diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,8670 +1,8635 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 5.0,
+  "epoch": 4.982278481012658,
   "eval_steps": 500,
-  "global_step": 1235,
+  "global_step": 1230,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.004048582995951417,
-      "grad_norm": 6.38890886607983,
-      "learning_rate": 3.2258064516129035e-07,
-      "loss": 0.8187,
+      "epoch": 0.004050632911392405,
+      "grad_norm": 6.321949967627026,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.8141,
       "step": 1
     },
     {
-      "epoch": 0.008097165991902834,
-      "grad_norm": 6.747916136480842,
-      "learning_rate": 6.451612903225807e-07,
-      "loss": 0.8621,
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.8199983661094405,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.8713,
       "step": 2
     },
     {
-      "epoch": 0.012145748987854251,
-      "grad_norm": 6.577493325606705,
-      "learning_rate": 9.67741935483871e-07,
-      "loss": 0.8328,
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.488320756546987,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.8253,
       "step": 3
     },
     {
-      "epoch": 0.016194331983805668,
-      "grad_norm": 6.295297987406139,
-      "learning_rate": 1.2903225806451614e-06,
-      "loss": 0.821,
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.14756306041393,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.8193,
       "step": 4
     },
     {
-      "epoch": 0.020242914979757085,
-      "grad_norm": 5.94247520710286,
-      "learning_rate": 1.6129032258064516e-06,
-      "loss": 0.7985,
+      "epoch": 0.020253164556962026,
+      "grad_norm": 5.744390117853706,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.7855,
       "step": 5
     },
     {
-      "epoch": 0.024291497975708502,
-      "grad_norm": 6.004472754314662,
-      "learning_rate": 1.935483870967742e-06,
-      "loss": 0.7959,
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.837831874637069,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.7919,
       "step": 6
     },
     {
-      "epoch": 0.02834008097165992,
-      "grad_norm": 4.819101412428404,
-      "learning_rate": 2.2580645161290324e-06,
-      "loss": 0.769,
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.760133383421948,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.7711,
       "step": 7
     },
     {
-      "epoch": 0.032388663967611336,
-      "grad_norm": 4.621120859113136,
-      "learning_rate": 2.580645161290323e-06,
-      "loss": 0.7746,
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.526236582456803,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.7808,
       "step": 8
     },
     {
-      "epoch": 0.03643724696356275,
-      "grad_norm": 2.652777888633462,
-      "learning_rate": 2.903225806451613e-06,
-      "loss": 0.7304,
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.503762766767712,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.729,
       "step": 9
     },
     {
-      "epoch": 0.04048582995951417,
-      "grad_norm": 2.277879871767528,
-      "learning_rate": 3.225806451612903e-06,
-      "loss": 0.7158,
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.201334010378042,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.7223,
       "step": 10
     },
     {
-      "epoch": 0.044534412955465584,
-      "grad_norm": 2.189714768329133,
-      "learning_rate": 3.548387096774194e-06,
-      "loss": 0.7241,
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.1499273953386484,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.7202,
       "step": 11
     },
     {
-      "epoch": 0.048582995951417005,
-      "grad_norm": 3.6058116417275223,
-      "learning_rate": 3.870967741935484e-06,
-      "loss": 0.684,
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.639093383403756,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.6952,
       "step": 12
     },
     {
-      "epoch": 0.05263157894736842,
-      "grad_norm": 4.029888525816398,
-      "learning_rate": 4.193548387096774e-06,
-      "loss": 0.7039,
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.9852149217528727,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.698,
       "step": 13
     },
     {
-      "epoch": 0.05668016194331984,
-      "grad_norm": 4.012832182382164,
-      "learning_rate": 4.516129032258065e-06,
-      "loss": 0.6808,
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.15534916687844,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.6919,
       "step": 14
     },
     {
-      "epoch": 0.06072874493927125,
-      "grad_norm": 3.65801396221574,
-      "learning_rate": 4.838709677419355e-06,
-      "loss": 0.6486,
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.754246201173943,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.655,
       "step": 15
     },
     {
-      "epoch": 0.06477732793522267,
-      "grad_norm": 2.9962213981537884,
-      "learning_rate": 5.161290322580646e-06,
-      "loss": 0.6303,
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.161136109475919,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.6466,
       "step": 16
     },
     {
-      "epoch": 0.06882591093117409,
-      "grad_norm": 2.92971903685758,
-      "learning_rate": 5.483870967741935e-06,
-      "loss": 0.6104,
+      "epoch": 0.06886075949367089,
+      "grad_norm": 3.1286336894242113,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.6278,
       "step": 17
     },
     {
-      "epoch": 0.0728744939271255,
-      "grad_norm": 2.5136951150556346,
-      "learning_rate": 5.806451612903226e-06,
-      "loss": 0.6246,
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.577346617949884,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.6306,
       "step": 18
     },
     {
-      "epoch": 0.07692307692307693,
-      "grad_norm": 1.8404982812017376,
-      "learning_rate": 6.129032258064517e-06,
-      "loss": 0.6022,
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.8025143025619421,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.6218,
       "step": 19
     },
     {
-      "epoch": 0.08097165991902834,
-      "grad_norm": 1.205795436529844,
-      "learning_rate": 6.451612903225806e-06,
-      "loss": 0.563,
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.2577613059384753,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.5791,
       "step": 20
     },
     {
-      "epoch": 0.08502024291497975,
-      "grad_norm": 1.5067251674522315,
-      "learning_rate": 6.774193548387097e-06,
-      "loss": 0.5861,
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.6181155189034264,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.6001,
       "step": 21
     },
     {
-      "epoch": 0.08906882591093117,
-      "grad_norm": 1.6445382357999714,
-      "learning_rate": 7.096774193548388e-06,
-      "loss": 0.5537,
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.793784411815894,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.567,
       "step": 22
     },
     {
-      "epoch": 0.0931174089068826,
-      "grad_norm": 1.4879587502212996,
-      "learning_rate": 7.4193548387096784e-06,
-      "loss": 0.5703,
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.6907198762228053,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.5983,
       "step": 23
     },
     {
-      "epoch": 0.09716599190283401,
-      "grad_norm": 1.0008004196105986,
-      "learning_rate": 7.741935483870968e-06,
-      "loss": 0.5433,
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.1032930466738173,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.5616,
       "step": 24
     },
     {
-      "epoch": 0.10121457489878542,
-      "grad_norm": 0.82674390687085,
-      "learning_rate": 8.064516129032258e-06,
-      "loss": 0.55,
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.8523637386284916,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.5588,
       "step": 25
     },
     {
-      "epoch": 0.10526315789473684,
-      "grad_norm": 1.0469832578906026,
-      "learning_rate": 8.387096774193549e-06,
-      "loss": 0.5366,
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.0833824522356301,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.5509,
       "step": 26
     },
     {
-      "epoch": 0.10931174089068826,
-      "grad_norm": 0.9539233719762708,
-      "learning_rate": 8.70967741935484e-06,
-      "loss": 0.5359,
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.9677882990698344,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.5417,
       "step": 27
     },
     {
-      "epoch": 0.11336032388663968,
-      "grad_norm": 0.7920469523540589,
-      "learning_rate": 9.03225806451613e-06,
-      "loss": 0.5614,
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.7657210985969838,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.5686,
       "step": 28
     },
     {
-      "epoch": 0.11740890688259109,
-      "grad_norm": 0.6785348639118282,
-      "learning_rate": 9.35483870967742e-06,
-      "loss": 0.5162,
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.7555845734088903,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.531,
       "step": 29
     },
     {
-      "epoch": 0.1214574898785425,
-      "grad_norm": 0.834588693390409,
-      "learning_rate": 9.67741935483871e-06,
-      "loss": 0.5104,
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8580210052381343,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.5224,
       "step": 30
     },
     {
-      "epoch": 0.12550607287449392,
-      "grad_norm": 0.7312720167742082,
-      "learning_rate": 1e-05,
-      "loss": 0.5401,
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.7472448625043672,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.5535,
       "step": 31
     },
     {
-      "epoch": 0.12955465587044535,
-      "grad_norm": 0.6652107883050867,
-      "learning_rate": 1.0322580645161291e-05,
-      "loss": 0.5215,
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.670670324834762,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.5337,
       "step": 32
     },
     {
-      "epoch": 0.13360323886639677,
-      "grad_norm": 0.6681069393839387,
-      "learning_rate": 1.0645161290322582e-05,
-      "loss": 0.5199,
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7915524694680136,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.5295,
       "step": 33
     },
     {
-      "epoch": 0.13765182186234817,
-      "grad_norm": 0.7609254975516906,
-      "learning_rate": 1.096774193548387e-05,
-      "loss": 0.5293,
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.7420731305225214,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.5458,
       "step": 34
     },
     {
-      "epoch": 0.1417004048582996,
-      "grad_norm": 0.5664771464524446,
-      "learning_rate": 1.1290322580645164e-05,
-      "loss": 0.5116,
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.5569955812943824,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.5211,
       "step": 35
     },
     {
-      "epoch": 0.145748987854251,
-      "grad_norm": 0.7436883043302881,
-      "learning_rate": 1.1612903225806453e-05,
-      "loss": 0.5079,
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.8909023535136663,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.5303,
       "step": 36
     },
     {
-      "epoch": 0.14979757085020243,
-      "grad_norm": 0.6820295633223774,
-      "learning_rate": 1.1935483870967743e-05,
-      "loss": 0.4957,
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.6756952861926832,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.5057,
       "step": 37
     },
     {
-      "epoch": 0.15384615384615385,
-      "grad_norm": 0.5659936603206349,
-      "learning_rate": 1.2258064516129034e-05,
-      "loss": 0.5154,
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6534758768977116,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.5234,
       "step": 38
     },
     {
-      "epoch": 0.15789473684210525,
-      "grad_norm": 0.5548809675147656,
-      "learning_rate": 1.2580645161290324e-05,
-      "loss": 0.506,
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.6540161113459456,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.52,
       "step": 39
     },
     {
-      "epoch": 0.16194331983805668,
-      "grad_norm": 0.5449546062984876,
-      "learning_rate": 1.2903225806451613e-05,
-      "loss": 0.4859,
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6080189331237126,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.4928,
       "step": 40
     },
     {
-      "epoch": 0.1659919028340081,
-      "grad_norm": 0.4749100725948443,
-      "learning_rate": 1.3225806451612903e-05,
-      "loss": 0.49,
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.5695849638215544,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5006,
       "step": 41
     },
     {
-      "epoch": 0.1700404858299595,
-      "grad_norm": 0.5185207794058557,
-      "learning_rate": 1.3548387096774194e-05,
-      "loss": 0.5054,
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.5598796639427857,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.5256,
       "step": 42
     },
     {
-      "epoch": 0.17408906882591094,
-      "grad_norm": 0.5312734151511638,
-      "learning_rate": 1.3870967741935486e-05,
-      "loss": 0.4629,
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.5260667616215371,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.4807,
       "step": 43
     },
     {
-      "epoch": 0.17813765182186234,
-      "grad_norm": 0.47477240887074174,
-      "learning_rate": 1.4193548387096776e-05,
-      "loss": 0.4778,
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.5277664304653565,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.4887,
       "step": 44
     },
     {
-      "epoch": 0.18218623481781376,
-      "grad_norm": 0.6727519518960435,
-      "learning_rate": 1.4516129032258066e-05,
-      "loss": 0.4827,
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7715052576716896,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.4925,
       "step": 45
     },
     {
-      "epoch": 0.1862348178137652,
-      "grad_norm": 0.5956370457417351,
-      "learning_rate": 1.4838709677419357e-05,
-      "loss": 0.4946,
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.5327793458049991,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.5137,
       "step": 46
     },
     {
-      "epoch": 0.1902834008097166,
-      "grad_norm": 0.6007443675691952,
-      "learning_rate": 1.5161290322580646e-05,
-      "loss": 0.4769,
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.8455866790570215,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.4967,
       "step": 47
     },
     {
-      "epoch": 0.19433198380566802,
-      "grad_norm": 0.5533554953629908,
-      "learning_rate": 1.5483870967741936e-05,
-      "loss": 0.4749,
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.5246032628806733,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.4827,
       "step": 48
     },
     {
-      "epoch": 0.19838056680161945,
-      "grad_norm": 0.6992242170010889,
-      "learning_rate": 1.5806451612903226e-05,
-      "loss": 0.4766,
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7017859091084974,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.485,
       "step": 49
     },
     {
-      "epoch": 0.20242914979757085,
-      "grad_norm": 0.463101526647117,
-      "learning_rate": 1.6129032258064517e-05,
-      "loss": 0.4683,
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.5419608444288071,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.4728,
       "step": 50
     },
     {
-      "epoch": 0.20647773279352227,
-      "grad_norm": 0.7642209207023005,
-      "learning_rate": 1.6451612903225807e-05,
-      "loss": 0.4705,
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7747289344509529,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.4805,
       "step": 51
     },
     {
-      "epoch": 0.21052631578947367,
-      "grad_norm": 0.5117243887010392,
-      "learning_rate": 1.6774193548387098e-05,
-      "loss": 0.4776,
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.5766170401965557,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.4886,
       "step": 52
     },
     {
-      "epoch": 0.2145748987854251,
-      "grad_norm": 0.660005677938164,
-      "learning_rate": 1.7096774193548388e-05,
-      "loss": 0.4842,
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.8567303949607963,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.4944,
       "step": 53
     },
     {
-      "epoch": 0.21862348178137653,
-      "grad_norm": 0.6592883430389442,
-      "learning_rate": 1.741935483870968e-05,
-      "loss": 0.489,
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.7924905871444238,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.5082,
       "step": 54
     },
     {
-      "epoch": 0.22267206477732793,
-      "grad_norm": 0.5088137692320291,
-      "learning_rate": 1.774193548387097e-05,
-      "loss": 0.471,
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6239719432421978,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.4789,
       "step": 55
     },
     {
-      "epoch": 0.22672064777327935,
-      "grad_norm": 0.6369432179232204,
-      "learning_rate": 1.806451612903226e-05,
-      "loss": 0.4874,
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7421717186887025,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.5003,
       "step": 56
     },
     {
-      "epoch": 0.23076923076923078,
-      "grad_norm": 0.48823786674077785,
-      "learning_rate": 1.838709677419355e-05,
-      "loss": 0.4737,
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.5281281080449155,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.4789,
       "step": 57
     },
     {
-      "epoch": 0.23481781376518218,
-      "grad_norm": 0.6712235681387004,
-      "learning_rate": 1.870967741935484e-05,
-      "loss": 0.469,
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.6729188603217315,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.4749,
       "step": 58
     },
     {
-      "epoch": 0.2388663967611336,
-      "grad_norm": 0.5504035599702746,
-      "learning_rate": 1.903225806451613e-05,
-      "loss": 0.4544,
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.6523756438492239,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.4581,
       "step": 59
     },
     {
-      "epoch": 0.242914979757085,
-      "grad_norm": 0.6580021256503613,
-      "learning_rate": 1.935483870967742e-05,
-      "loss": 0.4766,
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.6897005546219833,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.4909,
       "step": 60
     },
     {
-      "epoch": 0.24696356275303644,
-      "grad_norm": 0.5999402436531919,
-      "learning_rate": 1.967741935483871e-05,
-      "loss": 0.4459,
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.631717057092728,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.4586,
       "step": 61
     },
     {
-      "epoch": 0.25101214574898784,
-      "grad_norm": 0.671509109439001,
-      "learning_rate": 2e-05,
-      "loss": 0.4606,
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.8011112836439717,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.4645,
       "step": 62
     },
     {
-      "epoch": 0.2550607287449393,
-      "grad_norm": 0.5697580019542885,
-      "learning_rate": 2.0322580645161292e-05,
-      "loss": 0.4485,
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.5869506410782408,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.4478,
       "step": 63
     },
     {
-      "epoch": 0.2591093117408907,
-      "grad_norm": 0.5716149390697071,
-      "learning_rate": 2.0645161290322582e-05,
-      "loss": 0.4682,
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.6945130585336375,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.4766,
       "step": 64
     },
     {
-      "epoch": 0.2631578947368421,
-      "grad_norm": 0.6691139146511995,
-      "learning_rate": 2.0967741935483873e-05,
-      "loss": 0.4726,
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.7645702933354891,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.4843,
       "step": 65
     },
     {
-      "epoch": 0.26720647773279355,
-      "grad_norm": 0.566153479082296,
-      "learning_rate": 2.1290322580645163e-05,
-      "loss": 0.4591,
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.5971978894332436,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.4756,
       "step": 66
     },
     {
-      "epoch": 0.27125506072874495,
-      "grad_norm": 0.7005553670930837,
-      "learning_rate": 2.161290322580645e-05,
-      "loss": 0.463,
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.8121918029670555,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.4737,
       "step": 67
     },
     {
-      "epoch": 0.27530364372469635,
-      "grad_norm": 0.4774800646588968,
-      "learning_rate": 2.193548387096774e-05,
-      "loss": 0.4395,
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.6353249105066672,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.4556,
       "step": 68
     },
     {
-      "epoch": 0.2793522267206478,
-      "grad_norm": 0.6036476148488584,
-      "learning_rate": 2.225806451612903e-05,
-      "loss": 0.438,
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.7443156760151339,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.4515,
       "step": 69
     },
     {
-      "epoch": 0.2834008097165992,
-      "grad_norm": 0.6302033629223501,
-      "learning_rate": 2.2580645161290328e-05,
-      "loss": 0.4789,
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.6905504581294317,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.4816,
       "step": 70
     },
     {
-      "epoch": 0.2874493927125506,
-      "grad_norm": 0.5918100248627136,
-      "learning_rate": 2.2903225806451615e-05,
-      "loss": 0.4545,
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.7079802065086289,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.4681,
       "step": 71
     },
     {
-      "epoch": 0.291497975708502,
-      "grad_norm": 0.6671301157774234,
-      "learning_rate": 2.3225806451612906e-05,
-      "loss": 0.4793,
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.9216652823365556,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.4833,
       "step": 72
     },
     {
-      "epoch": 0.29554655870445345,
-      "grad_norm": 0.7234656926781394,
-      "learning_rate": 2.3548387096774196e-05,
-      "loss": 0.4723,
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.8691664428250214,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.4836,
       "step": 73
     },
     {
-      "epoch": 0.29959514170040485,
-      "grad_norm": 0.9517135086072501,
-      "learning_rate": 2.3870967741935486e-05,
-      "loss": 0.4514,
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.9202761372432186,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.4622,
       "step": 74
     },
     {
-      "epoch": 0.30364372469635625,
-      "grad_norm": 0.6751629869317226,
-      "learning_rate": 2.4193548387096777e-05,
-      "loss": 0.4505,
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.0776468919217712,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.4545,
       "step": 75
     },
     {
-      "epoch": 0.3076923076923077,
-      "grad_norm": 0.7506801828588169,
-      "learning_rate": 2.4516129032258067e-05,
-      "loss": 0.4531,
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.952394253984643,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.4639,
       "step": 76
     },
     {
-      "epoch": 0.3117408906882591,
-      "grad_norm": 0.6049402123753981,
-      "learning_rate": 2.4838709677419358e-05,
-      "loss": 0.4711,
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.9983502506500074,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.4729,
       "step": 77
     },
     {
-      "epoch": 0.3157894736842105,
-      "grad_norm": 0.6975197600840494,
-      "learning_rate": 2.5161290322580648e-05,
-      "loss": 0.4424,
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.0077484068793887,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.4608,
       "step": 78
     },
     {
-      "epoch": 0.31983805668016196,
-      "grad_norm": 0.7229324845700684,
-      "learning_rate": 2.548387096774194e-05,
-      "loss": 0.473,
+      "epoch": 0.32,
+      "grad_norm": 0.8174019841834125,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.4875,
       "step": 79
     },
     {
-      "epoch": 0.32388663967611336,
-      "grad_norm": 0.7643323264375167,
-      "learning_rate": 2.5806451612903226e-05,
-      "loss": 0.4578,
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.9950055496459871,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.4627,
       "step": 80
     },
     {
-      "epoch": 0.32793522267206476,
-      "grad_norm": 0.5646716115082011,
-      "learning_rate": 2.6129032258064516e-05,
-      "loss": 0.4546,
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.2612628579013692,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.4561,
       "step": 81
     },
     {
-      "epoch": 0.3319838056680162,
-      "grad_norm": 0.7145225608882887,
-      "learning_rate": 2.6451612903225806e-05,
-      "loss": 0.4458,
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.9165206986847768,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4622,
       "step": 82
     },
     {
-      "epoch": 0.3360323886639676,
-      "grad_norm": 0.6447957836917604,
-      "learning_rate": 2.6774193548387097e-05,
-      "loss": 0.4484,
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.941538407664318,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.4574,
       "step": 83
     },
     {
-      "epoch": 0.340080971659919,
-      "grad_norm": 0.722451796771732,
-      "learning_rate": 2.7096774193548387e-05,
-      "loss": 0.4541,
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.2694698822800556,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.4579,
       "step": 84
     },
     {
-      "epoch": 0.3441295546558704,
-      "grad_norm": 0.593893103951638,
-      "learning_rate": 2.7419354838709678e-05,
-      "loss": 0.4539,
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.6569643892669981,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.4618,
       "step": 85
     },
     {
-      "epoch": 0.3481781376518219,
-      "grad_norm": 0.6866408232755394,
-      "learning_rate": 2.774193548387097e-05,
-      "loss": 0.4721,
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.2796113052821287,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.4756,
       "step": 86
     },
     {
-      "epoch": 0.3522267206477733,
-      "grad_norm": 0.5847031205631198,
-      "learning_rate": 2.8064516129032262e-05,
-      "loss": 0.4605,
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.7672013060339831,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.4633,
       "step": 87
     },
     {
-      "epoch": 0.3562753036437247,
-      "grad_norm": 0.7656256333858148,
-      "learning_rate": 2.8387096774193552e-05,
-      "loss": 0.4604,
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.9450614178585994,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.4732,
       "step": 88
     },
     {
-      "epoch": 0.3603238866396761,
-      "grad_norm": 0.6723781516055587,
-      "learning_rate": 2.8709677419354843e-05,
-      "loss": 0.4582,
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.1054754152296005,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.4655,
       "step": 89
     },
     {
-      "epoch": 0.3643724696356275,
-      "grad_norm": 0.6753824035653645,
-      "learning_rate": 2.9032258064516133e-05,
-      "loss": 0.4367,
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.055952282424391,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.4446,
       "step": 90
     },
     {
-      "epoch": 0.3684210526315789,
-      "grad_norm": 0.5617709371960506,
-      "learning_rate": 2.9354838709677423e-05,
-      "loss": 0.4194,
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.9338167690491658,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.434,
       "step": 91
     },
     {
-      "epoch": 0.3724696356275304,
-      "grad_norm": 0.6555286275743986,
-      "learning_rate": 2.9677419354838714e-05,
-      "loss": 0.4556,
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.0322866760209968,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.4576,
       "step": 92
     },
     {
-      "epoch": 0.3765182186234818,
-      "grad_norm": 0.557787513554509,
-      "learning_rate": 3.0000000000000004e-05,
-      "loss": 0.4563,
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.003970762602946,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.4687,
       "step": 93
     },
     {
-      "epoch": 0.3805668016194332,
-      "grad_norm": 0.6903096004489547,
-      "learning_rate": 3.032258064516129e-05,
-      "loss": 0.461,
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.065734570184499,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.4713,
       "step": 94
     },
     {
-      "epoch": 0.38461538461538464,
-      "grad_norm": 0.5672596311672535,
-      "learning_rate": 3.0645161290322585e-05,
-      "loss": 0.4183,
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.777249002943878,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.425,
       "step": 95
     },
     {
-      "epoch": 0.38866396761133604,
-      "grad_norm": 0.6791454441027609,
-      "learning_rate": 3.096774193548387e-05,
-      "loss": 0.4523,
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.0062688351790485,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.4663,
       "step": 96
     },
     {
-      "epoch": 0.39271255060728744,
-      "grad_norm": 0.5235799242931867,
-      "learning_rate": 3.1290322580645166e-05,
-      "loss": 0.4586,
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.8269409863453074,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.4639,
       "step": 97
     },
     {
-      "epoch": 0.3967611336032389,
-      "grad_norm": 0.5734482765290675,
-      "learning_rate": 3.161290322580645e-05,
-      "loss": 0.4228,
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.9625596710107783,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.4345,
       "step": 98
     },
     {
-      "epoch": 0.4008097165991903,
-      "grad_norm": 0.5754614361810384,
-      "learning_rate": 3.193548387096774e-05,
-      "loss": 0.4428,
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9444181913846895,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.4536,
       "step": 99
     },
     {
-      "epoch": 0.4048582995951417,
-      "grad_norm": 0.6404891394410366,
-      "learning_rate": 3.2258064516129034e-05,
-      "loss": 0.4432,
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.1264072732182475,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.4467,
       "step": 100
     },
     {
-      "epoch": 0.4089068825910931,
-      "grad_norm": 0.6474215679166399,
-      "learning_rate": 3.258064516129033e-05,
-      "loss": 0.446,
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.9369043755410301,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.4553,
       "step": 101
     },
     {
-      "epoch": 0.41295546558704455,
-      "grad_norm": 0.6689520288805223,
-      "learning_rate": 3.2903225806451614e-05,
-      "loss": 0.456,
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.1523424036107508,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.4696,
       "step": 102
     },
     {
-      "epoch": 0.41700404858299595,
-      "grad_norm": 0.8554401202095955,
-      "learning_rate": 3.322580645161291e-05,
-      "loss": 0.4849,
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8777163014271671,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.481,
       "step": 103
     },
     {
-      "epoch": 0.42105263157894735,
-      "grad_norm": 0.8610528320864348,
-      "learning_rate": 3.3548387096774195e-05,
-      "loss": 0.4661,
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.0967852812965075,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.4803,
       "step": 104
     },
     {
-      "epoch": 0.4251012145748988,
-      "grad_norm": 0.578489552350898,
-      "learning_rate": 3.387096774193549e-05,
-      "loss": 0.4504,
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.8147757607633831,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.4678,
       "step": 105
     },
     {
-      "epoch": 0.4291497975708502,
-      "grad_norm": 0.6562651163306047,
-      "learning_rate": 3.4193548387096776e-05,
-      "loss": 0.431,
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.8853309357938409,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.4368,
       "step": 106
     },
     {
-      "epoch": 0.4331983805668016,
-      "grad_norm": 0.7553871379053245,
-      "learning_rate": 3.451612903225807e-05,
-      "loss": 0.4614,
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.9564210431734049,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.4701,
       "step": 107
     },
     {
-      "epoch": 0.43724696356275305,
-      "grad_norm": 0.5283311907538893,
-      "learning_rate": 3.483870967741936e-05,
-      "loss": 0.4771,
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.9690662289828335,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.4808,
       "step": 108
     },
     {
-      "epoch": 0.44129554655870445,
-      "grad_norm": 0.8229527423594084,
-      "learning_rate": 3.516129032258065e-05,
-      "loss": 0.4369,
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.7858176985420097,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.457,
       "step": 109
     },
     {
-      "epoch": 0.44534412955465585,
-      "grad_norm": 0.7931062793163733,
-      "learning_rate": 3.548387096774194e-05,
-      "loss": 0.4423,
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.083866081000804,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.4483,
       "step": 110
     },
     {
-      "epoch": 0.4493927125506073,
-      "grad_norm": 0.7918137877459142,
-      "learning_rate": 3.5806451612903225e-05,
-      "loss": 0.4464,
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.8656989772716427,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.455,
       "step": 111
     },
     {
-      "epoch": 0.4534412955465587,
-      "grad_norm": 0.7768895449136866,
-      "learning_rate": 3.612903225806452e-05,
-      "loss": 0.4468,
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.9734460499136482,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.4536,
       "step": 112
     },
     {
-      "epoch": 0.4574898785425101,
-      "grad_norm": 0.8109813487310678,
-      "learning_rate": 3.6451612903225805e-05,
-      "loss": 0.4559,
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.0629044896780213,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.4564,
       "step": 113
     },
     {
-      "epoch": 0.46153846153846156,
-      "grad_norm": 0.9845400557110856,
-      "learning_rate": 3.67741935483871e-05,
-      "loss": 0.442,
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.9472612738313319,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.4399,
       "step": 114
     },
     {
-      "epoch": 0.46558704453441296,
-      "grad_norm": 1.003042055152213,
-      "learning_rate": 3.7096774193548386e-05,
-      "loss": 0.452,
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.7700681900785087,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.4481,
       "step": 115
     },
     {
-      "epoch": 0.46963562753036436,
-      "grad_norm": 0.7817066487104143,
-      "learning_rate": 3.741935483870968e-05,
-      "loss": 0.4578,
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.8600872899913012,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.4588,
       "step": 116
     },
     {
-      "epoch": 0.47368421052631576,
-      "grad_norm": 0.6790114458769083,
-      "learning_rate": 3.7741935483870974e-05,
-      "loss": 0.4598,
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.7869178801024028,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.4614,
       "step": 117
     },
     {
-      "epoch": 0.4777327935222672,
-      "grad_norm": 0.6950734412396813,
-      "learning_rate": 3.806451612903226e-05,
-      "loss": 0.4568,
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.9409557851649036,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.4608,
       "step": 118
     },
     {
-      "epoch": 0.4817813765182186,
-      "grad_norm": 0.6310598236091703,
-      "learning_rate": 3.8387096774193555e-05,
-      "loss": 0.4268,
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.9508703852174584,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.4334,
       "step": 119
     },
     {
-      "epoch": 0.48582995951417,
-      "grad_norm": 0.8268877298317439,
-      "learning_rate": 3.870967741935484e-05,
-      "loss": 0.4531,
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.2869892158810599,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.4602,
       "step": 120
     },
     {
-      "epoch": 0.4898785425101215,
-      "grad_norm": 0.8608059617477575,
-      "learning_rate": 3.9032258064516135e-05,
-      "loss": 0.4632,
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.861135831418999,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.4673,
       "step": 121
     },
     {
-      "epoch": 0.4939271255060729,
-      "grad_norm": 0.6959437765530498,
-      "learning_rate": 3.935483870967742e-05,
-      "loss": 0.4563,
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.2362980881789531,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.4645,
       "step": 122
     },
     {
-      "epoch": 0.4979757085020243,
-      "grad_norm": 0.5536111113512044,
-      "learning_rate": 3.9677419354838716e-05,
-      "loss": 0.441,
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.1135472161270028,
+      "learning_rate": 4e-05,
+      "loss": 0.45,
       "step": 123
     },
     {
-      "epoch": 0.5020242914979757,
-      "grad_norm": 0.6994367200209375,
-      "learning_rate": 4e-05,
-      "loss": 0.4477,
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.7333454071757762,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.4565,
       "step": 124
     },
     {
-      "epoch": 0.5060728744939271,
-      "grad_norm": 0.552529443342419,
-      "learning_rate": 3.9999920040266474e-05,
-      "loss": 0.445,
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.019242570249519,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.453,
       "step": 125
     },
     {
-      "epoch": 0.5101214574898786,
-      "grad_norm": 0.6241656917746695,
-      "learning_rate": 3.999968016170525e-05,
-      "loss": 0.4476,
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.7322353273688116,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.456,
       "step": 126
     },
     {
-      "epoch": 0.5141700404858299,
-      "grad_norm": 0.6578110011474876,
-      "learning_rate": 3.9999280366234394e-05,
-      "loss": 0.4415,
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.7946693623120755,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.4498,
       "step": 127
     },
     {
-      "epoch": 0.5182186234817814,
-      "grad_norm": 0.5333835798006835,
-      "learning_rate": 3.999872065705065e-05,
-      "loss": 0.4446,
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.0388449239066677,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.4498,
       "step": 128
     },
     {
-      "epoch": 0.5222672064777328,
-      "grad_norm": 0.5991056006299307,
-      "learning_rate": 3.999800103862945e-05,
-      "loss": 0.4276,
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.135110124648114,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.4404,
       "step": 129
     },
     {
-      "epoch": 0.5263157894736842,
-      "grad_norm": 0.6007573224544371,
-      "learning_rate": 3.9997121516724834e-05,
-      "loss": 0.45,
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.1372737967625657,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.4592,
       "step": 130
     },
     {
-      "epoch": 0.5303643724696356,
-      "grad_norm": 0.7427149281512079,
-      "learning_rate": 3.999608209836944e-05,
-      "loss": 0.456,
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.8630816426425227,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.4644,
       "step": 131
     },
     {
-      "epoch": 0.5344129554655871,
-      "grad_norm": 0.7360838323937663,
-      "learning_rate": 3.999488279187443e-05,
-      "loss": 0.4747,
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.8500736727727053,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.4879,
       "step": 132
     },
     {
-      "epoch": 0.5384615384615384,
-      "grad_norm": 0.6860472813016516,
-      "learning_rate": 3.9993523606829425e-05,
-      "loss": 0.4586,
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.8947651963339159,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.4588,
       "step": 133
     },
     {
-      "epoch": 0.5425101214574899,
-      "grad_norm": 0.556815394069109,
-      "learning_rate": 3.9992004554102435e-05,
-      "loss": 0.4493,
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.9039882713956191,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.4505,
       "step": 134
     },
     {
-      "epoch": 0.5465587044534413,
-      "grad_norm": 0.6016393925154496,
-      "learning_rate": 3.999032564583976e-05,
-      "loss": 0.4584,
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.235880416312397,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.4733,
       "step": 135
     },
     {
-      "epoch": 0.5506072874493927,
-      "grad_norm": 0.6997702921000107,
-      "learning_rate": 3.998848689546591e-05,
-      "loss": 0.447,
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.10516076615186,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.4522,
       "step": 136
     },
     {
-      "epoch": 0.5546558704453441,
-      "grad_norm": 0.664426564578934,
-      "learning_rate": 3.9986488317683496e-05,
-      "loss": 0.4665,
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.7969459482130627,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.4657,
       "step": 137
     },
     {
-      "epoch": 0.5587044534412956,
-      "grad_norm": 0.9341623188738303,
-      "learning_rate": 3.9984329928473067e-05,
-      "loss": 0.4696,
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.340455735884462,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.4816,
       "step": 138
     },
     {
-      "epoch": 0.562753036437247,
-      "grad_norm": 0.8284445426813348,
-      "learning_rate": 3.998201174509306e-05,
-      "loss": 0.4321,
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.6699872282344217,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.438,
       "step": 139
     },
     {
-      "epoch": 0.5668016194331984,
-      "grad_norm": 0.6963359375393712,
-      "learning_rate": 3.997953378607961e-05,
-      "loss": 0.4512,
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.9923431056316986,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.4561,
       "step": 140
     },
     {
-      "epoch": 0.5708502024291497,
-      "grad_norm": 0.591586822770761,
-      "learning_rate": 3.9976896071246403e-05,
-      "loss": 0.4262,
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.7495611181405863,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.4332,
       "step": 141
     },
     {
-      "epoch": 0.5748987854251012,
-      "grad_norm": 1.8154412368115942,
-      "learning_rate": 3.997409862168454e-05,
-      "loss": 0.4562,
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.7674138712027692,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.4558,
       "step": 142
     },
     {
-      "epoch": 0.5789473684210527,
-      "grad_norm": 0.7040644741222044,
-      "learning_rate": 3.997114145976237e-05,
-      "loss": 0.4634,
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.8579357755761065,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.4642,
       "step": 143
     },
     {
-      "epoch": 0.582995951417004,
-      "grad_norm": 0.7259336136835651,
-      "learning_rate": 3.9968024609125245e-05,
-      "loss": 0.4484,
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.7466935715508728,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.452,
       "step": 144
     },
     {
-      "epoch": 0.5870445344129555,
-      "grad_norm": 0.867893051550791,
-      "learning_rate": 3.996474809469545e-05,
-      "loss": 0.4486,
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.8146596329769796,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.4573,
       "step": 145
     },
     {
-      "epoch": 0.5910931174089069,
-      "grad_norm": 0.6654390870019895,
-      "learning_rate": 3.996131194267189e-05,
-      "loss": 0.4322,
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.6654182512833235,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.4515,
       "step": 146
     },
     {
-      "epoch": 0.5951417004048583,
-      "grad_norm": 0.821022458037405,
-      "learning_rate": 3.995771618052995e-05,
-      "loss": 0.4477,
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.0291498511074015,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.4525,
       "step": 147
     },
     {
-      "epoch": 0.5991902834008097,
-      "grad_norm": 1.1262045652638697,
-      "learning_rate": 3.9953960837021256e-05,
-      "loss": 0.4604,
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.9119972665400535,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.4662,
       "step": 148
     },
     {
-      "epoch": 0.6032388663967612,
-      "grad_norm": 0.6465181652444291,
-      "learning_rate": 3.995004594217343e-05,
-      "loss": 0.4539,
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.1943249324314062,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.4591,
       "step": 149
     },
     {
-      "epoch": 0.6072874493927125,
-      "grad_norm": 0.7503765556558867,
-      "learning_rate": 3.994597152728987e-05,
-      "loss": 0.4416,
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.9196475326700626,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.4501,
       "step": 150
     },
     {
-      "epoch": 0.611336032388664,
-      "grad_norm": 0.6426516705115229,
-      "learning_rate": 3.994173762494947e-05,
-      "loss": 0.442,
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.0217956144731273,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.4504,
       "step": 151
     },
     {
-      "epoch": 0.6153846153846154,
-      "grad_norm": 0.6496051605195963,
-      "learning_rate": 3.9937344269006424e-05,
-      "loss": 0.466,
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8151062564999557,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.4746,
       "step": 152
     },
     {
-      "epoch": 0.6194331983805668,
-      "grad_norm": 0.7217291717406101,
-      "learning_rate": 3.993279149458988e-05,
-      "loss": 0.4386,
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.9561310208920438,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.4477,
       "step": 153
     },
     {
-      "epoch": 0.6234817813765182,
-      "grad_norm": 0.5699314787574253,
-      "learning_rate": 3.9928079338103695e-05,
-      "loss": 0.452,
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.8005929855598839,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.4539,
       "step": 154
     },
     {
-      "epoch": 0.6275303643724697,
-      "grad_norm": 0.6536943733459252,
-      "learning_rate": 3.992320783722616e-05,
-      "loss": 0.4433,
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.9343914591308471,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.4494,
       "step": 155
     },
     {
-      "epoch": 0.631578947368421,
-      "grad_norm": 0.5391965049095635,
-      "learning_rate": 3.991817703090965e-05,
-      "loss": 0.438,
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.7559765076377314,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.4448,
       "step": 156
     },
     {
-      "epoch": 0.6356275303643725,
-      "grad_norm": 0.6087615720346508,
-      "learning_rate": 3.991298695938038e-05,
-      "loss": 0.4534,
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.9686746165212624,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.4576,
       "step": 157
     },
     {
-      "epoch": 0.6396761133603239,
-      "grad_norm": 0.5333489596412007,
-      "learning_rate": 3.9907637664138e-05,
-      "loss": 0.4475,
+      "epoch": 0.64,
+      "grad_norm": 0.8624291865765256,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.4587,
       "step": 158
     },
     {
-      "epoch": 0.6437246963562753,
-      "grad_norm": 0.528236745877187,
-      "learning_rate": 3.990212918795534e-05,
-      "loss": 0.4305,
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.7522461516967455,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.4387,
       "step": 159
     },
     {
-      "epoch": 0.6477732793522267,
-      "grad_norm": 0.5168826785061815,
-      "learning_rate": 3.9896461574878036e-05,
-      "loss": 0.4467,
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.9943549183174756,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.4641,
       "step": 160
     },
     {
-      "epoch": 0.6518218623481782,
-      "grad_norm": 0.5432166311702046,
-      "learning_rate": 3.989063487022416e-05,
-      "loss": 0.4314,
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.8648214233500979,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.4333,
       "step": 161
     },
     {
-      "epoch": 0.6558704453441295,
-      "grad_norm": 0.5046526008009887,
-      "learning_rate": 3.9884649120583903e-05,
-      "loss": 0.4307,
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.9144199833628481,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.4311,
       "step": 162
     },
     {
-      "epoch": 0.659919028340081,
-      "grad_norm": 0.5002939082484585,
-      "learning_rate": 3.987850437381915e-05,
-      "loss": 0.4186,
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.6160331606022642,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.4195,
       "step": 163
     },
     {
-      "epoch": 0.6639676113360324,
-      "grad_norm": 0.5501922249635752,
-      "learning_rate": 3.987220067906313e-05,
-      "loss": 0.4275,
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.9840925294923062,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.438,
       "step": 164
     },
     {
-      "epoch": 0.6680161943319838,
-      "grad_norm": 0.586928959623514,
-      "learning_rate": 3.986573808672002e-05,
-      "loss": 0.4307,
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.6896596247092802,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.4408,
       "step": 165
     },
     {
-      "epoch": 0.6720647773279352,
-      "grad_norm": 0.6771080858759494,
-      "learning_rate": 3.9859116648464534e-05,
-      "loss": 0.4313,
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.8786166641505603,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.433,
       "step": 166
     },
     {
-      "epoch": 0.6761133603238867,
-      "grad_norm": 0.4361637917449624,
-      "learning_rate": 3.985233641724152e-05,
-      "loss": 0.4411,
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.9294817890537319,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.4464,
       "step": 167
     },
     {
-      "epoch": 0.680161943319838,
-      "grad_norm": 0.6225724081907832,
-      "learning_rate": 3.984539744726554e-05,
-      "loss": 0.4428,
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.7712211401634979,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.4453,
       "step": 168
     },
     {
-      "epoch": 0.6842105263157895,
-      "grad_norm": 0.6432253591421742,
-      "learning_rate": 3.983829979402038e-05,
-      "loss": 0.4325,
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.8776465716094822,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.445,
       "step": 169
     },
     {
-      "epoch": 0.6882591093117408,
-      "grad_norm": 0.5942877426279993,
-      "learning_rate": 3.983104351425871e-05,
-      "loss": 0.4813,
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.8703531138467249,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.4899,
       "step": 170
     },
     {
-      "epoch": 0.6923076923076923,
-      "grad_norm": 0.5983067077733915,
-      "learning_rate": 3.982362866600154e-05,
-      "loss": 0.4369,
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.9213802713321433,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.4411,
       "step": 171
     },
     {
-      "epoch": 0.6963562753036437,
-      "grad_norm": 0.5575582969779534,
-      "learning_rate": 3.9816055308537806e-05,
-      "loss": 0.4454,
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.1166837882667735,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.4539,
       "step": 172
     },
     {
-      "epoch": 0.7004048582995951,
-      "grad_norm": 0.537941788625238,
-      "learning_rate": 3.980832350242388e-05,
-      "loss": 0.4374,
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.9519590857159009,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.4351,
       "step": 173
     },
     {
-      "epoch": 0.7044534412955465,
-      "grad_norm": 0.5626358634899484,
-      "learning_rate": 3.980043330948306e-05,
-      "loss": 0.45,
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.6923171154268185,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.4587,
       "step": 174
     },
     {
-      "epoch": 0.708502024291498,
-      "grad_norm": 0.5688083811145629,
-      "learning_rate": 3.9792384792805124e-05,
-      "loss": 0.442,
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.0816732564236065,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.4458,
       "step": 175
     },
     {
-      "epoch": 0.7125506072874493,
-      "grad_norm": 0.5630893224543854,
-      "learning_rate": 3.978417801674579e-05,
-      "loss": 0.4474,
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.9351491146863047,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.4548,
       "step": 176
     },
     {
-      "epoch": 0.7165991902834008,
-      "grad_norm": 0.5240622516898986,
-      "learning_rate": 3.9775813046926244e-05,
-      "loss": 0.4601,
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.786837290901519,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.4664,
       "step": 177
     },
     {
-      "epoch": 0.7206477732793523,
-      "grad_norm": 0.4961703781825216,
-      "learning_rate": 3.976728995023254e-05,
-      "loss": 0.4443,
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6461030383201106,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.456,
       "step": 178
     },
     {
-      "epoch": 0.7246963562753036,
-      "grad_norm": 0.5736004582845314,
-      "learning_rate": 3.975860879481514e-05,
-      "loss": 0.4323,
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.8028222141406482,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.435,
       "step": 179
     },
     {
-      "epoch": 0.728744939271255,
-      "grad_norm": 0.5468739610792731,
-      "learning_rate": 3.974976965008833e-05,
-      "loss": 0.4528,
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7100026473934186,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.4622,
       "step": 180
     },
     {
-      "epoch": 0.7327935222672065,
-      "grad_norm": 0.6110410299151724,
-      "learning_rate": 3.9740772586729674e-05,
-      "loss": 0.4182,
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.7375145537972061,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.4268,
       "step": 181
     },
     {
-      "epoch": 0.7368421052631579,
-      "grad_norm": 0.625541178769375,
-      "learning_rate": 3.973161767667946e-05,
-      "loss": 0.4362,
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.5820506433646312,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.441,
       "step": 182
     },
     {
-      "epoch": 0.7408906882591093,
-      "grad_norm": 0.5018618437286215,
-      "learning_rate": 3.972230499314009e-05,
-      "loss": 0.425,
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.7625667547273004,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.4433,
       "step": 183
     },
     {
-      "epoch": 0.7449392712550608,
-      "grad_norm": 0.5408341312942263,
-      "learning_rate": 3.9712834610575545e-05,
-      "loss": 0.4399,
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.7502233311018757,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.4525,
       "step": 184
     },
     {
-      "epoch": 0.7489878542510121,
-      "grad_norm": 0.5869029231194034,
-      "learning_rate": 3.9703206604710746e-05,
-      "loss": 0.445,
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.894604683458091,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.4526,
       "step": 185
     },
     {
-      "epoch": 0.7530364372469636,
-      "grad_norm": 0.4526367212939427,
-      "learning_rate": 3.969342105253098e-05,
-      "loss": 0.4372,
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.8287719382727782,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.4472,
       "step": 186
     },
     {
-      "epoch": 0.757085020242915,
-      "grad_norm": 0.46412774456331146,
-      "learning_rate": 3.9683478032281246e-05,
-      "loss": 0.4402,
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.711131139808416,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.4482,
       "step": 187
     },
     {
-      "epoch": 0.7611336032388664,
-      "grad_norm": 0.47331086468317757,
-      "learning_rate": 3.967337762346568e-05,
-      "loss": 0.4481,
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.9777462845344229,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.4621,
       "step": 188
     },
     {
-      "epoch": 0.7651821862348178,
-      "grad_norm": 0.4503366644126491,
-      "learning_rate": 3.9663119906846885e-05,
-      "loss": 0.4304,
+      "epoch": 0.7655696202531646,
+      "grad_norm": 1.250352414503313,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.4463,
       "step": 189
     },
     {
-      "epoch": 0.7692307692307693,
-      "grad_norm": 0.5448055578366483,
-      "learning_rate": 3.965270496444528e-05,
-      "loss": 0.4606,
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.6005486076518204,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.4674,
       "step": 190
     },
     {
-      "epoch": 0.7732793522267206,
-      "grad_norm": 0.610049151729368,
-      "learning_rate": 3.9642132879538465e-05,
-      "loss": 0.4325,
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.8057912255057355,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.4453,
       "step": 191
     },
     {
-      "epoch": 0.7773279352226721,
-      "grad_norm": 0.4768340228015045,
-      "learning_rate": 3.9631403736660566e-05,
-      "loss": 0.4558,
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.713221240764434,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.4542,
       "step": 192
     },
     {
-      "epoch": 0.7813765182186235,
-      "grad_norm": 0.4521727963891733,
-      "learning_rate": 3.962051762160151e-05,
-      "loss": 0.4517,
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.6831748835092057,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.4595,
       "step": 193
     },
     {
-      "epoch": 0.7854251012145749,
-      "grad_norm": 0.47780668775336443,
-      "learning_rate": 3.960947462140638e-05,
-      "loss": 0.4481,
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6829569879923053,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.4538,
       "step": 194
     },
     {
-      "epoch": 0.7894736842105263,
-      "grad_norm": 0.43942143491756364,
-      "learning_rate": 3.959827482437472e-05,
-      "loss": 0.434,
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.7565038644117508,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.4364,
       "step": 195
     },
     {
-      "epoch": 0.7935222672064778,
-      "grad_norm": 0.4009005580652296,
-      "learning_rate": 3.958691832005981e-05,
-      "loss": 0.439,
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.681551621984661,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.4448,
       "step": 196
     },
     {
-      "epoch": 0.7975708502024291,
-      "grad_norm": 0.41835716244319204,
-      "learning_rate": 3.9575405199267944e-05,
-      "loss": 0.4378,
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.7116603526580798,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.4413,
       "step": 197
     },
     {
-      "epoch": 0.8016194331983806,
-      "grad_norm": 0.4567113317418262,
-      "learning_rate": 3.956373555405774e-05,
-      "loss": 0.4525,
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6586362577180774,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.4606,
       "step": 198
     },
     {
-      "epoch": 0.805668016194332,
-      "grad_norm": 0.5045407017294614,
-      "learning_rate": 3.955190947773937e-05,
-      "loss": 0.4275,
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.7193161060190371,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.4318,
       "step": 199
     },
     {
-      "epoch": 0.8097165991902834,
-      "grad_norm": 0.4460323040229284,
-      "learning_rate": 3.9539927064873824e-05,
-      "loss": 0.449,
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.6164839371157329,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.4586,
       "step": 200
     },
     {
-      "epoch": 0.8137651821862348,
-      "grad_norm": 0.48726368697771305,
-      "learning_rate": 3.952778841127214e-05,
-      "loss": 0.4443,
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.849329431146727,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.4476,
       "step": 201
     },
     {
-      "epoch": 0.8178137651821862,
-      "grad_norm": 0.47687403576363474,
-      "learning_rate": 3.95154936139947e-05,
-      "loss": 0.4312,
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.8337537239344638,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.446,
       "step": 202
     },
     {
-      "epoch": 0.8218623481781376,
-      "grad_norm": 0.5465666902430841,
-      "learning_rate": 3.950304277135034e-05,
-      "loss": 0.4335,
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.8891521719348002,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.4456,
       "step": 203
     },
     {
-      "epoch": 0.8259109311740891,
-      "grad_norm": 0.5028234465626226,
-      "learning_rate": 3.94904359828957e-05,
-      "loss": 0.4399,
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.7275452200919326,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.4577,
       "step": 204
     },
     {
-      "epoch": 0.8299595141700404,
-      "grad_norm": 0.4377224747177585,
-      "learning_rate": 3.94776733494343e-05,
-      "loss": 0.4265,
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.7677006463582865,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.4296,
       "step": 205
     },
     {
-      "epoch": 0.8340080971659919,
-      "grad_norm": 0.5863581496152291,
-      "learning_rate": 3.9464754973015826e-05,
-      "loss": 0.4494,
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.726219044247643,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.4598,
       "step": 206
     },
     {
-      "epoch": 0.8380566801619433,
-      "grad_norm": 0.5109528156806592,
-      "learning_rate": 3.9451680956935275e-05,
-      "loss": 0.4171,
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.8712519666180191,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.4269,
       "step": 207
     },
     {
-      "epoch": 0.8421052631578947,
-      "grad_norm": 0.46367879751545155,
-      "learning_rate": 3.943845140573213e-05,
-      "loss": 0.4418,
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.7736034660576943,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.455,
       "step": 208
     },
     {
-      "epoch": 0.8461538461538461,
-      "grad_norm": 0.4557726172809232,
-      "learning_rate": 3.942506642518952e-05,
-      "loss": 0.4101,
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5639469627986099,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.4289,
       "step": 209
     },
     {
-      "epoch": 0.8502024291497976,
-      "grad_norm": 0.5019640708209037,
-      "learning_rate": 3.9411526122333396e-05,
-      "loss": 0.4072,
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.8109880420039349,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.4226,
       "step": 210
     },
     {
-      "epoch": 0.854251012145749,
-      "grad_norm": 0.4333564525252833,
-      "learning_rate": 3.939783060543167e-05,
-      "loss": 0.4135,
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5989567522977639,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.427,
       "step": 211
     },
     {
-      "epoch": 0.8582995951417004,
-      "grad_norm": 0.4094637774333867,
-      "learning_rate": 3.938397998399333e-05,
-      "loss": 0.416,
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5812948529796553,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.4183,
       "step": 212
     },
     {
-      "epoch": 0.8623481781376519,
-      "grad_norm": 0.4600187913955183,
-      "learning_rate": 3.936997436876756e-05,
-      "loss": 0.4143,
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.7890820798188667,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.4171,
       "step": 213
     },
     {
-      "epoch": 0.8663967611336032,
-      "grad_norm": 0.373363798373944,
-      "learning_rate": 3.935581387174289e-05,
-      "loss": 0.4424,
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.7594574392233743,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.441,
       "step": 214
     },
     {
-      "epoch": 0.8704453441295547,
-      "grad_norm": 0.39091006479781826,
-      "learning_rate": 3.9341498606146295e-05,
-      "loss": 0.4364,
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.5768984502401866,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.44,
       "step": 215
     },
     {
-      "epoch": 0.8744939271255061,
-      "grad_norm": 0.45776599019751363,
-      "learning_rate": 3.932702868644224e-05,
-      "loss": 0.4323,
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.8051037948139778,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.4357,
       "step": 216
     },
     {
-      "epoch": 0.8785425101214575,
-      "grad_norm": 0.4981860703772371,
-      "learning_rate": 3.931240422833181e-05,
-      "loss": 0.4522,
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.9534768918141695,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.4684,
       "step": 217
     },
     {
-      "epoch": 0.8825910931174089,
-      "grad_norm": 0.44101504090581095,
-      "learning_rate": 3.9297625348751806e-05,
-      "loss": 0.4395,
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.5672340060184335,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.4476,
       "step": 218
     },
     {
-      "epoch": 0.8866396761133604,
-      "grad_norm": 0.4786888557772545,
-      "learning_rate": 3.9282692165873744e-05,
-      "loss": 0.4398,
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.7729358434145178,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.4617,
       "step": 219
     },
     {
-      "epoch": 0.8906882591093117,
-      "grad_norm": 0.5034020731179247,
-      "learning_rate": 3.9267604799102945e-05,
-      "loss": 0.4264,
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.73363675365063,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.4325,
       "step": 220
     },
     {
-      "epoch": 0.8947368421052632,
-      "grad_norm": 0.42648788449552244,
-      "learning_rate": 3.9252363369077615e-05,
-      "loss": 0.4644,
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6816762357696874,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.4702,
       "step": 221
     },
     {
-      "epoch": 0.8987854251012146,
-      "grad_norm": 0.469901247795006,
-      "learning_rate": 3.9236967997667796e-05,
-      "loss": 0.4376,
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.7804679701912035,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.4426,
       "step": 222
     },
     {
-      "epoch": 0.902834008097166,
-      "grad_norm": 0.4261383876049414,
-      "learning_rate": 3.9221418807974496e-05,
-      "loss": 0.4295,
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.8071790122937856,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.4406,
       "step": 223
     },
     {
-      "epoch": 0.9068825910931174,
-      "grad_norm": 0.46036042926277265,
-      "learning_rate": 3.92057159243286e-05,
-      "loss": 0.4281,
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6148119450322652,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.4369,
       "step": 224
     },
     {
-      "epoch": 0.9109311740890689,
-      "grad_norm": 0.4049097118196568,
-      "learning_rate": 3.9189859472289956e-05,
-      "loss": 0.4105,
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6845736818913868,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.4154,
       "step": 225
     },
     {
-      "epoch": 0.9149797570850202,
-      "grad_norm": 0.4736214397913212,
-      "learning_rate": 3.9173849578646324e-05,
-      "loss": 0.4373,
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.6977925119330053,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.4457,
       "step": 226
     },
     {
-      "epoch": 0.9190283400809717,
-      "grad_norm": 0.5373257486256438,
-      "learning_rate": 3.9157686371412395e-05,
-      "loss": 0.4349,
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.5733329379222073,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.439,
       "step": 227
     },
     {
-      "epoch": 0.9230769230769231,
-      "grad_norm": 0.5789010488974032,
-      "learning_rate": 3.914136997982874e-05,
-      "loss": 0.4389,
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9146722529543119,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.4467,
       "step": 228
     },
     {
-      "epoch": 0.9271255060728745,
-      "grad_norm": 0.613302914232418,
-      "learning_rate": 3.912490053436079e-05,
-      "loss": 0.4345,
+      "epoch": 0.9275949367088607,
+      "grad_norm": 1.1862523912499803,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.4399,
       "step": 229
     },
     {
-      "epoch": 0.9311740890688259,
-      "grad_norm": 0.44238979210800766,
-      "learning_rate": 3.9108278166697805e-05,
-      "loss": 0.4282,
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.7031886799401439,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.4365,
       "step": 230
     },
     {
-      "epoch": 0.9352226720647774,
-      "grad_norm": 0.5312279636441718,
-      "learning_rate": 3.909150300975178e-05,
-      "loss": 0.4479,
+      "epoch": 0.9356962025316455,
+      "grad_norm": 1.0276939467382085,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.4577,
       "step": 231
     },
     {
-      "epoch": 0.9392712550607287,
-      "grad_norm": 0.5042796870235429,
-      "learning_rate": 3.907457519765642e-05,
-      "loss": 0.4189,
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.7552633433962823,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.4374,
       "step": 232
     },
     {
-      "epoch": 0.9433198380566802,
-      "grad_norm": 0.49667151407247295,
-      "learning_rate": 3.905749486576607e-05,
-      "loss": 0.4328,
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.8137481666394772,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.4444,
       "step": 233
     },
     {
-      "epoch": 0.9473684210526315,
-      "grad_norm": 0.4672324245931207,
-      "learning_rate": 3.90402621506546e-05,
-      "loss": 0.4332,
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.7716880563723987,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.4405,
       "step": 234
     },
     {
-      "epoch": 0.951417004048583,
-      "grad_norm": 0.4679935880886033,
-      "learning_rate": 3.902287719011435e-05,
-      "loss": 0.417,
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.789436580699524,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.4231,
       "step": 235
     },
     {
-      "epoch": 0.9554655870445344,
-      "grad_norm": 0.3907772195673455,
-      "learning_rate": 3.900534012315499e-05,
-      "loss": 0.429,
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.5431259468129148,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.4347,
       "step": 236
     },
     {
-      "epoch": 0.9595141700404858,
-      "grad_norm": 0.46998236290016066,
-      "learning_rate": 3.8987651090002454e-05,
-      "loss": 0.4074,
+      "epoch": 0.96,
+      "grad_norm": 0.6307559995378647,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.4119,
       "step": 237
     },
     {
-      "epoch": 0.9635627530364372,
-      "grad_norm": 0.39022286885350754,
-      "learning_rate": 3.896981023209777e-05,
-      "loss": 0.4484,
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5319816816300302,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.4574,
       "step": 238
     },
     {
-      "epoch": 0.9676113360323887,
-      "grad_norm": 0.5278122169637925,
-      "learning_rate": 3.8951817692095966e-05,
-      "loss": 0.4436,
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.5244026649382629,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.4592,
       "step": 239
     },
     {
-      "epoch": 0.97165991902834,
-      "grad_norm": 0.41188102714276525,
-      "learning_rate": 3.893367361386491e-05,
-      "loss": 0.438,
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.585928885455211,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.4369,
       "step": 240
     },
     {
-      "epoch": 0.9757085020242915,
-      "grad_norm": 0.500037905109882,
-      "learning_rate": 3.891537814248417e-05,
-      "loss": 0.4253,
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.5723493604296689,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.4287,
       "step": 241
     },
     {
-      "epoch": 0.979757085020243,
-      "grad_norm": 0.40419475647499165,
-      "learning_rate": 3.889693142424384e-05,
-      "loss": 0.4269,
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.5975318664501996,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.4312,
       "step": 242
     },
     {
-      "epoch": 0.9838056680161943,
-      "grad_norm": 0.5082802486554187,
-      "learning_rate": 3.8878333606643405e-05,
-      "loss": 0.4336,
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.6768588704055118,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.441,
       "step": 243
     },
     {
-      "epoch": 0.9878542510121457,
-      "grad_norm": 0.5144520891995027,
-      "learning_rate": 3.885958483839049e-05,
-      "loss": 0.4063,
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6010752967507942,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.4136,
       "step": 244
     },
     {
-      "epoch": 0.9919028340080972,
-      "grad_norm": 0.3927946142631165,
-      "learning_rate": 3.8840685269399786e-05,
-      "loss": 0.4041,
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6197608643777256,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.4218,
       "step": 245
     },
     {
-      "epoch": 0.9959514170040485,
-      "grad_norm": 0.5015012976691898,
-      "learning_rate": 3.882163505079171e-05,
-      "loss": 0.434,
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.5902047119679896,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.4406,
       "step": 246
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.4161618428337255,
-      "learning_rate": 3.880243433489132e-05,
-      "loss": 0.4165,
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.6612460734303174,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.4935,
       "step": 247
     },
     {
-      "epoch": 1.0040485829959513,
-      "grad_norm": 0.4622298440967577,
-      "learning_rate": 3.878308327522702e-05,
-      "loss": 0.3952,
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.5932922847514694,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.3662,
       "step": 248
     },
     {
-      "epoch": 1.008097165991903,
-      "grad_norm": 0.4273965905907388,
-      "learning_rate": 3.876358202652936e-05,
-      "loss": 0.3903,
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.5301115451685868,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.3684,
       "step": 249
     },
     {
-      "epoch": 1.0121457489878543,
-      "grad_norm": 0.46025736090938835,
-      "learning_rate": 3.8743930744729835e-05,
-      "loss": 0.3412,
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.6070925202795017,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.3749,
       "step": 250
     },
     {
-      "epoch": 1.0161943319838056,
-      "grad_norm": 0.458546726644252,
-      "learning_rate": 3.8724129586959535e-05,
-      "loss": 0.3678,
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.6223103378368309,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.3693,
       "step": 251
     },
     {
-      "epoch": 1.0202429149797572,
-      "grad_norm": 0.4511178872213509,
-      "learning_rate": 3.870417871154801e-05,
-      "loss": 0.3824,
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.5926131873276659,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.3822,
       "step": 252
     },
     {
-      "epoch": 1.0242914979757085,
-      "grad_norm": 0.4495997062773142,
-      "learning_rate": 3.868407827802192e-05,
-      "loss": 0.3618,
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.5822583703652573,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.3525,
       "step": 253
     },
     {
-      "epoch": 1.0283400809716599,
-      "grad_norm": 0.43849049306894494,
-      "learning_rate": 3.8663828447103804e-05,
-      "loss": 0.3756,
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.6642197900714935,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.3877,
       "step": 254
     },
     {
-      "epoch": 1.0323886639676114,
-      "grad_norm": 0.4286176981640959,
-      "learning_rate": 3.8643429380710765e-05,
-      "loss": 0.3702,
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.5393015463691317,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.3652,
       "step": 255
     },
     {
-      "epoch": 1.0364372469635628,
-      "grad_norm": 0.4602106917839205,
-      "learning_rate": 3.862288124195319e-05,
-      "loss": 0.371,
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.7091442404104473,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.3943,
       "step": 256
     },
     {
-      "epoch": 1.040485829959514,
-      "grad_norm": 0.4533721495685298,
-      "learning_rate": 3.8602184195133464e-05,
-      "loss": 0.3777,
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.6795099681960469,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.3727,
       "step": 257
     },
     {
-      "epoch": 1.0445344129554657,
-      "grad_norm": 0.4708636929031252,
-      "learning_rate": 3.8581338405744604e-05,
-      "loss": 0.3624,
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.7071539459434137,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.3957,
       "step": 258
     },
     {
-      "epoch": 1.048582995951417,
-      "grad_norm": 0.5041734017917592,
-      "learning_rate": 3.8560344040468996e-05,
-      "loss": 0.3932,
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.6693365089263168,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.3893,
       "step": 259
     },
     {
-      "epoch": 1.0526315789473684,
-      "grad_norm": 0.5103459100580762,
-      "learning_rate": 3.853920126717703e-05,
-      "loss": 0.3857,
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.7122192116507396,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.4033,
       "step": 260
     },
     {
-      "epoch": 1.05668016194332,
-      "grad_norm": 0.505921571434978,
-      "learning_rate": 3.851791025492574e-05,
-      "loss": 0.3709,
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.639231562089442,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.3741,
       "step": 261
     },
     {
-      "epoch": 1.0607287449392713,
-      "grad_norm": 0.5012809475943975,
-      "learning_rate": 3.8496471173957515e-05,
-      "loss": 0.3833,
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.7027477942731284,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.3954,
       "step": 262
     },
     {
-      "epoch": 1.0647773279352226,
-      "grad_norm": 0.4541601143310206,
-      "learning_rate": 3.847488419569866e-05,
-      "loss": 0.3598,
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.5444536598691878,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.3475,
       "step": 263
     },
     {
-      "epoch": 1.0688259109311742,
-      "grad_norm": 0.5453425691751234,
-      "learning_rate": 3.8453149492758077e-05,
-      "loss": 0.3723,
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.6523494542191236,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.3655,
       "step": 264
     },
     {
-      "epoch": 1.0728744939271255,
-      "grad_norm": 0.3973811631733981,
-      "learning_rate": 3.8431267238925886e-05,
-      "loss": 0.3403,
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.6626519871848646,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.367,
       "step": 265
     },
     {
-      "epoch": 1.0769230769230769,
-      "grad_norm": 0.5426938748750524,
-      "learning_rate": 3.840923760917199e-05,
-      "loss": 0.3766,
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.7644467399750277,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.3656,
       "step": 266
     },
     {
-      "epoch": 1.0809716599190284,
-      "grad_norm": 0.4069740020946505,
-      "learning_rate": 3.838706077964473e-05,
-      "loss": 0.3383,
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.6973611727954419,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.3752,
       "step": 267
     },
     {
-      "epoch": 1.0850202429149798,
-      "grad_norm": 0.4172293011476944,
-      "learning_rate": 3.836473692766944e-05,
-      "loss": 0.3782,
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.5858740716211058,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.3667,
       "step": 268
     },
     {
-      "epoch": 1.0890688259109311,
-      "grad_norm": 0.43708781285658554,
-      "learning_rate": 3.834226623174705e-05,
-      "loss": 0.3433,
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.6101456636712379,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.3827,
       "step": 269
     },
     {
-      "epoch": 1.0931174089068827,
-      "grad_norm": 0.46609805044304803,
-      "learning_rate": 3.831964887155264e-05,
-      "loss": 0.3793,
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.5292638190420572,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.3679,
       "step": 270
     },
     {
-      "epoch": 1.097165991902834,
-      "grad_norm": 0.503138948418934,
-      "learning_rate": 3.829688502793403e-05,
-      "loss": 0.3669,
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.6337058668543881,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.3847,
       "step": 271
     },
     {
-      "epoch": 1.1012145748987854,
-      "grad_norm": 0.46090120289878994,
-      "learning_rate": 3.827397488291029e-05,
-      "loss": 0.3788,
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.6122962027854479,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.3794,
       "step": 272
     },
     {
-      "epoch": 1.1052631578947367,
-      "grad_norm": 0.4928127101905674,
-      "learning_rate": 3.8250918619670343e-05,
-      "loss": 0.3879,
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.5789757692251489,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.3739,
       "step": 273
     },
     {
-      "epoch": 1.1093117408906883,
-      "grad_norm": 0.4177227737773772,
-      "learning_rate": 3.822771642257145e-05,
-      "loss": 0.3511,
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.5827618687766665,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.3493,
       "step": 274
     },
     {
-      "epoch": 1.1133603238866396,
-      "grad_norm": 0.5097701315655183,
-      "learning_rate": 3.820436847713776e-05,
-      "loss": 0.3647,
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.5241311744105478,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.3631,
       "step": 275
     },
     {
-      "epoch": 1.117408906882591,
-      "grad_norm": 0.46290444726245245,
-      "learning_rate": 3.8180874970058826e-05,
-      "loss": 0.3558,
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.6650795720874478,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.3902,
       "step": 276
     },
     {
-      "epoch": 1.1214574898785425,
-      "grad_norm": 0.43918695003619634,
-      "learning_rate": 3.8157236089188103e-05,
-      "loss": 0.3729,
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.47903471214970306,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.3627,
       "step": 277
     },
     {
-      "epoch": 1.125506072874494,
-      "grad_norm": 0.514597048401357,
-      "learning_rate": 3.813345202354145e-05,
-      "loss": 0.3514,
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.6415214755319258,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.353,
       "step": 278
     },
     {
-      "epoch": 1.1295546558704452,
-      "grad_norm": 0.412323412455217,
-      "learning_rate": 3.810952296329563e-05,
-      "loss": 0.3632,
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.6745891003534598,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.3861,
       "step": 279
     },
     {
-      "epoch": 1.1336032388663968,
-      "grad_norm": 0.4286356355329969,
-      "learning_rate": 3.808544909978676e-05,
-      "loss": 0.3635,
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.4712793149747154,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.3423,
       "step": 280
     },
     {
-      "epoch": 1.1376518218623481,
-      "grad_norm": 0.4159491785796094,
-      "learning_rate": 3.806123062550882e-05,
-      "loss": 0.3795,
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.6371351586212599,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.3938,
       "step": 281
     },
     {
-      "epoch": 1.1417004048582995,
-      "grad_norm": 0.3948641972733011,
-      "learning_rate": 3.803686773411208e-05,
-      "loss": 0.3593,
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.5027575051790194,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.3608,
       "step": 282
     },
     {
-      "epoch": 1.145748987854251,
-      "grad_norm": 0.4811275133564173,
-      "learning_rate": 3.801236062040158e-05,
-      "loss": 0.3666,
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.62036324908681,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.3712,
       "step": 283
     },
     {
-      "epoch": 1.1497975708502024,
-      "grad_norm": 0.37593595275566816,
-      "learning_rate": 3.798770948033553e-05,
-      "loss": 0.3687,
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.4696951764734086,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.364,
       "step": 284
     },
     {
-      "epoch": 1.1538461538461537,
-      "grad_norm": 0.44753403594644825,
-      "learning_rate": 3.796291451102381e-05,
-      "loss": 0.3687,
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.6847070791153437,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4222,
       "step": 285
     },
     {
-      "epoch": 1.1578947368421053,
-      "grad_norm": 0.4213417149629224,
-      "learning_rate": 3.793797591072631e-05,
-      "loss": 0.3694,
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.5028974924625986,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.3457,
       "step": 286
     },
     {
-      "epoch": 1.1619433198380567,
-      "grad_norm": 0.36889732845000706,
-      "learning_rate": 3.791289387885144e-05,
-      "loss": 0.3596,
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.6357976911009015,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.3808,
       "step": 287
     },
     {
-      "epoch": 1.165991902834008,
-      "grad_norm": 0.4283574867414021,
-      "learning_rate": 3.788766861595443e-05,
-      "loss": 0.3743,
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.6710094894371826,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.3614,
       "step": 288
     },
     {
-      "epoch": 1.1700404858299596,
-      "grad_norm": 0.4380167029524801,
-      "learning_rate": 3.7862300323735835e-05,
-      "loss": 0.3448,
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.6979652247242085,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.3539,
       "step": 289
     },
     {
-      "epoch": 1.174089068825911,
-      "grad_norm": 0.3914346762375412,
-      "learning_rate": 3.783678920503983e-05,
-      "loss": 0.3422,
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.6412796422091839,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.3616,
       "step": 290
     },
     {
-      "epoch": 1.1781376518218623,
-      "grad_norm": 0.40913413012130895,
-      "learning_rate": 3.781113546385265e-05,
-      "loss": 0.3281,
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.6288252839057961,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.3548,
       "step": 291
     },
     {
-      "epoch": 1.1821862348178138,
-      "grad_norm": 0.5257949316626421,
-      "learning_rate": 3.7785339305300916e-05,
-      "loss": 0.3954,
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.6303812700893479,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.3851,
       "step": 292
     },
     {
-      "epoch": 1.1862348178137652,
-      "grad_norm": 0.47292207767154426,
-      "learning_rate": 3.7759400935650025e-05,
-      "loss": 0.3489,
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.6443938531880744,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.3537,
       "step": 293
     },
     {
-      "epoch": 1.1902834008097165,
-      "grad_norm": 0.42827599319690396,
-      "learning_rate": 3.77333205623025e-05,
-      "loss": 0.3798,
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.6647476441531147,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.3654,
       "step": 294
     },
     {
-      "epoch": 1.194331983805668,
-      "grad_norm": 0.4506577364252486,
-      "learning_rate": 3.7707098393796295e-05,
-      "loss": 0.3848,
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.6651477281386762,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.3838,
       "step": 295
     },
     {
-      "epoch": 1.1983805668016194,
-      "grad_norm": 0.4319287009511947,
-      "learning_rate": 3.7680734639803185e-05,
-      "loss": 0.3686,
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.6822201752044151,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.3828,
       "step": 296
     },
     {
-      "epoch": 1.2024291497975708,
-      "grad_norm": 0.3644693733150136,
-      "learning_rate": 3.765422951112704e-05,
-      "loss": 0.3704,
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.7452462739950095,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.3819,
       "step": 297
     },
     {
-      "epoch": 1.2064777327935223,
-      "grad_norm": 0.43439878032811335,
-      "learning_rate": 3.762758321970216e-05,
-      "loss": 0.3657,
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.6734446509912946,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.3628,
       "step": 298
     },
     {
-      "epoch": 1.2105263157894737,
-      "grad_norm": 0.38752194815426527,
-      "learning_rate": 3.7600795978591584e-05,
-      "loss": 0.3419,
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.7842013804470751,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.3488,
       "step": 299
     },
     {
-      "epoch": 1.214574898785425,
-      "grad_norm": 0.47819350257242244,
-      "learning_rate": 3.757386800198538e-05,
-      "loss": 0.3668,
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.6564415347716268,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.4075,
       "step": 300
     },
     {
-      "epoch": 1.2186234817813766,
-      "grad_norm": 0.4036991893879403,
-      "learning_rate": 3.7546799505198925e-05,
-      "loss": 0.3719,
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.702215320335481,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.3397,
       "step": 301
     },
     {
-      "epoch": 1.222672064777328,
-      "grad_norm": 0.38777154347247267,
-      "learning_rate": 3.7519590704671197e-05,
-      "loss": 0.3549,
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.5996239900881898,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.3644,
       "step": 302
     },
     {
-      "epoch": 1.2267206477732793,
-      "grad_norm": 0.411459019226064,
-      "learning_rate": 3.749224181796305e-05,
-      "loss": 0.3605,
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7985092822728791,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.3533,
       "step": 303
     },
     {
-      "epoch": 1.2307692307692308,
-      "grad_norm": 0.4828452647381776,
-      "learning_rate": 3.746475306375545e-05,
-      "loss": 0.3848,
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.7108749109970185,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.4022,
       "step": 304
     },
     {
-      "epoch": 1.2348178137651822,
-      "grad_norm": 0.41798021065136404,
-      "learning_rate": 3.743712466184774e-05,
-      "loss": 0.3718,
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.7894282890259561,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.3889,
       "step": 305
     },
     {
-      "epoch": 1.2388663967611335,
-      "grad_norm": 0.4803895452073825,
-      "learning_rate": 3.7409356833155885e-05,
-      "loss": 0.3727,
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.5722671265684276,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.3852,
       "step": 306
     },
     {
-      "epoch": 1.242914979757085,
-      "grad_norm": 0.3978456750878927,
-      "learning_rate": 3.73814497997107e-05,
-      "loss": 0.3452,
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.7260180956667716,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.3486,
       "step": 307
     },
     {
-      "epoch": 1.2469635627530364,
-      "grad_norm": 0.5695255728316996,
-      "learning_rate": 3.7353403784656095e-05,
-      "loss": 0.3729,
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.6148201704838376,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.3872,
       "step": 308
     },
     {
-      "epoch": 1.2510121457489878,
-      "grad_norm": 0.38680067923299466,
-      "learning_rate": 3.732521901224724e-05,
-      "loss": 0.3357,
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.6973012911229917,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.3286,
       "step": 309
     },
     {
-      "epoch": 1.2550607287449393,
-      "grad_norm": 0.4646634581625683,
-      "learning_rate": 3.7296895707848844e-05,
-      "loss": 0.3928,
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.6927303295301698,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.4217,
       "step": 310
     },
     {
-      "epoch": 1.2591093117408907,
-      "grad_norm": 0.41467982758957717,
-      "learning_rate": 3.7268434097933275e-05,
-      "loss": 0.3711,
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.685915499159247,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.3805,
       "step": 311
     },
     {
-      "epoch": 1.263157894736842,
-      "grad_norm": 0.42641078575239855,
-      "learning_rate": 3.723983441007882e-05,
-      "loss": 0.3841,
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.6528232638481228,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.3799,
       "step": 312
     },
     {
-      "epoch": 1.2672064777327936,
-      "grad_norm": 0.41862045472521475,
-      "learning_rate": 3.721109687296781e-05,
-      "loss": 0.3798,
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.5294444785209557,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.3702,
       "step": 313
     },
     {
-      "epoch": 1.271255060728745,
-      "grad_norm": 0.5231729068306928,
-      "learning_rate": 3.718222171638484e-05,
-      "loss": 0.379,
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.6356293012245876,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4098,
       "step": 314
     },
     {
-      "epoch": 1.2753036437246963,
-      "grad_norm": 0.39462599979263513,
-      "learning_rate": 3.715320917121488e-05,
-      "loss": 0.3544,
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.6016229467266194,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.3505,
       "step": 315
     },
     {
-      "epoch": 1.2793522267206479,
-      "grad_norm": 0.4820600184900401,
-      "learning_rate": 3.712405946944149e-05,
-      "loss": 0.3547,
+      "epoch": 1.28,
+      "grad_norm": 0.7315418258500226,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.3742,
       "step": 316
     },
     {
-      "epoch": 1.2834008097165992,
-      "grad_norm": 0.4830654303285895,
-      "learning_rate": 3.709477284414488e-05,
-      "loss": 0.3664,
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.6410776599431915,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.3452,
       "step": 317
     },
     {
-      "epoch": 1.2874493927125505,
-      "grad_norm": 0.45203757525418714,
-      "learning_rate": 3.7065349529500136e-05,
-      "loss": 0.3512,
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.5691344411543786,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.3641,
       "step": 318
     },
     {
-      "epoch": 1.291497975708502,
-      "grad_norm": 0.5088739588325081,
-      "learning_rate": 3.7035789760775305e-05,
-      "loss": 0.3815,
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.754905396260898,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.3931,
       "step": 319
     },
     {
-      "epoch": 1.2955465587044535,
-      "grad_norm": 0.4286377837538106,
-      "learning_rate": 3.70060937743295e-05,
-      "loss": 0.3686,
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.6579048292188042,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.3689,
       "step": 320
     },
     {
-      "epoch": 1.2995951417004048,
-      "grad_norm": 0.4070637519776669,
-      "learning_rate": 3.6976261807611045e-05,
-      "loss": 0.3651,
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.5701287157677051,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.3685,
       "step": 321
     },
     {
-      "epoch": 1.3036437246963564,
-      "grad_norm": 0.4691212471507432,
-      "learning_rate": 3.694629409915555e-05,
-      "loss": 0.3628,
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.6370159192308493,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.3848,
       "step": 322
     },
     {
-      "epoch": 1.3076923076923077,
-      "grad_norm": 0.368296791964032,
-      "learning_rate": 3.691619088858401e-05,
-      "loss": 0.3692,
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.5027845963458968,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.3619,
       "step": 323
     },
     {
-      "epoch": 1.311740890688259,
-      "grad_norm": 0.44339205438699725,
-      "learning_rate": 3.68859524166009e-05,
-      "loss": 0.3643,
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.5194413979343931,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.3744,
       "step": 324
     },
     {
-      "epoch": 1.3157894736842106,
-      "grad_norm": 0.41754816303371967,
-      "learning_rate": 3.6855578924992226e-05,
-      "loss": 0.3623,
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.5077262308475304,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.3615,
       "step": 325
     },
     {
-      "epoch": 1.319838056680162,
-      "grad_norm": 0.3963970775401275,
-      "learning_rate": 3.6825070656623626e-05,
-      "loss": 0.3866,
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.5672673530659689,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.3958,
       "step": 326
     },
     {
-      "epoch": 1.3238866396761133,
-      "grad_norm": 0.4386009895850681,
-      "learning_rate": 3.67944278554384e-05,
-      "loss": 0.36,
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.6070969870187868,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.3406,
       "step": 327
     },
     {
-      "epoch": 1.3279352226720649,
-      "grad_norm": 0.4160419301418215,
-      "learning_rate": 3.676365076645557e-05,
-      "loss": 0.3673,
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.6333247815022708,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.3978,
       "step": 328
     },
     {
-      "epoch": 1.3319838056680162,
-      "grad_norm": 0.45214167488238494,
-      "learning_rate": 3.673273963576791e-05,
-      "loss": 0.3725,
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.5493541847755329,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.3642,
       "step": 329
     },
     {
-      "epoch": 1.3360323886639676,
-      "grad_norm": 0.43767518160488705,
-      "learning_rate": 3.670169471054001e-05,
-      "loss": 0.3622,
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.6470167702031546,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.3742,
       "step": 330
     },
     {
-      "epoch": 1.3400809716599191,
-      "grad_norm": 0.3964779759863151,
-      "learning_rate": 3.6670516239006254e-05,
-      "loss": 0.3727,
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.5723736586321807,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.4043,
       "step": 331
     },
     {
-      "epoch": 1.3441295546558705,
-      "grad_norm": 0.4381246515141964,
-      "learning_rate": 3.663920447046888e-05,
-      "loss": 0.3887,
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.670128455838125,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.3941,
       "step": 332
     },
     {
-      "epoch": 1.3481781376518218,
-      "grad_norm": 0.4821008848995016,
-      "learning_rate": 3.6607759655295954e-05,
-      "loss": 0.3711,
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.6389286930803958,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.3427,
       "step": 333
     },
     {
-      "epoch": 1.3522267206477734,
-      "grad_norm": 0.4365019890100636,
-      "learning_rate": 3.657618204491937e-05,
-      "loss": 0.3566,
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.813759126336383,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.3913,
       "step": 334
     },
     {
-      "epoch": 1.3562753036437247,
-      "grad_norm": 0.48299568239897367,
-      "learning_rate": 3.6544471891832875e-05,
-      "loss": 0.3784,
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.6648340562731355,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.3861,
       "step": 335
     },
     {
-      "epoch": 1.360323886639676,
-      "grad_norm": 0.4240112651710822,
-      "learning_rate": 3.6512629449589983e-05,
-      "loss": 0.3895,
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.6279827554457417,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.3686,
       "step": 336
     },
     {
-      "epoch": 1.3643724696356276,
-      "grad_norm": 0.4637631228288687,
-      "learning_rate": 3.6480654972802045e-05,
-      "loss": 0.3678,
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.5609390115470597,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.3885,
       "step": 337
     },
     {
-      "epoch": 1.368421052631579,
-      "grad_norm": 0.458950618180715,
-      "learning_rate": 3.644854871713611e-05,
-      "loss": 0.3678,
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.7131205506748883,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.3677,
       "step": 338
     },
     {
-      "epoch": 1.3724696356275303,
-      "grad_norm": 0.46585752804732244,
-      "learning_rate": 3.6416310939312924e-05,
-      "loss": 0.3815,
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.658020647764234,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.3691,
       "step": 339
     },
     {
-      "epoch": 1.376518218623482,
-      "grad_norm": 0.5175938535351634,
-      "learning_rate": 3.638394189710493e-05,
-      "loss": 0.3438,
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.5619366320421245,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.3714,
       "step": 340
     },
     {
-      "epoch": 1.3805668016194332,
-      "grad_norm": 0.44026284592125114,
-      "learning_rate": 3.635144184933412e-05,
-      "loss": 0.3867,
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.7581286623523374,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.3874,
       "step": 341
     },
     {
-      "epoch": 1.3846153846153846,
-      "grad_norm": 0.5580112348755426,
-      "learning_rate": 3.6318811055869994e-05,
-      "loss": 0.3551,
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.6338884085407034,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.3566,
       "step": 342
     },
     {
-      "epoch": 1.3886639676113361,
-      "grad_norm": 0.4544368210907052,
-      "learning_rate": 3.628604977762752e-05,
-      "loss": 0.3933,
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.7708176378333006,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.3872,
       "step": 343
     },
     {
-      "epoch": 1.3927125506072875,
-      "grad_norm": 0.5013773902390593,
-      "learning_rate": 3.6253158276565004e-05,
-      "loss": 0.363,
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.7423045083463986,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.3579,
       "step": 344
     },
     {
-      "epoch": 1.3967611336032388,
-      "grad_norm": 0.4179186478088046,
-      "learning_rate": 3.6220136815682016e-05,
-      "loss": 0.3629,
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.7162037125133675,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.3637,
       "step": 345
     },
     {
-      "epoch": 1.4008097165991904,
-      "grad_norm": 0.4471894486833656,
-      "learning_rate": 3.618698565901727e-05,
-      "loss": 0.359,
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.7602099116406451,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.3684,
       "step": 346
     },
     {
-      "epoch": 1.4048582995951417,
-      "grad_norm": 0.5124892118921237,
-      "learning_rate": 3.615370507164654e-05,
-      "loss": 0.3612,
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.7184879230750214,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.373,
       "step": 347
     },
     {
-      "epoch": 1.408906882591093,
-      "grad_norm": 0.390637139882488,
-      "learning_rate": 3.61202953196805e-05,
-      "loss": 0.3672,
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.848128338272079,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.3734,
       "step": 348
     },
     {
-      "epoch": 1.4129554655870447,
-      "grad_norm": 0.4679435155013486,
-      "learning_rate": 3.6086756670262654e-05,
-      "loss": 0.3755,
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.7263424805763969,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.3705,
       "step": 349
     },
     {
-      "epoch": 1.417004048582996,
-      "grad_norm": 0.4161105240826457,
-      "learning_rate": 3.6053089391567135e-05,
-      "loss": 0.3616,
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.7028777869669544,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.4049,
       "step": 350
     },
     {
-      "epoch": 1.4210526315789473,
-      "grad_norm": 0.3983900302405366,
-      "learning_rate": 3.601929375279662e-05,
-      "loss": 0.3577,
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.5045229808849552,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.3271,
       "step": 351
     },
     {
-      "epoch": 1.425101214574899,
-      "grad_norm": 0.42325859530948917,
-      "learning_rate": 3.598537002418012e-05,
-      "loss": 0.3864,
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.8549355187969817,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.4126,
       "step": 352
     },
     {
-      "epoch": 1.4291497975708503,
-      "grad_norm": 0.4334173136738058,
-      "learning_rate": 3.5951318476970885e-05,
-      "loss": 0.3893,
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.7727190655873146,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.392,
       "step": 353
     },
     {
-      "epoch": 1.4331983805668016,
-      "grad_norm": 0.40201638743002244,
-      "learning_rate": 3.591713938344416e-05,
-      "loss": 0.3698,
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.8237553535731215,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.3603,
       "step": 354
     },
     {
-      "epoch": 1.4372469635627532,
-      "grad_norm": 0.46390711281092667,
-      "learning_rate": 3.588283301689507e-05,
-      "loss": 0.3599,
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.8019415066924191,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.3989,
       "step": 355
     },
     {
-      "epoch": 1.4412955465587045,
-      "grad_norm": 0.5957345597578366,
-      "learning_rate": 3.5848399651636424e-05,
-      "loss": 0.3728,
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.9438078801936826,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.3572,
       "step": 356
     },
     {
-      "epoch": 1.4453441295546559,
-      "grad_norm": 0.41968941038000307,
-      "learning_rate": 3.5813839562996473e-05,
-      "loss": 0.3807,
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.7897761430190083,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.4,
       "step": 357
     },
     {
-      "epoch": 1.4493927125506074,
-      "grad_norm": 0.5244383927415321,
-      "learning_rate": 3.5779153027316775e-05,
-      "loss": 0.3891,
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.6925314962969537,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.3516,
       "step": 358
     },
     {
-      "epoch": 1.4534412955465588,
-      "grad_norm": 0.45655142246334673,
-      "learning_rate": 3.574434032194994e-05,
-      "loss": 0.3564,
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.9489998015024487,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.3822,
       "step": 359
     },
     {
-      "epoch": 1.45748987854251,
-      "grad_norm": 0.5246725930912993,
-      "learning_rate": 3.5709401725257436e-05,
-      "loss": 0.3604,
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.5624048306355566,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.3565,
       "step": 360
     },
     {
-      "epoch": 1.4615384615384617,
-      "grad_norm": 0.39744066661771876,
-      "learning_rate": 3.5674337516607346e-05,
-      "loss": 0.3675,
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.8280906844224754,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.3874,
       "step": 361
     },
     {
-      "epoch": 1.465587044534413,
-      "grad_norm": 0.5496954529045435,
-      "learning_rate": 3.563914797637216e-05,
-      "loss": 0.3733,
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.6828403594096736,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.3693,
       "step": 362
     },
     {
-      "epoch": 1.4696356275303644,
-      "grad_norm": 0.4442957389769822,
-      "learning_rate": 3.560383338592649e-05,
-      "loss": 0.3711,
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.7578497815337718,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.3844,
       "step": 363
     },
     {
-      "epoch": 1.4736842105263157,
-      "grad_norm": 0.4796740188616433,
-      "learning_rate": 3.556839402764487e-05,
-      "loss": 0.3639,
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.7470592093279689,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.353,
       "step": 364
     },
     {
-      "epoch": 1.4777327935222673,
-      "grad_norm": 0.5979812771700712,
-      "learning_rate": 3.553283018489946e-05,
-      "loss": 0.3584,
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.6944291210616202,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.3525,
       "step": 365
     },
     {
-      "epoch": 1.4817813765182186,
-      "grad_norm": 0.4623317140434867,
-      "learning_rate": 3.54971421420578e-05,
-      "loss": 0.3731,
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.6942545685247483,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.3836,
       "step": 366
     },
     {
-      "epoch": 1.48582995951417,
-      "grad_norm": 0.5467750134612158,
-      "learning_rate": 3.5461330184480524e-05,
-      "loss": 0.3761,
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.8267553019390819,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.3968,
       "step": 367
     },
     {
-      "epoch": 1.4898785425101215,
-      "grad_norm": 0.44136891949047574,
-      "learning_rate": 3.54253945985191e-05,
-      "loss": 0.3773,
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.6506326819719855,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.3929,
       "step": 368
     },
     {
-      "epoch": 1.4939271255060729,
-      "grad_norm": 0.4761739844965191,
-      "learning_rate": 3.5389335671513524e-05,
-      "loss": 0.3865,
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.639686441506425,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.3706,
       "step": 369
     },
     {
-      "epoch": 1.4979757085020242,
-      "grad_norm": 0.35791969788357353,
-      "learning_rate": 3.5353153691789986e-05,
-      "loss": 0.373,
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.5311097733277504,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.3706,
       "step": 370
     },
     {
-      "epoch": 1.5020242914979756,
-      "grad_norm": 0.46712247568300874,
-      "learning_rate": 3.531684894865866e-05,
-      "loss": 0.3775,
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.7085498526182703,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.3912,
       "step": 371
     },
     {
-      "epoch": 1.5060728744939271,
-      "grad_norm": 0.37244052956051427,
-      "learning_rate": 3.528042173241128e-05,
-      "loss": 0.3576,
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.4762416495602507,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.3287,
       "step": 372
     },
     {
-      "epoch": 1.5101214574898787,
-      "grad_norm": 0.4530900596025655,
-      "learning_rate": 3.524387233431892e-05,
-      "loss": 0.3688,
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.7116776216671409,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.379,
       "step": 373
     },
     {
-      "epoch": 1.5141700404858298,
-      "grad_norm": 0.35665597762748263,
-      "learning_rate": 3.520720104662958e-05,
-      "loss": 0.3813,
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.5518904428559207,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.3965,
       "step": 374
     },
     {
-      "epoch": 1.5182186234817814,
-      "grad_norm": 0.4751809273340378,
-      "learning_rate": 3.5170408162565904e-05,
-      "loss": 0.3792,
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6921137434829213,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.3822,
       "step": 375
     },
     {
-      "epoch": 1.522267206477733,
-      "grad_norm": 0.34243088673677863,
-      "learning_rate": 3.513349397632281e-05,
-      "loss": 0.3817,
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.5122919199591476,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.3906,
       "step": 376
     },
     {
-      "epoch": 1.526315789473684,
-      "grad_norm": 0.49040128029250774,
-      "learning_rate": 3.5096458783065145e-05,
-      "loss": 0.3626,
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.6211698991267575,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.3651,
       "step": 377
     },
     {
-      "epoch": 1.5303643724696356,
-      "grad_norm": 0.38244269102640255,
-      "learning_rate": 3.505930287892533e-05,
-      "loss": 0.3639,
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.5455202404064395,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.3936,
       "step": 378
     },
     {
-      "epoch": 1.5344129554655872,
-      "grad_norm": 0.41694419834718793,
-      "learning_rate": 3.502202656100099e-05,
-      "loss": 0.3481,
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.48537962197966916,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.3277,
       "step": 379
     },
     {
-      "epoch": 1.5384615384615383,
-      "grad_norm": 0.43961477591565246,
-      "learning_rate": 3.498463012735256e-05,
-      "loss": 0.3602,
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.6064638020338912,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.379,
       "step": 380
     },
     {
-      "epoch": 1.54251012145749,
-      "grad_norm": 0.3978737125816165,
-      "learning_rate": 3.494711387700094e-05,
-      "loss": 0.3558,
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.5513304792733938,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.3638,
       "step": 381
     },
     {
-      "epoch": 1.5465587044534415,
-      "grad_norm": 0.41652203191136894,
-      "learning_rate": 3.490947810992505e-05,
-      "loss": 0.3549,
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.5431331472577596,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.3599,
       "step": 382
     },
     {
-      "epoch": 1.5506072874493926,
-      "grad_norm": 0.43242563404370693,
-      "learning_rate": 3.48717231270595e-05,
-      "loss": 0.3731,
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.6230431545056061,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.3694,
       "step": 383
     },
     {
-      "epoch": 1.5546558704453441,
-      "grad_norm": 0.4062881151095278,
-      "learning_rate": 3.483384923029211e-05,
-      "loss": 0.3524,
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.5052767800249827,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.3637,
       "step": 384
     },
     {
-      "epoch": 1.5587044534412957,
-      "grad_norm": 0.4088934447468767,
-      "learning_rate": 3.479585672246156e-05,
-      "loss": 0.3725,
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6046185739426602,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.4046,
       "step": 385
     },
     {
-      "epoch": 1.5627530364372468,
-      "grad_norm": 0.3882814921036605,
-      "learning_rate": 3.4757745907354924e-05,
-      "loss": 0.3824,
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6198513569733549,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.3566,
       "step": 386
     },
     {
-      "epoch": 1.5668016194331984,
-      "grad_norm": 0.40811454049876333,
-      "learning_rate": 3.471951708970527e-05,
-      "loss": 0.3562,
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.5410715231452782,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.3563,
       "step": 387
     },
     {
-      "epoch": 1.5708502024291497,
-      "grad_norm": 0.43203597041258757,
-      "learning_rate": 3.468117057518921e-05,
-      "loss": 0.3812,
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6723255398222208,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.3915,
       "step": 388
     },
     {
-      "epoch": 1.574898785425101,
-      "grad_norm": 0.4698189782604624,
-      "learning_rate": 3.4642706670424426e-05,
-      "loss": 0.3602,
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.5618668941118388,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.3815,
       "step": 389
     },
     {
-      "epoch": 1.5789473684210527,
-      "grad_norm": 0.46962844527211844,
-      "learning_rate": 3.460412568296731e-05,
-      "loss": 0.399,
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.604309103743189,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.3909,
       "step": 390
     },
     {
-      "epoch": 1.582995951417004,
-      "grad_norm": 0.36069737573445115,
-      "learning_rate": 3.456542792131039e-05,
-      "loss": 0.3636,
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.5545189079850223,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.3598,
       "step": 391
     },
     {
-      "epoch": 1.5870445344129553,
-      "grad_norm": 0.43743417594025247,
-      "learning_rate": 3.4526613694879936e-05,
-      "loss": 0.4029,
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.6295926777135528,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.4072,
       "step": 392
     },
     {
-      "epoch": 1.591093117408907,
-      "grad_norm": 0.3317773024221916,
-      "learning_rate": 3.448768331403348e-05,
-      "loss": 0.3576,
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.5846643270127403,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.3795,
       "step": 393
     },
     {
-      "epoch": 1.5951417004048583,
-      "grad_norm": 0.40404924356564276,
-      "learning_rate": 3.44486370900573e-05,
-      "loss": 0.364,
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.6059541198640588,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.3884,
       "step": 394
     },
     {
-      "epoch": 1.5991902834008096,
-      "grad_norm": 0.3542452134700563,
-      "learning_rate": 3.440947533516396e-05,
-      "loss": 0.3684,
+      "epoch": 1.6,
+      "grad_norm": 0.6089980662317442,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.383,
       "step": 395
     },
     {
-      "epoch": 1.6032388663967612,
-      "grad_norm": 0.46704044487870117,
-      "learning_rate": 3.4370198362489816e-05,
-      "loss": 0.3775,
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.5699557429027935,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.3541,
       "step": 396
     },
     {
-      "epoch": 1.6072874493927125,
-      "grad_norm": 0.3915755719703622,
-      "learning_rate": 3.43308064860925e-05,
-      "loss": 0.3643,
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.5845858988338071,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.3957,
       "step": 397
     },
     {
-      "epoch": 1.6113360323886639,
-      "grad_norm": 0.3786883537278133,
-      "learning_rate": 3.4291300020948393e-05,
-      "loss": 0.3582,
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.5775730383175001,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.368,
       "step": 398
     },
     {
-      "epoch": 1.6153846153846154,
-      "grad_norm": 0.43624156629921673,
-      "learning_rate": 3.425167928295015e-05,
-      "loss": 0.3806,
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.4680672894451932,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.3495,
       "step": 399
     },
     {
-      "epoch": 1.6194331983805668,
-      "grad_norm": 0.37966431225214153,
-      "learning_rate": 3.421194458890411e-05,
-      "loss": 0.3853,
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6379508805745,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.4089,
       "step": 400
     },
     {
-      "epoch": 1.623481781376518,
-      "grad_norm": 0.45304637024823796,
-      "learning_rate": 3.4172096256527865e-05,
-      "loss": 0.3698,
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.5100733786994194,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.3719,
       "step": 401
     },
     {
-      "epoch": 1.6275303643724697,
-      "grad_norm": 0.42410436633410337,
-      "learning_rate": 3.4132134604447595e-05,
-      "loss": 0.3766,
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.5765699750578449,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.3633,
       "step": 402
     },
     {
-      "epoch": 1.631578947368421,
-      "grad_norm": 0.49449171281212123,
-      "learning_rate": 3.409205995219561e-05,
-      "loss": 0.3604,
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.47710084467917785,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.3641,
       "step": 403
     },
     {
-      "epoch": 1.6356275303643724,
-      "grad_norm": 0.43765583433988225,
-      "learning_rate": 3.4051872620207765e-05,
-      "loss": 0.4008,
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6663263071231945,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.4089,
       "step": 404
     },
     {
-      "epoch": 1.639676113360324,
-      "grad_norm": 0.4660814129603057,
-      "learning_rate": 3.40115729298209e-05,
-      "loss": 0.376,
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.5397536860894295,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.3775,
       "step": 405
     },
     {
-      "epoch": 1.6437246963562753,
-      "grad_norm": 0.38042567063839056,
-      "learning_rate": 3.397116120327025e-05,
-      "loss": 0.3646,
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.5295626494837684,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.3676,
       "step": 406
     },
     {
-      "epoch": 1.6477732793522266,
-      "grad_norm": 0.3962203290785257,
-      "learning_rate": 3.3930637763686916e-05,
-      "loss": 0.3697,
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.5731544053759818,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.382,
       "step": 407
     },
     {
-      "epoch": 1.6518218623481782,
-      "grad_norm": 0.4030802551586152,
-      "learning_rate": 3.389000293509524e-05,
-      "loss": 0.3841,
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.578332213328555,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.3726,
       "step": 408
     },
     {
-      "epoch": 1.6558704453441295,
-      "grad_norm": 0.4231504426559868,
-      "learning_rate": 3.3849257042410225e-05,
-      "loss": 0.3899,
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.4950916022540452,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.3955,
       "step": 409
     },
     {
-      "epoch": 1.6599190283400809,
-      "grad_norm": 0.36309827737962996,
-      "learning_rate": 3.380840041143494e-05,
-      "loss": 0.3486,
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.5097314345822723,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.3673,
       "step": 410
     },
     {
-      "epoch": 1.6639676113360324,
-      "grad_norm": 0.4893334136860676,
-      "learning_rate": 3.376743336885792e-05,
-      "loss": 0.3801,
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.5936361324057824,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.4053,
       "step": 411
     },
     {
-      "epoch": 1.6680161943319838,
-      "grad_norm": 0.34731667536725275,
-      "learning_rate": 3.372635624225055e-05,
-      "loss": 0.3706,
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.4857911033248024,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.3531,
       "step": 412
     },
     {
-      "epoch": 1.6720647773279351,
-      "grad_norm": 0.438220261768144,
-      "learning_rate": 3.3685169360064445e-05,
-      "loss": 0.3701,
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.5260042077488933,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.3856,
       "step": 413
     },
     {
-      "epoch": 1.6761133603238867,
-      "grad_norm": 0.3197101884181713,
-      "learning_rate": 3.3643873051628804e-05,
-      "loss": 0.3788,
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.5417361001125799,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.3729,
       "step": 414
     },
     {
-      "epoch": 1.680161943319838,
-      "grad_norm": 0.4650622614174161,
-      "learning_rate": 3.3602467647147814e-05,
-      "loss": 0.3891,
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.5814360675306575,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.3739,
       "step": 415
     },
     {
-      "epoch": 1.6842105263157894,
-      "grad_norm": 0.41626629651448616,
-      "learning_rate": 3.356095347769799e-05,
-      "loss": 0.3636,
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.47000010429312095,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.3682,
       "step": 416
     },
     {
-      "epoch": 1.688259109311741,
-      "grad_norm": 0.43018252886725533,
-      "learning_rate": 3.351933087522552e-05,
-      "loss": 0.3734,
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.6468705930036914,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.4003,
       "step": 417
     },
     {
-      "epoch": 1.6923076923076923,
-      "grad_norm": 0.4138229732407728,
-      "learning_rate": 3.3477600172543626e-05,
-      "loss": 0.363,
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5067219971347707,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.366,
       "step": 418
     },
     {
-      "epoch": 1.6963562753036436,
-      "grad_norm": 0.41299201352664794,
-      "learning_rate": 3.3435761703329894e-05,
-      "loss": 0.3815,
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.47629712467535473,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.351,
       "step": 419
     },
     {
-      "epoch": 1.7004048582995952,
-      "grad_norm": 0.42303317087843817,
-      "learning_rate": 3.3393815802123606e-05,
-      "loss": 0.3623,
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.5523034974776373,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.3948,
       "step": 420
     },
     {
-      "epoch": 1.7044534412955465,
-      "grad_norm": 0.3928354358506378,
-      "learning_rate": 3.335176280432308e-05,
-      "loss": 0.3885,
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.489811047204724,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.3901,
       "step": 421
     },
     {
-      "epoch": 1.708502024291498,
-      "grad_norm": 0.48751075720338827,
-      "learning_rate": 3.330960304618295e-05,
-      "loss": 0.358,
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.5342996834213659,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.3459,
       "step": 422
     },
     {
-      "epoch": 1.7125506072874495,
-      "grad_norm": 0.38509346248918064,
-      "learning_rate": 3.326733686481153e-05,
-      "loss": 0.3755,
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.5431926630102648,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.3796,
       "step": 423
     },
     {
-      "epoch": 1.7165991902834008,
-      "grad_norm": 0.4440007618458125,
-      "learning_rate": 3.3224964598168084e-05,
-      "loss": 0.3629,
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.4559862541955908,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.3526,
       "step": 424
     },
     {
-      "epoch": 1.7206477732793521,
-      "grad_norm": 0.3482674671718291,
-      "learning_rate": 3.318248658506012e-05,
-      "loss": 0.358,
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.5797114411110708,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.3888,
       "step": 425
     },
     {
-      "epoch": 1.7246963562753037,
-      "grad_norm": 0.47068145324224236,
-      "learning_rate": 3.313990316514069e-05,
-      "loss": 0.3822,
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.5043779907174283,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.3837,
       "step": 426
     },
     {
-      "epoch": 1.728744939271255,
-      "grad_norm": 0.38662345009697097,
-      "learning_rate": 3.309721467890571e-05,
-      "loss": 0.3601,
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.6845579563626579,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.4108,
       "step": 427
     },
     {
-      "epoch": 1.7327935222672064,
-      "grad_norm": 0.4422478897566049,
-      "learning_rate": 3.305442146769115e-05,
-      "loss": 0.3695,
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.49762999420579973,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.3326,
       "step": 428
     },
     {
-      "epoch": 1.736842105263158,
-      "grad_norm": 0.42756034233926343,
-      "learning_rate": 3.30115238736704e-05,
-      "loss": 0.3764,
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.5891402247247638,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.3746,
       "step": 429
     },
     {
-      "epoch": 1.7408906882591093,
-      "grad_norm": 0.3971096287470518,
-      "learning_rate": 3.296852223985148e-05,
-      "loss": 0.3498,
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.5140421265435756,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.3657,
       "step": 430
     },
     {
-      "epoch": 1.7449392712550607,
-      "grad_norm": 0.47349873661077724,
-      "learning_rate": 3.2925416910074315e-05,
-      "loss": 0.3679,
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.5334256252935604,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.3651,
       "step": 431
     },
     {
-      "epoch": 1.7489878542510122,
-      "grad_norm": 0.35065803480972013,
-      "learning_rate": 3.288220822900796e-05,
-      "loss": 0.3824,
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.531139981740974,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.3677,
       "step": 432
     },
     {
-      "epoch": 1.7530364372469636,
-      "grad_norm": 0.4755758677448353,
-      "learning_rate": 3.283889654214788e-05,
-      "loss": 0.3672,
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.4405552493678426,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.3791,
       "step": 433
     },
     {
-      "epoch": 1.757085020242915,
-      "grad_norm": 0.41114392955695794,
-      "learning_rate": 3.279548219581317e-05,
-      "loss": 0.3736,
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.5483242106349978,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.3916,
       "step": 434
     },
     {
-      "epoch": 1.7611336032388665,
-      "grad_norm": 0.44520014888579834,
-      "learning_rate": 3.275196553714379e-05,
-      "loss": 0.3602,
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.4397977629042849,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.3651,
       "step": 435
     },
     {
-      "epoch": 1.7651821862348178,
-      "grad_norm": 0.5006474124990422,
-      "learning_rate": 3.270834691409778e-05,
-      "loss": 0.3752,
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.5198827273321849,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.3758,
       "step": 436
     },
     {
-      "epoch": 1.7692307692307692,
-      "grad_norm": 0.36605039677363393,
-      "learning_rate": 3.266462667544849e-05,
-      "loss": 0.3711,
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.5188965785556219,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.373,
       "step": 437
     },
     {
-      "epoch": 1.7732793522267207,
-      "grad_norm": 0.48853560644108285,
-      "learning_rate": 3.262080517078178e-05,
-      "loss": 0.3771,
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.573604526321178,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.3758,
       "step": 438
     },
     {
-      "epoch": 1.777327935222672,
-      "grad_norm": 0.5060508164862538,
-      "learning_rate": 3.257688275049323e-05,
-      "loss": 0.3662,
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.4904091276181341,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.3817,
       "step": 439
     },
     {
-      "epoch": 1.7813765182186234,
-      "grad_norm": 0.3974205791721758,
-      "learning_rate": 3.253285976578535e-05,
-      "loss": 0.3908,
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.6418551864783973,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.4068,
       "step": 440
     },
     {
-      "epoch": 1.785425101214575,
-      "grad_norm": 0.5112296991785555,
-      "learning_rate": 3.2488736568664756e-05,
-      "loss": 0.375,
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.5716766604932132,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.3759,
       "step": 441
     },
     {
-      "epoch": 1.7894736842105263,
-      "grad_norm": 0.46689786860900323,
-      "learning_rate": 3.244451351193935e-05,
-      "loss": 0.382,
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.6256480959774622,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.3779,
       "step": 442
     },
     {
-      "epoch": 1.7935222672064777,
-      "grad_norm": 0.5250067066705535,
-      "learning_rate": 3.240019094921551e-05,
-      "loss": 0.3665,
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.7870799910202776,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.3708,
       "step": 443
     },
     {
-      "epoch": 1.7975708502024292,
-      "grad_norm": 0.6358937133928813,
-      "learning_rate": 3.235576923489527e-05,
-      "loss": 0.3742,
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.5591004426216337,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.3547,
       "step": 444
     },
     {
-      "epoch": 1.8016194331983806,
-      "grad_norm": 0.33977913769200396,
-      "learning_rate": 3.2311248724173476e-05,
-      "loss": 0.3647,
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.7442181870529674,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.4077,
       "step": 445
     },
     {
-      "epoch": 1.805668016194332,
-      "grad_norm": 0.6195469853411087,
-      "learning_rate": 3.226662977303494e-05,
-      "loss": 0.3812,
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.5071374600495145,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.4163,
       "step": 446
     },
     {
-      "epoch": 1.8097165991902835,
-      "grad_norm": 0.5129892203665206,
-      "learning_rate": 3.2221912738251626e-05,
-      "loss": 0.3822,
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.577473946625859,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.3805,
       "step": 447
     },
     {
-      "epoch": 1.8137651821862348,
-      "grad_norm": 0.4571620013409168,
-      "learning_rate": 3.2177097977379736e-05,
-      "loss": 0.3847,
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.5213570473780911,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.3534,
       "step": 448
     },
     {
-      "epoch": 1.8178137651821862,
-      "grad_norm": 0.48832915043039343,
-      "learning_rate": 3.21321858487569e-05,
-      "loss": 0.3669,
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.4911233633145584,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.3666,
       "step": 449
     },
     {
-      "epoch": 1.8218623481781377,
-      "grad_norm": 0.376788778442922,
-      "learning_rate": 3.208717671149931e-05,
-      "loss": 0.3708,
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.5714084857494942,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.3778,
       "step": 450
     },
     {
-      "epoch": 1.825910931174089,
-      "grad_norm": 0.4453142432423007,
-      "learning_rate": 3.204207092549882e-05,
-      "loss": 0.3663,
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.4711721972964105,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.3676,
       "step": 451
     },
     {
-      "epoch": 1.8299595141700404,
-      "grad_norm": 0.3951103208610028,
-      "learning_rate": 3.19968688514201e-05,
-      "loss": 0.374,
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.5972918685466335,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.3839,
       "step": 452
     },
     {
-      "epoch": 1.834008097165992,
-      "grad_norm": 0.3927483238840978,
-      "learning_rate": 3.195157085069774e-05,
-      "loss": 0.3595,
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.6231327543317907,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.3515,
       "step": 453
     },
     {
-      "epoch": 1.8380566801619433,
-      "grad_norm": 0.4230975743106898,
-      "learning_rate": 3.190617728553332e-05,
-      "loss": 0.3617,
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.5365055918467982,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.3946,
       "step": 454
     },
     {
-      "epoch": 1.8421052631578947,
-      "grad_norm": 0.33148310549363164,
-      "learning_rate": 3.1860688518892606e-05,
-      "loss": 0.3574,
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.5183045687230502,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.3508,
       "step": 455
     },
     {
-      "epoch": 1.8461538461538463,
-      "grad_norm": 0.44859310814151526,
-      "learning_rate": 3.181510491450254e-05,
-      "loss": 0.3667,
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.5709262199011997,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.3624,
       "step": 456
     },
     {
-      "epoch": 1.8502024291497976,
-      "grad_norm": 0.4253243203588643,
-      "learning_rate": 3.176942683684842e-05,
-      "loss": 0.3585,
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.5456976038511028,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.3574,
       "step": 457
     },
     {
-      "epoch": 1.854251012145749,
-      "grad_norm": 0.3769380943781389,
-      "learning_rate": 3.1723654651170934e-05,
-      "loss": 0.3601,
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.5554342270306488,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.3796,
       "step": 458
     },
     {
-      "epoch": 1.8582995951417005,
-      "grad_norm": 0.5403073581782615,
-      "learning_rate": 3.167778872346327e-05,
-      "loss": 0.3605,
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.6423204438691494,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.3994,
       "step": 459
     },
     {
-      "epoch": 1.8623481781376519,
-      "grad_norm": 0.375829960050978,
-      "learning_rate": 3.163182942046815e-05,
-      "loss": 0.3692,
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.46149449426401207,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.3513,
       "step": 460
     },
     {
-      "epoch": 1.8663967611336032,
-      "grad_norm": 0.5050553523718886,
-      "learning_rate": 3.158577710967494e-05,
-      "loss": 0.346,
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.5725559475340252,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.3485,
       "step": 461
     },
     {
-      "epoch": 1.8704453441295548,
-      "grad_norm": 0.3610357947818454,
-      "learning_rate": 3.153963215931669e-05,
-      "loss": 0.3597,
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.47013469438249095,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.3594,
       "step": 462
     },
     {
-      "epoch": 1.874493927125506,
-      "grad_norm": 0.4690872005317614,
-      "learning_rate": 3.14933949383672e-05,
-      "loss": 0.3774,
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.4516902661278115,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.3806,
       "step": 463
     },
     {
-      "epoch": 1.8785425101214575,
-      "grad_norm": 0.38242584506495775,
-      "learning_rate": 3.144706581653805e-05,
-      "loss": 0.3661,
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.49313863292068383,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.3671,
       "step": 464
     },
     {
-      "epoch": 1.882591093117409,
-      "grad_norm": 0.5253338242180909,
-      "learning_rate": 3.140064516427566e-05,
-      "loss": 0.3615,
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.48658193203203526,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.3667,
       "step": 465
     },
     {
-      "epoch": 1.8866396761133604,
-      "grad_norm": 0.3636936504763626,
-      "learning_rate": 3.135413335275833e-05,
-      "loss": 0.3729,
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.4952997422151929,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.3733,
       "step": 466
     },
     {
-      "epoch": 1.8906882591093117,
-      "grad_norm": 0.5748350699701886,
-      "learning_rate": 3.130753075389327e-05,
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.5096746521040342,
+      "learning_rate": 3.120267288911952e-05,
       "loss": 0.3595,
       "step": 467
     },
     {
-      "epoch": 1.8947368421052633,
-      "grad_norm": 0.3816058435232405,
-      "learning_rate": 3.1260837740313625e-05,
-      "loss": 0.3773,
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.5574979513032707,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.3915,
       "step": 468
     },
     {
-      "epoch": 1.8987854251012146,
-      "grad_norm": 0.5555740426039257,
-      "learning_rate": 3.121405468537547e-05,
-      "loss": 0.3798,
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.49870863628648027,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.3648,
       "step": 469
     },
     {
-      "epoch": 1.902834008097166,
-      "grad_norm": 0.3887998527575953,
-      "learning_rate": 3.116718196315488e-05,
-      "loss": 0.3449,
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6025653444308415,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.3731,
       "step": 470
     },
     {
-      "epoch": 1.9068825910931175,
-      "grad_norm": 0.4240183166899678,
-      "learning_rate": 3.1120219948444894e-05,
-      "loss": 0.3579,
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.46770279907822626,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.3678,
       "step": 471
     },
     {
-      "epoch": 1.9109311740890689,
-      "grad_norm": 0.40834442146265615,
-      "learning_rate": 3.107316901675252e-05,
-      "loss": 0.3815,
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.62141502625386,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.3743,
       "step": 472
     },
     {
-      "epoch": 1.9149797570850202,
-      "grad_norm": 0.3743423640461393,
-      "learning_rate": 3.1026029544295745e-05,
-      "loss": 0.3669,
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.5077986967531527,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.3638,
       "step": 473
     },
     {
-      "epoch": 1.9190283400809718,
-      "grad_norm": 0.3815907345912122,
-      "learning_rate": 3.097880190800057e-05,
-      "loss": 0.3644,
+      "epoch": 1.92,
+      "grad_norm": 0.5627789947841831,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.3786,
       "step": 474
     },
     {
-      "epoch": 1.9230769230769231,
-      "grad_norm": 0.3458941547528241,
-      "learning_rate": 3.093148648549788e-05,
-      "loss": 0.3758,
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.6050096071798715,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.3992,
       "step": 475
     },
     {
-      "epoch": 1.9271255060728745,
-      "grad_norm": 0.35081801277943647,
-      "learning_rate": 3.088408365512055e-05,
-      "loss": 0.3839,
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.6721391729061993,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.3538,
       "step": 476
     },
     {
-      "epoch": 1.931174089068826,
-      "grad_norm": 0.3518915023552267,
-      "learning_rate": 3.083659379590034e-05,
-      "loss": 0.3616,
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.5999797243621217,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.3788,
       "step": 477
     },
     {
-      "epoch": 1.9352226720647774,
-      "grad_norm": 0.32252928390215596,
-      "learning_rate": 3.0789017287564913e-05,
-      "loss": 0.3589,
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.6986835565894822,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.3887,
       "step": 478
     },
     {
-      "epoch": 1.9392712550607287,
-      "grad_norm": 0.35819669122568826,
-      "learning_rate": 3.074135451053475e-05,
-      "loss": 0.3852,
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.5378988215362247,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.3505,
       "step": 479
     },
     {
-      "epoch": 1.9433198380566803,
-      "grad_norm": 0.37959145407217204,
-      "learning_rate": 3.069360584592015e-05,
-      "loss": 0.3655,
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.6035634994875964,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.3797,
       "step": 480
     },
     {
-      "epoch": 1.9473684210526314,
-      "grad_norm": 0.34615446582216836,
-      "learning_rate": 3.064577167551816e-05,
-      "loss": 0.3815,
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5328491648119086,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.3885,
       "step": 481
     },
     {
-      "epoch": 1.951417004048583,
-      "grad_norm": 0.3791099512204315,
-      "learning_rate": 3.059785238180953e-05,
-      "loss": 0.3721,
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.5901491018290067,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.3882,
       "step": 482
     },
     {
-      "epoch": 1.9554655870445345,
-      "grad_norm": 0.34705959482538684,
-      "learning_rate": 3.054984834795566e-05,
-      "loss": 0.381,
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.4547757655530949,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.3899,
       "step": 483
     },
     {
-      "epoch": 1.9595141700404857,
-      "grad_norm": 0.33469864689404655,
-      "learning_rate": 3.0501759957795537e-05,
-      "loss": 0.3655,
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.43725575077592166,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.368,
       "step": 484
     },
     {
-      "epoch": 1.9635627530364372,
-      "grad_norm": 0.3672991806335059,
-      "learning_rate": 3.0453587595842627e-05,
-      "loss": 0.3672,
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.5297611626390727,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.3779,
       "step": 485
     },
     {
-      "epoch": 1.9676113360323888,
-      "grad_norm": 0.3721309966722606,
-      "learning_rate": 3.0405331647281856e-05,
-      "loss": 0.3884,
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.477166338795374,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.3949,
       "step": 486
     },
     {
-      "epoch": 1.97165991902834,
-      "grad_norm": 0.395400960555198,
-      "learning_rate": 3.0356992497966508e-05,
-      "loss": 0.3553,
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.5106969175175945,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.3454,
       "step": 487
     },
     {
-      "epoch": 1.9757085020242915,
-      "grad_norm": 0.3487267244308189,
-      "learning_rate": 3.0308570534415137e-05,
-      "loss": 0.3587,
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.5237675224403795,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.3932,
       "step": 488
     },
     {
-      "epoch": 1.979757085020243,
-      "grad_norm": 0.36177444137006537,
-      "learning_rate": 3.0260066143808467e-05,
-      "loss": 0.3882,
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.5018250454888903,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.353,
       "step": 489
     },
     {
-      "epoch": 1.9838056680161942,
-      "grad_norm": 0.38629318842931215,
-      "learning_rate": 3.021147971398632e-05,
-      "loss": 0.3665,
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.4563181603833332,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.3651,
       "step": 490
     },
     {
-      "epoch": 1.9878542510121457,
-      "grad_norm": 0.3249711394012302,
-      "learning_rate": 3.0162811633444478e-05,
-      "loss": 0.3703,
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.5811080663473511,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.402,
       "step": 491
     },
     {
-      "epoch": 1.9919028340080973,
-      "grad_norm": 0.37662581597787714,
-      "learning_rate": 3.0114062291331635e-05,
-      "loss": 0.3637,
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.4382085751836705,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3596,
       "step": 492
     },
     {
-      "epoch": 1.9959514170040484,
-      "grad_norm": 0.3645202288930804,
-      "learning_rate": 3.0065232077446217e-05,
-      "loss": 0.3566,
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.4922053397427758,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.3779,
       "step": 493
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.3905445088482649,
-      "learning_rate": 3.001632138223332e-05,
-      "loss": 0.3689,
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.593093739525878,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3756,
       "step": 494
     },
     {
-      "epoch": 2.0040485829959516,
-      "grad_norm": 0.4547137742572888,
-      "learning_rate": 2.9967330596781565e-05,
-      "loss": 0.2874,
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.5538959854870248,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3027,
       "step": 495
     },
     {
-      "epoch": 2.0080971659919027,
-      "grad_norm": 0.4702707650715754,
-      "learning_rate": 2.9918260112819952e-05,
-      "loss": 0.2869,
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.673865705032606,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.2929,
       "step": 496
     },
     {
-      "epoch": 2.0121457489878543,
-      "grad_norm": 0.5363224425182969,
-      "learning_rate": 2.9869110322714778e-05,
-      "loss": 0.3077,
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.5038399335144141,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.2659,
       "step": 497
     },
     {
-      "epoch": 2.016194331983806,
-      "grad_norm": 0.414706201658724,
-      "learning_rate": 2.9819881619466447e-05,
-      "loss": 0.2818,
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.48885097579864534,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.2897,
       "step": 498
     },
     {
-      "epoch": 2.020242914979757,
-      "grad_norm": 0.4297620257015429,
-      "learning_rate": 2.9770574396706364e-05,
-      "loss": 0.2833,
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.6080462027271569,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3093,
       "step": 499
     },
     {
-      "epoch": 2.0242914979757085,
-      "grad_norm": 0.36652762376651077,
-      "learning_rate": 2.972118904869376e-05,
-      "loss": 0.2726,
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.5384879099143883,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.272,
       "step": 500
     },
     {
-      "epoch": 2.02834008097166,
-      "grad_norm": 0.3878761163568221,
-      "learning_rate": 2.9671725970312563e-05,
-      "loss": 0.268,
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.44079217583128627,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.272,
       "step": 501
     },
     {
-      "epoch": 2.032388663967611,
-      "grad_norm": 0.45808129763948874,
-      "learning_rate": 2.9622185557068238e-05,
-      "loss": 0.2925,
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.5497734893916537,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3098,
       "step": 502
     },
     {
-      "epoch": 2.0364372469635628,
-      "grad_norm": 0.411628469074801,
-      "learning_rate": 2.95725682050846e-05,
-      "loss": 0.2716,
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.39209424482394517,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.2455,
       "step": 503
     },
     {
-      "epoch": 2.0404858299595143,
-      "grad_norm": 0.4295089664372783,
-      "learning_rate": 2.9522874311100684e-05,
-      "loss": 0.2812,
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.5898620467370751,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.2962,
       "step": 504
     },
     {
-      "epoch": 2.0445344129554655,
-      "grad_norm": 0.3932191322203993,
-      "learning_rate": 2.9473104272467526e-05,
-      "loss": 0.2811,
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.46905699987094907,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.2973,
       "step": 505
     },
     {
-      "epoch": 2.048582995951417,
-      "grad_norm": 0.43828843086543573,
-      "learning_rate": 2.942325848714505e-05,
-      "loss": 0.2711,
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.5249500860974198,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.2565,
       "step": 506
     },
     {
-      "epoch": 2.0526315789473686,
-      "grad_norm": 0.4290960163226014,
-      "learning_rate": 2.9373337353698804e-05,
-      "loss": 0.2869,
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.6439663370389226,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3001,
       "step": 507
     },
     {
-      "epoch": 2.0566801619433197,
-      "grad_norm": 0.46836095577645087,
-      "learning_rate": 2.932334127129686e-05,
-      "loss": 0.2958,
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.587555970688215,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3044,
       "step": 508
     },
     {
-      "epoch": 2.0607287449392713,
-      "grad_norm": 0.39125811665279503,
-      "learning_rate": 2.9273270639706547e-05,
-      "loss": 0.2884,
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.5712393204054836,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.2834,
       "step": 509
     },
     {
-      "epoch": 2.064777327935223,
-      "grad_norm": 0.4346352975126504,
-      "learning_rate": 2.922312585929131e-05,
-      "loss": 0.2823,
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.48078530418199455,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.2635,
       "step": 510
     },
     {
-      "epoch": 2.068825910931174,
-      "grad_norm": 0.37577235637591433,
-      "learning_rate": 2.9172907331007478e-05,
-      "loss": 0.3129,
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6305415181396017,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.321,
       "step": 511
     },
     {
-      "epoch": 2.0728744939271255,
-      "grad_norm": 0.42503327369706084,
-      "learning_rate": 2.9122615456401053e-05,
-      "loss": 0.2842,
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.5735906834945632,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.2573,
       "step": 512
     },
     {
-      "epoch": 2.076923076923077,
-      "grad_norm": 0.34557422676373756,
-      "learning_rate": 2.9072250637604535e-05,
-      "loss": 0.2734,
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.5048125863712263,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.283,
       "step": 513
     },
     {
-      "epoch": 2.080971659919028,
-      "grad_norm": 0.42642777920155867,
-      "learning_rate": 2.9021813277333667e-05,
-      "loss": 0.2795,
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.5771311715981097,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.2629,
       "step": 514
     },
     {
-      "epoch": 2.08502024291498,
-      "grad_norm": 0.42154624052077205,
-      "learning_rate": 2.8971303778884244e-05,
-      "loss": 0.2686,
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.505156258462438,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.2853,
       "step": 515
     },
     {
-      "epoch": 2.0890688259109313,
-      "grad_norm": 0.37698563467846613,
-      "learning_rate": 2.8920722546128863e-05,
-      "loss": 0.3021,
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.6013734391608286,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2907,
       "step": 516
     },
     {
-      "epoch": 2.0931174089068825,
-      "grad_norm": 0.4054172109196136,
-      "learning_rate": 2.887006998351372e-05,
-      "loss": 0.2668,
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.3757545464005975,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.2777,
       "step": 517
     },
     {
-      "epoch": 2.097165991902834,
-      "grad_norm": 0.371114056264615,
-      "learning_rate": 2.881934649605535e-05,
-      "loss": 0.2845,
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.5780568316112885,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.2742,
       "step": 518
     },
     {
-      "epoch": 2.1012145748987856,
-      "grad_norm": 0.4282458992971882,
-      "learning_rate": 2.8768552489337408e-05,
-      "loss": 0.2836,
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.46405442848213085,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.311,
       "step": 519
     },
     {
-      "epoch": 2.1052631578947367,
-      "grad_norm": 0.3499396897667106,
-      "learning_rate": 2.8717688369507425e-05,
-      "loss": 0.2683,
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.48236471009341964,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.2474,
       "step": 520
     },
     {
-      "epoch": 2.1093117408906883,
-      "grad_norm": 0.3805454180589256,
-      "learning_rate": 2.8666754543273535e-05,
-      "loss": 0.2561,
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.4894258456721074,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.294,
       "step": 521
     },
     {
-      "epoch": 2.11336032388664,
-      "grad_norm": 0.3989939501875296,
-      "learning_rate": 2.8615751417901273e-05,
-      "loss": 0.2833,
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.3804278155783774,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.2661,
       "step": 522
     },
     {
-      "epoch": 2.117408906882591,
-      "grad_norm": 0.36816249714330496,
-      "learning_rate": 2.856467940121025e-05,
-      "loss": 0.2903,
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.5161510673870723,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2931,
       "step": 523
     },
     {
-      "epoch": 2.1214574898785425,
-      "grad_norm": 0.38675786653271954,
-      "learning_rate": 2.8513538901570964e-05,
-      "loss": 0.2719,
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.47294926374256985,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.2644,
       "step": 524
     },
     {
-      "epoch": 2.125506072874494,
-      "grad_norm": 0.3740936235955749,
-      "learning_rate": 2.846233032790149e-05,
-      "loss": 0.2924,
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5301634773070928,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.2892,
       "step": 525
     },
     {
-      "epoch": 2.1295546558704452,
-      "grad_norm": 0.3731983372967956,
-      "learning_rate": 2.84110540896642e-05,
-      "loss": 0.2832,
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.43494814671782417,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.2714,
       "step": 526
     },
     {
-      "epoch": 2.133603238866397,
-      "grad_norm": 0.4054080764284514,
-      "learning_rate": 2.8359710596862555e-05,
-      "loss": 0.29,
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.4768542495789591,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.2889,
       "step": 527
     },
     {
-      "epoch": 2.1376518218623484,
-      "grad_norm": 0.37243123196437333,
-      "learning_rate": 2.8308300260037734e-05,
-      "loss": 0.2868,
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.44917709767549946,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.2895,
       "step": 528
     },
     {
-      "epoch": 2.1417004048582995,
-      "grad_norm": 0.3631845878901921,
-      "learning_rate": 2.8256823490265424e-05,
-      "loss": 0.288,
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.4681611162223507,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3083,
       "step": 529
     },
     {
-      "epoch": 2.145748987854251,
-      "grad_norm": 0.3494468853230752,
-      "learning_rate": 2.8205280699152515e-05,
-      "loss": 0.279,
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.4326487812211854,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.2748,
       "step": 530
     },
     {
-      "epoch": 2.1497975708502026,
-      "grad_norm": 0.37460551690163546,
-      "learning_rate": 2.8153672298833777e-05,
-      "loss": 0.2806,
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.40461783465747436,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2679,
       "step": 531
     },
     {
-      "epoch": 2.1538461538461537,
-      "grad_norm": 0.3291112260163799,
-      "learning_rate": 2.8101998701968615e-05,
-      "loss": 0.2652,
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.4602497626590127,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.2982,
       "step": 532
     },
     {
-      "epoch": 2.1578947368421053,
-      "grad_norm": 0.39174397764025976,
-      "learning_rate": 2.805026032173773e-05,
-      "loss": 0.2834,
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.4158026794906077,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.2643,
       "step": 533
     },
     {
-      "epoch": 2.161943319838057,
-      "grad_norm": 0.32723867284814784,
-      "learning_rate": 2.7998457571839832e-05,
-      "loss": 0.2689,
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.39922433133197477,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.257,
       "step": 534
     },
     {
-      "epoch": 2.165991902834008,
-      "grad_norm": 0.44135352839130604,
-      "learning_rate": 2.7946590866488322e-05,
-      "loss": 0.2982,
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.4463871631574887,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.296,
       "step": 535
     },
     {
-      "epoch": 2.1700404858299596,
-      "grad_norm": 0.33055595286250944,
-      "learning_rate": 2.7894660620408e-05,
-      "loss": 0.2932,
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.39923365542690586,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.2635,
       "step": 536
     },
     {
-      "epoch": 2.174089068825911,
-      "grad_norm": 0.40071874464373064,
-      "learning_rate": 2.784266724883173e-05,
-      "loss": 0.2667,
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.49091645018374835,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.2945,
       "step": 537
     },
     {
-      "epoch": 2.1781376518218623,
-      "grad_norm": 0.37189779821639746,
-      "learning_rate": 2.7790611167497128e-05,
-      "loss": 0.2838,
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.40577509231911546,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.2569,
       "step": 538
     },
     {
-      "epoch": 2.182186234817814,
-      "grad_norm": 0.3613431114822131,
-      "learning_rate": 2.7738492792643224e-05,
-      "loss": 0.2878,
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.49742619896883133,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3117,
       "step": 539
     },
     {
-      "epoch": 2.1862348178137654,
-      "grad_norm": 0.3604489903368459,
-      "learning_rate": 2.768631254100716e-05,
-      "loss": 0.2928,
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.44373660187789804,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.2668,
       "step": 540
     },
     {
-      "epoch": 2.1902834008097165,
-      "grad_norm": 0.32757082834270734,
-      "learning_rate": 2.7634070829820836e-05,
-      "loss": 0.2861,
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.4268279200532454,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.2796,
       "step": 541
     },
     {
-      "epoch": 2.194331983805668,
-      "grad_norm": 0.35049436157758856,
-      "learning_rate": 2.7581768076807592e-05,
-      "loss": 0.2711,
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.48588023788458623,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.2691,
       "step": 542
     },
     {
-      "epoch": 2.1983805668016196,
-      "grad_norm": 0.3672858693957036,
-      "learning_rate": 2.7529404700178832e-05,
-      "loss": 0.2959,
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.5363467661660773,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3421,
       "step": 543
     },
     {
-      "epoch": 2.2024291497975708,
-      "grad_norm": 0.3753534946143875,
-      "learning_rate": 2.7476981118630735e-05,
-      "loss": 0.3121,
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.4037882144645396,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.2749,
       "step": 544
     },
     {
-      "epoch": 2.2064777327935223,
-      "grad_norm": 0.360459506379204,
-      "learning_rate": 2.742449775134085e-05,
-      "loss": 0.2863,
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.5326243592531117,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.2767,
       "step": 545
     },
     {
-      "epoch": 2.2105263157894735,
-      "grad_norm": 0.3649469608612791,
-      "learning_rate": 2.7371955017964788e-05,
-      "loss": 0.3079,
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.45171916819682023,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.2964,
       "step": 546
     },
     {
-      "epoch": 2.214574898785425,
-      "grad_norm": 0.3752520723130487,
-      "learning_rate": 2.7319353338632845e-05,
-      "loss": 0.2841,
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.522107836213339,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3181,
       "step": 547
     },
     {
-      "epoch": 2.2186234817813766,
-      "grad_norm": 0.3210881439793421,
-      "learning_rate": 2.7266693133946645e-05,
-      "loss": 0.2789,
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.4146238073474943,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.2648,
       "step": 548
     },
     {
-      "epoch": 2.2226720647773277,
-      "grad_norm": 0.40718125723800835,
-      "learning_rate": 2.7213974824975785e-05,
-      "loss": 0.272,
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.48507414204137594,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.2676,
       "step": 549
     },
     {
-      "epoch": 2.2267206477732793,
-      "grad_norm": 0.3636567765863038,
-      "learning_rate": 2.7161198833254464e-05,
-      "loss": 0.2898,
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.4548578287811572,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.2747,
       "step": 550
     },
     {
-      "epoch": 2.230769230769231,
-      "grad_norm": 0.3653323384320538,
-      "learning_rate": 2.7108365580778095e-05,
-      "loss": 0.2819,
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.46061773501856995,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.2731,
       "step": 551
     },
     {
-      "epoch": 2.234817813765182,
-      "grad_norm": 0.3508615079996267,
-      "learning_rate": 2.705547548999996e-05,
-      "loss": 0.2777,
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.520861760288944,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3054,
       "step": 552
     },
     {
-      "epoch": 2.2388663967611335,
-      "grad_norm": 0.3341294926311874,
-      "learning_rate": 2.7002528983827817e-05,
-      "loss": 0.2669,
+      "epoch": 2.24,
+      "grad_norm": 0.4179535650037209,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.2664,
       "step": 553
     },
     {
-      "epoch": 2.242914979757085,
-      "grad_norm": 0.36109424984343874,
-      "learning_rate": 2.6949526485620517e-05,
-      "loss": 0.2856,
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.46896180356185097,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.2952,
       "step": 554
     },
     {
-      "epoch": 2.246963562753036,
-      "grad_norm": 0.33794667398811457,
-      "learning_rate": 2.6896468419184623e-05,
-      "loss": 0.2897,
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.4735363051256582,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.2948,
       "step": 555
     },
     {
-      "epoch": 2.251012145748988,
-      "grad_norm": 0.3407886360282343,
-      "learning_rate": 2.6843355208771024e-05,
-      "loss": 0.2894,
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.4232937039750707,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.2624,
       "step": 556
     },
     {
-      "epoch": 2.2550607287449393,
-      "grad_norm": 0.3330353112393106,
-      "learning_rate": 2.6790187279071536e-05,
-      "loss": 0.2996,
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.5196940894807031,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3036,
       "step": 557
     },
     {
-      "epoch": 2.2591093117408905,
-      "grad_norm": 0.3897784661067059,
-      "learning_rate": 2.67369650552155e-05,
-      "loss": 0.2784,
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.5071138086359843,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2909,
       "step": 558
     },
     {
-      "epoch": 2.263157894736842,
-      "grad_norm": 0.3991445009507451,
-      "learning_rate": 2.6683688962766407e-05,
-      "loss": 0.2928,
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.5146427989986204,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.2762,
       "step": 559
     },
     {
-      "epoch": 2.2672064777327936,
-      "grad_norm": 0.3801365676476558,
-      "learning_rate": 2.6630359427718468e-05,
-      "loss": 0.2739,
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.5354532326077166,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3105,
       "step": 560
     },
     {
-      "epoch": 2.2712550607287447,
-      "grad_norm": 0.3472520056528079,
-      "learning_rate": 2.6576976876493225e-05,
-      "loss": 0.2722,
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.443082597353416,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.2457,
       "step": 561
     },
     {
-      "epoch": 2.2753036437246963,
-      "grad_norm": 0.3930094072159701,
-      "learning_rate": 2.652354173593614e-05,
-      "loss": 0.2871,
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.49647525542453036,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3052,
       "step": 562
     },
     {
-      "epoch": 2.279352226720648,
-      "grad_norm": 0.3273713549105774,
-      "learning_rate": 2.647005443331317e-05,
-      "loss": 0.2869,
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.4060061839556909,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.2596,
       "step": 563
     },
     {
-      "epoch": 2.283400809716599,
-      "grad_norm": 0.3760252130817158,
-      "learning_rate": 2.6416515396307354e-05,
-      "loss": 0.2794,
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.473254927025622,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2812,
       "step": 564
     },
     {
-      "epoch": 2.2874493927125505,
-      "grad_norm": 0.3448659818899541,
-      "learning_rate": 2.636292505301542e-05,
-      "loss": 0.2956,
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.4290575398830281,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.2914,
       "step": 565
     },
     {
-      "epoch": 2.291497975708502,
-      "grad_norm": 0.36854874938254556,
-      "learning_rate": 2.630928383194431e-05,
-      "loss": 0.2819,
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.512355115735953,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.2795,
       "step": 566
     },
     {
-      "epoch": 2.2955465587044532,
-      "grad_norm": 0.4289984610197063,
-      "learning_rate": 2.625559216200781e-05,
-      "loss": 0.2997,
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.4849182590053955,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3046,
       "step": 567
     },
     {
-      "epoch": 2.299595141700405,
-      "grad_norm": 0.3379359175107265,
-      "learning_rate": 2.620185047252308e-05,
-      "loss": 0.2879,
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.4652580975889665,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.2776,
       "step": 568
     },
     {
-      "epoch": 2.3036437246963564,
-      "grad_norm": 0.4747942747498032,
-      "learning_rate": 2.614805919320723e-05,
-      "loss": 0.2867,
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.5054995590668088,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.2841,
       "step": 569
     },
     {
-      "epoch": 2.3076923076923075,
-      "grad_norm": 0.35552912307818374,
-      "learning_rate": 2.6094218754173895e-05,
-      "loss": 0.2844,
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.40720863977550853,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2997,
       "step": 570
     },
     {
-      "epoch": 2.311740890688259,
-      "grad_norm": 0.3976757924982598,
-      "learning_rate": 2.60403295859298e-05,
-      "loss": 0.2864,
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.5564492883739593,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.2953,
       "step": 571
     },
     {
-      "epoch": 2.3157894736842106,
-      "grad_norm": 0.4441911437369727,
-      "learning_rate": 2.5986392119371297e-05,
-      "loss": 0.3025,
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.39609946528207157,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.2949,
       "step": 572
     },
     {
-      "epoch": 2.3198380566801617,
-      "grad_norm": 0.4334387543042122,
-      "learning_rate": 2.5932406785780928e-05,
-      "loss": 0.2921,
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.6572882994104452,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.304,
       "step": 573
     },
     {
-      "epoch": 2.3238866396761133,
-      "grad_norm": 0.542290531269228,
-      "learning_rate": 2.5878374016823987e-05,
-      "loss": 0.2807,
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.38970923154733966,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.2529,
       "step": 574
     },
     {
-      "epoch": 2.327935222672065,
-      "grad_norm": 0.37069150632032466,
-      "learning_rate": 2.5824294244545045e-05,
-      "loss": 0.2823,
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5868013564536844,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3134,
       "step": 575
     },
     {
-      "epoch": 2.331983805668016,
-      "grad_norm": 0.43668207676934645,
-      "learning_rate": 2.577016790136453e-05,
-      "loss": 0.2693,
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.4633872753042094,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.2651,
       "step": 576
     },
     {
-      "epoch": 2.3360323886639676,
-      "grad_norm": 0.37978571622323476,
-      "learning_rate": 2.5715995420075225e-05,
-      "loss": 0.2822,
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.4908438542203886,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.2532,
       "step": 577
     },
     {
-      "epoch": 2.340080971659919,
-      "grad_norm": 0.43902450627080414,
-      "learning_rate": 2.5661777233838864e-05,
-      "loss": 0.2934,
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.47208486504666963,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3105,
       "step": 578
     },
     {
-      "epoch": 2.3441295546558703,
-      "grad_norm": 0.37771144605870544,
-      "learning_rate": 2.5607513776182616e-05,
-      "loss": 0.2708,
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.564565537034114,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.2812,
       "step": 579
     },
     {
-      "epoch": 2.348178137651822,
-      "grad_norm": 0.43468598871205744,
-      "learning_rate": 2.5553205480995633e-05,
-      "loss": 0.2897,
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.4106061508874553,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3084,
       "step": 580
     },
     {
-      "epoch": 2.3522267206477734,
-      "grad_norm": 0.34397522983634676,
-      "learning_rate": 2.5498852782525604e-05,
-      "loss": 0.3007,
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.4786843298858375,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.2772,
       "step": 581
     },
     {
-      "epoch": 2.3562753036437245,
-      "grad_norm": 0.3627985209956064,
-      "learning_rate": 2.5444456115375252e-05,
-      "loss": 0.2998,
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.4194564879815133,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2965,
       "step": 582
     },
     {
-      "epoch": 2.360323886639676,
-      "grad_norm": 0.35015806774592156,
-      "learning_rate": 2.539001591449889e-05,
-      "loss": 0.2987,
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.4130393227546154,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.2819,
       "step": 583
     },
     {
-      "epoch": 2.3643724696356276,
-      "grad_norm": 0.3696897793474707,
-      "learning_rate": 2.53355326151989e-05,
-      "loss": 0.2771,
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.4678320355660151,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.2951,
       "step": 584
     },
     {
-      "epoch": 2.3684210526315788,
-      "grad_norm": 0.38184889714846365,
-      "learning_rate": 2.5281006653122288e-05,
-      "loss": 0.2681,
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.4198026120276588,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.2776,
       "step": 585
     },
     {
-      "epoch": 2.3724696356275303,
-      "grad_norm": 0.34675335736483526,
-      "learning_rate": 2.5226438464257217e-05,
-      "loss": 0.2928,
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.4324821539705962,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.2952,
       "step": 586
     },
     {
-      "epoch": 2.376518218623482,
-      "grad_norm": 0.3465514033985015,
-      "learning_rate": 2.5171828484929447e-05,
-      "loss": 0.2883,
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.39135633655407837,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.2956,
       "step": 587
     },
     {
-      "epoch": 2.380566801619433,
-      "grad_norm": 0.36605939621420913,
-      "learning_rate": 2.5117177151798934e-05,
-      "loss": 0.2719,
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.47148090243237756,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.2733,
       "step": 588
     },
     {
-      "epoch": 2.3846153846153846,
-      "grad_norm": 0.37765963615658926,
-      "learning_rate": 2.506248490185627e-05,
-      "loss": 0.2766,
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.4298285179750515,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3112,
       "step": 589
     },
     {
-      "epoch": 2.388663967611336,
-      "grad_norm": 0.37308957770479745,
-      "learning_rate": 2.500775217241923e-05,
-      "loss": 0.2771,
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.3835993090114841,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.2429,
       "step": 590
     },
     {
-      "epoch": 2.3927125506072873,
-      "grad_norm": 0.3708349168272413,
-      "learning_rate": 2.4952979401129266e-05,
-      "loss": 0.2905,
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.4320143826366396,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.2803,
       "step": 591
     },
     {
-      "epoch": 2.396761133603239,
-      "grad_norm": 0.4182441697535505,
-      "learning_rate": 2.4898167025948e-05,
-      "loss": 0.2812,
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.4094315005131263,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.292,
       "step": 592
     },
     {
-      "epoch": 2.4008097165991904,
-      "grad_norm": 0.35325432539400836,
-      "learning_rate": 2.4843315485153703e-05,
-      "loss": 0.2839,
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.440375048742197,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.2867,
       "step": 593
     },
     {
-      "epoch": 2.4048582995951415,
-      "grad_norm": 0.41707173497370326,
-      "learning_rate": 2.4788425217337856e-05,
-      "loss": 0.277,
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.4928371367658382,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.289,
       "step": 594
     },
     {
-      "epoch": 2.408906882591093,
-      "grad_norm": 0.38132535490812297,
-      "learning_rate": 2.4733496661401572e-05,
-      "loss": 0.271,
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.4340500056097791,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2772,
       "step": 595
     },
     {
-      "epoch": 2.4129554655870447,
-      "grad_norm": 0.36217824464335097,
-      "learning_rate": 2.4678530256552114e-05,
-      "loss": 0.2931,
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.3980291957675104,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.2668,
       "step": 596
     },
     {
-      "epoch": 2.417004048582996,
-      "grad_norm": 0.3522226778848643,
-      "learning_rate": 2.46235264422994e-05,
-      "loss": 0.2851,
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.41543339083449476,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.2904,
       "step": 597
     },
     {
-      "epoch": 2.4210526315789473,
-      "grad_norm": 0.29969679847083797,
-      "learning_rate": 2.4568485658452452e-05,
-      "loss": 0.2938,
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.44856637869572913,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.2793,
       "step": 598
     },
     {
-      "epoch": 2.425101214574899,
-      "grad_norm": 0.3429401973557726,
-      "learning_rate": 2.451340834511592e-05,
-      "loss": 0.2943,
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.4562052204229528,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3071,
       "step": 599
     },
     {
-      "epoch": 2.42914979757085,
-      "grad_norm": 0.32962718040034955,
-      "learning_rate": 2.4458294942686528e-05,
-      "loss": 0.3047,
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.5959302866213294,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3283,
       "step": 600
     },
     {
-      "epoch": 2.4331983805668016,
-      "grad_norm": 0.3298753028644587,
-      "learning_rate": 2.440314589184958e-05,
-      "loss": 0.28,
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.3741770683748045,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.2645,
       "step": 601
     },
     {
-      "epoch": 2.437246963562753,
-      "grad_norm": 0.35995118761669676,
-      "learning_rate": 2.434796163357541e-05,
-      "loss": 0.2673,
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.45667394024324337,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.2674,
       "step": 602
     },
     {
-      "epoch": 2.4412955465587043,
-      "grad_norm": 0.33222477386340454,
-      "learning_rate": 2.429274260911588e-05,
-      "loss": 0.291,
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.46063229625144747,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.2716,
       "step": 603
     },
     {
-      "epoch": 2.445344129554656,
-      "grad_norm": 0.35375268240682206,
-      "learning_rate": 2.423748926000084e-05,
-      "loss": 0.2778,
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.43970384507369575,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3052,
       "step": 604
     },
     {
-      "epoch": 2.4493927125506074,
-      "grad_norm": 0.3402942441917847,
-      "learning_rate": 2.418220202803459e-05,
-      "loss": 0.2829,
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.38661011435995857,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.2564,
       "step": 605
     },
     {
-      "epoch": 2.4534412955465585,
-      "grad_norm": 0.3265838484653659,
-      "learning_rate": 2.412688135529237e-05,
-      "loss": 0.2737,
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.3952789468905448,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.2719,
       "step": 606
     },
     {
-      "epoch": 2.45748987854251,
-      "grad_norm": 0.35059894986085494,
-      "learning_rate": 2.4071527684116807e-05,
-      "loss": 0.2771,
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.3866003509597451,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.2818,
       "step": 607
     },
     {
-      "epoch": 2.4615384615384617,
-      "grad_norm": 0.3542764956562334,
-      "learning_rate": 2.4016141457114372e-05,
-      "loss": 0.2879,
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.4249803576555808,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3025,
       "step": 608
     },
     {
-      "epoch": 2.465587044534413,
-      "grad_norm": 0.3794152668009615,
-      "learning_rate": 2.396072311715187e-05,
-      "loss": 0.2776,
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.4220466976945481,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.2831,
       "step": 609
     },
     {
-      "epoch": 2.4696356275303644,
-      "grad_norm": 0.35483458954759917,
-      "learning_rate": 2.3905273107352863e-05,
-      "loss": 0.2742,
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.35497046932730186,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2596,
       "step": 610
     },
     {
-      "epoch": 2.473684210526316,
-      "grad_norm": 0.33958001219573086,
-      "learning_rate": 2.3849791871094157e-05,
-      "loss": 0.3132,
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.42127218531199895,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3195,
       "step": 611
     },
     {
-      "epoch": 2.477732793522267,
-      "grad_norm": 0.37224036550787687,
-      "learning_rate": 2.379427985200223e-05,
-      "loss": 0.2757,
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.38805030738043556,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.2906,
       "step": 612
     },
     {
-      "epoch": 2.4817813765182186,
-      "grad_norm": 0.3840586023079233,
-      "learning_rate": 2.373873749394972e-05,
-      "loss": 0.2889,
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.3315743982735547,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.254,
       "step": 613
     },
     {
-      "epoch": 2.48582995951417,
-      "grad_norm": 0.3679886516727151,
-      "learning_rate": 2.3683165241051836e-05,
-      "loss": 0.2971,
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.38013961574533955,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.2811,
       "step": 614
     },
     {
-      "epoch": 2.4898785425101213,
-      "grad_norm": 0.3938591682299967,
-      "learning_rate": 2.3627563537662823e-05,
-      "loss": 0.2825,
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.40030740877649773,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3056,
       "step": 615
     },
     {
-      "epoch": 2.493927125506073,
-      "grad_norm": 0.34548201266219397,
-      "learning_rate": 2.3571932828372433e-05,
-      "loss": 0.3069,
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.378837141033372,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3081,
       "step": 616
     },
     {
-      "epoch": 2.4979757085020244,
-      "grad_norm": 0.35225361185918846,
-      "learning_rate": 2.351627355800233e-05,
-      "loss": 0.2858,
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.3776478865324526,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.2763,
       "step": 617
     },
     {
-      "epoch": 2.5020242914979756,
-      "grad_norm": 0.32323838201147026,
-      "learning_rate": 2.346058617160256e-05,
-      "loss": 0.2954,
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.3927727439578414,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.286,
       "step": 618
     },
     {
-      "epoch": 2.506072874493927,
-      "grad_norm": 0.3682592861519381,
-      "learning_rate": 2.3404871114447976e-05,
-      "loss": 0.283,
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.3544326210682019,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.2676,
       "step": 619
     },
     {
-      "epoch": 2.5101214574898787,
-      "grad_norm": 0.3221374829484278,
-      "learning_rate": 2.334912883203469e-05,
-      "loss": 0.2797,
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.36617029268471696,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3015,
       "step": 620
     },
     {
-      "epoch": 2.51417004048583,
-      "grad_norm": 0.3396826546982308,
-      "learning_rate": 2.3293359770076505e-05,
-      "loss": 0.2886,
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.32282547047554155,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2582,
       "step": 621
     },
     {
-      "epoch": 2.5182186234817814,
-      "grad_norm": 0.33086110829376303,
-      "learning_rate": 2.323756437450136e-05,
-      "loss": 0.2909,
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.38001447152372375,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.2889,
       "step": 622
     },
     {
-      "epoch": 2.522267206477733,
-      "grad_norm": 0.32819304051479237,
-      "learning_rate": 2.3181743091447754e-05,
-      "loss": 0.2763,
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.3995208038526333,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.301,
       "step": 623
     },
     {
-      "epoch": 2.526315789473684,
-      "grad_norm": 0.3667867660652519,
-      "learning_rate": 2.3125896367261164e-05,
-      "loss": 0.2746,
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.3683266320349205,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.2562,
       "step": 624
     },
     {
-      "epoch": 2.5303643724696356,
-      "grad_norm": 0.3457868713606491,
-      "learning_rate": 2.307002464849052e-05,
-      "loss": 0.2633,
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.4079616069291649,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3164,
       "step": 625
     },
     {
-      "epoch": 2.534412955465587,
-      "grad_norm": 0.34987398192872377,
-      "learning_rate": 2.30141283818846e-05,
-      "loss": 0.2926,
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.32981703454268513,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.2421,
       "step": 626
     },
     {
-      "epoch": 2.5384615384615383,
-      "grad_norm": 0.3701421607802869,
-      "learning_rate": 2.2958208014388462e-05,
-      "loss": 0.2851,
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.4255864047759635,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3163,
       "step": 627
     },
     {
-      "epoch": 2.54251012145749,
-      "grad_norm": 0.3428190753850153,
-      "learning_rate": 2.2902263993139857e-05,
-      "loss": 0.2933,
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.4061085652124247,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2826,
       "step": 628
     },
     {
-      "epoch": 2.5465587044534415,
-      "grad_norm": 0.3423612222395882,
-      "learning_rate": 2.2846296765465708e-05,
-      "loss": 0.2804,
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.4047488432742045,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.2801,
       "step": 629
     },
     {
-      "epoch": 2.5506072874493926,
-      "grad_norm": 0.42288883885715745,
-      "learning_rate": 2.2790306778878465e-05,
-      "loss": 0.2855,
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.3714180599209369,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2833,
       "step": 630
     },
     {
-      "epoch": 2.554655870445344,
-      "grad_norm": 0.3210592756961385,
-      "learning_rate": 2.273429448107257e-05,
-      "loss": 0.3015,
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.48434481198829926,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.2853,
       "step": 631
     },
     {
-      "epoch": 2.5587044534412957,
-      "grad_norm": 0.38891694528596815,
-      "learning_rate": 2.2678260319920868e-05,
-      "loss": 0.2858,
+      "epoch": 2.56,
+      "grad_norm": 0.4020387923903253,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.2849,
       "step": 632
     },
     {
-      "epoch": 2.562753036437247,
-      "grad_norm": 0.36901246039732666,
-      "learning_rate": 2.262220474347102e-05,
-      "loss": 0.2848,
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.4362417585172134,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.2973,
       "step": 633
     },
     {
-      "epoch": 2.5668016194331984,
-      "grad_norm": 0.3529328779589883,
-      "learning_rate": 2.2566128199941918e-05,
-      "loss": 0.2728,
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.5489057726408276,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3164,
       "step": 634
     },
     {
-      "epoch": 2.57085020242915,
-      "grad_norm": 0.3712830807781546,
-      "learning_rate": 2.2510031137720104e-05,
-      "loss": 0.2854,
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.36985399471488123,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.2943,
       "step": 635
     },
     {
-      "epoch": 2.574898785425101,
-      "grad_norm": 0.3047194296394481,
-      "learning_rate": 2.2453914005356206e-05,
-      "loss": 0.2833,
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.35835533356045157,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2463,
       "step": 636
     },
     {
-      "epoch": 2.5789473684210527,
-      "grad_norm": 0.33611856962773,
-      "learning_rate": 2.2397777251561302e-05,
-      "loss": 0.2877,
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.3802873562364821,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.2992,
       "step": 637
     },
     {
-      "epoch": 2.582995951417004,
-      "grad_norm": 0.3299134791546615,
-      "learning_rate": 2.2341621325203393e-05,
-      "loss": 0.3121,
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.37507945035050516,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.2854,
       "step": 638
     },
     {
-      "epoch": 2.5870445344129553,
-      "grad_norm": 0.3584788645266628,
-      "learning_rate": 2.2285446675303764e-05,
-      "loss": 0.2869,
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.40480106137928634,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.294,
       "step": 639
     },
     {
-      "epoch": 2.591093117408907,
-      "grad_norm": 0.31682955031384075,
-      "learning_rate": 2.2229253751033423e-05,
-      "loss": 0.2656,
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.3466099727691219,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.2829,
       "step": 640
     },
     {
-      "epoch": 2.5951417004048585,
-      "grad_norm": 0.34637083354882653,
-      "learning_rate": 2.2173043001709492e-05,
-      "loss": 0.2824,
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.3891405147954072,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.2989,
       "step": 641
     },
     {
-      "epoch": 2.5991902834008096,
-      "grad_norm": 0.3186056169362435,
-      "learning_rate": 2.2116814876791627e-05,
-      "loss": 0.2938,
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.4132357320826912,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.2618,
       "step": 642
     },
     {
-      "epoch": 2.603238866396761,
-      "grad_norm": 0.32960589622735,
-      "learning_rate": 2.2060569825878413e-05,
-      "loss": 0.2887,
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.3900917568124276,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3144,
       "step": 643
     },
     {
-      "epoch": 2.6072874493927127,
-      "grad_norm": 0.31582478190489804,
-      "learning_rate": 2.2004308298703777e-05,
-      "loss": 0.3005,
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.358101745683651,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.2878,
       "step": 644
     },
     {
-      "epoch": 2.611336032388664,
-      "grad_norm": 0.27965788807823566,
-      "learning_rate": 2.1948030745133396e-05,
-      "loss": 0.2873,
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.3477300438804496,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.2699,
       "step": 645
     },
     {
-      "epoch": 2.6153846153846154,
-      "grad_norm": 0.32819302003635037,
-      "learning_rate": 2.1891737615161083e-05,
-      "loss": 0.2749,
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.3866787472286775,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.2714,
       "step": 646
     },
     {
-      "epoch": 2.619433198380567,
-      "grad_norm": 0.3507725415420072,
-      "learning_rate": 2.1835429358905218e-05,
-      "loss": 0.2834,
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.37343111841630233,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.2936,
       "step": 647
     },
     {
-      "epoch": 2.623481781376518,
-      "grad_norm": 0.3092720969876203,
-      "learning_rate": 2.17791064266051e-05,
-      "loss": 0.2794,
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.37381450559996626,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.2899,
       "step": 648
     },
     {
-      "epoch": 2.6275303643724697,
-      "grad_norm": 0.3177197763499792,
-      "learning_rate": 2.172276926861741e-05,
-      "loss": 0.29,
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.39377728116098054,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.2902,
       "step": 649
     },
     {
-      "epoch": 2.6315789473684212,
-      "grad_norm": 0.33221610418929026,
-      "learning_rate": 2.166641833541255e-05,
-      "loss": 0.2943,
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.3671876985649111,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.274,
       "step": 650
     },
     {
-      "epoch": 2.6356275303643724,
-      "grad_norm": 0.32786719417706717,
-      "learning_rate": 2.161005407757109e-05,
-      "loss": 0.2654,
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.41051229226746877,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3057,
       "step": 651
     },
     {
-      "epoch": 2.639676113360324,
-      "grad_norm": 0.35941900694150564,
-      "learning_rate": 2.155367694578013e-05,
-      "loss": 0.3065,
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.45051344454766024,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3014,
       "step": 652
     },
     {
-      "epoch": 2.6437246963562755,
-      "grad_norm": 0.3175771727629365,
-      "learning_rate": 2.1497287390829712e-05,
-      "loss": 0.2994,
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.4109785157331815,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.2579,
       "step": 653
     },
     {
-      "epoch": 2.6477732793522266,
-      "grad_norm": 0.32800265679743495,
-      "learning_rate": 2.1440885863609208e-05,
-      "loss": 0.2656,
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.44695211051783246,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.2685,
       "step": 654
     },
     {
-      "epoch": 2.651821862348178,
-      "grad_norm": 0.3574644607645087,
-      "learning_rate": 2.1384472815103737e-05,
-      "loss": 0.3018,
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.4557215230925412,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3026,
       "step": 655
     },
     {
-      "epoch": 2.6558704453441297,
-      "grad_norm": 0.35553813768940096,
-      "learning_rate": 2.1328048696390534e-05,
-      "loss": 0.2878,
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.4539460445825398,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.2762,
       "step": 656
     },
     {
-      "epoch": 2.659919028340081,
-      "grad_norm": 0.35481563612517214,
-      "learning_rate": 2.1271613958635342e-05,
-      "loss": 0.2942,
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.5264259693317037,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.308,
       "step": 657
     },
     {
-      "epoch": 2.6639676113360324,
-      "grad_norm": 0.35268282311624,
-      "learning_rate": 2.1215169053088825e-05,
-      "loss": 0.2682,
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.44543036408656983,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.2633,
       "step": 658
     },
     {
-      "epoch": 2.668016194331984,
-      "grad_norm": 0.348976089442499,
-      "learning_rate": 2.1158714431082937e-05,
-      "loss": 0.3041,
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.42670515306005696,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3149,
       "step": 659
     },
     {
-      "epoch": 2.672064777327935,
-      "grad_norm": 0.33348552012922333,
-      "learning_rate": 2.110225054402733e-05,
-      "loss": 0.2906,
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.4817430336149499,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.2883,
       "step": 660
     },
     {
-      "epoch": 2.6761133603238867,
-      "grad_norm": 0.3454583832196458,
-      "learning_rate": 2.1045777843405747e-05,
-      "loss": 0.2656,
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.40983849369515485,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.2987,
       "step": 661
     },
     {
-      "epoch": 2.6801619433198383,
-      "grad_norm": 0.33030930446614576,
-      "learning_rate": 2.09892967807724e-05,
-      "loss": 0.2989,
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.36970532474235995,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.2554,
       "step": 662
     },
     {
-      "epoch": 2.6842105263157894,
-      "grad_norm": 0.3117204436485248,
-      "learning_rate": 2.093280780774835e-05,
-      "loss": 0.2752,
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.3752489456838842,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.2739,
       "step": 663
     },
     {
-      "epoch": 2.688259109311741,
-      "grad_norm": 0.3181356426617412,
-      "learning_rate": 2.087631137601793e-05,
-      "loss": 0.2934,
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.341224852156229,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.2822,
       "step": 664
     },
     {
-      "epoch": 2.6923076923076925,
-      "grad_norm": 0.3142135559381296,
-      "learning_rate": 2.0819807937325104e-05,
-      "loss": 0.264,
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.34113297343219817,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2653,
       "step": 665
     },
     {
-      "epoch": 2.6963562753036436,
-      "grad_norm": 0.3096644784458591,
-      "learning_rate": 2.0763297943469847e-05,
-      "loss": 0.2648,
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.3570666289525047,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.2878,
       "step": 666
     },
     {
-      "epoch": 2.700404858299595,
-      "grad_norm": 0.30123750494847745,
-      "learning_rate": 2.070678184630458e-05,
-      "loss": 0.2865,
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.3529330112371475,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.2721,
       "step": 667
     },
     {
-      "epoch": 2.7044534412955468,
-      "grad_norm": 0.30985103200859543,
-      "learning_rate": 2.06502600977305e-05,
-      "loss": 0.2991,
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.36615525653480974,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.2929,
       "step": 668
     },
     {
-      "epoch": 2.708502024291498,
-      "grad_norm": 0.35310905026546724,
-      "learning_rate": 2.0593733149694008e-05,
-      "loss": 0.2768,
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.3308640258881196,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.2504,
       "step": 669
     },
     {
-      "epoch": 2.7125506072874495,
-      "grad_norm": 0.31025203440529814,
-      "learning_rate": 2.0537201454183074e-05,
-      "loss": 0.2873,
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.402899373034655,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3059,
       "step": 670
     },
     {
-      "epoch": 2.716599190283401,
-      "grad_norm": 0.37409835624647114,
-      "learning_rate": 2.0480665463223626e-05,
-      "loss": 0.303,
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.3754585891881446,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2845,
       "step": 671
     },
     {
-      "epoch": 2.720647773279352,
-      "grad_norm": 0.2943316550918794,
-      "learning_rate": 2.042412562887594e-05,
-      "loss": 0.2758,
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.4219300354913725,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.2981,
       "step": 672
     },
     {
-      "epoch": 2.7246963562753037,
-      "grad_norm": 0.34106384393653877,
-      "learning_rate": 2.0367582403231033e-05,
-      "loss": 0.2969,
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.3747487242247928,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3071,
       "step": 673
     },
     {
-      "epoch": 2.7287449392712553,
-      "grad_norm": 0.3399006776471924,
-      "learning_rate": 2.0311036238407023e-05,
-      "loss": 0.3135,
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.40297142278616027,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.2849,
       "step": 674
     },
     {
-      "epoch": 2.7327935222672064,
-      "grad_norm": 0.3028977573330079,
-      "learning_rate": 2.0254487586545534e-05,
-      "loss": 0.2823,
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.42195909584567604,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3006,
       "step": 675
     },
     {
-      "epoch": 2.736842105263158,
-      "grad_norm": 0.3554358021095428,
-      "learning_rate": 2.0197936899808088e-05,
-      "loss": 0.2814,
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.3795628554658524,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2757,
       "step": 676
     },
     {
-      "epoch": 2.7408906882591095,
-      "grad_norm": 0.33905960857432704,
-      "learning_rate": 2.0141384630372466e-05,
-      "loss": 0.2831,
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.40412625879847486,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.2966,
       "step": 677
     },
     {
-      "epoch": 2.7449392712550607,
-      "grad_norm": 0.34172091883303896,
-      "learning_rate": 2.0084831230429106e-05,
-      "loss": 0.2777,
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.42034283889260926,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.2615,
       "step": 678
     },
     {
-      "epoch": 2.748987854251012,
-      "grad_norm": 0.35935842733675893,
-      "learning_rate": 2.0028277152177492e-05,
-      "loss": 0.2825,
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.36164754395761955,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2873,
       "step": 679
     },
     {
-      "epoch": 2.753036437246964,
-      "grad_norm": 0.35041882115708683,
-      "learning_rate": 1.9971722847822518e-05,
-      "loss": 0.2812,
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.38435321002785144,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.2777,
       "step": 680
     },
     {
-      "epoch": 2.757085020242915,
-      "grad_norm": 0.326990603352221,
-      "learning_rate": 1.99151687695709e-05,
-      "loss": 0.263,
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.3840939395096417,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.2775,
       "step": 681
     },
     {
-      "epoch": 2.7611336032388665,
-      "grad_norm": 0.3715982505282083,
-      "learning_rate": 1.9858615369627537e-05,
-      "loss": 0.2782,
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.39238754686185523,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.2746,
       "step": 682
     },
     {
-      "epoch": 2.765182186234818,
-      "grad_norm": 0.3484130404909403,
-      "learning_rate": 1.9802063100191916e-05,
-      "loss": 0.2796,
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.4008545714323698,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2754,
       "step": 683
     },
     {
-      "epoch": 2.769230769230769,
-      "grad_norm": 0.357022914789281,
-      "learning_rate": 1.974551241345447e-05,
-      "loss": 0.3007,
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.3433750776144352,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.2735,
       "step": 684
     },
     {
-      "epoch": 2.7732793522267207,
-      "grad_norm": 0.35748117657114353,
-      "learning_rate": 1.9688963761592984e-05,
-      "loss": 0.2808,
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.3999610758353946,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.2887,
       "step": 685
     },
     {
-      "epoch": 2.7773279352226723,
-      "grad_norm": 0.33616858693469054,
-      "learning_rate": 1.963241759676897e-05,
-      "loss": 0.2899,
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.36306407650088524,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.2914,
       "step": 686
     },
     {
-      "epoch": 2.7813765182186234,
-      "grad_norm": 0.37449969419846807,
-      "learning_rate": 1.9575874371124066e-05,
-      "loss": 0.2811,
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.4727077178333078,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.306,
       "step": 687
     },
     {
-      "epoch": 2.785425101214575,
-      "grad_norm": 0.31290032114325156,
-      "learning_rate": 1.951933453677638e-05,
-      "loss": 0.2878,
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.3557658623865322,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.27,
       "step": 688
     },
     {
-      "epoch": 2.7894736842105265,
-      "grad_norm": 0.3823169735130646,
-      "learning_rate": 1.9462798545816932e-05,
-      "loss": 0.2845,
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.4692384610547744,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3049,
       "step": 689
     },
     {
-      "epoch": 2.7935222672064777,
-      "grad_norm": 0.32470030472673606,
-      "learning_rate": 1.9406266850305996e-05,
-      "loss": 0.2982,
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.3656205333114426,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.2682,
       "step": 690
     },
     {
-      "epoch": 2.7975708502024292,
-      "grad_norm": 0.3514875077103973,
-      "learning_rate": 1.9349739902269503e-05,
-      "loss": 0.2882,
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.509054237210849,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.2804,
       "step": 691
     },
     {
-      "epoch": 2.801619433198381,
-      "grad_norm": 0.3347015794045389,
-      "learning_rate": 1.9293218153695425e-05,
-      "loss": 0.2619,
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.3949548113001569,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.2818,
       "step": 692
     },
     {
-      "epoch": 2.805668016194332,
-      "grad_norm": 0.3200404843210326,
-      "learning_rate": 1.9236702056530153e-05,
-      "loss": 0.2728,
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.4086513643714326,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.2871,
       "step": 693
     },
     {
-      "epoch": 2.8097165991902835,
-      "grad_norm": 0.3623481634169084,
-      "learning_rate": 1.9180192062674907e-05,
-      "loss": 0.2756,
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.47858137092813735,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.2558,
       "step": 694
     },
     {
-      "epoch": 2.813765182186235,
-      "grad_norm": 0.32585218188540926,
-      "learning_rate": 1.9123688623982076e-05,
-      "loss": 0.2925,
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.39793590965016695,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3024,
       "step": 695
     },
     {
-      "epoch": 2.817813765182186,
-      "grad_norm": 0.33865167628839715,
-      "learning_rate": 1.906719219225165e-05,
-      "loss": 0.2852,
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.5651284751751702,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2827,
       "step": 696
     },
     {
-      "epoch": 2.8218623481781377,
-      "grad_norm": 0.3802223906073882,
-      "learning_rate": 1.9010703219227603e-05,
-      "loss": 0.2832,
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.3696402717881142,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2637,
       "step": 697
     },
     {
-      "epoch": 2.8259109311740893,
-      "grad_norm": 0.3106814294675215,
-      "learning_rate": 1.8954222156594256e-05,
-      "loss": 0.2881,
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.4490188601890236,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3013,
       "step": 698
     },
     {
-      "epoch": 2.8299595141700404,
-      "grad_norm": 0.3426040680296811,
-      "learning_rate": 1.8897749455972673e-05,
-      "loss": 0.2836,
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.4749337414873632,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.308,
       "step": 699
     },
     {
-      "epoch": 2.834008097165992,
-      "grad_norm": 0.32809815057580144,
-      "learning_rate": 1.884128556891707e-05,
-      "loss": 0.2945,
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.3551323674605517,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2747,
       "step": 700
     },
     {
-      "epoch": 2.8380566801619436,
-      "grad_norm": 0.32372785212613603,
-      "learning_rate": 1.878483094691119e-05,
-      "loss": 0.275,
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.43394071460729716,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.2638,
       "step": 701
     },
     {
-      "epoch": 2.8421052631578947,
-      "grad_norm": 0.32424837211886254,
-      "learning_rate": 1.8728386041364664e-05,
-      "loss": 0.2943,
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.3943672226551438,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.302,
       "step": 702
     },
     {
-      "epoch": 2.8461538461538463,
-      "grad_norm": 0.31030539856964573,
-      "learning_rate": 1.867195130360947e-05,
-      "loss": 0.3062,
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.38193776260582096,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.287,
       "step": 703
     },
     {
-      "epoch": 2.850202429149798,
-      "grad_norm": 0.30529959558936487,
-      "learning_rate": 1.8615527184896263e-05,
-      "loss": 0.3046,
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.4027832200171129,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.2905,
       "step": 704
     },
     {
-      "epoch": 2.854251012145749,
-      "grad_norm": 0.2936748409529249,
-      "learning_rate": 1.8559114136390795e-05,
-      "loss": 0.2575,
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.34913817935652525,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.2792,
       "step": 705
     },
     {
-      "epoch": 2.8582995951417005,
-      "grad_norm": 0.2914033001389704,
-      "learning_rate": 1.8502712609170298e-05,
-      "loss": 0.2732,
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.35962937512287657,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.2841,
       "step": 706
     },
     {
-      "epoch": 2.862348178137652,
-      "grad_norm": 0.29435903695029597,
-      "learning_rate": 1.8446323054219876e-05,
-      "loss": 0.2749,
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.3860240365752726,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.267,
       "step": 707
     },
     {
-      "epoch": 2.866396761133603,
-      "grad_norm": 0.3377790163669105,
-      "learning_rate": 1.838994592242891e-05,
-      "loss": 0.3104,
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.4219953491673746,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3138,
       "step": 708
     },
     {
-      "epoch": 2.8704453441295548,
-      "grad_norm": 0.31129249060880965,
-      "learning_rate": 1.8333581664587453e-05,
-      "loss": 0.2976,
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.3654631320972551,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.2816,
       "step": 709
     },
     {
-      "epoch": 2.8744939271255063,
-      "grad_norm": 0.336180288393722,
-      "learning_rate": 1.8277230731382593e-05,
-      "loss": 0.2931,
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.3888057825575905,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.2843,
       "step": 710
     },
     {
-      "epoch": 2.8785425101214575,
-      "grad_norm": 0.3234136598402761,
-      "learning_rate": 1.82208935733949e-05,
-      "loss": 0.2832,
+      "epoch": 2.88,
+      "grad_norm": 0.3453588178308771,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.2855,
       "step": 711
     },
     {
-      "epoch": 2.882591093117409,
-      "grad_norm": 0.32915897180673415,
-      "learning_rate": 1.8164570641094793e-05,
-      "loss": 0.2761,
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.40339483991087344,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.2735,
       "step": 712
     },
     {
-      "epoch": 2.8866396761133606,
-      "grad_norm": 0.35455056008719266,
-      "learning_rate": 1.810826238483892e-05,
-      "loss": 0.2812,
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.37606984624863327,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3168,
       "step": 713
     },
     {
-      "epoch": 2.8906882591093117,
-      "grad_norm": 0.3165551709089881,
-      "learning_rate": 1.8051969254866608e-05,
-      "loss": 0.299,
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.30331086810556135,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.2474,
       "step": 714
     },
     {
-      "epoch": 2.8947368421052633,
-      "grad_norm": 0.3265329549199093,
-      "learning_rate": 1.7995691701296226e-05,
-      "loss": 0.2931,
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.4182463503798477,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3167,
       "step": 715
     },
     {
-      "epoch": 2.898785425101215,
-      "grad_norm": 0.28824552074803206,
-      "learning_rate": 1.7939430174121593e-05,
-      "loss": 0.2865,
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.32121121262837876,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.2699,
       "step": 716
     },
     {
-      "epoch": 2.902834008097166,
-      "grad_norm": 0.3074387181952338,
-      "learning_rate": 1.788318512320838e-05,
-      "loss": 0.2762,
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.388934107624758,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.2828,
       "step": 717
     },
     {
-      "epoch": 2.9068825910931175,
-      "grad_norm": 0.3200340433397477,
-      "learning_rate": 1.782695699829051e-05,
-      "loss": 0.274,
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.3343410609177138,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.2783,
       "step": 718
     },
     {
-      "epoch": 2.910931174089069,
-      "grad_norm": 0.29258745434936645,
-      "learning_rate": 1.7770746248966587e-05,
-      "loss": 0.2528,
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.35478551369844147,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.2717,
       "step": 719
     },
     {
-      "epoch": 2.91497975708502,
-      "grad_norm": 0.29728458773654415,
-      "learning_rate": 1.7714553324696243e-05,
-      "loss": 0.2752,
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.34965251001999614,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.2628,
       "step": 720
     },
     {
-      "epoch": 2.919028340080972,
-      "grad_norm": 0.3276778262297052,
-      "learning_rate": 1.7658378674796614e-05,
-      "loss": 0.2707,
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.3659394279282846,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.2721,
       "step": 721
     },
     {
-      "epoch": 2.9230769230769234,
-      "grad_norm": 0.32318242520787066,
-      "learning_rate": 1.76022227484387e-05,
-      "loss": 0.2833,
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.387394154035108,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.2937,
       "step": 722
     },
     {
-      "epoch": 2.9271255060728745,
-      "grad_norm": 0.328356077488891,
-      "learning_rate": 1.7546085994643807e-05,
-      "loss": 0.2774,
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.34966977342594396,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.2673,
       "step": 723
     },
     {
-      "epoch": 2.931174089068826,
-      "grad_norm": 0.29455534790365795,
-      "learning_rate": 1.7489968862279902e-05,
-      "loss": 0.2782,
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.42172612721604574,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3018,
       "step": 724
     },
     {
-      "epoch": 2.9352226720647776,
-      "grad_norm": 0.3221970617475892,
-      "learning_rate": 1.743387180005809e-05,
-      "loss": 0.2695,
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.3545010712860689,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.2815,
       "step": 725
     },
     {
-      "epoch": 2.9392712550607287,
-      "grad_norm": 0.33010471483802134,
-      "learning_rate": 1.737779525652899e-05,
-      "loss": 0.288,
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.3528236287028636,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2638,
       "step": 726
     },
     {
-      "epoch": 2.9433198380566803,
-      "grad_norm": 0.2870912479917241,
-      "learning_rate": 1.732173968007914e-05,
-      "loss": 0.3083,
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.4020107282387961,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3075,
       "step": 727
     },
     {
-      "epoch": 2.9473684210526314,
-      "grad_norm": 0.31805028673324415,
-      "learning_rate": 1.7265705518927437e-05,
-      "loss": 0.2665,
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.347893536407178,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.2816,
       "step": 728
     },
     {
-      "epoch": 2.951417004048583,
-      "grad_norm": 0.30099674657213504,
-      "learning_rate": 1.7209693221121542e-05,
-      "loss": 0.2776,
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.4018438494290215,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.2956,
       "step": 729
     },
     {
-      "epoch": 2.9554655870445345,
-      "grad_norm": 0.2892831129382038,
-      "learning_rate": 1.7153703234534302e-05,
-      "loss": 0.2943,
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.3595472096136101,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.298,
       "step": 730
     },
     {
-      "epoch": 2.9595141700404857,
-      "grad_norm": 0.31443971312690117,
-      "learning_rate": 1.709773600686015e-05,
-      "loss": 0.2913,
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.34243842282928755,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.2495,
       "step": 731
     },
     {
-      "epoch": 2.9635627530364372,
-      "grad_norm": 0.28448245147624424,
-      "learning_rate": 1.7041791985611548e-05,
-      "loss": 0.2993,
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.3464411698525787,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3105,
       "step": 732
     },
     {
-      "epoch": 2.967611336032389,
-      "grad_norm": 0.29200784647950095,
-      "learning_rate": 1.69858716181154e-05,
-      "loss": 0.2996,
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.34714613505077985,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.2874,
       "step": 733
     },
     {
-      "epoch": 2.97165991902834,
-      "grad_norm": 0.2893709618773031,
-      "learning_rate": 1.692997535150948e-05,
-      "loss": 0.2713,
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.3740363357313951,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.2933,
       "step": 734
     },
     {
-      "epoch": 2.9757085020242915,
-      "grad_norm": 0.3060862606332107,
-      "learning_rate": 1.687410363273884e-05,
-      "loss": 0.3135,
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.37356597425900867,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2801,
       "step": 735
     },
     {
-      "epoch": 2.979757085020243,
-      "grad_norm": 0.3013311307722476,
-      "learning_rate": 1.6818256908552257e-05,
-      "loss": 0.2889,
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.3361851992803513,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.2993,
       "step": 736
     },
     {
-      "epoch": 2.983805668016194,
-      "grad_norm": 0.30334938563931374,
-      "learning_rate": 1.6762435625498646e-05,
-      "loss": 0.2778,
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.42028589259894766,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3065,
       "step": 737
     },
     {
-      "epoch": 2.9878542510121457,
-      "grad_norm": 0.3195483831625941,
-      "learning_rate": 1.67066402299235e-05,
-      "loss": 0.2845,
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.3071700416488511,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2552,
       "step": 738
     },
     {
-      "epoch": 2.9919028340080973,
-      "grad_norm": 0.3086434713582764,
-      "learning_rate": 1.6650871167965313e-05,
-      "loss": 0.2726,
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.35611592311540813,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.2841,
       "step": 739
     },
     {
-      "epoch": 2.9959514170040484,
-      "grad_norm": 0.3204254827572802,
-      "learning_rate": 1.6595128885552028e-05,
-      "loss": 0.2957,
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.3647689086158393,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.304,
       "step": 740
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 0.32891311390446054,
-      "learning_rate": 1.653941382839745e-05,
-      "loss": 0.2701,
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.401309643919786,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.2726,
       "step": 741
     },
     {
-      "epoch": 3.0040485829959516,
-      "grad_norm": 0.45397855295747724,
-      "learning_rate": 1.6483726441997673e-05,
-      "loss": 0.2174,
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.4494879446247661,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2086,
       "step": 742
     },
     {
-      "epoch": 3.0080971659919027,
-      "grad_norm": 0.3230004679933198,
-      "learning_rate": 1.642806717162757e-05,
-      "loss": 0.2053,
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.35381986287484263,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1998,
       "step": 743
     },
     {
-      "epoch": 3.0121457489878543,
-      "grad_norm": 0.7769445241833571,
-      "learning_rate": 1.637243646233718e-05,
-      "loss": 0.2191,
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.6571210713631405,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2126,
       "step": 744
     },
     {
-      "epoch": 3.016194331983806,
-      "grad_norm": 0.35891993736557254,
-      "learning_rate": 1.6316834758948174e-05,
-      "loss": 0.2032,
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.38493755101395094,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1762,
       "step": 745
     },
     {
-      "epoch": 3.020242914979757,
-      "grad_norm": 0.5183943935505201,
-      "learning_rate": 1.6261262506050282e-05,
-      "loss": 0.1966,
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.43153963795665495,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1932,
       "step": 746
     },
     {
-      "epoch": 3.0242914979757085,
-      "grad_norm": 0.43546331801858507,
-      "learning_rate": 1.620572014799777e-05,
-      "loss": 0.2168,
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.4556834033422859,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.2118,
       "step": 747
     },
     {
-      "epoch": 3.02834008097166,
-      "grad_norm": 0.34984697945784105,
-      "learning_rate": 1.6150208128905857e-05,
-      "loss": 0.1822,
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.38457269229223884,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1792,
       "step": 748
     },
     {
-      "epoch": 3.032388663967611,
-      "grad_norm": 0.37966508783495206,
-      "learning_rate": 1.6094726892647147e-05,
-      "loss": 0.1962,
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.3827861040724339,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.192,
       "step": 749
     },
     {
-      "epoch": 3.0364372469635628,
-      "grad_norm": 0.39845492803884436,
-      "learning_rate": 1.6039276882848135e-05,
-      "loss": 0.1896,
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.40422200285693727,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1863,
       "step": 750
     },
     {
-      "epoch": 3.0404858299595143,
-      "grad_norm": 0.3520524508492352,
-      "learning_rate": 1.5983858542885635e-05,
-      "loss": 0.1955,
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.39608316826023104,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2117,
       "step": 751
     },
     {
-      "epoch": 3.0445344129554655,
-      "grad_norm": 0.35452107501027325,
-      "learning_rate": 1.5928472315883203e-05,
-      "loss": 0.2067,
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.37506560895077706,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1959,
       "step": 752
     },
     {
-      "epoch": 3.048582995951417,
-      "grad_norm": 0.32597728347781635,
-      "learning_rate": 1.5873118644707633e-05,
-      "loss": 0.2182,
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.3239979617907035,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.187,
       "step": 753
     },
     {
-      "epoch": 3.0526315789473686,
-      "grad_norm": 0.32651177343890075,
-      "learning_rate": 1.5817797971965413e-05,
-      "loss": 0.1901,
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.3526233318512141,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1885,
       "step": 754
     },
     {
-      "epoch": 3.0566801619433197,
-      "grad_norm": 0.3491731073227955,
-      "learning_rate": 1.576251073999917e-05,
-      "loss": 0.2047,
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.3643454883397354,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2096,
       "step": 755
     },
     {
-      "epoch": 3.0607287449392713,
-      "grad_norm": 0.3132879188170738,
-      "learning_rate": 1.5707257390884126e-05,
-      "loss": 0.1975,
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.3340850194985118,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.19,
       "step": 756
     },
     {
-      "epoch": 3.064777327935223,
-      "grad_norm": 0.33778010791898205,
-      "learning_rate": 1.5652038366424595e-05,
-      "loss": 0.196,
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.39603528545617583,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1922,
       "step": 757
     },
     {
-      "epoch": 3.068825910931174,
-      "grad_norm": 0.3451925819630269,
-      "learning_rate": 1.5596854108150424e-05,
-      "loss": 0.1886,
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.32947913485758346,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1834,
       "step": 758
     },
     {
-      "epoch": 3.0728744939271255,
-      "grad_norm": 0.3230287642198081,
-      "learning_rate": 1.5541705057313476e-05,
-      "loss": 0.2017,
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.3447655936199811,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1906,
       "step": 759
     },
     {
-      "epoch": 3.076923076923077,
-      "grad_norm": 0.304030138321163,
-      "learning_rate": 1.5486591654884086e-05,
-      "loss": 0.1959,
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.3490837474095369,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1837,
       "step": 760
     },
     {
-      "epoch": 3.080971659919028,
-      "grad_norm": 0.3498427938669405,
-      "learning_rate": 1.543151434154755e-05,
-      "loss": 0.1929,
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.3252215490168291,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1794,
       "step": 761
     },
     {
-      "epoch": 3.08502024291498,
-      "grad_norm": 0.30783253736037053,
-      "learning_rate": 1.5376473557700615e-05,
-      "loss": 0.1957,
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.3506484502188549,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2067,
       "step": 762
     },
     {
-      "epoch": 3.0890688259109313,
-      "grad_norm": 0.30670827521454014,
-      "learning_rate": 1.5321469743447893e-05,
-      "loss": 0.1993,
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.313253311171457,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1985,
       "step": 763
     },
     {
-      "epoch": 3.0931174089068825,
-      "grad_norm": 0.31504596724053113,
-      "learning_rate": 1.5266503338598434e-05,
-      "loss": 0.208,
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.3195519625607016,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.181,
       "step": 764
     },
     {
-      "epoch": 3.097165991902834,
-      "grad_norm": 0.2979200229637876,
-      "learning_rate": 1.5211574782662149e-05,
-      "loss": 0.1942,
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.31497916208230764,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1875,
       "step": 765
     },
     {
-      "epoch": 3.1012145748987856,
-      "grad_norm": 0.3148478195568172,
-      "learning_rate": 1.5156684514846307e-05,
-      "loss": 0.1875,
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.3214370030018471,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1937,
       "step": 766
     },
     {
-      "epoch": 3.1052631578947367,
-      "grad_norm": 0.31789615151728173,
-      "learning_rate": 1.5101832974052012e-05,
-      "loss": 0.1931,
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.3280108037126933,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1819,
       "step": 767
     },
     {
-      "epoch": 3.1093117408906883,
-      "grad_norm": 0.3104904446413981,
-      "learning_rate": 1.5047020598870737e-05,
-      "loss": 0.1993,
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.3281325064357066,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.195,
       "step": 768
     },
     {
-      "epoch": 3.11336032388664,
-      "grad_norm": 0.31121932532888075,
-      "learning_rate": 1.4992247827580778e-05,
-      "loss": 0.1841,
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.3633544380453866,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2028,
       "step": 769
     },
     {
-      "epoch": 3.117408906882591,
-      "grad_norm": 0.32450791274084767,
-      "learning_rate": 1.493751509814374e-05,
-      "loss": 0.1944,
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.30386641627069044,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1679,
       "step": 770
     },
     {
-      "epoch": 3.1214574898785425,
-      "grad_norm": 0.2946525586143942,
-      "learning_rate": 1.4882822848201073e-05,
-      "loss": 0.1844,
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.33964352912661855,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1955,
       "step": 771
     },
     {
-      "epoch": 3.125506072874494,
-      "grad_norm": 0.327447480860331,
-      "learning_rate": 1.4828171515070553e-05,
-      "loss": 0.1856,
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.3334583956202426,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1589,
       "step": 772
     },
     {
-      "epoch": 3.1295546558704452,
-      "grad_norm": 0.3119904652586277,
-      "learning_rate": 1.4773561535742793e-05,
-      "loss": 0.2182,
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.3464035542502072,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2142,
       "step": 773
     },
     {
-      "epoch": 3.133603238866397,
-      "grad_norm": 0.30453350804973056,
-      "learning_rate": 1.4718993346877715e-05,
-      "loss": 0.2134,
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.3543989297743814,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2043,
       "step": 774
     },
     {
-      "epoch": 3.1376518218623484,
-      "grad_norm": 0.3041910219243991,
-      "learning_rate": 1.466446738480111e-05,
-      "loss": 0.1955,
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.3280051263906032,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1863,
       "step": 775
     },
     {
-      "epoch": 3.1417004048582995,
-      "grad_norm": 0.3120429610222733,
-      "learning_rate": 1.460998408550112e-05,
-      "loss": 0.2098,
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.3421642474344214,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2002,
       "step": 776
     },
     {
-      "epoch": 3.145748987854251,
-      "grad_norm": 0.3138204380801465,
-      "learning_rate": 1.4555543884624751e-05,
-      "loss": 0.1918,
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.3271810313746488,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1882,
       "step": 777
     },
     {
-      "epoch": 3.1497975708502026,
-      "grad_norm": 0.28954652272922066,
-      "learning_rate": 1.4501147217474402e-05,
-      "loss": 0.2198,
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.3273951186875465,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1916,
       "step": 778
     },
     {
-      "epoch": 3.1538461538461537,
-      "grad_norm": 0.3185035566111182,
-      "learning_rate": 1.444679451900437e-05,
-      "loss": 0.208,
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.3279943208981525,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2165,
       "step": 779
     },
     {
-      "epoch": 3.1578947368421053,
-      "grad_norm": 0.2917474933724308,
-      "learning_rate": 1.4392486223817397e-05,
-      "loss": 0.2171,
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.31562835087780816,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1927,
       "step": 780
     },
     {
-      "epoch": 3.161943319838057,
-      "grad_norm": 0.3199279442836199,
-      "learning_rate": 1.4338222766161143e-05,
-      "loss": 0.2088,
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.303071191813408,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1843,
       "step": 781
     },
     {
-      "epoch": 3.165991902834008,
-      "grad_norm": 0.2937564864833695,
-      "learning_rate": 1.4284004579924777e-05,
-      "loss": 0.2001,
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.32843643189392013,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2031,
       "step": 782
     },
     {
-      "epoch": 3.1700404858299596,
-      "grad_norm": 0.29450200875990146,
-      "learning_rate": 1.4229832098635479e-05,
-      "loss": 0.207,
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.3602073809373371,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.2052,
       "step": 783
     },
     {
-      "epoch": 3.174089068825911,
-      "grad_norm": 0.3134830004391043,
-      "learning_rate": 1.4175705755454963e-05,
-      "loss": 0.1928,
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.31321416662455503,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1771,
       "step": 784
     },
     {
-      "epoch": 3.1781376518218623,
-      "grad_norm": 0.29819359487818803,
-      "learning_rate": 1.412162598317602e-05,
-      "loss": 0.1973,
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.3503656228243467,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2161,
       "step": 785
     },
     {
-      "epoch": 3.182186234817814,
-      "grad_norm": 0.31620777943210787,
-      "learning_rate": 1.406759321421907e-05,
-      "loss": 0.2152,
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.32715290923669305,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1925,
       "step": 786
     },
     {
-      "epoch": 3.1862348178137654,
-      "grad_norm": 0.28873817888732295,
-      "learning_rate": 1.401360788062871e-05,
-      "loss": 0.2013,
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.29752128349034096,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1752,
       "step": 787
     },
     {
-      "epoch": 3.1902834008097165,
-      "grad_norm": 0.31987064893416567,
-      "learning_rate": 1.3959670414070204e-05,
-      "loss": 0.1977,
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.31972021759865543,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1819,
       "step": 788
     },
     {
-      "epoch": 3.194331983805668,
-      "grad_norm": 0.3150811021968869,
-      "learning_rate": 1.3905781245826108e-05,
-      "loss": 0.1933,
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.35726438164714275,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1963,
       "step": 789
     },
     {
-      "epoch": 3.1983805668016196,
-      "grad_norm": 0.30641824778825566,
-      "learning_rate": 1.3851940806792778e-05,
-      "loss": 0.1915,
+      "epoch": 3.2,
+      "grad_norm": 0.32657403835389953,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1875,
       "step": 790
     },
     {
-      "epoch": 3.2024291497975708,
-      "grad_norm": 0.3238030888299841,
-      "learning_rate": 1.379814952747693e-05,
-      "loss": 0.1953,
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.36686751256191136,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.1827,
       "step": 791
     },
     {
-      "epoch": 3.2064777327935223,
-      "grad_norm": 0.30635107856165167,
-      "learning_rate": 1.3744407837992193e-05,
-      "loss": 0.1869,
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.3061089433884727,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1795,
       "step": 792
     },
     {
-      "epoch": 3.2105263157894735,
-      "grad_norm": 0.29226429840977736,
-      "learning_rate": 1.3690716168055692e-05,
-      "loss": 0.2048,
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.30810380778079943,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.187,
       "step": 793
     },
     {
-      "epoch": 3.214574898785425,
-      "grad_norm": 0.30149983563407895,
-      "learning_rate": 1.363707494698459e-05,
-      "loss": 0.2099,
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.3572034008758245,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2043,
       "step": 794
     },
     {
-      "epoch": 3.2186234817813766,
-      "grad_norm": 0.3194632868479304,
-      "learning_rate": 1.358348460369265e-05,
-      "loss": 0.1899,
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.3117954977640782,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1932,
       "step": 795
     },
     {
-      "epoch": 3.2226720647773277,
-      "grad_norm": 0.3036200385032639,
-      "learning_rate": 1.352994556668684e-05,
-      "loss": 0.211,
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.3466191479430294,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2103,
       "step": 796
     },
     {
-      "epoch": 3.2267206477732793,
-      "grad_norm": 0.29748891331028876,
-      "learning_rate": 1.3476458264063865e-05,
-      "loss": 0.2007,
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.3082684496422822,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1861,
       "step": 797
     },
     {
-      "epoch": 3.230769230769231,
-      "grad_norm": 0.2880672419946549,
-      "learning_rate": 1.342302312350678e-05,
-      "loss": 0.1925,
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.29947712149768213,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1713,
       "step": 798
     },
     {
-      "epoch": 3.234817813765182,
-      "grad_norm": 0.3007241466618202,
-      "learning_rate": 1.3369640572281537e-05,
-      "loss": 0.1861,
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.406977376750653,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2089,
       "step": 799
     },
     {
-      "epoch": 3.2388663967611335,
-      "grad_norm": 0.3236179949362517,
-      "learning_rate": 1.3316311037233596e-05,
-      "loss": 0.1895,
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.3173709159412448,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1884,
       "step": 800
     },
     {
-      "epoch": 3.242914979757085,
-      "grad_norm": 0.28658455623369705,
-      "learning_rate": 1.326303494478451e-05,
-      "loss": 0.189,
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3036748862583886,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1685,
       "step": 801
     },
     {
-      "epoch": 3.246963562753036,
-      "grad_norm": 0.2949631480913927,
-      "learning_rate": 1.3209812720928472e-05,
-      "loss": 0.195,
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.3378481692454428,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1951,
       "step": 802
     },
     {
-      "epoch": 3.251012145748988,
-      "grad_norm": 0.3011943861934074,
-      "learning_rate": 1.315664479122898e-05,
-      "loss": 0.1992,
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.3423786435754868,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1947,
       "step": 803
     },
     {
-      "epoch": 3.2550607287449393,
-      "grad_norm": 0.29791488115377485,
-      "learning_rate": 1.3103531580815378e-05,
-      "loss": 0.2104,
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.310644923713298,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1793,
       "step": 804
     },
     {
-      "epoch": 3.2591093117408905,
-      "grad_norm": 0.2909510519198696,
-      "learning_rate": 1.3050473514379488e-05,
-      "loss": 0.1957,
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.3375710704663777,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1871,
       "step": 805
     },
     {
-      "epoch": 3.263157894736842,
-      "grad_norm": 0.29409980427739046,
-      "learning_rate": 1.2997471016172188e-05,
-      "loss": 0.191,
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.3241117553714247,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1985,
       "step": 806
     },
     {
-      "epoch": 3.2672064777327936,
-      "grad_norm": 0.27009641196688117,
-      "learning_rate": 1.2944524510000042e-05,
-      "loss": 0.1973,
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.31781233005326803,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1954,
       "step": 807
     },
     {
-      "epoch": 3.2712550607287447,
-      "grad_norm": 0.280677022289714,
-      "learning_rate": 1.2891634419221908e-05,
-      "loss": 0.2138,
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.3001392963396394,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1647,
       "step": 808
     },
     {
-      "epoch": 3.2753036437246963,
-      "grad_norm": 0.289290898504646,
-      "learning_rate": 1.2838801166745545e-05,
-      "loss": 0.2021,
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.33180034602268044,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2065,
       "step": 809
     },
     {
-      "epoch": 3.279352226720648,
-      "grad_norm": 0.2867772244425207,
-      "learning_rate": 1.2786025175024216e-05,
-      "loss": 0.1854,
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.3138339001813041,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1906,
       "step": 810
     },
     {
-      "epoch": 3.283400809716599,
-      "grad_norm": 0.2776025723100309,
-      "learning_rate": 1.2733306866053357e-05,
-      "loss": 0.2017,
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.33717246587815036,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2073,
       "step": 811
     },
     {
-      "epoch": 3.2874493927125505,
-      "grad_norm": 0.26475258934409956,
-      "learning_rate": 1.2680646661367163e-05,
-      "loss": 0.2196,
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.28423208679217404,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1953,
       "step": 812
     },
     {
-      "epoch": 3.291497975708502,
-      "grad_norm": 0.2916358419355548,
-      "learning_rate": 1.2628044982035219e-05,
-      "loss": 0.2141,
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.30851047673086524,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1931,
       "step": 813
     },
     {
-      "epoch": 3.2955465587044532,
-      "grad_norm": 0.2698706864496747,
-      "learning_rate": 1.2575502248659155e-05,
-      "loss": 0.2275,
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.3145171436431576,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2309,
       "step": 814
     },
     {
-      "epoch": 3.299595141700405,
-      "grad_norm": 0.2959666138828329,
-      "learning_rate": 1.2523018881369269e-05,
-      "loss": 0.1992,
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.3223081171004941,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1727,
       "step": 815
     },
     {
-      "epoch": 3.3036437246963564,
-      "grad_norm": 0.2894142322372878,
-      "learning_rate": 1.2470595299821176e-05,
-      "loss": 0.1973,
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.3228554485027876,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1954,
       "step": 816
     },
     {
-      "epoch": 3.3076923076923075,
-      "grad_norm": 0.3049439876705841,
-      "learning_rate": 1.2418231923192416e-05,
-      "loss": 0.2212,
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.33149853624937964,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2085,
       "step": 817
     },
     {
-      "epoch": 3.311740890688259,
-      "grad_norm": 0.290539562173047,
-      "learning_rate": 1.2365929170179165e-05,
-      "loss": 0.1943,
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.33239332776375285,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2072,
       "step": 818
     },
     {
-      "epoch": 3.3157894736842106,
-      "grad_norm": 0.29454856653042844,
-      "learning_rate": 1.2313687458992848e-05,
-      "loss": 0.2148,
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.3151556169006731,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.184,
       "step": 819
     },
     {
-      "epoch": 3.3198380566801617,
-      "grad_norm": 0.30081135251703717,
-      "learning_rate": 1.2261507207356785e-05,
-      "loss": 0.2002,
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.36122777980596416,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2048,
       "step": 820
     },
     {
-      "epoch": 3.3238866396761133,
-      "grad_norm": 0.3084815756698952,
-      "learning_rate": 1.220938883250288e-05,
-      "loss": 0.2068,
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.3391039648143441,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.204,
       "step": 821
     },
     {
-      "epoch": 3.327935222672065,
-      "grad_norm": 0.2817559325629289,
-      "learning_rate": 1.215733275116827e-05,
-      "loss": 0.1935,
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.29154332300476044,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.17,
       "step": 822
     },
     {
-      "epoch": 3.331983805668016,
-      "grad_norm": 0.28497437379844903,
-      "learning_rate": 1.2105339379592005e-05,
-      "loss": 0.2134,
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.37247782957764125,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2017,
       "step": 823
     },
     {
-      "epoch": 3.3360323886639676,
-      "grad_norm": 0.28926267870346506,
-      "learning_rate": 1.2053409133511681e-05,
-      "loss": 0.1974,
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.3128434680774391,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1996,
       "step": 824
     },
     {
-      "epoch": 3.340080971659919,
-      "grad_norm": 0.2833560430025525,
-      "learning_rate": 1.2001542428160173e-05,
-      "loss": 0.2095,
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.31596160079454044,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1838,
       "step": 825
     },
     {
-      "epoch": 3.3441295546558703,
-      "grad_norm": 0.2828257532966344,
-      "learning_rate": 1.1949739678262278e-05,
-      "loss": 0.1916,
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.3514880734477051,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1926,
       "step": 826
     },
     {
-      "epoch": 3.348178137651822,
-      "grad_norm": 0.28837917170566474,
-      "learning_rate": 1.189800129803139e-05,
-      "loss": 0.1829,
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.29629312060832774,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1742,
       "step": 827
     },
     {
-      "epoch": 3.3522267206477734,
-      "grad_norm": 0.2886783537383735,
-      "learning_rate": 1.1846327701166228e-05,
-      "loss": 0.2,
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.3725673122948238,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2032,
       "step": 828
     },
     {
-      "epoch": 3.3562753036437245,
-      "grad_norm": 0.314078598312296,
-      "learning_rate": 1.1794719300847493e-05,
-      "loss": 0.1929,
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.3179295018010224,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1869,
       "step": 829
     },
     {
-      "epoch": 3.360323886639676,
-      "grad_norm": 0.30940239373763345,
-      "learning_rate": 1.1743176509734582e-05,
-      "loss": 0.215,
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.35456916784028575,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2074,
       "step": 830
     },
     {
-      "epoch": 3.3643724696356276,
-      "grad_norm": 0.2844507895522093,
-      "learning_rate": 1.1691699739962275e-05,
-      "loss": 0.19,
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.2816457152404054,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1675,
       "step": 831
     },
     {
-      "epoch": 3.3684210526315788,
-      "grad_norm": 0.3009549721986297,
-      "learning_rate": 1.1640289403137456e-05,
-      "loss": 0.2076,
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.3415124872732163,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2182,
       "step": 832
     },
     {
-      "epoch": 3.3724696356275303,
-      "grad_norm": 0.29152949485848767,
-      "learning_rate": 1.1588945910335802e-05,
-      "loss": 0.2088,
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.3052600764423698,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1836,
       "step": 833
     },
     {
-      "epoch": 3.376518218623482,
-      "grad_norm": 0.2835637112632266,
-      "learning_rate": 1.1537669672098517e-05,
-      "loss": 0.2091,
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.34908442255408284,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2043,
       "step": 834
     },
     {
-      "epoch": 3.380566801619433,
-      "grad_norm": 0.29518292238560045,
-      "learning_rate": 1.1486461098429036e-05,
-      "loss": 0.2037,
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.31787067969809163,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2075,
       "step": 835
     },
     {
-      "epoch": 3.3846153846153846,
-      "grad_norm": 0.2911065056607521,
-      "learning_rate": 1.1435320598789747e-05,
-      "loss": 0.2088,
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.2971433762374366,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1949,
       "step": 836
     },
     {
-      "epoch": 3.388663967611336,
-      "grad_norm": 0.28918043860597553,
-      "learning_rate": 1.1384248582098736e-05,
-      "loss": 0.2,
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.33136975313195804,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.185,
       "step": 837
     },
     {
-      "epoch": 3.3927125506072873,
-      "grad_norm": 0.31006639592233204,
-      "learning_rate": 1.1333245456726467e-05,
-      "loss": 0.1947,
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.3584672145064533,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1873,
       "step": 838
     },
     {
-      "epoch": 3.396761133603239,
-      "grad_norm": 0.2979993522979623,
-      "learning_rate": 1.128231163049258e-05,
-      "loss": 0.1885,
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.3348274763473414,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.205,
       "step": 839
     },
     {
-      "epoch": 3.4008097165991904,
-      "grad_norm": 0.2868400887014694,
-      "learning_rate": 1.123144751066259e-05,
-      "loss": 0.2077,
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.31005174635379884,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1706,
       "step": 840
     },
     {
-      "epoch": 3.4048582995951415,
-      "grad_norm": 0.2872838729043886,
-      "learning_rate": 1.1180653503944657e-05,
-      "loss": 0.1969,
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.3568890761984914,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1936,
       "step": 841
     },
     {
-      "epoch": 3.408906882591093,
-      "grad_norm": 0.31418360353032854,
-      "learning_rate": 1.1129930016486287e-05,
-      "loss": 0.2021,
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.28907935627872994,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1762,
       "step": 842
     },
     {
-      "epoch": 3.4129554655870447,
-      "grad_norm": 0.27665007104581235,
-      "learning_rate": 1.107927745387114e-05,
-      "loss": 0.2078,
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.32458059949344953,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2063,
       "step": 843
     },
     {
-      "epoch": 3.417004048582996,
-      "grad_norm": 0.2965303502565838,
-      "learning_rate": 1.1028696221115766e-05,
-      "loss": 0.2111,
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.34332649514479757,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1967,
       "step": 844
     },
     {
-      "epoch": 3.4210526315789473,
-      "grad_norm": 0.290002038241745,
-      "learning_rate": 1.0978186722666341e-05,
-      "loss": 0.2026,
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.2786500854730323,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1746,
       "step": 845
     },
     {
-      "epoch": 3.425101214574899,
-      "grad_norm": 0.2814487807185271,
-      "learning_rate": 1.0927749362395475e-05,
-      "loss": 0.1948,
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.3311157392292299,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2072,
       "step": 846
     },
     {
-      "epoch": 3.42914979757085,
-      "grad_norm": 0.2822870815945768,
-      "learning_rate": 1.0877384543598954e-05,
-      "loss": 0.1954,
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.30390316626074027,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1874,
       "step": 847
     },
     {
-      "epoch": 3.4331983805668016,
-      "grad_norm": 0.2701306828209743,
-      "learning_rate": 1.0827092668992529e-05,
-      "loss": 0.2128,
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.30774439446802043,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.203,
       "step": 848
     },
     {
-      "epoch": 3.437246963562753,
-      "grad_norm": 0.282987086906909,
-      "learning_rate": 1.077687414070869e-05,
-      "loss": 0.1893,
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.3160892865510697,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1824,
       "step": 849
     },
     {
-      "epoch": 3.4412955465587043,
-      "grad_norm": 0.2992661097327179,
-      "learning_rate": 1.0726729360293451e-05,
-      "loss": 0.1805,
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.3308126971387281,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1806,
       "step": 850
     },
     {
-      "epoch": 3.445344129554656,
-      "grad_norm": 0.2969196146962342,
-      "learning_rate": 1.0676658728703147e-05,
-      "loss": 0.2041,
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.3195998826434039,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1963,
       "step": 851
     },
     {
-      "epoch": 3.4493927125506074,
-      "grad_norm": 0.27963807284910586,
-      "learning_rate": 1.0626662646301198e-05,
-      "loss": 0.2055,
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.2917501342813611,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1704,
       "step": 852
     },
     {
-      "epoch": 3.4534412955465585,
-      "grad_norm": 0.31320162513604966,
-      "learning_rate": 1.0576741512854958e-05,
-      "loss": 0.2018,
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.36026427134491057,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2285,
       "step": 853
     },
     {
-      "epoch": 3.45748987854251,
-      "grad_norm": 0.31148516086049327,
-      "learning_rate": 1.052689572753247e-05,
-      "loss": 0.2033,
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.29912877103625146,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1856,
       "step": 854
     },
     {
-      "epoch": 3.4615384615384617,
-      "grad_norm": 0.284553384319045,
-      "learning_rate": 1.0477125688899328e-05,
-      "loss": 0.2159,
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.3169965387444753,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.208,
       "step": 855
     },
     {
-      "epoch": 3.465587044534413,
-      "grad_norm": 0.2754243759412652,
-      "learning_rate": 1.0427431794915406e-05,
-      "loss": 0.2049,
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.310433406732971,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.181,
       "step": 856
     },
     {
-      "epoch": 3.4696356275303644,
-      "grad_norm": 0.28233943619382645,
-      "learning_rate": 1.0377814442931769e-05,
-      "loss": 0.1955,
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.3004522245276902,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1946,
       "step": 857
     },
     {
-      "epoch": 3.473684210526316,
-      "grad_norm": 0.291723379354529,
-      "learning_rate": 1.0328274029687439e-05,
-      "loss": 0.1923,
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.3097588823988763,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1941,
       "step": 858
     },
     {
-      "epoch": 3.477732793522267,
-      "grad_norm": 0.2860832815631675,
-      "learning_rate": 1.0278810951306253e-05,
-      "loss": 0.2092,
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.31820842468511207,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.196,
       "step": 859
     },
     {
-      "epoch": 3.4817813765182186,
-      "grad_norm": 0.3030383542123505,
-      "learning_rate": 1.0229425603293647e-05,
-      "loss": 0.215,
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.33356981423137827,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2165,
       "step": 860
     },
     {
-      "epoch": 3.48582995951417,
-      "grad_norm": 0.2814512735138987,
-      "learning_rate": 1.0180118380533558e-05,
-      "loss": 0.2129,
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.28800031979569146,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.171,
       "step": 861
     },
     {
-      "epoch": 3.4898785425101213,
-      "grad_norm": 0.2757783484516759,
-      "learning_rate": 1.0130889677285225e-05,
-      "loss": 0.1844,
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.33065718033134617,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2055,
       "step": 862
     },
     {
-      "epoch": 3.493927125506073,
-      "grad_norm": 0.28505563771778436,
-      "learning_rate": 1.008173988718005e-05,
-      "loss": 0.1844,
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.31291653519652063,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1744,
       "step": 863
     },
     {
-      "epoch": 3.4979757085020244,
-      "grad_norm": 0.2962476223740099,
-      "learning_rate": 1.0032669403218439e-05,
-      "loss": 0.2078,
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.31192393862117695,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.197,
       "step": 864
     },
     {
-      "epoch": 3.5020242914979756,
-      "grad_norm": 0.2848598249002067,
-      "learning_rate": 9.983678617766677e-06,
-      "loss": 0.1982,
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.30411125682895074,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1864,
       "step": 865
     },
     {
-      "epoch": 3.506072874493927,
-      "grad_norm": 0.281237879486758,
-      "learning_rate": 9.934767922553788e-06,
-      "loss": 0.2192,
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.30448389622791167,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2062,
       "step": 866
     },
     {
-      "epoch": 3.5101214574898787,
-      "grad_norm": 0.27796071224267127,
-      "learning_rate": 9.885937708668371e-06,
-      "loss": 0.2078,
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.27657187249138393,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1897,
       "step": 867
     },
     {
-      "epoch": 3.51417004048583,
-      "grad_norm": 0.2850415137244732,
-      "learning_rate": 9.837188366555524e-06,
-      "loss": 0.1922,
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.3016953465130358,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1772,
       "step": 868
     },
     {
-      "epoch": 3.5182186234817814,
-      "grad_norm": 0.294609881889197,
-      "learning_rate": 9.788520286013691e-06,
-      "loss": 0.1789,
+      "epoch": 3.52,
+      "grad_norm": 0.3247755524259392,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1905,
       "step": 869
     },
     {
-      "epoch": 3.522267206477733,
-      "grad_norm": 0.29980267005437644,
-      "learning_rate": 9.739933856191537e-06,
-      "loss": 0.1901,
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.2983230668484799,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1952,
       "step": 870
     },
     {
-      "epoch": 3.526315789473684,
-      "grad_norm": 0.2919292062056388,
-      "learning_rate": 9.691429465584866e-06,
-      "loss": 0.2089,
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.2791566581949605,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1664,
       "step": 871
     },
     {
-      "epoch": 3.5303643724696356,
-      "grad_norm": 0.29908934523112,
-      "learning_rate": 9.643007502033492e-06,
-      "loss": 0.1919,
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.35017218642140213,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2249,
       "step": 872
     },
     {
-      "epoch": 3.534412955465587,
-      "grad_norm": 0.2952238513267275,
-      "learning_rate": 9.594668352718152e-06,
-      "loss": 0.1985,
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.28345241423673895,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1797,
       "step": 873
     },
     {
-      "epoch": 3.5384615384615383,
-      "grad_norm": 0.27759419736374785,
-      "learning_rate": 9.546412404157383e-06,
-      "loss": 0.1974,
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.31510783958162175,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.196,
       "step": 874
     },
     {
-      "epoch": 3.54251012145749,
-      "grad_norm": 0.295246189794401,
-      "learning_rate": 9.498240042204471e-06,
-      "loss": 0.1983,
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.2929298018229853,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1709,
       "step": 875
     },
     {
-      "epoch": 3.5465587044534415,
-      "grad_norm": 0.2862427026398589,
-      "learning_rate": 9.450151652044339e-06,
-      "loss": 0.2028,
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.32183200298485964,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2259,
       "step": 876
     },
     {
-      "epoch": 3.5506072874493926,
-      "grad_norm": 0.2906792317154318,
-      "learning_rate": 9.402147618190472e-06,
-      "loss": 0.1983,
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.29037135712232304,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1781,
       "step": 877
     },
     {
-      "epoch": 3.554655870445344,
-      "grad_norm": 0.272165761010565,
-      "learning_rate": 9.354228324481841e-06,
-      "loss": 0.1963,
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.2954694359343009,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2023,
       "step": 878
     },
     {
-      "epoch": 3.5587044534412957,
-      "grad_norm": 0.2617819351133759,
-      "learning_rate": 9.306394154079852e-06,
-      "loss": 0.203,
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.28392218674290415,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1817,
       "step": 879
     },
     {
-      "epoch": 3.562753036437247,
-      "grad_norm": 0.28370551764036966,
-      "learning_rate": 9.258645489465254e-06,
-      "loss": 0.1853,
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.3166746150962137,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.208,
       "step": 880
     },
     {
-      "epoch": 3.5668016194331984,
-      "grad_norm": 0.2824513455179297,
-      "learning_rate": 9.21098271243509e-06,
-      "loss": 0.2017,
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.26366872226419663,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1584,
       "step": 881
     },
     {
-      "epoch": 3.57085020242915,
-      "grad_norm": 0.28835743456769797,
-      "learning_rate": 9.163406204099659e-06,
-      "loss": 0.1944,
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.30064909486334657,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1866,
       "step": 882
     },
     {
-      "epoch": 3.574898785425101,
-      "grad_norm": 0.2835831226230746,
-      "learning_rate": 9.115916344879452e-06,
-      "loss": 0.1945,
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.29000392149457344,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1887,
       "step": 883
     },
     {
-      "epoch": 3.5789473684210527,
-      "grad_norm": 0.2961876863618075,
-      "learning_rate": 9.068513514502126e-06,
-      "loss": 0.2048,
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.31619291346988887,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2016,
       "step": 884
     },
     {
-      "epoch": 3.582995951417004,
-      "grad_norm": 0.3074452119492931,
-      "learning_rate": 9.021198091999439e-06,
-      "loss": 0.1979,
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.3152321705642546,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1902,
       "step": 885
     },
     {
-      "epoch": 3.5870445344129553,
-      "grad_norm": 0.2901781541282623,
-      "learning_rate": 8.973970455704251e-06,
-      "loss": 0.1814,
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.29210465260853147,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1811,
       "step": 886
     },
     {
-      "epoch": 3.591093117408907,
-      "grad_norm": 0.27904111785462254,
-      "learning_rate": 8.926830983247495e-06,
-      "loss": 0.1924,
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.2821742617092896,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1795,
       "step": 887
     },
     {
-      "epoch": 3.5951417004048585,
-      "grad_norm": 0.28702012626156315,
-      "learning_rate": 8.879780051555116e-06,
-      "loss": 0.2015,
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.30863737057768914,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2107,
       "step": 888
     },
     {
-      "epoch": 3.5991902834008096,
-      "grad_norm": 0.2824943921319869,
-      "learning_rate": 8.83281803684512e-06,
-      "loss": 0.2032,
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.2967985671369567,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1866,
       "step": 889
     },
     {
-      "epoch": 3.603238866396761,
-      "grad_norm": 0.30068424014733525,
-      "learning_rate": 8.785945314624532e-06,
-      "loss": 0.2015,
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.31083807726294765,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2044,
       "step": 890
     },
     {
-      "epoch": 3.6072874493927127,
-      "grad_norm": 0.2813194645302014,
-      "learning_rate": 8.739162259686385e-06,
-      "loss": 0.1948,
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.29925262315020507,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1839,
       "step": 891
     },
     {
-      "epoch": 3.611336032388664,
-      "grad_norm": 0.276736233185004,
-      "learning_rate": 8.692469246106734e-06,
-      "loss": 0.1884,
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.3058087682714933,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1937,
       "step": 892
     },
     {
-      "epoch": 3.6153846153846154,
-      "grad_norm": 0.29983677843975,
-      "learning_rate": 8.645866647241677e-06,
-      "loss": 0.2113,
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.2959978586081765,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1915,
       "step": 893
     },
     {
-      "epoch": 3.619433198380567,
-      "grad_norm": 0.2973205575882941,
-      "learning_rate": 8.599354835724348e-06,
-      "loss": 0.1916,
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.2994009310830393,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1918,
       "step": 894
     },
     {
-      "epoch": 3.623481781376518,
-      "grad_norm": 0.2793278061617111,
-      "learning_rate": 8.552934183461958e-06,
-      "loss": 0.2125,
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.30614594951124896,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1991,
       "step": 895
     },
     {
-      "epoch": 3.6275303643724697,
-      "grad_norm": 0.3142882188772036,
-      "learning_rate": 8.506605061632805e-06,
-      "loss": 0.2055,
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.2912263282848892,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1766,
       "step": 896
     },
     {
-      "epoch": 3.6315789473684212,
-      "grad_norm": 0.275226293927442,
-      "learning_rate": 8.460367840683308e-06,
-      "loss": 0.2083,
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.28702940481303857,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1911,
       "step": 897
     },
     {
-      "epoch": 3.6356275303643724,
-      "grad_norm": 0.2949560069566496,
-      "learning_rate": 8.414222890325065e-06,
-      "loss": 0.204,
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.30292694042839297,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1929,
       "step": 898
     },
     {
-      "epoch": 3.639676113360324,
-      "grad_norm": 0.2937673959520555,
-      "learning_rate": 8.368170579531855e-06,
-      "loss": 0.198,
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.3363721348732952,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2116,
       "step": 899
     },
     {
-      "epoch": 3.6437246963562755,
-      "grad_norm": 0.29809307724509315,
-      "learning_rate": 8.322211276536733e-06,
-      "loss": 0.1999,
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.30769324600182746,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1998,
       "step": 900
     },
     {
-      "epoch": 3.6477732793522266,
-      "grad_norm": 0.27973283270217686,
-      "learning_rate": 8.27634534882907e-06,
-      "loss": 0.2075,
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.2665452495739093,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.175,
       "step": 901
     },
     {
-      "epoch": 3.651821862348178,
-      "grad_norm": 0.2828287687529827,
-      "learning_rate": 8.230573163151585e-06,
-      "loss": 0.2004,
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.3309048251440096,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2076,
       "step": 902
     },
     {
-      "epoch": 3.6558704453441297,
-      "grad_norm": 0.2756938501607346,
-      "learning_rate": 8.184895085497466e-06,
-      "loss": 0.2006,
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.2982087067531819,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1961,
       "step": 903
     },
     {
-      "epoch": 3.659919028340081,
-      "grad_norm": 0.2842224375771387,
-      "learning_rate": 8.139311481107401e-06,
-      "loss": 0.1951,
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.2836320030639386,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1706,
       "step": 904
     },
     {
-      "epoch": 3.6639676113360324,
-      "grad_norm": 0.29923138518862696,
-      "learning_rate": 8.093822714466686e-06,
-      "loss": 0.1866,
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.35586942415067974,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.208,
       "step": 905
     },
     {
-      "epoch": 3.668016194331984,
-      "grad_norm": 0.2837420158393417,
-      "learning_rate": 8.048429149302271e-06,
-      "loss": 0.2226,
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.2815691121569045,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.187,
       "step": 906
     },
     {
-      "epoch": 3.672064777327935,
-      "grad_norm": 0.3071966502255487,
-      "learning_rate": 8.003131148579902e-06,
-      "loss": 0.2029,
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.31244157099153974,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2129,
       "step": 907
     },
     {
-      "epoch": 3.6761133603238867,
-      "grad_norm": 0.2880935986790524,
-      "learning_rate": 7.957929074501182e-06,
-      "loss": 0.2116,
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.3287067053160449,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1917,
       "step": 908
     },
     {
-      "epoch": 3.6801619433198383,
-      "grad_norm": 0.30352890757044665,
-      "learning_rate": 7.912823288500695e-06,
-      "loss": 0.1816,
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.29527485301754425,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1727,
       "step": 909
     },
     {
-      "epoch": 3.6842105263157894,
-      "grad_norm": 0.263099771668497,
-      "learning_rate": 7.867814151243102e-06,
-      "loss": 0.2114,
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.2949233211556494,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2023,
       "step": 910
     },
     {
-      "epoch": 3.688259109311741,
-      "grad_norm": 0.2661883561356836,
-      "learning_rate": 7.822902022620267e-06,
-      "loss": 0.2066,
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.2892121692711779,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1931,
       "step": 911
     },
     {
-      "epoch": 3.6923076923076925,
-      "grad_norm": 0.2886054638610827,
-      "learning_rate": 7.778087261748377e-06,
-      "loss": 0.2112,
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.31921025675722897,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2027,
       "step": 912
     },
     {
-      "epoch": 3.6963562753036436,
-      "grad_norm": 0.27588257495037005,
-      "learning_rate": 7.733370226965056e-06,
-      "loss": 0.2,
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.2968896195733594,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1986,
       "step": 913
     },
     {
-      "epoch": 3.700404858299595,
-      "grad_norm": 0.2873354808636462,
-      "learning_rate": 7.688751275826527e-06,
-      "loss": 0.2105,
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.2875040707705201,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1758,
       "step": 914
     },
     {
-      "epoch": 3.7044534412955468,
-      "grad_norm": 0.29249997908102116,
-      "learning_rate": 7.644230765104735e-06,
-      "loss": 0.1809,
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.3190329112177124,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1965,
       "step": 915
     },
     {
-      "epoch": 3.708502024291498,
-      "grad_norm": 0.2805131843963426,
-      "learning_rate": 7.599809050784502e-06,
-      "loss": 0.1944,
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.30296831487578185,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1806,
       "step": 916
     },
     {
-      "epoch": 3.7125506072874495,
-      "grad_norm": 0.2828500079921462,
-      "learning_rate": 7.555486488060661e-06,
-      "loss": 0.1814,
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.2989677867701075,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.184,
       "step": 917
     },
     {
-      "epoch": 3.716599190283401,
-      "grad_norm": 0.2693468734690399,
-      "learning_rate": 7.511263431335249e-06,
-      "loss": 0.1953,
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.30858261563547773,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2095,
       "step": 918
     },
     {
-      "epoch": 3.720647773279352,
-      "grad_norm": 0.29083681015277146,
-      "learning_rate": 7.467140234214656e-06,
-      "loss": 0.2038,
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.2984757369262141,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1708,
       "step": 919
     },
     {
-      "epoch": 3.7246963562753037,
-      "grad_norm": 0.3016517021778304,
-      "learning_rate": 7.423117249506777e-06,
-      "loss": 0.1851,
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.3001343895517145,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1806,
       "step": 920
     },
     {
-      "epoch": 3.7287449392712553,
-      "grad_norm": 0.29147564493282735,
-      "learning_rate": 7.379194829218231e-06,
-      "loss": 0.1866,
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.29363586955334803,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1779,
       "step": 921
     },
     {
-      "epoch": 3.7327935222672064,
-      "grad_norm": 0.2918559984660557,
-      "learning_rate": 7.335373324551518e-06,
-      "loss": 0.1946,
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.30968578964945676,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1784,
       "step": 922
     },
     {
-      "epoch": 3.736842105263158,
-      "grad_norm": 0.27271968461702345,
-      "learning_rate": 7.291653085902224e-06,
-      "loss": 0.192,
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.29082653174118994,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1965,
       "step": 923
     },
     {
-      "epoch": 3.7408906882591095,
-      "grad_norm": 0.28457431877505707,
-      "learning_rate": 7.248034462856213e-06,
-      "loss": 0.2131,
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.3161166327625153,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1938,
       "step": 924
     },
     {
-      "epoch": 3.7449392712550607,
-      "grad_norm": 0.2828271113281027,
-      "learning_rate": 7.20451780418683e-06,
-      "loss": 0.1892,
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.2778980112666072,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1871,
       "step": 925
     },
     {
-      "epoch": 3.748987854251012,
-      "grad_norm": 0.2644844488745109,
-      "learning_rate": 7.161103457852128e-06,
-      "loss": 0.2142,
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.2966650163803844,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2079,
       "step": 926
     },
     {
-      "epoch": 3.753036437246964,
-      "grad_norm": 0.2763234555008061,
-      "learning_rate": 7.117791770992046e-06,
-      "loss": 0.192,
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.2727125086958448,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1813,
       "step": 927
     },
     {
-      "epoch": 3.757085020242915,
-      "grad_norm": 0.2782181213067938,
-      "learning_rate": 7.074583089925688e-06,
-      "loss": 0.1996,
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.2955855841232463,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1997,
       "step": 928
     },
     {
-      "epoch": 3.7611336032388665,
-      "grad_norm": 0.2705146140835472,
-      "learning_rate": 7.0314777601485155e-06,
-      "loss": 0.211,
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.2545364928566164,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1591,
       "step": 929
     },
     {
-      "epoch": 3.765182186234818,
-      "grad_norm": 0.2617793420898279,
-      "learning_rate": 6.988476126329604e-06,
-      "loss": 0.1969,
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.3062250182200471,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2144,
       "step": 930
     },
     {
-      "epoch": 3.769230769230769,
-      "grad_norm": 0.26296040447795554,
-      "learning_rate": 6.945578532308857e-06,
-      "loss": 0.2021,
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.2871265807752657,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.188,
       "step": 931
     },
     {
-      "epoch": 3.7732793522267207,
-      "grad_norm": 0.27423911943955814,
-      "learning_rate": 6.902785321094301e-06,
-      "loss": 0.193,
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.31970451210835016,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2112,
       "step": 932
     },
     {
-      "epoch": 3.7773279352226723,
-      "grad_norm": 0.2971776137616127,
-      "learning_rate": 6.8600968348593115e-06,
-      "loss": 0.2042,
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.29026557012652804,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1981,
       "step": 933
     },
     {
-      "epoch": 3.7813765182186234,
-      "grad_norm": 0.2726871261700261,
-      "learning_rate": 6.817513414939892e-06,
-      "loss": 0.2083,
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.2596021997538026,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1599,
       "step": 934
     },
     {
-      "epoch": 3.785425101214575,
-      "grad_norm": 0.2795660038177764,
-      "learning_rate": 6.775035401831924e-06,
-      "loss": 0.2047,
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.3226744236070473,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2245,
       "step": 935
     },
     {
-      "epoch": 3.7894736842105265,
-      "grad_norm": 0.28480382459654296,
-      "learning_rate": 6.732663135188473e-06,
-      "loss": 0.1841,
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.2847793064786697,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1681,
       "step": 936
     },
     {
-      "epoch": 3.7935222672064777,
-      "grad_norm": 0.27904746361912647,
-      "learning_rate": 6.690396953817055e-06,
-      "loss": 0.2012,
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.3048641054079504,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.21,
       "step": 937
     },
     {
-      "epoch": 3.7975708502024292,
-      "grad_norm": 0.28023851342540096,
-      "learning_rate": 6.648237195676927e-06,
-      "loss": 0.1969,
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.2823660471380863,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1761,
       "step": 938
     },
     {
-      "epoch": 3.801619433198381,
-      "grad_norm": 0.2865287945707453,
-      "learning_rate": 6.606184197876395e-06,
-      "loss": 0.2012,
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.29849587350943424,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1915,
       "step": 939
     },
     {
-      "epoch": 3.805668016194332,
-      "grad_norm": 0.2742182374381058,
-      "learning_rate": 6.564238296670109e-06,
-      "loss": 0.2015,
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.2875693758074018,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.182,
       "step": 940
     },
     {
-      "epoch": 3.8097165991902835,
-      "grad_norm": 0.282217330866044,
-      "learning_rate": 6.52239982745638e-06,
-      "loss": 0.1941,
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.2979569039845659,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1968,
       "step": 941
     },
     {
-      "epoch": 3.813765182186235,
-      "grad_norm": 0.2818727954657328,
-      "learning_rate": 6.480669124774484e-06,
-      "loss": 0.1964,
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.284565109183977,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1674,
       "step": 942
     },
     {
-      "epoch": 3.817813765182186,
-      "grad_norm": 0.27071124306495115,
-      "learning_rate": 6.439046522302013e-06,
-      "loss": 0.2067,
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3156466114115207,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2032,
       "step": 943
     },
     {
-      "epoch": 3.8218623481781377,
-      "grad_norm": 0.2794004871728548,
-      "learning_rate": 6.397532352852191e-06,
-      "loss": 0.19,
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.3243927142657716,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2077,
       "step": 944
     },
     {
-      "epoch": 3.8259109311740893,
-      "grad_norm": 0.29649974236623616,
-      "learning_rate": 6.356126948371202e-06,
-      "loss": 0.1963,
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.33545119928304246,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1969,
       "step": 945
     },
     {
-      "epoch": 3.8299595141700404,
-      "grad_norm": 0.29177218439777036,
-      "learning_rate": 6.314830639935559e-06,
-      "loss": 0.1937,
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.2809887206645705,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1753,
       "step": 946
     },
     {
-      "epoch": 3.834008097165992,
-      "grad_norm": 0.28035106058697584,
-      "learning_rate": 6.273643757749448e-06,
-      "loss": 0.1941,
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.2656486392682715,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1778,
       "step": 947
     },
     {
-      "epoch": 3.8380566801619436,
-      "grad_norm": 0.2649923146184543,
-      "learning_rate": 6.2325666311420856e-06,
-      "loss": 0.2104,
+      "epoch": 3.84,
+      "grad_norm": 0.30729710993141873,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2058,
       "step": 948
     },
     {
-      "epoch": 3.8421052631578947,
-      "grad_norm": 0.27040818148864043,
-      "learning_rate": 6.1915995885650675e-06,
-      "loss": 0.1863,
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.28115403310953035,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1731,
       "step": 949
     },
     {
-      "epoch": 3.8461538461538463,
-      "grad_norm": 0.2725159109975051,
-      "learning_rate": 6.150742957589784e-06,
-      "loss": 0.1977,
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.2882922365627405,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1853,
       "step": 950
     },
     {
-      "epoch": 3.850202429149798,
-      "grad_norm": 0.2849824869364351,
-      "learning_rate": 6.109997064904774e-06,
-      "loss": 0.187,
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.31048135342854727,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2045,
       "step": 951
     },
     {
-      "epoch": 3.854251012145749,
-      "grad_norm": 0.27284473430626105,
-      "learning_rate": 6.069362236313092e-06,
-      "loss": 0.2067,
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.29347847074939143,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2006,
       "step": 952
     },
     {
-      "epoch": 3.8582995951417005,
-      "grad_norm": 0.2776064935116341,
-      "learning_rate": 6.0288387967297565e-06,
-      "loss": 0.1944,
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.2739576318077019,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1737,
       "step": 953
     },
     {
-      "epoch": 3.862348178137652,
-      "grad_norm": 0.2778683925756643,
-      "learning_rate": 5.988427070179108e-06,
-      "loss": 0.1991,
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.30444142480886927,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1944,
       "step": 954
     },
     {
-      "epoch": 3.866396761133603,
-      "grad_norm": 0.27240569054392927,
-      "learning_rate": 5.948127379792234e-06,
-      "loss": 0.2036,
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.29893231263164494,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1871,
       "step": 955
     },
     {
-      "epoch": 3.8704453441295548,
-      "grad_norm": 0.26164712424999537,
-      "learning_rate": 5.90794004780439e-06,
-      "loss": 0.1924,
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.2854602562825505,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.196,
       "step": 956
     },
     {
-      "epoch": 3.8744939271255063,
-      "grad_norm": 0.28285344771293364,
-      "learning_rate": 5.867865395552405e-06,
-      "loss": 0.2086,
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.3008248412136313,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1882,
       "step": 957
     },
     {
-      "epoch": 3.8785425101214575,
-      "grad_norm": 0.28584289471874674,
-      "learning_rate": 5.8279037434721364e-06,
-      "loss": 0.1823,
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.3042234115651594,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1864,
       "step": 958
     },
     {
-      "epoch": 3.882591093117409,
-      "grad_norm": 0.296336380037505,
-      "learning_rate": 5.788055411095892e-06,
-      "loss": 0.1828,
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.34602939807509464,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1951,
       "step": 959
     },
     {
-      "epoch": 3.8866396761133606,
-      "grad_norm": 0.2846619227894967,
-      "learning_rate": 5.7483207170498625e-06,
-      "loss": 0.1847,
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.29492181548124774,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1808,
       "step": 960
     },
     {
-      "epoch": 3.8906882591093117,
-      "grad_norm": 0.29697503168799333,
-      "learning_rate": 5.708699979051613e-06,
-      "loss": 0.1948,
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.28110583359541996,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1846,
       "step": 961
     },
     {
-      "epoch": 3.8947368421052633,
-      "grad_norm": 0.29389512880699564,
-      "learning_rate": 5.66919351390751e-06,
-      "loss": 0.1988,
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.28112620545717126,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1852,
       "step": 962
     },
     {
-      "epoch": 3.898785425101215,
-      "grad_norm": 0.28497334224573256,
-      "learning_rate": 5.629801637510188e-06,
-      "loss": 0.2013,
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.2780573155105353,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1827,
       "step": 963
     },
     {
-      "epoch": 3.902834008097166,
-      "grad_norm": 0.2879448737332081,
-      "learning_rate": 5.590524664836043e-06,
-      "loss": 0.2058,
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.31989118827293017,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1973,
       "step": 964
     },
     {
-      "epoch": 3.9068825910931175,
-      "grad_norm": 0.30096556124052476,
-      "learning_rate": 5.551362909942708e-06,
-      "loss": 0.2045,
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.30173947072324303,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2137,
       "step": 965
     },
     {
-      "epoch": 3.910931174089069,
-      "grad_norm": 0.27710551098613484,
-      "learning_rate": 5.51231668596653e-06,
-      "loss": 0.1913,
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.25447586746076944,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1646,
       "step": 966
     },
     {
-      "epoch": 3.91497975708502,
-      "grad_norm": 0.2709844047523216,
-      "learning_rate": 5.47338630512007e-06,
-      "loss": 0.1901,
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.28809770911547616,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.2023,
       "step": 967
     },
     {
-      "epoch": 3.919028340080972,
-      "grad_norm": 0.2563750374589138,
-      "learning_rate": 5.434572078689617e-06,
-      "loss": 0.2099,
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.25447155570297353,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1814,
       "step": 968
     },
     {
-      "epoch": 3.9230769230769234,
-      "grad_norm": 0.266900614134214,
-      "learning_rate": 5.395874317032694e-06,
-      "loss": 0.1957,
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.3039890117016232,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2015,
       "step": 969
     },
     {
-      "epoch": 3.9271255060728745,
-      "grad_norm": 0.2872887930573186,
-      "learning_rate": 5.357293329575573e-06,
-      "loss": 0.1947,
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.27167974508025344,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1706,
       "step": 970
     },
     {
-      "epoch": 3.931174089068826,
-      "grad_norm": 0.27188002460280075,
-      "learning_rate": 5.318829424810797e-06,
-      "loss": 0.1992,
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.2919896568042387,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1887,
       "step": 971
     },
     {
-      "epoch": 3.9352226720647776,
-      "grad_norm": 0.284548723991034,
-      "learning_rate": 5.280482910294728e-06,
-      "loss": 0.191,
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.28979190438938723,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1883,
       "step": 972
     },
     {
-      "epoch": 3.9392712550607287,
-      "grad_norm": 0.269870331577144,
-      "learning_rate": 5.242254092645078e-06,
-      "loss": 0.1798,
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.2929541662782835,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1886,
       "step": 973
     },
     {
-      "epoch": 3.9433198380566803,
-      "grad_norm": 0.26547450468256717,
-      "learning_rate": 5.2041432775384445e-06,
-      "loss": 0.1871,
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.26653511993806295,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1648,
       "step": 974
     },
     {
-      "epoch": 3.9473684210526314,
-      "grad_norm": 0.27202265898604655,
-      "learning_rate": 5.166150769707892e-06,
-      "loss": 0.183,
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.2889585686341497,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2008,
       "step": 975
     },
     {
-      "epoch": 3.951417004048583,
-      "grad_norm": 0.2514186765807129,
-      "learning_rate": 5.128276872940503e-06,
-      "loss": 0.2079,
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.2730512461799011,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1882,
       "step": 976
     },
     {
-      "epoch": 3.9554655870445345,
-      "grad_norm": 0.2699503520862887,
-      "learning_rate": 5.090521890074953e-06,
-      "loss": 0.2033,
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.2781127224884192,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1829,
       "step": 977
     },
     {
-      "epoch": 3.9595141700404857,
-      "grad_norm": 0.27784242390953295,
-      "learning_rate": 5.0528861229990675e-06,
-      "loss": 0.206,
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.29218309416945726,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1654,
       "step": 978
     },
     {
-      "epoch": 3.9635627530364372,
-      "grad_norm": 0.281161924360351,
-      "learning_rate": 5.015369872647442e-06,
-      "loss": 0.1738,
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.3194147318321097,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2067,
       "step": 979
     },
     {
-      "epoch": 3.967611336032389,
-      "grad_norm": 0.2901363736499241,
-      "learning_rate": 4.9779734389990195e-06,
-      "loss": 0.185,
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.28451262204031147,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1881,
       "step": 980
     },
     {
-      "epoch": 3.97165991902834,
-      "grad_norm": 0.2556628571205397,
-      "learning_rate": 4.940697121074675e-06,
-      "loss": 0.1944,
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.2534625168364004,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1752,
       "step": 981
     },
     {
-      "epoch": 3.9757085020242915,
-      "grad_norm": 0.26306289869766963,
-      "learning_rate": 4.903541216934864e-06,
-      "loss": 0.1991,
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.2951388665034386,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2096,
       "step": 982
     },
     {
-      "epoch": 3.979757085020243,
-      "grad_norm": 0.26682232130166145,
-      "learning_rate": 4.8665060236772e-06,
-      "loss": 0.2064,
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.27567794757903297,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.174,
       "step": 983
     },
     {
-      "epoch": 3.983805668016194,
-      "grad_norm": 0.30756846422881945,
-      "learning_rate": 4.829591837434102e-06,
-      "loss": 0.1856,
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.3218754739920183,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2058,
       "step": 984
     },
     {
-      "epoch": 3.9878542510121457,
-      "grad_norm": 0.27536717869631744,
-      "learning_rate": 4.792798953370424e-06,
-      "loss": 0.2101,
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.2625141028415917,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1806,
       "step": 985
     },
     {
-      "epoch": 3.9919028340080973,
-      "grad_norm": 0.2632737514579952,
-      "learning_rate": 4.756127665681083e-06,
-      "loss": 0.2045,
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.27518626502017246,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2026,
       "step": 986
     },
     {
-      "epoch": 3.9959514170040484,
-      "grad_norm": 0.26083256434720403,
-      "learning_rate": 4.719578267588725e-06,
-      "loss": 0.2101,
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.2841734726200722,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2049,
       "step": 987
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 0.2644789124992204,
-      "learning_rate": 4.6831510513413525e-06,
-      "loss": 0.1789,
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.3589457043110656,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1791,
       "step": 988
     },
     {
-      "epoch": 4.004048582995951,
-      "grad_norm": 0.42461815278568327,
-      "learning_rate": 4.6468463082100176e-06,
-      "loss": 0.1507,
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.39596384513274746,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1362,
       "step": 989
     },
     {
-      "epoch": 4.008097165991903,
-      "grad_norm": 0.3556645364819371,
-      "learning_rate": 4.610664328486485e-06,
-      "loss": 0.1393,
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.3287650535415528,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1221,
       "step": 990
     },
     {
-      "epoch": 4.012145748987854,
-      "grad_norm": 0.27356886327687635,
-      "learning_rate": 4.5746054014809e-06,
-      "loss": 0.1651,
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.29967411439591723,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1424,
       "step": 991
     },
     {
-      "epoch": 4.016194331983805,
-      "grad_norm": 0.2786473013809714,
-      "learning_rate": 4.5386698155194766e-06,
-      "loss": 0.1461,
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.27778102592244514,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1248,
       "step": 992
     },
     {
-      "epoch": 4.020242914979757,
-      "grad_norm": 0.34431862167815774,
-      "learning_rate": 4.502857857942204e-06,
-      "loss": 0.1502,
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.40607190928621323,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1465,
       "step": 993
     },
     {
-      "epoch": 4.0242914979757085,
-      "grad_norm": 0.41761566903314373,
-      "learning_rate": 4.467169815100549e-06,
-      "loss": 0.1425,
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.41268438093106974,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1235,
       "step": 994
     },
     {
-      "epoch": 4.02834008097166,
-      "grad_norm": 0.39509343396249524,
-      "learning_rate": 4.431605972355137e-06,
-      "loss": 0.1511,
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.4100640193291354,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.133,
       "step": 995
     },
     {
-      "epoch": 4.032388663967612,
-      "grad_norm": 0.291224351439545,
-      "learning_rate": 4.396166614073516e-06,
-      "loss": 0.1335,
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.3405599764955851,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1201,
       "step": 996
     },
     {
-      "epoch": 4.036437246963563,
-      "grad_norm": 0.3182538977475653,
-      "learning_rate": 4.360852023627846e-06,
-      "loss": 0.1303,
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.32813856782170825,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1333,
       "step": 997
     },
     {
-      "epoch": 4.040485829959514,
-      "grad_norm": 0.3470230246230391,
-      "learning_rate": 4.325662483392656e-06,
-      "loss": 0.1507,
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.3102510351504104,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1327,
       "step": 998
     },
     {
-      "epoch": 4.044534412955466,
-      "grad_norm": 0.3800570279443967,
-      "learning_rate": 4.29059827474257e-06,
-      "loss": 0.1485,
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.31918981670876106,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1298,
       "step": 999
     },
     {
-      "epoch": 4.048582995951417,
-      "grad_norm": 0.3328708243539084,
-      "learning_rate": 4.255659678050063e-06,
-      "loss": 0.1511,
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.3644829031172187,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1499,
       "step": 1000
     },
     {
-      "epoch": 4.052631578947368,
-      "grad_norm": 0.26009575961555426,
-      "learning_rate": 4.22084697268323e-06,
-      "loss": 0.1558,
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.28258614903805207,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1155,
       "step": 1001
     },
     {
-      "epoch": 4.05668016194332,
-      "grad_norm": 0.2698108228077193,
-      "learning_rate": 4.186160437003535e-06,
-      "loss": 0.1379,
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.29425430217510745,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1314,
       "step": 1002
     },
     {
-      "epoch": 4.060728744939271,
-      "grad_norm": 0.3154030544264742,
-      "learning_rate": 4.151600348363582e-06,
-      "loss": 0.1375,
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.29428146851287923,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1334,
       "step": 1003
     },
     {
-      "epoch": 4.064777327935222,
-      "grad_norm": 0.33775890892823474,
-      "learning_rate": 4.11716698310493e-06,
-      "loss": 0.1469,
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.28186963741661236,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1226,
       "step": 1004
     },
     {
-      "epoch": 4.068825910931174,
-      "grad_norm": 0.29657300538376014,
-      "learning_rate": 4.082860616555848e-06,
-      "loss": 0.1493,
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.30782725810299055,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1428,
       "step": 1005
     },
     {
-      "epoch": 4.0728744939271255,
-      "grad_norm": 0.2502576715034806,
-      "learning_rate": 4.048681523029125e-06,
-      "loss": 0.1443,
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.3017183401102668,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1494,
       "step": 1006
     },
     {
-      "epoch": 4.076923076923077,
-      "grad_norm": 0.2748173117145429,
-      "learning_rate": 4.014629975819882e-06,
-      "loss": 0.1591,
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.2581850240152474,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1126,
       "step": 1007
     },
     {
-      "epoch": 4.080971659919029,
-      "grad_norm": 0.26468430358919315,
-      "learning_rate": 3.9807062472033855e-06,
-      "loss": 0.1445,
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.2897310763971961,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1339,
       "step": 1008
     },
     {
-      "epoch": 4.08502024291498,
-      "grad_norm": 0.2765593860715269,
-      "learning_rate": 3.94691060843287e-06,
-      "loss": 0.1403,
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.28455620452264013,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1421,
       "step": 1009
     },
     {
-      "epoch": 4.089068825910931,
-      "grad_norm": 0.25750645197585037,
-      "learning_rate": 3.913243329737353e-06,
-      "loss": 0.1365,
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.2509460056179919,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1232,
       "step": 1010
     },
     {
-      "epoch": 4.093117408906883,
-      "grad_norm": 0.2715018527338675,
-      "learning_rate": 3.879704680319503e-06,
-      "loss": 0.1642,
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.270819084575095,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1339,
       "step": 1011
     },
     {
-      "epoch": 4.097165991902834,
-      "grad_norm": 0.26112136650778917,
-      "learning_rate": 3.846294928353471e-06,
-      "loss": 0.1422,
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.2804316539217467,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1312,
       "step": 1012
     },
     {
-      "epoch": 4.101214574898785,
-      "grad_norm": 0.2544294294131027,
-      "learning_rate": 3.8130143409827345e-06,
-      "loss": 0.1324,
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.24737886444298046,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1252,
       "step": 1013
     },
     {
-      "epoch": 4.105263157894737,
-      "grad_norm": 0.2579550078930912,
-      "learning_rate": 3.7798631843179887e-06,
-      "loss": 0.1479,
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.24447892542107744,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.114,
       "step": 1014
     },
     {
-      "epoch": 4.109311740890688,
-      "grad_norm": 0.26376041471566286,
-      "learning_rate": 3.7468417234349973e-06,
-      "loss": 0.1477,
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.29569668334185906,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1492,
       "step": 1015
     },
     {
-      "epoch": 4.113360323886639,
-      "grad_norm": 0.2787235105600079,
-      "learning_rate": 3.7139502223724823e-06,
-      "loss": 0.1478,
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.26039382291253754,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1162,
       "step": 1016
     },
     {
-      "epoch": 4.117408906882591,
-      "grad_norm": 0.24887112876361783,
-      "learning_rate": 3.6811889441300095e-06,
-      "loss": 0.1332,
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.26796659886696267,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.136,
       "step": 1017
     },
     {
-      "epoch": 4.1214574898785425,
-      "grad_norm": 0.257442198166874,
-      "learning_rate": 3.648558150665886e-06,
-      "loss": 0.1329,
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.27443189515482236,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1226,
       "step": 1018
     },
     {
-      "epoch": 4.125506072874494,
-      "grad_norm": 0.2627783014136314,
-      "learning_rate": 3.616058102895075e-06,
-      "loss": 0.1405,
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.2699612074955308,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1371,
       "step": 1019
     },
     {
-      "epoch": 4.129554655870446,
-      "grad_norm": 0.2512205317402381,
-      "learning_rate": 3.583689060687079e-06,
-      "loss": 0.1362,
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.24662335848294042,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.111,
       "step": 1020
     },
     {
-      "epoch": 4.133603238866397,
-      "grad_norm": 0.2532152567091869,
-      "learning_rate": 3.5514512828639023e-06,
-      "loss": 0.1448,
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.24586145781885288,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.116,
       "step": 1021
     },
     {
-      "epoch": 4.137651821862348,
-      "grad_norm": 0.24421121330617396,
-      "learning_rate": 3.51934502719796e-06,
-      "loss": 0.1496,
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.26895715314902124,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1399,
       "step": 1022
     },
     {
-      "epoch": 4.1417004048583,
-      "grad_norm": 0.2527216350701032,
-      "learning_rate": 3.4873705504100185e-06,
-      "loss": 0.1399,
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.29233486328267794,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1381,
       "step": 1023
     },
     {
-      "epoch": 4.145748987854251,
-      "grad_norm": 0.2621078284107286,
-      "learning_rate": 3.4555281081671345e-06,
-      "loss": 0.1297,
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.26218520577316584,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1275,
       "step": 1024
     },
     {
-      "epoch": 4.149797570850202,
-      "grad_norm": 0.25313382120795497,
-      "learning_rate": 3.4238179550806326e-06,
-      "loss": 0.1577,
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.24266021768997115,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1296,
       "step": 1025
     },
     {
-      "epoch": 4.153846153846154,
-      "grad_norm": 0.2577447515642222,
-      "learning_rate": 3.3922403447040497e-06,
-      "loss": 0.1467,
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.2745605116212587,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.139,
       "step": 1026
     },
     {
-      "epoch": 4.157894736842105,
-      "grad_norm": 0.24925865672351788,
-      "learning_rate": 3.3607955295311247e-06,
-      "loss": 0.1455,
+      "epoch": 4.16,
+      "grad_norm": 0.24658396712838926,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1108,
       "step": 1027
     },
     {
-      "epoch": 4.161943319838056,
-      "grad_norm": 0.2494957409120481,
-      "learning_rate": 3.32948376099375e-06,
-      "loss": 0.14,
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.2813535383495792,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1373,
       "step": 1028
     },
     {
-      "epoch": 4.165991902834008,
-      "grad_norm": 0.2605539888481595,
-      "learning_rate": 3.2983052894599976e-06,
-      "loss": 0.141,
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.25561439310277395,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1193,
       "step": 1029
     },
     {
-      "epoch": 4.17004048582996,
-      "grad_norm": 0.262805171018694,
-      "learning_rate": 3.2672603642320945e-06,
-      "loss": 0.1425,
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.28204992292062014,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1489,
       "step": 1030
     },
     {
-      "epoch": 4.174089068825911,
-      "grad_norm": 0.2559121286825748,
-      "learning_rate": 3.236349233544438e-06,
-      "loss": 0.1438,
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.25011757042485444,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1319,
       "step": 1031
     },
     {
-      "epoch": 4.178137651821863,
-      "grad_norm": 0.2737554215844273,
-      "learning_rate": 3.205572144561604e-06,
-      "loss": 0.1466,
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.26253966588369254,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.133,
       "step": 1032
     },
     {
-      "epoch": 4.182186234817814,
-      "grad_norm": 0.2684115976120087,
-      "learning_rate": 3.174929343376374e-06,
-      "loss": 0.1581,
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.24221371863159286,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1171,
       "step": 1033
     },
     {
-      "epoch": 4.186234817813765,
-      "grad_norm": 0.25688847446122176,
-      "learning_rate": 3.1444210750077777e-06,
-      "loss": 0.1462,
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.2603851173282285,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1401,
       "step": 1034
     },
     {
-      "epoch": 4.190283400809717,
-      "grad_norm": 0.25948913205563745,
-      "learning_rate": 3.1140475833991046e-06,
-      "loss": 0.1574,
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.2603837687393077,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.13,
       "step": 1035
     },
     {
-      "epoch": 4.194331983805668,
-      "grad_norm": 0.257231496405024,
-      "learning_rate": 3.0838091114159895e-06,
-      "loss": 0.1463,
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.26709468144933496,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1456,
       "step": 1036
     },
     {
-      "epoch": 4.198380566801619,
-      "grad_norm": 0.24224778880900893,
-      "learning_rate": 3.0537059008444524e-06,
-      "loss": 0.1549,
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.2538274862775357,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1193,
       "step": 1037
     },
     {
-      "epoch": 4.202429149797571,
-      "grad_norm": 0.2549944170580778,
-      "learning_rate": 3.0237381923889563e-06,
-      "loss": 0.1283,
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.27773396437019,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1363,
       "step": 1038
     },
     {
-      "epoch": 4.206477732793522,
-      "grad_norm": 0.25182152403469854,
-      "learning_rate": 2.9939062256704997e-06,
-      "loss": 0.1286,
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.2541197383392023,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1161,
       "step": 1039
     },
     {
-      "epoch": 4.2105263157894735,
-      "grad_norm": 0.25273789964474175,
-      "learning_rate": 2.964210239224696e-06,
-      "loss": 0.1453,
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.250961272561924,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1254,
       "step": 1040
     },
     {
-      "epoch": 4.2145748987854255,
-      "grad_norm": 0.2480669024983094,
-      "learning_rate": 2.934650470499869e-06,
-      "loss": 0.1501,
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.2609031546948184,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.136,
       "step": 1041
     },
     {
-      "epoch": 4.218623481781377,
-      "grad_norm": 0.25482169031836427,
-      "learning_rate": 2.905227155855128e-06,
-      "loss": 0.1502,
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.2515513535966848,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.126,
       "step": 1042
     },
     {
-      "epoch": 4.222672064777328,
-      "grad_norm": 0.2552025114995772,
-      "learning_rate": 2.875940530558519e-06,
-      "loss": 0.1443,
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.2807888742068889,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1398,
       "step": 1043
     },
     {
-      "epoch": 4.22672064777328,
-      "grad_norm": 0.2689828022061519,
-      "learning_rate": 2.8467908287851154e-06,
-      "loss": 0.1368,
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.2385032064523665,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1092,
       "step": 1044
     },
     {
-      "epoch": 4.230769230769231,
-      "grad_norm": 0.2545209566164527,
-      "learning_rate": 2.8177782836151603e-06,
-      "loss": 0.1446,
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.2579769642907314,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1289,
       "step": 1045
     },
     {
-      "epoch": 4.234817813765182,
-      "grad_norm": 0.25132732654302775,
-      "learning_rate": 2.7889031270321874e-06,
-      "loss": 0.1416,
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.26882355436731425,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1297,
       "step": 1046
     },
     {
-      "epoch": 4.238866396761134,
-      "grad_norm": 0.2656175157123964,
-      "learning_rate": 2.760165589921184e-06,
-      "loss": 0.1347,
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.25295257589528325,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1217,
       "step": 1047
     },
     {
-      "epoch": 4.242914979757085,
-      "grad_norm": 0.2543252675807152,
-      "learning_rate": 2.7315659020667305e-06,
-      "loss": 0.1334,
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.2739150895106317,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1322,
       "step": 1048
     },
     {
-      "epoch": 4.246963562753036,
-      "grad_norm": 0.24680727722496912,
-      "learning_rate": 2.703104292151164e-06,
-      "loss": 0.1406,
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.25912489380194964,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1195,
       "step": 1049
     },
     {
-      "epoch": 4.251012145748988,
-      "grad_norm": 0.25632804088492017,
-      "learning_rate": 2.6747809877527607e-06,
-      "loss": 0.1451,
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.2741673395973818,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1362,
       "step": 1050
     },
     {
-      "epoch": 4.255060728744939,
-      "grad_norm": 0.2580273042559127,
-      "learning_rate": 2.6465962153439082e-06,
-      "loss": 0.127,
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.2413401370696669,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1147,
       "step": 1051
     },
     {
-      "epoch": 4.2591093117408905,
-      "grad_norm": 0.25626653831828994,
-      "learning_rate": 2.618550200289303e-06,
-      "loss": 0.14,
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.26630958783866227,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1297,
       "step": 1052
     },
     {
-      "epoch": 4.2631578947368425,
-      "grad_norm": 0.2609580355798477,
-      "learning_rate": 2.59064316684412e-06,
-      "loss": 0.1418,
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.26291969125366876,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1324,
       "step": 1053
     },
     {
-      "epoch": 4.267206477732794,
-      "grad_norm": 0.24353252854517088,
-      "learning_rate": 2.562875338152264e-06,
-      "loss": 0.1431,
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.26378494295362404,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1271,
       "step": 1054
     },
     {
-      "epoch": 4.271255060728745,
-      "grad_norm": 0.25760597755423703,
-      "learning_rate": 2.5352469362445552e-06,
-      "loss": 0.1282,
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.23919728783132116,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1144,
       "step": 1055
     },
     {
-      "epoch": 4.275303643724697,
-      "grad_norm": 0.24890702364719633,
-      "learning_rate": 2.5077581820369523e-06,
-      "loss": 0.1429,
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.26684723426885637,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.139,
       "step": 1056
     },
     {
-      "epoch": 4.279352226720648,
-      "grad_norm": 0.24768339400164946,
-      "learning_rate": 2.480409295328803e-06,
-      "loss": 0.1404,
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.2743194037357143,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.128,
       "step": 1057
     },
     {
-      "epoch": 4.283400809716599,
-      "grad_norm": 0.2610460281857517,
-      "learning_rate": 2.45320049480108e-06,
-      "loss": 0.1355,
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.23999051767575408,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.11,
       "step": 1058
     },
     {
-      "epoch": 4.287449392712551,
-      "grad_norm": 0.2434556914102796,
-      "learning_rate": 2.4261319980146293e-06,
-      "loss": 0.1311,
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.24679988354720991,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1156,
       "step": 1059
     },
     {
-      "epoch": 4.291497975708502,
-      "grad_norm": 0.25150136181177607,
-      "learning_rate": 2.399204021408421e-06,
-      "loss": 0.1398,
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.27155453663236306,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1455,
       "step": 1060
     },
     {
-      "epoch": 4.295546558704453,
-      "grad_norm": 0.25691136385218977,
-      "learning_rate": 2.3724167802978395e-06,
-      "loss": 0.1391,
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.26491792073489767,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1382,
       "step": 1061
     },
     {
-      "epoch": 4.299595141700405,
-      "grad_norm": 0.2598516606447447,
-      "learning_rate": 2.3457704888729625e-06,
-      "loss": 0.151,
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.26570743942714775,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1166,
       "step": 1062
     },
     {
-      "epoch": 4.303643724696356,
-      "grad_norm": 0.261707316348782,
-      "learning_rate": 2.3192653601968162e-06,
-      "loss": 0.1327,
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.2554176181484278,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1265,
       "step": 1063
     },
     {
-      "epoch": 4.3076923076923075,
-      "grad_norm": 0.2412359140707734,
-      "learning_rate": 2.2929016062037057e-06,
-      "loss": 0.1361,
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.24741025710909176,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1139,
       "step": 1064
     },
     {
-      "epoch": 4.3117408906882595,
-      "grad_norm": 0.2592569248227437,
-      "learning_rate": 2.2666794376975055e-06,
-      "loss": 0.1303,
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.26284591325539014,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1236,
       "step": 1065
     },
     {
-      "epoch": 4.315789473684211,
-      "grad_norm": 0.26907443017474514,
-      "learning_rate": 2.24059906434998e-06,
-      "loss": 0.1361,
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.27335147818648414,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1401,
       "step": 1066
     },
     {
-      "epoch": 4.319838056680162,
-      "grad_norm": 0.26084146192718016,
-      "learning_rate": 2.214660694699091e-06,
-      "loss": 0.1484,
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.26883064529528083,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1368,
       "step": 1067
     },
     {
-      "epoch": 4.323886639676114,
-      "grad_norm": 0.2585409240006632,
-      "learning_rate": 2.188864536147355e-06,
-      "loss": 0.1439,
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.23853421212821707,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1109,
       "step": 1068
     },
     {
-      "epoch": 4.327935222672065,
-      "grad_norm": 0.25436105589264213,
-      "learning_rate": 2.1632107949601757e-06,
-      "loss": 0.1333,
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.2727370318547846,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1329,
       "step": 1069
     },
     {
-      "epoch": 4.331983805668016,
-      "grad_norm": 0.26863152594606393,
-      "learning_rate": 2.137699676264171e-06,
-      "loss": 0.1357,
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.2497770728835698,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1093,
       "step": 1070
     },
     {
-      "epoch": 4.336032388663968,
-      "grad_norm": 0.2533675056787129,
-      "learning_rate": 2.1123313840455738e-06,
-      "loss": 0.1342,
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.2677943872854694,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1285,
       "step": 1071
     },
     {
-      "epoch": 4.340080971659919,
-      "grad_norm": 0.24321014925433002,
-      "learning_rate": 2.08710612114857e-06,
-      "loss": 0.121,
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.2606505193123236,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1173,
       "step": 1072
     },
     {
-      "epoch": 4.34412955465587,
-      "grad_norm": 0.26635137546376525,
-      "learning_rate": 2.0620240892736953e-06,
-      "loss": 0.1281,
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.269928462235847,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1265,
       "step": 1073
     },
     {
-      "epoch": 4.348178137651822,
-      "grad_norm": 0.25270260618963714,
-      "learning_rate": 2.0370854889762e-06,
-      "loss": 0.1289,
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.259277247109757,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1252,
       "step": 1074
     },
     {
-      "epoch": 4.352226720647773,
-      "grad_norm": 0.26108548827427747,
-      "learning_rate": 2.0122905196644726e-06,
-      "loss": 0.1485,
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.25858115317568253,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1289,
       "step": 1075
     },
     {
-      "epoch": 4.3562753036437245,
-      "grad_norm": 0.2517426139269778,
-      "learning_rate": 1.987639379598427e-06,
-      "loss": 0.1387,
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.2840322774550965,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1254,
       "step": 1076
     },
     {
-      "epoch": 4.3603238866396765,
-      "grad_norm": 0.26550103882608556,
-      "learning_rate": 1.963132265887919e-06,
-      "loss": 0.1182,
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.24905006887573852,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1149,
       "step": 1077
     },
     {
-      "epoch": 4.364372469635628,
-      "grad_norm": 0.257086722592531,
-      "learning_rate": 1.9387693744911784e-06,
-      "loss": 0.1423,
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.2618379024740677,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1328,
       "step": 1078
     },
     {
-      "epoch": 4.368421052631579,
-      "grad_norm": 0.25003563752667973,
-      "learning_rate": 1.9145509002132368e-06,
-      "loss": 0.1371,
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.2518341518233997,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1174,
       "step": 1079
     },
     {
-      "epoch": 4.372469635627531,
-      "grad_norm": 0.2636850726766459,
-      "learning_rate": 1.8904770367043723e-06,
-      "loss": 0.1432,
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.27310513254048646,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.134,
       "step": 1080
     },
     {
-      "epoch": 4.376518218623482,
-      "grad_norm": 0.26014058982873994,
-      "learning_rate": 1.8665479764585504e-06,
-      "loss": 0.1455,
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.27032373546251937,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1346,
       "step": 1081
     },
     {
-      "epoch": 4.380566801619433,
-      "grad_norm": 0.2533094849766903,
-      "learning_rate": 1.8427639108118977e-06,
-      "loss": 0.1381,
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.24288458789157585,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1136,
       "step": 1082
     },
     {
-      "epoch": 4.384615384615385,
-      "grad_norm": 0.26644982101582854,
-      "learning_rate": 1.8191250299411755e-06,
-      "loss": 0.1444,
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.2570702251034397,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1257,
       "step": 1083
     },
     {
-      "epoch": 4.388663967611336,
-      "grad_norm": 0.24539736129428402,
-      "learning_rate": 1.7956315228622446e-06,
-      "loss": 0.1354,
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.2662443781196174,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1273,
       "step": 1084
     },
     {
-      "epoch": 4.392712550607287,
-      "grad_norm": 0.25658952443854965,
-      "learning_rate": 1.7722835774285553e-06,
-      "loss": 0.1267,
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.26797290578434907,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1233,
       "step": 1085
     },
     {
-      "epoch": 4.396761133603239,
-      "grad_norm": 0.25060463123617605,
-      "learning_rate": 1.7490813803296624e-06,
-      "loss": 0.1464,
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.23098102614957047,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1108,
       "step": 1086
     },
     {
-      "epoch": 4.40080971659919,
-      "grad_norm": 0.23614035923936402,
-      "learning_rate": 1.726025117089718e-06,
-      "loss": 0.1377,
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.2603599350853924,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1274,
       "step": 1087
     },
     {
-      "epoch": 4.4048582995951415,
-      "grad_norm": 0.24376398541041677,
-      "learning_rate": 1.70311497206598e-06,
-      "loss": 0.1382,
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.26011085568654807,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1304,
       "step": 1088
     },
     {
-      "epoch": 4.4089068825910935,
-      "grad_norm": 0.2628280357961318,
-      "learning_rate": 1.6803511284473638e-06,
-      "loss": 0.138,
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.25995164909524976,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1279,
       "step": 1089
     },
     {
-      "epoch": 4.412955465587045,
-      "grad_norm": 0.2641680350732329,
-      "learning_rate": 1.6577337682529538e-06,
-      "loss": 0.1417,
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.26909597068065433,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1307,
       "step": 1090
     },
     {
-      "epoch": 4.417004048582996,
-      "grad_norm": 0.24672637992827925,
-      "learning_rate": 1.6352630723305617e-06,
-      "loss": 0.1385,
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.2678303682341107,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.121,
       "step": 1091
     },
     {
-      "epoch": 4.421052631578947,
-      "grad_norm": 0.2602967443026303,
-      "learning_rate": 1.612939220355272e-06,
-      "loss": 0.1273,
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.2562235279652154,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1246,
       "step": 1092
     },
     {
-      "epoch": 4.425101214574899,
-      "grad_norm": 0.26967103947852955,
-      "learning_rate": 1.5907623908280112e-06,
-      "loss": 0.1469,
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.2694203877220934,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1204,
       "step": 1093
     },
     {
-      "epoch": 4.42914979757085,
-      "grad_norm": 0.2678226016988002,
-      "learning_rate": 1.568732761074121e-06,
-      "loss": 0.1498,
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.26386715051798615,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1458,
       "step": 1094
     },
     {
-      "epoch": 4.433198380566802,
-      "grad_norm": 0.2594513086137601,
-      "learning_rate": 1.5468505072419237e-06,
-      "loss": 0.1536,
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.26604644946907513,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1444,
       "step": 1095
     },
     {
-      "epoch": 4.437246963562753,
-      "grad_norm": 0.2514882803461518,
-      "learning_rate": 1.5251158043013448e-06,
-      "loss": 0.1498,
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.23369168784943897,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1073,
       "step": 1096
     },
     {
-      "epoch": 4.441295546558704,
-      "grad_norm": 0.26444778557036275,
-      "learning_rate": 1.5035288260424885e-06,
-      "loss": 0.1357,
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.2563614311000625,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1205,
       "step": 1097
     },
     {
-      "epoch": 4.445344129554655,
-      "grad_norm": 0.25659516970897883,
-      "learning_rate": 1.4820897450742623e-06,
-      "loss": 0.1395,
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.2882340031834417,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1406,
       "step": 1098
     },
     {
-      "epoch": 4.449392712550607,
-      "grad_norm": 0.2545370984111641,
-      "learning_rate": 1.4607987328229767e-06,
-      "loss": 0.1447,
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.2623106241099142,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.126,
       "step": 1099
     },
     {
-      "epoch": 4.4534412955465585,
-      "grad_norm": 0.25999787025974885,
-      "learning_rate": 1.4396559595310056e-06,
-      "loss": 0.1394,
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.2538155222692844,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1268,
       "step": 1100
     },
     {
-      "epoch": 4.4574898785425106,
-      "grad_norm": 0.25785170054677686,
-      "learning_rate": 1.4186615942553973e-06,
-      "loss": 0.1576,
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.2649170780803851,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1313,
       "step": 1101
     },
     {
-      "epoch": 4.461538461538462,
-      "grad_norm": 0.26407924818324086,
-      "learning_rate": 1.3978158048665424e-06,
-      "loss": 0.129,
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.24424216278693126,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1156,
       "step": 1102
     },
     {
-      "epoch": 4.465587044534413,
-      "grad_norm": 0.2415909492367587,
-      "learning_rate": 1.3771187580468115e-06,
-      "loss": 0.1527,
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.2560416290299474,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1424,
       "step": 1103
     },
     {
-      "epoch": 4.469635627530364,
-      "grad_norm": 0.25497199226299405,
-      "learning_rate": 1.3565706192892392e-06,
-      "loss": 0.1516,
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.251740630384699,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1233,
       "step": 1104
     },
     {
-      "epoch": 4.473684210526316,
-      "grad_norm": 0.25758175960979224,
-      "learning_rate": 1.3361715528961993e-06,
-      "loss": 0.1423,
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.2578490397897112,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1323,
       "step": 1105
     },
     {
-      "epoch": 4.477732793522267,
-      "grad_norm": 0.24531826941533041,
-      "learning_rate": 1.3159217219780817e-06,
-      "loss": 0.1468,
+      "epoch": 4.48,
+      "grad_norm": 0.25862543489960477,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1266,
       "step": 1106
     },
     {
-      "epoch": 4.481781376518219,
-      "grad_norm": 0.25563920083453406,
-      "learning_rate": 1.2958212884519949e-06,
-      "loss": 0.1411,
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.2647117959267297,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1293,
       "step": 1107
     },
     {
-      "epoch": 4.48582995951417,
-      "grad_norm": 0.2561932078067287,
-      "learning_rate": 1.2758704130404675e-06,
-      "loss": 0.1404,
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.2661612952781214,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.131,
       "step": 1108
     },
     {
-      "epoch": 4.489878542510121,
-      "grad_norm": 0.25025544616255696,
-      "learning_rate": 1.2560692552701738e-06,
-      "loss": 0.1479,
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.2544822550218989,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.118,
       "step": 1109
     },
     {
-      "epoch": 4.493927125506072,
-      "grad_norm": 0.2721515335345776,
-      "learning_rate": 1.2364179734706382e-06,
-      "loss": 0.1321,
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.2598701226488315,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1177,
       "step": 1110
     },
     {
-      "epoch": 4.497975708502024,
-      "grad_norm": 0.2544330306393475,
-      "learning_rate": 1.2169167247729852e-06,
-      "loss": 0.1431,
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.2613652093317228,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1295,
       "step": 1111
     },
     {
-      "epoch": 4.502024291497976,
-      "grad_norm": 0.24237093703294343,
-      "learning_rate": 1.1975656651086864e-06,
-      "loss": 0.1294,
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.2451563010347704,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1177,
       "step": 1112
     },
     {
-      "epoch": 4.506072874493928,
-      "grad_norm": 0.2393585613300664,
-      "learning_rate": 1.1783649492082926e-06,
-      "loss": 0.1371,
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.25742755452508254,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1276,
       "step": 1113
     },
     {
-      "epoch": 4.510121457489879,
-      "grad_norm": 0.2647767258876469,
-      "learning_rate": 1.1593147306002183e-06,
-      "loss": 0.1338,
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.26951924368393837,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1247,
       "step": 1114
     },
     {
-      "epoch": 4.51417004048583,
-      "grad_norm": 0.2534953737938607,
-      "learning_rate": 1.140415161609505e-06,
-      "loss": 0.1336,
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.28049822960859466,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1243,
       "step": 1115
     },
     {
-      "epoch": 4.518218623481781,
-      "grad_norm": 0.25987062070155675,
-      "learning_rate": 1.121666393356604e-06,
-      "loss": 0.1274,
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.2600987563614626,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1219,
       "step": 1116
     },
     {
-      "epoch": 4.522267206477733,
-      "grad_norm": 0.26093636799729053,
-      "learning_rate": 1.1030685757561632e-06,
-      "loss": 0.155,
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.25295215640469537,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1231,
       "step": 1117
     },
     {
-      "epoch": 4.526315789473684,
-      "grad_norm": 0.24828151903284465,
-      "learning_rate": 1.0846218575158373e-06,
-      "loss": 0.1279,
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.26519388319812437,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1303,
       "step": 1118
     },
     {
-      "epoch": 4.530364372469636,
-      "grad_norm": 0.2518888957881492,
-      "learning_rate": 1.066326386135097e-06,
-      "loss": 0.1282,
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.242708643877295,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1078,
       "step": 1119
     },
     {
-      "epoch": 4.534412955465587,
-      "grad_norm": 0.2596729494855252,
-      "learning_rate": 1.0481823079040398e-06,
-      "loss": 0.1315,
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.2711553579079075,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1323,
       "step": 1120
     },
     {
-      "epoch": 4.538461538461538,
-      "grad_norm": 0.25098671535579914,
-      "learning_rate": 1.0301897679022344e-06,
-      "loss": 0.1346,
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.2578196384637373,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1289,
       "step": 1121
     },
     {
-      "epoch": 4.5425101214574894,
-      "grad_norm": 0.24030358987225306,
-      "learning_rate": 1.0123489099975491e-06,
-      "loss": 0.1496,
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.26507357592670955,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1285,
       "step": 1122
     },
     {
-      "epoch": 4.5465587044534415,
-      "grad_norm": 0.26084136273768715,
-      "learning_rate": 9.9465987684501e-07,
-      "loss": 0.1286,
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.27129941042350764,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1346,
       "step": 1123
     },
     {
-      "epoch": 4.550607287449393,
-      "grad_norm": 0.2647483226594251,
-      "learning_rate": 9.771228098856534e-07,
-      "loss": 0.1339,
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.26709735293387654,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1268,
       "step": 1124
     },
     {
-      "epoch": 4.554655870445345,
-      "grad_norm": 0.2696840840065532,
-      "learning_rate": 9.597378493454013e-07,
-      "loss": 0.1254,
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.256236663541204,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1202,
       "step": 1125
     },
     {
-      "epoch": 4.558704453441296,
-      "grad_norm": 0.2460721501940823,
-      "learning_rate": 9.425051342339353e-07,
-      "loss": 0.1605,
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.2337779204936602,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1177,
       "step": 1126
     },
     {
-      "epoch": 4.562753036437247,
-      "grad_norm": 0.2631552147614341,
-      "learning_rate": 9.254248023435864e-07,
-      "loss": 0.1427,
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2914939532619924,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1469,
       "step": 1127
     },
     {
-      "epoch": 4.566801619433198,
-      "grad_norm": 0.26596723571002256,
-      "learning_rate": 9.084969902482288e-07,
-      "loss": 0.1409,
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.2708281410322519,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1274,
       "step": 1128
     },
     {
-      "epoch": 4.57085020242915,
-      "grad_norm": 0.2587918172176834,
-      "learning_rate": 8.917218333021993e-07,
-      "loss": 0.1255,
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.25997021858246977,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1322,
       "step": 1129
     },
     {
-      "epoch": 4.574898785425101,
-      "grad_norm": 0.24614218630345777,
-      "learning_rate": 8.750994656392109e-07,
-      "loss": 0.1435,
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.26019801818181765,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.143,
       "step": 1130
     },
     {
-      "epoch": 4.578947368421053,
-      "grad_norm": 0.2703664596624098,
-      "learning_rate": 8.58630020171265e-07,
-      "loss": 0.1536,
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.25855491162601285,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1334,
       "step": 1131
     },
     {
-      "epoch": 4.582995951417004,
-      "grad_norm": 0.26329660886922474,
-      "learning_rate": 8.423136285876099e-07,
-      "loss": 0.1483,
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.26190956258871806,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1133,
       "step": 1132
     },
     {
-      "epoch": 4.587044534412955,
-      "grad_norm": 0.26541481835182634,
-      "learning_rate": 8.26150421353682e-07,
-      "loss": 0.1144,
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.276632072381936,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1221,
       "step": 1133
     },
     {
-      "epoch": 4.5910931174089065,
-      "grad_norm": 0.26382637090784783,
-      "learning_rate": 8.101405277100549e-07,
-      "loss": 0.1359,
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.2570694119646122,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1312,
       "step": 1134
     },
     {
-      "epoch": 4.5951417004048585,
-      "grad_norm": 0.2630262393423851,
-      "learning_rate": 7.942840756714077e-07,
-      "loss": 0.1504,
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.24432997026382733,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1151,
       "step": 1135
     },
     {
-      "epoch": 4.59919028340081,
-      "grad_norm": 0.24435949484700942,
-      "learning_rate": 7.785811920255093e-07,
-      "loss": 0.1392,
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.24353983269347979,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1222,
       "step": 1136
     },
     {
-      "epoch": 4.603238866396762,
-      "grad_norm": 0.25253666045264395,
-      "learning_rate": 7.630320023322024e-07,
-      "loss": 0.1375,
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.26948207927554296,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1417,
       "step": 1137
     },
     {
-      "epoch": 4.607287449392713,
-      "grad_norm": 0.253501550963815,
-      "learning_rate": 7.476366309223903e-07,
-      "loss": 0.1475,
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.2551507976402425,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1244,
       "step": 1138
     },
     {
-      "epoch": 4.611336032388664,
-      "grad_norm": 0.2526852543014532,
-      "learning_rate": 7.323952008970537e-07,
-      "loss": 0.1316,
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.2541107685976849,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1158,
       "step": 1139
     },
     {
-      "epoch": 4.615384615384615,
-      "grad_norm": 0.25531528937395953,
-      "learning_rate": 7.173078341262618e-07,
-      "loss": 0.1346,
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.2754875790963283,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1448,
       "step": 1140
     },
     {
-      "epoch": 4.619433198380567,
-      "grad_norm": 0.26534058728433874,
-      "learning_rate": 7.023746512481989e-07,
-      "loss": 0.1576,
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.24853107236887742,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.123,
       "step": 1141
     },
     {
-      "epoch": 4.623481781376518,
-      "grad_norm": 0.247681528624369,
-      "learning_rate": 6.875957716681902e-07,
-      "loss": 0.1306,
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.24697525133813494,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1128,
       "step": 1142
     },
     {
-      "epoch": 4.62753036437247,
-      "grad_norm": 0.26010068996931357,
-      "learning_rate": 6.729713135577687e-07,
-      "loss": 0.1367,
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.2584977510487406,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1303,
       "step": 1143
     },
     {
-      "epoch": 4.631578947368421,
-      "grad_norm": 0.2657053880789342,
-      "learning_rate": 6.585013938537143e-07,
-      "loss": 0.1558,
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.25439249139182657,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1234,
       "step": 1144
     },
     {
-      "epoch": 4.635627530364372,
-      "grad_norm": 0.24070800422501726,
-      "learning_rate": 6.441861282571115e-07,
-      "loss": 0.1473,
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.2592165951680047,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1296,
       "step": 1145
     },
     {
-      "epoch": 4.6396761133603235,
-      "grad_norm": 0.24971157031228847,
-      "learning_rate": 6.300256312324471e-07,
-      "loss": 0.1385,
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.2518189800002725,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1304,
       "step": 1146
     },
     {
-      "epoch": 4.6437246963562755,
-      "grad_norm": 0.24420952078674596,
-      "learning_rate": 6.160200160066776e-07,
-      "loss": 0.136,
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.27001250698983703,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1468,
       "step": 1147
     },
     {
-      "epoch": 4.647773279352227,
-      "grad_norm": 0.2630631657643355,
-      "learning_rate": 6.021693945683327e-07,
-      "loss": 0.1472,
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.25190557916659384,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1184,
       "step": 1148
     },
     {
-      "epoch": 4.651821862348179,
-      "grad_norm": 0.24614319939225268,
-      "learning_rate": 5.884738776666044e-07,
-      "loss": 0.1383,
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.2762292291942114,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1499,
       "step": 1149
     },
     {
-      "epoch": 4.65587044534413,
-      "grad_norm": 0.2597922736956971,
-      "learning_rate": 5.749335748104878e-07,
-      "loss": 0.1477,
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.23833381724957228,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1056,
       "step": 1150
     },
     {
-      "epoch": 4.659919028340081,
-      "grad_norm": 0.2661290399322553,
-      "learning_rate": 5.615485942678778e-07,
-      "loss": 0.1176,
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.2704245662003936,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1245,
       "step": 1151
     },
     {
-      "epoch": 4.663967611336032,
-      "grad_norm": 0.25933578326106216,
-      "learning_rate": 5.483190430647289e-07,
-      "loss": 0.1308,
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.2547098951516238,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.12,
       "step": 1152
     },
     {
-      "epoch": 4.668016194331984,
-      "grad_norm": 0.24123557201559745,
-      "learning_rate": 5.352450269841747e-07,
-      "loss": 0.1497,
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.25162096422139796,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1339,
       "step": 1153
     },
     {
-      "epoch": 4.672064777327935,
-      "grad_norm": 0.2516384003992576,
-      "learning_rate": 5.22326650565701e-07,
-      "loss": 0.1378,
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.25531657048902096,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1252,
       "step": 1154
     },
     {
-      "epoch": 4.676113360323887,
-      "grad_norm": 0.25227358239389763,
-      "learning_rate": 5.095640171043048e-07,
-      "loss": 0.1451,
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.24718135243319847,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1194,
       "step": 1155
     },
     {
-      "epoch": 4.680161943319838,
-      "grad_norm": 0.23554833673685077,
-      "learning_rate": 4.969572286496594e-07,
-      "loss": 0.1429,
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.2863013918942534,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1517,
       "step": 1156
     },
     {
-      "epoch": 4.684210526315789,
-      "grad_norm": 0.2546437757515918,
-      "learning_rate": 4.845063860053079e-07,
-      "loss": 0.1437,
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.2426083749016937,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.119,
       "step": 1157
     },
     {
-      "epoch": 4.6882591093117405,
-      "grad_norm": 0.23873103905331866,
-      "learning_rate": 4.7221158872785997e-07,
-      "loss": 0.1406,
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.22895838557814863,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1118,
       "step": 1158
     },
     {
-      "epoch": 4.6923076923076925,
-      "grad_norm": 0.24567758177219937,
-      "learning_rate": 4.600729351261857e-07,
-      "loss": 0.136,
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.2533033359841471,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1225,
       "step": 1159
     },
     {
-      "epoch": 4.696356275303644,
-      "grad_norm": 0.25506156519118983,
-      "learning_rate": 4.4809052226063355e-07,
-      "loss": 0.1349,
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.26240079707084923,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1254,
       "step": 1160
     },
     {
-      "epoch": 4.700404858299595,
-      "grad_norm": 0.252138161814215,
-      "learning_rate": 4.3626444594226046e-07,
-      "loss": 0.1289,
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.26026805719850005,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1292,
       "step": 1161
     },
     {
-      "epoch": 4.704453441295547,
-      "grad_norm": 0.2560668991260997,
-      "learning_rate": 4.245948007320588e-07,
-      "loss": 0.1402,
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.2448139891501939,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1145,
       "step": 1162
     },
     {
-      "epoch": 4.708502024291498,
-      "grad_norm": 0.2540073933479402,
-      "learning_rate": 4.130816799401971e-07,
-      "loss": 0.1328,
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.24784074074466406,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1176,
       "step": 1163
     },
     {
-      "epoch": 4.712550607287449,
-      "grad_norm": 0.2456804761858607,
-      "learning_rate": 4.017251756252827e-07,
-      "loss": 0.1349,
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.2609601779646533,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1351,
       "step": 1164
     },
     {
-      "epoch": 4.716599190283401,
-      "grad_norm": 0.26487946019178704,
-      "learning_rate": 3.9052537859362473e-07,
-      "loss": 0.1563,
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.24870646850927244,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1265,
       "step": 1165
     },
     {
-      "epoch": 4.720647773279352,
-      "grad_norm": 0.2577167697518228,
-      "learning_rate": 3.7948237839849687e-07,
-      "loss": 0.1513,
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.2542850157850641,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1229,
       "step": 1166
     },
     {
-      "epoch": 4.724696356275303,
-      "grad_norm": 0.24746567816283327,
-      "learning_rate": 3.6859626333943797e-07,
-      "loss": 0.153,
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.24171322241447227,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.118,
       "step": 1167
     },
     {
-      "epoch": 4.728744939271255,
-      "grad_norm": 0.26035226082420726,
-      "learning_rate": 3.578671204615347e-07,
-      "loss": 0.1432,
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.26475052873619753,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1428,
       "step": 1168
     },
     {
-      "epoch": 4.732793522267206,
-      "grad_norm": 0.25192269187322786,
-      "learning_rate": 3.472950355547267e-07,
-      "loss": 0.1352,
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.25318325334786984,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.126,
       "step": 1169
     },
     {
-      "epoch": 4.7368421052631575,
-      "grad_norm": 0.24671569794064108,
-      "learning_rate": 3.3688009315312244e-07,
-      "loss": 0.1342,
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.24857028300028816,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1176,
       "step": 1170
     },
     {
-      "epoch": 4.7408906882591095,
-      "grad_norm": 0.25280344828315104,
-      "learning_rate": 3.266223765343224e-07,
-      "loss": 0.137,
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.25176464720805947,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1189,
       "step": 1171
     },
     {
-      "epoch": 4.744939271255061,
-      "grad_norm": 0.2509185663395771,
-      "learning_rate": 3.16521967718757e-07,
-      "loss": 0.1383,
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.26245734947622534,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1367,
       "step": 1172
     },
     {
-      "epoch": 4.748987854251012,
-      "grad_norm": 0.24482049636596975,
-      "learning_rate": 3.0657894746902726e-07,
-      "loss": 0.1436,
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.26545586487287304,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.14,
       "step": 1173
     },
     {
-      "epoch": 4.753036437246964,
-      "grad_norm": 0.2537984659976569,
-      "learning_rate": 2.9679339528925654e-07,
-      "loss": 0.1469,
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.23109472537362258,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1075,
       "step": 1174
     },
     {
-      "epoch": 4.757085020242915,
-      "grad_norm": 0.24492398111482258,
-      "learning_rate": 2.871653894244597e-07,
-      "loss": 0.1448,
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2658663303052665,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1438,
       "step": 1175
     },
     {
-      "epoch": 4.761133603238866,
-      "grad_norm": 0.2567951152272228,
-      "learning_rate": 2.7769500685991266e-07,
-      "loss": 0.1428,
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.2526916532900411,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.119,
       "step": 1176
     },
     {
-      "epoch": 4.765182186234818,
-      "grad_norm": 0.25812705583747086,
-      "learning_rate": 2.683823233205485e-07,
-      "loss": 0.1258,
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.25428749306683057,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1185,
       "step": 1177
     },
     {
-      "epoch": 4.769230769230769,
-      "grad_norm": 0.23967886253195203,
-      "learning_rate": 2.592274132703287e-07,
-      "loss": 0.1277,
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.23948945234154093,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1179,
       "step": 1178
     },
     {
-      "epoch": 4.77327935222672,
-      "grad_norm": 0.2506014599816903,
-      "learning_rate": 2.502303499116754e-07,
-      "loss": 0.1553,
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.2641467753564119,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1404,
       "step": 1179
     },
     {
-      "epoch": 4.777327935222672,
-      "grad_norm": 0.2507991138944762,
-      "learning_rate": 2.413912051848666e-07,
-      "loss": 0.1402,
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.24467646747499316,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1155,
       "step": 1180
     },
     {
-      "epoch": 4.781376518218623,
-      "grad_norm": 0.2600489517288205,
-      "learning_rate": 2.3271004976746613e-07,
-      "loss": 0.1372,
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.2517328299161815,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1218,
       "step": 1181
     },
     {
-      "epoch": 4.7854251012145745,
-      "grad_norm": 0.23919396182616307,
-      "learning_rate": 2.2418695307376393e-07,
-      "loss": 0.1445,
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.24992450681796907,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1426,
       "step": 1182
     },
     {
-      "epoch": 4.7894736842105265,
-      "grad_norm": 0.24376554983331916,
-      "learning_rate": 2.1582198325420968e-07,
-      "loss": 0.1384,
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.25802895683516724,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.1393,
       "step": 1183
     },
     {
-      "epoch": 4.793522267206478,
-      "grad_norm": 0.2411886539936394,
-      "learning_rate": 2.0761520719488447e-07,
-      "loss": 0.1451,
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.25373746104333555,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1241,
       "step": 1184
     },
     {
-      "epoch": 4.797570850202429,
-      "grad_norm": 0.26150377315566187,
-      "learning_rate": 1.9956669051694798e-07,
-      "loss": 0.1312,
+      "epoch": 4.8,
+      "grad_norm": 0.2648923215321204,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1248,
       "step": 1185
     },
     {
-      "epoch": 4.801619433198381,
-      "grad_norm": 0.25828402044745485,
-      "learning_rate": 1.916764975761254e-07,
-      "loss": 0.1403,
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.25140740088004593,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1226,
       "step": 1186
     },
     {
-      "epoch": 4.805668016194332,
-      "grad_norm": 0.23980064455174385,
-      "learning_rate": 1.839446914621923e-07,
-      "loss": 0.1476,
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.25849524396062906,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1377,
       "step": 1187
     },
     {
-      "epoch": 4.809716599190283,
-      "grad_norm": 0.2423972085131414,
-      "learning_rate": 1.7637133399845962e-07,
-      "loss": 0.1489,
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.23825372537211095,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1102,
       "step": 1188
     },
     {
-      "epoch": 4.813765182186235,
-      "grad_norm": 0.2550471806787937,
-      "learning_rate": 1.6895648574129397e-07,
-      "loss": 0.143,
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.27636552194474695,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1356,
       "step": 1189
     },
     {
-      "epoch": 4.817813765182186,
-      "grad_norm": 0.26310099841745044,
-      "learning_rate": 1.6170020597962465e-07,
-      "loss": 0.1343,
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.2538271768343137,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1155,
       "step": 1190
     },
     {
-      "epoch": 4.821862348178137,
-      "grad_norm": 0.2630075579535585,
-      "learning_rate": 1.5460255273447078e-07,
-      "loss": 0.1406,
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.2513505134624802,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1182,
       "step": 1191
     },
     {
-      "epoch": 4.825910931174089,
-      "grad_norm": 0.24858641148901744,
-      "learning_rate": 1.4766358275847715e-07,
-      "loss": 0.1378,
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.2539610676637153,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1381,
       "step": 1192
     },
     {
-      "epoch": 4.82995951417004,
-      "grad_norm": 0.25288568606468864,
-      "learning_rate": 1.4088335153546573e-07,
-      "loss": 0.1443,
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.2595875178765861,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.123,
       "step": 1193
     },
     {
-      "epoch": 4.834008097165992,
-      "grad_norm": 0.24576413140644296,
-      "learning_rate": 1.3426191327998496e-07,
-      "loss": 0.1461,
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.2585712344947893,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1277,
       "step": 1194
     },
     {
-      "epoch": 4.838056680161944,
-      "grad_norm": 0.25879480263382443,
-      "learning_rate": 1.277993209368744e-07,
-      "loss": 0.142,
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.23610808571399003,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1057,
       "step": 1195
     },
     {
-      "epoch": 4.842105263157895,
-      "grad_norm": 0.25818292186411423,
-      "learning_rate": 1.214956261808542e-07,
-      "loss": 0.1354,
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.2735641427233854,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1396,
       "step": 1196
     },
     {
-      "epoch": 4.846153846153846,
-      "grad_norm": 0.2545018814340851,
-      "learning_rate": 1.1535087941609846e-07,
-      "loss": 0.1334,
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.26211710742361943,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1361,
       "step": 1197
     },
     {
-      "epoch": 4.850202429149798,
-      "grad_norm": 0.24460067777196443,
-      "learning_rate": 1.0936512977583802e-07,
-      "loss": 0.1378,
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.24241183202555797,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1187,
       "step": 1198
     },
     {
-      "epoch": 4.854251012145749,
-      "grad_norm": 0.2517308865065694,
-      "learning_rate": 1.0353842512196732e-07,
-      "loss": 0.1448,
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.2425117743837619,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1113,
       "step": 1199
     },
     {
-      "epoch": 4.8582995951417,
-      "grad_norm": 0.2580603651558228,
-      "learning_rate": 9.787081204466253e-08,
-      "loss": 0.1289,
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.25953359183027047,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.131,
       "step": 1200
     },
     {
-      "epoch": 4.862348178137652,
-      "grad_norm": 0.2523583200500229,
-      "learning_rate": 9.236233586200627e-08,
-      "loss": 0.1367,
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.25705101342158604,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1342,
       "step": 1201
     },
     {
-      "epoch": 4.866396761133603,
-      "grad_norm": 0.25174901261561156,
-      "learning_rate": 8.70130406196279e-08,
-      "loss": 0.1532,
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.2783677188970859,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1434,
       "step": 1202
     },
     {
-      "epoch": 4.870445344129554,
-      "grad_norm": 0.2623829737840918,
-      "learning_rate": 8.182296909035047e-08,
-      "loss": 0.1335,
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.23536894615794246,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1067,
       "step": 1203
     },
     {
-      "epoch": 4.874493927125506,
-      "grad_norm": 0.24517762621457143,
-      "learning_rate": 7.679216277384439e-08,
-      "loss": 0.1352,
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.24321681870388476,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1189,
       "step": 1204
     },
     {
-      "epoch": 4.8785425101214575,
-      "grad_norm": 0.26410244832491625,
-      "learning_rate": 7.19206618963053e-08,
-      "loss": 0.1392,
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.28846256976901835,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1547,
       "step": 1205
     },
     {
-      "epoch": 4.882591093117409,
-      "grad_norm": 0.25464853201293697,
-      "learning_rate": 6.720850541012347e-08,
-      "loss": 0.1433,
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.23757926987780167,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1111,
       "step": 1206
     },
     {
-      "epoch": 4.886639676113361,
-      "grad_norm": 0.24980581268371166,
-      "learning_rate": 6.265573099357714e-08,
-      "loss": 0.1432,
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.26173623534339346,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1231,
       "step": 1207
     },
     {
-      "epoch": 4.890688259109312,
-      "grad_norm": 0.25281159839395234,
-      "learning_rate": 5.826237505053067e-08,
-      "loss": 0.1374,
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.25474628390137055,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1191,
       "step": 1208
     },
     {
-      "epoch": 4.894736842105263,
-      "grad_norm": 0.25055741952365534,
-      "learning_rate": 5.4028472710137e-08,
-      "loss": 0.1349,
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.24283676751545663,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1184,
       "step": 1209
     },
     {
-      "epoch": 4.898785425101215,
-      "grad_norm": 0.2448921591193139,
-      "learning_rate": 4.9954057826571146e-08,
-      "loss": 0.1422,
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.25368452672205216,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.126,
       "step": 1210
     },
     {
-      "epoch": 4.902834008097166,
-      "grad_norm": 0.2615966087581078,
-      "learning_rate": 4.6039162978743776e-08,
-      "loss": 0.1393,
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.27917065783505746,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1458,
       "step": 1211
     },
     {
-      "epoch": 4.906882591093117,
-      "grad_norm": 0.25775946224185226,
-      "learning_rate": 4.228381947004812e-08,
-      "loss": 0.1357,
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.23812029899313444,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1163,
       "step": 1212
     },
     {
-      "epoch": 4.910931174089069,
-      "grad_norm": 0.2511397813314095,
-      "learning_rate": 3.868805732811343e-08,
-      "loss": 0.1632,
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.2596356338526681,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.145,
       "step": 1213
     },
     {
-      "epoch": 4.91497975708502,
-      "grad_norm": 0.2525852468055962,
-      "learning_rate": 3.525190530455636e-08,
-      "loss": 0.1421,
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.24093858237607285,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1254,
       "step": 1214
     },
     {
-      "epoch": 4.919028340080971,
-      "grad_norm": 0.23439238219821834,
-      "learning_rate": 3.197539087475887e-08,
-      "loss": 0.1614,
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.25733136003635987,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1323,
       "step": 1215
     },
     {
-      "epoch": 4.923076923076923,
-      "grad_norm": 0.26122316939999735,
-      "learning_rate": 2.8858540237641786e-08,
-      "loss": 0.1365,
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.2638747442953262,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1417,
       "step": 1216
     },
     {
-      "epoch": 4.9271255060728745,
-      "grad_norm": 0.25045699252923054,
-      "learning_rate": 2.5901378315458247e-08,
-      "loss": 0.1421,
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.2404858304458219,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1083,
       "step": 1217
     },
     {
-      "epoch": 4.931174089068826,
-      "grad_norm": 0.2507636711240327,
-      "learning_rate": 2.3103928753598347e-08,
-      "loss": 0.1403,
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.24949097712968396,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.121,
       "step": 1218
     },
     {
-      "epoch": 4.935222672064778,
-      "grad_norm": 0.24644093246293994,
-      "learning_rate": 2.0466213920393718e-08,
-      "loss": 0.1354,
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.2624937512774275,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1338,
       "step": 1219
     },
     {
-      "epoch": 4.939271255060729,
-      "grad_norm": 0.25503354164122277,
-      "learning_rate": 1.7988254906942115e-08,
-      "loss": 0.1272,
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.2434212504771701,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1151,
       "step": 1220
     },
     {
-      "epoch": 4.94331983805668,
-      "grad_norm": 0.24041095875048663,
-      "learning_rate": 1.5670071526936447e-08,
-      "loss": 0.1373,
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.2423322969906,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1265,
       "step": 1221
     },
     {
-      "epoch": 4.947368421052632,
-      "grad_norm": 0.24819548649998086,
-      "learning_rate": 1.3511682316509344e-08,
-      "loss": 0.1459,
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.2727610283250997,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1321,
       "step": 1222
     },
     {
-      "epoch": 4.951417004048583,
-      "grad_norm": 0.26264635198816577,
-      "learning_rate": 1.1513104534086605e-08,
-      "loss": 0.137,
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.2537617996233366,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1267,
       "step": 1223
     },
     {
-      "epoch": 4.955465587044534,
-      "grad_norm": 0.24607846386911292,
-      "learning_rate": 9.67435416023843e-09,
-      "loss": 0.1596,
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.24306738166922487,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1231,
       "step": 1224
     },
     {
-      "epoch": 4.959514170040486,
-      "grad_norm": 0.2516725893130132,
-      "learning_rate": 7.995445897566178e-09,
-      "loss": 0.1308,
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.2603517242295158,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1303,
       "step": 1225
     },
     {
-      "epoch": 4.963562753036437,
-      "grad_norm": 0.23686316335633908,
-      "learning_rate": 6.476393170578021e-09,
-      "loss": 0.1381,
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.24096720721756443,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1116,
       "step": 1226
     },
     {
-      "epoch": 4.967611336032388,
-      "grad_norm": 0.2540954146812355,
-      "learning_rate": 5.11720812557348e-09,
-      "loss": 0.1338,
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.2660291728757945,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1263,
       "step": 1227
     },
     {
-      "epoch": 4.97165991902834,
-      "grad_norm": 0.2505187328730891,
-      "learning_rate": 3.917901630563492e-09,
-      "loss": 0.1349,
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.24622362431011433,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1177,
       "step": 1228
     },
     {
-      "epoch": 4.9757085020242915,
-      "grad_norm": 0.24922706645977655,
-      "learning_rate": 2.878483275170485e-09,
-      "loss": 0.1395,
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.249220573122714,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1251,
       "step": 1229
     },
     {
-      "epoch": 4.979757085020243,
-      "grad_norm": 0.2435016517951732,
-      "learning_rate": 1.998961370555108e-09,
-      "loss": 0.1398,
-      "step": 1230
-    },
-    {
-      "epoch": 4.983805668016195,
-      "grad_norm": 0.24888464355199189,
-      "learning_rate": 1.2793429493518361e-09,
-      "loss": 0.1445,
-      "step": 1231
-    },
-    {
-      "epoch": 4.987854251012146,
-      "grad_norm": 0.24638826053387872,
-      "learning_rate": 7.196337656112384e-10,
-      "loss": 0.1592,
-      "step": 1232
-    },
-    {
-      "epoch": 4.991902834008097,
-      "grad_norm": 0.24706128941903582,
-      "learning_rate": 3.198382947511292e-10,
-      "loss": 0.1352,
-      "step": 1233
-    },
-    {
-      "epoch": 4.995951417004049,
-      "grad_norm": 0.2416676189670063,
-      "learning_rate": 7.995973352770137e-11,
-      "loss": 0.1496,
-      "step": 1234
-    },
-    {
-      "epoch": 5.0,
-      "grad_norm": 0.2461701200164543,
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.2500548152601822,
       "learning_rate": 0.0,
-      "loss": 0.1402,
-      "step": 1235
+      "loss": 0.1359,
+      "step": 1230
     },
     {
-      "epoch": 5.0,
-      "step": 1235,
-      "total_flos": 1515609891143680.0,
-      "train_loss": 0.29405738967753614,
-      "train_runtime": 24654.457,
-      "train_samples_per_second": 6.409,
-      "train_steps_per_second": 0.05
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.312324413665575e+18,
+      "train_loss": 0.2930771143637537,
+      "train_runtime": 126133.8526,
+      "train_samples_per_second": 1.253,
+      "train_steps_per_second": 0.01
     }
   ],
-  "logging_steps": 1,
-  "max_steps": 1235,
+  "logging_steps": 1.0,
+  "max_steps": 1230,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
   "save_steps": 500,
@@ -8680,7 +8645,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1515609891143680.0,
+  "total_flos": 2.312324413665575e+18,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null