norm": 3.2491848468780518,
+      "grad_norm": 3.779052734375,
       "learning_rate": 4.722843649018146e-07,
-      "loss": 4.3766,
+      "loss": 4.3165,
       "step": 4485
     },
     {
       "epoch": 0.9927039575502985,
-      "grad_norm": 3.1869592666625977,
+      "grad_norm": 3.7316598892211914,
       "learning_rate": 4.1014168530947054e-07,
-      "loss": 4.393,
+      "loss": 4.3329,
       "step": 4490
     },
     {
       "epoch": 0.993809418527526,
-      "grad_norm": 3.4105918407440186,
+      "grad_norm": 3.936168670654297,
       "learning_rate": 3.4799900571712656e-07,
-      "loss": 4.2419,
+      "loss": 4.1988,
       "step": 4495
     },
     {
       "epoch": 0.9949148795047534,
-      "grad_norm": 3.1611382961273193,
+      "grad_norm": 3.7016172409057617,
       "learning_rate": 2.858563261247825e-07,
-      "loss": 4.3572,
+      "loss": 4.3067,
       "step": 4500
     },
     {
       "epoch": 0.996020340481981,
-      "grad_norm": 3.0471818447113037,
+      "grad_norm": 3.5614588260650635,
       "learning_rate": 2.2371364653243848e-07,
-      "loss": 4.3163,
+      "loss": 4.2676,
       "step": 4505
     },
     {
       "epoch": 0.9971258014592085,
-      "grad_norm": 2.9979894161224365,
+      "grad_norm": 3.4923853874206543,
       "learning_rate": 1.6157096694009447e-07,
-      "loss": 4.1885,
+      "loss": 4.137,
       "step": 4510
     },
     {
       "epoch": 0.998231262436436,
-      "grad_norm": 3.4176154136657715,
+      "grad_norm": 3.9990386962890625,
       "learning_rate": 9.942828734775043e-08,
-      "loss": 4.3076,
+      "loss": 4.2468,
       "step": 4515
     },
     {
       "epoch": 0.9993367234136635,
-      "grad_norm": 3.594446897506714,
+      "grad_norm": 4.192621231079102,
       "learning_rate": 3.728560775540641e-08,
-      "loss": 4.184,
+      "loss": 4.1348,
       "step": 4520
     }
   ],