diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,2570 +1,3116 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9876543209876543,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 363,
+  "global_step": 441,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.00823045267489712,
-      "grad_norm": 0.5614587726392962,
-      "learning_rate": 1.8181818181818183e-06,
-      "loss": 1.1353,
+      "epoch": 0.006802721088435374,
+      "grad_norm": 0.7543969951894244,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.9624,
       "step": 1
     },
     {
-      "epoch": 0.01646090534979424,
-      "grad_norm": 0.483289005719831,
-      "learning_rate": 3.6363636363636366e-06,
-      "loss": 1.0151,
+      "epoch": 0.013605442176870748,
+      "grad_norm": 0.44681902466146967,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.964,
       "step": 2
     },
     {
-      "epoch": 0.024691358024691357,
-      "grad_norm": 0.49045996521726837,
-      "learning_rate": 5.4545454545454545e-06,
-      "loss": 1.0192,
+      "epoch": 0.02040816326530612,
+      "grad_norm": 0.5151164741690802,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.0721,
       "step": 3
     },
     {
-      "epoch": 0.03292181069958848,
-      "grad_norm": 0.49189235549026494,
-      "learning_rate": 7.272727272727273e-06,
-      "loss": 1.0097,
+      "epoch": 0.027210884353741496,
+      "grad_norm": 0.5012473323244383,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.0208,
       "step": 4
     },
     {
-      "epoch": 0.0411522633744856,
-      "grad_norm": 0.49355540827357086,
-      "learning_rate": 9.090909090909091e-06,
-      "loss": 1.015,
+      "epoch": 0.034013605442176874,
+      "grad_norm": 0.47766942036038285,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.987,
       "step": 5
     },
     {
-      "epoch": 0.04938271604938271,
-      "grad_norm": 0.44300361161405877,
-      "learning_rate": 1.0909090909090909e-05,
-      "loss": 0.9427,
+      "epoch": 0.04081632653061224,
+      "grad_norm": 0.4930453957288559,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.0436,
       "step": 6
     },
     {
-      "epoch": 0.05761316872427984,
-      "grad_norm": 0.46007061319263887,
-      "learning_rate": 1.2727272727272728e-05,
-      "loss": 0.9611,
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.4570728289003672,
+      "learning_rate": 1e-05,
+      "loss": 0.9853,
       "step": 7
     },
     {
-      "epoch": 0.06584362139917696,
-      "grad_norm": 0.4496385962766769,
-      "learning_rate": 1.4545454545454546e-05,
-      "loss": 0.9448,
+      "epoch": 0.05442176870748299,
+      "grad_norm": 0.39629873030193696,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.8791,
       "step": 8
     },
     {
-      "epoch": 0.07407407407407407,
-      "grad_norm": 0.42419186524470937,
-      "learning_rate": 1.6363636363636366e-05,
-      "loss": 0.9037,
+      "epoch": 0.061224489795918366,
+      "grad_norm": 0.49589342201146847,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 1.0681,
       "step": 9
     },
     {
-      "epoch": 0.0823045267489712,
-      "grad_norm": 0.4097499692227412,
-      "learning_rate": 1.8181818181818182e-05,
-      "loss": 0.902,
+      "epoch": 0.06802721088435375,
+      "grad_norm": 0.39261176159121086,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.9111,
       "step": 10
     },
     {
-      "epoch": 0.09053497942386832,
-      "grad_norm": 0.8591537746989959,
-      "learning_rate": 2e-05,
-      "loss": 1.0221,
+      "epoch": 0.07482993197278912,
+      "grad_norm": 0.3716281292996566,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.8437,
       "step": 11
     },
     {
-      "epoch": 0.09876543209876543,
-      "grad_norm": 0.3815231813089009,
-      "learning_rate": 1.9999601726381415e-05,
-      "loss": 0.8661,
+      "epoch": 0.08163265306122448,
+      "grad_norm": 0.4413102077225305,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.9862,
       "step": 12
     },
     {
-      "epoch": 0.10699588477366255,
-      "grad_norm": 0.4037638540922456,
-      "learning_rate": 1.9998406937250035e-05,
-      "loss": 0.929,
+      "epoch": 0.08843537414965986,
+      "grad_norm": 0.42728639625517184,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 0.9318,
       "step": 13
     },
     {
-      "epoch": 0.11522633744855967,
-      "grad_norm": 0.3571057677122664,
-      "learning_rate": 1.9996415727776456e-05,
-      "loss": 0.7979,
+      "epoch": 0.09523809523809523,
+      "grad_norm": 0.3865701816492697,
+      "learning_rate": 2e-05,
+      "loss": 0.8794,
       "step": 14
     },
     {
-      "epoch": 0.12345679012345678,
-      "grad_norm": 0.36471196489511565,
-      "learning_rate": 1.999362825656992e-05,
-      "loss": 0.7981,
+      "epoch": 0.10204081632653061,
+      "grad_norm": 0.40870933803528964,
+      "learning_rate": 1.9999729347501484e-05,
+      "loss": 0.9287,
       "step": 15
     },
     {
-      "epoch": 0.13168724279835392,
-      "grad_norm": 0.3784154545749015,
-      "learning_rate": 1.9990044745665672e-05,
-      "loss": 0.8351,
+      "epoch": 0.10884353741496598,
+      "grad_norm": 0.3488567698718625,
+      "learning_rate": 1.9998917404656488e-05,
+      "loss": 0.802,
       "step": 16
     },
     {
-      "epoch": 0.13991769547325103,
-      "grad_norm": 0.3693847612005228,
-      "learning_rate": 1.998566548050729e-05,
-      "loss": 0.794,
+      "epoch": 0.11564625850340136,
+      "grad_norm": 0.32774310053561123,
+      "learning_rate": 1.9997564215415886e-05,
+      "loss": 0.7316,
       "step": 17
     },
     {
-      "epoch": 0.14814814814814814,
-      "grad_norm": 0.38170896488444844,
-      "learning_rate": 1.9980490809923928e-05,
-      "loss": 0.7724,
+      "epoch": 0.12244897959183673,
+      "grad_norm": 0.3962817834034302,
+      "learning_rate": 1.9995669853028485e-05,
+      "loss": 0.8542,
       "step": 18
     },
     {
-      "epoch": 0.15637860082304528,
-      "grad_norm": 0.3461401581424495,
-      "learning_rate": 1.9974521146102535e-05,
-      "loss": 0.7123,
+      "epoch": 0.1292517006802721,
+      "grad_norm": 0.3339991775697425,
+      "learning_rate": 1.9993234420037072e-05,
+      "loss": 0.7316,
       "step": 19
     },
     {
-      "epoch": 0.1646090534979424,
-      "grad_norm": 0.3461401581424495,
-      "learning_rate": 1.9974521146102535e-05,
-      "loss": 0.6782,
+      "epoch": 0.1360544217687075,
+      "grad_norm": 0.3613596947800494,
+      "learning_rate": 1.999025804827285e-05,
+      "loss": 0.7685,
       "step": 20
     },
     {
-      "epoch": 0.1728395061728395,
-      "grad_norm": 0.31652407939469995,
-      "learning_rate": 1.9967756964555044e-05,
-      "loss": 0.6283,
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.3427806922992633,
+      "learning_rate": 1.9986740898848306e-05,
+      "loss": 0.7005,
       "step": 21
     },
     {
-      "epoch": 0.18106995884773663,
-      "grad_norm": 0.3736799183333394,
-      "learning_rate": 1.9960198804080462e-05,
-      "loss": 0.7145,
+      "epoch": 0.14965986394557823,
+      "grad_norm": 0.34368082813141,
+      "learning_rate": 1.99826831621485e-05,
+      "loss": 0.6939,
       "step": 22
     },
     {
-      "epoch": 0.18930041152263374,
-      "grad_norm": 0.3384946983997503,
-      "learning_rate": 1.995184726672197e-05,
-      "loss": 0.6347,
+      "epoch": 0.1564625850340136,
+      "grad_norm": 0.3300141228352275,
+      "learning_rate": 1.997808505782075e-05,
+      "loss": 0.6756,
       "step": 23
     },
     {
-      "epoch": 0.19753086419753085,
-      "grad_norm": 0.34466790351123977,
-      "learning_rate": 1.9942703017718977e-05,
-      "loss": 0.6415,
+      "epoch": 0.16326530612244897,
+      "grad_norm": 0.3440934696087154,
+      "learning_rate": 1.9972946834762732e-05,
+      "loss": 0.6643,
       "step": 24
     },
     {
-      "epoch": 0.205761316872428,
-      "grad_norm": 0.3607668838667792,
-      "learning_rate": 1.99327667854541e-05,
-      "loss": 0.632,
+      "epoch": 0.17006802721088435,
+      "grad_norm": 0.35675869839992724,
+      "learning_rate": 1.9967268771109037e-05,
+      "loss": 0.6623,
       "step": 25
     },
     {
-      "epoch": 0.2139917695473251,
-      "grad_norm": 0.3605475740089386,
-      "learning_rate": 1.9922039361395186e-05,
-      "loss": 0.601,
+      "epoch": 0.17687074829931973,
+      "grad_norm": 0.34399594353909607,
+      "learning_rate": 1.996105117421608e-05,
+      "loss": 0.6182,
       "step": 26
     },
     {
-      "epoch": 0.2222222222222222,
-      "grad_norm": 0.3523858922945964,
-      "learning_rate": 1.991052160003223e-05,
-      "loss": 0.5866,
+      "epoch": 0.1836734693877551,
+      "grad_norm": 0.3634094292173927,
+      "learning_rate": 1.9954294380645497e-05,
+      "loss": 0.6286,
       "step": 27
     },
     {
-      "epoch": 0.23045267489711935,
-      "grad_norm": 0.33706327335583675,
-      "learning_rate": 1.989821441880933e-05,
-      "loss": 0.5301,
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.37888057387883767,
+      "learning_rate": 1.9946998756145894e-05,
+      "loss": 0.6367,
       "step": 28
     },
     {
-      "epoch": 0.23868312757201646,
-      "grad_norm": 0.34471550282355007,
-      "learning_rate": 1.9885118798051607e-05,
-      "loss": 0.5375,
+      "epoch": 0.19727891156462585,
+      "grad_norm": 0.36856282647607674,
+      "learning_rate": 1.9939164695633067e-05,
+      "loss": 0.5763,
       "step": 29
     },
     {
-      "epoch": 0.24691358024691357,
-      "grad_norm": 0.3705674808591847,
-      "learning_rate": 1.9871235780887114e-05,
-      "loss": 0.5271,
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.33791502181439126,
+      "learning_rate": 1.9930792623168638e-05,
+      "loss": 0.5111,
       "step": 30
     },
     {
-      "epoch": 0.2551440329218107,
-      "grad_norm": 0.34300917466537967,
-      "learning_rate": 1.9856566473163747e-05,
-      "loss": 0.4778,
+      "epoch": 0.2108843537414966,
+      "grad_norm": 0.3507039073715319,
+      "learning_rate": 1.992188299193706e-05,
+      "loss": 0.5191,
       "step": 31
     },
     {
-      "epoch": 0.26337448559670784,
-      "grad_norm": 0.3832902951135858,
-      "learning_rate": 1.984111204336116e-05,
-      "loss": 0.5004,
+      "epoch": 0.21768707482993196,
+      "grad_norm": 0.3530833992530626,
+      "learning_rate": 1.9912436284221134e-05,
+      "loss": 0.4889,
       "step": 32
     },
     {
-      "epoch": 0.2716049382716049,
-      "grad_norm": 0.3925208797697874,
-      "learning_rate": 1.9824873722497694e-05,
-      "loss": 0.4941,
+      "epoch": 0.22448979591836735,
+      "grad_norm": 0.3592183944530985,
+      "learning_rate": 1.9902453011375865e-05,
+      "loss": 0.4832,
       "step": 33
     },
     {
-      "epoch": 0.27983539094650206,
-      "grad_norm": 0.3440849883835382,
-      "learning_rate": 1.9807852804032306e-05,
-      "loss": 0.4051,
+      "epoch": 0.23129251700680273,
+      "grad_norm": 0.34918368964463997,
+      "learning_rate": 1.98919337138008e-05,
+      "loss": 0.444,
       "step": 34
     },
     {
-      "epoch": 0.2880658436213992,
-      "grad_norm": 0.3581815755056744,
-      "learning_rate": 1.9790050643761552e-05,
-      "loss": 0.4055,
+      "epoch": 0.23809523809523808,
+      "grad_norm": 0.38954703063221796,
+      "learning_rate": 1.9880878960910772e-05,
+      "loss": 0.4722,
       "step": 35
     },
     {
-      "epoch": 0.2962962962962963,
-      "grad_norm": 0.38137085766827344,
-      "learning_rate": 1.9771468659711595e-05,
-      "loss": 0.4006,
+      "epoch": 0.24489795918367346,
+      "grad_norm": 0.37889828104003,
+      "learning_rate": 1.9869289351105087e-05,
+      "loss": 0.4284,
       "step": 36
     },
     {
-      "epoch": 0.3045267489711934,
-      "grad_norm": 0.4214422944046215,
-      "learning_rate": 1.975210833202524e-05,
-      "loss": 0.4166,
+      "epoch": 0.25170068027210885,
+      "grad_norm": 0.3879345796013965,
+      "learning_rate": 1.9857165511735105e-05,
+      "loss": 0.4192,
       "step": 37
     },
     {
-      "epoch": 0.31275720164609055,
-      "grad_norm": 0.3347215677142363,
-      "learning_rate": 1.9731971202844036e-05,
-      "loss": 0.3259,
+      "epoch": 0.2585034013605442,
+      "grad_norm": 0.36130568820722925,
+      "learning_rate": 1.9844508099070313e-05,
+      "loss": 0.3716,
       "step": 38
     },
     {
-      "epoch": 0.32098765432098764,
-      "grad_norm": 0.4398950031636108,
-      "learning_rate": 1.9711058876185446e-05,
-      "loss": 0.3391,
+      "epoch": 0.2653061224489796,
+      "grad_norm": 0.36791653912252376,
+      "learning_rate": 1.9831317798262787e-05,
+      "loss": 0.3653,
       "step": 39
     },
     {
-      "epoch": 0.3292181069958848,
-      "grad_norm": 0.37618739383861566,
-      "learning_rate": 1.9689373017815076e-05,
-      "loss": 0.2942,
+      "epoch": 0.272108843537415,
+      "grad_norm": 0.3761185645158447,
+      "learning_rate": 1.98175953233101e-05,
+      "loss": 0.3465,
       "step": 40
     },
     {
-      "epoch": 0.3374485596707819,
-      "grad_norm": 0.3834784119471882,
-      "learning_rate": 1.9666915355113976e-05,
-      "loss": 0.2983,
+      "epoch": 0.2789115646258503,
+      "grad_norm": 0.4350230270095709,
+      "learning_rate": 1.980334141701667e-05,
+      "loss": 0.348,
       "step": 41
     },
     {
-      "epoch": 0.345679012345679,
-      "grad_norm": 0.43100085029424373,
-      "learning_rate": 1.964368767694107e-05,
-      "loss": 0.2924,
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.32213563899302,
+      "learning_rate": 1.978855685095358e-05,
+      "loss": 0.2679,
       "step": 42
     },
     {
-      "epoch": 0.35390946502057613,
-      "grad_norm": 0.3691111882891019,
-      "learning_rate": 1.9619691833490645e-05,
-      "loss": 0.2535,
+      "epoch": 0.2925170068027211,
+      "grad_norm": 0.3768678079788929,
+      "learning_rate": 1.977324242541677e-05,
+      "loss": 0.2842,
       "step": 43
     },
     {
-      "epoch": 0.36213991769547327,
-      "grad_norm": 0.39507998692048524,
-      "learning_rate": 1.9594929736144978e-05,
-      "loss": 0.2507,
+      "epoch": 0.29931972789115646,
+      "grad_norm": 0.3675295235354873,
+      "learning_rate": 1.9757398969383752e-05,
+      "loss": 0.25,
       "step": 44
     },
     {
-      "epoch": 0.37037037037037035,
-      "grad_norm": 0.37490188815841785,
-      "learning_rate": 1.956940335732209e-05,
-      "loss": 0.2242,
+      "epoch": 0.30612244897959184,
+      "grad_norm": 0.3702046286490698,
+      "learning_rate": 1.974102734046872e-05,
+      "loss": 0.2398,
       "step": 45
     },
     {
-      "epoch": 0.3786008230452675,
-      "grad_norm": 0.36721511207034846,
-      "learning_rate": 1.954311473031864e-05,
-      "loss": 0.1988,
+      "epoch": 0.3129251700680272,
+      "grad_norm": 0.39377885956971614,
+      "learning_rate": 1.9724128424876117e-05,
+      "loss": 0.2288,
       "step": 46
     },
     {
-      "epoch": 0.3868312757201646,
-      "grad_norm": 0.3415062420750618,
-      "learning_rate": 1.9516065949147945e-05,
-      "loss": 0.1776,
+      "epoch": 0.3197278911564626,
+      "grad_norm": 0.35607191659617016,
+      "learning_rate": 1.9706703137352695e-05,
+      "loss": 0.1986,
       "step": 47
     },
     {
-      "epoch": 0.3950617283950617,
-      "grad_norm": 0.40710342267665045,
-      "learning_rate": 1.9488259168373198e-05,
-      "loss": 0.1748,
+      "epoch": 0.32653061224489793,
+      "grad_norm": 0.37797721697894693,
+      "learning_rate": 1.968875242113798e-05,
+      "loss": 0.1921,
       "step": 48
     },
     {
-      "epoch": 0.40329218106995884,
-      "grad_norm": 0.46050780516822537,
-      "learning_rate": 1.9459696602935838e-05,
-      "loss": 0.1662,
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.4458500482446309,
+      "learning_rate": 1.9670277247913205e-05,
+      "loss": 0.1808,
       "step": 49
     },
     {
-      "epoch": 0.411522633744856,
-      "grad_norm": 0.5468567654504426,
-      "learning_rate": 1.9430380527979124e-05,
-      "loss": 0.1659,
+      "epoch": 0.3401360544217687,
+      "grad_norm": 0.4904856763671931,
+      "learning_rate": 1.965127861774873e-05,
+      "loss": 0.178,
       "step": 50
     },
     {
-      "epoch": 0.41975308641975306,
-      "grad_norm": 0.46777963678521145,
-      "learning_rate": 1.94003132786669e-05,
-      "loss": 0.1406,
+      "epoch": 0.3469387755102041,
+      "grad_norm": 0.42047437091507667,
+      "learning_rate": 1.96317575590499e-05,
+      "loss": 0.1556,
       "step": 51
     },
     {
-      "epoch": 0.4279835390946502,
-      "grad_norm": 0.4307828961334118,
-      "learning_rate": 1.936949724999762e-05,
-      "loss": 0.1274,
+      "epoch": 0.35374149659863946,
+      "grad_norm": 0.3659087858039602,
+      "learning_rate": 1.9611715128501378e-05,
+      "loss": 0.1316,
       "step": 52
     },
     {
-      "epoch": 0.43621399176954734,
-      "grad_norm": 0.5510758752645835,
-      "learning_rate": 1.9337934896613516e-05,
-      "loss": 0.1308,
+      "epoch": 0.36054421768707484,
+      "grad_norm": 0.41650852039639386,
+      "learning_rate": 1.9591152411009942e-05,
+      "loss": 0.1366,
       "step": 53
     },
     {
-      "epoch": 0.4444444444444444,
-      "grad_norm": 0.3532647331561503,
-      "learning_rate": 1.930562873260514e-05,
-      "loss": 0.1111,
+      "epoch": 0.3673469387755102,
+      "grad_norm": 0.42327758680290756,
+      "learning_rate": 1.9570070519645767e-05,
+      "loss": 0.1239,
       "step": 54
     },
     {
-      "epoch": 0.45267489711934156,
-      "grad_norm": 0.3646056510337883,
-      "learning_rate": 1.927258133131105e-05,
-      "loss": 0.1118,
+      "epoch": 0.3741496598639456,
+      "grad_norm": 0.4263037625862704,
+      "learning_rate": 1.9548470595582166e-05,
+      "loss": 0.1226,
       "step": 55
     },
     {
-      "epoch": 0.4609053497942387,
-      "grad_norm": 0.35329494842052633,
-      "learning_rate": 1.9238795325112867e-05,
-      "loss": 0.102,
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.4203561676384866,
+      "learning_rate": 1.9526353808033827e-05,
+      "loss": 0.1203,
       "step": 56
     },
     {
-      "epoch": 0.4691358024691358,
-      "grad_norm": 0.3556881186952675,
-      "learning_rate": 1.9204273405225588e-05,
-      "loss": 0.102,
+      "epoch": 0.3877551020408163,
+      "grad_norm": 0.3960865903458932,
+      "learning_rate": 1.9503721354193507e-05,
+      "loss": 0.1121,
       "step": 57
     },
     {
-      "epoch": 0.4773662551440329,
-      "grad_norm": 0.31364984090906534,
-      "learning_rate": 1.9169018321483198e-05,
-      "loss": 0.0975,
+      "epoch": 0.3945578231292517,
+      "grad_norm": 0.3811191794773781,
+      "learning_rate": 1.948057445916724e-05,
+      "loss": 0.1095,
       "step": 58
     },
     {
-      "epoch": 0.48559670781893005,
-      "grad_norm": 0.2918862188752059,
-      "learning_rate": 1.9133032882119656e-05,
-      "loss": 0.0926,
+      "epoch": 0.4013605442176871,
+      "grad_norm": 0.34427762247965304,
+      "learning_rate": 1.9456914375908026e-05,
+      "loss": 0.1019,
       "step": 59
     },
     {
-      "epoch": 0.49382716049382713,
-      "grad_norm": 0.303152901681533,
-      "learning_rate": 1.9096319953545186e-05,
-      "loss": 0.094,
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.3250633041475826,
+      "learning_rate": 1.9432742385147988e-05,
+      "loss": 0.0961,
       "step": 60
     },
     {
-      "epoch": 0.5020576131687243,
-      "grad_norm": 0.31149744809113017,
-      "learning_rate": 1.9058882460117972e-05,
-      "loss": 0.0921,
+      "epoch": 0.41496598639455784,
+      "grad_norm": 0.3391653472941772,
+      "learning_rate": 1.9408059795329073e-05,
+      "loss": 0.0907,
       "step": 61
     },
     {
-      "epoch": 0.5102880658436214,
-      "grad_norm": 0.2788528854959893,
-      "learning_rate": 1.9020723383911214e-05,
-      "loss": 0.0889,
+      "epoch": 0.4217687074829932,
+      "grad_norm": 0.3071859632068275,
+      "learning_rate": 1.9382867942532195e-05,
+      "loss": 0.0887,
       "step": 62
     },
     {
-      "epoch": 0.5185185185185185,
-      "grad_norm": 0.27906401151493954,
-      "learning_rate": 1.8981845764475585e-05,
-      "loss": 0.085,
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.28368480944915603,
+      "learning_rate": 1.9357168190404937e-05,
+      "loss": 0.0844,
       "step": 63
     },
     {
-      "epoch": 0.5267489711934157,
-      "grad_norm": 0.20914730450528948,
-      "learning_rate": 1.8942252698597113e-05,
-      "loss": 0.0812,
+      "epoch": 0.43537414965986393,
+      "grad_norm": 0.37481710014760405,
+      "learning_rate": 1.9330961930087724e-05,
+      "loss": 0.0891,
       "step": 64
     },
     {
-      "epoch": 0.5349794238683128,
-      "grad_norm": 0.2657919670053626,
-      "learning_rate": 1.890194734005053e-05,
-      "loss": 0.0805,
+      "epoch": 0.4421768707482993,
+      "grad_norm": 0.3263398981365655,
+      "learning_rate": 1.9304250580138524e-05,
+      "loss": 0.0866,
       "step": 65
     },
     {
-      "epoch": 0.5432098765432098,
-      "grad_norm": 0.2207043697689934,
-      "learning_rate": 1.8860932899348028e-05,
-      "loss": 0.0811,
+      "epoch": 0.4489795918367347,
+      "grad_norm": 0.2880232601443931,
+      "learning_rate": 1.9277035586456056e-05,
+      "loss": 0.0851,
       "step": 66
     },
     {
-      "epoch": 0.551440329218107,
-      "grad_norm": 0.20662366240442817,
-      "learning_rate": 1.881921264348355e-05,
-      "loss": 0.0731,
+      "epoch": 0.4557823129251701,
+      "grad_norm": 0.38682915757490255,
+      "learning_rate": 1.9249318422201524e-05,
+      "loss": 0.0813,
       "step": 67
     },
     {
-      "epoch": 0.5596707818930041,
-      "grad_norm": 0.23805984958759943,
-      "learning_rate": 1.8776789895672557e-05,
-      "loss": 0.0762,
+      "epoch": 0.46258503401360546,
+      "grad_norm": 0.20937614024944473,
+      "learning_rate": 1.9221100587718884e-05,
+      "loss": 0.0782,
       "step": 68
     },
     {
-      "epoch": 0.5679012345679012,
-      "grad_norm": 0.23465995041715484,
-      "learning_rate": 1.8733668035087302e-05,
-      "loss": 0.0777,
+      "epoch": 0.46938775510204084,
+      "grad_norm": 0.24997247747455145,
+      "learning_rate": 1.919238361045362e-05,
+      "loss": 0.0764,
       "step": 69
     },
     {
-      "epoch": 0.5761316872427984,
-      "grad_norm": 0.2432037840691274,
-      "learning_rate": 1.8689850496587674e-05,
-      "loss": 0.0758,
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.23920563707712375,
+      "learning_rate": 1.916316904487005e-05,
+      "loss": 0.0783,
       "step": 70
     },
     {
-      "epoch": 0.5843621399176955,
-      "grad_norm": 0.2529033680049068,
-      "learning_rate": 1.8645340770447595e-05,
-      "loss": 0.0733,
+      "epoch": 0.48299319727891155,
+      "grad_norm": 0.23894151439100778,
+      "learning_rate": 1.9133458472367216e-05,
+      "loss": 0.0761,
       "step": 71
     },
     {
-      "epoch": 0.5925925925925926,
-      "grad_norm": 0.19602645989777068,
-      "learning_rate": 1.8600142402077006e-05,
-      "loss": 0.0722,
+      "epoch": 0.4897959183673469,
+      "grad_norm": 0.24006873105227072,
+      "learning_rate": 1.9103253501193256e-05,
+      "loss": 0.076,
       "step": 72
     },
     {
-      "epoch": 0.6008230452674898,
-      "grad_norm": 0.2672568735624594,
-      "learning_rate": 1.8554258991739454e-05,
-      "loss": 0.0744,
+      "epoch": 0.4965986394557823,
+      "grad_norm": 0.2551200414134792,
+      "learning_rate": 1.9072555766358346e-05,
+      "loss": 0.0766,
       "step": 73
     },
     {
-      "epoch": 0.6090534979423868,
-      "grad_norm": 0.21656686479994547,
-      "learning_rate": 1.850769419426531e-05,
-      "loss": 0.0707,
+      "epoch": 0.5034013605442177,
+      "grad_norm": 0.22157625780163037,
+      "learning_rate": 1.904136692954622e-05,
+      "loss": 0.0736,
       "step": 74
     },
     {
-      "epoch": 0.6172839506172839,
-      "grad_norm": 0.22333708041398428,
-      "learning_rate": 1.8460451718760653e-05,
-      "loss": 0.0697,
+      "epoch": 0.5102040816326531,
+      "grad_norm": 0.22027442259887026,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.0714,
       "step": 75
     },
     {
-      "epoch": 0.6255144032921811,
-      "grad_norm": 0.496507124725327,
-      "learning_rate": 1.8412535328311813e-05,
-      "loss": 0.0745,
+      "epoch": 0.5170068027210885,
+      "grad_norm": 0.18204627775070914,
+      "learning_rate": 1.89775227295518e-05,
+      "loss": 0.0714,
       "step": 76
     },
     {
-      "epoch": 0.6337448559670782,
-      "grad_norm": 0.2375236493655952,
-      "learning_rate": 1.8363948839685638e-05,
-      "loss": 0.0671,
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.2620995135551117,
+      "learning_rate": 1.8944870822287957e-05,
+      "loss": 0.0729,
       "step": 77
     },
     {
-      "epoch": 0.6419753086419753,
-      "grad_norm": 0.2929529684537933,
-      "learning_rate": 1.8314696123025456e-05,
-      "loss": 0.0688,
+      "epoch": 0.5306122448979592,
+      "grad_norm": 0.29878189909698066,
+      "learning_rate": 1.891173472469672e-05,
+      "loss": 0.0744,
       "step": 78
     },
     {
-      "epoch": 0.6502057613168725,
-      "grad_norm": 0.22745788734012945,
-      "learning_rate": 1.8264781101542797e-05,
-      "loss": 0.0673,
+      "epoch": 0.5374149659863946,
+      "grad_norm": 0.20917807504184172,
+      "learning_rate": 1.8878116230451615e-05,
+      "loss": 0.0709,
       "step": 79
     },
     {
-      "epoch": 0.6584362139917695,
-      "grad_norm": 0.28594030546665017,
-      "learning_rate": 1.8214207751204917e-05,
-      "loss": 0.0688,
+      "epoch": 0.54421768707483,
+      "grad_norm": 0.2280652332605586,
+      "learning_rate": 1.884401715933853e-05,
+      "loss": 0.0691,
       "step": 80
     },
     {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 0.20264319707616257,
-      "learning_rate": 1.816298010041806e-05,
-      "loss": 0.0627,
+      "epoch": 0.5510204081632653,
+      "grad_norm": 0.16319230454307865,
+      "learning_rate": 1.8809439357157226e-05,
+      "loss": 0.0664,
       "step": 81
     },
     {
-      "epoch": 0.6748971193415638,
-      "grad_norm": 0.20693129554679163,
-      "learning_rate": 1.8111102229706593e-05,
-      "loss": 0.0678,
+      "epoch": 0.5578231292517006,
+      "grad_norm": 0.17045276025315298,
+      "learning_rate": 1.8774384695621407e-05,
+      "loss": 0.0672,
       "step": 82
     },
     {
-      "epoch": 0.6831275720164609,
-      "grad_norm": 0.30583374645284117,
-      "learning_rate": 1.805857827138798e-05,
+      "epoch": 0.564625850340136,
+      "grad_norm": 0.18557421713410696,
+      "learning_rate": 1.8738855072257428e-05,
       "loss": 0.0684,
       "step": 83
     },
     {
-      "epoch": 0.691358024691358,
-      "grad_norm": 0.23136460120570432,
-      "learning_rate": 1.8005412409243604e-05,
-      "loss": 0.0642,
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.1857928022764888,
+      "learning_rate": 1.8702852410301556e-05,
+      "loss": 0.0713,
       "step": 84
     },
     {
-      "epoch": 0.6995884773662552,
-      "grad_norm": 0.17354239204762031,
-      "learning_rate": 1.7951608878185533e-05,
-      "loss": 0.0655,
+      "epoch": 0.5782312925170068,
+      "grad_norm": 0.19053752905827265,
+      "learning_rate": 1.8666378658595863e-05,
+      "loss": 0.067,
       "step": 85
     },
     {
-      "epoch": 0.7078189300411523,
-      "grad_norm": 0.23756490177404446,
-      "learning_rate": 1.789717196391916e-05,
-      "loss": 0.0678,
+      "epoch": 0.5850340136054422,
+      "grad_norm": 0.26375035657754126,
+      "learning_rate": 1.8629435791482765e-05,
+      "loss": 0.0643,
       "step": 86
     },
     {
-      "epoch": 0.7160493827160493,
-      "grad_norm": 0.19874211076804596,
-      "learning_rate": 1.7842106002601854e-05,
-      "loss": 0.0665,
+      "epoch": 0.5918367346938775,
+      "grad_norm": 0.20416646049963377,
+      "learning_rate": 1.8592025808698116e-05,
+      "loss": 0.0664,
       "step": 87
     },
     {
-      "epoch": 0.7242798353909465,
-      "grad_norm": 0.19532547769316078,
-      "learning_rate": 1.778641538049755e-05,
-      "loss": 0.0658,
+      "epoch": 0.5986394557823129,
+      "grad_norm": 0.22529978535979228,
+      "learning_rate": 1.8554150735262975e-05,
+      "loss": 0.0666,
       "step": 88
     },
     {
-      "epoch": 0.7325102880658436,
-      "grad_norm": 0.22412310657600473,
-      "learning_rate": 1.773010453362737e-05,
-      "loss": 0.0673,
+      "epoch": 0.6054421768707483,
+      "grad_norm": 0.2343565249290847,
+      "learning_rate": 1.8515812621373998e-05,
+      "loss": 0.0671,
       "step": 89
     },
     {
-      "epoch": 0.7407407407407407,
-      "grad_norm": 0.3132368525031726,
-      "learning_rate": 1.7673177947416258e-05,
-      "loss": 0.0658,
+      "epoch": 0.6122448979591837,
+      "grad_norm": 0.23897319920733306,
+      "learning_rate": 1.8477013542292446e-05,
+      "loss": 0.066,
       "step": 90
     },
     {
-      "epoch": 0.7489711934156379,
-      "grad_norm": 0.19395421272200608,
-      "learning_rate": 1.7615640156335713e-05,
-      "loss": 0.0632,
+      "epoch": 0.6190476190476191,
+      "grad_norm": 0.24710547661980992,
+      "learning_rate": 1.8437755598231857e-05,
+      "loss": 0.0657,
       "step": 91
     },
     {
-      "epoch": 0.757201646090535,
-      "grad_norm": 0.28326260409072385,
-      "learning_rate": 1.7557495743542586e-05,
-      "loss": 0.0627,
+      "epoch": 0.6258503401360545,
+      "grad_norm": 0.25014419464910725,
+      "learning_rate": 1.8398040914244363e-05,
+      "loss": 0.0684,
       "step": 92
     },
     {
-      "epoch": 0.7654320987654321,
-      "grad_norm": 0.2546519603541441,
-      "learning_rate": 1.749874934051401e-05,
-      "loss": 0.0647,
+      "epoch": 0.6326530612244898,
+      "grad_norm": 0.17847236067983924,
+      "learning_rate": 1.8357871640105648e-05,
+      "loss": 0.0631,
       "step": 93
     },
     {
-      "epoch": 0.7736625514403292,
-      "grad_norm": 0.1539191916526331,
-      "learning_rate": 1.7439405626678496e-05,
-      "loss": 0.0611,
+      "epoch": 0.6394557823129252,
+      "grad_norm": 0.24262416375321524,
+      "learning_rate": 1.8317249950198598e-05,
+      "loss": 0.0657,
       "step": 94
     },
     {
-      "epoch": 0.7818930041152263,
-      "grad_norm": 0.18268548893439074,
-      "learning_rate": 1.7379469329043166e-05,
-      "loss": 0.0629,
+      "epoch": 0.6462585034013606,
+      "grad_norm": 0.19587079364109738,
+      "learning_rate": 1.8276178043395588e-05,
+      "loss": 0.0661,
       "step": 95
     },
     {
-      "epoch": 0.7901234567901234,
-      "grad_norm": 0.24868547007332967,
-      "learning_rate": 1.7318945221817255e-05,
-      "loss": 0.0647,
+      "epoch": 0.6530612244897959,
+      "grad_norm": 0.2511062200421889,
+      "learning_rate": 1.8234658142939454e-05,
+      "loss": 0.0626,
       "step": 96
     },
     {
-      "epoch": 0.7983539094650206,
-      "grad_norm": 0.24898521909085292,
-      "learning_rate": 1.7257838126031797e-05,
-      "loss": 0.0613,
+      "epoch": 0.6598639455782312,
+      "grad_norm": 0.19039392887739032,
+      "learning_rate": 1.8192692496323158e-05,
+      "loss": 0.0645,
       "step": 97
     },
     {
-      "epoch": 0.8065843621399177,
-      "grad_norm": 0.1743699589376782,
-      "learning_rate": 1.719615290915563e-05,
-      "loss": 0.06,
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.1999976259553887,
+      "learning_rate": 1.8150283375168112e-05,
+      "loss": 0.0657,
       "step": 98
     },
     {
-      "epoch": 0.8148148148148148,
-      "grad_norm": 0.22229965921820366,
-      "learning_rate": 1.7133894484707657e-05,
-      "loss": 0.0652,
+      "epoch": 0.673469387755102,
+      "grad_norm": 0.20368044763239615,
+      "learning_rate": 1.8107433075101254e-05,
+      "loss": 0.0665,
       "step": 99
     },
     {
-      "epoch": 0.823045267489712,
-      "grad_norm": 0.2630979598727429,
-      "learning_rate": 1.7071067811865477e-05,
-      "loss": 0.0615,
+      "epoch": 0.6802721088435374,
+      "grad_norm": 0.22553964175771693,
+      "learning_rate": 1.8064143915630723e-05,
+      "loss": 0.0661,
       "step": 100
     },
     {
-      "epoch": 0.831275720164609,
-      "grad_norm": 0.3877901930196216,
-      "learning_rate": 1.7007677895070358e-05,
-      "loss": 0.0646,
+      "epoch": 0.6870748299319728,
+      "grad_norm": 0.28951843622254164,
+      "learning_rate": 1.8020418240020362e-05,
+      "loss": 0.069,
       "step": 101
     },
     {
-      "epoch": 0.8395061728395061,
-      "grad_norm": 0.21159933483060928,
-      "learning_rate": 1.694372978362861e-05,
-      "loss": 0.0593,
+      "epoch": 0.6938775510204082,
+      "grad_norm": 0.2057412466161773,
+      "learning_rate": 1.7976258415162836e-05,
+      "loss": 0.0618,
       "step": 102
     },
     {
-      "epoch": 0.8477366255144033,
-      "grad_norm": 0.21650980432319827,
-      "learning_rate": 1.6879228571309377e-05,
-      "loss": 0.0638,
+      "epoch": 0.7006802721088435,
+      "grad_norm": 0.1836029567698062,
+      "learning_rate": 1.7931666831451536e-05,
+      "loss": 0.064,
       "step": 103
     },
     {
-      "epoch": 0.8559670781893004,
-      "grad_norm": 0.2082928545520928,
-      "learning_rate": 1.6814179395938915e-05,
-      "loss": 0.0614,
+      "epoch": 0.7074829931972789,
+      "grad_norm": 0.219074746713116,
+      "learning_rate": 1.7886645902651166e-05,
+      "loss": 0.0629,
       "step": 104
     },
     {
-      "epoch": 0.8641975308641975,
-      "grad_norm": 0.21356406026718303,
-      "learning_rate": 1.6748587438991303e-05,
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.1826642649944216,
+      "learning_rate": 1.7841198065767107e-05,
       "loss": 0.06,
       "step": 105
     },
     {
-      "epoch": 0.8724279835390947,
-      "grad_norm": 0.2167524487044743,
-      "learning_rate": 1.6682457925175762e-05,
-      "loss": 0.0614,
+      "epoch": 0.7210884353741497,
+      "grad_norm": 0.14348426307016002,
+      "learning_rate": 1.779532578091347e-05,
+      "loss": 0.0598,
       "step": 106
     },
     {
-      "epoch": 0.8806584362139918,
-      "grad_norm": 0.17102223171072273,
-      "learning_rate": 1.6615796122020443e-05,
-      "loss": 0.061,
+      "epoch": 0.7278911564625851,
+      "grad_norm": 0.1892487521305465,
+      "learning_rate": 1.7749031531179962e-05,
+      "loss": 0.0608,
       "step": 107
     },
     {
-      "epoch": 0.8888888888888888,
-      "grad_norm": 0.19479620476762075,
-      "learning_rate": 1.6548607339452853e-05,
-      "loss": 0.0617,
+      "epoch": 0.7346938775510204,
+      "grad_norm": 0.1852039687458771,
+      "learning_rate": 1.7702317822497457e-05,
+      "loss": 0.0612,
       "step": 108
     },
     {
-      "epoch": 0.897119341563786,
-      "grad_norm": 0.1749966663248173,
-      "learning_rate": 1.6480896929376905e-05,
-      "loss": 0.0569,
+      "epoch": 0.7414965986394558,
+      "grad_norm": 0.20838595163721002,
+      "learning_rate": 1.7655187183502344e-05,
+      "loss": 0.0636,
       "step": 109
     },
     {
-      "epoch": 0.9053497942386831,
-      "grad_norm": 0.34196359928268927,
-      "learning_rate": 1.641267028524661e-05,
-      "loss": 0.0595,
+      "epoch": 0.7482993197278912,
+      "grad_norm": 0.1981664363252372,
+      "learning_rate": 1.7607642165399665e-05,
+      "loss": 0.0586,
       "step": 110
     },
     {
-      "epoch": 0.9135802469135802,
-      "grad_norm": 0.1788419208008274,
-      "learning_rate": 1.6343932841636455e-05,
-      "loss": 0.0602,
+      "epoch": 0.7551020408163265,
+      "grad_norm": 0.2820295964012395,
+      "learning_rate": 1.755968534182501e-05,
+      "loss": 0.0627,
       "step": 111
     },
     {
-      "epoch": 0.9218106995884774,
-      "grad_norm": 0.2350017500402357,
-      "learning_rate": 1.627469007380852e-05,
-      "loss": 0.0599,
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.16519458200620668,
+      "learning_rate": 1.7511319308705198e-05,
+      "loss": 0.059,
       "step": 112
     },
     {
-      "epoch": 0.9300411522633745,
-      "grad_norm": 0.3414508983933339,
-      "learning_rate": 1.6204947497276346e-05,
-      "loss": 0.0617,
+      "epoch": 0.7687074829931972,
+      "grad_norm": 0.26268272830133443,
+      "learning_rate": 1.746254668411778e-05,
+      "loss": 0.0649,
       "step": 113
     },
     {
-      "epoch": 0.9382716049382716,
-      "grad_norm": 0.4326608785185992,
-      "learning_rate": 1.6134710667365598e-05,
-      "loss": 0.0609,
+      "epoch": 0.7755102040816326,
+      "grad_norm": 0.19201979033727093,
+      "learning_rate": 1.7413370108149288e-05,
+      "loss": 0.0639,
       "step": 114
     },
     {
-      "epoch": 0.9465020576131687,
-      "grad_norm": 0.2354361721822415,
-      "learning_rate": 1.6063985178771555e-05,
-      "loss": 0.0585,
+      "epoch": 0.782312925170068,
+      "grad_norm": 0.23222721849071837,
+      "learning_rate": 1.7363792242752354e-05,
+      "loss": 0.0611,
       "step": 115
     },
     {
-      "epoch": 0.9547325102880658,
-      "grad_norm": 0.2922490262824464,
-      "learning_rate": 1.599277666511347e-05,
-      "loss": 0.0558,
+      "epoch": 0.7891156462585034,
+      "grad_norm": 0.19754689775387121,
+      "learning_rate": 1.731381577160161e-05,
+      "loss": 0.0611,
       "step": 116
     },
     {
-      "epoch": 0.9629629629629629,
-      "grad_norm": 0.19407491372408675,
-      "learning_rate": 1.592109079848583e-05,
-      "loss": 0.0596,
+      "epoch": 0.7959183673469388,
+      "grad_norm": 0.26191489705008636,
+      "learning_rate": 1.726344339994841e-05,
+      "loss": 0.0628,
       "step": 117
     },
     {
-      "epoch": 0.9711934156378601,
-      "grad_norm": 0.30993778242867026,
-      "learning_rate": 1.584893328900653e-05,
-      "loss": 0.0626,
+      "epoch": 0.8027210884353742,
+      "grad_norm": 0.22002163098655456,
+      "learning_rate": 1.7212677854474402e-05,
+      "loss": 0.0597,
       "step": 118
     },
     {
-      "epoch": 0.9794238683127572,
-      "grad_norm": 0.2775225647285507,
-      "learning_rate": 1.577630988436206e-05,
-      "loss": 0.0542,
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.21168658257559084,
+      "learning_rate": 1.7161521883143936e-05,
+      "loss": 0.06,
       "step": 119
     },
     {
-      "epoch": 0.9876543209876543,
-      "grad_norm": 0.2713799446677433,
-      "learning_rate": 1.5703226369349642e-05,
-      "loss": 0.0599,
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.2222174632266417,
+      "learning_rate": 1.7109978255055295e-05,
+      "loss": 0.0603,
       "step": 120
     },
     {
-      "epoch": 0.9958847736625515,
-      "grad_norm": 0.2338145265628359,
-      "learning_rate": 1.562968856541648e-05,
-      "loss": 0.0581,
+      "epoch": 0.8231292517006803,
+      "grad_norm": 0.2520406444086263,
+      "learning_rate": 1.705804976029083e-05,
+      "loss": 0.0618,
       "step": 121
     },
     {
-      "epoch": 1.0041152263374487,
-      "grad_norm": 0.23843906135728044,
-      "learning_rate": 1.5555702330196024e-05,
-      "loss": 0.0496,
+      "epoch": 0.8299319727891157,
+      "grad_norm": 0.1873330783063175,
+      "learning_rate": 1.7005739209765906e-05,
+      "loss": 0.0599,
       "step": 122
     },
     {
-      "epoch": 1.0123456790123457,
-      "grad_norm": 0.26857872224906654,
-      "learning_rate": 1.5481273557041402e-05,
-      "loss": 0.0578,
+      "epoch": 0.8367346938775511,
+      "grad_norm": 0.24145186470793584,
+      "learning_rate": 1.6953049435076768e-05,
+      "loss": 0.0607,
       "step": 123
     },
     {
-      "epoch": 1.0205761316872428,
-      "grad_norm": 0.20838554012208246,
-      "learning_rate": 1.5406408174555978e-05,
-      "loss": 0.0546,
+      "epoch": 0.8435374149659864,
+      "grad_norm": 0.24291351924884128,
+      "learning_rate": 1.6899983288347248e-05,
+      "loss": 0.0588,
       "step": 124
     },
     {
-      "epoch": 1.02880658436214,
-      "grad_norm": 0.3624505696906226,
-      "learning_rate": 1.5331112146121104e-05,
-      "loss": 0.0587,
+      "epoch": 0.8503401360544217,
+      "grad_norm": 0.19252112685578338,
+      "learning_rate": 1.6846543642074382e-05,
+      "loss": 0.0605,
       "step": 125
     },
     {
-      "epoch": 1.037037037037037,
-      "grad_norm": 0.21407973065400665,
-      "learning_rate": 1.525539146942113e-05,
-      "loss": 0.058,
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.2211891440249708,
+      "learning_rate": 1.679273338897293e-05,
+      "loss": 0.0571,
       "step": 126
     },
     {
-      "epoch": 1.045267489711934,
-      "grad_norm": 0.2726097771070703,
-      "learning_rate": 1.5179252175965632e-05,
-      "loss": 0.0588,
+      "epoch": 0.8639455782312925,
+      "grad_norm": 0.22450038327268926,
+      "learning_rate": 1.6738555441818785e-05,
+      "loss": 0.0589,
       "step": 127
     },
     {
-      "epoch": 1.0534979423868314,
-      "grad_norm": 0.2453134661092687,
-      "learning_rate": 1.5102700330609e-05,
-      "loss": 0.0586,
+      "epoch": 0.8707482993197279,
+      "grad_norm": 0.21870971416073584,
+      "learning_rate": 1.668401273329129e-05,
+      "loss": 0.0596,
       "step": 128
     },
     {
-      "epoch": 1.0617283950617284,
-      "grad_norm": 0.26783720738273514,
-      "learning_rate": 1.5025742031067316e-05,
+      "epoch": 0.8775510204081632,
+      "grad_norm": 0.2290461205947771,
+      "learning_rate": 1.6629108215814523e-05,
       "loss": 0.0572,
       "step": 129
     },
     {
-      "epoch": 1.0699588477366255,
-      "grad_norm": 0.30091314099275257,
-      "learning_rate": 1.4948383407432678e-05,
-      "loss": 0.0598,
+      "epoch": 0.8843537414965986,
+      "grad_norm": 0.218824327466051,
+      "learning_rate": 1.6573844861397444e-05,
+      "loss": 0.059,
       "step": 130
     },
     {
-      "epoch": 1.0781893004115226,
-      "grad_norm": 0.202788829409372,
-      "learning_rate": 1.4870630621684873e-05,
-      "loss": 0.0547,
+      "epoch": 0.891156462585034,
+      "grad_norm": 0.25616970456692517,
+      "learning_rate": 1.6518225661473045e-05,
+      "loss": 0.0607,
       "step": 131
     },
     {
-      "epoch": 1.0864197530864197,
-      "grad_norm": 0.25881685206833793,
-      "learning_rate": 1.479248986720057e-05,
-      "loss": 0.0546,
+      "epoch": 0.8979591836734694,
+      "grad_norm": 0.22446082647506313,
+      "learning_rate": 1.6462253626736413e-05,
+      "loss": 0.0602,
       "step": 132
     },
     {
-      "epoch": 1.0946502057613168,
-      "grad_norm": 0.23337381367388124,
-      "learning_rate": 1.4713967368259981e-05,
-      "loss": 0.0567,
+      "epoch": 0.9047619047619048,
+      "grad_norm": 0.19346153297308857,
+      "learning_rate": 1.6405931786981753e-05,
+      "loss": 0.0598,
       "step": 133
     },
     {
-      "epoch": 1.102880658436214,
-      "grad_norm": 0.2150759869883221,
-      "learning_rate": 1.4635069379551054e-05,
-      "loss": 0.0561,
+      "epoch": 0.9115646258503401,
+      "grad_norm": 0.19485485871091884,
+      "learning_rate": 1.63492631909384e-05,
+      "loss": 0.0583,
       "step": 134
     },
     {
-      "epoch": 1.1111111111111112,
-      "grad_norm": 0.25301720928646987,
-      "learning_rate": 1.4555802185671297e-05,
-      "loss": 0.0535,
+      "epoch": 0.9183673469387755,
+      "grad_norm": 0.2164924347710122,
+      "learning_rate": 1.629225090610577e-05,
+      "loss": 0.0591,
       "step": 135
     },
     {
-      "epoch": 1.1193415637860082,
-      "grad_norm": 0.19088482016766886,
-      "learning_rate": 1.4476172100627127e-05,
-      "loss": 0.0565,
+      "epoch": 0.9251700680272109,
+      "grad_norm": 0.19853198517283607,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.0578,
       "step": 136
     },
     {
-      "epoch": 1.1275720164609053,
-      "grad_norm": 0.19790904566323414,
-      "learning_rate": 1.4396185467330974e-05,
-      "loss": 0.0509,
+      "epoch": 0.9319727891156463,
+      "grad_norm": 0.22323813166846368,
+      "learning_rate": 1.6177207632923558e-05,
+      "loss": 0.061,
       "step": 137
     },
     {
-      "epoch": 1.1358024691358024,
-      "grad_norm": 0.31296542394204196,
-      "learning_rate": 1.4315848657096006e-05,
-      "loss": 0.0581,
+      "epoch": 0.9387755102040817,
+      "grad_norm": 0.2832642613760529,
+      "learning_rate": 1.6119182871923834e-05,
+      "loss": 0.0576,
       "step": 138
     },
     {
-      "epoch": 1.1440329218106995,
-      "grad_norm": 0.21662238290061556,
-      "learning_rate": 1.4235168069128657e-05,
-      "loss": 0.0545,
+      "epoch": 0.9455782312925171,
+      "grad_norm": 0.1775474704870787,
+      "learning_rate": 1.606082687649748e-05,
+      "loss": 0.0575,
       "step": 139
     },
     {
-      "epoch": 1.1522633744855968,
-      "grad_norm": 0.21481184300158193,
-      "learning_rate": 1.4154150130018867e-05,
-      "loss": 0.0558,
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.1974886171121568,
+      "learning_rate": 1.6002142805483686e-05,
+      "loss": 0.057,
       "step": 140
     },
     {
-      "epoch": 1.1604938271604939,
-      "grad_norm": 0.2848681451806038,
-      "learning_rate": 1.407280129322819e-05,
-      "loss": 0.0535,
+      "epoch": 0.9591836734693877,
+      "grad_norm": 0.23607667812519084,
+      "learning_rate": 1.5943133835480536e-05,
+      "loss": 0.0563,
       "step": 141
     },
     {
-      "epoch": 1.168724279835391,
-      "grad_norm": 0.19329905937013103,
-      "learning_rate": 1.3991128038575741e-05,
-      "loss": 0.0552,
+      "epoch": 0.9659863945578231,
+      "grad_norm": 0.2558776300805798,
+      "learning_rate": 1.588380316067307e-05,
+      "loss": 0.0581,
       "step": 142
     },
     {
-      "epoch": 1.176954732510288,
-      "grad_norm": 0.27548799309496036,
-      "learning_rate": 1.3909136871722066e-05,
-      "loss": 0.0518,
+      "epoch": 0.9727891156462585,
+      "grad_norm": 0.2571443627011699,
+      "learning_rate": 1.582415399266036e-05,
+      "loss": 0.0628,
       "step": 143
     },
     {
-      "epoch": 1.1851851851851851,
-      "grad_norm": 0.26231556821331964,
-      "learning_rate": 1.3826834323650899e-05,
-      "loss": 0.0527,
+      "epoch": 0.9795918367346939,
+      "grad_norm": 0.3450295396522308,
+      "learning_rate": 1.5764189560281677e-05,
+      "loss": 0.0582,
       "step": 144
     },
     {
-      "epoch": 1.1934156378600824,
-      "grad_norm": 0.26557199976386675,
-      "learning_rate": 1.374422695014897e-05,
-      "loss": 0.0576,
+      "epoch": 0.9863945578231292,
+      "grad_norm": 0.26588782094294083,
+      "learning_rate": 1.5703913109441715e-05,
+      "loss": 0.0583,
       "step": 145
     },
     {
-      "epoch": 1.2016460905349795,
-      "grad_norm": 0.28136793485435446,
-      "learning_rate": 1.3661321331283796e-05,
-      "loss": 0.0554,
+      "epoch": 0.9931972789115646,
+      "grad_norm": 0.24694959346022663,
+      "learning_rate": 1.564332790293487e-05,
+      "loss": 0.055,
       "step": 146
     },
     {
-      "epoch": 1.2098765432098766,
-      "grad_norm": 0.2311600538669461,
-      "learning_rate": 1.3578124070879534e-05,
-      "loss": 0.0481,
+      "epoch": 1.0,
+      "grad_norm": 0.27312319281208475,
+      "learning_rate": 1.5582437220268648e-05,
+      "loss": 0.0626,
       "step": 147
     },
     {
-      "epoch": 1.2181069958847737,
-      "grad_norm": 0.23273286456201997,
-      "learning_rate": 1.3494641795990986e-05,
-      "loss": 0.0516,
+      "epoch": 1.0068027210884354,
+      "grad_norm": 0.30750819635972554,
+      "learning_rate": 1.5521244357486132e-05,
+      "loss": 0.0593,
       "step": 148
     },
     {
-      "epoch": 1.2263374485596708,
-      "grad_norm": 0.24493468908630536,
-      "learning_rate": 1.3410881156375684e-05,
-      "loss": 0.0544,
+      "epoch": 1.0136054421768708,
+      "grad_norm": 0.24494203240771864,
+      "learning_rate": 1.5459752626987563e-05,
+      "loss": 0.054,
       "step": 149
     },
     {
-      "epoch": 1.2345679012345678,
-      "grad_norm": 0.34141471042388616,
-      "learning_rate": 1.3326848823964243e-05,
-      "loss": 0.0517,
+      "epoch": 1.0204081632653061,
+      "grad_norm": 0.22210056662122235,
+      "learning_rate": 1.5397965357351035e-05,
+      "loss": 0.0558,
       "step": 150
     },
     {
-      "epoch": 1.242798353909465,
-      "grad_norm": 0.2576832529046746,
-      "learning_rate": 1.3242551492328875e-05,
-      "loss": 0.0543,
+      "epoch": 1.0272108843537415,
+      "grad_norm": 0.23782219601214885,
+      "learning_rate": 1.5335885893152335e-05,
+      "loss": 0.0564,
       "step": 151
     },
     {
-      "epoch": 1.2510288065843622,
-      "grad_norm": 0.2614124457629345,
-      "learning_rate": 1.3157995876150252e-05,
-      "loss": 0.0484,
+      "epoch": 1.034013605442177,
+      "grad_norm": 0.21427038780601312,
+      "learning_rate": 1.5273517594783878e-05,
+      "loss": 0.0523,
       "step": 152
     },
     {
-      "epoch": 1.2592592592592593,
-      "grad_norm": 0.19343668486412105,
-      "learning_rate": 1.3073188710682612e-05,
-      "loss": 0.0467,
+      "epoch": 1.0408163265306123,
+      "grad_norm": 0.32614094149577777,
+      "learning_rate": 1.521086383827282e-05,
+      "loss": 0.0568,
       "step": 153
     },
     {
-      "epoch": 1.2674897119341564,
-      "grad_norm": 0.19809805402327385,
-      "learning_rate": 1.2988136751217292e-05,
-      "loss": 0.0507,
+      "epoch": 1.0476190476190477,
+      "grad_norm": 0.1902175321443073,
+      "learning_rate": 1.5147928015098309e-05,
+      "loss": 0.0545,
       "step": 154
     },
     {
-      "epoch": 1.2757201646090535,
-      "grad_norm": 0.30398995200030304,
-      "learning_rate": 1.2902846772544625e-05,
-      "loss": 0.0519,
+      "epoch": 1.054421768707483,
+      "grad_norm": 0.2087024945007221,
+      "learning_rate": 1.5084713532007906e-05,
+      "loss": 0.0541,
       "step": 155
     },
     {
-      "epoch": 1.2839506172839505,
-      "grad_norm": 0.21948508334629585,
-      "learning_rate": 1.2817325568414299e-05,
-      "loss": 0.0493,
+      "epoch": 1.0612244897959184,
+      "grad_norm": 0.30608404804040057,
+      "learning_rate": 1.5021223810833165e-05,
+      "loss": 0.0556,
       "step": 156
     },
     {
-      "epoch": 1.2921810699588478,
-      "grad_norm": 0.26451695698307937,
-      "learning_rate": 1.27315799509942e-05,
-      "loss": 0.047,
+      "epoch": 1.0680272108843538,
+      "grad_norm": 0.27839206067610106,
+      "learning_rate": 1.4957462288304421e-05,
+      "loss": 0.0534,
       "step": 157
     },
     {
-      "epoch": 1.300411522633745,
-      "grad_norm": 0.19677120215727498,
-      "learning_rate": 1.2645616750327792e-05,
-      "loss": 0.0487,
+      "epoch": 1.0748299319727892,
+      "grad_norm": 0.2213087101191798,
+      "learning_rate": 1.489343241586475e-05,
+      "loss": 0.0547,
       "step": 158
     },
     {
-      "epoch": 1.308641975308642,
-      "grad_norm": 0.22984380885090203,
-      "learning_rate": 1.2559442813790077e-05,
-      "loss": 0.0468,
+      "epoch": 1.0816326530612246,
+      "grad_norm": 0.20871107281249643,
+      "learning_rate": 1.4829137659483144e-05,
+      "loss": 0.0543,
       "step": 159
     },
     {
-      "epoch": 1.316872427983539,
-      "grad_norm": 0.3364297762395688,
-      "learning_rate": 1.2473065005542155e-05,
-      "loss": 0.0515,
+      "epoch": 1.08843537414966,
+      "grad_norm": 0.23030377388618803,
+      "learning_rate": 1.4764581499466895e-05,
+      "loss": 0.0556,
       "step": 160
     },
     {
-      "epoch": 1.3251028806584362,
-      "grad_norm": 0.19763740027879062,
-      "learning_rate": 1.2386490205984488e-05,
-      "loss": 0.0525,
+      "epoch": 1.0952380952380953,
+      "grad_norm": 0.1932290087273829,
+      "learning_rate": 1.4699767430273202e-05,
+      "loss": 0.0524,
       "step": 161
     },
     {
-      "epoch": 1.3333333333333333,
-      "grad_norm": 0.22014348334833883,
-      "learning_rate": 1.2299725311208807e-05,
-      "loss": 0.0471,
+      "epoch": 1.1020408163265305,
+      "grad_norm": 0.40409724353966364,
+      "learning_rate": 1.4634698960320018e-05,
+      "loss": 0.0527,
       "step": 162
     },
     {
-      "epoch": 1.3415637860082303,
-      "grad_norm": 0.2805839794596332,
-      "learning_rate": 1.2212777232448837e-05,
-      "loss": 0.0525,
+      "epoch": 1.1088435374149659,
+      "grad_norm": 0.22092199900707887,
+      "learning_rate": 1.4569379611796137e-05,
+      "loss": 0.0531,
       "step": 163
     },
     {
-      "epoch": 1.3497942386831276,
-      "grad_norm": 0.27239111613326095,
-      "learning_rate": 1.2125652895529766e-05,
-      "loss": 0.0515,
+      "epoch": 1.1156462585034013,
+      "grad_norm": 0.23715254386252377,
+      "learning_rate": 1.4503812920470535e-05,
+      "loss": 0.052,
       "step": 164
     },
     {
-      "epoch": 1.3580246913580247,
-      "grad_norm": 0.2787079576089145,
-      "learning_rate": 1.2038359240316589e-05,
-      "loss": 0.0452,
+      "epoch": 1.1224489795918366,
+      "grad_norm": 0.25222406455269336,
+      "learning_rate": 1.443800243550098e-05,
+      "loss": 0.0507,
       "step": 165
     },
     {
-      "epoch": 1.3662551440329218,
-      "grad_norm": 0.33595144965399354,
-      "learning_rate": 1.1950903220161286e-05,
-      "loss": 0.053,
+      "epoch": 1.129251700680272,
+      "grad_norm": 0.20150273343577169,
+      "learning_rate": 1.4371951719241906e-05,
+      "loss": 0.0572,
       "step": 166
     },
     {
-      "epoch": 1.374485596707819,
-      "grad_norm": 0.2554338228011529,
-      "learning_rate": 1.186329180134898e-05,
-      "loss": 0.0439,
+      "epoch": 1.1360544217687074,
+      "grad_norm": 0.2981137590723175,
+      "learning_rate": 1.4305664347051586e-05,
+      "loss": 0.0557,
       "step": 167
     },
     {
-      "epoch": 1.382716049382716,
-      "grad_norm": 0.28715064089173337,
-      "learning_rate": 1.1775531962543036e-05,
-      "loss": 0.0528,
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.36363358730765705,
+      "learning_rate": 1.423914390709861e-05,
+      "loss": 0.0499,
       "step": 168
     },
     {
-      "epoch": 1.3909465020576133,
-      "grad_norm": 0.26836179486578365,
-      "learning_rate": 1.1687630694229159e-05,
-      "loss": 0.04,
+      "epoch": 1.1496598639455782,
+      "grad_norm": 0.3489948294285827,
+      "learning_rate": 1.4172394000167625e-05,
+      "loss": 0.0559,
       "step": 169
     },
     {
-      "epoch": 1.3991769547325104,
-      "grad_norm": 0.2672843060683665,
-      "learning_rate": 1.1599594998158602e-05,
-      "loss": 0.0479,
+      "epoch": 1.1564625850340136,
+      "grad_norm": 0.25108294665044295,
+      "learning_rate": 1.4105418239464452e-05,
+      "loss": 0.0516,
       "step": 170
     },
     {
-      "epoch": 1.4074074074074074,
-      "grad_norm": 0.2069347790736105,
-      "learning_rate": 1.1511431886790407e-05,
-      "loss": 0.0467,
+      "epoch": 1.163265306122449,
+      "grad_norm": 0.31647443965414784,
+      "learning_rate": 1.4038220250420487e-05,
+      "loss": 0.0535,
       "step": 171
     },
     {
-      "epoch": 1.4156378600823045,
-      "grad_norm": 0.2241979441649495,
-      "learning_rate": 1.1423148382732854e-05,
-      "loss": 0.0487,
+      "epoch": 1.1700680272108843,
+      "grad_norm": 0.3026605870439961,
+      "learning_rate": 1.3970803670496453e-05,
+      "loss": 0.0501,
       "step": 172
     },
     {
-      "epoch": 1.4238683127572016,
-      "grad_norm": 0.2129752744580926,
-      "learning_rate": 1.1334751518184062e-05,
-      "loss": 0.0448,
+      "epoch": 1.1768707482993197,
+      "grad_norm": 0.3310780887181829,
+      "learning_rate": 1.390317214898551e-05,
+      "loss": 0.0541,
       "step": 173
     },
     {
-      "epoch": 1.4320987654320987,
-      "grad_norm": 0.24709203655835915,
-      "learning_rate": 1.124624833437186e-05,
-      "loss": 0.047,
+      "epoch": 1.183673469387755,
+      "grad_norm": 0.23883640286772895,
+      "learning_rate": 1.3835329346815716e-05,
+      "loss": 0.052,
       "step": 174
     },
     {
-      "epoch": 1.4403292181069958,
-      "grad_norm": 0.21545097375940966,
-      "learning_rate": 1.1157645880992901e-05,
-      "loss": 0.0475,
+      "epoch": 1.1904761904761905,
+      "grad_norm": 0.3668380546284328,
+      "learning_rate": 1.3767278936351853e-05,
+      "loss": 0.0521,
       "step": 175
     },
     {
-      "epoch": 1.448559670781893,
-      "grad_norm": 0.2435965590800978,
-      "learning_rate": 1.1068951215651132e-05,
-      "loss": 0.0467,
+      "epoch": 1.1972789115646258,
+      "grad_norm": 0.3165490998015608,
+      "learning_rate": 1.3699024601196641e-05,
+      "loss": 0.05,
       "step": 176
     },
     {
-      "epoch": 1.4567901234567902,
-      "grad_norm": 0.26728310945356726,
-      "learning_rate": 1.098017140329561e-05,
-      "loss": 0.041,
+      "epoch": 1.2040816326530612,
+      "grad_norm": 0.28757144157751063,
+      "learning_rate": 1.3630570035991352e-05,
+      "loss": 0.0523,
       "step": 177
     },
     {
-      "epoch": 1.4650205761316872,
-      "grad_norm": 0.34434692596372063,
-      "learning_rate": 1.089131351565776e-05,
-      "loss": 0.0461,
+      "epoch": 1.2108843537414966,
+      "grad_norm": 0.24922250771310617,
+      "learning_rate": 1.3561918946215807e-05,
+      "loss": 0.0476,
       "step": 178
     },
     {
-      "epoch": 1.4732510288065843,
-      "grad_norm": 0.5887891834464638,
-      "learning_rate": 1.080238463068808e-05,
-      "loss": 0.0509,
+      "epoch": 1.217687074829932,
+      "grad_norm": 0.3109432126479953,
+      "learning_rate": 1.34930750479878e-05,
+      "loss": 0.0528,
       "step": 179
     },
     {
-      "epoch": 1.4814814814814814,
-      "grad_norm": 0.2529279349031601,
-      "learning_rate": 1.0713391831992324e-05,
-      "loss": 0.0447,
+      "epoch": 1.2244897959183674,
+      "grad_norm": 0.3191250919868236,
+      "learning_rate": 1.3424042067861944e-05,
+      "loss": 0.0488,
       "step": 180
     },
     {
-      "epoch": 1.4897119341563787,
-      "grad_norm": 0.23822774710376737,
-      "learning_rate": 1.0624342208267293e-05,
-      "loss": 0.0446,
+      "epoch": 1.2312925170068028,
+      "grad_norm": 0.33504846480374884,
+      "learning_rate": 1.335482374262795e-05,
+      "loss": 0.0485,
       "step": 181
     },
     {
-      "epoch": 1.4979423868312758,
-      "grad_norm": 0.19480949987814833,
-      "learning_rate": 1.0535242852736152e-05,
-      "loss": 0.0468,
+      "epoch": 1.2380952380952381,
+      "grad_norm": 0.41636172795109355,
+      "learning_rate": 1.3285423819108349e-05,
+      "loss": 0.0481,
       "step": 182
     },
     {
-      "epoch": 1.5061728395061729,
-      "grad_norm": 0.2327277827776109,
-      "learning_rate": 1.0446100862583459e-05,
-      "loss": 0.0487,
+      "epoch": 1.2448979591836735,
+      "grad_norm": 0.43892808658376165,
+      "learning_rate": 1.3215846053955683e-05,
+      "loss": 0.0486,
       "step": 183
     },
     {
-      "epoch": 1.51440329218107,
-      "grad_norm": 0.31275705518072855,
-      "learning_rate": 1.0356923338389807e-05,
-      "loss": 0.041,
+      "epoch": 1.251700680272109,
+      "grad_norm": 0.32837726008368157,
+      "learning_rate": 1.3146094213449148e-05,
+      "loss": 0.0451,
       "step": 184
     },
     {
-      "epoch": 1.522633744855967,
-      "grad_norm": 0.2891984367280337,
-      "learning_rate": 1.0267717383566247e-05,
-      "loss": 0.0457,
+      "epoch": 1.2585034013605443,
+      "grad_norm": 0.24228613984412564,
+      "learning_rate": 1.3076172073290726e-05,
+      "loss": 0.046,
       "step": 185
     },
     {
-      "epoch": 1.5308641975308643,
-      "grad_norm": 0.26111803859652866,
-      "learning_rate": 1.0178490103788462e-05,
-      "loss": 0.046,
+      "epoch": 1.2653061224489797,
+      "grad_norm": 0.42623267395045844,
+      "learning_rate": 1.3006083418400799e-05,
+      "loss": 0.0495,
       "step": 186
     },
     {
-      "epoch": 1.5390946502057612,
-      "grad_norm": 0.24672680710413208,
-      "learning_rate": 1.0089248606430775e-05,
-      "loss": 0.0467,
+      "epoch": 1.272108843537415,
+      "grad_norm": 0.3067088420671033,
+      "learning_rate": 1.2935832042713288e-05,
+      "loss": 0.0506,
       "step": 187
     },
     {
-      "epoch": 1.5473251028806585,
-      "grad_norm": 0.19925020919272812,
-      "learning_rate": 1e-05,
-      "loss": 0.044,
+      "epoch": 1.2789115646258504,
+      "grad_norm": 0.3652294320037095,
+      "learning_rate": 1.2865421748970257e-05,
+      "loss": 0.0493,
       "step": 188
     },
     {
-      "epoch": 1.5555555555555556,
-      "grad_norm": 0.24668853054896053,
-      "learning_rate": 9.910751393569228e-06,
-      "loss": 0.0447,
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.4171203977029359,
+      "learning_rate": 1.2794856348516095e-05,
+      "loss": 0.0458,
       "step": 189
     },
     {
-      "epoch": 1.5637860082304527,
-      "grad_norm": 0.39519256015191795,
-      "learning_rate": 9.82150989621154e-06,
-      "loss": 0.0512,
+      "epoch": 1.2925170068027212,
+      "grad_norm": 0.2701855372697879,
+      "learning_rate": 1.2724139661091188e-05,
+      "loss": 0.0432,
       "step": 190
     },
     {
-      "epoch": 1.5720164609053497,
-      "grad_norm": 0.2415354464323035,
-      "learning_rate": 9.732282616433756e-06,
-      "loss": 0.0441,
+      "epoch": 1.2993197278911564,
+      "grad_norm": 0.3259307296746982,
+      "learning_rate": 1.2653275514625165e-05,
+      "loss": 0.0463,
       "step": 191
     },
     {
-      "epoch": 1.5802469135802468,
-      "grad_norm": 0.3216164056977005,
-      "learning_rate": 9.643076661610197e-06,
-      "loss": 0.0471,
+      "epoch": 1.306122448979592,
+      "grad_norm": 0.2550766456669367,
+      "learning_rate": 1.2582267745029685e-05,
+      "loss": 0.0447,
       "step": 192
     },
     {
-      "epoch": 1.5884773662551441,
-      "grad_norm": 0.21343319214461834,
-      "learning_rate": 9.553899137416546e-06,
-      "loss": 0.0408,
+      "epoch": 1.3129251700680271,
+      "grad_norm": 0.4220940884592931,
+      "learning_rate": 1.2511120195990797e-05,
+      "loss": 0.048,
       "step": 193
     },
     {
-      "epoch": 1.596707818930041,
-      "grad_norm": 0.2763914169234503,
-      "learning_rate": 9.464757147263849e-06,
-      "loss": 0.0459,
+      "epoch": 1.3197278911564627,
+      "grad_norm": 0.36984656348421135,
+      "learning_rate": 1.2439836718760887e-05,
+      "loss": 0.0411,
       "step": 194
     },
     {
-      "epoch": 1.6049382716049383,
-      "grad_norm": 0.386915642708884,
-      "learning_rate": 9.37565779173271e-06,
-      "loss": 0.0433,
+      "epoch": 1.3265306122448979,
+      "grad_norm": 0.36136346549964,
+      "learning_rate": 1.2368421171950193e-05,
+      "loss": 0.046,
       "step": 195
     },
     {
-      "epoch": 1.6131687242798354,
-      "grad_norm": 0.22385455403266594,
-      "learning_rate": 9.286608168007678e-06,
-      "loss": 0.047,
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.3788271776627097,
+      "learning_rate": 1.2296877421317958e-05,
+      "loss": 0.0437,
       "step": 196
     },
     {
-      "epoch": 1.6213991769547325,
-      "grad_norm": 0.23465285981431852,
-      "learning_rate": 9.197615369311926e-06,
-      "loss": 0.0448,
+      "epoch": 1.3401360544217686,
+      "grad_norm": 0.28520654057079897,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.0418,
       "step": 197
     },
     {
-      "epoch": 1.6296296296296298,
-      "grad_norm": 0.26575989753627616,
-      "learning_rate": 9.108686484342241e-06,
-      "loss": 0.0446,
+      "epoch": 1.346938775510204,
+      "grad_norm": 0.49694987958709724,
+      "learning_rate": 1.215342080611484e-05,
+      "loss": 0.0448,
       "step": 198
     },
     {
-      "epoch": 1.6378600823045266,
-      "grad_norm": 0.24115234950722386,
-      "learning_rate": 9.019828596704394e-06,
-      "loss": 0.0445,
+      "epoch": 1.3537414965986394,
+      "grad_norm": 0.442980712849271,
+      "learning_rate": 1.2081515706922226e-05,
+      "loss": 0.0477,
       "step": 199
     },
     {
-      "epoch": 1.646090534979424,
-      "grad_norm": 0.259073918752676,
-      "learning_rate": 8.931048784348875e-06,
-      "loss": 0.0411,
+      "epoch": 1.3605442176870748,
+      "grad_norm": 0.30908251467526043,
+      "learning_rate": 1.2009497934244257e-05,
+      "loss": 0.0486,
       "step": 200
     },
     {
-      "epoch": 1.654320987654321,
-      "grad_norm": 0.25398810096959956,
-      "learning_rate": 8.8423541190071e-06,
-      "loss": 0.0446,
+      "epoch": 1.3673469387755102,
+      "grad_norm": 0.31690703603059645,
+      "learning_rate": 1.1937371386438954e-05,
+      "loss": 0.0434,
       "step": 201
     },
     {
-      "epoch": 1.662551440329218,
-      "grad_norm": 0.22819463464531642,
-      "learning_rate": 8.753751665628141e-06,
-      "loss": 0.0474,
+      "epoch": 1.3741496598639455,
+      "grad_norm": 0.3622937591780268,
+      "learning_rate": 1.186513996775239e-05,
+      "loss": 0.0442,
       "step": 202
     },
     {
-      "epoch": 1.6707818930041154,
-      "grad_norm": 0.23445138294853232,
-      "learning_rate": 8.665248481815941e-06,
-      "loss": 0.0441,
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.30716164525104495,
+      "learning_rate": 1.1792807588107358e-05,
+      "loss": 0.0436,
       "step": 203
     },
     {
-      "epoch": 1.6790123456790123,
-      "grad_norm": 0.30003106032342386,
-      "learning_rate": 8.576851617267151e-06,
-      "loss": 0.0421,
+      "epoch": 1.3877551020408163,
+      "grad_norm": 0.3321484566845795,
+      "learning_rate": 1.1720378162891709e-05,
+      "loss": 0.0464,
       "step": 204
     },
     {
-      "epoch": 1.6872427983539096,
-      "grad_norm": 0.2629347004565258,
-      "learning_rate": 8.488568113209593e-06,
-      "loss": 0.0427,
+      "epoch": 1.3945578231292517,
+      "grad_norm": 0.3019164729518717,
+      "learning_rate": 1.1647855612746423e-05,
+      "loss": 0.0421,
       "step": 205
     },
     {
-      "epoch": 1.6954732510288066,
-      "grad_norm": 0.20569571631682926,
-      "learning_rate": 8.4004050018414e-06,
-      "loss": 0.0391,
+      "epoch": 1.401360544217687,
+      "grad_norm": 0.28208298182130315,
+      "learning_rate": 1.1575243863353383e-05,
+      "loss": 0.0384,
       "step": 206
     },
     {
-      "epoch": 1.7037037037037037,
-      "grad_norm": 0.24708870255267373,
-      "learning_rate": 8.312369305770843e-06,
-      "loss": 0.0402,
+      "epoch": 1.4081632653061225,
+      "grad_norm": 0.27187917628455394,
+      "learning_rate": 1.150254684522286e-05,
+      "loss": 0.0395,
       "step": 207
     },
     {
-      "epoch": 1.7119341563786008,
-      "grad_norm": 0.19972858271830007,
-      "learning_rate": 8.224468037456969e-06,
-      "loss": 0.0394,
+      "epoch": 1.4149659863945578,
+      "grad_norm": 0.29967306578610386,
+      "learning_rate": 1.142976849348078e-05,
+      "loss": 0.0408,
       "step": 208
     },
     {
-      "epoch": 1.7201646090534979,
-      "grad_norm": 0.2811098028592767,
-      "learning_rate": 8.136708198651022e-06,
-      "loss": 0.0455,
+      "epoch": 1.4217687074829932,
+      "grad_norm": 0.22436230612440464,
+      "learning_rate": 1.1356912747655687e-05,
+      "loss": 0.0339,
       "step": 209
     },
     {
-      "epoch": 1.7283950617283952,
-      "grad_norm": 0.20819108343920742,
-      "learning_rate": 8.04909677983872e-06,
-      "loss": 0.0389,
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.42439154027724657,
+      "learning_rate": 1.1283983551465512e-05,
+      "loss": 0.0476,
       "step": 210
     },
     {
-      "epoch": 1.736625514403292,
-      "grad_norm": 0.2397253686121003,
-      "learning_rate": 7.961640759683416e-06,
-      "loss": 0.0433,
+      "epoch": 1.435374149659864,
+      "grad_norm": 0.31700756503083577,
+      "learning_rate": 1.1210984852604084e-05,
+      "loss": 0.0358,
       "step": 211
     },
     {
-      "epoch": 1.7448559670781894,
-      "grad_norm": 0.2119970563782489,
-      "learning_rate": 7.874347104470234e-06,
-      "loss": 0.0397,
+      "epoch": 1.4421768707482994,
+      "grad_norm": 0.3378079110420009,
+      "learning_rate": 1.1137920602527448e-05,
+      "loss": 0.0379,
       "step": 212
     },
     {
-      "epoch": 1.7530864197530864,
-      "grad_norm": 0.31924904460902614,
-      "learning_rate": 7.787222767551164e-06,
-      "loss": 0.039,
+      "epoch": 1.4489795918367347,
+      "grad_norm": 0.24349092505484088,
+      "learning_rate": 1.1064794756239978e-05,
+      "loss": 0.0363,
       "step": 213
     },
     {
-      "epoch": 1.7613168724279835,
-      "grad_norm": 0.294935157459512,
-      "learning_rate": 7.700274688791196e-06,
-      "loss": 0.0426,
+      "epoch": 1.4557823129251701,
+      "grad_norm": 0.3400711419864756,
+      "learning_rate": 1.099161127208027e-05,
+      "loss": 0.0377,
       "step": 214
     },
     {
-      "epoch": 1.7695473251028808,
-      "grad_norm": 0.24926608841845987,
-      "learning_rate": 7.613509794015517e-06,
-      "loss": 0.0451,
+      "epoch": 1.4625850340136055,
+      "grad_norm": 0.5100291242139185,
+      "learning_rate": 1.0918374111506893e-05,
+      "loss": 0.0382,
       "step": 215
     },
     {
-      "epoch": 1.7777777777777777,
-      "grad_norm": 0.32045764874703286,
-      "learning_rate": 7.5269349944578454e-06,
-      "loss": 0.0447,
+      "epoch": 1.469387755102041,
+      "grad_norm": 0.28802764191587965,
+      "learning_rate": 1.0845087238883945e-05,
+      "loss": 0.032,
       "step": 216
     },
     {
-      "epoch": 1.786008230452675,
-      "grad_norm": 0.2550384249500506,
-      "learning_rate": 7.440557186209927e-06,
-      "loss": 0.0408,
+      "epoch": 1.4761904761904763,
+      "grad_norm": 0.2607923120434538,
+      "learning_rate": 1.0771754621266466e-05,
+      "loss": 0.0365,
       "step": 217
     },
     {
-      "epoch": 1.794238683127572,
-      "grad_norm": 0.2772059288520359,
-      "learning_rate": 7.354383249672212e-06,
-      "loss": 0.0465,
+      "epoch": 1.4829931972789114,
+      "grad_norm": 0.2640244555130478,
+      "learning_rate": 1.0698380228185685e-05,
+      "loss": 0.0433,
       "step": 218
     },
     {
-      "epoch": 1.8024691358024691,
-      "grad_norm": 0.23303168593422405,
-      "learning_rate": 7.268420049005806e-06,
-      "loss": 0.0394,
+      "epoch": 1.489795918367347,
+      "grad_norm": 0.2849877896554485,
+      "learning_rate": 1.0624968031434174e-05,
+      "loss": 0.0426,
       "step": 219
     },
     {
-      "epoch": 1.8106995884773662,
-      "grad_norm": 0.3118047053593649,
-      "learning_rate": 7.182674431585703e-06,
-      "loss": 0.0447,
+      "epoch": 1.4965986394557822,
+      "grad_norm": 0.34138049373172324,
+      "learning_rate": 1.0551522004850821e-05,
+      "loss": 0.042,
       "step": 220
     },
     {
-      "epoch": 1.8189300411522633,
-      "grad_norm": 0.7455818122851706,
-      "learning_rate": 7.097153227455379e-06,
-      "loss": 0.0406,
+      "epoch": 1.5034013605442178,
+      "grad_norm": 0.4028334423985101,
+      "learning_rate": 1.0478046124105746e-05,
+      "loss": 0.032,
       "step": 221
     },
     {
-      "epoch": 1.8271604938271606,
-      "grad_norm": 0.2945463432870641,
-      "learning_rate": 7.011863248782711e-06,
-      "loss": 0.0443,
+      "epoch": 1.510204081632653,
+      "grad_norm": 0.29976572412890073,
+      "learning_rate": 1.0404544366485094e-05,
+      "loss": 0.0318,
       "step": 222
     },
     {
-      "epoch": 1.8353909465020575,
-      "grad_norm": 0.26264313360867836,
-      "learning_rate": 6.92681128931739e-06,
-      "loss": 0.0418,
+      "epoch": 1.5170068027210886,
+      "grad_norm": 0.3575056341024037,
+      "learning_rate": 1.033102071067573e-05,
+      "loss": 0.0328,
       "step": 223
     },
     {
-      "epoch": 1.8436213991769548,
-      "grad_norm": 0.3312208907352792,
-      "learning_rate": 6.8420041238497525e-06,
-      "loss": 0.0427,
+      "epoch": 1.5238095238095237,
+      "grad_norm": 0.42003758608591496,
+      "learning_rate": 1.0257479136549889e-05,
+      "loss": 0.0412,
       "step": 224
     },
     {
-      "epoch": 1.8518518518518519,
-      "grad_norm": 0.22603134209986628,
-      "learning_rate": 6.7574485076711285e-06,
-      "loss": 0.0381,
+      "epoch": 1.5306122448979593,
+      "grad_norm": 0.2646979747322718,
+      "learning_rate": 1.0183923624949721e-05,
+      "loss": 0.0359,
       "step": 225
     },
     {
-      "epoch": 1.860082304526749,
-      "grad_norm": 0.23100800341246475,
-      "learning_rate": 6.673151176035762e-06,
-      "loss": 0.0371,
+      "epoch": 1.5374149659863945,
+      "grad_norm": 0.2822553257788332,
+      "learning_rate": 1.0110358157471825e-05,
+      "loss": 0.0306,
       "step": 226
     },
     {
-      "epoch": 1.8683127572016462,
-      "grad_norm": 0.2847417603788892,
-      "learning_rate": 6.589118843624316e-06,
-      "loss": 0.0421,
+      "epoch": 1.54421768707483,
+      "grad_norm": 0.46122155273972015,
+      "learning_rate": 1.0036786716251721e-05,
+      "loss": 0.0457,
       "step": 227
     },
     {
-      "epoch": 1.876543209876543,
-      "grad_norm": 0.3507622075922086,
-      "learning_rate": 6.505358204009018e-06,
-      "loss": 0.0395,
+      "epoch": 1.5510204081632653,
+      "grad_norm": 0.44028200523172295,
+      "learning_rate": 9.963213283748282e-06,
+      "loss": 0.0348,
       "step": 228
     },
     {
-      "epoch": 1.8847736625514404,
-      "grad_norm": 0.23215767642461382,
-      "learning_rate": 6.421875929120469e-06,
-      "loss": 0.0371,
+      "epoch": 1.5578231292517006,
+      "grad_norm": 0.33390274335640047,
+      "learning_rate": 9.889641842528179e-06,
+      "loss": 0.0385,
       "step": 229
     },
     {
-      "epoch": 1.8930041152263375,
-      "grad_norm": 0.22019702385873458,
-      "learning_rate": 6.33867866871621e-06,
-      "loss": 0.0392,
+      "epoch": 1.564625850340136,
+      "grad_norm": 0.35548363557576734,
+      "learning_rate": 9.816076375050284e-06,
+      "loss": 0.0316,
       "step": 230
     },
     {
-      "epoch": 1.9012345679012346,
-      "grad_norm": 0.26909110946943715,
-      "learning_rate": 6.25577304985103e-06,
-      "loss": 0.039,
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.2988165557824134,
+      "learning_rate": 9.742520863450116e-06,
+      "loss": 0.0349,
       "step": 231
     },
     {
-      "epoch": 1.9094650205761317,
-      "grad_norm": 0.24691914088432093,
-      "learning_rate": 6.173165676349103e-06,
-      "loss": 0.0396,
+      "epoch": 1.5782312925170068,
+      "grad_norm": 0.31453672481100403,
+      "learning_rate": 9.668979289324274e-06,
+      "loss": 0.0318,
       "step": 232
     },
     {
-      "epoch": 1.9176954732510287,
-      "grad_norm": 0.1874264178548672,
-      "learning_rate": 6.090863128277938e-06,
-      "loss": 0.04,
+      "epoch": 1.5850340136054422,
+      "grad_norm": 0.2910996810127783,
+      "learning_rate": 9.595455633514908e-06,
+      "loss": 0.0254,
       "step": 233
     },
     {
-      "epoch": 1.925925925925926,
-      "grad_norm": 0.21884973858177934,
-      "learning_rate": 6.008871961424259e-06,
-      "loss": 0.0376,
+      "epoch": 1.5918367346938775,
+      "grad_norm": 0.27024499311466754,
+      "learning_rate": 9.521953875894256e-06,
+      "loss": 0.0317,
       "step": 234
     },
     {
-      "epoch": 1.934156378600823,
-      "grad_norm": 0.30857388990408324,
-      "learning_rate": 5.927198706771813e-06,
-      "loss": 0.0348,
+      "epoch": 1.598639455782313,
+      "grad_norm": 0.3192737285714962,
+      "learning_rate": 9.448477995149182e-06,
+      "loss": 0.0346,
       "step": 235
     },
     {
-      "epoch": 1.9423868312757202,
-      "grad_norm": 0.26813747360891743,
-      "learning_rate": 5.845849869981137e-06,
-      "loss": 0.0393,
+      "epoch": 1.6054421768707483,
+      "grad_norm": 0.31462197089436433,
+      "learning_rate": 9.37503196856583e-06,
+      "loss": 0.0347,
       "step": 236
     },
     {
-      "epoch": 1.9506172839506173,
-      "grad_norm": 0.3174546445476974,
-      "learning_rate": 5.764831930871346e-06,
-      "loss": 0.042,
+      "epoch": 1.6122448979591837,
+      "grad_norm": 0.4780207494130082,
+      "learning_rate": 9.301619771814317e-06,
+      "loss": 0.0344,
       "step": 237
     },
     {
-      "epoch": 1.9588477366255144,
-      "grad_norm": 0.39633765647261926,
-      "learning_rate": 5.684151342903992e-06,
-      "loss": 0.0368,
+      "epoch": 1.619047619047619,
+      "grad_norm": 0.3603665494734086,
+      "learning_rate": 9.228245378733537e-06,
+      "loss": 0.0284,
       "step": 238
     },
     {
-      "epoch": 1.9670781893004117,
-      "grad_norm": 0.27999699035314357,
-      "learning_rate": 5.603814532669032e-06,
-      "loss": 0.0355,
+      "epoch": 1.6258503401360545,
+      "grad_norm": 0.3481977799989387,
+      "learning_rate": 9.154912761116056e-06,
+      "loss": 0.0319,
       "step": 239
     },
     {
-      "epoch": 1.9753086419753085,
-      "grad_norm": 0.38461297456388893,
-      "learning_rate": 5.523827899372876e-06,
-      "loss": 0.0428,
+      "epoch": 1.6326530612244898,
+      "grad_norm": 0.31718948127093954,
+      "learning_rate": 9.081625888493107e-06,
+      "loss": 0.0347,
       "step": 240
     },
     {
-      "epoch": 1.9835390946502058,
-      "grad_norm": 0.24004040802554896,
-      "learning_rate": 5.444197814328707e-06,
-      "loss": 0.0374,
+      "epoch": 1.6394557823129252,
+      "grad_norm": 0.39602394734020263,
+      "learning_rate": 9.00838872791973e-06,
+      "loss": 0.0389,
       "step": 241
     },
     {
-      "epoch": 1.991769547325103,
-      "grad_norm": 0.3542552755665731,
-      "learning_rate": 5.364930620448946e-06,
-      "loss": 0.0364,
+      "epoch": 1.6462585034013606,
+      "grad_norm": 0.2811227178639146,
+      "learning_rate": 8.935205243760022e-06,
+      "loss": 0.0301,
       "step": 242
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.3483540336202611,
-      "learning_rate": 5.286032631740023e-06,
-      "loss": 0.0322,
+      "epoch": 1.6530612244897958,
+      "grad_norm": 0.2738235813523902,
+      "learning_rate": 8.862079397472552e-06,
+      "loss": 0.0295,
       "step": 243
     },
     {
-      "epoch": 2.0082304526748973,
-      "grad_norm": 0.2773566591891378,
-      "learning_rate": 5.207510132799436e-06,
-      "loss": 0.0378,
+      "epoch": 1.6598639455782314,
+      "grad_norm": 0.325061391409321,
+      "learning_rate": 8.78901514739592e-06,
+      "loss": 0.029,
       "step": 244
     },
     {
-      "epoch": 2.016460905349794,
-      "grad_norm": 0.24177358132222973,
-      "learning_rate": 5.129369378315128e-06,
-      "loss": 0.0393,
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.3027200222791084,
+      "learning_rate": 8.71601644853449e-06,
+      "loss": 0.0338,
       "step": 245
     },
     {
-      "epoch": 2.0246913580246915,
-      "grad_norm": 0.3169366023010318,
-      "learning_rate": 5.051616592567323e-06,
-      "loss": 0.0366,
+      "epoch": 1.6734693877551021,
+      "grad_norm": 0.3891063981452845,
+      "learning_rate": 8.643087252344313e-06,
+      "loss": 0.0299,
       "step": 246
     },
     {
-      "epoch": 2.0329218106995883,
-      "grad_norm": 0.35089036335173035,
-      "learning_rate": 4.974257968932687e-06,
-      "loss": 0.0377,
+      "epoch": 1.6802721088435373,
+      "grad_norm": 0.31302794618115964,
+      "learning_rate": 8.57023150651922e-06,
+      "loss": 0.03,
       "step": 247
     },
     {
-      "epoch": 2.0411522633744856,
-      "grad_norm": 0.29634232447720477,
-      "learning_rate": 4.897299669391006e-06,
-      "loss": 0.0373,
+      "epoch": 1.6870748299319729,
+      "grad_norm": 0.3449817518564228,
+      "learning_rate": 8.49745315477714e-06,
+      "loss": 0.0317,
       "step": 248
     },
     {
-      "epoch": 2.049382716049383,
-      "grad_norm": 0.1916448625769453,
-      "learning_rate": 4.820747824034369e-06,
-      "loss": 0.0342,
+      "epoch": 1.693877551020408,
+      "grad_norm": 0.29810889691448594,
+      "learning_rate": 8.424756136646624e-06,
+      "loss": 0.0291,
       "step": 249
     },
     {
-      "epoch": 2.05761316872428,
-      "grad_norm": 0.44747684852654307,
-      "learning_rate": 4.744608530578872e-06,
-      "loss": 0.0345,
+      "epoch": 1.7006802721088436,
+      "grad_norm": 0.3798859940115675,
+      "learning_rate": 8.352144387253582e-06,
+      "loss": 0.0322,
       "step": 250
     },
     {
-      "epoch": 2.065843621399177,
-      "grad_norm": 0.20682311235092382,
-      "learning_rate": 4.668887853878896e-06,
-      "loss": 0.0361,
+      "epoch": 1.7074829931972788,
+      "grad_norm": 0.21879122540865556,
+      "learning_rate": 8.279621837108295e-06,
+      "loss": 0.0317,
       "step": 251
     },
     {
-      "epoch": 2.074074074074074,
-      "grad_norm": 0.28869180974004,
-      "learning_rate": 4.593591825444028e-06,
-      "loss": 0.04,
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.35646635601921695,
+      "learning_rate": 8.207192411892645e-06,
+      "loss": 0.0266,
       "step": 252
     },
     {
-      "epoch": 2.0823045267489713,
-      "grad_norm": 0.23532131252759386,
-      "learning_rate": 4.518726442958599e-06,
-      "loss": 0.0381,
+      "epoch": 1.7210884353741496,
+      "grad_norm": 0.32902401726950553,
+      "learning_rate": 8.134860032247613e-06,
+      "loss": 0.0281,
       "step": 253
     },
     {
-      "epoch": 2.090534979423868,
-      "grad_norm": 0.4498707807244971,
-      "learning_rate": 4.444297669803981e-06,
-      "loss": 0.0356,
+      "epoch": 1.7278911564625852,
+      "grad_norm": 0.2954321296424936,
+      "learning_rate": 8.062628613561051e-06,
+      "loss": 0.0288,
       "step": 254
     },
     {
-      "epoch": 2.0987654320987654,
-      "grad_norm": 0.3317926606932244,
-      "learning_rate": 4.370311434583525e-06,
-      "loss": 0.0388,
+      "epoch": 1.7346938775510203,
+      "grad_norm": 0.33576585838310874,
+      "learning_rate": 7.990502065755748e-06,
+      "loss": 0.0283,
       "step": 255
     },
     {
-      "epoch": 2.1069958847736627,
-      "grad_norm": 0.2732518053786639,
-      "learning_rate": 4.296773630650358e-06,
-      "loss": 0.0335,
+      "epoch": 1.741496598639456,
+      "grad_norm": 0.372785587382324,
+      "learning_rate": 7.918484293077777e-06,
+      "loss": 0.0325,
       "step": 256
     },
     {
-      "epoch": 2.1152263374485596,
-      "grad_norm": 0.23205013937953725,
-      "learning_rate": 4.223690115637944e-06,
-      "loss": 0.0358,
+      "epoch": 1.748299319727891,
+      "grad_norm": 0.2699962005993758,
+      "learning_rate": 7.846579193885165e-06,
+      "loss": 0.0245,
       "step": 257
     },
     {
-      "epoch": 2.123456790123457,
-      "grad_norm": 0.2849463542938286,
-      "learning_rate": 4.15106671099347e-06,
-      "loss": 0.0402,
+      "epoch": 1.7551020408163265,
+      "grad_norm": 0.4069704653536772,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.0331,
       "step": 258
     },
     {
-      "epoch": 2.1316872427983538,
-      "grad_norm": 0.29038779306625356,
-      "learning_rate": 4.078909201514172e-06,
-      "loss": 0.0377,
+      "epoch": 1.7619047619047619,
+      "grad_norm": 0.2543285657041178,
+      "learning_rate": 7.703122578682047e-06,
+      "loss": 0.0195,
       "step": 259
     },
     {
-      "epoch": 2.139917695473251,
-      "grad_norm": 0.333361843536132,
-      "learning_rate": 4.007223334886531e-06,
-      "loss": 0.0376,
+      "epoch": 1.7687074829931972,
+      "grad_norm": 0.49851302483614535,
+      "learning_rate": 7.631578828049809e-06,
+      "loss": 0.0228,
       "step": 260
     },
     {
-      "epoch": 2.148148148148148,
-      "grad_norm": 0.2443167653220952,
-      "learning_rate": 3.936014821228448e-06,
-      "loss": 0.0363,
+      "epoch": 1.7755102040816326,
+      "grad_norm": 0.34904705107956824,
+      "learning_rate": 7.560163281239116e-06,
+      "loss": 0.0248,
       "step": 261
     },
     {
-      "epoch": 2.156378600823045,
-      "grad_norm": 0.31829486832422227,
-      "learning_rate": 3.865289332634407e-06,
-      "loss": 0.0358,
+      "epoch": 1.782312925170068,
+      "grad_norm": 0.3079169045046351,
+      "learning_rate": 7.488879804009206e-06,
+      "loss": 0.029,
       "step": 262
     },
     {
-      "epoch": 2.1646090534979425,
-      "grad_norm": 0.25652876596392077,
-      "learning_rate": 3.7950525027236585e-06,
-      "loss": 0.0351,
+      "epoch": 1.7891156462585034,
+      "grad_norm": 0.32573423282858954,
+      "learning_rate": 7.4177322549703165e-06,
+      "loss": 0.0246,
       "step": 263
     },
     {
-      "epoch": 2.1728395061728394,
-      "grad_norm": 0.4391324686498631,
-      "learning_rate": 3.7253099261914794e-06,
-      "loss": 0.0322,
+      "epoch": 1.7959183673469388,
+      "grad_norm": 0.3112809895061056,
+      "learning_rate": 7.346724485374837e-06,
+      "loss": 0.0299,
       "step": 264
     },
     {
-      "epoch": 2.1810699588477367,
-      "grad_norm": 0.27988686477069163,
-      "learning_rate": 3.6560671583635467e-06,
-      "loss": 0.0329,
+      "epoch": 1.8027210884353742,
+      "grad_norm": 0.2874275900338384,
+      "learning_rate": 7.275860338908815e-06,
+      "loss": 0.0299,
       "step": 265
     },
     {
-      "epoch": 2.1893004115226335,
-      "grad_norm": 0.271620195292481,
-      "learning_rate": 3.5873297147533913e-06,
-      "loss": 0.0359,
+      "epoch": 1.8095238095238095,
+      "grad_norm": 0.37722251427133746,
+      "learning_rate": 7.2051436514839064e-06,
+      "loss": 0.0295,
       "step": 266
     },
     {
-      "epoch": 2.197530864197531,
-      "grad_norm": 0.2899553325044773,
-      "learning_rate": 3.5191030706230967e-06,
-      "loss": 0.0318,
+      "epoch": 1.816326530612245,
+      "grad_norm": 0.4495052816684312,
+      "learning_rate": 7.134578251029745e-06,
+      "loss": 0.0287,
       "step": 267
     },
     {
-      "epoch": 2.205761316872428,
-      "grad_norm": 0.2816230189984451,
-      "learning_rate": 3.4513926605471504e-06,
-      "loss": 0.0364,
+      "epoch": 1.8231292517006803,
+      "grad_norm": 0.4383184010849996,
+      "learning_rate": 7.064167957286714e-06,
+      "loss": 0.0276,
       "step": 268
     },
     {
-      "epoch": 2.213991769547325,
-      "grad_norm": 0.2635920928214212,
-      "learning_rate": 3.3842038779795594e-06,
-      "loss": 0.0356,
+      "epoch": 1.8299319727891157,
+      "grad_norm": 0.34023669062637585,
+      "learning_rate": 6.993916581599203e-06,
+      "loss": 0.0271,
       "step": 269
     },
     {
-      "epoch": 2.2222222222222223,
-      "grad_norm": 0.36048326165593647,
-      "learning_rate": 3.3175420748242405e-06,
-      "loss": 0.0393,
+      "epoch": 1.836734693877551,
+      "grad_norm": 0.4710800088561054,
+      "learning_rate": 6.923827926709277e-06,
+      "loss": 0.0409,
       "step": 270
     },
     {
-      "epoch": 2.230452674897119,
-      "grad_norm": 0.23321488261836193,
-      "learning_rate": 3.2514125610086957e-06,
-      "loss": 0.0366,
+      "epoch": 1.8435374149659864,
+      "grad_norm": 0.5275268904537102,
+      "learning_rate": 6.853905786550855e-06,
+      "loss": 0.0296,
       "step": 271
     },
     {
-      "epoch": 2.2386831275720165,
-      "grad_norm": 0.23619597803672424,
-      "learning_rate": 3.1858206040610883e-06,
-      "loss": 0.0319,
+      "epoch": 1.8503401360544216,
+      "grad_norm": 0.33912114616353456,
+      "learning_rate": 6.784153946044321e-06,
+      "loss": 0.0292,
       "step": 272
     },
     {
-      "epoch": 2.246913580246914,
-      "grad_norm": 0.32718410423908234,
-      "learning_rate": 3.1207714286906253e-06,
-      "loss": 0.0323,
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.5301308829618651,
+      "learning_rate": 6.714576180891653e-06,
+      "loss": 0.0279,
       "step": 273
     },
     {
-      "epoch": 2.2551440329218106,
-      "grad_norm": 0.25330884010417526,
-      "learning_rate": 3.0562702163713954e-06,
-      "loss": 0.0305,
+      "epoch": 1.8639455782312924,
+      "grad_norm": 0.36662643442740994,
+      "learning_rate": 6.645176257372054e-06,
+      "loss": 0.0279,
       "step": 274
     },
     {
-      "epoch": 2.263374485596708,
-      "grad_norm": 0.3057301438355615,
-      "learning_rate": 2.9923221049296448e-06,
-      "loss": 0.0307,
+      "epoch": 1.870748299319728,
+      "grad_norm": 0.6307178798575221,
+      "learning_rate": 6.5759579321380576e-06,
+      "loss": 0.0255,
       "step": 275
     },
     {
-      "epoch": 2.271604938271605,
-      "grad_norm": 0.2716663498442403,
-      "learning_rate": 2.9289321881345257e-06,
-      "loss": 0.0271,
+      "epoch": 1.8775510204081631,
+      "grad_norm": 0.3391198098709979,
+      "learning_rate": 6.5069249520122026e-06,
+      "loss": 0.0299,
       "step": 276
     },
     {
-      "epoch": 2.279835390946502,
-      "grad_norm": 0.26549969017008895,
-      "learning_rate": 2.8661055152923456e-06,
-      "loss": 0.0381,
+      "epoch": 1.8843537414965987,
+      "grad_norm": 0.4817806870006288,
+      "learning_rate": 6.438081053784197e-06,
+      "loss": 0.0293,
       "step": 277
     },
     {
-      "epoch": 2.288065843621399,
-      "grad_norm": 0.36572210412714806,
-      "learning_rate": 2.8038470908443717e-06,
-      "loss": 0.0308,
+      "epoch": 1.891156462585034,
+      "grad_norm": 0.4572796789157729,
+      "learning_rate": 6.36942996400865e-06,
+      "loss": 0.0309,
       "step": 278
     },
     {
-      "epoch": 2.2962962962962963,
-      "grad_norm": 0.3750500553391758,
-      "learning_rate": 2.742161873968202e-06,
-      "loss": 0.031,
+      "epoch": 1.8979591836734695,
+      "grad_norm": 0.3943436654831781,
+      "learning_rate": 6.300975398803362e-06,
+      "loss": 0.0274,
       "step": 279
     },
     {
-      "epoch": 2.3045267489711936,
-      "grad_norm": 0.2917432327230002,
-      "learning_rate": 2.681054778182748e-06,
-      "loss": 0.0345,
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.39513158230874584,
+      "learning_rate": 6.232721063648148e-06,
+      "loss": 0.0232,
       "step": 280
     },
     {
-      "epoch": 2.3127572016460904,
-      "grad_norm": 0.26239122275469073,
-      "learning_rate": 2.6205306709568358e-06,
-      "loss": 0.0339,
+      "epoch": 1.9115646258503403,
+      "grad_norm": 0.3405190746199494,
+      "learning_rate": 6.1646706531842845e-06,
+      "loss": 0.0249,
       "step": 281
     },
     {
-      "epoch": 2.3209876543209877,
-      "grad_norm": 0.26177687774458025,
-      "learning_rate": 2.5605943733215044e-06,
-      "loss": 0.0317,
+      "epoch": 1.9183673469387754,
+      "grad_norm": 0.3938020668564346,
+      "learning_rate": 6.09682785101449e-06,
+      "loss": 0.0277,
       "step": 282
     },
     {
-      "epoch": 2.3292181069958846,
-      "grad_norm": 0.27506190906633227,
-      "learning_rate": 2.501250659485992e-06,
-      "loss": 0.0334,
+      "epoch": 1.925170068027211,
+      "grad_norm": 0.3802867002071383,
+      "learning_rate": 6.029196329503548e-06,
+      "loss": 0.0264,
       "step": 283
     },
     {
-      "epoch": 2.337448559670782,
-      "grad_norm": 0.24944589535813674,
-      "learning_rate": 2.4425042564574186e-06,
-      "loss": 0.0359,
+      "epoch": 1.9319727891156462,
+      "grad_norm": 0.28045602092848787,
+      "learning_rate": 5.961779749579516e-06,
+      "loss": 0.0209,
       "step": 284
     },
     {
-      "epoch": 2.3456790123456788,
-      "grad_norm": 0.26548114143765855,
-      "learning_rate": 2.38435984366429e-06,
-      "loss": 0.0312,
+      "epoch": 1.9387755102040818,
+      "grad_norm": 0.33135168744952337,
+      "learning_rate": 5.8945817605355495e-06,
+      "loss": 0.0192,
       "step": 285
     },
     {
-      "epoch": 2.353909465020576,
-      "grad_norm": 0.2972312195538176,
-      "learning_rate": 2.3268220525837436e-06,
-      "loss": 0.0365,
+      "epoch": 1.945578231292517,
+      "grad_norm": 0.310525514277901,
+      "learning_rate": 5.827605999832375e-06,
+      "loss": 0.0179,
       "step": 286
     },
     {
-      "epoch": 2.3621399176954734,
-      "grad_norm": 0.4570974860281035,
-      "learning_rate": 2.26989546637263e-06,
-      "loss": 0.0362,
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.4333347605749311,
+      "learning_rate": 5.760856092901394e-06,
+      "loss": 0.0299,
       "step": 287
     },
     {
-      "epoch": 2.3703703703703702,
-      "grad_norm": 0.3126236268623473,
-      "learning_rate": 2.213584619502451e-06,
-      "loss": 0.0344,
+      "epoch": 1.9591836734693877,
+      "grad_norm": 0.2600623965752405,
+      "learning_rate": 5.694335652948415e-06,
+      "loss": 0.0222,
       "step": 288
     },
     {
-      "epoch": 2.3786008230452675,
-      "grad_norm": 0.29642687322126127,
-      "learning_rate": 2.157893997398146e-06,
-      "loss": 0.0351,
+      "epoch": 1.965986394557823,
+      "grad_norm": 0.3654991906697444,
+      "learning_rate": 5.628048280758096e-06,
+      "loss": 0.0299,
       "step": 289
     },
     {
-      "epoch": 2.386831275720165,
-      "grad_norm": 0.29453794842791514,
-      "learning_rate": 2.1028280360808405e-06,
-      "loss": 0.0297,
+      "epoch": 1.9727891156462585,
+      "grad_norm": 0.2836245750881348,
+      "learning_rate": 5.561997564499024e-06,
+      "loss": 0.0266,
       "step": 290
     },
     {
-      "epoch": 2.3950617283950617,
-      "grad_norm": 0.26602023500151717,
-      "learning_rate": 2.0483911218144713e-06,
-      "loss": 0.031,
+      "epoch": 1.9795918367346939,
+      "grad_norm": 0.3772557920038287,
+      "learning_rate": 5.4961870795294644e-06,
+      "loss": 0.028,
       "step": 291
     },
     {
-      "epoch": 2.403292181069959,
-      "grad_norm": 0.265514494777824,
-      "learning_rate": 1.994587590756397e-06,
-      "loss": 0.0338,
+      "epoch": 1.9863945578231292,
+      "grad_norm": 0.2390418620516205,
+      "learning_rate": 5.430620388203866e-06,
+      "loss": 0.0181,
       "step": 292
     },
     {
-      "epoch": 2.411522633744856,
-      "grad_norm": 0.34135343141394575,
-      "learning_rate": 1.941421728612023e-06,
-      "loss": 0.0327,
+      "epoch": 1.9931972789115646,
+      "grad_norm": 0.26917774514431425,
+      "learning_rate": 5.365301039679985e-06,
+      "loss": 0.0252,
       "step": 293
     },
     {
-      "epoch": 2.419753086419753,
-      "grad_norm": 0.3467441938509079,
-      "learning_rate": 1.8888977702934086e-06,
-      "loss": 0.03,
+      "epoch": 2.0,
+      "grad_norm": 0.3027370695631011,
+      "learning_rate": 5.300232569726805e-06,
+      "loss": 0.0187,
       "step": 294
     },
     {
-      "epoch": 2.42798353909465,
-      "grad_norm": 0.29821869856510014,
-      "learning_rate": 1.8370198995819432e-06,
-      "loss": 0.0317,
+      "epoch": 2.006802721088435,
+      "grad_norm": 0.5374555027441916,
+      "learning_rate": 5.2354185005331095e-06,
+      "loss": 0.0282,
       "step": 295
     },
     {
-      "epoch": 2.4362139917695473,
-      "grad_norm": 0.3056847220574039,
-      "learning_rate": 1.7857922487950873e-06,
-      "loss": 0.033,
+      "epoch": 2.0136054421768708,
+      "grad_norm": 0.23930806738660212,
+      "learning_rate": 5.170862340516858e-06,
+      "loss": 0.0149,
       "step": 296
     },
     {
-      "epoch": 2.4444444444444446,
-      "grad_norm": 0.3168972116272592,
-      "learning_rate": 1.7352188984572026e-06,
-      "loss": 0.0316,
+      "epoch": 2.020408163265306,
+      "grad_norm": 0.32072539854622056,
+      "learning_rate": 5.106567584135251e-06,
+      "loss": 0.0255,
       "step": 297
     },
     {
-      "epoch": 2.4526748971193415,
-      "grad_norm": 0.29877332862932127,
-      "learning_rate": 1.6853038769745466e-06,
-      "loss": 0.0343,
+      "epoch": 2.0272108843537415,
+      "grad_norm": 0.3765856363755405,
+      "learning_rate": 5.042537711695584e-06,
+      "loss": 0.0265,
       "step": 298
     },
     {
-      "epoch": 2.460905349794239,
-      "grad_norm": 0.32443955464300556,
-      "learning_rate": 1.6360511603143648e-06,
-      "loss": 0.0341,
+      "epoch": 2.0340136054421767,
+      "grad_norm": 0.375918284328471,
+      "learning_rate": 4.97877618916684e-06,
+      "loss": 0.0296,
       "step": 299
     },
     {
-      "epoch": 2.4691358024691357,
-      "grad_norm": 0.27303073439051084,
-      "learning_rate": 1.587464671688187e-06,
-      "loss": 0.0318,
+      "epoch": 2.0408163265306123,
+      "grad_norm": 0.28544218268890903,
+      "learning_rate": 4.915286467992098e-06,
+      "loss": 0.0266,
       "step": 300
     },
     {
-      "epoch": 2.477366255144033,
-      "grad_norm": 0.3397831672229159,
-      "learning_rate": 1.5395482812393513e-06,
-      "loss": 0.0343,
+      "epoch": 2.0476190476190474,
+      "grad_norm": 0.29041194755799743,
+      "learning_rate": 4.852071984901696e-06,
+      "loss": 0.0243,
       "step": 301
     },
     {
-      "epoch": 2.48559670781893,
-      "grad_norm": 0.2498007989451009,
-      "learning_rate": 1.492305805734693e-06,
-      "loss": 0.0306,
+      "epoch": 2.054421768707483,
+      "grad_norm": 0.289138208611435,
+      "learning_rate": 4.789136161727184e-06,
+      "loss": 0.0193,
       "step": 302
     },
     {
-      "epoch": 2.493827160493827,
-      "grad_norm": 0.2817138724568094,
-      "learning_rate": 1.4457410082605483e-06,
-      "loss": 0.0326,
+      "epoch": 2.061224489795918,
+      "grad_norm": 0.39699377159750365,
+      "learning_rate": 4.7264824052161255e-06,
+      "loss": 0.0213,
       "step": 303
     },
     {
-      "epoch": 2.5020576131687244,
-      "grad_norm": 0.2584140929552954,
-      "learning_rate": 1.3998575979229944e-06,
-      "loss": 0.0342,
+      "epoch": 2.068027210884354,
+      "grad_norm": 0.2981351983066942,
+      "learning_rate": 4.664114106847667e-06,
+      "loss": 0.0157,
       "step": 304
     },
     {
-      "epoch": 2.5102880658436213,
-      "grad_norm": 0.2948187429496015,
-      "learning_rate": 1.3546592295524075e-06,
-      "loss": 0.0331,
+      "epoch": 2.074829931972789,
+      "grad_norm": 0.41346371414139993,
+      "learning_rate": 4.602034642648968e-06,
+      "loss": 0.0222,
       "step": 305
     },
     {
-      "epoch": 2.5185185185185186,
-      "grad_norm": 0.3041583210158896,
-      "learning_rate": 1.3101495034123313e-06,
-      "loss": 0.0364,
+      "epoch": 2.0816326530612246,
+      "grad_norm": 0.46281644559656343,
+      "learning_rate": 4.5402473730124395e-06,
+      "loss": 0.0228,
       "step": 306
     },
     {
-      "epoch": 2.526748971193416,
-      "grad_norm": 0.2668123815457834,
-      "learning_rate": 1.2663319649127025e-06,
-      "loss": 0.0313,
+      "epoch": 2.0884353741496597,
+      "grad_norm": 0.36453555154639505,
+      "learning_rate": 4.478755642513868e-06,
+      "loss": 0.0214,
       "step": 307
     },
     {
-      "epoch": 2.5349794238683128,
-      "grad_norm": 0.31297488610646934,
-      "learning_rate": 1.2232101043274437e-06,
-      "loss": 0.0319,
+      "epoch": 2.0952380952380953,
+      "grad_norm": 0.3747842217307271,
+      "learning_rate": 4.417562779731355e-06,
+      "loss": 0.0237,
       "step": 308
     },
     {
-      "epoch": 2.5432098765432096,
-      "grad_norm": 0.34249562678045614,
-      "learning_rate": 1.1807873565164507e-06,
-      "loss": 0.0347,
+      "epoch": 2.1020408163265305,
+      "grad_norm": 0.25516596155835825,
+      "learning_rate": 4.356672097065134e-06,
+      "loss": 0.0227,
       "step": 309
     },
     {
-      "epoch": 2.551440329218107,
-      "grad_norm": 0.300856829760047,
-      "learning_rate": 1.139067100651976e-06,
-      "loss": 0.0338,
+      "epoch": 2.108843537414966,
+      "grad_norm": 0.3978996621254664,
+      "learning_rate": 4.2960868905582895e-06,
+      "loss": 0.0251,
       "step": 310
     },
     {
-      "epoch": 2.5596707818930042,
-      "grad_norm": 0.28196907702239665,
-      "learning_rate": 1.0980526599494733e-06,
-      "loss": 0.0276,
+      "epoch": 2.1156462585034013,
+      "grad_norm": 0.37774170305452065,
+      "learning_rate": 4.235810439718327e-06,
+      "loss": 0.0185,
       "step": 311
     },
     {
-      "epoch": 2.567901234567901,
-      "grad_norm": 0.24385678172428948,
-      "learning_rate": 1.0577473014028872e-06,
-      "loss": 0.0305,
+      "epoch": 2.122448979591837,
+      "grad_norm": 0.5276653229249155,
+      "learning_rate": 4.175846007339644e-06,
+      "loss": 0.0175,
       "step": 312
     },
     {
-      "epoch": 2.5761316872427984,
-      "grad_norm": 0.26824511630947134,
-      "learning_rate": 1.0181542355244167e-06,
-      "loss": 0.0274,
+      "epoch": 2.129251700680272,
+      "grad_norm": 0.45335169134199393,
+      "learning_rate": 4.1161968393269324e-06,
+      "loss": 0.0176,
       "step": 313
     },
     {
-      "epoch": 2.5843621399176957,
-      "grad_norm": 0.360571686445375,
-      "learning_rate": 9.792766160887868e-07,
-      "loss": 0.0308,
+      "epoch": 2.1360544217687076,
+      "grad_norm": 0.3501639808861222,
+      "learning_rate": 4.0568661645194656e-06,
+      "loss": 0.0229,
       "step": 314
     },
     {
-      "epoch": 2.5925925925925926,
-      "grad_norm": 0.2644116563787061,
-      "learning_rate": 9.411175398820271e-07,
-      "loss": 0.0302,
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.36922699495113326,
+      "learning_rate": 3.997857194516319e-06,
+      "loss": 0.0247,
       "step": 315
     },
     {
-      "epoch": 2.60082304526749,
-      "grad_norm": 0.2728520330553383,
-      "learning_rate": 9.036800464548157e-07,
-      "loss": 0.033,
+      "epoch": 2.1496598639455784,
+      "grad_norm": 0.3844861347028257,
+      "learning_rate": 3.939173123502523e-06,
+      "loss": 0.0199,
       "step": 316
     },
     {
-      "epoch": 2.6090534979423867,
-      "grad_norm": 0.4186449774266214,
-      "learning_rate": 8.669671178803485e-07,
-      "loss": 0.03,
+      "epoch": 2.1564625850340136,
+      "grad_norm": 0.645854225510273,
+      "learning_rate": 3.8808171280761665e-06,
+      "loss": 0.0208,
       "step": 317
     },
     {
-      "epoch": 2.617283950617284,
-      "grad_norm": 0.2786905984699959,
-      "learning_rate": 8.309816785168035e-07,
-      "loss": 0.0276,
+      "epoch": 2.163265306122449,
+      "grad_norm": 0.3101470346726489,
+      "learning_rate": 3.822792367076446e-06,
+      "loss": 0.0168,
       "step": 318
     },
     {
-      "epoch": 2.625514403292181,
-      "grad_norm": 0.27629461864767646,
-      "learning_rate": 7.957265947744131e-07,
-      "loss": 0.0265,
+      "epoch": 2.1700680272108843,
+      "grad_norm": 0.4327622148598966,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.0194,
       "step": 319
     },
     {
-      "epoch": 2.633744855967078,
-      "grad_norm": 0.5344854955583579,
-      "learning_rate": 7.612046748871327e-07,
-      "loss": 0.0273,
+      "epoch": 2.17687074829932,
+      "grad_norm": 0.3994172850300765,
+      "learning_rate": 3.7077490938942307e-06,
+      "loss": 0.0189,
       "step": 320
     },
     {
-      "epoch": 2.6419753086419755,
-      "grad_norm": 0.35311473470046706,
-      "learning_rate": 7.274186686889539e-07,
-      "loss": 0.0319,
+      "epoch": 2.183673469387755,
+      "grad_norm": 0.23341486544176343,
+      "learning_rate": 3.6507368090616014e-06,
+      "loss": 0.0154,
       "step": 321
     },
     {
-      "epoch": 2.6502057613168724,
-      "grad_norm": 0.2666014135758544,
-      "learning_rate": 6.943712673948643e-07,
-      "loss": 0.032,
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.2800782727579594,
+      "learning_rate": 3.594068213018249e-06,
+      "loss": 0.0271,
       "step": 322
     },
     {
-      "epoch": 2.6584362139917697,
-      "grad_norm": 0.2967717858448913,
-      "learning_rate": 6.620651033864844e-07,
-      "loss": 0.034,
+      "epoch": 2.197278911564626,
+      "grad_norm": 0.41245009605814564,
+      "learning_rate": 3.53774637326359e-06,
+      "loss": 0.0211,
       "step": 323
     },
     {
-      "epoch": 2.6666666666666665,
-      "grad_norm": 0.24817731685513944,
-      "learning_rate": 6.305027500023841e-07,
-      "loss": 0.0303,
+      "epoch": 2.204081632653061,
+      "grad_norm": 0.44102425000184575,
+      "learning_rate": 3.481774338526954e-06,
+      "loss": 0.0203,
       "step": 324
     },
     {
-      "epoch": 2.674897119341564,
-      "grad_norm": 0.2851225185275127,
-      "learning_rate": 5.996867213330993e-07,
-      "loss": 0.0283,
+      "epoch": 2.2108843537414966,
+      "grad_norm": 0.3612731375372648,
+      "learning_rate": 3.426155138602558e-06,
+      "loss": 0.0109,
       "step": 325
     },
     {
-      "epoch": 2.6831275720164607,
-      "grad_norm": 0.26333175021397076,
-      "learning_rate": 5.696194720208792e-07,
-      "loss": 0.0285,
+      "epoch": 2.2176870748299318,
+      "grad_norm": 0.41003267633028256,
+      "learning_rate": 3.3708917841854782e-06,
+      "loss": 0.0254,
       "step": 326
     },
     {
-      "epoch": 2.691358024691358,
-      "grad_norm": 0.271644698776009,
-      "learning_rate": 5.403033970641647e-07,
-      "loss": 0.0262,
+      "epoch": 2.2244897959183674,
+      "grad_norm": 0.44409978093074687,
+      "learning_rate": 3.3159872667087077e-06,
+      "loss": 0.0219,
       "step": 327
     },
     {
-      "epoch": 2.6995884773662553,
-      "grad_norm": 0.2591873823608315,
-      "learning_rate": 5.117408316268047e-07,
-      "loss": 0.0307,
+      "epoch": 2.2312925170068025,
+      "grad_norm": 0.4310743734288755,
+      "learning_rate": 3.2614445581812183e-06,
+      "loss": 0.0201,
       "step": 328
     },
     {
-      "epoch": 2.707818930041152,
-      "grad_norm": 0.33885144865871347,
-      "learning_rate": 4.839340508520563e-07,
-      "loss": 0.0304,
+      "epoch": 2.238095238095238,
+      "grad_norm": 0.3035091737111279,
+      "learning_rate": 3.207266611027069e-06,
+      "loss": 0.024,
       "step": 329
     },
     {
-      "epoch": 2.7160493827160495,
-      "grad_norm": 0.3218551816126933,
-      "learning_rate": 4.5688526968136193e-07,
-      "loss": 0.0317,
+      "epoch": 2.2448979591836733,
+      "grad_norm": 0.3671044820236742,
+      "learning_rate": 3.1534563579256172e-06,
+      "loss": 0.024,
       "step": 330
     },
     {
-      "epoch": 2.7242798353909468,
-      "grad_norm": 0.2753036002553041,
-      "learning_rate": 4.305966426779118e-07,
-      "loss": 0.0331,
+      "epoch": 2.251700680272109,
+      "grad_norm": 0.3619582161667144,
+      "learning_rate": 3.1000167116527525e-06,
+      "loss": 0.0203,
       "step": 331
     },
     {
-      "epoch": 2.7325102880658436,
-      "grad_norm": 0.29341125008567315,
-      "learning_rate": 4.0507026385502747e-07,
-      "loss": 0.0317,
+      "epoch": 2.258503401360544,
+      "grad_norm": 0.450706470571488,
+      "learning_rate": 3.0469505649232333e-06,
+      "loss": 0.029,
       "step": 332
     },
     {
-      "epoch": 2.7407407407407405,
-      "grad_norm": 0.3417963876466253,
-      "learning_rate": 3.8030816650935777e-07,
-      "loss": 0.0332,
+      "epoch": 2.2653061224489797,
+      "grad_norm": 0.3953038026535905,
+      "learning_rate": 2.9942607902340946e-06,
+      "loss": 0.021,
       "step": 333
     },
     {
-      "epoch": 2.748971193415638,
-      "grad_norm": 0.23819240441996148,
-      "learning_rate": 3.5631232305893047e-07,
-      "loss": 0.033,
+      "epoch": 2.272108843537415,
+      "grad_norm": 0.45052845721729323,
+      "learning_rate": 2.9419502397091715e-06,
+      "loss": 0.0239,
       "step": 334
     },
     {
-      "epoch": 2.757201646090535,
-      "grad_norm": 0.3485988195720003,
-      "learning_rate": 3.3308464488602587e-07,
-      "loss": 0.0348,
+      "epoch": 2.2789115646258504,
+      "grad_norm": 0.40042405402317627,
+      "learning_rate": 2.8900217449447077e-06,
+      "loss": 0.0162,
       "step": 335
     },
     {
-      "epoch": 2.765432098765432,
-      "grad_norm": 0.2745564158660002,
-      "learning_rate": 3.106269821849273e-07,
-      "loss": 0.0281,
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.26311574813539457,
+      "learning_rate": 2.8384781168560693e-06,
+      "loss": 0.022,
       "step": 336
     },
     {
-      "epoch": 2.7736625514403292,
-      "grad_norm": 0.40955337075489795,
-      "learning_rate": 2.889411238145545e-07,
-      "loss": 0.0302,
+      "epoch": 2.292517006802721,
+      "grad_norm": 0.3897033453197049,
+      "learning_rate": 2.7873221455256006e-06,
+      "loss": 0.0207,
       "step": 337
     },
     {
-      "epoch": 2.7818930041152266,
-      "grad_norm": 0.2913163904501433,
-      "learning_rate": 2.6802879715596585e-07,
-      "loss": 0.0305,
+      "epoch": 2.2993197278911564,
+      "grad_norm": 0.3677858687841903,
+      "learning_rate": 2.736556600051593e-06,
+      "loss": 0.023,
       "step": 338
     },
     {
-      "epoch": 2.7901234567901234,
-      "grad_norm": 0.3140479694030182,
-      "learning_rate": 2.478916679747623e-07,
-      "loss": 0.0307,
+      "epoch": 2.306122448979592,
+      "grad_norm": 0.3577516109284015,
+      "learning_rate": 2.6861842283983953e-06,
+      "loss": 0.0219,
       "step": 339
     },
     {
-      "epoch": 2.7983539094650207,
-      "grad_norm": 0.2452162407259469,
-      "learning_rate": 2.2853134028840594e-07,
-      "loss": 0.0293,
+      "epoch": 2.312925170068027,
+      "grad_norm": 0.24621923670001875,
+      "learning_rate": 2.6362077572476495e-06,
+      "loss": 0.0166,
       "step": 340
     },
     {
-      "epoch": 2.8065843621399176,
-      "grad_norm": 0.3024785946569477,
-      "learning_rate": 2.099493562384469e-07,
-      "loss": 0.0269,
+      "epoch": 2.3197278911564627,
+      "grad_norm": 0.26774100034818,
+      "learning_rate": 2.586629891850716e-06,
+      "loss": 0.0149,
       "step": 341
     },
     {
-      "epoch": 2.814814814814815,
-      "grad_norm": 0.7761466702382431,
-      "learning_rate": 1.921471959676957e-07,
-      "loss": 0.0345,
+      "epoch": 2.326530612244898,
+      "grad_norm": 0.3498362662110048,
+      "learning_rate": 2.5374533158822225e-06,
+      "loss": 0.0203,
       "step": 342
     },
     {
-      "epoch": 2.8230452674897117,
-      "grad_norm": 0.3019994183119137,
-      "learning_rate": 1.7512627750230772e-07,
-      "loss": 0.0326,
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.4379991319756946,
+      "learning_rate": 2.4886806912948034e-06,
+      "loss": 0.0234,
       "step": 343
     },
     {
-      "epoch": 2.831275720164609,
-      "grad_norm": 0.30325683843695656,
-      "learning_rate": 1.5888795663883904e-07,
-      "loss": 0.0296,
+      "epoch": 2.3401360544217686,
+      "grad_norm": 0.4484515975309493,
+      "learning_rate": 2.4403146581749925e-06,
+      "loss": 0.023,
       "step": 344
     },
     {
-      "epoch": 2.8395061728395063,
-      "grad_norm": 0.3002527524913964,
-      "learning_rate": 1.4343352683625412e-07,
-      "loss": 0.0326,
+      "epoch": 2.3469387755102042,
+      "grad_norm": 0.24634210193643918,
+      "learning_rate": 2.392357834600336e-06,
+      "loss": 0.0151,
       "step": 345
     },
     {
-      "epoch": 2.847736625514403,
-      "grad_norm": 0.3585727711861594,
-      "learning_rate": 1.2876421911288906e-07,
-      "loss": 0.0346,
+      "epoch": 2.3537414965986394,
+      "grad_norm": 0.29145191787124425,
+      "learning_rate": 2.3448128164976593e-06,
+      "loss": 0.0212,
       "step": 346
     },
     {
-      "epoch": 2.8559670781893005,
-      "grad_norm": 0.38136282407374994,
-      "learning_rate": 1.148812019483958e-07,
-      "loss": 0.0322,
+      "epoch": 2.360544217687075,
+      "grad_norm": 0.40253909206686106,
+      "learning_rate": 2.297682177502546e-06,
+      "loss": 0.025,
       "step": 347
     },
     {
-      "epoch": 2.8641975308641974,
-      "grad_norm": 0.24965464911826693,
-      "learning_rate": 1.0178558119067316e-07,
-      "loss": 0.0327,
+      "epoch": 2.36734693877551,
+      "grad_norm": 0.3130871435447182,
+      "learning_rate": 2.2509684688200385e-06,
+      "loss": 0.0189,
       "step": 348
     },
     {
-      "epoch": 2.8724279835390947,
-      "grad_norm": 0.3737627173539169,
-      "learning_rate": 8.947839996777286e-08,
-      "loss": 0.0292,
+      "epoch": 2.3741496598639458,
+      "grad_norm": 0.2359349890142643,
+      "learning_rate": 2.204674219086531e-06,
+      "loss": 0.0164,
       "step": 349
     },
     {
-      "epoch": 2.8806584362139915,
-      "grad_norm": 0.28058540051709246,
-      "learning_rate": 7.796063860481595e-08,
-      "loss": 0.035,
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.3417163765611956,
+      "learning_rate": 2.158801934232897e-06,
+      "loss": 0.023,
       "step": 350
     },
     {
-      "epoch": 2.888888888888889,
-      "grad_norm": 0.24352705124392576,
-      "learning_rate": 6.723321454590093e-08,
-      "loss": 0.0319,
+      "epoch": 2.387755102040816,
+      "grad_norm": 0.2911079105732259,
+      "learning_rate": 2.113354097348834e-06,
+      "loss": 0.0222,
       "step": 351
     },
     {
-      "epoch": 2.897119341563786,
-      "grad_norm": 0.3211122649341077,
-      "learning_rate": 5.7296982281026534e-08,
-      "loss": 0.0329,
+      "epoch": 2.3945578231292517,
+      "grad_norm": 0.3515229678204925,
+      "learning_rate": 2.0683331685484655e-06,
+      "loss": 0.0226,
       "step": 352
     },
     {
-      "epoch": 2.905349794238683,
-      "grad_norm": 0.29724093186258926,
-      "learning_rate": 4.815273327803183e-08,
-      "loss": 0.026,
+      "epoch": 2.4013605442176873,
+      "grad_norm": 0.2769437555973887,
+      "learning_rate": 2.0237415848371666e-06,
+      "loss": 0.0131,
       "step": 353
     },
     {
-      "epoch": 2.9135802469135803,
-      "grad_norm": 0.2860118524127369,
-      "learning_rate": 3.980119591954101e-08,
-      "loss": 0.0315,
+      "epoch": 2.4081632653061225,
+      "grad_norm": 0.33922796737817396,
+      "learning_rate": 1.979581759979642e-06,
+      "loss": 0.0241,
       "step": 354
     },
     {
-      "epoch": 2.9218106995884776,
-      "grad_norm": 0.2786243137693698,
-      "learning_rate": 3.224303544495766e-08,
-      "loss": 0.0346,
+      "epoch": 2.4149659863945576,
+      "grad_norm": 0.31493116786046815,
+      "learning_rate": 1.9358560843692787e-06,
+      "loss": 0.0172,
       "step": 355
     },
     {
-      "epoch": 2.9300411522633745,
-      "grad_norm": 0.32949520090124634,
-      "learning_rate": 2.547885389746485e-08,
-      "loss": 0.0324,
+      "epoch": 2.421768707482993,
+      "grad_norm": 0.3073302581281779,
+      "learning_rate": 1.892566924898751e-06,
+      "loss": 0.0216,
       "step": 356
     },
     {
-      "epoch": 2.9382716049382713,
-      "grad_norm": 0.30923281308689177,
-      "learning_rate": 1.9509190076074657e-08,
-      "loss": 0.0296,
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.40858128963320417,
+      "learning_rate": 1.8497166248318876e-06,
+      "loss": 0.0225,
       "step": 357
     },
     {
-      "epoch": 2.9465020576131686,
-      "grad_norm": 0.26184165763062317,
-      "learning_rate": 1.4334519492711362e-08,
-      "loss": 0.0287,
+      "epoch": 2.435374149659864,
+      "grad_norm": 0.38355656039601815,
+      "learning_rate": 1.807307503676846e-06,
+      "loss": 0.0187,
       "step": 358
     },
     {
-      "epoch": 2.954732510288066,
-      "grad_norm": 0.30914713329556287,
-      "learning_rate": 9.955254334328424e-09,
-      "loss": 0.0303,
+      "epoch": 2.442176870748299,
+      "grad_norm": 0.31714048371803,
+      "learning_rate": 1.7653418570605474e-06,
+      "loss": 0.0181,
       "step": 359
     },
     {
-      "epoch": 2.962962962962963,
-      "grad_norm": 0.258895938100784,
-      "learning_rate": 6.371743430082511e-09,
-      "loss": 0.0285,
+      "epoch": 2.4489795918367347,
+      "grad_norm": 0.25496567244495894,
+      "learning_rate": 1.7238219566044145e-06,
+      "loss": 0.0249,
       "step": 360
     },
     {
-      "epoch": 2.97119341563786,
-      "grad_norm": 0.29900176595284766,
-      "learning_rate": 3.5842722235468475e-09,
-      "loss": 0.0347,
+      "epoch": 2.45578231292517,
+      "grad_norm": 0.2810148889000208,
+      "learning_rate": 1.6827500498014026e-06,
+      "loss": 0.0195,
       "step": 361
     },
     {
-      "epoch": 2.9794238683127574,
-      "grad_norm": 0.23314693152924557,
-      "learning_rate": 1.593062749967178e-09,
-      "loss": 0.0273,
+      "epoch": 2.4625850340136055,
+      "grad_norm": 0.34079615823562126,
+      "learning_rate": 1.6421283598943526e-06,
+      "loss": 0.0242,
       "step": 362
     },
     {
-      "epoch": 2.9876543209876543,
-      "grad_norm": 0.2446521510361699,
-      "learning_rate": 3.982736185859093e-10,
-      "loss": 0.0312,
+      "epoch": 2.4693877551020407,
+      "grad_norm": 0.27849042979199606,
+      "learning_rate": 1.601959085755641e-06,
+      "loss": 0.0215,
       "step": 363
     },
     {
-      "epoch": 2.9876543209876543,
-      "step": 363,
-      "total_flos": 28212527521792.0,
-      "train_loss": 0.12437393488080213,
-      "train_runtime": 3566.9961,
-      "train_samples_per_second": 1.63,
-      "train_steps_per_second": 0.102
+      "epoch": 2.4761904761904763,
+      "grad_norm": 0.31547844011000803,
+      "learning_rate": 1.5622444017681438e-06,
+      "loss": 0.0209,
+      "step": 364
+    },
+    {
+      "epoch": 2.4829931972789114,
+      "grad_norm": 0.3280004034375483,
+      "learning_rate": 1.5229864577075548e-06,
+      "loss": 0.0206,
+      "step": 365
+    },
+    {
+      "epoch": 2.489795918367347,
+      "grad_norm": 0.36138147578834934,
+      "learning_rate": 1.4841873786260019e-06,
+      "loss": 0.0206,
+      "step": 366
+    },
+    {
+      "epoch": 2.496598639455782,
+      "grad_norm": 0.34871288825979213,
+      "learning_rate": 1.445849264737026e-06,
+      "loss": 0.0197,
+      "step": 367
+    },
+    {
+      "epoch": 2.503401360544218,
+      "grad_norm": 0.2929843679149074,
+      "learning_rate": 1.4079741913018863e-06,
+      "loss": 0.0191,
+      "step": 368
+    },
+    {
+      "epoch": 2.510204081632653,
+      "grad_norm": 0.3729046126118177,
+      "learning_rate": 1.3705642085172367e-06,
+      "loss": 0.017,
+      "step": 369
+    },
+    {
+      "epoch": 2.5170068027210886,
+      "grad_norm": 0.23294615156735132,
+      "learning_rate": 1.3336213414041387e-06,
+      "loss": 0.0129,
+      "step": 370
+    },
+    {
+      "epoch": 2.5238095238095237,
+      "grad_norm": 0.28575171918313647,
+      "learning_rate": 1.2971475896984475e-06,
+      "loss": 0.02,
+      "step": 371
+    },
+    {
+      "epoch": 2.5306122448979593,
+      "grad_norm": 0.24125146893347207,
+      "learning_rate": 1.2611449277425715e-06,
+      "loss": 0.0143,
+      "step": 372
+    },
+    {
+      "epoch": 2.5374149659863945,
+      "grad_norm": 0.32756177057759744,
+      "learning_rate": 1.2256153043785911e-06,
+      "loss": 0.0162,
+      "step": 373
+    },
+    {
+      "epoch": 2.54421768707483,
+      "grad_norm": 0.3138440305938392,
+      "learning_rate": 1.1905606428427775e-06,
+      "loss": 0.0193,
+      "step": 374
+    },
+    {
+      "epoch": 2.5510204081632653,
+      "grad_norm": 0.3278031624256675,
+      "learning_rate": 1.1559828406614716e-06,
+      "loss": 0.0236,
+      "step": 375
+    },
+    {
+      "epoch": 2.557823129251701,
+      "grad_norm": 0.3496069896696055,
+      "learning_rate": 1.1218837695483853e-06,
+      "loss": 0.017,
+      "step": 376
+    },
+    {
+      "epoch": 2.564625850340136,
+      "grad_norm": 0.35806348572299274,
+      "learning_rate": 1.0882652753032797e-06,
+      "loss": 0.0183,
+      "step": 377
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.3416680123893772,
+      "learning_rate": 1.0551291777120465e-06,
+      "loss": 0.0159,
+      "step": 378
+    },
+    {
+      "epoch": 2.578231292517007,
+      "grad_norm": 0.2848539912430437,
+      "learning_rate": 1.0224772704482033e-06,
+      "loss": 0.021,
+      "step": 379
+    },
+    {
+      "epoch": 2.5850340136054424,
+      "grad_norm": 0.34478335165414964,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.02,
+      "step": 380
+    },
+    {
+      "epoch": 2.5918367346938775,
+      "grad_norm": 0.30259868082341945,
+      "learning_rate": 9.58633070453785e-07,
+      "loss": 0.0224,
+      "step": 381
+    },
+    {
+      "epoch": 2.5986394557823127,
+      "grad_norm": 0.29749976015408053,
+      "learning_rate": 9.274442336416567e-07,
+      "loss": 0.0226,
+      "step": 382
+    },
+    {
+      "epoch": 2.6054421768707483,
+      "grad_norm": 0.3195969770154066,
+      "learning_rate": 8.967464988067476e-07,
+      "loss": 0.0168,
+      "step": 383
+    },
+    {
+      "epoch": 2.612244897959184,
+      "grad_norm": 0.2600960483615313,
+      "learning_rate": 8.665415276327871e-07,
+      "loss": 0.0137,
+      "step": 384
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.26403678914149437,
+      "learning_rate": 8.368309551299536e-07,
+      "loss": 0.0178,
+      "step": 385
+    },
+    {
+      "epoch": 2.6258503401360542,
+      "grad_norm": 0.2748498822235369,
+      "learning_rate": 8.076163895463862e-07,
+      "loss": 0.0144,
+      "step": 386
+    },
+    {
+      "epoch": 2.63265306122449,
+      "grad_norm": 0.3305010052406547,
+      "learning_rate": 7.788994122811178e-07,
+      "loss": 0.0219,
+      "step": 387
+    },
+    {
+      "epoch": 2.6394557823129254,
+      "grad_norm": 0.246882891291933,
+      "learning_rate": 7.506815777984788e-07,
+      "loss": 0.0158,
+      "step": 388
+    },
+    {
+      "epoch": 2.6462585034013606,
+      "grad_norm": 0.25439177703144955,
+      "learning_rate": 7.229644135439473e-07,
+      "loss": 0.0148,
+      "step": 389
+    },
+    {
+      "epoch": 2.6530612244897958,
+      "grad_norm": 0.36357656004842953,
+      "learning_rate": 6.957494198614778e-07,
+      "loss": 0.0238,
+      "step": 390
+    },
+    {
+      "epoch": 2.6598639455782314,
+      "grad_norm": 0.2872101510976723,
+      "learning_rate": 6.690380699122767e-07,
+      "loss": 0.0204,
+      "step": 391
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.24331315521025246,
+      "learning_rate": 6.428318095950648e-07,
+      "loss": 0.0155,
+      "step": 392
+    },
+    {
+      "epoch": 2.673469387755102,
+      "grad_norm": 0.25145814555745893,
+      "learning_rate": 6.171320574678064e-07,
+      "loss": 0.0175,
+      "step": 393
+    },
+    {
+      "epoch": 2.6802721088435373,
+      "grad_norm": 0.3359588828649143,
+      "learning_rate": 5.919402046709288e-07,
+      "loss": 0.02,
+      "step": 394
+    },
+    {
+      "epoch": 2.687074829931973,
+      "grad_norm": 0.29084282580528553,
+      "learning_rate": 5.672576148520136e-07,
+      "loss": 0.0183,
+      "step": 395
+    },
+    {
+      "epoch": 2.693877551020408,
+      "grad_norm": 0.344918984609033,
+      "learning_rate": 5.430856240919779e-07,
+      "loss": 0.0235,
+      "step": 396
+    },
+    {
+      "epoch": 2.7006802721088436,
+      "grad_norm": 0.2711450817646753,
+      "learning_rate": 5.19425540832762e-07,
+      "loss": 0.016,
+      "step": 397
+    },
+    {
+      "epoch": 2.707482993197279,
+      "grad_norm": 0.40375159495643054,
+      "learning_rate": 4.962786458064972e-07,
+      "loss": 0.0228,
+      "step": 398
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.24410300959704656,
+      "learning_rate": 4.73646191966175e-07,
+      "loss": 0.0169,
+      "step": 399
+    },
+    {
+      "epoch": 2.7210884353741496,
+      "grad_norm": 0.3592873823963376,
+      "learning_rate": 4.515294044178331e-07,
+      "loss": 0.0192,
+      "step": 400
+    },
+    {
+      "epoch": 2.727891156462585,
+      "grad_norm": 0.47656899007676473,
+      "learning_rate": 4.299294803542331e-07,
+      "loss": 0.0165,
+      "step": 401
+    },
+    {
+      "epoch": 2.7346938775510203,
+      "grad_norm": 0.3541427597027374,
+      "learning_rate": 4.0884758899006007e-07,
+      "loss": 0.0185,
+      "step": 402
+    },
+    {
+      "epoch": 2.741496598639456,
+      "grad_norm": 0.4056943661965022,
+      "learning_rate": 3.882848714986243e-07,
+      "loss": 0.0194,
+      "step": 403
+    },
+    {
+      "epoch": 2.748299319727891,
+      "grad_norm": 0.32241558571045703,
+      "learning_rate": 3.6824244095010064e-07,
+      "loss": 0.0248,
+      "step": 404
+    },
+    {
+      "epoch": 2.7551020408163263,
+      "grad_norm": 0.30581795904421494,
+      "learning_rate": 3.4872138225127137e-07,
+      "loss": 0.016,
+      "step": 405
+    },
+    {
+      "epoch": 2.761904761904762,
+      "grad_norm": 0.34578165087246426,
+      "learning_rate": 3.2972275208679625e-07,
+      "loss": 0.0213,
+      "step": 406
+    },
+    {
+      "epoch": 2.7687074829931975,
+      "grad_norm": 0.3136844616082194,
+      "learning_rate": 3.112475788620217e-07,
+      "loss": 0.0186,
+      "step": 407
+    },
+    {
+      "epoch": 2.7755102040816326,
+      "grad_norm": 0.3330644974135503,
+      "learning_rate": 2.932968626473065e-07,
+      "loss": 0.0214,
+      "step": 408
+    },
+    {
+      "epoch": 2.782312925170068,
+      "grad_norm": 0.2734645865898147,
+      "learning_rate": 2.758715751238872e-07,
+      "loss": 0.0167,
+      "step": 409
+    },
+    {
+      "epoch": 2.7891156462585034,
+      "grad_norm": 0.3235976546346094,
+      "learning_rate": 2.589726595312858e-07,
+      "loss": 0.0205,
+      "step": 410
+    },
+    {
+      "epoch": 2.795918367346939,
+      "grad_norm": 0.3446098604740126,
+      "learning_rate": 2.426010306162485e-07,
+      "loss": 0.0169,
+      "step": 411
+    },
+    {
+      "epoch": 2.802721088435374,
+      "grad_norm": 0.3270040316033607,
+      "learning_rate": 2.2675757458323066e-07,
+      "loss": 0.0184,
+      "step": 412
+    },
+    {
+      "epoch": 2.8095238095238093,
+      "grad_norm": 0.2879016695364431,
+      "learning_rate": 2.1144314904642194e-07,
+      "loss": 0.0177,
+      "step": 413
+    },
+    {
+      "epoch": 2.816326530612245,
+      "grad_norm": 0.33267463209181264,
+      "learning_rate": 1.9665858298333006e-07,
+      "loss": 0.0203,
+      "step": 414
+    },
+    {
+      "epoch": 2.8231292517006805,
+      "grad_norm": 0.23836799115278284,
+      "learning_rate": 1.824046766899046e-07,
+      "loss": 0.0144,
+      "step": 415
+    },
+    {
+      "epoch": 2.8299319727891157,
+      "grad_norm": 0.35726854776874595,
+      "learning_rate": 1.6868220173721472e-07,
+      "loss": 0.0239,
+      "step": 416
+    },
+    {
+      "epoch": 2.836734693877551,
+      "grad_norm": 0.3457061306540936,
+      "learning_rate": 1.5549190092968736e-07,
+      "loss": 0.0229,
+      "step": 417
+    },
+    {
+      "epoch": 2.8435374149659864,
+      "grad_norm": 0.30955214004843584,
+      "learning_rate": 1.4283448826489798e-07,
+      "loss": 0.0211,
+      "step": 418
+    },
+    {
+      "epoch": 2.8503401360544216,
+      "grad_norm": 0.3263601366659666,
+      "learning_rate": 1.3071064889491723e-07,
+      "loss": 0.02,
+      "step": 419
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.32256000869883056,
+      "learning_rate": 1.1912103908922945e-07,
+      "loss": 0.0196,
+      "step": 420
+    },
+    {
+      "epoch": 2.8639455782312924,
+      "grad_norm": 0.28683991197300734,
+      "learning_rate": 1.0806628619920322e-07,
+      "loss": 0.0163,
+      "step": 421
+    },
+    {
+      "epoch": 2.870748299319728,
+      "grad_norm": 0.2903311126798398,
+      "learning_rate": 9.754698862413758e-08,
+      "loss": 0.0202,
+      "step": 422
+    },
+    {
+      "epoch": 2.877551020408163,
+      "grad_norm": 0.2239071167217727,
+      "learning_rate": 8.756371577886891e-08,
+      "loss": 0.0134,
+      "step": 423
+    },
+    {
+      "epoch": 2.8843537414965987,
+      "grad_norm": 0.29808971100347104,
+      "learning_rate": 7.81170080629412e-08,
+      "loss": 0.0204,
+      "step": 424
+    },
+    {
+      "epoch": 2.891156462585034,
+      "grad_norm": 0.2789565484019186,
+      "learning_rate": 6.920737683136614e-08,
+      "loss": 0.0187,
+      "step": 425
+    },
+    {
+      "epoch": 2.8979591836734695,
+      "grad_norm": 0.3158830313532607,
+      "learning_rate": 6.083530436693408e-08,
+      "loss": 0.0213,
+      "step": 426
+    },
+    {
+      "epoch": 2.9047619047619047,
+      "grad_norm": 0.2652235833848647,
+      "learning_rate": 5.300124385410943e-08,
+      "loss": 0.0184,
+      "step": 427
+    },
+    {
+      "epoch": 2.9115646258503403,
+      "grad_norm": 0.303708914505213,
+      "learning_rate": 4.570561935450468e-08,
+      "loss": 0.0228,
+      "step": 428
+    },
+    {
+      "epoch": 2.9183673469387754,
+      "grad_norm": 0.2552716618292082,
+      "learning_rate": 3.894882578391879e-08,
+      "loss": 0.0168,
+      "step": 429
+    },
+    {
+      "epoch": 2.925170068027211,
+      "grad_norm": 0.3437542248362107,
+      "learning_rate": 3.273122889096536e-08,
+      "loss": 0.018,
+      "step": 430
+    },
+    {
+      "epoch": 2.931972789115646,
+      "grad_norm": 0.41181823304903403,
+      "learning_rate": 2.705316523726853e-08,
+      "loss": 0.0175,
+      "step": 431
+    },
+    {
+      "epoch": 2.938775510204082,
+      "grad_norm": 0.31165983112820045,
+      "learning_rate": 2.1914942179253052e-08,
+      "loss": 0.02,
+      "step": 432
+    },
+    {
+      "epoch": 2.945578231292517,
+      "grad_norm": 0.37990351221219176,
+      "learning_rate": 1.7316837851499845e-08,
+      "loss": 0.0208,
+      "step": 433
+    },
+    {
+      "epoch": 2.9523809523809526,
+      "grad_norm": 0.42236543353263456,
+      "learning_rate": 1.325910115169471e-08,
+      "loss": 0.0211,
+      "step": 434
+    },
+    {
+      "epoch": 2.9591836734693877,
+      "grad_norm": 0.3667740733661891,
+      "learning_rate": 9.74195172715242e-09,
+      "loss": 0.0185,
+      "step": 435
+    },
+    {
+      "epoch": 2.965986394557823,
+      "grad_norm": 0.2491767102688122,
+      "learning_rate": 6.7655799629284815e-09,
+      "loss": 0.0216,
+      "step": 436
+    },
+    {
+      "epoch": 2.9727891156462585,
+      "grad_norm": 0.27827297191105266,
+      "learning_rate": 4.330146971515126e-09,
+      "loss": 0.0135,
+      "step": 437
+    },
+    {
+      "epoch": 2.979591836734694,
+      "grad_norm": 0.2479372965359603,
+      "learning_rate": 2.435784584114975e-09,
+      "loss": 0.0197,
+      "step": 438
+    },
+    {
+      "epoch": 2.9863945578231292,
+      "grad_norm": 0.3342210953903805,
+      "learning_rate": 1.0825953435122938e-09,
+      "loss": 0.0224,
+      "step": 439
+    },
+    {
+      "epoch": 2.9931972789115644,
+      "grad_norm": 0.3251029484247515,
+      "learning_rate": 2.706524985174319e-10,
+      "loss": 0.0255,
+      "step": 440
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.30791966633936196,
+      "learning_rate": 0.0,
+      "loss": 0.0177,
+      "step": 441
+    },
+    {
+      "epoch": 3.0,
+      "step": 441,
+      "total_flos": 34413652910080.0,
+      "train_loss": 0.10652120225131512,
+      "train_runtime": 8437.5347,
+      "train_samples_per_second": 0.836,
+      "train_steps_per_second": 0.052
     }
   ],
   "logging_steps": 1.0,
-  "max_steps": 363,
+  "max_steps": 441,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 50000,
-  "total_flos": 28212527521792.0,
+  "total_flos": 34413652910080.0,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null