diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16388 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2335,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00042826552462526765,
+      "grad_norm": 3.495242708432237,
+      "learning_rate": 0.0,
+      "loss": 0.2728,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008565310492505353,
+      "grad_norm": 4.638281573997755,
+      "learning_rate": 1.4084507042253522e-07,
+      "loss": 0.3781,
+      "step": 2
+    },
+    {
+      "epoch": 0.001284796573875803,
+      "grad_norm": 3.118142705183361,
+      "learning_rate": 2.8169014084507043e-07,
+      "loss": 0.3131,
+      "step": 3
+    },
+    {
+      "epoch": 0.0017130620985010706,
+      "grad_norm": 3.6820776103295643,
+      "learning_rate": 4.225352112676057e-07,
+      "loss": 0.3603,
+      "step": 4
+    },
+    {
+      "epoch": 0.0021413276231263384,
+      "grad_norm": 2.894448924221148,
+      "learning_rate": 5.633802816901409e-07,
+      "loss": 0.3063,
+      "step": 5
+    },
+    {
+      "epoch": 0.002569593147751606,
+      "grad_norm": 3.3561317032799893,
+      "learning_rate": 7.042253521126762e-07,
+      "loss": 0.3353,
+      "step": 6
+    },
+    {
+      "epoch": 0.0029978586723768737,
+      "grad_norm": 2.4266932045452134,
+      "learning_rate": 8.450704225352114e-07,
+      "loss": 0.2583,
+      "step": 7
+    },
+    {
+      "epoch": 0.003426124197002141,
+      "grad_norm": 3.9653214695611068,
+      "learning_rate": 9.859154929577465e-07,
+      "loss": 0.3509,
+      "step": 8
+    },
+    {
+      "epoch": 0.003854389721627409,
+      "grad_norm": 4.123396270202061,
+      "learning_rate": 1.1267605633802817e-06,
+      "loss": 0.3989,
+      "step": 9
+    },
+    {
+      "epoch": 0.004282655246252677,
+      "grad_norm": 3.140025644289672,
+      "learning_rate": 1.267605633802817e-06,
+      "loss": 0.3265,
+      "step": 10
+    },
+    {
+      "epoch": 0.004710920770877944,
+      "grad_norm": 3.534612665039247,
+      "learning_rate": 1.4084507042253523e-06,
+      "loss": 0.314,
+      "step": 11
+    },
+    {
+      "epoch": 0.005139186295503212,
+      "grad_norm": 3.444887295621346,
+      "learning_rate": 1.5492957746478873e-06,
+      "loss": 0.3084,
+      "step": 12
+    },
+    {
+      "epoch": 0.00556745182012848,
+      "grad_norm": 3.1833799988364793,
+      "learning_rate": 1.6901408450704227e-06,
+      "loss": 0.3059,
+      "step": 13
+    },
+    {
+      "epoch": 0.0059957173447537475,
+      "grad_norm": 4.466831262926489,
+      "learning_rate": 1.8309859154929579e-06,
+      "loss": 0.453,
+      "step": 14
+    },
+    {
+      "epoch": 0.006423982869379015,
+      "grad_norm": 3.96430535422971,
+      "learning_rate": 1.971830985915493e-06,
+      "loss": 0.3775,
+      "step": 15
+    },
+    {
+      "epoch": 0.006852248394004282,
+      "grad_norm": 2.4895674459781776,
+      "learning_rate": 2.1126760563380285e-06,
+      "loss": 0.2555,
+      "step": 16
+    },
+    {
+      "epoch": 0.007280513918629551,
+      "grad_norm": 2.423590220886769,
+      "learning_rate": 2.2535211267605635e-06,
+      "loss": 0.3001,
+      "step": 17
+    },
+    {
+      "epoch": 0.007708779443254818,
+      "grad_norm": 3.110070936920792,
+      "learning_rate": 2.3943661971830984e-06,
+      "loss": 0.2822,
+      "step": 18
+    },
+    {
+      "epoch": 0.008137044967880086,
+      "grad_norm": 3.4960215982348926,
+      "learning_rate": 2.535211267605634e-06,
+      "loss": 0.3627,
+      "step": 19
+    },
+    {
+      "epoch": 0.008565310492505354,
+      "grad_norm": 2.950148810053227,
+      "learning_rate": 2.676056338028169e-06,
+      "loss": 0.3494,
+      "step": 20
+    },
+    {
+      "epoch": 0.008993576017130621,
+      "grad_norm": 2.3744748194512404,
+      "learning_rate": 2.8169014084507046e-06,
+      "loss": 0.2995,
+      "step": 21
+    },
+    {
+      "epoch": 0.009421841541755889,
+      "grad_norm": 3.0650567768965806,
+      "learning_rate": 2.9577464788732396e-06,
+      "loss": 0.322,
+      "step": 22
+    },
+    {
+      "epoch": 0.009850107066381156,
+      "grad_norm": 3.399592276583163,
+      "learning_rate": 3.0985915492957746e-06,
+      "loss": 0.3269,
+      "step": 23
+    },
+    {
+      "epoch": 0.010278372591006424,
+      "grad_norm": 2.9738051574549567,
+      "learning_rate": 3.2394366197183104e-06,
+      "loss": 0.3242,
+      "step": 24
+    },
+    {
+      "epoch": 0.010706638115631691,
+      "grad_norm": 2.5469719434746176,
+      "learning_rate": 3.3802816901408454e-06,
+      "loss": 0.3056,
+      "step": 25
+    },
+    {
+      "epoch": 0.01113490364025696,
+      "grad_norm": 2.4851474725256124,
+      "learning_rate": 3.5211267605633804e-06,
+      "loss": 0.3138,
+      "step": 26
+    },
+    {
+      "epoch": 0.011563169164882228,
+      "grad_norm": 2.5057828562687376,
+      "learning_rate": 3.6619718309859158e-06,
+      "loss": 0.2818,
+      "step": 27
+    },
+    {
+      "epoch": 0.011991434689507495,
+      "grad_norm": 2.7844074581050617,
+      "learning_rate": 3.8028169014084508e-06,
+      "loss": 0.3611,
+      "step": 28
+    },
+    {
+      "epoch": 0.012419700214132762,
+      "grad_norm": 2.5273775703520727,
+      "learning_rate": 3.943661971830986e-06,
+      "loss": 0.3489,
+      "step": 29
+    },
+    {
+      "epoch": 0.01284796573875803,
+      "grad_norm": 2.463642532563213,
+      "learning_rate": 4.0845070422535216e-06,
+      "loss": 0.3159,
+      "step": 30
+    },
+    {
+      "epoch": 0.013276231263383297,
+      "grad_norm": 2.3284108077952297,
+      "learning_rate": 4.225352112676057e-06,
+      "loss": 0.272,
+      "step": 31
+    },
+    {
+      "epoch": 0.013704496788008565,
+      "grad_norm": 2.5172747100024204,
+      "learning_rate": 4.3661971830985915e-06,
+      "loss": 0.3084,
+      "step": 32
+    },
+    {
+      "epoch": 0.014132762312633832,
+      "grad_norm": 2.9672207621794024,
+      "learning_rate": 4.507042253521127e-06,
+      "loss": 0.3594,
+      "step": 33
+    },
+    {
+      "epoch": 0.014561027837259101,
+      "grad_norm": 2.5731428168412904,
+      "learning_rate": 4.647887323943662e-06,
+      "loss": 0.2903,
+      "step": 34
+    },
+    {
+      "epoch": 0.014989293361884369,
+      "grad_norm": 2.554497052559098,
+      "learning_rate": 4.788732394366197e-06,
+      "loss": 0.2922,
+      "step": 35
+    },
+    {
+      "epoch": 0.015417558886509636,
+      "grad_norm": 2.5984524291910387,
+      "learning_rate": 4.929577464788733e-06,
+      "loss": 0.234,
+      "step": 36
+    },
+    {
+      "epoch": 0.015845824411134905,
+      "grad_norm": 2.824630111977682,
+      "learning_rate": 5.070422535211268e-06,
+      "loss": 0.3357,
+      "step": 37
+    },
+    {
+      "epoch": 0.016274089935760173,
+      "grad_norm": 2.531379899009298,
+      "learning_rate": 5.211267605633803e-06,
+      "loss": 0.3085,
+      "step": 38
+    },
+    {
+      "epoch": 0.01670235546038544,
+      "grad_norm": 2.563353209339304,
+      "learning_rate": 5.352112676056338e-06,
+      "loss": 0.2972,
+      "step": 39
+    },
+    {
+      "epoch": 0.017130620985010708,
+      "grad_norm": 2.6013610171571253,
+      "learning_rate": 5.492957746478874e-06,
+      "loss": 0.2698,
+      "step": 40
+    },
+    {
+      "epoch": 0.017558886509635975,
+      "grad_norm": 2.52337234419715,
+      "learning_rate": 5.633802816901409e-06,
+      "loss": 0.2675,
+      "step": 41
+    },
+    {
+      "epoch": 0.017987152034261242,
+      "grad_norm": 2.785875388624029,
+      "learning_rate": 5.774647887323944e-06,
+      "loss": 0.3093,
+      "step": 42
+    },
+    {
+      "epoch": 0.01841541755888651,
+      "grad_norm": 2.651604737986769,
+      "learning_rate": 5.915492957746479e-06,
+      "loss": 0.3419,
+      "step": 43
+    },
+    {
+      "epoch": 0.018843683083511777,
+      "grad_norm": 2.787476483700921,
+      "learning_rate": 6.056338028169015e-06,
+      "loss": 0.2723,
+      "step": 44
+    },
+    {
+      "epoch": 0.019271948608137045,
+      "grad_norm": 2.6274139618248444,
+      "learning_rate": 6.197183098591549e-06,
+      "loss": 0.3583,
+      "step": 45
+    },
+    {
+      "epoch": 0.019700214132762312,
+      "grad_norm": 2.536029752538332,
+      "learning_rate": 6.3380281690140855e-06,
+      "loss": 0.283,
+      "step": 46
+    },
+    {
+      "epoch": 0.02012847965738758,
+      "grad_norm": 2.898960378828785,
+      "learning_rate": 6.478873239436621e-06,
+      "loss": 0.3009,
+      "step": 47
+    },
+    {
+      "epoch": 0.020556745182012847,
+      "grad_norm": 2.660476222394167,
+      "learning_rate": 6.619718309859155e-06,
+      "loss": 0.3068,
+      "step": 48
+    },
+    {
+      "epoch": 0.020985010706638114,
+      "grad_norm": 2.5209759175683657,
+      "learning_rate": 6.760563380281691e-06,
+      "loss": 0.2901,
+      "step": 49
+    },
+    {
+      "epoch": 0.021413276231263382,
+      "grad_norm": 2.3025669802270046,
+      "learning_rate": 6.901408450704225e-06,
+      "loss": 0.3282,
+      "step": 50
+    },
+    {
+      "epoch": 0.02184154175588865,
+      "grad_norm": 2.5147025289722387,
+      "learning_rate": 7.042253521126761e-06,
+      "loss": 0.3336,
+      "step": 51
+    },
+    {
+      "epoch": 0.02226980728051392,
+      "grad_norm": 2.3954558765799017,
+      "learning_rate": 7.183098591549297e-06,
+      "loss": 0.2732,
+      "step": 52
+    },
+    {
+      "epoch": 0.022698072805139188,
+      "grad_norm": 2.3632126013139008,
+      "learning_rate": 7.3239436619718316e-06,
+      "loss": 0.2673,
+      "step": 53
+    },
+    {
+      "epoch": 0.023126338329764455,
+      "grad_norm": 2.4807175566792132,
+      "learning_rate": 7.464788732394367e-06,
+      "loss": 0.3073,
+      "step": 54
+    },
+    {
+      "epoch": 0.023554603854389723,
+      "grad_norm": 2.6519539002682597,
+      "learning_rate": 7.6056338028169015e-06,
+      "loss": 0.2642,
+      "step": 55
+    },
+    {
+      "epoch": 0.02398286937901499,
+      "grad_norm": 2.390249402350393,
+      "learning_rate": 7.746478873239436e-06,
+      "loss": 0.2853,
+      "step": 56
+    },
+    {
+      "epoch": 0.024411134903640257,
+      "grad_norm": 2.572703503867353,
+      "learning_rate": 7.887323943661972e-06,
+      "loss": 0.2686,
+      "step": 57
+    },
+    {
+      "epoch": 0.024839400428265525,
+      "grad_norm": 3.063578288025629,
+      "learning_rate": 8.028169014084509e-06,
+      "loss": 0.2721,
+      "step": 58
+    },
+    {
+      "epoch": 0.025267665952890792,
+      "grad_norm": 3.1544047968395765,
+      "learning_rate": 8.169014084507043e-06,
+      "loss": 0.3007,
+      "step": 59
+    },
+    {
+      "epoch": 0.02569593147751606,
+      "grad_norm": 2.9281420429531155,
+      "learning_rate": 8.309859154929578e-06,
+      "loss": 0.2849,
+      "step": 60
+    },
+    {
+      "epoch": 0.026124197002141327,
+      "grad_norm": 2.6132864390090083,
+      "learning_rate": 8.450704225352114e-06,
+      "loss": 0.2462,
+      "step": 61
+    },
+    {
+      "epoch": 0.026552462526766595,
+      "grad_norm": 2.6974996610853172,
+      "learning_rate": 8.591549295774648e-06,
+      "loss": 0.2858,
+      "step": 62
+    },
+    {
+      "epoch": 0.026980728051391862,
+      "grad_norm": 2.804746791259205,
+      "learning_rate": 8.732394366197183e-06,
+      "loss": 0.3225,
+      "step": 63
+    },
+    {
+      "epoch": 0.02740899357601713,
+      "grad_norm": 2.4038882502242966,
+      "learning_rate": 8.87323943661972e-06,
+      "loss": 0.2615,
+      "step": 64
+    },
+    {
+      "epoch": 0.027837259100642397,
+      "grad_norm": 2.52960086041787,
+      "learning_rate": 9.014084507042254e-06,
+      "loss": 0.2581,
+      "step": 65
+    },
+    {
+      "epoch": 0.028265524625267664,
+      "grad_norm": 2.5114902160148387,
+      "learning_rate": 9.15492957746479e-06,
+      "loss": 0.2805,
+      "step": 66
+    },
+    {
+      "epoch": 0.028693790149892935,
+      "grad_norm": 2.507320477920401,
+      "learning_rate": 9.295774647887325e-06,
+      "loss": 0.2521,
+      "step": 67
+    },
+    {
+      "epoch": 0.029122055674518203,
+      "grad_norm": 2.77673977600708,
+      "learning_rate": 9.43661971830986e-06,
+      "loss": 0.3132,
+      "step": 68
+    },
+    {
+      "epoch": 0.02955032119914347,
+      "grad_norm": 2.9256652374769425,
+      "learning_rate": 9.577464788732394e-06,
+      "loss": 0.3084,
+      "step": 69
+    },
+    {
+      "epoch": 0.029978586723768737,
+      "grad_norm": 2.7870377223756924,
+      "learning_rate": 9.71830985915493e-06,
+      "loss": 0.2778,
+      "step": 70
+    },
+    {
+      "epoch": 0.030406852248394005,
+      "grad_norm": 2.685304120783897,
+      "learning_rate": 9.859154929577466e-06,
+      "loss": 0.3245,
+      "step": 71
+    },
+    {
+      "epoch": 0.030835117773019272,
+      "grad_norm": 2.65142039610452,
+      "learning_rate": 1e-05,
+      "loss": 0.2907,
+      "step": 72
+    },
+    {
+      "epoch": 0.03126338329764454,
+      "grad_norm": 2.56815609074316,
+      "learning_rate": 9.999995186213338e-06,
+      "loss": 0.2325,
+      "step": 73
+    },
+    {
+      "epoch": 0.03169164882226981,
+      "grad_norm": 2.582387908897847,
+      "learning_rate": 9.999980744862618e-06,
+      "loss": 0.3202,
+      "step": 74
+    },
+    {
+      "epoch": 0.032119914346895075,
+      "grad_norm": 2.7392147586895494,
+      "learning_rate": 9.999956675975649e-06,
+      "loss": 0.302,
+      "step": 75
+    },
+    {
+      "epoch": 0.032548179871520345,
+      "grad_norm": 2.463388689648983,
+      "learning_rate": 9.999922979598773e-06,
+      "loss": 0.297,
+      "step": 76
+    },
+    {
+      "epoch": 0.03297644539614561,
+      "grad_norm": 2.547650039009328,
+      "learning_rate": 9.999879655796875e-06,
+      "loss": 0.2531,
+      "step": 77
+    },
+    {
+      "epoch": 0.03340471092077088,
+      "grad_norm": 2.5858498276101365,
+      "learning_rate": 9.999826704653376e-06,
+      "loss": 0.3427,
+      "step": 78
+    },
+    {
+      "epoch": 0.033832976445396144,
+      "grad_norm": 2.6176961257699887,
+      "learning_rate": 9.999764126270236e-06,
+      "loss": 0.2469,
+      "step": 79
+    },
+    {
+      "epoch": 0.034261241970021415,
+      "grad_norm": 2.8235409226471657,
+      "learning_rate": 9.999691920767945e-06,
+      "loss": 0.3085,
+      "step": 80
+    },
+    {
+      "epoch": 0.03468950749464668,
+      "grad_norm": 3.0153344013831376,
+      "learning_rate": 9.99961008828554e-06,
+      "loss": 0.303,
+      "step": 81
+    },
+    {
+      "epoch": 0.03511777301927195,
+      "grad_norm": 2.6356963497037684,
+      "learning_rate": 9.999518628980589e-06,
+      "loss": 0.3267,
+      "step": 82
+    },
+    {
+      "epoch": 0.035546038543897214,
+      "grad_norm": 2.397496738916966,
+      "learning_rate": 9.9994175430292e-06,
+      "loss": 0.2762,
+      "step": 83
+    },
+    {
+      "epoch": 0.035974304068522485,
+      "grad_norm": 2.4575959074374305,
+      "learning_rate": 9.999306830626015e-06,
+      "loss": 0.2984,
+      "step": 84
+    },
+    {
+      "epoch": 0.03640256959314775,
+      "grad_norm": 2.5400525559089155,
+      "learning_rate": 9.99918649198421e-06,
+      "loss": 0.3057,
+      "step": 85
+    },
+    {
+      "epoch": 0.03683083511777302,
+      "grad_norm": 3.0998548248226627,
+      "learning_rate": 9.999056527335503e-06,
+      "loss": 0.3118,
+      "step": 86
+    },
+    {
+      "epoch": 0.037259100642398284,
+      "grad_norm": 2.671795826900986,
+      "learning_rate": 9.998916936930137e-06,
+      "loss": 0.3079,
+      "step": 87
+    },
+    {
+      "epoch": 0.037687366167023555,
+      "grad_norm": 2.599469504248622,
+      "learning_rate": 9.998767721036901e-06,
+      "loss": 0.2553,
+      "step": 88
+    },
+    {
+      "epoch": 0.038115631691648826,
+      "grad_norm": 2.831204743711164,
+      "learning_rate": 9.998608879943111e-06,
+      "loss": 0.292,
+      "step": 89
+    },
+    {
+      "epoch": 0.03854389721627409,
+      "grad_norm": 2.936752646142617,
+      "learning_rate": 9.998440413954615e-06,
+      "loss": 0.3234,
+      "step": 90
+    },
+    {
+      "epoch": 0.03897216274089936,
+      "grad_norm": 2.903048427939846,
+      "learning_rate": 9.9982623233958e-06,
+      "loss": 0.3007,
+      "step": 91
+    },
+    {
+      "epoch": 0.039400428265524624,
+      "grad_norm": 2.88116483711928,
+      "learning_rate": 9.998074608609579e-06,
+      "loss": 0.2576,
+      "step": 92
+    },
+    {
+      "epoch": 0.039828693790149895,
+      "grad_norm": 2.8630144336084915,
+      "learning_rate": 9.997877269957403e-06,
+      "loss": 0.331,
+      "step": 93
+    },
+    {
+      "epoch": 0.04025695931477516,
+      "grad_norm": 3.131681690486848,
+      "learning_rate": 9.997670307819248e-06,
+      "loss": 0.3377,
+      "step": 94
+    },
+    {
+      "epoch": 0.04068522483940043,
+      "grad_norm": 2.835449539444292,
+      "learning_rate": 9.997453722593624e-06,
+      "loss": 0.288,
+      "step": 95
+    },
+    {
+      "epoch": 0.041113490364025694,
+      "grad_norm": 2.5837546123376534,
+      "learning_rate": 9.997227514697568e-06,
+      "loss": 0.2706,
+      "step": 96
+    },
+    {
+      "epoch": 0.041541755888650965,
+      "grad_norm": 2.5788156481483377,
+      "learning_rate": 9.996991684566647e-06,
+      "loss": 0.299,
+      "step": 97
+    },
+    {
+      "epoch": 0.04197002141327623,
+      "grad_norm": 2.6630165854589087,
+      "learning_rate": 9.996746232654956e-06,
+      "loss": 0.2947,
+      "step": 98
+    },
+    {
+      "epoch": 0.0423982869379015,
+      "grad_norm": 2.6486003547853456,
+      "learning_rate": 9.996491159435116e-06,
+      "loss": 0.2987,
+      "step": 99
+    },
+    {
+      "epoch": 0.042826552462526764,
+      "grad_norm": 2.6416320151587276,
+      "learning_rate": 9.996226465398272e-06,
+      "loss": 0.283,
+      "step": 100
+    },
+    {
+      "epoch": 0.043254817987152035,
+      "grad_norm": 2.778054306345007,
+      "learning_rate": 9.9959521510541e-06,
+      "loss": 0.3246,
+      "step": 101
+    },
+    {
+      "epoch": 0.0436830835117773,
+      "grad_norm": 2.7821403878643847,
+      "learning_rate": 9.995668216930793e-06,
+      "loss": 0.2645,
+      "step": 102
+    },
+    {
+      "epoch": 0.04411134903640257,
+      "grad_norm": 2.8213158217028127,
+      "learning_rate": 9.995374663575072e-06,
+      "loss": 0.3328,
+      "step": 103
+    },
+    {
+      "epoch": 0.04453961456102784,
+      "grad_norm": 3.0483388203823667,
+      "learning_rate": 9.99507149155218e-06,
+      "loss": 0.3522,
+      "step": 104
+    },
+    {
+      "epoch": 0.044967880085653104,
+      "grad_norm": 2.527122131156574,
+      "learning_rate": 9.994758701445876e-06,
+      "loss": 0.3332,
+      "step": 105
+    },
+    {
+      "epoch": 0.045396145610278375,
+      "grad_norm": 2.466828860948755,
+      "learning_rate": 9.994436293858442e-06,
+      "loss": 0.2746,
+      "step": 106
+    },
+    {
+      "epoch": 0.04582441113490364,
+      "grad_norm": 2.727781956348546,
+      "learning_rate": 9.994104269410681e-06,
+      "loss": 0.3209,
+      "step": 107
+    },
+    {
+      "epoch": 0.04625267665952891,
+      "grad_norm": 2.595558515850621,
+      "learning_rate": 9.99376262874191e-06,
+      "loss": 0.2843,
+      "step": 108
+    },
+    {
+      "epoch": 0.046680942184154174,
+      "grad_norm": 2.757588379550716,
+      "learning_rate": 9.993411372509962e-06,
+      "loss": 0.2723,
+      "step": 109
+    },
+    {
+      "epoch": 0.047109207708779445,
+      "grad_norm": 2.7998523382468092,
+      "learning_rate": 9.993050501391188e-06,
+      "loss": 0.2675,
+      "step": 110
+    },
+    {
+      "epoch": 0.04753747323340471,
+      "grad_norm": 2.8334539157618552,
+      "learning_rate": 9.992680016080447e-06,
+      "loss": 0.3125,
+      "step": 111
+    },
+    {
+      "epoch": 0.04796573875802998,
+      "grad_norm": 2.9416461161665035,
+      "learning_rate": 9.992299917291118e-06,
+      "loss": 0.3242,
+      "step": 112
+    },
+    {
+      "epoch": 0.048394004282655244,
+      "grad_norm": 2.9292714992763975,
+      "learning_rate": 9.991910205755085e-06,
+      "loss": 0.3324,
+      "step": 113
+    },
+    {
+      "epoch": 0.048822269807280515,
+      "grad_norm": 2.5131683522611747,
+      "learning_rate": 9.991510882222743e-06,
+      "loss": 0.2987,
+      "step": 114
+    },
+    {
+      "epoch": 0.04925053533190578,
+      "grad_norm": 2.4823253578665243,
+      "learning_rate": 9.991101947462995e-06,
+      "loss": 0.3082,
+      "step": 115
+    },
+    {
+      "epoch": 0.04967880085653105,
+      "grad_norm": 2.4943234816260595,
+      "learning_rate": 9.990683402263254e-06,
+      "loss": 0.2935,
+      "step": 116
+    },
+    {
+      "epoch": 0.050107066381156314,
+      "grad_norm": 2.8921221227900444,
+      "learning_rate": 9.99025524742943e-06,
+      "loss": 0.2793,
+      "step": 117
+    },
+    {
+      "epoch": 0.050535331905781584,
+      "grad_norm": 2.6514077783611154,
+      "learning_rate": 9.989817483785946e-06,
+      "loss": 0.2807,
+      "step": 118
+    },
+    {
+      "epoch": 0.050963597430406855,
+      "grad_norm": 2.6384669972627792,
+      "learning_rate": 9.989370112175718e-06,
+      "loss": 0.3557,
+      "step": 119
+    },
+    {
+      "epoch": 0.05139186295503212,
+      "grad_norm": 2.864209650805259,
+      "learning_rate": 9.98891313346017e-06,
+      "loss": 0.2862,
+      "step": 120
+    },
+    {
+      "epoch": 0.05182012847965739,
+      "grad_norm": 3.00295414993546,
+      "learning_rate": 9.988446548519221e-06,
+      "loss": 0.2275,
+      "step": 121
+    },
+    {
+      "epoch": 0.052248394004282654,
+      "grad_norm": 2.9117366905319515,
+      "learning_rate": 9.987970358251285e-06,
+      "loss": 0.3188,
+      "step": 122
+    },
+    {
+      "epoch": 0.052676659528907925,
+      "grad_norm": 2.7958944449380128,
+      "learning_rate": 9.987484563573275e-06,
+      "loss": 0.2733,
+      "step": 123
+    },
+    {
+      "epoch": 0.05310492505353319,
+      "grad_norm": 2.7542944222854335,
+      "learning_rate": 9.986989165420596e-06,
+      "loss": 0.3177,
+      "step": 124
+    },
+    {
+      "epoch": 0.05353319057815846,
+      "grad_norm": 2.7841279839250133,
+      "learning_rate": 9.986484164747142e-06,
+      "loss": 0.3159,
+      "step": 125
+    },
+    {
+      "epoch": 0.053961456102783724,
+      "grad_norm": 2.5874697095588792,
+      "learning_rate": 9.985969562525301e-06,
+      "loss": 0.2983,
+      "step": 126
+    },
+    {
+      "epoch": 0.054389721627408995,
+      "grad_norm": 2.7567891031171845,
+      "learning_rate": 9.985445359745949e-06,
+      "loss": 0.2797,
+      "step": 127
+    },
+    {
+      "epoch": 0.05481798715203426,
+      "grad_norm": 2.586310626210313,
+      "learning_rate": 9.984911557418444e-06,
+      "loss": 0.2981,
+      "step": 128
+    },
+    {
+      "epoch": 0.05524625267665953,
+      "grad_norm": 2.760318818159857,
+      "learning_rate": 9.984368156570628e-06,
+      "loss": 0.314,
+      "step": 129
+    },
+    {
+      "epoch": 0.055674518201284794,
+      "grad_norm": 3.0933762751026115,
+      "learning_rate": 9.983815158248831e-06,
+      "loss": 0.3392,
+      "step": 130
+    },
+    {
+      "epoch": 0.056102783725910065,
+      "grad_norm": 2.6974605045732347,
+      "learning_rate": 9.98325256351786e-06,
+      "loss": 0.3173,
+      "step": 131
+    },
+    {
+      "epoch": 0.05653104925053533,
+      "grad_norm": 2.805516469943714,
+      "learning_rate": 9.982680373460996e-06,
+      "loss": 0.291,
+      "step": 132
+    },
+    {
+      "epoch": 0.0569593147751606,
+      "grad_norm": 2.6362195118615754,
+      "learning_rate": 9.98209858918e-06,
+      "loss": 0.2931,
+      "step": 133
+    },
+    {
+      "epoch": 0.05738758029978587,
+      "grad_norm": 2.565658407724938,
+      "learning_rate": 9.981507211795107e-06,
+      "loss": 0.3161,
+      "step": 134
+    },
+    {
+      "epoch": 0.057815845824411134,
+      "grad_norm": 2.6715205250201137,
+      "learning_rate": 9.980906242445023e-06,
+      "loss": 0.278,
+      "step": 135
+    },
+    {
+      "epoch": 0.058244111349036405,
+      "grad_norm": 2.857539534894141,
+      "learning_rate": 9.980295682286924e-06,
+      "loss": 0.3209,
+      "step": 136
+    },
+    {
+      "epoch": 0.05867237687366167,
+      "grad_norm": 2.571594821233033,
+      "learning_rate": 9.979675532496451e-06,
+      "loss": 0.2968,
+      "step": 137
+    },
+    {
+      "epoch": 0.05910064239828694,
+      "grad_norm": 2.6230413271803945,
+      "learning_rate": 9.979045794267713e-06,
+      "loss": 0.2711,
+      "step": 138
+    },
+    {
+      "epoch": 0.059528907922912204,
+      "grad_norm": 2.8614540712851846,
+      "learning_rate": 9.97840646881328e-06,
+      "loss": 0.3357,
+      "step": 139
+    },
+    {
+      "epoch": 0.059957173447537475,
+      "grad_norm": 2.8766883948853392,
+      "learning_rate": 9.97775755736418e-06,
+      "loss": 0.3708,
+      "step": 140
+    },
+    {
+      "epoch": 0.06038543897216274,
+      "grad_norm": 2.672057589648402,
+      "learning_rate": 9.977099061169904e-06,
+      "loss": 0.2857,
+      "step": 141
+    },
+    {
+      "epoch": 0.06081370449678801,
+      "grad_norm": 2.8690728186866874,
+      "learning_rate": 9.976430981498395e-06,
+      "loss": 0.3034,
+      "step": 142
+    },
+    {
+      "epoch": 0.061241970021413274,
+      "grad_norm": 2.617117388595669,
+      "learning_rate": 9.975753319636053e-06,
+      "loss": 0.3154,
+      "step": 143
+    },
+    {
+      "epoch": 0.061670235546038545,
+      "grad_norm": 2.6492627325613154,
+      "learning_rate": 9.97506607688772e-06,
+      "loss": 0.2935,
+      "step": 144
+    },
+    {
+      "epoch": 0.06209850107066381,
+      "grad_norm": 2.771205938907388,
+      "learning_rate": 9.974369254576698e-06,
+      "loss": 0.3034,
+      "step": 145
+    },
+    {
+      "epoch": 0.06252676659528908,
+      "grad_norm": 2.759921713678794,
+      "learning_rate": 9.973662854044725e-06,
+      "loss": 0.2899,
+      "step": 146
+    },
+    {
+      "epoch": 0.06295503211991435,
+      "grad_norm": 2.7717107874726516,
+      "learning_rate": 9.972946876651988e-06,
+      "loss": 0.3077,
+      "step": 147
+    },
+    {
+      "epoch": 0.06338329764453962,
+      "grad_norm": 2.771872133004264,
+      "learning_rate": 9.97222132377711e-06,
+      "loss": 0.3336,
+      "step": 148
+    },
+    {
+      "epoch": 0.06381156316916488,
+      "grad_norm": 2.82474695509753,
+      "learning_rate": 9.971486196817154e-06,
+      "loss": 0.3253,
+      "step": 149
+    },
+    {
+      "epoch": 0.06423982869379015,
+      "grad_norm": 2.690629440560037,
+      "learning_rate": 9.97074149718762e-06,
+      "loss": 0.3034,
+      "step": 150
+    },
+    {
+      "epoch": 0.06466809421841542,
+      "grad_norm": 2.642910577267685,
+      "learning_rate": 9.969987226322435e-06,
+      "loss": 0.29,
+      "step": 151
+    },
+    {
+      "epoch": 0.06509635974304069,
+      "grad_norm": 2.61995478899865,
+      "learning_rate": 9.969223385673958e-06,
+      "loss": 0.3398,
+      "step": 152
+    },
+    {
+      "epoch": 0.06552462526766595,
+      "grad_norm": 2.9137313981088173,
+      "learning_rate": 9.96844997671298e-06,
+      "loss": 0.3405,
+      "step": 153
+    },
+    {
+      "epoch": 0.06595289079229122,
+      "grad_norm": 2.801308571978124,
+      "learning_rate": 9.967667000928706e-06,
+      "loss": 0.3112,
+      "step": 154
+    },
+    {
+      "epoch": 0.06638115631691649,
+      "grad_norm": 3.1288431419069744,
+      "learning_rate": 9.966874459828773e-06,
+      "loss": 0.3056,
+      "step": 155
+    },
+    {
+      "epoch": 0.06680942184154176,
+      "grad_norm": 2.669342453677056,
+      "learning_rate": 9.966072354939225e-06,
+      "loss": 0.3055,
+      "step": 156
+    },
+    {
+      "epoch": 0.06723768736616702,
+      "grad_norm": 2.6343102718061013,
+      "learning_rate": 9.96526068780453e-06,
+      "loss": 0.2942,
+      "step": 157
+    },
+    {
+      "epoch": 0.06766595289079229,
+      "grad_norm": 2.5729404184167266,
+      "learning_rate": 9.964439459987563e-06,
+      "loss": 0.3057,
+      "step": 158
+    },
+    {
+      "epoch": 0.06809421841541756,
+      "grad_norm": 2.669747730306156,
+      "learning_rate": 9.963608673069611e-06,
+      "loss": 0.3679,
+      "step": 159
+    },
+    {
+      "epoch": 0.06852248394004283,
+      "grad_norm": 2.7023141423947408,
+      "learning_rate": 9.962768328650367e-06,
+      "loss": 0.2991,
+      "step": 160
+    },
+    {
+      "epoch": 0.06895074946466809,
+      "grad_norm": 2.6820360180602534,
+      "learning_rate": 9.961918428347927e-06,
+      "loss": 0.2687,
+      "step": 161
+    },
+    {
+      "epoch": 0.06937901498929336,
+      "grad_norm": 2.624811770753171,
+      "learning_rate": 9.961058973798785e-06,
+      "loss": 0.288,
+      "step": 162
+    },
+    {
+      "epoch": 0.06980728051391863,
+      "grad_norm": 3.016328861012562,
+      "learning_rate": 9.960189966657833e-06,
+      "loss": 0.3177,
+      "step": 163
+    },
+    {
+      "epoch": 0.0702355460385439,
+      "grad_norm": 2.9053647957406623,
+      "learning_rate": 9.95931140859836e-06,
+      "loss": 0.3075,
+      "step": 164
+    },
+    {
+      "epoch": 0.07066381156316917,
+      "grad_norm": 3.1261317157932953,
+      "learning_rate": 9.95842330131204e-06,
+      "loss": 0.3422,
+      "step": 165
+    },
+    {
+      "epoch": 0.07109207708779443,
+      "grad_norm": 2.7560447849547676,
+      "learning_rate": 9.957525646508936e-06,
+      "loss": 0.3327,
+      "step": 166
+    },
+    {
+      "epoch": 0.0715203426124197,
+      "grad_norm": 2.8677939427715087,
+      "learning_rate": 9.956618445917496e-06,
+      "loss": 0.3411,
+      "step": 167
+    },
+    {
+      "epoch": 0.07194860813704497,
+      "grad_norm": 2.95920530791143,
+      "learning_rate": 9.95570170128455e-06,
+      "loss": 0.3119,
+      "step": 168
+    },
+    {
+      "epoch": 0.07237687366167024,
+      "grad_norm": 2.7486355000526705,
+      "learning_rate": 9.954775414375302e-06,
+      "loss": 0.3415,
+      "step": 169
+    },
+    {
+      "epoch": 0.0728051391862955,
+      "grad_norm": 2.846490857304893,
+      "learning_rate": 9.95383958697333e-06,
+      "loss": 0.3309,
+      "step": 170
+    },
+    {
+      "epoch": 0.07323340471092077,
+      "grad_norm": 2.655854977079735,
+      "learning_rate": 9.952894220880583e-06,
+      "loss": 0.2939,
+      "step": 171
+    },
+    {
+      "epoch": 0.07366167023554604,
+      "grad_norm": 2.561129996280741,
+      "learning_rate": 9.951939317917381e-06,
+      "loss": 0.2983,
+      "step": 172
+    },
+    {
+      "epoch": 0.07408993576017131,
+      "grad_norm": 2.823340571591925,
+      "learning_rate": 9.950974879922399e-06,
+      "loss": 0.3321,
+      "step": 173
+    },
+    {
+      "epoch": 0.07451820128479657,
+      "grad_norm": 2.7056642741627703,
+      "learning_rate": 9.95000090875268e-06,
+      "loss": 0.3556,
+      "step": 174
+    },
+    {
+      "epoch": 0.07494646680942184,
+      "grad_norm": 2.9465018093436073,
+      "learning_rate": 9.949017406283616e-06,
+      "loss": 0.3336,
+      "step": 175
+    },
+    {
+      "epoch": 0.07537473233404711,
+      "grad_norm": 2.762811161549083,
+      "learning_rate": 9.94802437440896e-06,
+      "loss": 0.2731,
+      "step": 176
+    },
+    {
+      "epoch": 0.07580299785867238,
+      "grad_norm": 2.811747599487674,
+      "learning_rate": 9.947021815040807e-06,
+      "loss": 0.2802,
+      "step": 177
+    },
+    {
+      "epoch": 0.07623126338329765,
+      "grad_norm": 2.7163312156585135,
+      "learning_rate": 9.9460097301096e-06,
+      "loss": 0.2731,
+      "step": 178
+    },
+    {
+      "epoch": 0.07665952890792291,
+      "grad_norm": 2.801185268878393,
+      "learning_rate": 9.944988121564123e-06,
+      "loss": 0.3458,
+      "step": 179
+    },
+    {
+      "epoch": 0.07708779443254818,
+      "grad_norm": 2.6899289945613947,
+      "learning_rate": 9.9439569913715e-06,
+      "loss": 0.2896,
+      "step": 180
+    },
+    {
+      "epoch": 0.07751605995717345,
+      "grad_norm": 2.9136994975520354,
+      "learning_rate": 9.942916341517184e-06,
+      "loss": 0.3479,
+      "step": 181
+    },
+    {
+      "epoch": 0.07794432548179872,
+      "grad_norm": 2.681082828452022,
+      "learning_rate": 9.941866174004964e-06,
+      "loss": 0.2813,
+      "step": 182
+    },
+    {
+      "epoch": 0.07837259100642398,
+      "grad_norm": 2.854487370222728,
+      "learning_rate": 9.940806490856953e-06,
+      "loss": 0.3282,
+      "step": 183
+    },
+    {
+      "epoch": 0.07880085653104925,
+      "grad_norm": 2.791957005888773,
+      "learning_rate": 9.939737294113585e-06,
+      "loss": 0.3413,
+      "step": 184
+    },
+    {
+      "epoch": 0.07922912205567452,
+      "grad_norm": 2.6518427491219323,
+      "learning_rate": 9.938658585833616e-06,
+      "loss": 0.2998,
+      "step": 185
+    },
+    {
+      "epoch": 0.07965738758029979,
+      "grad_norm": 2.6988002411471497,
+      "learning_rate": 9.937570368094113e-06,
+      "loss": 0.3285,
+      "step": 186
+    },
+    {
+      "epoch": 0.08008565310492505,
+      "grad_norm": 2.5894362095479244,
+      "learning_rate": 9.936472642990456e-06,
+      "loss": 0.2778,
+      "step": 187
+    },
+    {
+      "epoch": 0.08051391862955032,
+      "grad_norm": 2.7043545948283305,
+      "learning_rate": 9.93536541263633e-06,
+      "loss": 0.2788,
+      "step": 188
+    },
+    {
+      "epoch": 0.08094218415417559,
+      "grad_norm": 3.0185008707686767,
+      "learning_rate": 9.934248679163724e-06,
+      "loss": 0.3242,
+      "step": 189
+    },
+    {
+      "epoch": 0.08137044967880086,
+      "grad_norm": 2.678430491448051,
+      "learning_rate": 9.933122444722924e-06,
+      "loss": 0.279,
+      "step": 190
+    },
+    {
+      "epoch": 0.08179871520342613,
+      "grad_norm": 2.695054085304221,
+      "learning_rate": 9.931986711482511e-06,
+      "loss": 0.3094,
+      "step": 191
+    },
+    {
+      "epoch": 0.08222698072805139,
+      "grad_norm": 3.007027133704269,
+      "learning_rate": 9.930841481629358e-06,
+      "loss": 0.3023,
+      "step": 192
+    },
+    {
+      "epoch": 0.08265524625267666,
+      "grad_norm": 2.6842970122702465,
+      "learning_rate": 9.929686757368619e-06,
+      "loss": 0.2889,
+      "step": 193
+    },
+    {
+      "epoch": 0.08308351177730193,
+      "grad_norm": 2.7246884125494324,
+      "learning_rate": 9.928522540923736e-06,
+      "loss": 0.3003,
+      "step": 194
+    },
+    {
+      "epoch": 0.0835117773019272,
+      "grad_norm": 2.720594310920424,
+      "learning_rate": 9.927348834536421e-06,
+      "loss": 0.3437,
+      "step": 195
+    },
+    {
+      "epoch": 0.08394004282655246,
+      "grad_norm": 2.560950647523178,
+      "learning_rate": 9.926165640466664e-06,
+      "loss": 0.2821,
+      "step": 196
+    },
+    {
+      "epoch": 0.08436830835117773,
+      "grad_norm": 2.401579664814493,
+      "learning_rate": 9.924972960992722e-06,
+      "loss": 0.3122,
+      "step": 197
+    },
+    {
+      "epoch": 0.084796573875803,
+      "grad_norm": 2.8643542052991506,
+      "learning_rate": 9.923770798411121e-06,
+      "loss": 0.2675,
+      "step": 198
+    },
+    {
+      "epoch": 0.08522483940042827,
+      "grad_norm": 3.334578073887408,
+      "learning_rate": 9.922559155036637e-06,
+      "loss": 0.2819,
+      "step": 199
+    },
+    {
+      "epoch": 0.08565310492505353,
+      "grad_norm": 2.5693110884042927,
+      "learning_rate": 9.92133803320231e-06,
+      "loss": 0.2559,
+      "step": 200
+    },
+    {
+      "epoch": 0.0860813704496788,
+      "grad_norm": 2.7125597015724243,
+      "learning_rate": 9.92010743525943e-06,
+      "loss": 0.3052,
+      "step": 201
+    },
+    {
+      "epoch": 0.08650963597430407,
+      "grad_norm": 3.034952718107863,
+      "learning_rate": 9.918867363577526e-06,
+      "loss": 0.3314,
+      "step": 202
+    },
+    {
+      "epoch": 0.08693790149892934,
+      "grad_norm": 3.029160536671787,
+      "learning_rate": 9.917617820544378e-06,
+      "loss": 0.315,
+      "step": 203
+    },
+    {
+      "epoch": 0.0873661670235546,
+      "grad_norm": 2.787644862581071,
+      "learning_rate": 9.916358808565999e-06,
+      "loss": 0.3303,
+      "step": 204
+    },
+    {
+      "epoch": 0.08779443254817987,
+      "grad_norm": 3.1109289510736287,
+      "learning_rate": 9.915090330066634e-06,
+      "loss": 0.3111,
+      "step": 205
+    },
+    {
+      "epoch": 0.08822269807280514,
+      "grad_norm": 2.6255256528911906,
+      "learning_rate": 9.91381238748876e-06,
+      "loss": 0.3479,
+      "step": 206
+    },
+    {
+      "epoch": 0.08865096359743041,
+      "grad_norm": 2.5054851521365467,
+      "learning_rate": 9.912524983293068e-06,
+      "loss": 0.3142,
+      "step": 207
+    },
+    {
+      "epoch": 0.08907922912205568,
+      "grad_norm": 2.715557260510155,
+      "learning_rate": 9.91122811995848e-06,
+      "loss": 0.2814,
+      "step": 208
+    },
+    {
+      "epoch": 0.08950749464668094,
+      "grad_norm": 2.8738939747726358,
+      "learning_rate": 9.909921799982123e-06,
+      "loss": 0.3088,
+      "step": 209
+    },
+    {
+      "epoch": 0.08993576017130621,
+      "grad_norm": 2.3838901305900744,
+      "learning_rate": 9.908606025879333e-06,
+      "loss": 0.2704,
+      "step": 210
+    },
+    {
+      "epoch": 0.09036402569593148,
+      "grad_norm": 2.686001973335655,
+      "learning_rate": 9.907280800183656e-06,
+      "loss": 0.2972,
+      "step": 211
+    },
+    {
+      "epoch": 0.09079229122055675,
+      "grad_norm": 2.5481986130642356,
+      "learning_rate": 9.905946125446832e-06,
+      "loss": 0.2949,
+      "step": 212
+    },
+    {
+      "epoch": 0.09122055674518201,
+      "grad_norm": 2.9516226514429165,
+      "learning_rate": 9.904602004238797e-06,
+      "loss": 0.2635,
+      "step": 213
+    },
+    {
+      "epoch": 0.09164882226980728,
+      "grad_norm": 2.9448578946133304,
+      "learning_rate": 9.903248439147674e-06,
+      "loss": 0.2973,
+      "step": 214
+    },
+    {
+      "epoch": 0.09207708779443255,
+      "grad_norm": 3.3340644272873345,
+      "learning_rate": 9.901885432779777e-06,
+      "loss": 0.3178,
+      "step": 215
+    },
+    {
+      "epoch": 0.09250535331905782,
+      "grad_norm": 3.0808505904427945,
+      "learning_rate": 9.90051298775959e-06,
+      "loss": 0.2783,
+      "step": 216
+    },
+    {
+      "epoch": 0.09293361884368308,
+      "grad_norm": 2.7474121058436896,
+      "learning_rate": 9.899131106729778e-06,
+      "loss": 0.3253,
+      "step": 217
+    },
+    {
+      "epoch": 0.09336188436830835,
+      "grad_norm": 2.656402648097636,
+      "learning_rate": 9.897739792351174e-06,
+      "loss": 0.3301,
+      "step": 218
+    },
+    {
+      "epoch": 0.09379014989293362,
+      "grad_norm": 2.621054458533641,
+      "learning_rate": 9.896339047302772e-06,
+      "loss": 0.34,
+      "step": 219
+    },
+    {
+      "epoch": 0.09421841541755889,
+      "grad_norm": 2.6546842516489355,
+      "learning_rate": 9.89492887428173e-06,
+      "loss": 0.3055,
+      "step": 220
+    },
+    {
+      "epoch": 0.09464668094218416,
+      "grad_norm": 2.94449663462326,
+      "learning_rate": 9.893509276003354e-06,
+      "loss": 0.3138,
+      "step": 221
+    },
+    {
+      "epoch": 0.09507494646680942,
+      "grad_norm": 2.9269850070533763,
+      "learning_rate": 9.892080255201105e-06,
+      "loss": 0.3286,
+      "step": 222
+    },
+    {
+      "epoch": 0.09550321199143469,
+      "grad_norm": 2.6164774453999162,
+      "learning_rate": 9.89064181462658e-06,
+      "loss": 0.3237,
+      "step": 223
+    },
+    {
+      "epoch": 0.09593147751605996,
+      "grad_norm": 2.96080196110219,
+      "learning_rate": 9.88919395704952e-06,
+      "loss": 0.358,
+      "step": 224
+    },
+    {
+      "epoch": 0.09635974304068523,
+      "grad_norm": 2.649785628867192,
+      "learning_rate": 9.887736685257791e-06,
+      "loss": 0.2984,
+      "step": 225
+    },
+    {
+      "epoch": 0.09678800856531049,
+      "grad_norm": 2.6624649894072165,
+      "learning_rate": 9.886270002057399e-06,
+      "loss": 0.3109,
+      "step": 226
+    },
+    {
+      "epoch": 0.09721627408993576,
+      "grad_norm": 2.7125957815484214,
+      "learning_rate": 9.884793910272457e-06,
+      "loss": 0.2804,
+      "step": 227
+    },
+    {
+      "epoch": 0.09764453961456103,
+      "grad_norm": 2.831634773567625,
+      "learning_rate": 9.883308412745206e-06,
+      "loss": 0.3478,
+      "step": 228
+    },
+    {
+      "epoch": 0.0980728051391863,
+      "grad_norm": 3.0630982697502196,
+      "learning_rate": 9.881813512335989e-06,
+      "loss": 0.2955,
+      "step": 229
+    },
+    {
+      "epoch": 0.09850107066381156,
+      "grad_norm": 2.9021472519989784,
+      "learning_rate": 9.880309211923263e-06,
+      "loss": 0.2932,
+      "step": 230
+    },
+    {
+      "epoch": 0.09892933618843683,
+      "grad_norm": 2.770624350418408,
+      "learning_rate": 9.878795514403579e-06,
+      "loss": 0.2828,
+      "step": 231
+    },
+    {
+      "epoch": 0.0993576017130621,
+      "grad_norm": 3.1974983603627214,
+      "learning_rate": 9.877272422691583e-06,
+      "loss": 0.281,
+      "step": 232
+    },
+    {
+      "epoch": 0.09978586723768737,
+      "grad_norm": 2.8748007764094754,
+      "learning_rate": 9.87573993972001e-06,
+      "loss": 0.2795,
+      "step": 233
+    },
+    {
+      "epoch": 0.10021413276231263,
+      "grad_norm": 2.909639398619177,
+      "learning_rate": 9.874198068439682e-06,
+      "loss": 0.3093,
+      "step": 234
+    },
+    {
+      "epoch": 0.1006423982869379,
+      "grad_norm": 2.723783228658883,
+      "learning_rate": 9.872646811819489e-06,
+      "loss": 0.2899,
+      "step": 235
+    },
+    {
+      "epoch": 0.10107066381156317,
+      "grad_norm": 2.5234557322839017,
+      "learning_rate": 9.871086172846403e-06,
+      "loss": 0.2594,
+      "step": 236
+    },
+    {
+      "epoch": 0.10149892933618844,
+      "grad_norm": 2.802296396376116,
+      "learning_rate": 9.869516154525455e-06,
+      "loss": 0.2819,
+      "step": 237
+    },
+    {
+      "epoch": 0.10192719486081371,
+      "grad_norm": 2.66226927196503,
+      "learning_rate": 9.86793675987974e-06,
+      "loss": 0.2864,
+      "step": 238
+    },
+    {
+      "epoch": 0.10235546038543897,
+      "grad_norm": 2.8000967162009665,
+      "learning_rate": 9.866347991950404e-06,
+      "loss": 0.3166,
+      "step": 239
+    },
+    {
+      "epoch": 0.10278372591006424,
+      "grad_norm": 2.945534638620589,
+      "learning_rate": 9.864749853796642e-06,
+      "loss": 0.31,
+      "step": 240
+    },
+    {
+      "epoch": 0.10321199143468951,
+      "grad_norm": 2.6294947090777336,
+      "learning_rate": 9.863142348495696e-06,
+      "loss": 0.2559,
+      "step": 241
+    },
+    {
+      "epoch": 0.10364025695931478,
+      "grad_norm": 3.020344481814639,
+      "learning_rate": 9.861525479142839e-06,
+      "loss": 0.3039,
+      "step": 242
+    },
+    {
+      "epoch": 0.10406852248394004,
+      "grad_norm": 2.7925900037377973,
+      "learning_rate": 9.859899248851374e-06,
+      "loss": 0.2271,
+      "step": 243
+    },
+    {
+      "epoch": 0.10449678800856531,
+      "grad_norm": 2.6771271735470497,
+      "learning_rate": 9.858263660752637e-06,
+      "loss": 0.3511,
+      "step": 244
+    },
+    {
+      "epoch": 0.10492505353319058,
+      "grad_norm": 2.7339576559924637,
+      "learning_rate": 9.856618717995972e-06,
+      "loss": 0.3279,
+      "step": 245
+    },
+    {
+      "epoch": 0.10535331905781585,
+      "grad_norm": 2.751387240394507,
+      "learning_rate": 9.854964423748743e-06,
+      "loss": 0.3013,
+      "step": 246
+    },
+    {
+      "epoch": 0.10578158458244111,
+      "grad_norm": 2.7222182983701413,
+      "learning_rate": 9.853300781196316e-06,
+      "loss": 0.3225,
+      "step": 247
+    },
+    {
+      "epoch": 0.10620985010706638,
+      "grad_norm": 2.6673677573080354,
+      "learning_rate": 9.85162779354206e-06,
+      "loss": 0.3379,
+      "step": 248
+    },
+    {
+      "epoch": 0.10663811563169165,
+      "grad_norm": 2.634312588040267,
+      "learning_rate": 9.84994546400734e-06,
+      "loss": 0.2789,
+      "step": 249
+    },
+    {
+      "epoch": 0.10706638115631692,
+      "grad_norm": 2.9548383252603063,
+      "learning_rate": 9.848253795831501e-06,
+      "loss": 0.3129,
+      "step": 250
+    },
+    {
+      "epoch": 0.10749464668094219,
+      "grad_norm": 2.735648574756119,
+      "learning_rate": 9.846552792271878e-06,
+      "loss": 0.2997,
+      "step": 251
+    },
+    {
+      "epoch": 0.10792291220556745,
+      "grad_norm": 2.934767439552992,
+      "learning_rate": 9.844842456603779e-06,
+      "loss": 0.3242,
+      "step": 252
+    },
+    {
+      "epoch": 0.10835117773019272,
+      "grad_norm": 3.4004494941766628,
+      "learning_rate": 9.843122792120478e-06,
+      "loss": 0.3226,
+      "step": 253
+    },
+    {
+      "epoch": 0.10877944325481799,
+      "grad_norm": 2.737399774677961,
+      "learning_rate": 9.841393802133214e-06,
+      "loss": 0.2702,
+      "step": 254
+    },
+    {
+      "epoch": 0.10920770877944326,
+      "grad_norm": 2.5849839371425176,
+      "learning_rate": 9.839655489971186e-06,
+      "loss": 0.2724,
+      "step": 255
+    },
+    {
+      "epoch": 0.10963597430406852,
+      "grad_norm": 2.6869678487015407,
+      "learning_rate": 9.837907858981536e-06,
+      "loss": 0.2738,
+      "step": 256
+    },
+    {
+      "epoch": 0.11006423982869379,
+      "grad_norm": 2.999876564800314,
+      "learning_rate": 9.836150912529357e-06,
+      "loss": 0.3389,
+      "step": 257
+    },
+    {
+      "epoch": 0.11049250535331906,
+      "grad_norm": 2.812020484265115,
+      "learning_rate": 9.83438465399767e-06,
+      "loss": 0.2871,
+      "step": 258
+    },
+    {
+      "epoch": 0.11092077087794433,
+      "grad_norm": 2.658456632268773,
+      "learning_rate": 9.832609086787437e-06,
+      "loss": 0.3262,
+      "step": 259
+    },
+    {
+      "epoch": 0.11134903640256959,
+      "grad_norm": 2.437719014201623,
+      "learning_rate": 9.830824214317533e-06,
+      "loss": 0.2866,
+      "step": 260
+    },
+    {
+      "epoch": 0.11177730192719486,
+      "grad_norm": 2.73865020132295,
+      "learning_rate": 9.829030040024762e-06,
+      "loss": 0.2928,
+      "step": 261
+    },
+    {
+      "epoch": 0.11220556745182013,
+      "grad_norm": 3.01166736631592,
+      "learning_rate": 9.82722656736383e-06,
+      "loss": 0.3146,
+      "step": 262
+    },
+    {
+      "epoch": 0.1126338329764454,
+      "grad_norm": 2.7678710085116136,
+      "learning_rate": 9.825413799807348e-06,
+      "loss": 0.3273,
+      "step": 263
+    },
+    {
+      "epoch": 0.11306209850107066,
+      "grad_norm": 2.721298107680012,
+      "learning_rate": 9.823591740845831e-06,
+      "loss": 0.2686,
+      "step": 264
+    },
+    {
+      "epoch": 0.11349036402569593,
+      "grad_norm": 2.6653320599632813,
+      "learning_rate": 9.821760393987679e-06,
+      "loss": 0.2699,
+      "step": 265
+    },
+    {
+      "epoch": 0.1139186295503212,
+      "grad_norm": 2.806614929132685,
+      "learning_rate": 9.819919762759174e-06,
+      "loss": 0.3031,
+      "step": 266
+    },
+    {
+      "epoch": 0.11434689507494647,
+      "grad_norm": 2.9452593541761924,
+      "learning_rate": 9.81806985070448e-06,
+      "loss": 0.2916,
+      "step": 267
+    },
+    {
+      "epoch": 0.11477516059957174,
+      "grad_norm": 2.9609984373431875,
+      "learning_rate": 9.816210661385633e-06,
+      "loss": 0.2832,
+      "step": 268
+    },
+    {
+      "epoch": 0.115203426124197,
+      "grad_norm": 3.067736079191241,
+      "learning_rate": 9.814342198382524e-06,
+      "loss": 0.3456,
+      "step": 269
+    },
+    {
+      "epoch": 0.11563169164882227,
+      "grad_norm": 2.6731133040136195,
+      "learning_rate": 9.81246446529291e-06,
+      "loss": 0.3489,
+      "step": 270
+    },
+    {
+      "epoch": 0.11605995717344754,
+      "grad_norm": 2.765460225885666,
+      "learning_rate": 9.810577465732393e-06,
+      "loss": 0.3139,
+      "step": 271
+    },
+    {
+      "epoch": 0.11648822269807281,
+      "grad_norm": 2.6525210851096865,
+      "learning_rate": 9.808681203334416e-06,
+      "loss": 0.2603,
+      "step": 272
+    },
+    {
+      "epoch": 0.11691648822269807,
+      "grad_norm": 2.5465747438651425,
+      "learning_rate": 9.806775681750262e-06,
+      "loss": 0.304,
+      "step": 273
+    },
+    {
+      "epoch": 0.11734475374732334,
+      "grad_norm": 2.5815373748697894,
+      "learning_rate": 9.804860904649044e-06,
+      "loss": 0.2854,
+      "step": 274
+    },
+    {
+      "epoch": 0.11777301927194861,
+      "grad_norm": 2.7712026380273795,
+      "learning_rate": 9.802936875717685e-06,
+      "loss": 0.3079,
+      "step": 275
+    },
+    {
+      "epoch": 0.11820128479657388,
+      "grad_norm": 2.580379751231512,
+      "learning_rate": 9.801003598660937e-06,
+      "loss": 0.313,
+      "step": 276
+    },
+    {
+      "epoch": 0.11862955032119914,
+      "grad_norm": 2.8426196149874436,
+      "learning_rate": 9.799061077201353e-06,
+      "loss": 0.3364,
+      "step": 277
+    },
+    {
+      "epoch": 0.11905781584582441,
+      "grad_norm": 2.5998496078385176,
+      "learning_rate": 9.797109315079285e-06,
+      "loss": 0.2865,
+      "step": 278
+    },
+    {
+      "epoch": 0.11948608137044968,
+      "grad_norm": 2.841981114685137,
+      "learning_rate": 9.79514831605288e-06,
+      "loss": 0.3094,
+      "step": 279
+    },
+    {
+      "epoch": 0.11991434689507495,
+      "grad_norm": 2.754151136459746,
+      "learning_rate": 9.793178083898073e-06,
+      "loss": 0.2904,
+      "step": 280
+    },
+    {
+      "epoch": 0.12034261241970022,
+      "grad_norm": 2.7828445713976757,
+      "learning_rate": 9.791198622408568e-06,
+      "loss": 0.2884,
+      "step": 281
+    },
+    {
+      "epoch": 0.12077087794432548,
+      "grad_norm": 2.543277730627548,
+      "learning_rate": 9.789209935395855e-06,
+      "loss": 0.2674,
+      "step": 282
+    },
+    {
+      "epoch": 0.12119914346895075,
+      "grad_norm": 2.5968118143410375,
+      "learning_rate": 9.787212026689174e-06,
+      "loss": 0.2767,
+      "step": 283
+    },
+    {
+      "epoch": 0.12162740899357602,
+      "grad_norm": 2.6766947019712943,
+      "learning_rate": 9.785204900135533e-06,
+      "loss": 0.3223,
+      "step": 284
+    },
+    {
+      "epoch": 0.12205567451820129,
+      "grad_norm": 2.6525834638779537,
+      "learning_rate": 9.78318855959968e-06,
+      "loss": 0.3081,
+      "step": 285
+    },
+    {
+      "epoch": 0.12248394004282655,
+      "grad_norm": 2.8576587731959506,
+      "learning_rate": 9.781163008964108e-06,
+      "loss": 0.2964,
+      "step": 286
+    },
+    {
+      "epoch": 0.12291220556745182,
+      "grad_norm": 2.8139798539335277,
+      "learning_rate": 9.779128252129046e-06,
+      "loss": 0.2901,
+      "step": 287
+    },
+    {
+      "epoch": 0.12334047109207709,
+      "grad_norm": 2.890013250585725,
+      "learning_rate": 9.777084293012448e-06,
+      "loss": 0.3257,
+      "step": 288
+    },
+    {
+      "epoch": 0.12376873661670236,
+      "grad_norm": 2.6169459883564072,
+      "learning_rate": 9.775031135549987e-06,
+      "loss": 0.259,
+      "step": 289
+    },
+    {
+      "epoch": 0.12419700214132762,
+      "grad_norm": 2.8358980850093163,
+      "learning_rate": 9.772968783695046e-06,
+      "loss": 0.2693,
+      "step": 290
+    },
+    {
+      "epoch": 0.12462526766595289,
+      "grad_norm": 3.0267519225632746,
+      "learning_rate": 9.770897241418717e-06,
+      "loss": 0.2997,
+      "step": 291
+    },
+    {
+      "epoch": 0.12505353319057816,
+      "grad_norm": 2.9006473465753064,
+      "learning_rate": 9.768816512709782e-06,
+      "loss": 0.2747,
+      "step": 292
+    },
+    {
+      "epoch": 0.12548179871520343,
+      "grad_norm": 2.724823390582959,
+      "learning_rate": 9.766726601574717e-06,
+      "loss": 0.2835,
+      "step": 293
+    },
+    {
+      "epoch": 0.1259100642398287,
+      "grad_norm": 2.786952612657992,
+      "learning_rate": 9.764627512037676e-06,
+      "loss": 0.3396,
+      "step": 294
+    },
+    {
+      "epoch": 0.12633832976445397,
+      "grad_norm": 2.7903310825495518,
+      "learning_rate": 9.762519248140484e-06,
+      "loss": 0.2681,
+      "step": 295
+    },
+    {
+      "epoch": 0.12676659528907924,
+      "grad_norm": 2.7155788390638915,
+      "learning_rate": 9.760401813942641e-06,
+      "loss": 0.3185,
+      "step": 296
+    },
+    {
+      "epoch": 0.12719486081370449,
+      "grad_norm": 2.677761585724147,
+      "learning_rate": 9.75827521352129e-06,
+      "loss": 0.2917,
+      "step": 297
+    },
+    {
+      "epoch": 0.12762312633832976,
+      "grad_norm": 2.7615810376869363,
+      "learning_rate": 9.756139450971234e-06,
+      "loss": 0.3697,
+      "step": 298
+    },
+    {
+      "epoch": 0.12805139186295503,
+      "grad_norm": 2.69629020069806,
+      "learning_rate": 9.753994530404915e-06,
+      "loss": 0.2924,
+      "step": 299
+    },
+    {
+      "epoch": 0.1284796573875803,
+      "grad_norm": 2.8978165537763214,
+      "learning_rate": 9.751840455952411e-06,
+      "loss": 0.3163,
+      "step": 300
+    },
+    {
+      "epoch": 0.12890792291220557,
+      "grad_norm": 3.0266167825272032,
+      "learning_rate": 9.749677231761421e-06,
+      "loss": 0.273,
+      "step": 301
+    },
+    {
+      "epoch": 0.12933618843683084,
+      "grad_norm": 2.959497866209961,
+      "learning_rate": 9.747504861997269e-06,
+      "loss": 0.3357,
+      "step": 302
+    },
+    {
+      "epoch": 0.1297644539614561,
+      "grad_norm": 2.8794089039958584,
+      "learning_rate": 9.745323350842877e-06,
+      "loss": 0.3602,
+      "step": 303
+    },
+    {
+      "epoch": 0.13019271948608138,
+      "grad_norm": 2.7642159408835414,
+      "learning_rate": 9.743132702498785e-06,
+      "loss": 0.3298,
+      "step": 304
+    },
+    {
+      "epoch": 0.13062098501070663,
+      "grad_norm": 2.751604847164902,
+      "learning_rate": 9.740932921183115e-06,
+      "loss": 0.3215,
+      "step": 305
+    },
+    {
+      "epoch": 0.1310492505353319,
+      "grad_norm": 2.706873214963062,
+      "learning_rate": 9.738724011131578e-06,
+      "loss": 0.3185,
+      "step": 306
+    },
+    {
+      "epoch": 0.13147751605995717,
+      "grad_norm": 2.7169312151027727,
+      "learning_rate": 9.736505976597463e-06,
+      "loss": 0.261,
+      "step": 307
+    },
+    {
+      "epoch": 0.13190578158458244,
+      "grad_norm": 2.7239036143337523,
+      "learning_rate": 9.734278821851631e-06,
+      "loss": 0.3495,
+      "step": 308
+    },
+    {
+      "epoch": 0.1323340471092077,
+      "grad_norm": 2.8121802867224197,
+      "learning_rate": 9.732042551182495e-06,
+      "loss": 0.3032,
+      "step": 309
+    },
+    {
+      "epoch": 0.13276231263383298,
+      "grad_norm": 2.846129881442966,
+      "learning_rate": 9.729797168896032e-06,
+      "loss": 0.3008,
+      "step": 310
+    },
+    {
+      "epoch": 0.13319057815845825,
+      "grad_norm": 3.106864486065552,
+      "learning_rate": 9.727542679315756e-06,
+      "loss": 0.3584,
+      "step": 311
+    },
+    {
+      "epoch": 0.13361884368308352,
+      "grad_norm": 2.9942521573665384,
+      "learning_rate": 9.725279086782719e-06,
+      "loss": 0.3077,
+      "step": 312
+    },
+    {
+      "epoch": 0.1340471092077088,
+      "grad_norm": 3.2441977046066404,
+      "learning_rate": 9.723006395655505e-06,
+      "loss": 0.3317,
+      "step": 313
+    },
+    {
+      "epoch": 0.13447537473233404,
+      "grad_norm": 2.943806648708445,
+      "learning_rate": 9.720724610310212e-06,
+      "loss": 0.3192,
+      "step": 314
+    },
+    {
+      "epoch": 0.1349036402569593,
+      "grad_norm": 2.6929595037143814,
+      "learning_rate": 9.718433735140454e-06,
+      "loss": 0.3123,
+      "step": 315
+    },
+    {
+      "epoch": 0.13533190578158458,
+      "grad_norm": 2.504578020337929,
+      "learning_rate": 9.716133774557337e-06,
+      "loss": 0.3189,
+      "step": 316
+    },
+    {
+      "epoch": 0.13576017130620985,
+      "grad_norm": 2.758670304524646,
+      "learning_rate": 9.713824732989479e-06,
+      "loss": 0.27,
+      "step": 317
+    },
+    {
+      "epoch": 0.13618843683083512,
+      "grad_norm": 2.5694635115602633,
+      "learning_rate": 9.711506614882965e-06,
+      "loss": 0.2594,
+      "step": 318
+    },
+    {
+      "epoch": 0.1366167023554604,
+      "grad_norm": 2.5835947472656935,
+      "learning_rate": 9.70917942470137e-06,
+      "loss": 0.3248,
+      "step": 319
+    },
+    {
+      "epoch": 0.13704496788008566,
+      "grad_norm": 2.7065841302676343,
+      "learning_rate": 9.706843166925733e-06,
+      "loss": 0.2716,
+      "step": 320
+    },
+    {
+      "epoch": 0.13747323340471093,
+      "grad_norm": 2.897700311484326,
+      "learning_rate": 9.704497846054548e-06,
+      "loss": 0.3194,
+      "step": 321
+    },
+    {
+      "epoch": 0.13790149892933617,
+      "grad_norm": 3.000499224123425,
+      "learning_rate": 9.70214346660377e-06,
+      "loss": 0.316,
+      "step": 322
+    },
+    {
+      "epoch": 0.13832976445396145,
+      "grad_norm": 2.5483862613520945,
+      "learning_rate": 9.69978003310679e-06,
+      "loss": 0.3002,
+      "step": 323
+    },
+    {
+      "epoch": 0.13875802997858672,
+      "grad_norm": 2.7839566014950985,
+      "learning_rate": 9.69740755011443e-06,
+      "loss": 0.2681,
+      "step": 324
+    },
+    {
+      "epoch": 0.139186295503212,
+      "grad_norm": 2.517006313802462,
+      "learning_rate": 9.695026022194947e-06,
+      "loss": 0.2773,
+      "step": 325
+    },
+    {
+      "epoch": 0.13961456102783726,
+      "grad_norm": 2.745961983153208,
+      "learning_rate": 9.692635453934003e-06,
+      "loss": 0.2855,
+      "step": 326
+    },
+    {
+      "epoch": 0.14004282655246253,
+      "grad_norm": 2.754716156209928,
+      "learning_rate": 9.690235849934675e-06,
+      "loss": 0.2683,
+      "step": 327
+    },
+    {
+      "epoch": 0.1404710920770878,
+      "grad_norm": 3.066139543777544,
+      "learning_rate": 9.687827214817433e-06,
+      "loss": 0.3003,
+      "step": 328
+    },
+    {
+      "epoch": 0.14089935760171307,
+      "grad_norm": 2.4814805111071645,
+      "learning_rate": 9.685409553220142e-06,
+      "loss": 0.2923,
+      "step": 329
+    },
+    {
+      "epoch": 0.14132762312633834,
+      "grad_norm": 2.8438591865593352,
+      "learning_rate": 9.682982869798043e-06,
+      "loss": 0.2802,
+      "step": 330
+    },
+    {
+      "epoch": 0.14175588865096359,
+      "grad_norm": 2.8930168706747486,
+      "learning_rate": 9.680547169223752e-06,
+      "loss": 0.2805,
+      "step": 331
+    },
+    {
+      "epoch": 0.14218415417558886,
+      "grad_norm": 2.5214393917315374,
+      "learning_rate": 9.678102456187246e-06,
+      "loss": 0.2962,
+      "step": 332
+    },
+    {
+      "epoch": 0.14261241970021413,
+      "grad_norm": 2.718051490957702,
+      "learning_rate": 9.675648735395855e-06,
+      "loss": 0.3173,
+      "step": 333
+    },
+    {
+      "epoch": 0.1430406852248394,
+      "grad_norm": 2.633837926324993,
+      "learning_rate": 9.673186011574254e-06,
+      "loss": 0.3098,
+      "step": 334
+    },
+    {
+      "epoch": 0.14346895074946467,
+      "grad_norm": 2.8756312002346944,
+      "learning_rate": 9.670714289464454e-06,
+      "loss": 0.2704,
+      "step": 335
+    },
+    {
+      "epoch": 0.14389721627408994,
+      "grad_norm": 2.8333231817008886,
+      "learning_rate": 9.668233573825794e-06,
+      "loss": 0.2802,
+      "step": 336
+    },
+    {
+      "epoch": 0.1443254817987152,
+      "grad_norm": 3.0091705925428482,
+      "learning_rate": 9.665743869434925e-06,
+      "loss": 0.2715,
+      "step": 337
+    },
+    {
+      "epoch": 0.14475374732334048,
+      "grad_norm": 2.8970462741430443,
+      "learning_rate": 9.663245181085812e-06,
+      "loss": 0.2455,
+      "step": 338
+    },
+    {
+      "epoch": 0.14518201284796575,
+      "grad_norm": 3.086386934133251,
+      "learning_rate": 9.660737513589717e-06,
+      "loss": 0.2927,
+      "step": 339
+    },
+    {
+      "epoch": 0.145610278372591,
+      "grad_norm": 3.069082257492287,
+      "learning_rate": 9.658220871775188e-06,
+      "loss": 0.2824,
+      "step": 340
+    },
+    {
+      "epoch": 0.14603854389721627,
+      "grad_norm": 2.6917123662194973,
+      "learning_rate": 9.655695260488057e-06,
+      "loss": 0.2598,
+      "step": 341
+    },
+    {
+      "epoch": 0.14646680942184154,
+      "grad_norm": 2.8487596445161154,
+      "learning_rate": 9.653160684591423e-06,
+      "loss": 0.2907,
+      "step": 342
+    },
+    {
+      "epoch": 0.1468950749464668,
+      "grad_norm": 2.69477854677915,
+      "learning_rate": 9.650617148965653e-06,
+      "loss": 0.3246,
+      "step": 343
+    },
+    {
+      "epoch": 0.14732334047109208,
+      "grad_norm": 2.6471586692859335,
+      "learning_rate": 9.64806465850836e-06,
+      "loss": 0.2862,
+      "step": 344
+    },
+    {
+      "epoch": 0.14775160599571735,
+      "grad_norm": 2.6761382180066704,
+      "learning_rate": 9.645503218134402e-06,
+      "loss": 0.2552,
+      "step": 345
+    },
+    {
+      "epoch": 0.14817987152034262,
+      "grad_norm": 2.5999493129943905,
+      "learning_rate": 9.642932832775871e-06,
+      "loss": 0.3155,
+      "step": 346
+    },
+    {
+      "epoch": 0.1486081370449679,
+      "grad_norm": 2.667772391056551,
+      "learning_rate": 9.640353507382081e-06,
+      "loss": 0.2984,
+      "step": 347
+    },
+    {
+      "epoch": 0.14903640256959313,
+      "grad_norm": 2.6827137294255254,
+      "learning_rate": 9.637765246919559e-06,
+      "loss": 0.2916,
+      "step": 348
+    },
+    {
+      "epoch": 0.1494646680942184,
+      "grad_norm": 2.8945163701201495,
+      "learning_rate": 9.635168056372041e-06,
+      "loss": 0.3062,
+      "step": 349
+    },
+    {
+      "epoch": 0.14989293361884368,
+      "grad_norm": 2.9505889215310623,
+      "learning_rate": 9.632561940740456e-06,
+      "loss": 0.3245,
+      "step": 350
+    },
+    {
+      "epoch": 0.15032119914346895,
+      "grad_norm": 2.765240723148734,
+      "learning_rate": 9.629946905042917e-06,
+      "loss": 0.3114,
+      "step": 351
+    },
+    {
+      "epoch": 0.15074946466809422,
+      "grad_norm": 2.777321690212769,
+      "learning_rate": 9.62732295431471e-06,
+      "loss": 0.2817,
+      "step": 352
+    },
+    {
+      "epoch": 0.1511777301927195,
+      "grad_norm": 2.581893041596423,
+      "learning_rate": 9.624690093608297e-06,
+      "loss": 0.2813,
+      "step": 353
+    },
+    {
+      "epoch": 0.15160599571734476,
+      "grad_norm": 2.808516542661495,
+      "learning_rate": 9.622048327993285e-06,
+      "loss": 0.2672,
+      "step": 354
+    },
+    {
+      "epoch": 0.15203426124197003,
+      "grad_norm": 2.9156296752032587,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.2808,
+      "step": 355
+    },
+    {
+      "epoch": 0.1524625267665953,
+      "grad_norm": 2.66616816068596,
+      "learning_rate": 9.616738102401641e-06,
+      "loss": 0.2941,
+      "step": 356
+    },
+    {
+      "epoch": 0.15289079229122055,
+      "grad_norm": 2.5642454890184596,
+      "learning_rate": 9.614069652649925e-06,
+      "loss": 0.3594,
+      "step": 357
+    },
+    {
+      "epoch": 0.15331905781584582,
+      "grad_norm": 2.7900784774806833,
+      "learning_rate": 9.611392318439427e-06,
+      "loss": 0.3698,
+      "step": 358
+    },
+    {
+      "epoch": 0.1537473233404711,
+      "grad_norm": 2.749639394306116,
+      "learning_rate": 9.608706104925391e-06,
+      "loss": 0.3108,
+      "step": 359
+    },
+    {
+      "epoch": 0.15417558886509636,
+      "grad_norm": 2.595297416582149,
+      "learning_rate": 9.606011017280166e-06,
+      "loss": 0.2672,
+      "step": 360
+    },
+    {
+      "epoch": 0.15460385438972163,
+      "grad_norm": 2.7749152091777547,
+      "learning_rate": 9.603307060693176e-06,
+      "loss": 0.3004,
+      "step": 361
+    },
+    {
+      "epoch": 0.1550321199143469,
+      "grad_norm": 2.628844557708143,
+      "learning_rate": 9.600594240370933e-06,
+      "loss": 0.2898,
+      "step": 362
+    },
+    {
+      "epoch": 0.15546038543897217,
+      "grad_norm": 2.7967956306349957,
+      "learning_rate": 9.597872561537011e-06,
+      "loss": 0.3183,
+      "step": 363
+    },
+    {
+      "epoch": 0.15588865096359744,
+      "grad_norm": 2.8780150649135483,
+      "learning_rate": 9.595142029432044e-06,
+      "loss": 0.2927,
+      "step": 364
+    },
+    {
+      "epoch": 0.15631691648822268,
+      "grad_norm": 3.3900621795015993,
+      "learning_rate": 9.592402649313711e-06,
+      "loss": 0.2892,
+      "step": 365
+    },
+    {
+      "epoch": 0.15674518201284796,
+      "grad_norm": 3.441900130002736,
+      "learning_rate": 9.589654426456726e-06,
+      "loss": 0.3022,
+      "step": 366
+    },
+    {
+      "epoch": 0.15717344753747323,
+      "grad_norm": 3.0060204585425807,
+      "learning_rate": 9.586897366152836e-06,
+      "loss": 0.3113,
+      "step": 367
+    },
+    {
+      "epoch": 0.1576017130620985,
+      "grad_norm": 2.782992805444639,
+      "learning_rate": 9.5841314737108e-06,
+      "loss": 0.3256,
+      "step": 368
+    },
+    {
+      "epoch": 0.15802997858672377,
+      "grad_norm": 2.9038655985318087,
+      "learning_rate": 9.581356754456384e-06,
+      "loss": 0.31,
+      "step": 369
+    },
+    {
+      "epoch": 0.15845824411134904,
+      "grad_norm": 2.59806689807304,
+      "learning_rate": 9.57857321373235e-06,
+      "loss": 0.3455,
+      "step": 370
+    },
+    {
+      "epoch": 0.1588865096359743,
+      "grad_norm": 2.709553745045623,
+      "learning_rate": 9.57578085689845e-06,
+      "loss": 0.3167,
+      "step": 371
+    },
+    {
+      "epoch": 0.15931477516059958,
+      "grad_norm": 2.582029866473802,
+      "learning_rate": 9.572979689331402e-06,
+      "loss": 0.318,
+      "step": 372
+    },
+    {
+      "epoch": 0.15974304068522485,
+      "grad_norm": 2.67013882624578,
+      "learning_rate": 9.570169716424899e-06,
+      "loss": 0.3283,
+      "step": 373
+    },
+    {
+      "epoch": 0.1601713062098501,
+      "grad_norm": 2.602878279394627,
+      "learning_rate": 9.567350943589585e-06,
+      "loss": 0.2551,
+      "step": 374
+    },
+    {
+      "epoch": 0.16059957173447537,
+      "grad_norm": 2.7482517974172436,
+      "learning_rate": 9.564523376253047e-06,
+      "loss": 0.2928,
+      "step": 375
+    },
+    {
+      "epoch": 0.16102783725910064,
+      "grad_norm": 2.8501392462065196,
+      "learning_rate": 9.56168701985981e-06,
+      "loss": 0.2689,
+      "step": 376
+    },
+    {
+      "epoch": 0.1614561027837259,
+      "grad_norm": 3.0193479203885585,
+      "learning_rate": 9.558841879871316e-06,
+      "loss": 0.2871,
+      "step": 377
+    },
+    {
+      "epoch": 0.16188436830835118,
+      "grad_norm": 2.891660304392641,
+      "learning_rate": 9.555987961765929e-06,
+      "loss": 0.2587,
+      "step": 378
+    },
+    {
+      "epoch": 0.16231263383297645,
+      "grad_norm": 3.0477518929806475,
+      "learning_rate": 9.553125271038905e-06,
+      "loss": 0.2919,
+      "step": 379
+    },
+    {
+      "epoch": 0.16274089935760172,
+      "grad_norm": 3.282166749133352,
+      "learning_rate": 9.5502538132024e-06,
+      "loss": 0.3056,
+      "step": 380
+    },
+    {
+      "epoch": 0.163169164882227,
+      "grad_norm": 3.2957211971706717,
+      "learning_rate": 9.547373593785447e-06,
+      "loss": 0.3557,
+      "step": 381
+    },
+    {
+      "epoch": 0.16359743040685226,
+      "grad_norm": 2.6592266411926238,
+      "learning_rate": 9.54448461833395e-06,
+      "loss": 0.2582,
+      "step": 382
+    },
+    {
+      "epoch": 0.1640256959314775,
+      "grad_norm": 2.7433710705818646,
+      "learning_rate": 9.541586892410674e-06,
+      "loss": 0.2773,
+      "step": 383
+    },
+    {
+      "epoch": 0.16445396145610278,
+      "grad_norm": 2.700179813984929,
+      "learning_rate": 9.538680421595236e-06,
+      "loss": 0.274,
+      "step": 384
+    },
+    {
+      "epoch": 0.16488222698072805,
+      "grad_norm": 2.53026612487461,
+      "learning_rate": 9.535765211484083e-06,
+      "loss": 0.2358,
+      "step": 385
+    },
+    {
+      "epoch": 0.16531049250535332,
+      "grad_norm": 3.138091388945966,
+      "learning_rate": 9.532841267690499e-06,
+      "loss": 0.3676,
+      "step": 386
+    },
+    {
+      "epoch": 0.1657387580299786,
+      "grad_norm": 2.585803647887988,
+      "learning_rate": 9.52990859584458e-06,
+      "loss": 0.3044,
+      "step": 387
+    },
+    {
+      "epoch": 0.16616702355460386,
+      "grad_norm": 2.5322733040064445,
+      "learning_rate": 9.526967201593225e-06,
+      "loss": 0.2733,
+      "step": 388
+    },
+    {
+      "epoch": 0.16659528907922913,
+      "grad_norm": 2.8551833071379615,
+      "learning_rate": 9.524017090600135e-06,
+      "loss": 0.2889,
+      "step": 389
+    },
+    {
+      "epoch": 0.1670235546038544,
+      "grad_norm": 2.6891507553347465,
+      "learning_rate": 9.521058268545792e-06,
+      "loss": 0.2963,
+      "step": 390
+    },
+    {
+      "epoch": 0.16745182012847964,
+      "grad_norm": 3.1802706896710466,
+      "learning_rate": 9.518090741127451e-06,
+      "loss": 0.3164,
+      "step": 391
+    },
+    {
+      "epoch": 0.16788008565310492,
+      "grad_norm": 2.839364438841274,
+      "learning_rate": 9.515114514059127e-06,
+      "loss": 0.2631,
+      "step": 392
+    },
+    {
+      "epoch": 0.1683083511777302,
+      "grad_norm": 2.7370040666976037,
+      "learning_rate": 9.512129593071593e-06,
+      "loss": 0.2856,
+      "step": 393
+    },
+    {
+      "epoch": 0.16873661670235546,
+      "grad_norm": 2.638797872092445,
+      "learning_rate": 9.509135983912356e-06,
+      "loss": 0.2676,
+      "step": 394
+    },
+    {
+      "epoch": 0.16916488222698073,
+      "grad_norm": 2.996836288605135,
+      "learning_rate": 9.506133692345655e-06,
+      "loss": 0.2906,
+      "step": 395
+    },
+    {
+      "epoch": 0.169593147751606,
+      "grad_norm": 3.0016211801643995,
+      "learning_rate": 9.503122724152445e-06,
+      "loss": 0.3078,
+      "step": 396
+    },
+    {
+      "epoch": 0.17002141327623127,
+      "grad_norm": 2.839457258188891,
+      "learning_rate": 9.500103085130391e-06,
+      "loss": 0.2929,
+      "step": 397
+    },
+    {
+      "epoch": 0.17044967880085654,
+      "grad_norm": 2.556126097010031,
+      "learning_rate": 9.497074781093851e-06,
+      "loss": 0.3245,
+      "step": 398
+    },
+    {
+      "epoch": 0.1708779443254818,
+      "grad_norm": 2.730086174938734,
+      "learning_rate": 9.49403781787387e-06,
+      "loss": 0.2919,
+      "step": 399
+    },
+    {
+      "epoch": 0.17130620985010706,
+      "grad_norm": 2.7661879079272884,
+      "learning_rate": 9.490992201318165e-06,
+      "loss": 0.2898,
+      "step": 400
+    },
+    {
+      "epoch": 0.17173447537473233,
+      "grad_norm": 2.713817879712303,
+      "learning_rate": 9.487937937291113e-06,
+      "loss": 0.2839,
+      "step": 401
+    },
+    {
+      "epoch": 0.1721627408993576,
+      "grad_norm": 2.629056862090074,
+      "learning_rate": 9.48487503167375e-06,
+      "loss": 0.268,
+      "step": 402
+    },
+    {
+      "epoch": 0.17259100642398287,
+      "grad_norm": 2.7325159020937515,
+      "learning_rate": 9.481803490363739e-06,
+      "loss": 0.3011,
+      "step": 403
+    },
+    {
+      "epoch": 0.17301927194860814,
+      "grad_norm": 3.0889786120683684,
+      "learning_rate": 9.47872331927538e-06,
+      "loss": 0.3057,
+      "step": 404
+    },
+    {
+      "epoch": 0.1734475374732334,
+      "grad_norm": 2.8927795008135746,
+      "learning_rate": 9.47563452433959e-06,
+      "loss": 0.269,
+      "step": 405
+    },
+    {
+      "epoch": 0.17387580299785868,
+      "grad_norm": 3.045137322567884,
+      "learning_rate": 9.472537111503887e-06,
+      "loss": 0.3005,
+      "step": 406
+    },
+    {
+      "epoch": 0.17430406852248395,
+      "grad_norm": 2.8353936590649873,
+      "learning_rate": 9.469431086732385e-06,
+      "loss": 0.2887,
+      "step": 407
+    },
+    {
+      "epoch": 0.1747323340471092,
+      "grad_norm": 2.754140180590903,
+      "learning_rate": 9.466316456005783e-06,
+      "loss": 0.2971,
+      "step": 408
+    },
+    {
+      "epoch": 0.17516059957173447,
+      "grad_norm": 2.962677150288555,
+      "learning_rate": 9.46319322532134e-06,
+      "loss": 0.3472,
+      "step": 409
+    },
+    {
+      "epoch": 0.17558886509635974,
+      "grad_norm": 2.5475723641278063,
+      "learning_rate": 9.460061400692892e-06,
+      "loss": 0.2878,
+      "step": 410
+    },
+    {
+      "epoch": 0.176017130620985,
+      "grad_norm": 2.7419333566902027,
+      "learning_rate": 9.456920988150809e-06,
+      "loss": 0.3281,
+      "step": 411
+    },
+    {
+      "epoch": 0.17644539614561028,
+      "grad_norm": 2.5240073923683393,
+      "learning_rate": 9.453771993742e-06,
+      "loss": 0.2221,
+      "step": 412
+    },
+    {
+      "epoch": 0.17687366167023555,
+      "grad_norm": 2.546823466775969,
+      "learning_rate": 9.450614423529903e-06,
+      "loss": 0.2514,
+      "step": 413
+    },
+    {
+      "epoch": 0.17730192719486082,
+      "grad_norm": 2.5636903460818763,
+      "learning_rate": 9.447448283594462e-06,
+      "loss": 0.2792,
+      "step": 414
+    },
+    {
+      "epoch": 0.1777301927194861,
+      "grad_norm": 2.928601970221406,
+      "learning_rate": 9.444273580032131e-06,
+      "loss": 0.3303,
+      "step": 415
+    },
+    {
+      "epoch": 0.17815845824411136,
+      "grad_norm": 2.732904407005547,
+      "learning_rate": 9.441090318955843e-06,
+      "loss": 0.2554,
+      "step": 416
+    },
+    {
+      "epoch": 0.1785867237687366,
+      "grad_norm": 2.857294808752289,
+      "learning_rate": 9.437898506495017e-06,
+      "loss": 0.2886,
+      "step": 417
+    },
+    {
+      "epoch": 0.17901498929336188,
+      "grad_norm": 3.2654998593234947,
+      "learning_rate": 9.434698148795534e-06,
+      "loss": 0.362,
+      "step": 418
+    },
+    {
+      "epoch": 0.17944325481798715,
+      "grad_norm": 2.731272621997461,
+      "learning_rate": 9.431489252019731e-06,
+      "loss": 0.2763,
+      "step": 419
+    },
+    {
+      "epoch": 0.17987152034261242,
+      "grad_norm": 2.6276937787402486,
+      "learning_rate": 9.428271822346384e-06,
+      "loss": 0.2955,
+      "step": 420
+    },
+    {
+      "epoch": 0.1802997858672377,
+      "grad_norm": 2.6030068966618716,
+      "learning_rate": 9.425045865970702e-06,
+      "loss": 0.2662,
+      "step": 421
+    },
+    {
+      "epoch": 0.18072805139186296,
+      "grad_norm": 2.6409437938311715,
+      "learning_rate": 9.42181138910431e-06,
+      "loss": 0.3144,
+      "step": 422
+    },
+    {
+      "epoch": 0.18115631691648823,
+      "grad_norm": 2.9340026596200914,
+      "learning_rate": 9.418568397975241e-06,
+      "loss": 0.3332,
+      "step": 423
+    },
+    {
+      "epoch": 0.1815845824411135,
+      "grad_norm": 2.8125969363697636,
+      "learning_rate": 9.415316898827923e-06,
+      "loss": 0.3017,
+      "step": 424
+    },
+    {
+      "epoch": 0.18201284796573874,
+      "grad_norm": 2.871668226012186,
+      "learning_rate": 9.412056897923163e-06,
+      "loss": 0.2695,
+      "step": 425
+    },
+    {
+      "epoch": 0.18244111349036402,
+      "grad_norm": 2.712741584619187,
+      "learning_rate": 9.408788401538145e-06,
+      "loss": 0.3003,
+      "step": 426
+    },
+    {
+      "epoch": 0.1828693790149893,
+      "grad_norm": 2.608522991577543,
+      "learning_rate": 9.405511415966403e-06,
+      "loss": 0.2672,
+      "step": 427
+    },
+    {
+      "epoch": 0.18329764453961456,
+      "grad_norm": 2.7666912423822265,
+      "learning_rate": 9.402225947517822e-06,
+      "loss": 0.2956,
+      "step": 428
+    },
+    {
+      "epoch": 0.18372591006423983,
+      "grad_norm": 3.1078195807771913,
+      "learning_rate": 9.398932002518619e-06,
+      "loss": 0.2678,
+      "step": 429
+    },
+    {
+      "epoch": 0.1841541755888651,
+      "grad_norm": 2.491069662996318,
+      "learning_rate": 9.395629587311334e-06,
+      "loss": 0.2338,
+      "step": 430
+    },
+    {
+      "epoch": 0.18458244111349037,
+      "grad_norm": 2.6419449138886444,
+      "learning_rate": 9.392318708254818e-06,
+      "loss": 0.2694,
+      "step": 431
+    },
+    {
+      "epoch": 0.18501070663811564,
+      "grad_norm": 2.870646620414496,
+      "learning_rate": 9.388999371724212e-06,
+      "loss": 0.2698,
+      "step": 432
+    },
+    {
+      "epoch": 0.1854389721627409,
+      "grad_norm": 2.874540177176191,
+      "learning_rate": 9.385671584110952e-06,
+      "loss": 0.273,
+      "step": 433
+    },
+    {
+      "epoch": 0.18586723768736615,
+      "grad_norm": 2.93441715492951,
+      "learning_rate": 9.382335351822738e-06,
+      "loss": 0.2578,
+      "step": 434
+    },
+    {
+      "epoch": 0.18629550321199143,
+      "grad_norm": 2.6926411654427502,
+      "learning_rate": 9.378990681283538e-06,
+      "loss": 0.2906,
+      "step": 435
+    },
+    {
+      "epoch": 0.1867237687366167,
+      "grad_norm": 2.907626158579969,
+      "learning_rate": 9.37563757893356e-06,
+      "loss": 0.2851,
+      "step": 436
+    },
+    {
+      "epoch": 0.18715203426124197,
+      "grad_norm": 2.7080622182675635,
+      "learning_rate": 9.372276051229257e-06,
+      "loss": 0.2696,
+      "step": 437
+    },
+    {
+      "epoch": 0.18758029978586724,
+      "grad_norm": 2.715006133169729,
+      "learning_rate": 9.368906104643295e-06,
+      "loss": 0.2756,
+      "step": 438
+    },
+    {
+      "epoch": 0.1880085653104925,
+      "grad_norm": 2.6769516042409216,
+      "learning_rate": 9.365527745664558e-06,
+      "loss": 0.2994,
+      "step": 439
+    },
+    {
+      "epoch": 0.18843683083511778,
+      "grad_norm": 2.7718719302187482,
+      "learning_rate": 9.362140980798127e-06,
+      "loss": 0.3008,
+      "step": 440
+    },
+    {
+      "epoch": 0.18886509635974305,
+      "grad_norm": 2.7522962198137146,
+      "learning_rate": 9.358745816565264e-06,
+      "loss": 0.2738,
+      "step": 441
+    },
+    {
+      "epoch": 0.18929336188436832,
+      "grad_norm": 2.5730941874861086,
+      "learning_rate": 9.355342259503409e-06,
+      "loss": 0.2596,
+      "step": 442
+    },
+    {
+      "epoch": 0.18972162740899357,
+      "grad_norm": 2.76012632240573,
+      "learning_rate": 9.351930316166162e-06,
+      "loss": 0.2661,
+      "step": 443
+    },
+    {
+      "epoch": 0.19014989293361884,
+      "grad_norm": 2.702863973589719,
+      "learning_rate": 9.34850999312327e-06,
+      "loss": 0.3119,
+      "step": 444
+    },
+    {
+      "epoch": 0.1905781584582441,
+      "grad_norm": 2.8359298412516956,
+      "learning_rate": 9.345081296960613e-06,
+      "loss": 0.3261,
+      "step": 445
+    },
+    {
+      "epoch": 0.19100642398286938,
+      "grad_norm": 2.722001348512975,
+      "learning_rate": 9.341644234280198e-06,
+      "loss": 0.2875,
+      "step": 446
+    },
+    {
+      "epoch": 0.19143468950749465,
+      "grad_norm": 2.55240186204932,
+      "learning_rate": 9.33819881170014e-06,
+      "loss": 0.2794,
+      "step": 447
+    },
+    {
+      "epoch": 0.19186295503211992,
+      "grad_norm": 2.706781809311813,
+      "learning_rate": 9.334745035854646e-06,
+      "loss": 0.3265,
+      "step": 448
+    },
+    {
+      "epoch": 0.1922912205567452,
+      "grad_norm": 2.842290589649882,
+      "learning_rate": 9.331282913394017e-06,
+      "loss": 0.3022,
+      "step": 449
+    },
+    {
+      "epoch": 0.19271948608137046,
+      "grad_norm": 2.7160832488410445,
+      "learning_rate": 9.32781245098462e-06,
+      "loss": 0.291,
+      "step": 450
+    },
+    {
+      "epoch": 0.1931477516059957,
+      "grad_norm": 2.7800977198341488,
+      "learning_rate": 9.324333655308881e-06,
+      "loss": 0.2749,
+      "step": 451
+    },
+    {
+      "epoch": 0.19357601713062098,
+      "grad_norm": 3.0639358366547644,
+      "learning_rate": 9.32084653306527e-06,
+      "loss": 0.3093,
+      "step": 452
+    },
+    {
+      "epoch": 0.19400428265524625,
+      "grad_norm": 2.758106120000733,
+      "learning_rate": 9.317351090968294e-06,
+      "loss": 0.2872,
+      "step": 453
+    },
+    {
+      "epoch": 0.19443254817987152,
+      "grad_norm": 2.4881662407164296,
+      "learning_rate": 9.313847335748479e-06,
+      "loss": 0.256,
+      "step": 454
+    },
+    {
+      "epoch": 0.1948608137044968,
+      "grad_norm": 2.65375148985648,
+      "learning_rate": 9.310335274152353e-06,
+      "loss": 0.2894,
+      "step": 455
+    },
+    {
+      "epoch": 0.19528907922912206,
+      "grad_norm": 2.891011423305799,
+      "learning_rate": 9.306814912942445e-06,
+      "loss": 0.3106,
+      "step": 456
+    },
+    {
+      "epoch": 0.19571734475374733,
+      "grad_norm": 2.667916565905268,
+      "learning_rate": 9.303286258897261e-06,
+      "loss": 0.3039,
+      "step": 457
+    },
+    {
+      "epoch": 0.1961456102783726,
+      "grad_norm": 2.9698080084204226,
+      "learning_rate": 9.29974931881128e-06,
+      "loss": 0.2787,
+      "step": 458
+    },
+    {
+      "epoch": 0.19657387580299787,
+      "grad_norm": 2.6231457879079785,
+      "learning_rate": 9.296204099494926e-06,
+      "loss": 0.2794,
+      "step": 459
+    },
+    {
+      "epoch": 0.19700214132762311,
+      "grad_norm": 2.8378624661954124,
+      "learning_rate": 9.292650607774576e-06,
+      "loss": 0.2832,
+      "step": 460
+    },
+    {
+      "epoch": 0.19743040685224839,
+      "grad_norm": 2.9837251823934574,
+      "learning_rate": 9.289088850492526e-06,
+      "loss": 0.2757,
+      "step": 461
+    },
+    {
+      "epoch": 0.19785867237687366,
+      "grad_norm": 2.8985624617324928,
+      "learning_rate": 9.285518834506994e-06,
+      "loss": 0.2974,
+      "step": 462
+    },
+    {
+      "epoch": 0.19828693790149893,
+      "grad_norm": 2.8368612681165066,
+      "learning_rate": 9.2819405666921e-06,
+      "loss": 0.3178,
+      "step": 463
+    },
+    {
+      "epoch": 0.1987152034261242,
+      "grad_norm": 2.732334501356763,
+      "learning_rate": 9.278354053937848e-06,
+      "loss": 0.3246,
+      "step": 464
+    },
+    {
+      "epoch": 0.19914346895074947,
+      "grad_norm": 2.4932142258920207,
+      "learning_rate": 9.274759303150123e-06,
+      "loss": 0.2689,
+      "step": 465
+    },
+    {
+      "epoch": 0.19957173447537474,
+      "grad_norm": 2.7261407686428862,
+      "learning_rate": 9.271156321250669e-06,
+      "loss": 0.2596,
+      "step": 466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7133101480245987,
+      "learning_rate": 9.26754511517708e-06,
+      "loss": 0.2827,
+      "step": 467
+    },
+    {
+      "epoch": 0.20042826552462525,
+      "grad_norm": 2.813205957360494,
+      "learning_rate": 9.26392569188279e-06,
+      "loss": 0.2973,
+      "step": 468
+    },
+    {
+      "epoch": 0.20085653104925053,
+      "grad_norm": 2.7237082688070893,
+      "learning_rate": 9.260298058337046e-06,
+      "loss": 0.3169,
+      "step": 469
+    },
+    {
+      "epoch": 0.2012847965738758,
+      "grad_norm": 2.6679298718976234,
+      "learning_rate": 9.256662221524913e-06,
+      "loss": 0.2658,
+      "step": 470
+    },
+    {
+      "epoch": 0.20171306209850107,
+      "grad_norm": 3.404621768018313,
+      "learning_rate": 9.25301818844725e-06,
+      "loss": 0.3285,
+      "step": 471
+    },
+    {
+      "epoch": 0.20214132762312634,
+      "grad_norm": 2.6199490573721045,
+      "learning_rate": 9.249365966120692e-06,
+      "loss": 0.2499,
+      "step": 472
+    },
+    {
+      "epoch": 0.2025695931477516,
+      "grad_norm": 2.9966402258073797,
+      "learning_rate": 9.245705561577647e-06,
+      "loss": 0.2814,
+      "step": 473
+    },
+    {
+      "epoch": 0.20299785867237688,
+      "grad_norm": 2.807056592715456,
+      "learning_rate": 9.24203698186628e-06,
+      "loss": 0.2864,
+      "step": 474
+    },
+    {
+      "epoch": 0.20342612419700215,
+      "grad_norm": 2.7082932370826045,
+      "learning_rate": 9.238360234050496e-06,
+      "loss": 0.3248,
+      "step": 475
+    },
+    {
+      "epoch": 0.20385438972162742,
+      "grad_norm": 2.726914694256749,
+      "learning_rate": 9.234675325209923e-06,
+      "loss": 0.2749,
+      "step": 476
+    },
+    {
+      "epoch": 0.20428265524625266,
+      "grad_norm": 2.9105177927160035,
+      "learning_rate": 9.230982262439909e-06,
+      "loss": 0.292,
+      "step": 477
+    },
+    {
+      "epoch": 0.20471092077087794,
+      "grad_norm": 2.6660428672251775,
+      "learning_rate": 9.2272810528515e-06,
+      "loss": 0.256,
+      "step": 478
+    },
+    {
+      "epoch": 0.2051391862955032,
+      "grad_norm": 2.4234955324416436,
+      "learning_rate": 9.22357170357143e-06,
+      "loss": 0.264,
+      "step": 479
+    },
+    {
+      "epoch": 0.20556745182012848,
+      "grad_norm": 2.615169467158398,
+      "learning_rate": 9.219854221742106e-06,
+      "loss": 0.3332,
+      "step": 480
+    },
+    {
+      "epoch": 0.20599571734475375,
+      "grad_norm": 2.77798421215127,
+      "learning_rate": 9.216128614521593e-06,
+      "loss": 0.2928,
+      "step": 481
+    },
+    {
+      "epoch": 0.20642398286937902,
+      "grad_norm": 2.87300808185799,
+      "learning_rate": 9.212394889083602e-06,
+      "loss": 0.293,
+      "step": 482
+    },
+    {
+      "epoch": 0.2068522483940043,
+      "grad_norm": 2.655305382164315,
+      "learning_rate": 9.208653052617475e-06,
+      "loss": 0.2407,
+      "step": 483
+    },
+    {
+      "epoch": 0.20728051391862956,
+      "grad_norm": 2.8663212374181115,
+      "learning_rate": 9.204903112328177e-06,
+      "loss": 0.3509,
+      "step": 484
+    },
+    {
+      "epoch": 0.20770877944325483,
+      "grad_norm": 2.617725879271642,
+      "learning_rate": 9.201145075436269e-06,
+      "loss": 0.2525,
+      "step": 485
+    },
+    {
+      "epoch": 0.20813704496788007,
+      "grad_norm": 2.785239825315732,
+      "learning_rate": 9.197378949177908e-06,
+      "loss": 0.2673,
+      "step": 486
+    },
+    {
+      "epoch": 0.20856531049250535,
+      "grad_norm": 2.8687553597089535,
+      "learning_rate": 9.193604740804825e-06,
+      "loss": 0.2636,
+      "step": 487
+    },
+    {
+      "epoch": 0.20899357601713062,
+      "grad_norm": 2.8344396376911067,
+      "learning_rate": 9.189822457584311e-06,
+      "loss": 0.2556,
+      "step": 488
+    },
+    {
+      "epoch": 0.2094218415417559,
+      "grad_norm": 2.586551019365263,
+      "learning_rate": 9.186032106799214e-06,
+      "loss": 0.292,
+      "step": 489
+    },
+    {
+      "epoch": 0.20985010706638116,
+      "grad_norm": 3.0259303641461637,
+      "learning_rate": 9.182233695747905e-06,
+      "loss": 0.282,
+      "step": 490
+    },
+    {
+      "epoch": 0.21027837259100643,
+      "grad_norm": 2.8391174465456617,
+      "learning_rate": 9.17842723174428e-06,
+      "loss": 0.2817,
+      "step": 491
+    },
+    {
+      "epoch": 0.2107066381156317,
+      "grad_norm": 2.7308244721849553,
+      "learning_rate": 9.174612722117744e-06,
+      "loss": 0.2606,
+      "step": 492
+    },
+    {
+      "epoch": 0.21113490364025697,
+      "grad_norm": 2.6838280594419754,
+      "learning_rate": 9.17079017421319e-06,
+      "loss": 0.2829,
+      "step": 493
+    },
+    {
+      "epoch": 0.21156316916488221,
+      "grad_norm": 3.0057860512993373,
+      "learning_rate": 9.16695959539099e-06,
+      "loss": 0.3245,
+      "step": 494
+    },
+    {
+      "epoch": 0.21199143468950749,
+      "grad_norm": 2.830474213273471,
+      "learning_rate": 9.163120993026979e-06,
+      "loss": 0.2857,
+      "step": 495
+    },
+    {
+      "epoch": 0.21241970021413276,
+      "grad_norm": 2.7520297941840375,
+      "learning_rate": 9.159274374512444e-06,
+      "loss": 0.2848,
+      "step": 496
+    },
+    {
+      "epoch": 0.21284796573875803,
+      "grad_norm": 2.7072726398893963,
+      "learning_rate": 9.155419747254103e-06,
+      "loss": 0.2527,
+      "step": 497
+    },
+    {
+      "epoch": 0.2132762312633833,
+      "grad_norm": 2.7443696345389186,
+      "learning_rate": 9.1515571186741e-06,
+      "loss": 0.2713,
+      "step": 498
+    },
+    {
+      "epoch": 0.21370449678800857,
+      "grad_norm": 2.6537934854510876,
+      "learning_rate": 9.14768649620998e-06,
+      "loss": 0.2755,
+      "step": 499
+    },
+    {
+      "epoch": 0.21413276231263384,
+      "grad_norm": 2.5833945302834853,
+      "learning_rate": 9.143807887314686e-06,
+      "loss": 0.2756,
+      "step": 500
+    },
+    {
+      "epoch": 0.2145610278372591,
+      "grad_norm": 2.508810572824063,
+      "learning_rate": 9.139921299456536e-06,
+      "loss": 0.2217,
+      "step": 501
+    },
+    {
+      "epoch": 0.21498929336188438,
+      "grad_norm": 2.9089742527913285,
+      "learning_rate": 9.13602674011921e-06,
+      "loss": 0.2941,
+      "step": 502
+    },
+    {
+      "epoch": 0.21541755888650962,
+      "grad_norm": 2.7871474201045743,
+      "learning_rate": 9.13212421680174e-06,
+      "loss": 0.2471,
+      "step": 503
+    },
+    {
+      "epoch": 0.2158458244111349,
+      "grad_norm": 2.7836896917286027,
+      "learning_rate": 9.128213737018493e-06,
+      "loss": 0.2624,
+      "step": 504
+    },
+    {
+      "epoch": 0.21627408993576017,
+      "grad_norm": 2.9392457547766004,
+      "learning_rate": 9.124295308299152e-06,
+      "loss": 0.291,
+      "step": 505
+    },
+    {
+      "epoch": 0.21670235546038544,
+      "grad_norm": 3.1083243736645354,
+      "learning_rate": 9.120368938188713e-06,
+      "loss": 0.2977,
+      "step": 506
+    },
+    {
+      "epoch": 0.2171306209850107,
+      "grad_norm": 2.9813868858293957,
+      "learning_rate": 9.116434634247457e-06,
+      "loss": 0.2724,
+      "step": 507
+    },
+    {
+      "epoch": 0.21755888650963598,
+      "grad_norm": 3.097945924789147,
+      "learning_rate": 9.112492404050944e-06,
+      "loss": 0.3101,
+      "step": 508
+    },
+    {
+      "epoch": 0.21798715203426125,
+      "grad_norm": 2.6912774132555555,
+      "learning_rate": 9.108542255189998e-06,
+      "loss": 0.263,
+      "step": 509
+    },
+    {
+      "epoch": 0.21841541755888652,
+      "grad_norm": 2.657780624839184,
+      "learning_rate": 9.104584195270685e-06,
+      "loss": 0.2434,
+      "step": 510
+    },
+    {
+      "epoch": 0.21884368308351176,
+      "grad_norm": 2.6989421901773047,
+      "learning_rate": 9.10061823191431e-06,
+      "loss": 0.2743,
+      "step": 511
+    },
+    {
+      "epoch": 0.21927194860813704,
+      "grad_norm": 2.369942512592271,
+      "learning_rate": 9.096644372757393e-06,
+      "loss": 0.2789,
+      "step": 512
+    },
+    {
+      "epoch": 0.2197002141327623,
+      "grad_norm": 2.6161843406619765,
+      "learning_rate": 9.092662625451659e-06,
+      "loss": 0.2138,
+      "step": 513
+    },
+    {
+      "epoch": 0.22012847965738758,
+      "grad_norm": 2.493041086942503,
+      "learning_rate": 9.08867299766402e-06,
+      "loss": 0.227,
+      "step": 514
+    },
+    {
+      "epoch": 0.22055674518201285,
+      "grad_norm": 2.398253943020624,
+      "learning_rate": 9.084675497076562e-06,
+      "loss": 0.264,
+      "step": 515
+    },
+    {
+      "epoch": 0.22098501070663812,
+      "grad_norm": 2.8282682350602593,
+      "learning_rate": 9.08067013138653e-06,
+      "loss": 0.2619,
+      "step": 516
+    },
+    {
+      "epoch": 0.2214132762312634,
+      "grad_norm": 2.686938794926156,
+      "learning_rate": 9.07665690830632e-06,
+      "loss": 0.2123,
+      "step": 517
+    },
+    {
+      "epoch": 0.22184154175588866,
+      "grad_norm": 2.9810019686493283,
+      "learning_rate": 9.072635835563446e-06,
+      "loss": 0.2221,
+      "step": 518
+    },
+    {
+      "epoch": 0.22226980728051393,
+      "grad_norm": 3.660695931171815,
+      "learning_rate": 9.068606920900544e-06,
+      "loss": 0.3404,
+      "step": 519
+    },
+    {
+      "epoch": 0.22269807280513917,
+      "grad_norm": 3.042182759827938,
+      "learning_rate": 9.064570172075349e-06,
+      "loss": 0.2771,
+      "step": 520
+    },
+    {
+      "epoch": 0.22312633832976445,
+      "grad_norm": 2.621419678483974,
+      "learning_rate": 9.060525596860681e-06,
+      "loss": 0.2197,
+      "step": 521
+    },
+    {
+      "epoch": 0.22355460385438972,
+      "grad_norm": 2.942350416287861,
+      "learning_rate": 9.056473203044428e-06,
+      "loss": 0.2866,
+      "step": 522
+    },
+    {
+      "epoch": 0.223982869379015,
+      "grad_norm": 2.773803532229209,
+      "learning_rate": 9.05241299842953e-06,
+      "loss": 0.2627,
+      "step": 523
+    },
+    {
+      "epoch": 0.22441113490364026,
+      "grad_norm": 2.7514379765673356,
+      "learning_rate": 9.048344990833978e-06,
+      "loss": 0.2895,
+      "step": 524
+    },
+    {
+      "epoch": 0.22483940042826553,
+      "grad_norm": 2.8043725085329583,
+      "learning_rate": 9.044269188090774e-06,
+      "loss": 0.288,
+      "step": 525
+    },
+    {
+      "epoch": 0.2252676659528908,
+      "grad_norm": 2.7276345504060875,
+      "learning_rate": 9.040185598047939e-06,
+      "loss": 0.2614,
+      "step": 526
+    },
+    {
+      "epoch": 0.22569593147751607,
+      "grad_norm": 2.8349540267755478,
+      "learning_rate": 9.036094228568485e-06,
+      "loss": 0.273,
+      "step": 527
+    },
+    {
+      "epoch": 0.22612419700214131,
+      "grad_norm": 2.921354775581877,
+      "learning_rate": 9.031995087530403e-06,
+      "loss": 0.2909,
+      "step": 528
+    },
+    {
+      "epoch": 0.22655246252676658,
+      "grad_norm": 2.746487448732942,
+      "learning_rate": 9.027888182826652e-06,
+      "loss": 0.2753,
+      "step": 529
+    },
+    {
+      "epoch": 0.22698072805139186,
+      "grad_norm": 2.5841852196510113,
+      "learning_rate": 9.023773522365133e-06,
+      "loss": 0.261,
+      "step": 530
+    },
+    {
+      "epoch": 0.22740899357601713,
+      "grad_norm": 2.9583828966731547,
+      "learning_rate": 9.019651114068689e-06,
+      "loss": 0.2827,
+      "step": 531
+    },
+    {
+      "epoch": 0.2278372591006424,
+      "grad_norm": 2.8200390113879688,
+      "learning_rate": 9.015520965875073e-06,
+      "loss": 0.2425,
+      "step": 532
+    },
+    {
+      "epoch": 0.22826552462526767,
+      "grad_norm": 2.9305060386116337,
+      "learning_rate": 9.011383085736951e-06,
+      "loss": 0.2704,
+      "step": 533
+    },
+    {
+      "epoch": 0.22869379014989294,
+      "grad_norm": 2.768148619691505,
+      "learning_rate": 9.007237481621869e-06,
+      "loss": 0.2998,
+      "step": 534
+    },
+    {
+      "epoch": 0.2291220556745182,
+      "grad_norm": 2.703874033408789,
+      "learning_rate": 9.003084161512249e-06,
+      "loss": 0.3059,
+      "step": 535
+    },
+    {
+      "epoch": 0.22955032119914348,
+      "grad_norm": 2.7254003838740912,
+      "learning_rate": 8.99892313340537e-06,
+      "loss": 0.2621,
+      "step": 536
+    },
+    {
+      "epoch": 0.22997858672376872,
+      "grad_norm": 2.7164757971250975,
+      "learning_rate": 8.994754405313353e-06,
+      "loss": 0.2825,
+      "step": 537
+    },
+    {
+      "epoch": 0.230406852248394,
+      "grad_norm": 2.5708047327060073,
+      "learning_rate": 8.990577985263144e-06,
+      "loss": 0.2815,
+      "step": 538
+    },
+    {
+      "epoch": 0.23083511777301927,
+      "grad_norm": 2.775849385531009,
+      "learning_rate": 8.986393881296502e-06,
+      "loss": 0.2446,
+      "step": 539
+    },
+    {
+      "epoch": 0.23126338329764454,
+      "grad_norm": 2.556984883612775,
+      "learning_rate": 8.98220210146998e-06,
+      "loss": 0.2961,
+      "step": 540
+    },
+    {
+      "epoch": 0.2316916488222698,
+      "grad_norm": 2.7532968904105397,
+      "learning_rate": 8.978002653854913e-06,
+      "loss": 0.2632,
+      "step": 541
+    },
+    {
+      "epoch": 0.23211991434689508,
+      "grad_norm": 2.5220759904608903,
+      "learning_rate": 8.973795546537396e-06,
+      "loss": 0.2218,
+      "step": 542
+    },
+    {
+      "epoch": 0.23254817987152035,
+      "grad_norm": 3.063622977631629,
+      "learning_rate": 8.96958078761828e-06,
+      "loss": 0.2806,
+      "step": 543
+    },
+    {
+      "epoch": 0.23297644539614562,
+      "grad_norm": 3.0743762447891876,
+      "learning_rate": 8.96535838521314e-06,
+      "loss": 0.299,
+      "step": 544
+    },
+    {
+      "epoch": 0.2334047109207709,
+      "grad_norm": 2.7755576262001727,
+      "learning_rate": 8.961128347452278e-06,
+      "loss": 0.2611,
+      "step": 545
+    },
+    {
+      "epoch": 0.23383297644539613,
+      "grad_norm": 2.5137418776530547,
+      "learning_rate": 8.956890682480692e-06,
+      "loss": 0.2563,
+      "step": 546
+    },
+    {
+      "epoch": 0.2342612419700214,
+      "grad_norm": 2.4805398553498175,
+      "learning_rate": 8.952645398458067e-06,
+      "loss": 0.2706,
+      "step": 547
+    },
+    {
+      "epoch": 0.23468950749464668,
+      "grad_norm": 2.8302167346803118,
+      "learning_rate": 8.948392503558763e-06,
+      "loss": 0.2769,
+      "step": 548
+    },
+    {
+      "epoch": 0.23511777301927195,
+      "grad_norm": 2.716525926661055,
+      "learning_rate": 8.944132005971788e-06,
+      "loss": 0.278,
+      "step": 549
+    },
+    {
+      "epoch": 0.23554603854389722,
+      "grad_norm": 2.832578270948385,
+      "learning_rate": 8.939863913900794e-06,
+      "loss": 0.2597,
+      "step": 550
+    },
+    {
+      "epoch": 0.2359743040685225,
+      "grad_norm": 2.8437517932334746,
+      "learning_rate": 8.935588235564054e-06,
+      "loss": 0.3014,
+      "step": 551
+    },
+    {
+      "epoch": 0.23640256959314776,
+      "grad_norm": 2.572357752206744,
+      "learning_rate": 8.931304979194452e-06,
+      "loss": 0.2138,
+      "step": 552
+    },
+    {
+      "epoch": 0.23683083511777303,
+      "grad_norm": 2.876648230962267,
+      "learning_rate": 8.927014153039458e-06,
+      "loss": 0.2647,
+      "step": 553
+    },
+    {
+      "epoch": 0.23725910064239827,
+      "grad_norm": 2.7587933168790824,
+      "learning_rate": 8.922715765361124e-06,
+      "loss": 0.2754,
+      "step": 554
+    },
+    {
+      "epoch": 0.23768736616702354,
+      "grad_norm": 2.754845940079021,
+      "learning_rate": 8.918409824436053e-06,
+      "loss": 0.2691,
+      "step": 555
+    },
+    {
+      "epoch": 0.23811563169164882,
+      "grad_norm": 2.6992976473310195,
+      "learning_rate": 8.914096338555402e-06,
+      "loss": 0.2979,
+      "step": 556
+    },
+    {
+      "epoch": 0.2385438972162741,
+      "grad_norm": 2.5372209744830947,
+      "learning_rate": 8.90977531602485e-06,
+      "loss": 0.2486,
+      "step": 557
+    },
+    {
+      "epoch": 0.23897216274089936,
+      "grad_norm": 2.466514725378207,
+      "learning_rate": 8.90544676516459e-06,
+      "loss": 0.2512,
+      "step": 558
+    },
+    {
+      "epoch": 0.23940042826552463,
+      "grad_norm": 2.7010148539119507,
+      "learning_rate": 8.90111069430931e-06,
+      "loss": 0.2425,
+      "step": 559
+    },
+    {
+      "epoch": 0.2398286937901499,
+      "grad_norm": 2.6887596119213346,
+      "learning_rate": 8.896767111808177e-06,
+      "loss": 0.255,
+      "step": 560
+    },
+    {
+      "epoch": 0.24025695931477517,
+      "grad_norm": 2.9975852142222723,
+      "learning_rate": 8.892416026024823e-06,
+      "loss": 0.2472,
+      "step": 561
+    },
+    {
+      "epoch": 0.24068522483940044,
+      "grad_norm": 2.8742360437603987,
+      "learning_rate": 8.888057445337328e-06,
+      "loss": 0.2519,
+      "step": 562
+    },
+    {
+      "epoch": 0.24111349036402568,
+      "grad_norm": 3.056945966573546,
+      "learning_rate": 8.883691378138201e-06,
+      "loss": 0.2766,
+      "step": 563
+    },
+    {
+      "epoch": 0.24154175588865096,
+      "grad_norm": 2.9224187087350555,
+      "learning_rate": 8.879317832834372e-06,
+      "loss": 0.2467,
+      "step": 564
+    },
+    {
+      "epoch": 0.24197002141327623,
+      "grad_norm": 2.4723684593081314,
+      "learning_rate": 8.874936817847165e-06,
+      "loss": 0.239,
+      "step": 565
+    },
+    {
+      "epoch": 0.2423982869379015,
+      "grad_norm": 2.874835582049693,
+      "learning_rate": 8.870548341612288e-06,
+      "loss": 0.3201,
+      "step": 566
+    },
+    {
+      "epoch": 0.24282655246252677,
+      "grad_norm": 2.551611678570096,
+      "learning_rate": 8.866152412579818e-06,
+      "loss": 0.2596,
+      "step": 567
+    },
+    {
+      "epoch": 0.24325481798715204,
+      "grad_norm": 2.2974389919986247,
+      "learning_rate": 8.861749039214177e-06,
+      "loss": 0.2549,
+      "step": 568
+    },
+    {
+      "epoch": 0.2436830835117773,
+      "grad_norm": 2.457688044445882,
+      "learning_rate": 8.85733822999413e-06,
+      "loss": 0.2301,
+      "step": 569
+    },
+    {
+      "epoch": 0.24411134903640258,
+      "grad_norm": 2.6746153899168403,
+      "learning_rate": 8.852919993412751e-06,
+      "loss": 0.3203,
+      "step": 570
+    },
+    {
+      "epoch": 0.24453961456102782,
+      "grad_norm": 2.3917817747723613,
+      "learning_rate": 8.848494337977422e-06,
+      "loss": 0.1947,
+      "step": 571
+    },
+    {
+      "epoch": 0.2449678800856531,
+      "grad_norm": 2.5326094953081606,
+      "learning_rate": 8.844061272209807e-06,
+      "loss": 0.2339,
+      "step": 572
+    },
+    {
+      "epoch": 0.24539614561027837,
+      "grad_norm": 2.845925097700425,
+      "learning_rate": 8.839620804645837e-06,
+      "loss": 0.2152,
+      "step": 573
+    },
+    {
+      "epoch": 0.24582441113490364,
+      "grad_norm": 2.939647193398751,
+      "learning_rate": 8.835172943835699e-06,
+      "loss": 0.2575,
+      "step": 574
+    },
+    {
+      "epoch": 0.2462526766595289,
+      "grad_norm": 3.650596380535361,
+      "learning_rate": 8.830717698343813e-06,
+      "loss": 0.273,
+      "step": 575
+    },
+    {
+      "epoch": 0.24668094218415418,
+      "grad_norm": 2.717749688871503,
+      "learning_rate": 8.826255076748823e-06,
+      "loss": 0.2102,
+      "step": 576
+    },
+    {
+      "epoch": 0.24710920770877945,
+      "grad_norm": 2.714939559406906,
+      "learning_rate": 8.821785087643566e-06,
+      "loss": 0.2791,
+      "step": 577
+    },
+    {
+      "epoch": 0.24753747323340472,
+      "grad_norm": 2.8592978359878773,
+      "learning_rate": 8.817307739635079e-06,
+      "loss": 0.2973,
+      "step": 578
+    },
+    {
+      "epoch": 0.24796573875803,
+      "grad_norm": 2.730960030534237,
+      "learning_rate": 8.812823041344557e-06,
+      "loss": 0.2845,
+      "step": 579
+    },
+    {
+      "epoch": 0.24839400428265523,
+      "grad_norm": 2.831720922754365,
+      "learning_rate": 8.808331001407352e-06,
+      "loss": 0.2947,
+      "step": 580
+    },
+    {
+      "epoch": 0.2488222698072805,
+      "grad_norm": 2.5300429301007443,
+      "learning_rate": 8.803831628472955e-06,
+      "loss": 0.2813,
+      "step": 581
+    },
+    {
+      "epoch": 0.24925053533190578,
+      "grad_norm": 2.6216424316055766,
+      "learning_rate": 8.799324931204972e-06,
+      "loss": 0.2903,
+      "step": 582
+    },
+    {
+      "epoch": 0.24967880085653105,
+      "grad_norm": 2.553355913477896,
+      "learning_rate": 8.794810918281117e-06,
+      "loss": 0.2323,
+      "step": 583
+    },
+    {
+      "epoch": 0.2501070663811563,
+      "grad_norm": 2.5411084274598035,
+      "learning_rate": 8.790289598393186e-06,
+      "loss": 0.2715,
+      "step": 584
+    },
+    {
+      "epoch": 0.2505353319057816,
+      "grad_norm": 2.7187278994009647,
+      "learning_rate": 8.785760980247051e-06,
+      "loss": 0.2702,
+      "step": 585
+    },
+    {
+      "epoch": 0.25096359743040686,
+      "grad_norm": 2.685660110859678,
+      "learning_rate": 8.781225072562627e-06,
+      "loss": 0.2135,
+      "step": 586
+    },
+    {
+      "epoch": 0.25139186295503213,
+      "grad_norm": 3.1042378895412637,
+      "learning_rate": 8.776681884073872e-06,
+      "loss": 0.2916,
+      "step": 587
+    },
+    {
+      "epoch": 0.2518201284796574,
+      "grad_norm": 2.9562441119873344,
+      "learning_rate": 8.772131423528766e-06,
+      "loss": 0.2766,
+      "step": 588
+    },
+    {
+      "epoch": 0.25224839400428267,
+      "grad_norm": 2.6964095339195944,
+      "learning_rate": 8.767573699689284e-06,
+      "loss": 0.2423,
+      "step": 589
+    },
+    {
+      "epoch": 0.25267665952890794,
+      "grad_norm": 2.8304771385600405,
+      "learning_rate": 8.76300872133139e-06,
+      "loss": 0.2822,
+      "step": 590
+    },
+    {
+      "epoch": 0.2531049250535332,
+      "grad_norm": 2.822787579549225,
+      "learning_rate": 8.758436497245019e-06,
+      "loss": 0.2467,
+      "step": 591
+    },
+    {
+      "epoch": 0.2535331905781585,
+      "grad_norm": 2.8118302650801055,
+      "learning_rate": 8.753857036234055e-06,
+      "loss": 0.1982,
+      "step": 592
+    },
+    {
+      "epoch": 0.2539614561027837,
+      "grad_norm": 3.024144367043807,
+      "learning_rate": 8.749270347116315e-06,
+      "loss": 0.2719,
+      "step": 593
+    },
+    {
+      "epoch": 0.25438972162740897,
+      "grad_norm": 2.836820486473772,
+      "learning_rate": 8.744676438723538e-06,
+      "loss": 0.2529,
+      "step": 594
+    },
+    {
+      "epoch": 0.25481798715203424,
+      "grad_norm": 2.604855197880593,
+      "learning_rate": 8.740075319901362e-06,
+      "loss": 0.3038,
+      "step": 595
+    },
+    {
+      "epoch": 0.2552462526766595,
+      "grad_norm": 2.528126449219841,
+      "learning_rate": 8.73546699950931e-06,
+      "loss": 0.2661,
+      "step": 596
+    },
+    {
+      "epoch": 0.2556745182012848,
+      "grad_norm": 2.5231806322741552,
+      "learning_rate": 8.730851486420768e-06,
+      "loss": 0.2623,
+      "step": 597
+    },
+    {
+      "epoch": 0.25610278372591005,
+      "grad_norm": 2.4985469916758554,
+      "learning_rate": 8.726228789522974e-06,
+      "loss": 0.2454,
+      "step": 598
+    },
+    {
+      "epoch": 0.2565310492505353,
+      "grad_norm": 2.557731837817599,
+      "learning_rate": 8.721598917717e-06,
+      "loss": 0.2516,
+      "step": 599
+    },
+    {
+      "epoch": 0.2569593147751606,
+      "grad_norm": 2.778021361752478,
+      "learning_rate": 8.716961879917734e-06,
+      "loss": 0.2686,
+      "step": 600
+    },
+    {
+      "epoch": 0.25738758029978587,
+      "grad_norm": 2.8278899679556195,
+      "learning_rate": 8.712317685053857e-06,
+      "loss": 0.2012,
+      "step": 601
+    },
+    {
+      "epoch": 0.25781584582441114,
+      "grad_norm": 2.8257899336252392,
+      "learning_rate": 8.707666342067835e-06,
+      "loss": 0.2558,
+      "step": 602
+    },
+    {
+      "epoch": 0.2582441113490364,
+      "grad_norm": 2.9227335096789724,
+      "learning_rate": 8.703007859915897e-06,
+      "loss": 0.2622,
+      "step": 603
+    },
+    {
+      "epoch": 0.2586723768736617,
+      "grad_norm": 3.004332302678826,
+      "learning_rate": 8.698342247568021e-06,
+      "loss": 0.2941,
+      "step": 604
+    },
+    {
+      "epoch": 0.25910064239828695,
+      "grad_norm": 3.1063148937053016,
+      "learning_rate": 8.69366951400791e-06,
+      "loss": 0.2781,
+      "step": 605
+    },
+    {
+      "epoch": 0.2595289079229122,
+      "grad_norm": 2.7099631132505633,
+      "learning_rate": 8.688989668232982e-06,
+      "loss": 0.2441,
+      "step": 606
+    },
+    {
+      "epoch": 0.2599571734475375,
+      "grad_norm": 2.38249676533993,
+      "learning_rate": 8.684302719254346e-06,
+      "loss": 0.1791,
+      "step": 607
+    },
+    {
+      "epoch": 0.26038543897216276,
+      "grad_norm": 2.543865613012074,
+      "learning_rate": 8.679608676096793e-06,
+      "loss": 0.2077,
+      "step": 608
+    },
+    {
+      "epoch": 0.26081370449678803,
+      "grad_norm": 2.870770327916057,
+      "learning_rate": 8.674907547798775e-06,
+      "loss": 0.2765,
+      "step": 609
+    },
+    {
+      "epoch": 0.26124197002141325,
+      "grad_norm": 2.732220233542284,
+      "learning_rate": 8.670199343412377e-06,
+      "loss": 0.2454,
+      "step": 610
+    },
+    {
+      "epoch": 0.2616702355460385,
+      "grad_norm": 2.445416874546525,
+      "learning_rate": 8.665484072003321e-06,
+      "loss": 0.2074,
+      "step": 611
+    },
+    {
+      "epoch": 0.2620985010706638,
+      "grad_norm": 2.721048190617264,
+      "learning_rate": 8.660761742650928e-06,
+      "loss": 0.2381,
+      "step": 612
+    },
+    {
+      "epoch": 0.26252676659528906,
+      "grad_norm": 2.8589556181159073,
+      "learning_rate": 8.656032364448113e-06,
+      "loss": 0.2851,
+      "step": 613
+    },
+    {
+      "epoch": 0.26295503211991433,
+      "grad_norm": 2.646664710494815,
+      "learning_rate": 8.651295946501366e-06,
+      "loss": 0.2694,
+      "step": 614
+    },
+    {
+      "epoch": 0.2633832976445396,
+      "grad_norm": 2.926561427528524,
+      "learning_rate": 8.646552497930725e-06,
+      "loss": 0.2835,
+      "step": 615
+    },
+    {
+      "epoch": 0.2638115631691649,
+      "grad_norm": 2.6117296474052436,
+      "learning_rate": 8.641802027869774e-06,
+      "loss": 0.2631,
+      "step": 616
+    },
+    {
+      "epoch": 0.26423982869379015,
+      "grad_norm": 2.5353432357422343,
+      "learning_rate": 8.637044545465609e-06,
+      "loss": 0.2925,
+      "step": 617
+    },
+    {
+      "epoch": 0.2646680942184154,
+      "grad_norm": 2.6022554407363763,
+      "learning_rate": 8.632280059878835e-06,
+      "loss": 0.2616,
+      "step": 618
+    },
+    {
+      "epoch": 0.2650963597430407,
+      "grad_norm": 2.3538138944898104,
+      "learning_rate": 8.627508580283536e-06,
+      "loss": 0.2787,
+      "step": 619
+    },
+    {
+      "epoch": 0.26552462526766596,
+      "grad_norm": 2.69600784967288,
+      "learning_rate": 8.622730115867268e-06,
+      "loss": 0.278,
+      "step": 620
+    },
+    {
+      "epoch": 0.26595289079229123,
+      "grad_norm": 2.8910291661940093,
+      "learning_rate": 8.617944675831034e-06,
+      "loss": 0.2902,
+      "step": 621
+    },
+    {
+      "epoch": 0.2663811563169165,
+      "grad_norm": 2.8932934364271077,
+      "learning_rate": 8.613152269389269e-06,
+      "loss": 0.291,
+      "step": 622
+    },
+    {
+      "epoch": 0.26680942184154177,
+      "grad_norm": 2.6409204077453645,
+      "learning_rate": 8.608352905769821e-06,
+      "loss": 0.2406,
+      "step": 623
+    },
+    {
+      "epoch": 0.26723768736616704,
+      "grad_norm": 2.7086959810803557,
+      "learning_rate": 8.603546594213935e-06,
+      "loss": 0.2572,
+      "step": 624
+    },
+    {
+      "epoch": 0.2676659528907923,
+      "grad_norm": 2.7305330992361445,
+      "learning_rate": 8.598733343976236e-06,
+      "loss": 0.2605,
+      "step": 625
+    },
+    {
+      "epoch": 0.2680942184154176,
+      "grad_norm": 2.9569887697999904,
+      "learning_rate": 8.593913164324708e-06,
+      "loss": 0.2978,
+      "step": 626
+    },
+    {
+      "epoch": 0.2685224839400428,
+      "grad_norm": 2.646670126624379,
+      "learning_rate": 8.589086064540675e-06,
+      "loss": 0.2339,
+      "step": 627
+    },
+    {
+      "epoch": 0.26895074946466807,
+      "grad_norm": 2.8792642161625603,
+      "learning_rate": 8.58425205391879e-06,
+      "loss": 0.2703,
+      "step": 628
+    },
+    {
+      "epoch": 0.26937901498929334,
+      "grad_norm": 2.8425607674029956,
+      "learning_rate": 8.579411141767013e-06,
+      "loss": 0.2518,
+      "step": 629
+    },
+    {
+      "epoch": 0.2698072805139186,
+      "grad_norm": 2.8681762499112033,
+      "learning_rate": 8.574563337406587e-06,
+      "loss": 0.2677,
+      "step": 630
+    },
+    {
+      "epoch": 0.2702355460385439,
+      "grad_norm": 2.4763214996668266,
+      "learning_rate": 8.569708650172037e-06,
+      "loss": 0.2373,
+      "step": 631
+    },
+    {
+      "epoch": 0.27066381156316915,
+      "grad_norm": 2.6063276048909976,
+      "learning_rate": 8.564847089411128e-06,
+      "loss": 0.26,
+      "step": 632
+    },
+    {
+      "epoch": 0.2710920770877944,
+      "grad_norm": 2.636022388967572,
+      "learning_rate": 8.55997866448487e-06,
+      "loss": 0.2231,
+      "step": 633
+    },
+    {
+      "epoch": 0.2715203426124197,
+      "grad_norm": 2.7517202504756995,
+      "learning_rate": 8.555103384767485e-06,
+      "loss": 0.2425,
+      "step": 634
+    },
+    {
+      "epoch": 0.27194860813704497,
+      "grad_norm": 2.846118330018806,
+      "learning_rate": 8.550221259646395e-06,
+      "loss": 0.2379,
+      "step": 635
+    },
+    {
+      "epoch": 0.27237687366167024,
+      "grad_norm": 2.7753423876247703,
+      "learning_rate": 8.545332298522207e-06,
+      "loss": 0.2771,
+      "step": 636
+    },
+    {
+      "epoch": 0.2728051391862955,
+      "grad_norm": 2.857373452787854,
+      "learning_rate": 8.540436510808684e-06,
+      "loss": 0.3414,
+      "step": 637
+    },
+    {
+      "epoch": 0.2732334047109208,
+      "grad_norm": 2.932314281465467,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.2935,
+      "step": 638
+    },
+    {
+      "epoch": 0.27366167023554605,
+      "grad_norm": 2.633888895218832,
+      "learning_rate": 8.530624493334408e-06,
+      "loss": 0.2675,
+      "step": 639
+    },
+    {
+      "epoch": 0.2740899357601713,
+      "grad_norm": 2.704996107958476,
+      "learning_rate": 8.525708282466839e-06,
+      "loss": 0.2784,
+      "step": 640
+    },
+    {
+      "epoch": 0.2745182012847966,
+      "grad_norm": 2.753428699530911,
+      "learning_rate": 8.520785282796265e-06,
+      "loss": 0.2692,
+      "step": 641
+    },
+    {
+      "epoch": 0.27494646680942186,
+      "grad_norm": 2.8076912882391105,
+      "learning_rate": 8.515855503801996e-06,
+      "loss": 0.227,
+      "step": 642
+    },
+    {
+      "epoch": 0.27537473233404713,
+      "grad_norm": 2.9277788964177436,
+      "learning_rate": 8.510918954976394e-06,
+      "loss": 0.281,
+      "step": 643
+    },
+    {
+      "epoch": 0.27580299785867235,
+      "grad_norm": 2.547218994049393,
+      "learning_rate": 8.505975645824858e-06,
+      "loss": 0.2133,
+      "step": 644
+    },
+    {
+      "epoch": 0.2762312633832976,
+      "grad_norm": 2.742158091727297,
+      "learning_rate": 8.501025585865797e-06,
+      "loss": 0.2318,
+      "step": 645
+    },
+    {
+      "epoch": 0.2766595289079229,
+      "grad_norm": 2.7909735909331936,
+      "learning_rate": 8.49606878463063e-06,
+      "loss": 0.245,
+      "step": 646
+    },
+    {
+      "epoch": 0.27708779443254816,
+      "grad_norm": 2.7561772366416193,
+      "learning_rate": 8.491105251663746e-06,
+      "loss": 0.2522,
+      "step": 647
+    },
+    {
+      "epoch": 0.27751605995717343,
+      "grad_norm": 2.5907461668081364,
+      "learning_rate": 8.486134996522502e-06,
+      "loss": 0.2698,
+      "step": 648
+    },
+    {
+      "epoch": 0.2779443254817987,
+      "grad_norm": 2.7280941050607965,
+      "learning_rate": 8.481158028777197e-06,
+      "loss": 0.2364,
+      "step": 649
+    },
+    {
+      "epoch": 0.278372591006424,
+      "grad_norm": 2.75760305870101,
+      "learning_rate": 8.476174358011056e-06,
+      "loss": 0.2566,
+      "step": 650
+    },
+    {
+      "epoch": 0.27880085653104925,
+      "grad_norm": 2.6104009135437796,
+      "learning_rate": 8.471183993820211e-06,
+      "loss": 0.2402,
+      "step": 651
+    },
+    {
+      "epoch": 0.2792291220556745,
+      "grad_norm": 2.4986741218559025,
+      "learning_rate": 8.46618694581368e-06,
+      "loss": 0.2136,
+      "step": 652
+    },
+    {
+      "epoch": 0.2796573875802998,
+      "grad_norm": 2.856414083117755,
+      "learning_rate": 8.46118322361335e-06,
+      "loss": 0.2263,
+      "step": 653
+    },
+    {
+      "epoch": 0.28008565310492506,
+      "grad_norm": 2.844166642571859,
+      "learning_rate": 8.456172836853967e-06,
+      "loss": 0.216,
+      "step": 654
+    },
+    {
+      "epoch": 0.28051391862955033,
+      "grad_norm": 3.0066001057739613,
+      "learning_rate": 8.4511557951831e-06,
+      "loss": 0.3161,
+      "step": 655
+    },
+    {
+      "epoch": 0.2809421841541756,
+      "grad_norm": 2.5797272371890037,
+      "learning_rate": 8.446132108261136e-06,
+      "loss": 0.2029,
+      "step": 656
+    },
+    {
+      "epoch": 0.28137044967880087,
+      "grad_norm": 2.871824771303358,
+      "learning_rate": 8.441101785761261e-06,
+      "loss": 0.2765,
+      "step": 657
+    },
+    {
+      "epoch": 0.28179871520342614,
+      "grad_norm": 2.8116295161651017,
+      "learning_rate": 8.436064837369433e-06,
+      "loss": 0.2367,
+      "step": 658
+    },
+    {
+      "epoch": 0.2822269807280514,
+      "grad_norm": 2.4204129218396613,
+      "learning_rate": 8.43102127278437e-06,
+      "loss": 0.2092,
+      "step": 659
+    },
+    {
+      "epoch": 0.2826552462526767,
+      "grad_norm": 2.7085338564807073,
+      "learning_rate": 8.425971101717528e-06,
+      "loss": 0.2836,
+      "step": 660
+    },
+    {
+      "epoch": 0.28308351177730195,
+      "grad_norm": 2.576712631333003,
+      "learning_rate": 8.420914333893089e-06,
+      "loss": 0.2366,
+      "step": 661
+    },
+    {
+      "epoch": 0.28351177730192717,
+      "grad_norm": 2.63407798004816,
+      "learning_rate": 8.41585097904793e-06,
+      "loss": 0.2471,
+      "step": 662
+    },
+    {
+      "epoch": 0.28394004282655244,
+      "grad_norm": 2.490928721303279,
+      "learning_rate": 8.410781046931618e-06,
+      "loss": 0.2531,
+      "step": 663
+    },
+    {
+      "epoch": 0.2843683083511777,
+      "grad_norm": 2.9020208014633173,
+      "learning_rate": 8.405704547306379e-06,
+      "loss": 0.2465,
+      "step": 664
+    },
+    {
+      "epoch": 0.284796573875803,
+      "grad_norm": 2.637459635791023,
+      "learning_rate": 8.40062148994709e-06,
+      "loss": 0.2357,
+      "step": 665
+    },
+    {
+      "epoch": 0.28522483940042825,
+      "grad_norm": 3.1565866759963987,
+      "learning_rate": 8.395531884641252e-06,
+      "loss": 0.272,
+      "step": 666
+    },
+    {
+      "epoch": 0.2856531049250535,
+      "grad_norm": 2.544406917487968,
+      "learning_rate": 8.390435741188973e-06,
+      "loss": 0.1915,
+      "step": 667
+    },
+    {
+      "epoch": 0.2860813704496788,
+      "grad_norm": 2.857388564673147,
+      "learning_rate": 8.385333069402952e-06,
+      "loss": 0.268,
+      "step": 668
+    },
+    {
+      "epoch": 0.28650963597430407,
+      "grad_norm": 3.035444296592344,
+      "learning_rate": 8.38022387910846e-06,
+      "loss": 0.2832,
+      "step": 669
+    },
+    {
+      "epoch": 0.28693790149892934,
+      "grad_norm": 3.0031507133638935,
+      "learning_rate": 8.375108180143317e-06,
+      "loss": 0.2566,
+      "step": 670
+    },
+    {
+      "epoch": 0.2873661670235546,
+      "grad_norm": 2.8725296031624854,
+      "learning_rate": 8.369985982357877e-06,
+      "loss": 0.2777,
+      "step": 671
+    },
+    {
+      "epoch": 0.2877944325481799,
+      "grad_norm": 2.748092760918569,
+      "learning_rate": 8.364857295615006e-06,
+      "loss": 0.234,
+      "step": 672
+    },
+    {
+      "epoch": 0.28822269807280515,
+      "grad_norm": 2.6980601399612683,
+      "learning_rate": 8.359722129790065e-06,
+      "loss": 0.2508,
+      "step": 673
+    },
+    {
+      "epoch": 0.2886509635974304,
+      "grad_norm": 2.4694488375293613,
+      "learning_rate": 8.354580494770894e-06,
+      "loss": 0.2357,
+      "step": 674
+    },
+    {
+      "epoch": 0.2890792291220557,
+      "grad_norm": 2.330118874428035,
+      "learning_rate": 8.349432400457783e-06,
+      "loss": 0.2445,
+      "step": 675
+    },
+    {
+      "epoch": 0.28950749464668096,
+      "grad_norm": 2.4333542284099487,
+      "learning_rate": 8.344277856763465e-06,
+      "loss": 0.2228,
+      "step": 676
+    },
+    {
+      "epoch": 0.28993576017130623,
+      "grad_norm": 2.542531887123611,
+      "learning_rate": 8.339116873613092e-06,
+      "loss": 0.2348,
+      "step": 677
+    },
+    {
+      "epoch": 0.2903640256959315,
+      "grad_norm": 2.5017105389150127,
+      "learning_rate": 8.333949460944206e-06,
+      "loss": 0.2409,
+      "step": 678
+    },
+    {
+      "epoch": 0.2907922912205567,
+      "grad_norm": 2.950151118861109,
+      "learning_rate": 8.328775628706741e-06,
+      "loss": 0.2765,
+      "step": 679
+    },
+    {
+      "epoch": 0.291220556745182,
+      "grad_norm": 2.5633018043499782,
+      "learning_rate": 8.323595386862985e-06,
+      "loss": 0.2167,
+      "step": 680
+    },
+    {
+      "epoch": 0.29164882226980726,
+      "grad_norm": 2.837507455743504,
+      "learning_rate": 8.318408745387574e-06,
+      "loss": 0.2125,
+      "step": 681
+    },
+    {
+      "epoch": 0.29207708779443253,
+      "grad_norm": 2.6632625282648337,
+      "learning_rate": 8.313215714267454e-06,
+      "loss": 0.189,
+      "step": 682
+    },
+    {
+      "epoch": 0.2925053533190578,
+      "grad_norm": 3.18778161362762,
+      "learning_rate": 8.308016303501891e-06,
+      "loss": 0.3308,
+      "step": 683
+    },
+    {
+      "epoch": 0.2929336188436831,
+      "grad_norm": 3.0056912826526374,
+      "learning_rate": 8.302810523102422e-06,
+      "loss": 0.2303,
+      "step": 684
+    },
+    {
+      "epoch": 0.29336188436830835,
+      "grad_norm": 2.8448319112185265,
+      "learning_rate": 8.297598383092852e-06,
+      "loss": 0.1918,
+      "step": 685
+    },
+    {
+      "epoch": 0.2937901498929336,
+      "grad_norm": 3.02010894631402,
+      "learning_rate": 8.292379893509238e-06,
+      "loss": 0.2513,
+      "step": 686
+    },
+    {
+      "epoch": 0.2942184154175589,
+      "grad_norm": 2.9985742015875774,
+      "learning_rate": 8.287155064399855e-06,
+      "loss": 0.255,
+      "step": 687
+    },
+    {
+      "epoch": 0.29464668094218416,
+      "grad_norm": 2.6312479416780503,
+      "learning_rate": 8.281923905825188e-06,
+      "loss": 0.2717,
+      "step": 688
+    },
+    {
+      "epoch": 0.29507494646680943,
+      "grad_norm": 2.628760430795605,
+      "learning_rate": 8.276686427857909e-06,
+      "loss": 0.2184,
+      "step": 689
+    },
+    {
+      "epoch": 0.2955032119914347,
+      "grad_norm": 2.566113774288066,
+      "learning_rate": 8.271442640582863e-06,
+      "loss": 0.2684,
+      "step": 690
+    },
+    {
+      "epoch": 0.29593147751605997,
+      "grad_norm": 2.508802118322248,
+      "learning_rate": 8.266192554097035e-06,
+      "loss": 0.2665,
+      "step": 691
+    },
+    {
+      "epoch": 0.29635974304068524,
+      "grad_norm": 2.457965919249878,
+      "learning_rate": 8.260936178509543e-06,
+      "loss": 0.231,
+      "step": 692
+    },
+    {
+      "epoch": 0.2967880085653105,
+      "grad_norm": 2.469227826596439,
+      "learning_rate": 8.255673523941619e-06,
+      "loss": 0.2064,
+      "step": 693
+    },
+    {
+      "epoch": 0.2972162740899358,
+      "grad_norm": 3.2497648195472943,
+      "learning_rate": 8.250404600526579e-06,
+      "loss": 0.2959,
+      "step": 694
+    },
+    {
+      "epoch": 0.29764453961456105,
+      "grad_norm": 3.2585641829338017,
+      "learning_rate": 8.245129418409811e-06,
+      "loss": 0.2281,
+      "step": 695
+    },
+    {
+      "epoch": 0.29807280513918627,
+      "grad_norm": 2.576215775504094,
+      "learning_rate": 8.23984798774876e-06,
+      "loss": 0.23,
+      "step": 696
+    },
+    {
+      "epoch": 0.29850107066381154,
+      "grad_norm": 2.724421998477764,
+      "learning_rate": 8.234560318712894e-06,
+      "loss": 0.2041,
+      "step": 697
+    },
+    {
+      "epoch": 0.2989293361884368,
+      "grad_norm": 2.739898998139238,
+      "learning_rate": 8.2292664214837e-06,
+      "loss": 0.2149,
+      "step": 698
+    },
+    {
+      "epoch": 0.2993576017130621,
+      "grad_norm": 2.8059030976971826,
+      "learning_rate": 8.223966306254652e-06,
+      "loss": 0.2786,
+      "step": 699
+    },
+    {
+      "epoch": 0.29978586723768735,
+      "grad_norm": 2.7095496420953906,
+      "learning_rate": 8.218659983231203e-06,
+      "loss": 0.2351,
+      "step": 700
+    },
+    {
+      "epoch": 0.3002141327623126,
+      "grad_norm": 2.5537612315641485,
+      "learning_rate": 8.213347462630753e-06,
+      "loss": 0.2384,
+      "step": 701
+    },
+    {
+      "epoch": 0.3006423982869379,
+      "grad_norm": 2.793519391509802,
+      "learning_rate": 8.208028754682637e-06,
+      "loss": 0.2525,
+      "step": 702
+    },
+    {
+      "epoch": 0.30107066381156317,
+      "grad_norm": 2.8461648345294503,
+      "learning_rate": 8.20270386962811e-06,
+      "loss": 0.2508,
+      "step": 703
+    },
+    {
+      "epoch": 0.30149892933618844,
+      "grad_norm": 2.536967477099415,
+      "learning_rate": 8.197372817720314e-06,
+      "loss": 0.2414,
+      "step": 704
+    },
+    {
+      "epoch": 0.3019271948608137,
+      "grad_norm": 2.4970543683181736,
+      "learning_rate": 8.192035609224268e-06,
+      "loss": 0.2265,
+      "step": 705
+    },
+    {
+      "epoch": 0.302355460385439,
+      "grad_norm": 2.5506404104513782,
+      "learning_rate": 8.186692254416841e-06,
+      "loss": 0.2736,
+      "step": 706
+    },
+    {
+      "epoch": 0.30278372591006425,
+      "grad_norm": 2.582082036682954,
+      "learning_rate": 8.181342763586747e-06,
+      "loss": 0.2795,
+      "step": 707
+    },
+    {
+      "epoch": 0.3032119914346895,
+      "grad_norm": 2.5320500573374805,
+      "learning_rate": 8.175987147034505e-06,
+      "loss": 0.2301,
+      "step": 708
+    },
+    {
+      "epoch": 0.3036402569593148,
+      "grad_norm": 2.683586637788079,
+      "learning_rate": 8.170625415072437e-06,
+      "loss": 0.2434,
+      "step": 709
+    },
+    {
+      "epoch": 0.30406852248394006,
+      "grad_norm": 2.455483252030546,
+      "learning_rate": 8.165257578024635e-06,
+      "loss": 0.2662,
+      "step": 710
+    },
+    {
+      "epoch": 0.30449678800856533,
+      "grad_norm": 2.5148160143497016,
+      "learning_rate": 8.159883646226944e-06,
+      "loss": 0.1963,
+      "step": 711
+    },
+    {
+      "epoch": 0.3049250535331906,
+      "grad_norm": 2.83992913789392,
+      "learning_rate": 8.154503630026955e-06,
+      "loss": 0.2467,
+      "step": 712
+    },
+    {
+      "epoch": 0.3053533190578158,
+      "grad_norm": 3.1840310771136084,
+      "learning_rate": 8.149117539783964e-06,
+      "loss": 0.2143,
+      "step": 713
+    },
+    {
+      "epoch": 0.3057815845824411,
+      "grad_norm": 3.098593364752467,
+      "learning_rate": 8.143725385868967e-06,
+      "loss": 0.2449,
+      "step": 714
+    },
+    {
+      "epoch": 0.30620985010706636,
+      "grad_norm": 2.9004250148911126,
+      "learning_rate": 8.138327178664636e-06,
+      "loss": 0.232,
+      "step": 715
+    },
+    {
+      "epoch": 0.30663811563169163,
+      "grad_norm": 2.9315899162563523,
+      "learning_rate": 8.1329229285653e-06,
+      "loss": 0.2375,
+      "step": 716
+    },
+    {
+      "epoch": 0.3070663811563169,
+      "grad_norm": 2.834788897928478,
+      "learning_rate": 8.127512645976918e-06,
+      "loss": 0.2679,
+      "step": 717
+    },
+    {
+      "epoch": 0.3074946466809422,
+      "grad_norm": 2.770020837837968,
+      "learning_rate": 8.122096341317071e-06,
+      "loss": 0.238,
+      "step": 718
+    },
+    {
+      "epoch": 0.30792291220556745,
+      "grad_norm": 2.8161848475675235,
+      "learning_rate": 8.116674025014933e-06,
+      "loss": 0.2395,
+      "step": 719
+    },
+    {
+      "epoch": 0.3083511777301927,
+      "grad_norm": 2.5195277890184977,
+      "learning_rate": 8.111245707511253e-06,
+      "loss": 0.2339,
+      "step": 720
+    },
+    {
+      "epoch": 0.308779443254818,
+      "grad_norm": 2.4877494903790303,
+      "learning_rate": 8.105811399258336e-06,
+      "loss": 0.2446,
+      "step": 721
+    },
+    {
+      "epoch": 0.30920770877944326,
+      "grad_norm": 2.4180401662496758,
+      "learning_rate": 8.100371110720024e-06,
+      "loss": 0.2083,
+      "step": 722
+    },
+    {
+      "epoch": 0.30963597430406853,
+      "grad_norm": 2.5947996988616486,
+      "learning_rate": 8.094924852371669e-06,
+      "loss": 0.2167,
+      "step": 723
+    },
+    {
+      "epoch": 0.3100642398286938,
+      "grad_norm": 2.5894560034741185,
+      "learning_rate": 8.089472634700123e-06,
+      "loss": 0.2101,
+      "step": 724
+    },
+    {
+      "epoch": 0.31049250535331907,
+      "grad_norm": 2.719332054149059,
+      "learning_rate": 8.084014468203712e-06,
+      "loss": 0.217,
+      "step": 725
+    },
+    {
+      "epoch": 0.31092077087794434,
+      "grad_norm": 3.141719866714662,
+      "learning_rate": 8.078550363392214e-06,
+      "loss": 0.218,
+      "step": 726
+    },
+    {
+      "epoch": 0.3113490364025696,
+      "grad_norm": 2.8083607750753057,
+      "learning_rate": 8.073080330786843e-06,
+      "loss": 0.2419,
+      "step": 727
+    },
+    {
+      "epoch": 0.3117773019271949,
+      "grad_norm": 3.109808849630512,
+      "learning_rate": 8.067604380920228e-06,
+      "loss": 0.2798,
+      "step": 728
+    },
+    {
+      "epoch": 0.31220556745182015,
+      "grad_norm": 2.769000378494559,
+      "learning_rate": 8.062122524336392e-06,
+      "loss": 0.2544,
+      "step": 729
+    },
+    {
+      "epoch": 0.31263383297644537,
+      "grad_norm": 3.0704798706746286,
+      "learning_rate": 8.056634771590728e-06,
+      "loss": 0.2153,
+      "step": 730
+    },
+    {
+      "epoch": 0.31306209850107064,
+      "grad_norm": 2.732198626037193,
+      "learning_rate": 8.051141133249984e-06,
+      "loss": 0.2395,
+      "step": 731
+    },
+    {
+      "epoch": 0.3134903640256959,
+      "grad_norm": 2.7548409767380084,
+      "learning_rate": 8.045641619892243e-06,
+      "loss": 0.2375,
+      "step": 732
+    },
+    {
+      "epoch": 0.3139186295503212,
+      "grad_norm": 2.5128649719465743,
+      "learning_rate": 8.040136242106898e-06,
+      "loss": 0.2054,
+      "step": 733
+    },
+    {
+      "epoch": 0.31434689507494645,
+      "grad_norm": 2.458254095411625,
+      "learning_rate": 8.034625010494634e-06,
+      "loss": 0.1915,
+      "step": 734
+    },
+    {
+      "epoch": 0.3147751605995717,
+      "grad_norm": 2.413763138603634,
+      "learning_rate": 8.02910793566741e-06,
+      "loss": 0.228,
+      "step": 735
+    },
+    {
+      "epoch": 0.315203426124197,
+      "grad_norm": 2.5363118757333156,
+      "learning_rate": 8.023585028248435e-06,
+      "loss": 0.2085,
+      "step": 736
+    },
+    {
+      "epoch": 0.31563169164882227,
+      "grad_norm": 2.6284870944362257,
+      "learning_rate": 8.018056298872143e-06,
+      "loss": 0.2286,
+      "step": 737
+    },
+    {
+      "epoch": 0.31605995717344754,
+      "grad_norm": 2.6352010294875896,
+      "learning_rate": 8.01252175818419e-06,
+      "loss": 0.2236,
+      "step": 738
+    },
+    {
+      "epoch": 0.3164882226980728,
+      "grad_norm": 2.767686296356742,
+      "learning_rate": 8.006981416841411e-06,
+      "loss": 0.2448,
+      "step": 739
+    },
+    {
+      "epoch": 0.3169164882226981,
+      "grad_norm": 2.92177265122786,
+      "learning_rate": 8.001435285511815e-06,
+      "loss": 0.2177,
+      "step": 740
+    },
+    {
+      "epoch": 0.31734475374732335,
+      "grad_norm": 2.682347926771566,
+      "learning_rate": 7.99588337487456e-06,
+      "loss": 0.2527,
+      "step": 741
+    },
+    {
+      "epoch": 0.3177730192719486,
+      "grad_norm": 3.0105204337611706,
+      "learning_rate": 7.990325695619933e-06,
+      "loss": 0.2355,
+      "step": 742
+    },
+    {
+      "epoch": 0.3182012847965739,
+      "grad_norm": 2.588097776150605,
+      "learning_rate": 7.984762258449325e-06,
+      "loss": 0.2366,
+      "step": 743
+    },
+    {
+      "epoch": 0.31862955032119916,
+      "grad_norm": 2.8442969935584657,
+      "learning_rate": 7.979193074075216e-06,
+      "loss": 0.1896,
+      "step": 744
+    },
+    {
+      "epoch": 0.31905781584582443,
+      "grad_norm": 2.976847624774794,
+      "learning_rate": 7.97361815322115e-06,
+      "loss": 0.2975,
+      "step": 745
+    },
+    {
+      "epoch": 0.3194860813704497,
+      "grad_norm": 2.636343907475271,
+      "learning_rate": 7.968037506621724e-06,
+      "loss": 0.207,
+      "step": 746
+    },
+    {
+      "epoch": 0.3199143468950749,
+      "grad_norm": 2.7898556464932738,
+      "learning_rate": 7.962451145022552e-06,
+      "loss": 0.2473,
+      "step": 747
+    },
+    {
+      "epoch": 0.3203426124197002,
+      "grad_norm": 2.3865494143856263,
+      "learning_rate": 7.956859079180255e-06,
+      "loss": 0.206,
+      "step": 748
+    },
+    {
+      "epoch": 0.32077087794432546,
+      "grad_norm": 2.6834514843300976,
+      "learning_rate": 7.951261319862437e-06,
+      "loss": 0.2477,
+      "step": 749
+    },
+    {
+      "epoch": 0.32119914346895073,
+      "grad_norm": 2.3343761561570044,
+      "learning_rate": 7.945657877847669e-06,
+      "loss": 0.2011,
+      "step": 750
+    },
+    {
+      "epoch": 0.321627408993576,
+      "grad_norm": 2.395826600085926,
+      "learning_rate": 7.940048763925456e-06,
+      "loss": 0.1919,
+      "step": 751
+    },
+    {
+      "epoch": 0.3220556745182013,
+      "grad_norm": 2.6318673831670334,
+      "learning_rate": 7.934433988896233e-06,
+      "loss": 0.2829,
+      "step": 752
+    },
+    {
+      "epoch": 0.32248394004282654,
+      "grad_norm": 2.760497816896426,
+      "learning_rate": 7.92881356357133e-06,
+      "loss": 0.2389,
+      "step": 753
+    },
+    {
+      "epoch": 0.3229122055674518,
+      "grad_norm": 2.63451595981931,
+      "learning_rate": 7.923187498772959e-06,
+      "loss": 0.2052,
+      "step": 754
+    },
+    {
+      "epoch": 0.3233404710920771,
+      "grad_norm": 3.000663928948494,
+      "learning_rate": 7.91755580533419e-06,
+      "loss": 0.2538,
+      "step": 755
+    },
+    {
+      "epoch": 0.32376873661670236,
+      "grad_norm": 2.868834854682641,
+      "learning_rate": 7.911918494098928e-06,
+      "loss": 0.2452,
+      "step": 756
+    },
+    {
+      "epoch": 0.32419700214132763,
+      "grad_norm": 3.0352073854941426,
+      "learning_rate": 7.906275575921906e-06,
+      "loss": 0.3025,
+      "step": 757
+    },
+    {
+      "epoch": 0.3246252676659529,
+      "grad_norm": 2.5720124180583435,
+      "learning_rate": 7.90062706166864e-06,
+      "loss": 0.2012,
+      "step": 758
+    },
+    {
+      "epoch": 0.32505353319057817,
+      "grad_norm": 2.6853685533243925,
+      "learning_rate": 7.894972962215428e-06,
+      "loss": 0.2475,
+      "step": 759
+    },
+    {
+      "epoch": 0.32548179871520344,
+      "grad_norm": 2.6774556475278497,
+      "learning_rate": 7.889313288449323e-06,
+      "loss": 0.2885,
+      "step": 760
+    },
+    {
+      "epoch": 0.3259100642398287,
+      "grad_norm": 2.7544752474111145,
+      "learning_rate": 7.883648051268107e-06,
+      "loss": 0.2515,
+      "step": 761
+    },
+    {
+      "epoch": 0.326338329764454,
+      "grad_norm": 2.6470357024062325,
+      "learning_rate": 7.877977261580281e-06,
+      "loss": 0.2482,
+      "step": 762
+    },
+    {
+      "epoch": 0.32676659528907925,
+      "grad_norm": 2.482446229250705,
+      "learning_rate": 7.87230093030503e-06,
+      "loss": 0.2341,
+      "step": 763
+    },
+    {
+      "epoch": 0.3271948608137045,
+      "grad_norm": 2.760588538952661,
+      "learning_rate": 7.866619068372217e-06,
+      "loss": 0.2553,
+      "step": 764
+    },
+    {
+      "epoch": 0.32762312633832974,
+      "grad_norm": 2.6019999566135157,
+      "learning_rate": 7.860931686722346e-06,
+      "loss": 0.2594,
+      "step": 765
+    },
+    {
+      "epoch": 0.328051391862955,
+      "grad_norm": 2.6424619893843144,
+      "learning_rate": 7.855238796306558e-06,
+      "loss": 0.2109,
+      "step": 766
+    },
+    {
+      "epoch": 0.3284796573875803,
+      "grad_norm": 2.335450297648302,
+      "learning_rate": 7.849540408086592e-06,
+      "loss": 0.2216,
+      "step": 767
+    },
+    {
+      "epoch": 0.32890792291220555,
+      "grad_norm": 2.380581442416178,
+      "learning_rate": 7.843836533034784e-06,
+      "loss": 0.2229,
+      "step": 768
+    },
+    {
+      "epoch": 0.3293361884368308,
+      "grad_norm": 2.8052554744599565,
+      "learning_rate": 7.838127182134023e-06,
+      "loss": 0.2878,
+      "step": 769
+    },
+    {
+      "epoch": 0.3297644539614561,
+      "grad_norm": 2.441354266597544,
+      "learning_rate": 7.832412366377755e-06,
+      "loss": 0.1831,
+      "step": 770
+    },
+    {
+      "epoch": 0.33019271948608137,
+      "grad_norm": 2.5387454691973446,
+      "learning_rate": 7.826692096769935e-06,
+      "loss": 0.2374,
+      "step": 771
+    },
+    {
+      "epoch": 0.33062098501070664,
+      "grad_norm": 3.1416417217819417,
+      "learning_rate": 7.82096638432503e-06,
+      "loss": 0.2765,
+      "step": 772
+    },
+    {
+      "epoch": 0.3310492505353319,
+      "grad_norm": 2.774958602309695,
+      "learning_rate": 7.81523524006798e-06,
+      "loss": 0.2304,
+      "step": 773
+    },
+    {
+      "epoch": 0.3314775160599572,
+      "grad_norm": 2.306383299294921,
+      "learning_rate": 7.809498675034191e-06,
+      "loss": 0.1742,
+      "step": 774
+    },
+    {
+      "epoch": 0.33190578158458245,
+      "grad_norm": 2.8049487003742186,
+      "learning_rate": 7.803756700269501e-06,
+      "loss": 0.196,
+      "step": 775
+    },
+    {
+      "epoch": 0.3323340471092077,
+      "grad_norm": 2.6640915421263833,
+      "learning_rate": 7.798009326830167e-06,
+      "loss": 0.2512,
+      "step": 776
+    },
+    {
+      "epoch": 0.332762312633833,
+      "grad_norm": 2.8753907523825455,
+      "learning_rate": 7.792256565782841e-06,
+      "loss": 0.2487,
+      "step": 777
+    },
+    {
+      "epoch": 0.33319057815845826,
+      "grad_norm": 2.69166038088852,
+      "learning_rate": 7.786498428204547e-06,
+      "loss": 0.2538,
+      "step": 778
+    },
+    {
+      "epoch": 0.33361884368308353,
+      "grad_norm": 3.103868520291351,
+      "learning_rate": 7.780734925182666e-06,
+      "loss": 0.2851,
+      "step": 779
+    },
+    {
+      "epoch": 0.3340471092077088,
+      "grad_norm": 2.683604651112579,
+      "learning_rate": 7.774966067814906e-06,
+      "loss": 0.2647,
+      "step": 780
+    },
+    {
+      "epoch": 0.3344753747323341,
+      "grad_norm": 2.460845079556041,
+      "learning_rate": 7.76919186720929e-06,
+      "loss": 0.2187,
+      "step": 781
+    },
+    {
+      "epoch": 0.3349036402569593,
+      "grad_norm": 2.293937189428902,
+      "learning_rate": 7.76341233448412e-06,
+      "loss": 0.2039,
+      "step": 782
+    },
+    {
+      "epoch": 0.33533190578158456,
+      "grad_norm": 2.676324777757,
+      "learning_rate": 7.757627480767973e-06,
+      "loss": 0.2201,
+      "step": 783
+    },
+    {
+      "epoch": 0.33576017130620983,
+      "grad_norm": 2.5130815761397254,
+      "learning_rate": 7.751837317199673e-06,
+      "loss": 0.2263,
+      "step": 784
+    },
+    {
+      "epoch": 0.3361884368308351,
+      "grad_norm": 2.6395054818886017,
+      "learning_rate": 7.74604185492826e-06,
+      "loss": 0.2177,
+      "step": 785
+    },
+    {
+      "epoch": 0.3366167023554604,
+      "grad_norm": 2.6574450808494614,
+      "learning_rate": 7.740241105112987e-06,
+      "loss": 0.2891,
+      "step": 786
+    },
+    {
+      "epoch": 0.33704496788008564,
+      "grad_norm": 2.740865640111425,
+      "learning_rate": 7.734435078923276e-06,
+      "loss": 0.206,
+      "step": 787
+    },
+    {
+      "epoch": 0.3374732334047109,
+      "grad_norm": 3.0383623072291353,
+      "learning_rate": 7.728623787538722e-06,
+      "loss": 0.265,
+      "step": 788
+    },
+    {
+      "epoch": 0.3379014989293362,
+      "grad_norm": 2.6754521966242355,
+      "learning_rate": 7.722807242149051e-06,
+      "loss": 0.2214,
+      "step": 789
+    },
+    {
+      "epoch": 0.33832976445396146,
+      "grad_norm": 2.8706342149065027,
+      "learning_rate": 7.716985453954103e-06,
+      "loss": 0.1947,
+      "step": 790
+    },
+    {
+      "epoch": 0.3387580299785867,
+      "grad_norm": 2.7131229302173767,
+      "learning_rate": 7.711158434163817e-06,
+      "loss": 0.2048,
+      "step": 791
+    },
+    {
+      "epoch": 0.339186295503212,
+      "grad_norm": 2.5436582392989093,
+      "learning_rate": 7.705326193998207e-06,
+      "loss": 0.2072,
+      "step": 792
+    },
+    {
+      "epoch": 0.33961456102783727,
+      "grad_norm": 2.584102283522472,
+      "learning_rate": 7.699488744687337e-06,
+      "loss": 0.1721,
+      "step": 793
+    },
+    {
+      "epoch": 0.34004282655246254,
+      "grad_norm": 2.954477286705158,
+      "learning_rate": 7.6936460974713e-06,
+      "loss": 0.2664,
+      "step": 794
+    },
+    {
+      "epoch": 0.3404710920770878,
+      "grad_norm": 2.7449494514626935,
+      "learning_rate": 7.6877982636002e-06,
+      "loss": 0.1972,
+      "step": 795
+    },
+    {
+      "epoch": 0.3408993576017131,
+      "grad_norm": 2.7119472980015336,
+      "learning_rate": 7.681945254334126e-06,
+      "loss": 0.2648,
+      "step": 796
+    },
+    {
+      "epoch": 0.34132762312633835,
+      "grad_norm": 2.6057975881659687,
+      "learning_rate": 7.676087080943133e-06,
+      "loss": 0.1898,
+      "step": 797
+    },
+    {
+      "epoch": 0.3417558886509636,
+      "grad_norm": 2.8234723220891724,
+      "learning_rate": 7.67022375470722e-06,
+      "loss": 0.2738,
+      "step": 798
+    },
+    {
+      "epoch": 0.34218415417558884,
+      "grad_norm": 2.435000254909467,
+      "learning_rate": 7.664355286916308e-06,
+      "loss": 0.2547,
+      "step": 799
+    },
+    {
+      "epoch": 0.3426124197002141,
+      "grad_norm": 2.4658709164097394,
+      "learning_rate": 7.658481688870218e-06,
+      "loss": 0.2293,
+      "step": 800
+    },
+    {
+      "epoch": 0.3430406852248394,
+      "grad_norm": 2.5881855046448026,
+      "learning_rate": 7.652602971878649e-06,
+      "loss": 0.2438,
+      "step": 801
+    },
+    {
+      "epoch": 0.34346895074946465,
+      "grad_norm": 2.4116697252487747,
+      "learning_rate": 7.646719147261156e-06,
+      "loss": 0.2101,
+      "step": 802
+    },
+    {
+      "epoch": 0.3438972162740899,
+      "grad_norm": 2.4902774182610963,
+      "learning_rate": 7.640830226347133e-06,
+      "loss": 0.2342,
+      "step": 803
+    },
+    {
+      "epoch": 0.3443254817987152,
+      "grad_norm": 2.942996743161323,
+      "learning_rate": 7.634936220475777e-06,
+      "loss": 0.2644,
+      "step": 804
+    },
+    {
+      "epoch": 0.34475374732334046,
+      "grad_norm": 3.047263616032456,
+      "learning_rate": 7.629037140996089e-06,
+      "loss": 0.2266,
+      "step": 805
+    },
+    {
+      "epoch": 0.34518201284796574,
+      "grad_norm": 2.5976550135406984,
+      "learning_rate": 7.6231329992668305e-06,
+      "loss": 0.2103,
+      "step": 806
+    },
+    {
+      "epoch": 0.345610278372591,
+      "grad_norm": 2.7697911505542665,
+      "learning_rate": 7.617223806656513e-06,
+      "loss": 0.2309,
+      "step": 807
+    },
+    {
+      "epoch": 0.3460385438972163,
+      "grad_norm": 2.597981943824152,
+      "learning_rate": 7.611309574543373e-06,
+      "loss": 0.2119,
+      "step": 808
+    },
+    {
+      "epoch": 0.34646680942184155,
+      "grad_norm": 3.1188098042456636,
+      "learning_rate": 7.605390314315353e-06,
+      "loss": 0.2518,
+      "step": 809
+    },
+    {
+      "epoch": 0.3468950749464668,
+      "grad_norm": 2.517509765380658,
+      "learning_rate": 7.599466037370073e-06,
+      "loss": 0.1897,
+      "step": 810
+    },
+    {
+      "epoch": 0.3473233404710921,
+      "grad_norm": 2.4226219764938075,
+      "learning_rate": 7.593536755114817e-06,
+      "loss": 0.2294,
+      "step": 811
+    },
+    {
+      "epoch": 0.34775160599571736,
+      "grad_norm": 2.730692591458189,
+      "learning_rate": 7.587602478966503e-06,
+      "loss": 0.2275,
+      "step": 812
+    },
+    {
+      "epoch": 0.34817987152034263,
+      "grad_norm": 3.0417425496244967,
+      "learning_rate": 7.581663220351669e-06,
+      "loss": 0.2402,
+      "step": 813
+    },
+    {
+      "epoch": 0.3486081370449679,
+      "grad_norm": 2.6628290059224513,
+      "learning_rate": 7.575718990706442e-06,
+      "loss": 0.2142,
+      "step": 814
+    },
+    {
+      "epoch": 0.3490364025695932,
+      "grad_norm": 2.4900785516889634,
+      "learning_rate": 7.569769801476527e-06,
+      "loss": 0.2074,
+      "step": 815
+    },
+    {
+      "epoch": 0.3494646680942184,
+      "grad_norm": 2.513681901230348,
+      "learning_rate": 7.563815664117173e-06,
+      "loss": 0.2418,
+      "step": 816
+    },
+    {
+      "epoch": 0.34989293361884366,
+      "grad_norm": 2.373118438572732,
+      "learning_rate": 7.557856590093158e-06,
+      "loss": 0.1969,
+      "step": 817
+    },
+    {
+      "epoch": 0.35032119914346893,
+      "grad_norm": 2.382816150654525,
+      "learning_rate": 7.551892590878766e-06,
+      "loss": 0.2197,
+      "step": 818
+    },
+    {
+      "epoch": 0.3507494646680942,
+      "grad_norm": 2.880317034753188,
+      "learning_rate": 7.545923677957769e-06,
+      "loss": 0.1966,
+      "step": 819
+    },
+    {
+      "epoch": 0.3511777301927195,
+      "grad_norm": 2.438366574328022,
+      "learning_rate": 7.5399498628233925e-06,
+      "loss": 0.174,
+      "step": 820
+    },
+    {
+      "epoch": 0.35160599571734474,
+      "grad_norm": 2.7000709527410267,
+      "learning_rate": 7.5339711569783054e-06,
+      "loss": 0.2437,
+      "step": 821
+    },
+    {
+      "epoch": 0.35203426124197,
+      "grad_norm": 2.890852072267884,
+      "learning_rate": 7.527987571934596e-06,
+      "loss": 0.2119,
+      "step": 822
+    },
+    {
+      "epoch": 0.3524625267665953,
+      "grad_norm": 2.377905288229827,
+      "learning_rate": 7.521999119213741e-06,
+      "loss": 0.1815,
+      "step": 823
+    },
+    {
+      "epoch": 0.35289079229122056,
+      "grad_norm": 2.634073051760519,
+      "learning_rate": 7.5160058103465985e-06,
+      "loss": 0.2132,
+      "step": 824
+    },
+    {
+      "epoch": 0.3533190578158458,
+      "grad_norm": 2.604933526660534,
+      "learning_rate": 7.5100076568733705e-06,
+      "loss": 0.2453,
+      "step": 825
+    },
+    {
+      "epoch": 0.3537473233404711,
+      "grad_norm": 2.671406183115142,
+      "learning_rate": 7.504004670343588e-06,
+      "loss": 0.2375,
+      "step": 826
+    },
+    {
+      "epoch": 0.35417558886509637,
+      "grad_norm": 2.5649675079378764,
+      "learning_rate": 7.497996862316091e-06,
+      "loss": 0.1771,
+      "step": 827
+    },
+    {
+      "epoch": 0.35460385438972164,
+      "grad_norm": 2.581791705899053,
+      "learning_rate": 7.491984244359003e-06,
+      "loss": 0.2324,
+      "step": 828
+    },
+    {
+      "epoch": 0.3550321199143469,
+      "grad_norm": 2.6418610681081804,
+      "learning_rate": 7.485966828049707e-06,
+      "loss": 0.1874,
+      "step": 829
+    },
+    {
+      "epoch": 0.3554603854389722,
+      "grad_norm": 2.413222311052329,
+      "learning_rate": 7.479944624974824e-06,
+      "loss": 0.1988,
+      "step": 830
+    },
+    {
+      "epoch": 0.35588865096359745,
+      "grad_norm": 2.4910359943704754,
+      "learning_rate": 7.473917646730199e-06,
+      "loss": 0.2163,
+      "step": 831
+    },
+    {
+      "epoch": 0.3563169164882227,
+      "grad_norm": 2.6887838077379635,
+      "learning_rate": 7.467885904920864e-06,
+      "loss": 0.2577,
+      "step": 832
+    },
+    {
+      "epoch": 0.35674518201284794,
+      "grad_norm": 2.3874491939415745,
+      "learning_rate": 7.461849411161027e-06,
+      "loss": 0.2084,
+      "step": 833
+    },
+    {
+      "epoch": 0.3571734475374732,
+      "grad_norm": 2.6868352240644318,
+      "learning_rate": 7.455808177074046e-06,
+      "loss": 0.1757,
+      "step": 834
+    },
+    {
+      "epoch": 0.3576017130620985,
+      "grad_norm": 2.6299010234644613,
+      "learning_rate": 7.449762214292404e-06,
+      "loss": 0.2313,
+      "step": 835
+    },
+    {
+      "epoch": 0.35802997858672375,
+      "grad_norm": 2.575065185049202,
+      "learning_rate": 7.4437115344576935e-06,
+      "loss": 0.227,
+      "step": 836
+    },
+    {
+      "epoch": 0.358458244111349,
+      "grad_norm": 2.772284933736181,
+      "learning_rate": 7.437656149220584e-06,
+      "loss": 0.2065,
+      "step": 837
+    },
+    {
+      "epoch": 0.3588865096359743,
+      "grad_norm": 2.663960274446012,
+      "learning_rate": 7.431596070240812e-06,
+      "loss": 0.1988,
+      "step": 838
+    },
+    {
+      "epoch": 0.35931477516059956,
+      "grad_norm": 2.492564726956308,
+      "learning_rate": 7.425531309187146e-06,
+      "loss": 0.2031,
+      "step": 839
+    },
+    {
+      "epoch": 0.35974304068522484,
+      "grad_norm": 2.387499540512249,
+      "learning_rate": 7.419461877737373e-06,
+      "loss": 0.2264,
+      "step": 840
+    },
+    {
+      "epoch": 0.3601713062098501,
+      "grad_norm": 2.879183475037556,
+      "learning_rate": 7.413387787578274e-06,
+      "loss": 0.2067,
+      "step": 841
+    },
+    {
+      "epoch": 0.3605995717344754,
+      "grad_norm": 2.5338673715894497,
+      "learning_rate": 7.4073090504055964e-06,
+      "loss": 0.1792,
+      "step": 842
+    },
+    {
+      "epoch": 0.36102783725910065,
+      "grad_norm": 2.88511607720729,
+      "learning_rate": 7.401225677924038e-06,
+      "loss": 0.2408,
+      "step": 843
+    },
+    {
+      "epoch": 0.3614561027837259,
+      "grad_norm": 2.57188490833629,
+      "learning_rate": 7.395137681847223e-06,
+      "loss": 0.1768,
+      "step": 844
+    },
+    {
+      "epoch": 0.3618843683083512,
+      "grad_norm": 2.8449206082833034,
+      "learning_rate": 7.389045073897676e-06,
+      "loss": 0.2003,
+      "step": 845
+    },
+    {
+      "epoch": 0.36231263383297646,
+      "grad_norm": 2.4760676551707874,
+      "learning_rate": 7.3829478658068035e-06,
+      "loss": 0.2111,
+      "step": 846
+    },
+    {
+      "epoch": 0.36274089935760173,
+      "grad_norm": 2.761734373386148,
+      "learning_rate": 7.376846069314869e-06,
+      "loss": 0.2487,
+      "step": 847
+    },
+    {
+      "epoch": 0.363169164882227,
+      "grad_norm": 2.6773914857678993,
+      "learning_rate": 7.370739696170971e-06,
+      "loss": 0.2486,
+      "step": 848
+    },
+    {
+      "epoch": 0.3635974304068523,
+      "grad_norm": 2.4958531128970893,
+      "learning_rate": 7.36462875813302e-06,
+      "loss": 0.2338,
+      "step": 849
+    },
+    {
+      "epoch": 0.3640256959314775,
+      "grad_norm": 2.3885400677742545,
+      "learning_rate": 7.358513266967717e-06,
+      "loss": 0.1859,
+      "step": 850
+    },
+    {
+      "epoch": 0.36445396145610276,
+      "grad_norm": 2.288002040474808,
+      "learning_rate": 7.3523932344505325e-06,
+      "loss": 0.2003,
+      "step": 851
+    },
+    {
+      "epoch": 0.36488222698072803,
+      "grad_norm": 2.4245656580901955,
+      "learning_rate": 7.346268672365675e-06,
+      "loss": 0.1937,
+      "step": 852
+    },
+    {
+      "epoch": 0.3653104925053533,
+      "grad_norm": 2.674148743268735,
+      "learning_rate": 7.340139592506082e-06,
+      "loss": 0.2217,
+      "step": 853
+    },
+    {
+      "epoch": 0.3657387580299786,
+      "grad_norm": 2.7935797140176506,
+      "learning_rate": 7.334006006673383e-06,
+      "loss": 0.2167,
+      "step": 854
+    },
+    {
+      "epoch": 0.36616702355460384,
+      "grad_norm": 2.422330853236713,
+      "learning_rate": 7.32786792667789e-06,
+      "loss": 0.1724,
+      "step": 855
+    },
+    {
+      "epoch": 0.3665952890792291,
+      "grad_norm": 3.1410197045466433,
+      "learning_rate": 7.321725364338566e-06,
+      "loss": 0.2593,
+      "step": 856
+    },
+    {
+      "epoch": 0.3670235546038544,
+      "grad_norm": 2.911272233719246,
+      "learning_rate": 7.315578331483005e-06,
+      "loss": 0.2447,
+      "step": 857
+    },
+    {
+      "epoch": 0.36745182012847966,
+      "grad_norm": 2.751249292921079,
+      "learning_rate": 7.309426839947407e-06,
+      "loss": 0.1936,
+      "step": 858
+    },
+    {
+      "epoch": 0.3678800856531049,
+      "grad_norm": 2.675794505285213,
+      "learning_rate": 7.30327090157656e-06,
+      "loss": 0.1916,
+      "step": 859
+    },
+    {
+      "epoch": 0.3683083511777302,
+      "grad_norm": 2.6736473218854084,
+      "learning_rate": 7.297110528223817e-06,
+      "loss": 0.1914,
+      "step": 860
+    },
+    {
+      "epoch": 0.36873661670235547,
+      "grad_norm": 2.6205282300608137,
+      "learning_rate": 7.2909457317510615e-06,
+      "loss": 0.2033,
+      "step": 861
+    },
+    {
+      "epoch": 0.36916488222698074,
+      "grad_norm": 2.719008378586701,
+      "learning_rate": 7.284776524028703e-06,
+      "loss": 0.2215,
+      "step": 862
+    },
+    {
+      "epoch": 0.369593147751606,
+      "grad_norm": 2.7866597958422883,
+      "learning_rate": 7.278602916935641e-06,
+      "loss": 0.2388,
+      "step": 863
+    },
+    {
+      "epoch": 0.3700214132762313,
+      "grad_norm": 3.113407791389773,
+      "learning_rate": 7.272424922359246e-06,
+      "loss": 0.2773,
+      "step": 864
+    },
+    {
+      "epoch": 0.37044967880085655,
+      "grad_norm": 2.5068544488653575,
+      "learning_rate": 7.266242552195337e-06,
+      "loss": 0.1968,
+      "step": 865
+    },
+    {
+      "epoch": 0.3708779443254818,
+      "grad_norm": 2.7479703648622125,
+      "learning_rate": 7.260055818348159e-06,
+      "loss": 0.2388,
+      "step": 866
+    },
+    {
+      "epoch": 0.3713062098501071,
+      "grad_norm": 2.2339647493601142,
+      "learning_rate": 7.253864732730359e-06,
+      "loss": 0.1352,
+      "step": 867
+    },
+    {
+      "epoch": 0.3717344753747323,
+      "grad_norm": 2.9348987949327956,
+      "learning_rate": 7.247669307262964e-06,
+      "loss": 0.2271,
+      "step": 868
+    },
+    {
+      "epoch": 0.3721627408993576,
+      "grad_norm": 2.43646412638881,
+      "learning_rate": 7.2414695538753535e-06,
+      "loss": 0.1735,
+      "step": 869
+    },
+    {
+      "epoch": 0.37259100642398285,
+      "grad_norm": 2.697073756157046,
+      "learning_rate": 7.235265484505245e-06,
+      "loss": 0.2418,
+      "step": 870
+    },
+    {
+      "epoch": 0.3730192719486081,
+      "grad_norm": 2.7058423829355376,
+      "learning_rate": 7.2290571110986665e-06,
+      "loss": 0.236,
+      "step": 871
+    },
+    {
+      "epoch": 0.3734475374732334,
+      "grad_norm": 2.4048361416108532,
+      "learning_rate": 7.222844445609931e-06,
+      "loss": 0.2022,
+      "step": 872
+    },
+    {
+      "epoch": 0.37387580299785866,
+      "grad_norm": 2.5875908312860756,
+      "learning_rate": 7.216627500001616e-06,
+      "loss": 0.1981,
+      "step": 873
+    },
+    {
+      "epoch": 0.37430406852248393,
+      "grad_norm": 2.4705789046782183,
+      "learning_rate": 7.210406286244543e-06,
+      "loss": 0.211,
+      "step": 874
+    },
+    {
+      "epoch": 0.3747323340471092,
+      "grad_norm": 2.770005216408,
+      "learning_rate": 7.20418081631775e-06,
+      "loss": 0.2207,
+      "step": 875
+    },
+    {
+      "epoch": 0.3751605995717345,
+      "grad_norm": 2.341177825496814,
+      "learning_rate": 7.19795110220847e-06,
+      "loss": 0.1863,
+      "step": 876
+    },
+    {
+      "epoch": 0.37558886509635975,
+      "grad_norm": 2.778785013941377,
+      "learning_rate": 7.191717155912111e-06,
+      "loss": 0.2298,
+      "step": 877
+    },
+    {
+      "epoch": 0.376017130620985,
+      "grad_norm": 2.6469023146952932,
+      "learning_rate": 7.185478989432225e-06,
+      "loss": 0.1816,
+      "step": 878
+    },
+    {
+      "epoch": 0.3764453961456103,
+      "grad_norm": 2.6408186369946405,
+      "learning_rate": 7.179236614780494e-06,
+      "loss": 0.2091,
+      "step": 879
+    },
+    {
+      "epoch": 0.37687366167023556,
+      "grad_norm": 2.4280099605384438,
+      "learning_rate": 7.172990043976703e-06,
+      "loss": 0.2623,
+      "step": 880
+    },
+    {
+      "epoch": 0.37730192719486083,
+      "grad_norm": 2.7142226082972334,
+      "learning_rate": 7.166739289048715e-06,
+      "loss": 0.2047,
+      "step": 881
+    },
+    {
+      "epoch": 0.3777301927194861,
+      "grad_norm": 2.9249528330039327,
+      "learning_rate": 7.160484362032449e-06,
+      "loss": 0.178,
+      "step": 882
+    },
+    {
+      "epoch": 0.3781584582441114,
+      "grad_norm": 2.4638173487548083,
+      "learning_rate": 7.15422527497186e-06,
+      "loss": 0.1925,
+      "step": 883
+    },
+    {
+      "epoch": 0.37858672376873664,
+      "grad_norm": 2.8607108398040078,
+      "learning_rate": 7.147962039918913e-06,
+      "loss": 0.2503,
+      "step": 884
+    },
+    {
+      "epoch": 0.37901498929336186,
+      "grad_norm": 2.696820565436154,
+      "learning_rate": 7.141694668933558e-06,
+      "loss": 0.2534,
+      "step": 885
+    },
+    {
+      "epoch": 0.37944325481798713,
+      "grad_norm": 2.598314713225295,
+      "learning_rate": 7.135423174083708e-06,
+      "loss": 0.2113,
+      "step": 886
+    },
+    {
+      "epoch": 0.3798715203426124,
+      "grad_norm": 2.659992935964774,
+      "learning_rate": 7.129147567445221e-06,
+      "loss": 0.2071,
+      "step": 887
+    },
+    {
+      "epoch": 0.38029978586723767,
+      "grad_norm": 2.707274345411228,
+      "learning_rate": 7.122867861101868e-06,
+      "loss": 0.2397,
+      "step": 888
+    },
+    {
+      "epoch": 0.38072805139186294,
+      "grad_norm": 2.482224324913828,
+      "learning_rate": 7.116584067145317e-06,
+      "loss": 0.1798,
+      "step": 889
+    },
+    {
+      "epoch": 0.3811563169164882,
+      "grad_norm": 2.549656583182633,
+      "learning_rate": 7.110296197675104e-06,
+      "loss": 0.2545,
+      "step": 890
+    },
+    {
+      "epoch": 0.3815845824411135,
+      "grad_norm": 2.5227176311971173,
+      "learning_rate": 7.104004264798614e-06,
+      "loss": 0.2312,
+      "step": 891
+    },
+    {
+      "epoch": 0.38201284796573876,
+      "grad_norm": 2.499631554919232,
+      "learning_rate": 7.097708280631057e-06,
+      "loss": 0.2151,
+      "step": 892
+    },
+    {
+      "epoch": 0.382441113490364,
+      "grad_norm": 2.463858613407602,
+      "learning_rate": 7.091408257295443e-06,
+      "loss": 0.227,
+      "step": 893
+    },
+    {
+      "epoch": 0.3828693790149893,
+      "grad_norm": 2.603529429224385,
+      "learning_rate": 7.085104206922557e-06,
+      "loss": 0.2186,
+      "step": 894
+    },
+    {
+      "epoch": 0.38329764453961457,
+      "grad_norm": 2.383699442272966,
+      "learning_rate": 7.078796141650943e-06,
+      "loss": 0.2026,
+      "step": 895
+    },
+    {
+      "epoch": 0.38372591006423984,
+      "grad_norm": 2.757362867188944,
+      "learning_rate": 7.072484073626872e-06,
+      "loss": 0.2087,
+      "step": 896
+    },
+    {
+      "epoch": 0.3841541755888651,
+      "grad_norm": 2.7479458338166003,
+      "learning_rate": 7.066168015004324e-06,
+      "loss": 0.1939,
+      "step": 897
+    },
+    {
+      "epoch": 0.3845824411134904,
+      "grad_norm": 2.596208001107085,
+      "learning_rate": 7.059847977944962e-06,
+      "loss": 0.2223,
+      "step": 898
+    },
+    {
+      "epoch": 0.38501070663811565,
+      "grad_norm": 2.93736227234744,
+      "learning_rate": 7.0535239746181115e-06,
+      "loss": 0.2152,
+      "step": 899
+    },
+    {
+      "epoch": 0.3854389721627409,
+      "grad_norm": 2.893701521848815,
+      "learning_rate": 7.047196017200731e-06,
+      "loss": 0.2232,
+      "step": 900
+    },
+    {
+      "epoch": 0.3858672376873662,
+      "grad_norm": 3.2701103480068854,
+      "learning_rate": 7.040864117877398e-06,
+      "loss": 0.2787,
+      "step": 901
+    },
+    {
+      "epoch": 0.3862955032119914,
+      "grad_norm": 2.7568921760071836,
+      "learning_rate": 7.034528288840276e-06,
+      "loss": 0.2169,
+      "step": 902
+    },
+    {
+      "epoch": 0.3867237687366167,
+      "grad_norm": 2.7726797158320013,
+      "learning_rate": 7.028188542289096e-06,
+      "loss": 0.1997,
+      "step": 903
+    },
+    {
+      "epoch": 0.38715203426124195,
+      "grad_norm": 2.508376837819604,
+      "learning_rate": 7.021844890431136e-06,
+      "loss": 0.2154,
+      "step": 904
+    },
+    {
+      "epoch": 0.3875802997858672,
+      "grad_norm": 2.288406984031857,
+      "learning_rate": 7.015497345481187e-06,
+      "loss": 0.1863,
+      "step": 905
+    },
+    {
+      "epoch": 0.3880085653104925,
+      "grad_norm": 2.455673471714184,
+      "learning_rate": 7.009145919661542e-06,
+      "loss": 0.2192,
+      "step": 906
+    },
+    {
+      "epoch": 0.38843683083511776,
+      "grad_norm": 2.630655277480143,
+      "learning_rate": 7.002790625201964e-06,
+      "loss": 0.1814,
+      "step": 907
+    },
+    {
+      "epoch": 0.38886509635974303,
+      "grad_norm": 2.7654968206310833,
+      "learning_rate": 6.996431474339666e-06,
+      "loss": 0.2206,
+      "step": 908
+    },
+    {
+      "epoch": 0.3892933618843683,
+      "grad_norm": 2.6314050904041495,
+      "learning_rate": 6.9900684793192864e-06,
+      "loss": 0.1938,
+      "step": 909
+    },
+    {
+      "epoch": 0.3897216274089936,
+      "grad_norm": 2.9919450595672004,
+      "learning_rate": 6.983701652392864e-06,
+      "loss": 0.2175,
+      "step": 910
+    },
+    {
+      "epoch": 0.39014989293361885,
+      "grad_norm": 2.4996239328693717,
+      "learning_rate": 6.977331005819821e-06,
+      "loss": 0.2308,
+      "step": 911
+    },
+    {
+      "epoch": 0.3905781584582441,
+      "grad_norm": 2.691253287152104,
+      "learning_rate": 6.970956551866925e-06,
+      "loss": 0.2023,
+      "step": 912
+    },
+    {
+      "epoch": 0.3910064239828694,
+      "grad_norm": 2.6697061997689584,
+      "learning_rate": 6.9645783028082884e-06,
+      "loss": 0.2385,
+      "step": 913
+    },
+    {
+      "epoch": 0.39143468950749466,
+      "grad_norm": 2.9123596342631144,
+      "learning_rate": 6.958196270925315e-06,
+      "loss": 0.2508,
+      "step": 914
+    },
+    {
+      "epoch": 0.39186295503211993,
+      "grad_norm": 2.5739456766022766,
+      "learning_rate": 6.951810468506706e-06,
+      "loss": 0.2331,
+      "step": 915
+    },
+    {
+      "epoch": 0.3922912205567452,
+      "grad_norm": 2.4736244706042503,
+      "learning_rate": 6.945420907848415e-06,
+      "loss": 0.2329,
+      "step": 916
+    },
+    {
+      "epoch": 0.3927194860813705,
+      "grad_norm": 2.7251034408151606,
+      "learning_rate": 6.939027601253636e-06,
+      "loss": 0.2109,
+      "step": 917
+    },
+    {
+      "epoch": 0.39314775160599574,
+      "grad_norm": 2.287371446384498,
+      "learning_rate": 6.932630561032776e-06,
+      "loss": 0.2256,
+      "step": 918
+    },
+    {
+      "epoch": 0.39357601713062096,
+      "grad_norm": 2.4042628028713042,
+      "learning_rate": 6.926229799503428e-06,
+      "loss": 0.1756,
+      "step": 919
+    },
+    {
+      "epoch": 0.39400428265524623,
+      "grad_norm": 2.6665043787459752,
+      "learning_rate": 6.9198253289903515e-06,
+      "loss": 0.2217,
+      "step": 920
+    },
+    {
+      "epoch": 0.3944325481798715,
+      "grad_norm": 2.8262841305826467,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.2194,
+      "step": 921
+    },
+    {
+      "epoch": 0.39486081370449677,
+      "grad_norm": 2.780504550892749,
+      "learning_rate": 6.907005310347742e-06,
+      "loss": 0.2566,
+      "step": 922
+    },
+    {
+      "epoch": 0.39528907922912204,
+      "grad_norm": 2.626223306545082,
+      "learning_rate": 6.900589786903343e-06,
+      "loss": 0.1921,
+      "step": 923
+    },
+    {
+      "epoch": 0.3957173447537473,
+      "grad_norm": 2.7152036741929404,
+      "learning_rate": 6.894170603845436e-06,
+      "loss": 0.2169,
+      "step": 924
+    },
+    {
+      "epoch": 0.3961456102783726,
+      "grad_norm": 2.6634411801124953,
+      "learning_rate": 6.887747773534255e-06,
+      "loss": 0.2453,
+      "step": 925
+    },
+    {
+      "epoch": 0.39657387580299786,
+      "grad_norm": 2.521897196312532,
+      "learning_rate": 6.881321308337049e-06,
+      "loss": 0.1941,
+      "step": 926
+    },
+    {
+      "epoch": 0.3970021413276231,
+      "grad_norm": 2.6980241847816813,
+      "learning_rate": 6.874891220628076e-06,
+      "loss": 0.2657,
+      "step": 927
+    },
+    {
+      "epoch": 0.3974304068522484,
+      "grad_norm": 2.3721449425974352,
+      "learning_rate": 6.868457522788561e-06,
+      "loss": 0.19,
+      "step": 928
+    },
+    {
+      "epoch": 0.39785867237687367,
+      "grad_norm": 2.65081145776317,
+      "learning_rate": 6.862020227206684e-06,
+      "loss": 0.2114,
+      "step": 929
+    },
+    {
+      "epoch": 0.39828693790149894,
+      "grad_norm": 2.8983416360532566,
+      "learning_rate": 6.855579346277554e-06,
+      "loss": 0.2199,
+      "step": 930
+    },
+    {
+      "epoch": 0.3987152034261242,
+      "grad_norm": 2.4936133927750816,
+      "learning_rate": 6.849134892403179e-06,
+      "loss": 0.1782,
+      "step": 931
+    },
+    {
+      "epoch": 0.3991434689507495,
+      "grad_norm": 2.6225688951539445,
+      "learning_rate": 6.842686877992453e-06,
+      "loss": 0.1798,
+      "step": 932
+    },
+    {
+      "epoch": 0.39957173447537475,
+      "grad_norm": 2.3797864342409927,
+      "learning_rate": 6.836235315461117e-06,
+      "loss": 0.2086,
+      "step": 933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.297603442830615,
+      "learning_rate": 6.829780217231754e-06,
+      "loss": 0.165,
+      "step": 934
+    },
+    {
+      "epoch": 0.4004282655246253,
+      "grad_norm": 2.5174049768965157,
+      "learning_rate": 6.823321595733749e-06,
+      "loss": 0.2024,
+      "step": 935
+    },
+    {
+      "epoch": 0.4008565310492505,
+      "grad_norm": 2.9104583140794924,
+      "learning_rate": 6.816859463403271e-06,
+      "loss": 0.2214,
+      "step": 936
+    },
+    {
+      "epoch": 0.4012847965738758,
+      "grad_norm": 2.7436765804461647,
+      "learning_rate": 6.810393832683252e-06,
+      "loss": 0.2061,
+      "step": 937
+    },
+    {
+      "epoch": 0.40171306209850105,
+      "grad_norm": 2.3765889058279623,
+      "learning_rate": 6.803924716023358e-06,
+      "loss": 0.1636,
+      "step": 938
+    },
+    {
+      "epoch": 0.4021413276231263,
+      "grad_norm": 2.695626930559915,
+      "learning_rate": 6.7974521258799686e-06,
+      "loss": 0.1956,
+      "step": 939
+    },
+    {
+      "epoch": 0.4025695931477516,
+      "grad_norm": 2.6287375217239766,
+      "learning_rate": 6.790976074716151e-06,
+      "loss": 0.184,
+      "step": 940
+    },
+    {
+      "epoch": 0.40299785867237686,
+      "grad_norm": 2.4256142809364896,
+      "learning_rate": 6.784496575001636e-06,
+      "loss": 0.2638,
+      "step": 941
+    },
+    {
+      "epoch": 0.40342612419700213,
+      "grad_norm": 2.793534074150348,
+      "learning_rate": 6.778013639212796e-06,
+      "loss": 0.2125,
+      "step": 942
+    },
+    {
+      "epoch": 0.4038543897216274,
+      "grad_norm": 2.792768329417172,
+      "learning_rate": 6.771527279832618e-06,
+      "loss": 0.2009,
+      "step": 943
+    },
+    {
+      "epoch": 0.4042826552462527,
+      "grad_norm": 2.7169852585801335,
+      "learning_rate": 6.765037509350685e-06,
+      "loss": 0.2293,
+      "step": 944
+    },
+    {
+      "epoch": 0.40471092077087795,
+      "grad_norm": 2.5742831235732204,
+      "learning_rate": 6.758544340263141e-06,
+      "loss": 0.2294,
+      "step": 945
+    },
+    {
+      "epoch": 0.4051391862955032,
+      "grad_norm": 2.552857880410138,
+      "learning_rate": 6.752047785072682e-06,
+      "loss": 0.2148,
+      "step": 946
+    },
+    {
+      "epoch": 0.4055674518201285,
+      "grad_norm": 2.6381382543313285,
+      "learning_rate": 6.745547856288517e-06,
+      "loss": 0.1818,
+      "step": 947
+    },
+    {
+      "epoch": 0.40599571734475376,
+      "grad_norm": 2.3420670872807454,
+      "learning_rate": 6.7390445664263586e-06,
+      "loss": 0.1663,
+      "step": 948
+    },
+    {
+      "epoch": 0.40642398286937903,
+      "grad_norm": 2.4967110847987075,
+      "learning_rate": 6.732537928008382e-06,
+      "loss": 0.1744,
+      "step": 949
+    },
+    {
+      "epoch": 0.4068522483940043,
+      "grad_norm": 2.8549497535059736,
+      "learning_rate": 6.7260279535632165e-06,
+      "loss": 0.199,
+      "step": 950
+    },
+    {
+      "epoch": 0.40728051391862957,
+      "grad_norm": 2.4907870788601056,
+      "learning_rate": 6.719514655625914e-06,
+      "loss": 0.2273,
+      "step": 951
+    },
+    {
+      "epoch": 0.40770877944325484,
+      "grad_norm": 2.4617793825548913,
+      "learning_rate": 6.7129980467379265e-06,
+      "loss": 0.1703,
+      "step": 952
+    },
+    {
+      "epoch": 0.40813704496788006,
+      "grad_norm": 2.5591827086705936,
+      "learning_rate": 6.706478139447077e-06,
+      "loss": 0.194,
+      "step": 953
+    },
+    {
+      "epoch": 0.40856531049250533,
+      "grad_norm": 2.4983503432981733,
+      "learning_rate": 6.699954946307545e-06,
+      "loss": 0.2261,
+      "step": 954
+    },
+    {
+      "epoch": 0.4089935760171306,
+      "grad_norm": 2.5501238854232926,
+      "learning_rate": 6.693428479879833e-06,
+      "loss": 0.221,
+      "step": 955
+    },
+    {
+      "epoch": 0.40942184154175587,
+      "grad_norm": 2.2421349824310335,
+      "learning_rate": 6.686898752730751e-06,
+      "loss": 0.157,
+      "step": 956
+    },
+    {
+      "epoch": 0.40985010706638114,
+      "grad_norm": 2.6755997420855615,
+      "learning_rate": 6.680365777433379e-06,
+      "loss": 0.2007,
+      "step": 957
+    },
+    {
+      "epoch": 0.4102783725910064,
+      "grad_norm": 2.386184731115499,
+      "learning_rate": 6.673829566567059e-06,
+      "loss": 0.2011,
+      "step": 958
+    },
+    {
+      "epoch": 0.4107066381156317,
+      "grad_norm": 2.7792613427657806,
+      "learning_rate": 6.667290132717365e-06,
+      "loss": 0.207,
+      "step": 959
+    },
+    {
+      "epoch": 0.41113490364025695,
+      "grad_norm": 2.5929014170207267,
+      "learning_rate": 6.660747488476066e-06,
+      "loss": 0.1656,
+      "step": 960
+    },
+    {
+      "epoch": 0.4115631691648822,
+      "grad_norm": 3.0166822538363993,
+      "learning_rate": 6.6542016464411254e-06,
+      "loss": 0.2108,
+      "step": 961
+    },
+    {
+      "epoch": 0.4119914346895075,
+      "grad_norm": 2.593607743766415,
+      "learning_rate": 6.6476526192166525e-06,
+      "loss": 0.1772,
+      "step": 962
+    },
+    {
+      "epoch": 0.41241970021413277,
+      "grad_norm": 2.5354611486891505,
+      "learning_rate": 6.6411004194129e-06,
+      "loss": 0.1865,
+      "step": 963
+    },
+    {
+      "epoch": 0.41284796573875804,
+      "grad_norm": 2.6981154226446864,
+      "learning_rate": 6.6345450596462224e-06,
+      "loss": 0.2063,
+      "step": 964
+    },
+    {
+      "epoch": 0.4132762312633833,
+      "grad_norm": 2.74239729795775,
+      "learning_rate": 6.6279865525390605e-06,
+      "loss": 0.2157,
+      "step": 965
+    },
+    {
+      "epoch": 0.4137044967880086,
+      "grad_norm": 3.2357715910027474,
+      "learning_rate": 6.62142491071992e-06,
+      "loss": 0.22,
+      "step": 966
+    },
+    {
+      "epoch": 0.41413276231263385,
+      "grad_norm": 2.70462666899732,
+      "learning_rate": 6.614860146823333e-06,
+      "loss": 0.2083,
+      "step": 967
+    },
+    {
+      "epoch": 0.4145610278372591,
+      "grad_norm": 2.206917977542963,
+      "learning_rate": 6.608292273489851e-06,
+      "loss": 0.1856,
+      "step": 968
+    },
+    {
+      "epoch": 0.4149892933618844,
+      "grad_norm": 2.7987046948822862,
+      "learning_rate": 6.601721303366011e-06,
+      "loss": 0.2261,
+      "step": 969
+    },
+    {
+      "epoch": 0.41541755888650966,
+      "grad_norm": 2.5524281278778482,
+      "learning_rate": 6.595147249104311e-06,
+      "loss": 0.2312,
+      "step": 970
+    },
+    {
+      "epoch": 0.4158458244111349,
+      "grad_norm": 2.7628830833275972,
+      "learning_rate": 6.588570123363192e-06,
+      "loss": 0.2004,
+      "step": 971
+    },
+    {
+      "epoch": 0.41627408993576015,
+      "grad_norm": 2.4347827238551356,
+      "learning_rate": 6.581989938807001e-06,
+      "loss": 0.1666,
+      "step": 972
+    },
+    {
+      "epoch": 0.4167023554603854,
+      "grad_norm": 2.2894829910300323,
+      "learning_rate": 6.575406708105985e-06,
+      "loss": 0.2008,
+      "step": 973
+    },
+    {
+      "epoch": 0.4171306209850107,
+      "grad_norm": 2.4167896048641273,
+      "learning_rate": 6.568820443936246e-06,
+      "loss": 0.1836,
+      "step": 974
+    },
+    {
+      "epoch": 0.41755888650963596,
+      "grad_norm": 2.655055492551505,
+      "learning_rate": 6.562231158979737e-06,
+      "loss": 0.1824,
+      "step": 975
+    },
+    {
+      "epoch": 0.41798715203426123,
+      "grad_norm": 2.7021329145370667,
+      "learning_rate": 6.555638865924221e-06,
+      "loss": 0.2376,
+      "step": 976
+    },
+    {
+      "epoch": 0.4184154175588865,
+      "grad_norm": 2.660897560302102,
+      "learning_rate": 6.549043577463254e-06,
+      "loss": 0.2421,
+      "step": 977
+    },
+    {
+      "epoch": 0.4188436830835118,
+      "grad_norm": 2.883569277982489,
+      "learning_rate": 6.542445306296163e-06,
+      "loss": 0.1739,
+      "step": 978
+    },
+    {
+      "epoch": 0.41927194860813705,
+      "grad_norm": 2.880720445703276,
+      "learning_rate": 6.535844065128012e-06,
+      "loss": 0.2134,
+      "step": 979
+    },
+    {
+      "epoch": 0.4197002141327623,
+      "grad_norm": 2.746731361552014,
+      "learning_rate": 6.529239866669592e-06,
+      "loss": 0.1983,
+      "step": 980
+    },
+    {
+      "epoch": 0.4201284796573876,
+      "grad_norm": 2.226449102412465,
+      "learning_rate": 6.522632723637381e-06,
+      "loss": 0.1787,
+      "step": 981
+    },
+    {
+      "epoch": 0.42055674518201286,
+      "grad_norm": 2.484027619715287,
+      "learning_rate": 6.516022648753533e-06,
+      "loss": 0.2161,
+      "step": 982
+    },
+    {
+      "epoch": 0.42098501070663813,
+      "grad_norm": 2.8093354042561165,
+      "learning_rate": 6.509409654745841e-06,
+      "loss": 0.1949,
+      "step": 983
+    },
+    {
+      "epoch": 0.4214132762312634,
+      "grad_norm": 2.6134002836139594,
+      "learning_rate": 6.502793754347721e-06,
+      "loss": 0.1975,
+      "step": 984
+    },
+    {
+      "epoch": 0.42184154175588867,
+      "grad_norm": 2.648238442444048,
+      "learning_rate": 6.4961749602981916e-06,
+      "loss": 0.2161,
+      "step": 985
+    },
+    {
+      "epoch": 0.42226980728051394,
+      "grad_norm": 2.417179168155311,
+      "learning_rate": 6.4895532853418315e-06,
+      "loss": 0.2042,
+      "step": 986
+    },
+    {
+      "epoch": 0.4226980728051392,
+      "grad_norm": 2.5058019321159737,
+      "learning_rate": 6.4829287422287765e-06,
+      "loss": 0.195,
+      "step": 987
+    },
+    {
+      "epoch": 0.42312633832976443,
+      "grad_norm": 2.7918984540675216,
+      "learning_rate": 6.476301343714682e-06,
+      "loss": 0.1976,
+      "step": 988
+    },
+    {
+      "epoch": 0.4235546038543897,
+      "grad_norm": 2.6838860232558504,
+      "learning_rate": 6.469671102560698e-06,
+      "loss": 0.2367,
+      "step": 989
+    },
+    {
+      "epoch": 0.42398286937901497,
+      "grad_norm": 2.319605024812522,
+      "learning_rate": 6.463038031533455e-06,
+      "loss": 0.2287,
+      "step": 990
+    },
+    {
+      "epoch": 0.42441113490364024,
+      "grad_norm": 2.4043514813159264,
+      "learning_rate": 6.456402143405026e-06,
+      "loss": 0.2054,
+      "step": 991
+    },
+    {
+      "epoch": 0.4248394004282655,
+      "grad_norm": 2.385835650986772,
+      "learning_rate": 6.449763450952912e-06,
+      "loss": 0.1999,
+      "step": 992
+    },
+    {
+      "epoch": 0.4252676659528908,
+      "grad_norm": 2.509819687127884,
+      "learning_rate": 6.443121966960011e-06,
+      "loss": 0.203,
+      "step": 993
+    },
+    {
+      "epoch": 0.42569593147751605,
+      "grad_norm": 2.322492900819722,
+      "learning_rate": 6.436477704214599e-06,
+      "loss": 0.1838,
+      "step": 994
+    },
+    {
+      "epoch": 0.4261241970021413,
+      "grad_norm": 2.5776410243319825,
+      "learning_rate": 6.429830675510301e-06,
+      "loss": 0.2438,
+      "step": 995
+    },
+    {
+      "epoch": 0.4265524625267666,
+      "grad_norm": 2.3538078943597993,
+      "learning_rate": 6.423180893646068e-06,
+      "loss": 0.1878,
+      "step": 996
+    },
+    {
+      "epoch": 0.42698072805139187,
+      "grad_norm": 2.813234510526458,
+      "learning_rate": 6.416528371426155e-06,
+      "loss": 0.1932,
+      "step": 997
+    },
+    {
+      "epoch": 0.42740899357601714,
+      "grad_norm": 2.866946153638045,
+      "learning_rate": 6.409873121660088e-06,
+      "loss": 0.2173,
+      "step": 998
+    },
+    {
+      "epoch": 0.4278372591006424,
+      "grad_norm": 2.6808304731697468,
+      "learning_rate": 6.40321515716265e-06,
+      "loss": 0.1652,
+      "step": 999
+    },
+    {
+      "epoch": 0.4282655246252677,
+      "grad_norm": 2.2330409883688884,
+      "learning_rate": 6.396554490753848e-06,
+      "loss": 0.1498,
+      "step": 1000
+    },
+    {
+      "epoch": 0.42869379014989295,
+      "grad_norm": 2.7538899739170586,
+      "learning_rate": 6.389891135258893e-06,
+      "loss": 0.254,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4291220556745182,
+      "grad_norm": 2.663350851056386,
+      "learning_rate": 6.383225103508175e-06,
+      "loss": 0.2277,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4295503211991435,
+      "grad_norm": 2.712401190079425,
+      "learning_rate": 6.376556408337233e-06,
+      "loss": 0.1763,
+      "step": 1003
+    },
+    {
+      "epoch": 0.42997858672376876,
+      "grad_norm": 2.7901463452216437,
+      "learning_rate": 6.369885062586741e-06,
+      "loss": 0.2002,
+      "step": 1004
+    },
+    {
+      "epoch": 0.430406852248394,
+      "grad_norm": 2.6065059179536423,
+      "learning_rate": 6.36321107910247e-06,
+      "loss": 0.1957,
+      "step": 1005
+    },
+    {
+      "epoch": 0.43083511777301925,
+      "grad_norm": 2.500841862886674,
+      "learning_rate": 6.356534470735274e-06,
+      "loss": 0.1765,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4312633832976445,
+      "grad_norm": 2.5652519669800045,
+      "learning_rate": 6.349855250341063e-06,
+      "loss": 0.1935,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4316916488222698,
+      "grad_norm": 2.208472030647767,
+      "learning_rate": 6.343173430780769e-06,
+      "loss": 0.1321,
+      "step": 1008
+    },
+    {
+      "epoch": 0.43211991434689506,
+      "grad_norm": 2.3073006175299087,
+      "learning_rate": 6.336489024920338e-06,
+      "loss": 0.1379,
+      "step": 1009
+    },
+    {
+      "epoch": 0.43254817987152033,
+      "grad_norm": 2.730992738390745,
+      "learning_rate": 6.3298020456306885e-06,
+      "loss": 0.1809,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4329764453961456,
+      "grad_norm": 2.3216025221462386,
+      "learning_rate": 6.3231125057876996e-06,
+      "loss": 0.1605,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4334047109207709,
+      "grad_norm": 2.5275899077572346,
+      "learning_rate": 6.316420418272176e-06,
+      "loss": 0.2097,
+      "step": 1012
+    },
+    {
+      "epoch": 0.43383297644539615,
+      "grad_norm": 2.946082288732787,
+      "learning_rate": 6.309725795969832e-06,
+      "loss": 0.2029,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4342612419700214,
+      "grad_norm": 2.5567826795463815,
+      "learning_rate": 6.30302865177126e-06,
+      "loss": 0.169,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4346895074946467,
+      "grad_norm": 2.796213231319827,
+      "learning_rate": 6.29632899857191e-06,
+      "loss": 0.218,
+      "step": 1015
+    },
+    {
+      "epoch": 0.43511777301927196,
+      "grad_norm": 2.8040535572865863,
+      "learning_rate": 6.289626849272062e-06,
+      "loss": 0.2138,
+      "step": 1016
+    },
+    {
+      "epoch": 0.43554603854389723,
+      "grad_norm": 2.775118567472346,
+      "learning_rate": 6.2829222167768035e-06,
+      "loss": 0.2134,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4359743040685225,
+      "grad_norm": 2.7602196761671682,
+      "learning_rate": 6.2762151139960026e-06,
+      "loss": 0.1948,
+      "step": 1018
+    },
+    {
+      "epoch": 0.43640256959314777,
+      "grad_norm": 2.5174080439393247,
+      "learning_rate": 6.269505553844284e-06,
+      "loss": 0.1842,
+      "step": 1019
+    },
+    {
+      "epoch": 0.43683083511777304,
+      "grad_norm": 2.2430038611627383,
+      "learning_rate": 6.262793549241003e-06,
+      "loss": 0.1774,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4372591006423983,
+      "grad_norm": 2.305059612776913,
+      "learning_rate": 6.256079113110225e-06,
+      "loss": 0.1511,
+      "step": 1021
+    },
+    {
+      "epoch": 0.43768736616702353,
+      "grad_norm": 2.496110150545391,
+      "learning_rate": 6.249362258380692e-06,
+      "loss": 0.2326,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4381156316916488,
+      "grad_norm": 2.229451285731679,
+      "learning_rate": 6.2426429979858085e-06,
+      "loss": 0.1601,
+      "step": 1023
+    },
+    {
+      "epoch": 0.43854389721627407,
+      "grad_norm": 2.400973508478108,
+      "learning_rate": 6.2359213448636104e-06,
+      "loss": 0.182,
+      "step": 1024
+    },
+    {
+      "epoch": 0.43897216274089934,
+      "grad_norm": 2.4018102463814777,
+      "learning_rate": 6.229197311956736e-06,
+      "loss": 0.2026,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4394004282655246,
+      "grad_norm": 2.993372393994128,
+      "learning_rate": 6.222470912212413e-06,
+      "loss": 0.1772,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4398286937901499,
+      "grad_norm": 2.322576894122534,
+      "learning_rate": 6.215742158582419e-06,
+      "loss": 0.1541,
+      "step": 1027
+    },
+    {
+      "epoch": 0.44025695931477515,
+      "grad_norm": 2.8753946634125342,
+      "learning_rate": 6.209011064023072e-06,
+      "loss": 0.2088,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4406852248394004,
+      "grad_norm": 2.6794267096771573,
+      "learning_rate": 6.202277641495188e-06,
+      "loss": 0.1976,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4411134903640257,
+      "grad_norm": 2.579476193237105,
+      "learning_rate": 6.195541903964074e-06,
+      "loss": 0.1694,
+      "step": 1030
+    },
+    {
+      "epoch": 0.44154175588865097,
+      "grad_norm": 2.5049099672637367,
+      "learning_rate": 6.188803864399492e-06,
+      "loss": 0.1867,
+      "step": 1031
+    },
+    {
+      "epoch": 0.44197002141327624,
+      "grad_norm": 2.6272262299572464,
+      "learning_rate": 6.182063535775634e-06,
+      "loss": 0.1403,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4423982869379015,
+      "grad_norm": 2.4503247627818876,
+      "learning_rate": 6.175320931071105e-06,
+      "loss": 0.1777,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4428265524625268,
+      "grad_norm": 2.5728787111764952,
+      "learning_rate": 6.168576063268884e-06,
+      "loss": 0.2045,
+      "step": 1034
+    },
+    {
+      "epoch": 0.44325481798715205,
+      "grad_norm": 2.761346244425715,
+      "learning_rate": 6.161828945356318e-06,
+      "loss": 0.194,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4436830835117773,
+      "grad_norm": 2.2593691666523887,
+      "learning_rate": 6.155079590325079e-06,
+      "loss": 0.1972,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4441113490364026,
+      "grad_norm": 2.538153914032363,
+      "learning_rate": 6.148328011171147e-06,
+      "loss": 0.1996,
+      "step": 1037
+    },
+    {
+      "epoch": 0.44453961456102786,
+      "grad_norm": 2.5718458119629535,
+      "learning_rate": 6.141574220894793e-06,
+      "loss": 0.2006,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4449678800856531,
+      "grad_norm": 2.1421922481215785,
+      "learning_rate": 6.1348182325005314e-06,
+      "loss": 0.1712,
+      "step": 1039
+    },
+    {
+      "epoch": 0.44539614561027835,
+      "grad_norm": 2.7712032771392874,
+      "learning_rate": 6.1280600589971225e-06,
+      "loss": 0.2035,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4458244111349036,
+      "grad_norm": 2.377108896263671,
+      "learning_rate": 6.121299713397526e-06,
+      "loss": 0.1943,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4462526766595289,
+      "grad_norm": 2.3702882765685636,
+      "learning_rate": 6.114537208718888e-06,
+      "loss": 0.224,
+      "step": 1042
+    },
+    {
+      "epoch": 0.44668094218415416,
+      "grad_norm": 2.173809515313738,
+      "learning_rate": 6.10777255798251e-06,
+      "loss": 0.1706,
+      "step": 1043
+    },
+    {
+      "epoch": 0.44710920770877943,
+      "grad_norm": 2.4866767418035916,
+      "learning_rate": 6.1010057742138255e-06,
+      "loss": 0.188,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4475374732334047,
+      "grad_norm": 2.8688519408249444,
+      "learning_rate": 6.094236870442376e-06,
+      "loss": 0.2245,
+      "step": 1045
+    },
+    {
+      "epoch": 0.44796573875803,
+      "grad_norm": 2.55650726813499,
+      "learning_rate": 6.087465859701784e-06,
+      "loss": 0.2118,
+      "step": 1046
+    },
+    {
+      "epoch": 0.44839400428265525,
+      "grad_norm": 2.488382184415331,
+      "learning_rate": 6.080692755029734e-06,
+      "loss": 0.2045,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4488222698072805,
+      "grad_norm": 2.3551818763413124,
+      "learning_rate": 6.073917569467934e-06,
+      "loss": 0.1793,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4492505353319058,
+      "grad_norm": 2.7578800296141495,
+      "learning_rate": 6.0671403160621045e-06,
+      "loss": 0.2106,
+      "step": 1049
+    },
+    {
+      "epoch": 0.44967880085653106,
+      "grad_norm": 2.6888579079373405,
+      "learning_rate": 6.060361007861948e-06,
+      "loss": 0.2058,
+      "step": 1050
+    },
+    {
+      "epoch": 0.45010706638115633,
+      "grad_norm": 2.7389336436922282,
+      "learning_rate": 6.053579657921118e-06,
+      "loss": 0.196,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4505353319057816,
+      "grad_norm": 2.6998708216443963,
+      "learning_rate": 6.046796279297208e-06,
+      "loss": 0.1722,
+      "step": 1052
+    },
+    {
+      "epoch": 0.45096359743040687,
+      "grad_norm": 2.743903903842347,
+      "learning_rate": 6.040010885051711e-06,
+      "loss": 0.2065,
+      "step": 1053
+    },
+    {
+      "epoch": 0.45139186295503214,
+      "grad_norm": 2.7561614101880787,
+      "learning_rate": 6.033223488250001e-06,
+      "loss": 0.2175,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4518201284796574,
+      "grad_norm": 2.3232416942130456,
+      "learning_rate": 6.026434101961314e-06,
+      "loss": 0.1788,
+      "step": 1055
+    },
+    {
+      "epoch": 0.45224839400428263,
+      "grad_norm": 2.830505766052931,
+      "learning_rate": 6.0196427392587085e-06,
+      "loss": 0.2468,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4526766595289079,
+      "grad_norm": 2.2917393235442542,
+      "learning_rate": 6.012849413219057e-06,
+      "loss": 0.1456,
+      "step": 1057
+    },
+    {
+      "epoch": 0.45310492505353317,
+      "grad_norm": 2.4474753944881416,
+      "learning_rate": 6.0060541369230055e-06,
+      "loss": 0.2038,
+      "step": 1058
+    },
+    {
+      "epoch": 0.45353319057815844,
+      "grad_norm": 2.5169538828812117,
+      "learning_rate": 5.999256923454959e-06,
+      "loss": 0.159,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4539614561027837,
+      "grad_norm": 2.6837048503104124,
+      "learning_rate": 5.992457785903054e-06,
+      "loss": 0.2119,
+      "step": 1060
+    },
+    {
+      "epoch": 0.454389721627409,
+      "grad_norm": 2.4529850490522365,
+      "learning_rate": 5.985656737359125e-06,
+      "loss": 0.1898,
+      "step": 1061
+    },
+    {
+      "epoch": 0.45481798715203425,
+      "grad_norm": 2.474429009243246,
+      "learning_rate": 5.978853790918696e-06,
+      "loss": 0.1893,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4552462526766595,
+      "grad_norm": 2.725251997649122,
+      "learning_rate": 5.972048959680936e-06,
+      "loss": 0.2002,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4556745182012848,
+      "grad_norm": 2.6339385261579484,
+      "learning_rate": 5.96524225674865e-06,
+      "loss": 0.1857,
+      "step": 1064
+    },
+    {
+      "epoch": 0.45610278372591007,
+      "grad_norm": 2.4726323367911505,
+      "learning_rate": 5.958433695228242e-06,
+      "loss": 0.2038,
+      "step": 1065
+    },
+    {
+      "epoch": 0.45653104925053534,
+      "grad_norm": 2.219615305752368,
+      "learning_rate": 5.951623288229699e-06,
+      "loss": 0.1319,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4569593147751606,
+      "grad_norm": 2.5244632425727405,
+      "learning_rate": 5.9448110488665576e-06,
+      "loss": 0.2102,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4573875802997859,
+      "grad_norm": 2.276734834949852,
+      "learning_rate": 5.937996990255886e-06,
+      "loss": 0.1899,
+      "step": 1068
+    },
+    {
+      "epoch": 0.45781584582441115,
+      "grad_norm": 2.5603989651544037,
+      "learning_rate": 5.931181125518255e-06,
+      "loss": 0.1908,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4582441113490364,
+      "grad_norm": 2.7942273026063837,
+      "learning_rate": 5.924363467777709e-06,
+      "loss": 0.1506,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4586723768736617,
+      "grad_norm": 2.71634395100704,
+      "learning_rate": 5.917544030161752e-06,
+      "loss": 0.204,
+      "step": 1071
+    },
+    {
+      "epoch": 0.45910064239828696,
+      "grad_norm": 2.5411048149825377,
+      "learning_rate": 5.9107228258013085e-06,
+      "loss": 0.1668,
+      "step": 1072
+    },
+    {
+      "epoch": 0.45952890792291223,
+      "grad_norm": 2.8242222200567006,
+      "learning_rate": 5.903899867830707e-06,
+      "loss": 0.2229,
+      "step": 1073
+    },
+    {
+      "epoch": 0.45995717344753745,
+      "grad_norm": 2.6715047128856004,
+      "learning_rate": 5.897075169387655e-06,
+      "loss": 0.2207,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4603854389721627,
+      "grad_norm": 2.514129611820294,
+      "learning_rate": 5.890248743613207e-06,
+      "loss": 0.1683,
+      "step": 1075
+    },
+    {
+      "epoch": 0.460813704496788,
+      "grad_norm": 3.6605580982129498,
+      "learning_rate": 5.883420603651749e-06,
+      "loss": 0.1638,
+      "step": 1076
+    },
+    {
+      "epoch": 0.46124197002141326,
+      "grad_norm": 2.540626929611411,
+      "learning_rate": 5.8765907626509625e-06,
+      "loss": 0.2208,
+      "step": 1077
+    },
+    {
+      "epoch": 0.46167023554603853,
+      "grad_norm": 2.659327924725805,
+      "learning_rate": 5.869759233761807e-06,
+      "loss": 0.1632,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4620985010706638,
+      "grad_norm": 2.63608970700839,
+      "learning_rate": 5.8629260301384924e-06,
+      "loss": 0.1768,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4625267665952891,
+      "grad_norm": 2.206067706314631,
+      "learning_rate": 5.856091164938451e-06,
+      "loss": 0.1506,
+      "step": 1080
+    },
+    {
+      "epoch": 0.46295503211991434,
+      "grad_norm": 2.9145804822844465,
+      "learning_rate": 5.849254651322317e-06,
+      "loss": 0.2132,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4633832976445396,
+      "grad_norm": 2.4756199676617663,
+      "learning_rate": 5.842416502453897e-06,
+      "loss": 0.2039,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4638115631691649,
+      "grad_norm": 2.5225140475086714,
+      "learning_rate": 5.8355767315001485e-06,
+      "loss": 0.218,
+      "step": 1083
+    },
+    {
+      "epoch": 0.46423982869379016,
+      "grad_norm": 2.4210439606739445,
+      "learning_rate": 5.828735351631149e-06,
+      "loss": 0.1902,
+      "step": 1084
+    },
+    {
+      "epoch": 0.46466809421841543,
+      "grad_norm": 2.6430036841327476,
+      "learning_rate": 5.821892376020075e-06,
+      "loss": 0.1402,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4650963597430407,
+      "grad_norm": 2.12461662415332,
+      "learning_rate": 5.815047817843179e-06,
+      "loss": 0.1241,
+      "step": 1086
+    },
+    {
+      "epoch": 0.46552462526766597,
+      "grad_norm": 2.3577110854579892,
+      "learning_rate": 5.808201690279757e-06,
+      "loss": 0.1796,
+      "step": 1087
+    },
+    {
+      "epoch": 0.46595289079229124,
+      "grad_norm": 2.4931640161723587,
+      "learning_rate": 5.801354006512127e-06,
+      "loss": 0.1498,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4663811563169165,
+      "grad_norm": 2.660646947308086,
+      "learning_rate": 5.794504779725607e-06,
+      "loss": 0.2052,
+      "step": 1089
+    },
+    {
+      "epoch": 0.4668094218415418,
+      "grad_norm": 2.718440068007936,
+      "learning_rate": 5.787654023108481e-06,
+      "loss": 0.1999,
+      "step": 1090
+    },
+    {
+      "epoch": 0.467237687366167,
+      "grad_norm": 2.4969117311308433,
+      "learning_rate": 5.780801749851983e-06,
+      "loss": 0.1786,
+      "step": 1091
+    },
+    {
+      "epoch": 0.46766595289079227,
+      "grad_norm": 2.837765269601434,
+      "learning_rate": 5.773947973150265e-06,
+      "loss": 0.2423,
+      "step": 1092
+    },
+    {
+      "epoch": 0.46809421841541754,
+      "grad_norm": 2.5052723678948365,
+      "learning_rate": 5.767092706200375e-06,
+      "loss": 0.1738,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4685224839400428,
+      "grad_norm": 2.4480994828841887,
+      "learning_rate": 5.7602359622022276e-06,
+      "loss": 0.1517,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4689507494646681,
+      "grad_norm": 2.5291021643025444,
+      "learning_rate": 5.753377754358587e-06,
+      "loss": 0.196,
+      "step": 1095
+    },
+    {
+      "epoch": 0.46937901498929335,
+      "grad_norm": 2.505007833475143,
+      "learning_rate": 5.746518095875033e-06,
+      "loss": 0.1657,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4698072805139186,
+      "grad_norm": 2.7105799475170675,
+      "learning_rate": 5.739656999959936e-06,
+      "loss": 0.1576,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4702355460385439,
+      "grad_norm": 2.6829362075492633,
+      "learning_rate": 5.73279447982444e-06,
+      "loss": 0.2101,
+      "step": 1098
+    },
+    {
+      "epoch": 0.47066381156316917,
+      "grad_norm": 2.651240608848617,
+      "learning_rate": 5.725930548682425e-06,
+      "loss": 0.1943,
+      "step": 1099
+    },
+    {
+      "epoch": 0.47109207708779444,
+      "grad_norm": 2.6364981115196096,
+      "learning_rate": 5.719065219750493e-06,
+      "loss": 0.1822,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4715203426124197,
+      "grad_norm": 2.4204264410788396,
+      "learning_rate": 5.712198506247935e-06,
+      "loss": 0.1759,
+      "step": 1101
+    },
+    {
+      "epoch": 0.471948608137045,
+      "grad_norm": 2.341481407828839,
+      "learning_rate": 5.70533042139671e-06,
+      "loss": 0.1938,
+      "step": 1102
+    },
+    {
+      "epoch": 0.47237687366167025,
+      "grad_norm": 2.7077794595710207,
+      "learning_rate": 5.698460978421413e-06,
+      "loss": 0.1904,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4728051391862955,
+      "grad_norm": 2.2557959667242966,
+      "learning_rate": 5.6915901905492586e-06,
+      "loss": 0.1424,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4732334047109208,
+      "grad_norm": 3.1824196063529624,
+      "learning_rate": 5.6847180710100526e-06,
+      "loss": 0.2427,
+      "step": 1105
+    },
+    {
+      "epoch": 0.47366167023554606,
+      "grad_norm": 2.524223632077738,
+      "learning_rate": 5.677844633036157e-06,
+      "loss": 0.1674,
+      "step": 1106
+    },
+    {
+      "epoch": 0.47408993576017133,
+      "grad_norm": 2.652593659263161,
+      "learning_rate": 5.670969889862481e-06,
+      "loss": 0.1791,
+      "step": 1107
+    },
+    {
+      "epoch": 0.47451820128479655,
+      "grad_norm": 2.5710734934052084,
+      "learning_rate": 5.664093854726442e-06,
+      "loss": 0.1823,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4749464668094218,
+      "grad_norm": 2.388498336589055,
+      "learning_rate": 5.6572165408679454e-06,
+      "loss": 0.216,
+      "step": 1109
+    },
+    {
+      "epoch": 0.4753747323340471,
+      "grad_norm": 2.1303521217620474,
+      "learning_rate": 5.650337961529364e-06,
+      "loss": 0.1506,
+      "step": 1110
+    },
+    {
+      "epoch": 0.47580299785867236,
+      "grad_norm": 2.3279828641656706,
+      "learning_rate": 5.643458129955497e-06,
+      "loss": 0.1637,
+      "step": 1111
+    },
+    {
+      "epoch": 0.47623126338329763,
+      "grad_norm": 2.841189866432095,
+      "learning_rate": 5.6365770593935665e-06,
+      "loss": 0.1833,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4766595289079229,
+      "grad_norm": 2.358789356174836,
+      "learning_rate": 5.629694763093172e-06,
+      "loss": 0.1519,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4770877944325482,
+      "grad_norm": 2.952200427997332,
+      "learning_rate": 5.622811254306275e-06,
+      "loss": 0.2593,
+      "step": 1114
+    },
+    {
+      "epoch": 0.47751605995717344,
+      "grad_norm": 2.6652087278861085,
+      "learning_rate": 5.6159265462871775e-06,
+      "loss": 0.2102,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4779443254817987,
+      "grad_norm": 2.7814163294936556,
+      "learning_rate": 5.609040652292479e-06,
+      "loss": 0.1784,
+      "step": 1116
+    },
+    {
+      "epoch": 0.478372591006424,
+      "grad_norm": 2.372052359799946,
+      "learning_rate": 5.602153585581075e-06,
+      "loss": 0.1664,
+      "step": 1117
+    },
+    {
+      "epoch": 0.47880085653104926,
+      "grad_norm": 2.752611534379897,
+      "learning_rate": 5.59526535941411e-06,
+      "loss": 0.1511,
+      "step": 1118
+    },
+    {
+      "epoch": 0.47922912205567453,
+      "grad_norm": 2.660967619766456,
+      "learning_rate": 5.588375987054967e-06,
+      "loss": 0.2228,
+      "step": 1119
+    },
+    {
+      "epoch": 0.4796573875802998,
+      "grad_norm": 2.6766210678510416,
+      "learning_rate": 5.581485481769231e-06,
+      "loss": 0.1707,
+      "step": 1120
+    },
+    {
+      "epoch": 0.48008565310492507,
+      "grad_norm": 2.62781295748319,
+      "learning_rate": 5.574593856824673e-06,
+      "loss": 0.1866,
+      "step": 1121
+    },
+    {
+      "epoch": 0.48051391862955034,
+      "grad_norm": 2.6491539679063547,
+      "learning_rate": 5.567701125491217e-06,
+      "loss": 0.2016,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4809421841541756,
+      "grad_norm": 2.442654025680973,
+      "learning_rate": 5.560807301040918e-06,
+      "loss": 0.1384,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4813704496788009,
+      "grad_norm": 2.5143625607786246,
+      "learning_rate": 5.553912396747938e-06,
+      "loss": 0.203,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4817987152034261,
+      "grad_norm": 2.7409314585209272,
+      "learning_rate": 5.547016425888514e-06,
+      "loss": 0.1778,
+      "step": 1125
+    },
+    {
+      "epoch": 0.48222698072805137,
+      "grad_norm": 2.421195589047659,
+      "learning_rate": 5.540119401740939e-06,
+      "loss": 0.1609,
+      "step": 1126
+    },
+    {
+      "epoch": 0.48265524625267664,
+      "grad_norm": 2.6159323795825298,
+      "learning_rate": 5.533221337585536e-06,
+      "loss": 0.1845,
+      "step": 1127
+    },
+    {
+      "epoch": 0.4830835117773019,
+      "grad_norm": 2.5042929575893593,
+      "learning_rate": 5.526322246704628e-06,
+      "loss": 0.1683,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4835117773019272,
+      "grad_norm": 2.6494484653472106,
+      "learning_rate": 5.519422142382514e-06,
+      "loss": 0.1894,
+      "step": 1129
+    },
+    {
+      "epoch": 0.48394004282655245,
+      "grad_norm": 2.78193828674508,
+      "learning_rate": 5.512521037905447e-06,
+      "loss": 0.1944,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4843683083511777,
+      "grad_norm": 2.5375907051181508,
+      "learning_rate": 5.505618946561608e-06,
+      "loss": 0.1539,
+      "step": 1131
+    },
+    {
+      "epoch": 0.484796573875803,
+      "grad_norm": 2.858337600460061,
+      "learning_rate": 5.498715881641069e-06,
+      "loss": 0.2162,
+      "step": 1132
+    },
+    {
+      "epoch": 0.48522483940042827,
+      "grad_norm": 2.6544062997270554,
+      "learning_rate": 5.491811856435788e-06,
+      "loss": 0.1755,
+      "step": 1133
+    },
+    {
+      "epoch": 0.48565310492505354,
+      "grad_norm": 2.0602263137702974,
+      "learning_rate": 5.484906884239564e-06,
+      "loss": 0.1313,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4860813704496788,
+      "grad_norm": 2.3988916353730128,
+      "learning_rate": 5.478000978348024e-06,
+      "loss": 0.1534,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4865096359743041,
+      "grad_norm": 2.438267193652537,
+      "learning_rate": 5.471094152058592e-06,
+      "loss": 0.1724,
+      "step": 1136
+    },
+    {
+      "epoch": 0.48693790149892935,
+      "grad_norm": 2.457965436585699,
+      "learning_rate": 5.464186418670458e-06,
+      "loss": 0.1765,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4873661670235546,
+      "grad_norm": 2.510408305943574,
+      "learning_rate": 5.45727779148457e-06,
+      "loss": 0.1697,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4877944325481799,
+      "grad_norm": 2.3849421744123553,
+      "learning_rate": 5.450368283803587e-06,
+      "loss": 0.1991,
+      "step": 1139
+    },
+    {
+      "epoch": 0.48822269807280516,
+      "grad_norm": 2.7367180822643142,
+      "learning_rate": 5.443457908931868e-06,
+      "loss": 0.2185,
+      "step": 1140
+    },
+    {
+      "epoch": 0.48865096359743043,
+      "grad_norm": 2.3992479693642195,
+      "learning_rate": 5.436546680175444e-06,
+      "loss": 0.1897,
+      "step": 1141
+    },
+    {
+      "epoch": 0.48907922912205565,
+      "grad_norm": 2.6034079861868458,
+      "learning_rate": 5.429634610841984e-06,
+      "loss": 0.1593,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4895074946466809,
+      "grad_norm": 2.5397318001774862,
+      "learning_rate": 5.4227217142407805e-06,
+      "loss": 0.1724,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4899357601713062,
+      "grad_norm": 2.514944449247344,
+      "learning_rate": 5.415808003682717e-06,
+      "loss": 0.1643,
+      "step": 1144
+    },
+    {
+      "epoch": 0.49036402569593146,
+      "grad_norm": 2.4625367408449033,
+      "learning_rate": 5.408893492480245e-06,
+      "loss": 0.1547,
+      "step": 1145
+    },
+    {
+      "epoch": 0.49079229122055673,
+      "grad_norm": 2.561437905856827,
+      "learning_rate": 5.401978193947357e-06,
+      "loss": 0.1922,
+      "step": 1146
+    },
+    {
+      "epoch": 0.491220556745182,
+      "grad_norm": 2.599706176150563,
+      "learning_rate": 5.39506212139956e-06,
+      "loss": 0.1769,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4916488222698073,
+      "grad_norm": 2.8666879903167293,
+      "learning_rate": 5.388145288153855e-06,
+      "loss": 0.1874,
+      "step": 1148
+    },
+    {
+      "epoch": 0.49207708779443254,
+      "grad_norm": 2.752340518072656,
+      "learning_rate": 5.381227707528705e-06,
+      "loss": 0.1801,
+      "step": 1149
+    },
+    {
+      "epoch": 0.4925053533190578,
+      "grad_norm": 2.4806394869717154,
+      "learning_rate": 5.374309392844014e-06,
+      "loss": 0.2078,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4929336188436831,
+      "grad_norm": 2.3019584337931027,
+      "learning_rate": 5.367390357421098e-06,
+      "loss": 0.145,
+      "step": 1151
+    },
+    {
+      "epoch": 0.49336188436830836,
+      "grad_norm": 2.653689764246234,
+      "learning_rate": 5.360470614582661e-06,
+      "loss": 0.1779,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4937901498929336,
+      "grad_norm": 2.598908257486413,
+      "learning_rate": 5.35355017765277e-06,
+      "loss": 0.1921,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4942184154175589,
+      "grad_norm": 2.064450121202305,
+      "learning_rate": 5.346629059956825e-06,
+      "loss": 0.144,
+      "step": 1154
+    },
+    {
+      "epoch": 0.49464668094218417,
+      "grad_norm": 2.4005021739626757,
+      "learning_rate": 5.339707274821543e-06,
+      "loss": 0.1698,
+      "step": 1155
+    },
+    {
+      "epoch": 0.49507494646680944,
+      "grad_norm": 2.1861012438086145,
+      "learning_rate": 5.33278483557492e-06,
+      "loss": 0.1566,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4955032119914347,
+      "grad_norm": 2.708877459882879,
+      "learning_rate": 5.325861755546216e-06,
+      "loss": 0.169,
+      "step": 1157
+    },
+    {
+      "epoch": 0.49593147751606,
+      "grad_norm": 2.8450759016756804,
+      "learning_rate": 5.318938048065926e-06,
+      "loss": 0.1735,
+      "step": 1158
+    },
+    {
+      "epoch": 0.4963597430406852,
+      "grad_norm": 2.506456334524834,
+      "learning_rate": 5.312013726465744e-06,
+      "loss": 0.1882,
+      "step": 1159
+    },
+    {
+      "epoch": 0.49678800856531047,
+      "grad_norm": 2.6284484254402867,
+      "learning_rate": 5.305088804078559e-06,
+      "loss": 0.1653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.49721627408993574,
+      "grad_norm": 3.658927290302993,
+      "learning_rate": 5.298163294238405e-06,
+      "loss": 0.228,
+      "step": 1161
+    },
+    {
+      "epoch": 0.497644539614561,
+      "grad_norm": 2.7409747948594654,
+      "learning_rate": 5.291237210280455e-06,
+      "loss": 0.1862,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4980728051391863,
+      "grad_norm": 2.344078890582227,
+      "learning_rate": 5.284310565540987e-06,
+      "loss": 0.1693,
+      "step": 1163
+    },
+    {
+      "epoch": 0.49850107066381155,
+      "grad_norm": 2.605315707968823,
+      "learning_rate": 5.277383373357353e-06,
+      "loss": 0.1537,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4989293361884368,
+      "grad_norm": 2.332721196882709,
+      "learning_rate": 5.270455647067968e-06,
+      "loss": 0.1554,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4993576017130621,
+      "grad_norm": 2.347690314271193,
+      "learning_rate": 5.263527400012268e-06,
+      "loss": 0.1816,
+      "step": 1166
+    },
+    {
+      "epoch": 0.49978586723768736,
+      "grad_norm": 2.1378218583151685,
+      "learning_rate": 5.256598645530695e-06,
+      "loss": 0.145,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5002141327623126,
+      "grad_norm": 2.045434281103789,
+      "learning_rate": 5.249669396964665e-06,
+      "loss": 0.1391,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5006423982869379,
+      "grad_norm": 2.4763051701138123,
+      "learning_rate": 5.242739667656551e-06,
+      "loss": 0.1971,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5010706638115632,
+      "grad_norm": 2.440445519730131,
+      "learning_rate": 5.235809470949647e-06,
+      "loss": 0.1678,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5014989293361884,
+      "grad_norm": 2.6851116213957518,
+      "learning_rate": 5.228878820188149e-06,
+      "loss": 0.168,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5019271948608137,
+      "grad_norm": 2.7800859080651383,
+      "learning_rate": 5.221947728717126e-06,
+      "loss": 0.1824,
+      "step": 1172
+    },
+    {
+      "epoch": 0.502355460385439,
+      "grad_norm": 2.5593960530193103,
+      "learning_rate": 5.215016209882496e-06,
+      "loss": 0.1604,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5027837259100643,
+      "grad_norm": 2.948123513057818,
+      "learning_rate": 5.208084277031001e-06,
+      "loss": 0.1845,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5032119914346895,
+      "grad_norm": 2.4832280399696804,
+      "learning_rate": 5.201151943510179e-06,
+      "loss": 0.1666,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5036402569593148,
+      "grad_norm": 2.5291104917942033,
+      "learning_rate": 5.1942192226683385e-06,
+      "loss": 0.1759,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5040685224839401,
+      "grad_norm": 2.4751714779879133,
+      "learning_rate": 5.187286127854538e-06,
+      "loss": 0.2053,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5044967880085653,
+      "grad_norm": 2.222204204983334,
+      "learning_rate": 5.180352672418553e-06,
+      "loss": 0.1376,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5049250535331906,
+      "grad_norm": 2.3227701191871515,
+      "learning_rate": 5.17341886971085e-06,
+      "loss": 0.1867,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5053533190578159,
+      "grad_norm": 2.2175322273479594,
+      "learning_rate": 5.166484733082572e-06,
+      "loss": 0.1412,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5057815845824412,
+      "grad_norm": 2.694886350241414,
+      "learning_rate": 5.1595502758854966e-06,
+      "loss": 0.1897,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5062098501070664,
+      "grad_norm": 2.4309920863940144,
+      "learning_rate": 5.1526155114720265e-06,
+      "loss": 0.1948,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5066381156316917,
+      "grad_norm": 2.5508118711021983,
+      "learning_rate": 5.145680453195151e-06,
+      "loss": 0.1824,
+      "step": 1183
+    },
+    {
+      "epoch": 0.507066381156317,
+      "grad_norm": 2.7447940944608025,
+      "learning_rate": 5.138745114408427e-06,
+      "loss": 0.1907,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5074946466809421,
+      "grad_norm": 2.729567538990291,
+      "learning_rate": 5.131809508465949e-06,
+      "loss": 0.1796,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5079229122055674,
+      "grad_norm": 2.7285883509842632,
+      "learning_rate": 5.124873648722329e-06,
+      "loss": 0.194,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5083511777301927,
+      "grad_norm": 2.5878819103256956,
+      "learning_rate": 5.1179375485326685e-06,
+      "loss": 0.1569,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5087794432548179,
+      "grad_norm": 2.53867217461566,
+      "learning_rate": 5.111001221252528e-06,
+      "loss": 0.1841,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5092077087794432,
+      "grad_norm": 2.5672236061934086,
+      "learning_rate": 5.10406468023791e-06,
+      "loss": 0.1525,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5096359743040685,
+      "grad_norm": 2.21342755835096,
+      "learning_rate": 5.097127938845222e-06,
+      "loss": 0.1415,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5100642398286938,
+      "grad_norm": 2.5125466438507074,
+      "learning_rate": 5.0901910104312645e-06,
+      "loss": 0.1837,
+      "step": 1191
+    },
+    {
+      "epoch": 0.510492505353319,
+      "grad_norm": 2.528879172143243,
+      "learning_rate": 5.083253908353193e-06,
+      "loss": 0.1831,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5109207708779443,
+      "grad_norm": 2.050502841827448,
+      "learning_rate": 5.0763166459684985e-06,
+      "loss": 0.1028,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5113490364025696,
+      "grad_norm": 2.394268591906985,
+      "learning_rate": 5.069379236634986e-06,
+      "loss": 0.1623,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5117773019271948,
+      "grad_norm": 2.5628991001460686,
+      "learning_rate": 5.062441693710735e-06,
+      "loss": 0.1658,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5122055674518201,
+      "grad_norm": 2.2872996970343484,
+      "learning_rate": 5.055504030554088e-06,
+      "loss": 0.208,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5126338329764454,
+      "grad_norm": 2.2393377203157856,
+      "learning_rate": 5.048566260523614e-06,
+      "loss": 0.1838,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5130620985010707,
+      "grad_norm": 2.325288217431444,
+      "learning_rate": 5.041628396978093e-06,
+      "loss": 0.1394,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5134903640256959,
+      "grad_norm": 2.324014527852947,
+      "learning_rate": 5.034690453276485e-06,
+      "loss": 0.1793,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5139186295503212,
+      "grad_norm": 2.4674036005329487,
+      "learning_rate": 5.0277524427778986e-06,
+      "loss": 0.1542,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5143468950749465,
+      "grad_norm": 2.3928410458985603,
+      "learning_rate": 5.020814378841579e-06,
+      "loss": 0.1606,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5147751605995717,
+      "grad_norm": 2.640829402744844,
+      "learning_rate": 5.013876274826866e-06,
+      "loss": 0.1639,
+      "step": 1202
+    },
+    {
+      "epoch": 0.515203426124197,
+      "grad_norm": 2.2210166314371764,
+      "learning_rate": 5.006938144093183e-06,
+      "loss": 0.1505,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5156316916488223,
+      "grad_norm": 2.318171291477901,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5160599571734475,
+      "grad_norm": 2.3493192691385327,
+      "learning_rate": 4.993061855906819e-06,
+      "loss": 0.1922,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5164882226980728,
+      "grad_norm": 2.399203523529749,
+      "learning_rate": 4.986123725173136e-06,
+      "loss": 0.154,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5169164882226981,
+      "grad_norm": 2.7588946588597842,
+      "learning_rate": 4.979185621158423e-06,
+      "loss": 0.1843,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5173447537473234,
+      "grad_norm": 2.468021218458666,
+      "learning_rate": 4.972247557222102e-06,
+      "loss": 0.1311,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5177730192719486,
+      "grad_norm": 2.2195503600893343,
+      "learning_rate": 4.965309546723516e-06,
+      "loss": 0.1275,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5182012847965739,
+      "grad_norm": 2.3828311604994723,
+      "learning_rate": 4.958371603021908e-06,
+      "loss": 0.139,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5186295503211992,
+      "grad_norm": 2.1685830908392685,
+      "learning_rate": 4.951433739476388e-06,
+      "loss": 0.1755,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5190578158458244,
+      "grad_norm": 2.434866802895821,
+      "learning_rate": 4.944495969445914e-06,
+      "loss": 0.1497,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5194860813704497,
+      "grad_norm": 3.0812692922108575,
+      "learning_rate": 4.937558306289266e-06,
+      "loss": 0.2098,
+      "step": 1213
+    },
+    {
+      "epoch": 0.519914346895075,
+      "grad_norm": 2.220740386822536,
+      "learning_rate": 4.930620763365014e-06,
+      "loss": 0.2369,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5203426124197003,
+      "grad_norm": 2.587454223653688,
+      "learning_rate": 4.923683354031501e-06,
+      "loss": 0.1589,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5207708779443255,
+      "grad_norm": 2.174826071621821,
+      "learning_rate": 4.916746091646808e-06,
+      "loss": 0.1346,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5211991434689508,
+      "grad_norm": 2.1441594506497554,
+      "learning_rate": 4.909808989568737e-06,
+      "loss": 0.2052,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5216274089935761,
+      "grad_norm": 2.370418277698325,
+      "learning_rate": 4.902872061154779e-06,
+      "loss": 0.1356,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5220556745182013,
+      "grad_norm": 2.4298081543676906,
+      "learning_rate": 4.895935319762091e-06,
+      "loss": 0.1641,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5224839400428265,
+      "grad_norm": 2.5665936977122916,
+      "learning_rate": 4.8889987787474716e-06,
+      "loss": 0.1853,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5229122055674518,
+      "grad_norm": 2.423206893727968,
+      "learning_rate": 4.882062451467331e-06,
+      "loss": 0.1691,
+      "step": 1221
+    },
+    {
+      "epoch": 0.523340471092077,
+      "grad_norm": 2.7315741812459726,
+      "learning_rate": 4.875126351277672e-06,
+      "loss": 0.1797,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5237687366167023,
+      "grad_norm": 2.5363513467147403,
+      "learning_rate": 4.868190491534054e-06,
+      "loss": 0.154,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5241970021413276,
+      "grad_norm": 2.3403266428925615,
+      "learning_rate": 4.8612548855915755e-06,
+      "loss": 0.1725,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5246252676659529,
+      "grad_norm": 2.0820801550715333,
+      "learning_rate": 4.854319546804851e-06,
+      "loss": 0.1595,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5250535331905781,
+      "grad_norm": 2.653891221679795,
+      "learning_rate": 4.847384488527975e-06,
+      "loss": 0.1892,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5254817987152034,
+      "grad_norm": 2.3764751467189043,
+      "learning_rate": 4.840449724114505e-06,
+      "loss": 0.1462,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5259100642398287,
+      "grad_norm": 2.5257188604763483,
+      "learning_rate": 4.833515266917431e-06,
+      "loss": 0.1486,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5263383297644539,
+      "grad_norm": 2.2267362317281445,
+      "learning_rate": 4.826581130289151e-06,
+      "loss": 0.1257,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5267665952890792,
+      "grad_norm": 2.4329794243726575,
+      "learning_rate": 4.81964732758145e-06,
+      "loss": 0.1411,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5271948608137045,
+      "grad_norm": 2.446723479687642,
+      "learning_rate": 4.812713872145463e-06,
+      "loss": 0.1609,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5276231263383298,
+      "grad_norm": 2.2546916563483617,
+      "learning_rate": 4.805780777331662e-06,
+      "loss": 0.1497,
+      "step": 1232
+    },
+    {
+      "epoch": 0.528051391862955,
+      "grad_norm": 2.270605077044052,
+      "learning_rate": 4.798848056489823e-06,
+      "loss": 0.146,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5284796573875803,
+      "grad_norm": 2.3513584721452023,
+      "learning_rate": 4.791915722969e-06,
+      "loss": 0.1568,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5289079229122056,
+      "grad_norm": 2.2635958411104227,
+      "learning_rate": 4.784983790117505e-06,
+      "loss": 0.144,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5293361884368308,
+      "grad_norm": 2.1697840733580076,
+      "learning_rate": 4.778052271282875e-06,
+      "loss": 0.1169,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5297644539614561,
+      "grad_norm": 2.4962776280701053,
+      "learning_rate": 4.771121179811852e-06,
+      "loss": 0.1405,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5301927194860814,
+      "grad_norm": 3.571171963746266,
+      "learning_rate": 4.7641905290503535e-06,
+      "loss": 0.2297,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5306209850107066,
+      "grad_norm": 2.7732914448785055,
+      "learning_rate": 4.7572603323434505e-06,
+      "loss": 0.1485,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5310492505353319,
+      "grad_norm": 2.635478084136686,
+      "learning_rate": 4.750330603035336e-06,
+      "loss": 0.153,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5314775160599572,
+      "grad_norm": 2.4754492688007255,
+      "learning_rate": 4.743401354469307e-06,
+      "loss": 0.1407,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5319057815845825,
+      "grad_norm": 2.4976945357068723,
+      "learning_rate": 4.736472599987733e-06,
+      "loss": 0.1617,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5323340471092077,
+      "grad_norm": 2.3719337706254886,
+      "learning_rate": 4.729544352932033e-06,
+      "loss": 0.1782,
+      "step": 1243
+    },
+    {
+      "epoch": 0.532762312633833,
+      "grad_norm": 2.4197852198757857,
+      "learning_rate": 4.722616626642648e-06,
+      "loss": 0.1865,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5331905781584583,
+      "grad_norm": 2.5681246000181455,
+      "learning_rate": 4.715689434459016e-06,
+      "loss": 0.1267,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5336188436830835,
+      "grad_norm": 2.2278885351176334,
+      "learning_rate": 4.708762789719547e-06,
+      "loss": 0.1372,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5340471092077088,
+      "grad_norm": 2.315238569831459,
+      "learning_rate": 4.701836705761598e-06,
+      "loss": 0.1826,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5344753747323341,
+      "grad_norm": 2.562435007821753,
+      "learning_rate": 4.694911195921443e-06,
+      "loss": 0.1921,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5349036402569594,
+      "grad_norm": 2.254287508109515,
+      "learning_rate": 4.687986273534257e-06,
+      "loss": 0.137,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5353319057815846,
+      "grad_norm": 2.340022587464616,
+      "learning_rate": 4.681061951934075e-06,
+      "loss": 0.1469,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5357601713062099,
+      "grad_norm": 2.562080716989251,
+      "learning_rate": 4.674138244453783e-06,
+      "loss": 0.231,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5361884368308352,
+      "grad_norm": 2.4282521111174145,
+      "learning_rate": 4.66721516442508e-06,
+      "loss": 0.1298,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5366167023554604,
+      "grad_norm": 2.339866742638222,
+      "learning_rate": 4.660292725178459e-06,
+      "loss": 0.184,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5370449678800856,
+      "grad_norm": 2.6437928338396945,
+      "learning_rate": 4.653370940043175e-06,
+      "loss": 0.158,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5374732334047109,
+      "grad_norm": 2.284535673788142,
+      "learning_rate": 4.646449822347231e-06,
+      "loss": 0.1166,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5379014989293361,
+      "grad_norm": 2.467740789844648,
+      "learning_rate": 4.6395293854173395e-06,
+      "loss": 0.1383,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5383297644539614,
+      "grad_norm": 2.8151800579752435,
+      "learning_rate": 4.632609642578902e-06,
+      "loss": 0.2095,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5387580299785867,
+      "grad_norm": 2.7590500558038875,
+      "learning_rate": 4.625690607155987e-06,
+      "loss": 0.1744,
+      "step": 1258
+    },
+    {
+      "epoch": 0.539186295503212,
+      "grad_norm": 2.3366027261722975,
+      "learning_rate": 4.618772292471297e-06,
+      "loss": 0.1559,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5396145610278372,
+      "grad_norm": 2.626126247879457,
+      "learning_rate": 4.611854711846147e-06,
+      "loss": 0.2025,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5400428265524625,
+      "grad_norm": 2.2848318064154096,
+      "learning_rate": 4.604937878600443e-06,
+      "loss": 0.1617,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5404710920770878,
+      "grad_norm": 2.0858183266543366,
+      "learning_rate": 4.598021806052646e-06,
+      "loss": 0.1473,
+      "step": 1262
+    },
+    {
+      "epoch": 0.540899357601713,
+      "grad_norm": 2.695806608721385,
+      "learning_rate": 4.591106507519756e-06,
+      "loss": 0.1764,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5413276231263383,
+      "grad_norm": 2.659258741990666,
+      "learning_rate": 4.584191996317285e-06,
+      "loss": 0.1815,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5417558886509636,
+      "grad_norm": 2.3365048539112583,
+      "learning_rate": 4.577278285759221e-06,
+      "loss": 0.1242,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5421841541755889,
+      "grad_norm": 2.016313378494844,
+      "learning_rate": 4.570365389158018e-06,
+      "loss": 0.1443,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5426124197002141,
+      "grad_norm": 2.4532449054992336,
+      "learning_rate": 4.563453319824558e-06,
+      "loss": 0.1607,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5430406852248394,
+      "grad_norm": 2.2393326498715878,
+      "learning_rate": 4.5565420910681334e-06,
+      "loss": 0.1283,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5434689507494647,
+      "grad_norm": 2.4648117704206456,
+      "learning_rate": 4.549631716196415e-06,
+      "loss": 0.1551,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5438972162740899,
+      "grad_norm": 2.6990505883020326,
+      "learning_rate": 4.542722208515432e-06,
+      "loss": 0.167,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5443254817987152,
+      "grad_norm": 2.370486564623986,
+      "learning_rate": 4.535813581329543e-06,
+      "loss": 0.1271,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5447537473233405,
+      "grad_norm": 2.550095462872023,
+      "learning_rate": 4.528905847941411e-06,
+      "loss": 0.1407,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5451820128479657,
+      "grad_norm": 1.9557049637666493,
+      "learning_rate": 4.521999021651977e-06,
+      "loss": 0.1106,
+      "step": 1273
+    },
+    {
+      "epoch": 0.545610278372591,
+      "grad_norm": 2.6053184527450326,
+      "learning_rate": 4.515093115760436e-06,
+      "loss": 0.1617,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5460385438972163,
+      "grad_norm": 2.303789466773574,
+      "learning_rate": 4.508188143564214e-06,
+      "loss": 0.1952,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5464668094218416,
+      "grad_norm": 2.3095581898523756,
+      "learning_rate": 4.501284118358932e-06,
+      "loss": 0.137,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5468950749464668,
+      "grad_norm": 2.619520340057349,
+      "learning_rate": 4.494381053438394e-06,
+      "loss": 0.1692,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5473233404710921,
+      "grad_norm": 2.45156529443492,
+      "learning_rate": 4.487478962094554e-06,
+      "loss": 0.1475,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5477516059957174,
+      "grad_norm": 2.336844328682401,
+      "learning_rate": 4.480577857617487e-06,
+      "loss": 0.1518,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5481798715203426,
+      "grad_norm": 3.0047218223745715,
+      "learning_rate": 4.473677753295375e-06,
+      "loss": 0.1714,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5486081370449679,
+      "grad_norm": 2.5773678664774637,
+      "learning_rate": 4.466778662414465e-06,
+      "loss": 0.167,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5490364025695932,
+      "grad_norm": 2.5195341900864374,
+      "learning_rate": 4.459880598259062e-06,
+      "loss": 0.1761,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5494646680942185,
+      "grad_norm": 2.404236305433376,
+      "learning_rate": 4.452983574111488e-06,
+      "loss": 0.1464,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5498929336188437,
+      "grad_norm": 2.426258324338087,
+      "learning_rate": 4.446087603252063e-06,
+      "loss": 0.1669,
+      "step": 1284
+    },
+    {
+      "epoch": 0.550321199143469,
+      "grad_norm": 2.57556843680859,
+      "learning_rate": 4.439192698959082e-06,
+      "loss": 0.1417,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5507494646680943,
+      "grad_norm": 2.728380395424712,
+      "learning_rate": 4.432298874508783e-06,
+      "loss": 0.1365,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5511777301927195,
+      "grad_norm": 2.3942584160281437,
+      "learning_rate": 4.425406143175327e-06,
+      "loss": 0.1486,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5516059957173447,
+      "grad_norm": 2.5722394410080596,
+      "learning_rate": 4.418514518230769e-06,
+      "loss": 0.2373,
+      "step": 1288
+    },
+    {
+      "epoch": 0.55203426124197,
+      "grad_norm": 2.630923320265094,
+      "learning_rate": 4.4116240129450335e-06,
+      "loss": 0.1831,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5524625267665952,
+      "grad_norm": 2.3221034675872043,
+      "learning_rate": 4.40473464058589e-06,
+      "loss": 0.1609,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5528907922912205,
+      "grad_norm": 2.173838297353612,
+      "learning_rate": 4.397846414418925e-06,
+      "loss": 0.1472,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5533190578158458,
+      "grad_norm": 2.228906860017907,
+      "learning_rate": 4.390959347707521e-06,
+      "loss": 0.1482,
+      "step": 1292
+    },
+    {
+      "epoch": 0.553747323340471,
+      "grad_norm": 2.2965550003848323,
+      "learning_rate": 4.384073453712825e-06,
+      "loss": 0.2063,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5541755888650963,
+      "grad_norm": 2.3905780255947566,
+      "learning_rate": 4.377188745693725e-06,
+      "loss": 0.1507,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5546038543897216,
+      "grad_norm": 2.302275509802097,
+      "learning_rate": 4.370305236906831e-06,
+      "loss": 0.1468,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5550321199143469,
+      "grad_norm": 2.260050251285635,
+      "learning_rate": 4.363422940606435e-06,
+      "loss": 0.1578,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5554603854389721,
+      "grad_norm": 2.2261941557536202,
+      "learning_rate": 4.356541870044505e-06,
+      "loss": 0.1661,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5558886509635974,
+      "grad_norm": 2.922876412855788,
+      "learning_rate": 4.349662038470639e-06,
+      "loss": 0.1881,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5563169164882227,
+      "grad_norm": 2.0893114901094547,
+      "learning_rate": 4.342783459132055e-06,
+      "loss": 0.1516,
+      "step": 1299
+    },
+    {
+      "epoch": 0.556745182012848,
+      "grad_norm": 2.7475029821646233,
+      "learning_rate": 4.33590614527356e-06,
+      "loss": 0.1457,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5571734475374732,
+      "grad_norm": 2.701629922060087,
+      "learning_rate": 4.329030110137521e-06,
+      "loss": 0.1615,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5576017130620985,
+      "grad_norm": 2.4475140609574684,
+      "learning_rate": 4.322155366963845e-06,
+      "loss": 0.1337,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5580299785867238,
+      "grad_norm": 2.2002437100159113,
+      "learning_rate": 4.315281928989949e-06,
+      "loss": 0.1333,
+      "step": 1303
+    },
+    {
+      "epoch": 0.558458244111349,
+      "grad_norm": 3.183510098652473,
+      "learning_rate": 4.308409809450742e-06,
+      "loss": 0.1741,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5588865096359743,
+      "grad_norm": 2.4127213814294066,
+      "learning_rate": 4.301539021578588e-06,
+      "loss": 0.1905,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5593147751605996,
+      "grad_norm": 2.522099273884657,
+      "learning_rate": 4.294669578603292e-06,
+      "loss": 0.1609,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5597430406852248,
+      "grad_norm": 2.1400835933796714,
+      "learning_rate": 4.287801493752066e-06,
+      "loss": 0.1266,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5601713062098501,
+      "grad_norm": 2.150951919725811,
+      "learning_rate": 4.280934780249508e-06,
+      "loss": 0.1198,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5605995717344754,
+      "grad_norm": 2.050725459573684,
+      "learning_rate": 4.274069451317577e-06,
+      "loss": 0.1285,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5610278372591007,
+      "grad_norm": 2.4441530657708017,
+      "learning_rate": 4.267205520175562e-06,
+      "loss": 0.1282,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5614561027837259,
+      "grad_norm": 2.3549999970726794,
+      "learning_rate": 4.260343000040065e-06,
+      "loss": 0.1596,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5618843683083512,
+      "grad_norm": 2.1856226566935604,
+      "learning_rate": 4.253481904124968e-06,
+      "loss": 0.1416,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5623126338329765,
+      "grad_norm": 2.6582165240445534,
+      "learning_rate": 4.246622245641413e-06,
+      "loss": 0.1716,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5627408993576017,
+      "grad_norm": 2.2305415602296526,
+      "learning_rate": 4.239764037797773e-06,
+      "loss": 0.1343,
+      "step": 1314
+    },
+    {
+      "epoch": 0.563169164882227,
+      "grad_norm": 2.5609879233691104,
+      "learning_rate": 4.232907293799627e-06,
+      "loss": 0.1679,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5635974304068523,
+      "grad_norm": 3.011655847021986,
+      "learning_rate": 4.226052026849737e-06,
+      "loss": 0.1987,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5640256959314776,
+      "grad_norm": 2.551637158609248,
+      "learning_rate": 4.219198250148018e-06,
+      "loss": 0.1393,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5644539614561028,
+      "grad_norm": 2.619326583571539,
+      "learning_rate": 4.21234597689152e-06,
+      "loss": 0.1519,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5648822269807281,
+      "grad_norm": 2.294226133594756,
+      "learning_rate": 4.205495220274394e-06,
+      "loss": 0.1524,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5653104925053534,
+      "grad_norm": 2.512790837899654,
+      "learning_rate": 4.198645993487872e-06,
+      "loss": 0.1272,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5657387580299786,
+      "grad_norm": 2.984998432530334,
+      "learning_rate": 4.191798309720244e-06,
+      "loss": 0.2155,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5661670235546039,
+      "grad_norm": 2.3707929563093555,
+      "learning_rate": 4.184952182156821e-06,
+      "loss": 0.1491,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5665952890792291,
+      "grad_norm": 2.2104882053322057,
+      "learning_rate": 4.178107623979926e-06,
+      "loss": 0.1304,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5670235546038543,
+      "grad_norm": 2.630968556500284,
+      "learning_rate": 4.171264648368852e-06,
+      "loss": 0.1523,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5674518201284796,
+      "grad_norm": 2.447706278352564,
+      "learning_rate": 4.164423268499853e-06,
+      "loss": 0.1432,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5678800856531049,
+      "grad_norm": 2.3173614285530495,
+      "learning_rate": 4.157583497546103e-06,
+      "loss": 0.1331,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5683083511777302,
+      "grad_norm": 2.404183474327912,
+      "learning_rate": 4.150745348677683e-06,
+      "loss": 0.1744,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5687366167023554,
+      "grad_norm": 2.90735100730342,
+      "learning_rate": 4.143908835061551e-06,
+      "loss": 0.1654,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5691648822269807,
+      "grad_norm": 2.0742371078614723,
+      "learning_rate": 4.13707396986151e-06,
+      "loss": 0.1624,
+      "step": 1329
+    },
+    {
+      "epoch": 0.569593147751606,
+      "grad_norm": 2.2566341952147564,
+      "learning_rate": 4.130240766238195e-06,
+      "loss": 0.1611,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5700214132762312,
+      "grad_norm": 2.2387896492813995,
+      "learning_rate": 4.12340923734904e-06,
+      "loss": 0.1327,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5704496788008565,
+      "grad_norm": 2.4598580775606433,
+      "learning_rate": 4.116579396348253e-06,
+      "loss": 0.1262,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5708779443254818,
+      "grad_norm": 2.5682215820476686,
+      "learning_rate": 4.1097512563867944e-06,
+      "loss": 0.1601,
+      "step": 1333
+    },
+    {
+      "epoch": 0.571306209850107,
+      "grad_norm": 2.453810287570149,
+      "learning_rate": 4.102924830612348e-06,
+      "loss": 0.139,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5717344753747323,
+      "grad_norm": 2.0200320946403263,
+      "learning_rate": 4.0961001321692945e-06,
+      "loss": 0.1077,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5721627408993576,
+      "grad_norm": 2.304582777966451,
+      "learning_rate": 4.089277174198694e-06,
+      "loss": 0.1498,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5725910064239829,
+      "grad_norm": 1.9219419879730721,
+      "learning_rate": 4.082455969838249e-06,
+      "loss": 0.1871,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5730192719486081,
+      "grad_norm": 2.5628085523915933,
+      "learning_rate": 4.0756365322222915e-06,
+      "loss": 0.1471,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5734475374732334,
+      "grad_norm": 2.4930678949514564,
+      "learning_rate": 4.068818874481746e-06,
+      "loss": 0.1577,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5738758029978587,
+      "grad_norm": 2.656215758731573,
+      "learning_rate": 4.062003009744115e-06,
+      "loss": 0.1399,
+      "step": 1340
+    },
+    {
+      "epoch": 0.574304068522484,
+      "grad_norm": 2.584560690603714,
+      "learning_rate": 4.055188951133443e-06,
+      "loss": 0.1651,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5747323340471092,
+      "grad_norm": 2.5387708734708756,
+      "learning_rate": 4.048376711770302e-06,
+      "loss": 0.1189,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5751605995717345,
+      "grad_norm": 2.8490429919971905,
+      "learning_rate": 4.04156630477176e-06,
+      "loss": 0.1864,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5755888650963598,
+      "grad_norm": 2.3398358315970698,
+      "learning_rate": 4.0347577432513515e-06,
+      "loss": 0.1209,
+      "step": 1344
+    },
+    {
+      "epoch": 0.576017130620985,
+      "grad_norm": 2.1709534422980163,
+      "learning_rate": 4.027951040319065e-06,
+      "loss": 0.1865,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5764453961456103,
+      "grad_norm": 2.4444620405780926,
+      "learning_rate": 4.0211462090813056e-06,
+      "loss": 0.1487,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5768736616702356,
+      "grad_norm": 2.2552403672780104,
+      "learning_rate": 4.0143432626408756e-06,
+      "loss": 0.1238,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5773019271948608,
+      "grad_norm": 2.4573545951103806,
+      "learning_rate": 4.007542214096947e-06,
+      "loss": 0.1548,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5777301927194861,
+      "grad_norm": 2.3830257187975876,
+      "learning_rate": 4.0007430765450425e-06,
+      "loss": 0.1344,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5781584582441114,
+      "grad_norm": 2.3405710352469518,
+      "learning_rate": 3.993945863076996e-06,
+      "loss": 0.1272,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5785867237687367,
+      "grad_norm": 2.300720025904557,
+      "learning_rate": 3.987150586780945e-06,
+      "loss": 0.1296,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5790149892933619,
+      "grad_norm": 2.3105703637069563,
+      "learning_rate": 3.980357260741293e-06,
+      "loss": 0.1173,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5794432548179872,
+      "grad_norm": 2.4433433983270807,
+      "learning_rate": 3.973565898038688e-06,
+      "loss": 0.1398,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5798715203426125,
+      "grad_norm": 2.326648933726636,
+      "learning_rate": 3.96677651175e-06,
+      "loss": 0.1248,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5802997858672377,
+      "grad_norm": 2.688668964075498,
+      "learning_rate": 3.95998911494829e-06,
+      "loss": 0.1845,
+      "step": 1355
+    },
+    {
+      "epoch": 0.580728051391863,
+      "grad_norm": 2.3043536178790673,
+      "learning_rate": 3.953203720702793e-06,
+      "loss": 0.1388,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5811563169164882,
+      "grad_norm": 2.4867504649514314,
+      "learning_rate": 3.946420342078882e-06,
+      "loss": 0.154,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5815845824411134,
+      "grad_norm": 2.506470590411309,
+      "learning_rate": 3.939638992138053e-06,
+      "loss": 0.1536,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5820128479657387,
+      "grad_norm": 2.593084499904981,
+      "learning_rate": 3.9328596839378955e-06,
+      "loss": 0.1495,
+      "step": 1359
+    },
+    {
+      "epoch": 0.582441113490364,
+      "grad_norm": 2.1422920902380604,
+      "learning_rate": 3.926082430532067e-06,
+      "loss": 0.1776,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5828693790149893,
+      "grad_norm": 2.2094012440787147,
+      "learning_rate": 3.919307244970267e-06,
+      "loss": 0.1652,
+      "step": 1361
+    },
+    {
+      "epoch": 0.5832976445396145,
+      "grad_norm": 2.755182120835082,
+      "learning_rate": 3.912534140298216e-06,
+      "loss": 0.199,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5837259100642398,
+      "grad_norm": 2.533180454142921,
+      "learning_rate": 3.905763129557625e-06,
+      "loss": 0.1469,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5841541755888651,
+      "grad_norm": 2.7387510944035456,
+      "learning_rate": 3.898994225786178e-06,
+      "loss": 0.1583,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5845824411134903,
+      "grad_norm": 2.5333719602003524,
+      "learning_rate": 3.892227442017493e-06,
+      "loss": 0.1357,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5850107066381156,
+      "grad_norm": 2.395216678832063,
+      "learning_rate": 3.885462791281114e-06,
+      "loss": 0.1114,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5854389721627409,
+      "grad_norm": 2.4870279765476746,
+      "learning_rate": 3.878700286602476e-06,
+      "loss": 0.1582,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5858672376873661,
+      "grad_norm": 2.322942952072667,
+      "learning_rate": 3.87193994100288e-06,
+      "loss": 0.1216,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5862955032119914,
+      "grad_norm": 2.410413516730318,
+      "learning_rate": 3.865181767499471e-06,
+      "loss": 0.1634,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5867237687366167,
+      "grad_norm": 2.3569293978139214,
+      "learning_rate": 3.858425779105211e-06,
+      "loss": 0.1701,
+      "step": 1370
+    },
+    {
+      "epoch": 0.587152034261242,
+      "grad_norm": 2.4243804166228684,
+      "learning_rate": 3.851671988828854e-06,
+      "loss": 0.1742,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5875802997858672,
+      "grad_norm": 1.951844729954393,
+      "learning_rate": 3.8449204096749235e-06,
+      "loss": 0.096,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5880085653104925,
+      "grad_norm": 2.0859312509600927,
+      "learning_rate": 3.838171054643683e-06,
+      "loss": 0.1046,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5884368308351178,
+      "grad_norm": 2.2698754952198557,
+      "learning_rate": 3.831423936731117e-06,
+      "loss": 0.1345,
+      "step": 1374
+    },
+    {
+      "epoch": 0.588865096359743,
+      "grad_norm": 2.115005586730726,
+      "learning_rate": 3.824679068928897e-06,
+      "loss": 0.1343,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5892933618843683,
+      "grad_norm": 2.404559206638727,
+      "learning_rate": 3.817936464224367e-06,
+      "loss": 0.1331,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5897216274089936,
+      "grad_norm": 2.38055571645337,
+      "learning_rate": 3.8111961356005088e-06,
+      "loss": 0.1241,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5901498929336189,
+      "grad_norm": 2.1577492643744,
+      "learning_rate": 3.804458096035928e-06,
+      "loss": 0.1305,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5905781584582441,
+      "grad_norm": 2.16254552351315,
+      "learning_rate": 3.797722358504814e-06,
+      "loss": 0.1468,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5910064239828694,
+      "grad_norm": 2.6172600034583757,
+      "learning_rate": 3.79098893597693e-06,
+      "loss": 0.2029,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5914346895074947,
+      "grad_norm": 2.1578993087741485,
+      "learning_rate": 3.7842578414175824e-06,
+      "loss": 0.1296,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5918629550321199,
+      "grad_norm": 2.078386673539379,
+      "learning_rate": 3.777529087787588e-06,
+      "loss": 0.1413,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5922912205567452,
+      "grad_norm": 2.206480438268327,
+      "learning_rate": 3.7708026880432647e-06,
+      "loss": 0.13,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5927194860813705,
+      "grad_norm": 2.55298114056071,
+      "learning_rate": 3.764078655136391e-06,
+      "loss": 0.1337,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5931477516059958,
+      "grad_norm": 2.594153255732443,
+      "learning_rate": 3.7573570020141924e-06,
+      "loss": 0.1274,
+      "step": 1385
+    },
+    {
+      "epoch": 0.593576017130621,
+      "grad_norm": 2.496859284981384,
+      "learning_rate": 3.7506377416193096e-06,
+      "loss": 0.1396,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5940042826552463,
+      "grad_norm": 2.065812436011103,
+      "learning_rate": 3.7439208868897768e-06,
+      "loss": 0.1336,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5944325481798716,
+      "grad_norm": 2.1393695522627603,
+      "learning_rate": 3.737206450758999e-06,
+      "loss": 0.143,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5948608137044968,
+      "grad_norm": 2.727108323675969,
+      "learning_rate": 3.7304944461557173e-06,
+      "loss": 0.1372,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5952890792291221,
+      "grad_norm": 1.9682164029103848,
+      "learning_rate": 3.7237848860039983e-06,
+      "loss": 0.1256,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5957173447537473,
+      "grad_norm": 2.553738340403351,
+      "learning_rate": 3.717077783223197e-06,
+      "loss": 0.1432,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5961456102783725,
+      "grad_norm": 2.3326529412155828,
+      "learning_rate": 3.7103731507279383e-06,
+      "loss": 0.1184,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5965738758029978,
+      "grad_norm": 5.563084320011895,
+      "learning_rate": 3.703671001428091e-06,
+      "loss": 0.1273,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5970021413276231,
+      "grad_norm": 2.4479898028614575,
+      "learning_rate": 3.69697134822874e-06,
+      "loss": 0.1165,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5974304068522484,
+      "grad_norm": 2.2518827099082337,
+      "learning_rate": 3.690274204030169e-06,
+      "loss": 0.1136,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5978586723768736,
+      "grad_norm": 2.4301075042737783,
+      "learning_rate": 3.683579581727824e-06,
+      "loss": 0.1643,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5982869379014989,
+      "grad_norm": 2.5208375179990865,
+      "learning_rate": 3.6768874942123012e-06,
+      "loss": 0.1659,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5987152034261242,
+      "grad_norm": 2.0951464477527453,
+      "learning_rate": 3.670197954369311e-06,
+      "loss": 0.1181,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5991434689507494,
+      "grad_norm": 2.3464556066053834,
+      "learning_rate": 3.6635109750796637e-06,
+      "loss": 0.2021,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5995717344753747,
+      "grad_norm": 2.149138554165214,
+      "learning_rate": 3.656826569219233e-06,
+      "loss": 0.1437,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4916727249052975,
+      "learning_rate": 3.650144749658939e-06,
+      "loss": 0.1281,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6004282655246252,
+      "grad_norm": 2.3577187824548607,
+      "learning_rate": 3.643465529264727e-06,
+      "loss": 0.1405,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6008565310492505,
+      "grad_norm": 2.239449104765556,
+      "learning_rate": 3.6367889208975327e-06,
+      "loss": 0.1301,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6012847965738758,
+      "grad_norm": 2.388226985221442,
+      "learning_rate": 3.6301149374132615e-06,
+      "loss": 0.1281,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6017130620985011,
+      "grad_norm": 2.960945240587159,
+      "learning_rate": 3.6234435916627696e-06,
+      "loss": 0.16,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6021413276231263,
+      "grad_norm": 2.229562610702544,
+      "learning_rate": 3.6167748964918282e-06,
+      "loss": 0.1444,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6025695931477516,
+      "grad_norm": 3.0305482760540388,
+      "learning_rate": 3.6101088647411097e-06,
+      "loss": 0.1654,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6029978586723769,
+      "grad_norm": 2.1326270451056057,
+      "learning_rate": 3.603445509246154e-06,
+      "loss": 0.1153,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6034261241970021,
+      "grad_norm": 2.067967423214708,
+      "learning_rate": 3.5967848428373512e-06,
+      "loss": 0.1383,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6038543897216274,
+      "grad_norm": 2.2836810814579787,
+      "learning_rate": 3.5901268783399135e-06,
+      "loss": 0.1366,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6042826552462527,
+      "grad_norm": 2.6287494862987604,
+      "learning_rate": 3.583471628573846e-06,
+      "loss": 0.1434,
+      "step": 1411
+    },
+    {
+      "epoch": 0.604710920770878,
+      "grad_norm": 2.5374128414286194,
+      "learning_rate": 3.5768191063539326e-06,
+      "loss": 0.1304,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6051391862955032,
+      "grad_norm": 2.2476745529691855,
+      "learning_rate": 3.5701693244897005e-06,
+      "loss": 0.1598,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6055674518201285,
+      "grad_norm": 2.284166364945164,
+      "learning_rate": 3.563522295785403e-06,
+      "loss": 0.1373,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6059957173447538,
+      "grad_norm": 2.6016190373765324,
+      "learning_rate": 3.556878033039991e-06,
+      "loss": 0.1102,
+      "step": 1415
+    },
+    {
+      "epoch": 0.606423982869379,
+      "grad_norm": 2.3484020080934256,
+      "learning_rate": 3.55023654904709e-06,
+      "loss": 0.1191,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6068522483940043,
+      "grad_norm": 2.5639012259015574,
+      "learning_rate": 3.543597856594976e-06,
+      "loss": 0.138,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6072805139186296,
+      "grad_norm": 2.3969687008646052,
+      "learning_rate": 3.5369619684665456e-06,
+      "loss": 0.1626,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6077087794432549,
+      "grad_norm": 2.181074231393954,
+      "learning_rate": 3.530328897439302e-06,
+      "loss": 0.1262,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6081370449678801,
+      "grad_norm": 3.2304137710594047,
+      "learning_rate": 3.5236986562853193e-06,
+      "loss": 0.2178,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6085653104925054,
+      "grad_norm": 1.8766895205643324,
+      "learning_rate": 3.517071257771225e-06,
+      "loss": 0.0918,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6089935760171307,
+      "grad_norm": 2.1925731213079853,
+      "learning_rate": 3.51044671465817e-06,
+      "loss": 0.1496,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6094218415417559,
+      "grad_norm": 2.6088861814998823,
+      "learning_rate": 3.50382503970181e-06,
+      "loss": 0.1485,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6098501070663812,
+      "grad_norm": 2.1868850387126493,
+      "learning_rate": 3.49720624565228e-06,
+      "loss": 0.1174,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6102783725910065,
+      "grad_norm": 2.235683860698061,
+      "learning_rate": 3.4905903452541612e-06,
+      "loss": 0.1266,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6107066381156316,
+      "grad_norm": 2.7724357243056854,
+      "learning_rate": 3.4839773512464693e-06,
+      "loss": 0.2125,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6111349036402569,
+      "grad_norm": 2.042974718790244,
+      "learning_rate": 3.477367276362619e-06,
+      "loss": 0.1089,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6115631691648822,
+      "grad_norm": 2.1601939893669555,
+      "learning_rate": 3.4707601333304093e-06,
+      "loss": 0.1532,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6119914346895075,
+      "grad_norm": 2.020064821812364,
+      "learning_rate": 3.4641559348719885e-06,
+      "loss": 0.1173,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6124197002141327,
+      "grad_norm": 2.9005360093279497,
+      "learning_rate": 3.4575546937038387e-06,
+      "loss": 0.1606,
+      "step": 1430
+    },
+    {
+      "epoch": 0.612847965738758,
+      "grad_norm": 2.172697422469941,
+      "learning_rate": 3.450956422536747e-06,
+      "loss": 0.1004,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6132762312633833,
+      "grad_norm": 2.519462157281068,
+      "learning_rate": 3.44436113407578e-06,
+      "loss": 0.1418,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6137044967880085,
+      "grad_norm": 2.058211812912002,
+      "learning_rate": 3.4377688410202634e-06,
+      "loss": 0.0955,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6141327623126338,
+      "grad_norm": 2.859151154864237,
+      "learning_rate": 3.431179556063756e-06,
+      "loss": 0.1535,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6145610278372591,
+      "grad_norm": 2.293692869554135,
+      "learning_rate": 3.4245932918940184e-06,
+      "loss": 0.1356,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6149892933618843,
+      "grad_norm": 2.462936189542011,
+      "learning_rate": 3.4180100611930012e-06,
+      "loss": 0.1426,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6154175588865096,
+      "grad_norm": 2.981192153845732,
+      "learning_rate": 3.4114298766368117e-06,
+      "loss": 0.1492,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6158458244111349,
+      "grad_norm": 2.0057968904078063,
+      "learning_rate": 3.4048527508956896e-06,
+      "loss": 0.1372,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6162740899357602,
+      "grad_norm": 2.473787056660534,
+      "learning_rate": 3.398278696633991e-06,
+      "loss": 0.177,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6167023554603854,
+      "grad_norm": 2.457571408636888,
+      "learning_rate": 3.3917077265101505e-06,
+      "loss": 0.1361,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6171306209850107,
+      "grad_norm": 2.452395773319289,
+      "learning_rate": 3.3851398531766695e-06,
+      "loss": 0.1722,
+      "step": 1441
+    },
+    {
+      "epoch": 0.617558886509636,
+      "grad_norm": 2.545890480440564,
+      "learning_rate": 3.3785750892800823e-06,
+      "loss": 0.1883,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6179871520342612,
+      "grad_norm": 2.4011874467505536,
+      "learning_rate": 3.37201344746094e-06,
+      "loss": 0.1642,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6184154175588865,
+      "grad_norm": 2.297064966084557,
+      "learning_rate": 3.365454940353779e-06,
+      "loss": 0.1555,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6188436830835118,
+      "grad_norm": 2.6420574859079857,
+      "learning_rate": 3.3588995805871005e-06,
+      "loss": 0.1598,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6192719486081371,
+      "grad_norm": 2.473242268022817,
+      "learning_rate": 3.3523473807833483e-06,
+      "loss": 0.1212,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6197002141327623,
+      "grad_norm": 2.3065412802762486,
+      "learning_rate": 3.3457983535588766e-06,
+      "loss": 0.1337,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6201284796573876,
+      "grad_norm": 2.3012187301192157,
+      "learning_rate": 3.3392525115239353e-06,
+      "loss": 0.1422,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6205567451820129,
+      "grad_norm": 2.3163118632874524,
+      "learning_rate": 3.3327098672826373e-06,
+      "loss": 0.1427,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6209850107066381,
+      "grad_norm": 2.328045635731721,
+      "learning_rate": 3.3261704334329416e-06,
+      "loss": 0.1283,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6214132762312634,
+      "grad_norm": 2.522662315468329,
+      "learning_rate": 3.319634222566623e-06,
+      "loss": 0.1598,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6218415417558887,
+      "grad_norm": 2.1677961999343123,
+      "learning_rate": 3.3131012472692515e-06,
+      "loss": 0.1429,
+      "step": 1452
+    },
+    {
+      "epoch": 0.622269807280514,
+      "grad_norm": 2.651002657478796,
+      "learning_rate": 3.3065715201201676e-06,
+      "loss": 0.1358,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6226980728051392,
+      "grad_norm": 2.560958836546812,
+      "learning_rate": 3.3000450536924554e-06,
+      "loss": 0.1223,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6231263383297645,
+      "grad_norm": 2.353748488824112,
+      "learning_rate": 3.293521860552924e-06,
+      "loss": 0.1394,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6235546038543898,
+      "grad_norm": 2.4799327295979485,
+      "learning_rate": 3.2870019532620744e-06,
+      "loss": 0.1512,
+      "step": 1456
+    },
+    {
+      "epoch": 0.623982869379015,
+      "grad_norm": 2.076467722651473,
+      "learning_rate": 3.280485344374087e-06,
+      "loss": 0.134,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6244111349036403,
+      "grad_norm": 2.31427504673827,
+      "learning_rate": 3.2739720464367848e-06,
+      "loss": 0.1208,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6248394004282656,
+      "grad_norm": 2.079939366099188,
+      "learning_rate": 3.2674620719916196e-06,
+      "loss": 0.0914,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6252676659528907,
+      "grad_norm": 2.5749988412187523,
+      "learning_rate": 3.2609554335736435e-06,
+      "loss": 0.1549,
+      "step": 1460
+    },
+    {
+      "epoch": 0.625695931477516,
+      "grad_norm": 2.0030913945838305,
+      "learning_rate": 3.2544521437114828e-06,
+      "loss": 0.1331,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6261241970021413,
+      "grad_norm": 2.110900204142116,
+      "learning_rate": 3.2479522149273195e-06,
+      "loss": 0.1106,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6265524625267666,
+      "grad_norm": 2.459117747745039,
+      "learning_rate": 3.2414556597368598e-06,
+      "loss": 0.1227,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6269807280513918,
+      "grad_norm": 2.0793960116412,
+      "learning_rate": 3.2349624906493164e-06,
+      "loss": 0.1676,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6274089935760171,
+      "grad_norm": 2.4157318086597055,
+      "learning_rate": 3.2284727201673816e-06,
+      "loss": 0.1405,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6278372591006424,
+      "grad_norm": 2.2533070122805436,
+      "learning_rate": 3.2219863607872044e-06,
+      "loss": 0.1429,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6282655246252676,
+      "grad_norm": 2.3659317219998104,
+      "learning_rate": 3.2155034249983652e-06,
+      "loss": 0.1368,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6286937901498929,
+      "grad_norm": 2.4853733418560124,
+      "learning_rate": 3.2090239252838496e-06,
+      "loss": 0.1342,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6291220556745182,
+      "grad_norm": 2.167748750280194,
+      "learning_rate": 3.2025478741200323e-06,
+      "loss": 0.1325,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6295503211991434,
+      "grad_norm": 2.1744233776790356,
+      "learning_rate": 3.1960752839766447e-06,
+      "loss": 0.1029,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6299785867237687,
+      "grad_norm": 2.5347222161474185,
+      "learning_rate": 3.18960616731675e-06,
+      "loss": 0.1251,
+      "step": 1471
+    },
+    {
+      "epoch": 0.630406852248394,
+      "grad_norm": 2.387896544242817,
+      "learning_rate": 3.1831405365967315e-06,
+      "loss": 0.1577,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6308351177730193,
+      "grad_norm": 2.1371436414771967,
+      "learning_rate": 3.1766784042662534e-06,
+      "loss": 0.1505,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6312633832976445,
+      "grad_norm": 2.0262216836412974,
+      "learning_rate": 3.1702197827682476e-06,
+      "loss": 0.1047,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6316916488222698,
+      "grad_norm": 2.958805292942883,
+      "learning_rate": 3.163764684538885e-06,
+      "loss": 0.1652,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6321199143468951,
+      "grad_norm": 2.1071727352811975,
+      "learning_rate": 3.1573131220075494e-06,
+      "loss": 0.0971,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6325481798715203,
+      "grad_norm": 2.1548782198953327,
+      "learning_rate": 3.1508651075968223e-06,
+      "loss": 0.1258,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6329764453961456,
+      "grad_norm": 2.2495697642393457,
+      "learning_rate": 3.1444206537224477e-06,
+      "loss": 0.1203,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6334047109207709,
+      "grad_norm": 2.0857625146736924,
+      "learning_rate": 3.137979772793317e-06,
+      "loss": 0.1023,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6338329764453962,
+      "grad_norm": 2.323150803901988,
+      "learning_rate": 3.1315424772114404e-06,
+      "loss": 0.1594,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6342612419700214,
+      "grad_norm": 2.387466523705558,
+      "learning_rate": 3.1251087793719255e-06,
+      "loss": 0.1493,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6346895074946467,
+      "grad_norm": 2.3369646078486013,
+      "learning_rate": 3.1186786916629517e-06,
+      "loss": 0.1599,
+      "step": 1482
+    },
+    {
+      "epoch": 0.635117773019272,
+      "grad_norm": 2.0201126379924803,
+      "learning_rate": 3.1122522264657474e-06,
+      "loss": 0.113,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6355460385438972,
+      "grad_norm": 2.381944798482335,
+      "learning_rate": 3.1058293961545648e-06,
+      "loss": 0.1641,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6359743040685225,
+      "grad_norm": 2.6269759521509712,
+      "learning_rate": 3.0994102130966588e-06,
+      "loss": 0.1331,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6364025695931478,
+      "grad_norm": 2.4203215648292393,
+      "learning_rate": 3.09299468965226e-06,
+      "loss": 0.1593,
+      "step": 1486
+    },
+    {
+      "epoch": 0.636830835117773,
+      "grad_norm": 2.2388506869150753,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.1481,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6372591006423983,
+      "grad_norm": 2.374963990346724,
+      "learning_rate": 3.0801746710096497e-06,
+      "loss": 0.1306,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6376873661670236,
+      "grad_norm": 2.279968311404025,
+      "learning_rate": 3.0737702004965743e-06,
+      "loss": 0.1371,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6381156316916489,
+      "grad_norm": 2.1777836059532585,
+      "learning_rate": 3.0673694389672247e-06,
+      "loss": 0.117,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6385438972162741,
+      "grad_norm": 1.951397195419024,
+      "learning_rate": 3.060972398746365e-06,
+      "loss": 0.0942,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6389721627408994,
+      "grad_norm": 2.54799355701838,
+      "learning_rate": 3.054579092151586e-06,
+      "loss": 0.1211,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6394004282655247,
+      "grad_norm": 2.3359227557882023,
+      "learning_rate": 3.0481895314932965e-06,
+      "loss": 0.1074,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6398286937901498,
+      "grad_norm": 2.2456631857412064,
+      "learning_rate": 3.041803729074686e-06,
+      "loss": 0.117,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6402569593147751,
+      "grad_norm": 2.389151474947469,
+      "learning_rate": 3.035421697191714e-06,
+      "loss": 0.1789,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6406852248394004,
+      "grad_norm": 2.256650203191494,
+      "learning_rate": 3.0290434481330746e-06,
+      "loss": 0.1136,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6411134903640257,
+      "grad_norm": 1.8756213964567443,
+      "learning_rate": 3.0226689941801803e-06,
+      "loss": 0.1246,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6415417558886509,
+      "grad_norm": 2.0724067672525806,
+      "learning_rate": 3.016298347607136e-06,
+      "loss": 0.122,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6419700214132762,
+      "grad_norm": 2.4409852926500504,
+      "learning_rate": 3.0099315206807135e-06,
+      "loss": 0.1363,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6423982869379015,
+      "grad_norm": 2.2339131827097085,
+      "learning_rate": 3.003568525660334e-06,
+      "loss": 0.1076,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6428265524625267,
+      "grad_norm": 2.180389865509403,
+      "learning_rate": 2.9972093747980357e-06,
+      "loss": 0.1294,
+      "step": 1501
+    },
+    {
+      "epoch": 0.643254817987152,
+      "grad_norm": 4.358429464476535,
+      "learning_rate": 2.9908540803384577e-06,
+      "loss": 0.1619,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6436830835117773,
+      "grad_norm": 2.19453744383653,
+      "learning_rate": 2.9845026545188137e-06,
+      "loss": 0.1428,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6441113490364025,
+      "grad_norm": 2.287865680747469,
+      "learning_rate": 2.978155109568864e-06,
+      "loss": 0.1184,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6445396145610278,
+      "grad_norm": 2.8543745469745043,
+      "learning_rate": 2.9718114577109055e-06,
+      "loss": 0.1744,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6449678800856531,
+      "grad_norm": 2.9893973262562326,
+      "learning_rate": 2.9654717111597265e-06,
+      "loss": 0.148,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6453961456102784,
+      "grad_norm": 1.8641949988001123,
+      "learning_rate": 2.9591358821226035e-06,
+      "loss": 0.1219,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6458244111349036,
+      "grad_norm": 2.1661923402032195,
+      "learning_rate": 2.952803982799271e-06,
+      "loss": 0.1467,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6462526766595289,
+      "grad_norm": 1.9578036095808655,
+      "learning_rate": 2.94647602538189e-06,
+      "loss": 0.1301,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6466809421841542,
+      "grad_norm": 2.0918433622344437,
+      "learning_rate": 2.940152022055039e-06,
+      "loss": 0.1217,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6471092077087794,
+      "grad_norm": 2.5568698503065717,
+      "learning_rate": 2.9338319849956777e-06,
+      "loss": 0.1372,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6475374732334047,
+      "grad_norm": 2.0629176979345565,
+      "learning_rate": 2.927515926373129e-06,
+      "loss": 0.1058,
+      "step": 1512
+    },
+    {
+      "epoch": 0.64796573875803,
+      "grad_norm": 2.5119807196608948,
+      "learning_rate": 2.9212038583490588e-06,
+      "loss": 0.182,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6483940042826553,
+      "grad_norm": 2.3584547457429568,
+      "learning_rate": 2.9148957930774456e-06,
+      "loss": 0.15,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6488222698072805,
+      "grad_norm": 2.0334504061402314,
+      "learning_rate": 2.9085917427045598e-06,
+      "loss": 0.1355,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6492505353319058,
+      "grad_norm": 2.182498510002562,
+      "learning_rate": 2.902291719368945e-06,
+      "loss": 0.1291,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6496788008565311,
+      "grad_norm": 2.3173734140730584,
+      "learning_rate": 2.895995735201387e-06,
+      "loss": 0.1023,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6501070663811563,
+      "grad_norm": 2.2947589902337198,
+      "learning_rate": 2.889703802324897e-06,
+      "loss": 0.1164,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6505353319057816,
+      "grad_norm": 2.2314526063993774,
+      "learning_rate": 2.883415932854685e-06,
+      "loss": 0.1294,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6509635974304069,
+      "grad_norm": 2.3831556261767886,
+      "learning_rate": 2.8771321388981334e-06,
+      "loss": 0.1176,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6513918629550322,
+      "grad_norm": 1.9326023451772578,
+      "learning_rate": 2.8708524325547805e-06,
+      "loss": 0.1345,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6518201284796574,
+      "grad_norm": 2.3250939100708647,
+      "learning_rate": 2.8645768259162925e-06,
+      "loss": 0.1696,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6522483940042827,
+      "grad_norm": 2.0571063592540213,
+      "learning_rate": 2.858305331066445e-06,
+      "loss": 0.0994,
+      "step": 1523
+    },
+    {
+      "epoch": 0.652676659528908,
+      "grad_norm": 2.3083876796067604,
+      "learning_rate": 2.8520379600810886e-06,
+      "loss": 0.1371,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6531049250535332,
+      "grad_norm": 2.094774921524476,
+      "learning_rate": 2.845774725028141e-06,
+      "loss": 0.1396,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6535331905781585,
+      "grad_norm": 2.2137928534052618,
+      "learning_rate": 2.839515637967552e-06,
+      "loss": 0.0869,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6539614561027838,
+      "grad_norm": 2.5541305708978155,
+      "learning_rate": 2.8332607109512866e-06,
+      "loss": 0.1236,
+      "step": 1527
+    },
+    {
+      "epoch": 0.654389721627409,
+      "grad_norm": 2.1983186670949415,
+      "learning_rate": 2.8270099560232992e-06,
+      "loss": 0.1332,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6548179871520342,
+      "grad_norm": 2.5820266509241647,
+      "learning_rate": 2.8207633852195075e-06,
+      "loss": 0.1608,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6552462526766595,
+      "grad_norm": 2.385526378485658,
+      "learning_rate": 2.814521010567777e-06,
+      "loss": 0.128,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6556745182012848,
+      "grad_norm": 2.495930101918827,
+      "learning_rate": 2.8082828440878897e-06,
+      "loss": 0.1366,
+      "step": 1531
+    },
+    {
+      "epoch": 0.65610278372591,
+      "grad_norm": 2.074614667586663,
+      "learning_rate": 2.802048897791529e-06,
+      "loss": 0.1226,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6565310492505353,
+      "grad_norm": 2.6339330208077416,
+      "learning_rate": 2.7958191836822505e-06,
+      "loss": 0.1623,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6569593147751606,
+      "grad_norm": 2.3699250699650465,
+      "learning_rate": 2.7895937137554576e-06,
+      "loss": 0.1118,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6573875802997858,
+      "grad_norm": 1.9952289343248646,
+      "learning_rate": 2.783372499998384e-06,
+      "loss": 0.1005,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6578158458244111,
+      "grad_norm": 2.0003420134738956,
+      "learning_rate": 2.77715555439007e-06,
+      "loss": 0.1106,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6582441113490364,
+      "grad_norm": 1.9663672810795925,
+      "learning_rate": 2.770942888901335e-06,
+      "loss": 0.1123,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6586723768736616,
+      "grad_norm": 2.3490474344973507,
+      "learning_rate": 2.7647345154947557e-06,
+      "loss": 0.1026,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6591006423982869,
+      "grad_norm": 2.4488138714188183,
+      "learning_rate": 2.7585304461246477e-06,
+      "loss": 0.1277,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6595289079229122,
+      "grad_norm": 2.181643326585293,
+      "learning_rate": 2.7523306927370375e-06,
+      "loss": 0.1313,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6599571734475375,
+      "grad_norm": 2.183621378233564,
+      "learning_rate": 2.7461352672696423e-06,
+      "loss": 0.0843,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6603854389721627,
+      "grad_norm": 2.3299951639011516,
+      "learning_rate": 2.7399441816518417e-06,
+      "loss": 0.1562,
+      "step": 1542
+    },
+    {
+      "epoch": 0.660813704496788,
+      "grad_norm": 2.3410411585557793,
+      "learning_rate": 2.733757447804666e-06,
+      "loss": 0.1467,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6612419700214133,
+      "grad_norm": 2.392751552408249,
+      "learning_rate": 2.7275750776407568e-06,
+      "loss": 0.0963,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6616702355460385,
+      "grad_norm": 2.0522871121598065,
+      "learning_rate": 2.721397083064361e-06,
+      "loss": 0.1201,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6620985010706638,
+      "grad_norm": 1.9166524739030406,
+      "learning_rate": 2.7152234759712986e-06,
+      "loss": 0.1039,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6625267665952891,
+      "grad_norm": 2.249995162334823,
+      "learning_rate": 2.7090542682489398e-06,
+      "loss": 0.1099,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6629550321199144,
+      "grad_norm": 2.5096619819269153,
+      "learning_rate": 2.7028894717761867e-06,
+      "loss": 0.1402,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6633832976445396,
+      "grad_norm": 2.1664141332390043,
+      "learning_rate": 2.6967290984234413e-06,
+      "loss": 0.1322,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6638115631691649,
+      "grad_norm": 2.077055491480785,
+      "learning_rate": 2.6905731600525946e-06,
+      "loss": 0.1186,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6642398286937902,
+      "grad_norm": 1.884526954497474,
+      "learning_rate": 2.6844216685169966e-06,
+      "loss": 0.111,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6646680942184154,
+      "grad_norm": 2.0672620249804763,
+      "learning_rate": 2.6782746356614364e-06,
+      "loss": 0.1235,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6650963597430407,
+      "grad_norm": 1.6684173576442853,
+      "learning_rate": 2.672132073322112e-06,
+      "loss": 0.087,
+      "step": 1553
+    },
+    {
+      "epoch": 0.665524625267666,
+      "grad_norm": 1.9086501116942187,
+      "learning_rate": 2.665993993326619e-06,
+      "loss": 0.0753,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6659528907922913,
+      "grad_norm": 2.167075283548563,
+      "learning_rate": 2.65986040749392e-06,
+      "loss": 0.1494,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6663811563169165,
+      "grad_norm": 2.8685960600975675,
+      "learning_rate": 2.6537313276343255e-06,
+      "loss": 0.1577,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6668094218415418,
+      "grad_norm": 2.2098182014081633,
+      "learning_rate": 2.6476067655494696e-06,
+      "loss": 0.1401,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6672376873661671,
+      "grad_norm": 2.0752166477478635,
+      "learning_rate": 2.6414867330322835e-06,
+      "loss": 0.0717,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6676659528907923,
+      "grad_norm": 2.428453861134804,
+      "learning_rate": 2.6353712418669814e-06,
+      "loss": 0.1286,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6680942184154176,
+      "grad_norm": 2.3874296871620513,
+      "learning_rate": 2.6292603038290306e-06,
+      "loss": 0.1069,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6685224839400429,
+      "grad_norm": 2.3617123682253434,
+      "learning_rate": 2.6231539306851316e-06,
+      "loss": 0.1293,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6689507494646681,
+      "grad_norm": 2.5408719814720686,
+      "learning_rate": 2.6170521341931986e-06,
+      "loss": 0.119,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6693790149892933,
+      "grad_norm": 2.1280820740193436,
+      "learning_rate": 2.610954926102326e-06,
+      "loss": 0.0854,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6698072805139186,
+      "grad_norm": 2.4901767742636474,
+      "learning_rate": 2.604862318152778e-06,
+      "loss": 0.1124,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6702355460385439,
+      "grad_norm": 2.297667305901331,
+      "learning_rate": 2.598774322075962e-06,
+      "loss": 0.0946,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6706638115631691,
+      "grad_norm": 3.014260176318125,
+      "learning_rate": 2.5926909495944056e-06,
+      "loss": 0.1669,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6710920770877944,
+      "grad_norm": 2.0640467772183353,
+      "learning_rate": 2.5866122124217273e-06,
+      "loss": 0.1054,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6715203426124197,
+      "grad_norm": 2.2968528532871404,
+      "learning_rate": 2.580538122262627e-06,
+      "loss": 0.1066,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6719486081370449,
+      "grad_norm": 2.41145923383787,
+      "learning_rate": 2.574468690812854e-06,
+      "loss": 0.1326,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6723768736616702,
+      "grad_norm": 1.9124870958908013,
+      "learning_rate": 2.568403929759188e-06,
+      "loss": 0.1054,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6728051391862955,
+      "grad_norm": 2.0600210471611993,
+      "learning_rate": 2.562343850779417e-06,
+      "loss": 0.1208,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6732334047109207,
+      "grad_norm": 2.491197995964675,
+      "learning_rate": 2.556288465542308e-06,
+      "loss": 0.1672,
+      "step": 1572
+    },
+    {
+      "epoch": 0.673661670235546,
+      "grad_norm": 2.255296384015942,
+      "learning_rate": 2.550237785707596e-06,
+      "loss": 0.1334,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6740899357601713,
+      "grad_norm": 2.1337828614605145,
+      "learning_rate": 2.5441918229259543e-06,
+      "loss": 0.1243,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6745182012847966,
+      "grad_norm": 1.9690854329914536,
+      "learning_rate": 2.538150588838972e-06,
+      "loss": 0.1255,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6749464668094218,
+      "grad_norm": 2.207927367354541,
+      "learning_rate": 2.532114095079137e-06,
+      "loss": 0.1234,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6753747323340471,
+      "grad_norm": 2.1227697120050624,
+      "learning_rate": 2.526082353269803e-06,
+      "loss": 0.0918,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6758029978586724,
+      "grad_norm": 2.178324209637493,
+      "learning_rate": 2.520055375025178e-06,
+      "loss": 0.1428,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6762312633832976,
+      "grad_norm": 2.04673347679938,
+      "learning_rate": 2.5140331719502957e-06,
+      "loss": 0.0987,
+      "step": 1579
+    },
+    {
+      "epoch": 0.6766595289079229,
+      "grad_norm": 2.8615025275666266,
+      "learning_rate": 2.508015755640999e-06,
+      "loss": 0.1302,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6770877944325482,
+      "grad_norm": 2.325749336058442,
+      "learning_rate": 2.5020031376839115e-06,
+      "loss": 0.1403,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6775160599571735,
+      "grad_norm": 2.496621469420446,
+      "learning_rate": 2.4959953296564143e-06,
+      "loss": 0.097,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6779443254817987,
+      "grad_norm": 2.4844462349154477,
+      "learning_rate": 2.489992343126632e-06,
+      "loss": 0.1261,
+      "step": 1583
+    },
+    {
+      "epoch": 0.678372591006424,
+      "grad_norm": 1.9144008567171158,
+      "learning_rate": 2.4839941896534027e-06,
+      "loss": 0.0998,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6788008565310493,
+      "grad_norm": 2.7826209691729997,
+      "learning_rate": 2.478000880786259e-06,
+      "loss": 0.1772,
+      "step": 1585
+    },
+    {
+      "epoch": 0.6792291220556745,
+      "grad_norm": 2.0886486382834537,
+      "learning_rate": 2.4720124280654066e-06,
+      "loss": 0.1445,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6796573875802998,
+      "grad_norm": 2.170806545439487,
+      "learning_rate": 2.4660288430216954e-06,
+      "loss": 0.1085,
+      "step": 1587
+    },
+    {
+      "epoch": 0.6800856531049251,
+      "grad_norm": 2.1065136869117533,
+      "learning_rate": 2.4600501371766087e-06,
+      "loss": 0.1083,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6805139186295504,
+      "grad_norm": 2.3309240164289604,
+      "learning_rate": 2.4540763220422316e-06,
+      "loss": 0.1189,
+      "step": 1589
+    },
+    {
+      "epoch": 0.6809421841541756,
+      "grad_norm": 2.3478929397527817,
+      "learning_rate": 2.4481074091212333e-06,
+      "loss": 0.115,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6813704496788009,
+      "grad_norm": 2.2244080968028466,
+      "learning_rate": 2.442143409906844e-06,
+      "loss": 0.1471,
+      "step": 1591
+    },
+    {
+      "epoch": 0.6817987152034262,
+      "grad_norm": 2.4695638913570543,
+      "learning_rate": 2.4361843358828287e-06,
+      "loss": 0.1403,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6822269807280514,
+      "grad_norm": 2.0987747841080373,
+      "learning_rate": 2.4302301985234737e-06,
+      "loss": 0.0969,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6826552462526767,
+      "grad_norm": 2.4302325040139405,
+      "learning_rate": 2.4242810092935578e-06,
+      "loss": 0.1461,
+      "step": 1594
+    },
+    {
+      "epoch": 0.683083511777302,
+      "grad_norm": 2.024864587382046,
+      "learning_rate": 2.4183367796483337e-06,
+      "loss": 0.0966,
+      "step": 1595
+    },
+    {
+      "epoch": 0.6835117773019272,
+      "grad_norm": 2.1766076971579156,
+      "learning_rate": 2.4123975210334987e-06,
+      "loss": 0.1466,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6839400428265524,
+      "grad_norm": 2.1375958600080853,
+      "learning_rate": 2.4064632448851852e-06,
+      "loss": 0.1036,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6843683083511777,
+      "grad_norm": 2.1591745741803585,
+      "learning_rate": 2.4005339626299285e-06,
+      "loss": 0.1255,
+      "step": 1598
+    },
+    {
+      "epoch": 0.684796573875803,
+      "grad_norm": 1.8110164663790995,
+      "learning_rate": 2.3946096856846485e-06,
+      "loss": 0.1209,
+      "step": 1599
+    },
+    {
+      "epoch": 0.6852248394004282,
+      "grad_norm": 2.0772773288566238,
+      "learning_rate": 2.388690425456629e-06,
+      "loss": 0.1204,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6856531049250535,
+      "grad_norm": 1.8717301082854454,
+      "learning_rate": 2.3827761933434886e-06,
+      "loss": 0.101,
+      "step": 1601
+    },
+    {
+      "epoch": 0.6860813704496788,
+      "grad_norm": 2.2767928664577894,
+      "learning_rate": 2.37686700073317e-06,
+      "loss": 0.1502,
+      "step": 1602
+    },
+    {
+      "epoch": 0.686509635974304,
+      "grad_norm": 2.2872283766793573,
+      "learning_rate": 2.3709628590039106e-06,
+      "loss": 0.1383,
+      "step": 1603
+    },
+    {
+      "epoch": 0.6869379014989293,
+      "grad_norm": 2.196502468025969,
+      "learning_rate": 2.365063779524222e-06,
+      "loss": 0.1369,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6873661670235546,
+      "grad_norm": 1.9381322206379352,
+      "learning_rate": 2.359169773652869e-06,
+      "loss": 0.1294,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6877944325481798,
+      "grad_norm": 2.467806411078072,
+      "learning_rate": 2.3532808527388435e-06,
+      "loss": 0.1442,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6882226980728051,
+      "grad_norm": 2.55036105640822,
+      "learning_rate": 2.347397028121351e-06,
+      "loss": 0.1318,
+      "step": 1607
+    },
+    {
+      "epoch": 0.6886509635974304,
+      "grad_norm": 2.1757359560870384,
+      "learning_rate": 2.341518311129781e-06,
+      "loss": 0.1031,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6890792291220557,
+      "grad_norm": 2.1008347644103478,
+      "learning_rate": 2.335644713083693e-06,
+      "loss": 0.0935,
+      "step": 1609
+    },
+    {
+      "epoch": 0.6895074946466809,
+      "grad_norm": 2.007351593454887,
+      "learning_rate": 2.3297762452927814e-06,
+      "loss": 0.0978,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6899357601713062,
+      "grad_norm": 1.7850723156181898,
+      "learning_rate": 2.32391291905687e-06,
+      "loss": 0.0698,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6903640256959315,
+      "grad_norm": 2.5985342929544113,
+      "learning_rate": 2.318054745665877e-06,
+      "loss": 0.1803,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6907922912205567,
+      "grad_norm": 2.133091809474177,
+      "learning_rate": 2.312201736399802e-06,
+      "loss": 0.092,
+      "step": 1613
+    },
+    {
+      "epoch": 0.691220556745182,
+      "grad_norm": 2.1250219237344417,
+      "learning_rate": 2.3063539025287014e-06,
+      "loss": 0.1079,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6916488222698073,
+      "grad_norm": 2.5156672122495967,
+      "learning_rate": 2.300511255312666e-06,
+      "loss": 0.1433,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6920770877944326,
+      "grad_norm": 2.4600037135143027,
+      "learning_rate": 2.2946738060017947e-06,
+      "loss": 0.1247,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6925053533190578,
+      "grad_norm": 2.143173864217884,
+      "learning_rate": 2.2888415658361845e-06,
+      "loss": 0.0901,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6929336188436831,
+      "grad_norm": 1.9667405363998096,
+      "learning_rate": 2.283014546045899e-06,
+      "loss": 0.0964,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6933618843683084,
+      "grad_norm": 2.0429653541677184,
+      "learning_rate": 2.2771927578509508e-06,
+      "loss": 0.1182,
+      "step": 1619
+    },
+    {
+      "epoch": 0.6937901498929336,
+      "grad_norm": 2.0625862878165617,
+      "learning_rate": 2.2713762124612794e-06,
+      "loss": 0.0917,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6942184154175589,
+      "grad_norm": 2.367028476928625,
+      "learning_rate": 2.2655649210767243e-06,
+      "loss": 0.1086,
+      "step": 1621
+    },
+    {
+      "epoch": 0.6946466809421842,
+      "grad_norm": 1.8124927186455375,
+      "learning_rate": 2.259758894887015e-06,
+      "loss": 0.1175,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6950749464668095,
+      "grad_norm": 2.3159641810583858,
+      "learning_rate": 2.2539581450717403e-06,
+      "loss": 0.13,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6955032119914347,
+      "grad_norm": 2.08641544838903,
+      "learning_rate": 2.24816268280033e-06,
+      "loss": 0.1103,
+      "step": 1624
+    },
+    {
+      "epoch": 0.69593147751606,
+      "grad_norm": 2.5141132205262697,
+      "learning_rate": 2.2423725192320284e-06,
+      "loss": 0.1766,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6963597430406853,
+      "grad_norm": 2.3476780168907423,
+      "learning_rate": 2.2365876655158824e-06,
+      "loss": 0.1313,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6967880085653105,
+      "grad_norm": 1.8402619196618024,
+      "learning_rate": 2.230808132790712e-06,
+      "loss": 0.102,
+      "step": 1627
+    },
+    {
+      "epoch": 0.6972162740899358,
+      "grad_norm": 2.0907715582521633,
+      "learning_rate": 2.2250339321850934e-06,
+      "loss": 0.1244,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6976445396145611,
+      "grad_norm": 2.2023650665190724,
+      "learning_rate": 2.2192650748173354e-06,
+      "loss": 0.1318,
+      "step": 1629
+    },
+    {
+      "epoch": 0.6980728051391863,
+      "grad_norm": 2.2118036633022853,
+      "learning_rate": 2.213501571795454e-06,
+      "loss": 0.1087,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6985010706638116,
+      "grad_norm": 2.1589078723456323,
+      "learning_rate": 2.2077434342171605e-06,
+      "loss": 0.1158,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6989293361884368,
+      "grad_norm": 2.37599259670701,
+      "learning_rate": 2.2019906731698337e-06,
+      "loss": 0.0982,
+      "step": 1632
+    },
+    {
+      "epoch": 0.699357601713062,
+      "grad_norm": 2.091989450782926,
+      "learning_rate": 2.1962432997304993e-06,
+      "loss": 0.09,
+      "step": 1633
+    },
+    {
+      "epoch": 0.6997858672376873,
+      "grad_norm": 2.3803410316989475,
+      "learning_rate": 2.1905013249658104e-06,
+      "loss": 0.1372,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7002141327623126,
+      "grad_norm": 2.322082062838183,
+      "learning_rate": 2.1847647599320204e-06,
+      "loss": 0.0999,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7006423982869379,
+      "grad_norm": 2.2786088033104335,
+      "learning_rate": 2.179033615674971e-06,
+      "loss": 0.1201,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7010706638115631,
+      "grad_norm": 2.316959926848027,
+      "learning_rate": 2.173307903230065e-06,
+      "loss": 0.1127,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7014989293361884,
+      "grad_norm": 1.8531592308615872,
+      "learning_rate": 2.1675876336222475e-06,
+      "loss": 0.1275,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7019271948608137,
+      "grad_norm": 1.8970778246431528,
+      "learning_rate": 2.1618728178659772e-06,
+      "loss": 0.1026,
+      "step": 1639
+    },
+    {
+      "epoch": 0.702355460385439,
+      "grad_norm": 2.0349389477043873,
+      "learning_rate": 2.156163466965218e-06,
+      "loss": 0.1011,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7027837259100642,
+      "grad_norm": 2.1863239794048517,
+      "learning_rate": 2.1504595919134086e-06,
+      "loss": 0.1087,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7032119914346895,
+      "grad_norm": 2.431903114571287,
+      "learning_rate": 2.1447612036934436e-06,
+      "loss": 0.0994,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7036402569593148,
+      "grad_norm": 2.315635621560668,
+      "learning_rate": 2.139068313277656e-06,
+      "loss": 0.11,
+      "step": 1643
+    },
+    {
+      "epoch": 0.70406852248394,
+      "grad_norm": 2.756160317122608,
+      "learning_rate": 2.1333809316277854e-06,
+      "loss": 0.1324,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7044967880085653,
+      "grad_norm": 2.225248540707109,
+      "learning_rate": 2.12769906969497e-06,
+      "loss": 0.1028,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7049250535331906,
+      "grad_norm": 2.179974860774558,
+      "learning_rate": 2.1220227384197195e-06,
+      "loss": 0.0983,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7053533190578158,
+      "grad_norm": 2.235824115064231,
+      "learning_rate": 2.1163519487318936e-06,
+      "loss": 0.1131,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7057815845824411,
+      "grad_norm": 2.5224722559673407,
+      "learning_rate": 2.110686711550678e-06,
+      "loss": 0.1184,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7062098501070664,
+      "grad_norm": 2.2568282794435293,
+      "learning_rate": 2.105027037784574e-06,
+      "loss": 0.1217,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7066381156316917,
+      "grad_norm": 2.093174266181544,
+      "learning_rate": 2.0993729383313615e-06,
+      "loss": 0.131,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7070663811563169,
+      "grad_norm": 1.8795165537223157,
+      "learning_rate": 2.093724424078095e-06,
+      "loss": 0.1762,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7074946466809422,
+      "grad_norm": 2.230518127272181,
+      "learning_rate": 2.0880815059010716e-06,
+      "loss": 0.1118,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7079229122055675,
+      "grad_norm": 1.9731279153004047,
+      "learning_rate": 2.0824441946658134e-06,
+      "loss": 0.0748,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7083511777301927,
+      "grad_norm": 2.4471509750926237,
+      "learning_rate": 2.076812501227044e-06,
+      "loss": 0.1508,
+      "step": 1654
+    },
+    {
+      "epoch": 0.708779443254818,
+      "grad_norm": 2.050188941949392,
+      "learning_rate": 2.071186436428672e-06,
+      "loss": 0.0848,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7092077087794433,
+      "grad_norm": 2.1727267971241027,
+      "learning_rate": 2.0655660111037685e-06,
+      "loss": 0.1008,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7096359743040686,
+      "grad_norm": 2.267225744245954,
+      "learning_rate": 2.059951236074545e-06,
+      "loss": 0.1308,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7100642398286938,
+      "grad_norm": 2.288699542608484,
+      "learning_rate": 2.054342122152334e-06,
+      "loss": 0.101,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7104925053533191,
+      "grad_norm": 2.148082179248419,
+      "learning_rate": 2.0487386801375646e-06,
+      "loss": 0.1179,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7109207708779444,
+      "grad_norm": 2.3620009577802072,
+      "learning_rate": 2.043140920819747e-06,
+      "loss": 0.1095,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7113490364025696,
+      "grad_norm": 2.183978860614624,
+      "learning_rate": 2.0375488549774493e-06,
+      "loss": 0.1118,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7117773019271949,
+      "grad_norm": 2.044486113280406,
+      "learning_rate": 2.031962493378276e-06,
+      "loss": 0.0826,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7122055674518202,
+      "grad_norm": 2.008522499453445,
+      "learning_rate": 2.026381846778851e-06,
+      "loss": 0.1097,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7126338329764454,
+      "grad_norm": 1.8465320324277266,
+      "learning_rate": 2.0208069259247866e-06,
+      "loss": 0.1025,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7130620985010707,
+      "grad_norm": 2.4664839570140513,
+      "learning_rate": 2.0152377415506767e-06,
+      "loss": 0.1176,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7134903640256959,
+      "grad_norm": 2.3011099294803836,
+      "learning_rate": 2.0096743043800683e-06,
+      "loss": 0.101,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7139186295503211,
+      "grad_norm": 2.0178640680033246,
+      "learning_rate": 2.004116625125442e-06,
+      "loss": 0.1102,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7143468950749464,
+      "grad_norm": 2.0851121374920294,
+      "learning_rate": 1.998564714488187e-06,
+      "loss": 0.1144,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7147751605995717,
+      "grad_norm": 2.2561044595639492,
+      "learning_rate": 1.9930185831585912e-06,
+      "loss": 0.1065,
+      "step": 1669
+    },
+    {
+      "epoch": 0.715203426124197,
+      "grad_norm": 2.4975889922695442,
+      "learning_rate": 1.987478241815812e-06,
+      "loss": 0.1074,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7156316916488222,
+      "grad_norm": 2.0881135441645724,
+      "learning_rate": 1.981943701127857e-06,
+      "loss": 0.1225,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7160599571734475,
+      "grad_norm": 2.0479604940889358,
+      "learning_rate": 1.976414971751568e-06,
+      "loss": 0.089,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7164882226980728,
+      "grad_norm": 2.1080578291541574,
+      "learning_rate": 1.97089206433259e-06,
+      "loss": 0.1094,
+      "step": 1673
+    },
+    {
+      "epoch": 0.716916488222698,
+      "grad_norm": 2.2464931568362267,
+      "learning_rate": 1.9653749895053653e-06,
+      "loss": 0.1253,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7173447537473233,
+      "grad_norm": 2.0466478799030017,
+      "learning_rate": 1.9598637578931016e-06,
+      "loss": 0.0917,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7177730192719486,
+      "grad_norm": 1.9421081843084924,
+      "learning_rate": 1.9543583801077567e-06,
+      "loss": 0.1068,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7182012847965739,
+      "grad_norm": 1.9724478701067276,
+      "learning_rate": 1.948858866750017e-06,
+      "loss": 0.0865,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7186295503211991,
+      "grad_norm": 2.220519304300286,
+      "learning_rate": 1.943365228409273e-06,
+      "loss": 0.1018,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7190578158458244,
+      "grad_norm": 2.056870976318867,
+      "learning_rate": 1.9378774756636086e-06,
+      "loss": 0.0881,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7194860813704497,
+      "grad_norm": 2.5179228375951097,
+      "learning_rate": 1.932395619079771e-06,
+      "loss": 0.1555,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7199143468950749,
+      "grad_norm": 2.426236666720795,
+      "learning_rate": 1.926919669213158e-06,
+      "loss": 0.1016,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7203426124197002,
+      "grad_norm": 2.2343102427120196,
+      "learning_rate": 1.9214496366077893e-06,
+      "loss": 0.1078,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7207708779443255,
+      "grad_norm": 2.1416919919345885,
+      "learning_rate": 1.915985531796291e-06,
+      "loss": 0.1199,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7211991434689508,
+      "grad_norm": 2.3865547231591937,
+      "learning_rate": 1.910527365299879e-06,
+      "loss": 0.1269,
+      "step": 1684
+    },
+    {
+      "epoch": 0.721627408993576,
+      "grad_norm": 2.093193164019054,
+      "learning_rate": 1.905075147628333e-06,
+      "loss": 0.1253,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7220556745182013,
+      "grad_norm": 2.2452600956898388,
+      "learning_rate": 1.8996288892799775e-06,
+      "loss": 0.0984,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7224839400428266,
+      "grad_norm": 2.148567405731751,
+      "learning_rate": 1.8941886007416654e-06,
+      "loss": 0.0857,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7229122055674518,
+      "grad_norm": 2.4719827720402114,
+      "learning_rate": 1.8887542924887486e-06,
+      "loss": 0.1349,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7233404710920771,
+      "grad_norm": 2.2201500942586723,
+      "learning_rate": 1.883325974985068e-06,
+      "loss": 0.0877,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7237687366167024,
+      "grad_norm": 2.198646605403852,
+      "learning_rate": 1.8779036586829296e-06,
+      "loss": 0.0864,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7241970021413277,
+      "grad_norm": 2.2559627744041335,
+      "learning_rate": 1.8724873540230826e-06,
+      "loss": 0.1149,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7246252676659529,
+      "grad_norm": 2.2219302876318343,
+      "learning_rate": 1.8670770714347024e-06,
+      "loss": 0.1267,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7250535331905782,
+      "grad_norm": 2.0411610008836196,
+      "learning_rate": 1.8616728213353652e-06,
+      "loss": 0.1353,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7254817987152035,
+      "grad_norm": 1.995343856594939,
+      "learning_rate": 1.856274614131034e-06,
+      "loss": 0.099,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7259100642398287,
+      "grad_norm": 2.4472070528894516,
+      "learning_rate": 1.8508824602160375e-06,
+      "loss": 0.141,
+      "step": 1695
+    },
+    {
+      "epoch": 0.726338329764454,
+      "grad_norm": 2.1972712310512565,
+      "learning_rate": 1.8454963699730471e-06,
+      "loss": 0.1246,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7267665952890793,
+      "grad_norm": 2.0839993117653663,
+      "learning_rate": 1.8401163537730566e-06,
+      "loss": 0.1061,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7271948608137045,
+      "grad_norm": 2.4017355244538905,
+      "learning_rate": 1.8347424219753674e-06,
+      "loss": 0.1465,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7276231263383298,
+      "grad_norm": 2.0938288243223075,
+      "learning_rate": 1.8293745849275635e-06,
+      "loss": 0.1072,
+      "step": 1699
+    },
+    {
+      "epoch": 0.728051391862955,
+      "grad_norm": 2.077649515725739,
+      "learning_rate": 1.8240128529654944e-06,
+      "loss": 0.1079,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7284796573875802,
+      "grad_norm": 2.257793424547906,
+      "learning_rate": 1.8186572364132548e-06,
+      "loss": 0.1502,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7289079229122055,
+      "grad_norm": 2.118798470311055,
+      "learning_rate": 1.8133077455831605e-06,
+      "loss": 0.0984,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7293361884368308,
+      "grad_norm": 2.159690098336262,
+      "learning_rate": 1.8079643907757349e-06,
+      "loss": 0.089,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7297644539614561,
+      "grad_norm": 2.0380633214025305,
+      "learning_rate": 1.802627182279687e-06,
+      "loss": 0.0904,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7301927194860813,
+      "grad_norm": 1.9316180660598707,
+      "learning_rate": 1.7972961303718899e-06,
+      "loss": 0.121,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7306209850107066,
+      "grad_norm": 1.8759878289071652,
+      "learning_rate": 1.7919712453173637e-06,
+      "loss": 0.0699,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7310492505353319,
+      "grad_norm": 1.9207148928750877,
+      "learning_rate": 1.7866525373692495e-06,
+      "loss": 0.1092,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7314775160599571,
+      "grad_norm": 2.1197148725993156,
+      "learning_rate": 1.781340016768799e-06,
+      "loss": 0.127,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7319057815845824,
+      "grad_norm": 2.099455328685759,
+      "learning_rate": 1.7760336937453487e-06,
+      "loss": 0.1476,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7323340471092077,
+      "grad_norm": 2.0163166459452784,
+      "learning_rate": 1.7707335785163027e-06,
+      "loss": 0.1407,
+      "step": 1710
+    },
+    {
+      "epoch": 0.732762312633833,
+      "grad_norm": 1.969008748205219,
+      "learning_rate": 1.7654396812871072e-06,
+      "loss": 0.1099,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7331905781584582,
+      "grad_norm": 2.457448635223921,
+      "learning_rate": 1.760152012251241e-06,
+      "loss": 0.116,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7336188436830835,
+      "grad_norm": 1.9718514959904916,
+      "learning_rate": 1.7548705815901885e-06,
+      "loss": 0.0952,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7340471092077088,
+      "grad_norm": 1.8128104689408087,
+      "learning_rate": 1.7495953994734216e-06,
+      "loss": 0.0807,
+      "step": 1714
+    },
+    {
+      "epoch": 0.734475374732334,
+      "grad_norm": 2.3301141157760243,
+      "learning_rate": 1.7443264760583829e-06,
+      "loss": 0.1345,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7349036402569593,
+      "grad_norm": 2.0921630637365967,
+      "learning_rate": 1.7390638214904576e-06,
+      "loss": 0.0824,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7353319057815846,
+      "grad_norm": 2.1875529547251307,
+      "learning_rate": 1.733807445902968e-06,
+      "loss": 0.1149,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7357601713062099,
+      "grad_norm": 1.9832813221178773,
+      "learning_rate": 1.728557359417139e-06,
+      "loss": 0.1094,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7361884368308351,
+      "grad_norm": 2.056950387696738,
+      "learning_rate": 1.7233135721420908e-06,
+      "loss": 0.1122,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7366167023554604,
+      "grad_norm": 2.0015430714652225,
+      "learning_rate": 1.7180760941748132e-06,
+      "loss": 0.0958,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7370449678800857,
+      "grad_norm": 2.525215139182262,
+      "learning_rate": 1.7128449356001476e-06,
+      "loss": 0.1564,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7374732334047109,
+      "grad_norm": 2.199182282295698,
+      "learning_rate": 1.7076201064907637e-06,
+      "loss": 0.1214,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7379014989293362,
+      "grad_norm": 2.1430825481204177,
+      "learning_rate": 1.7024016169071483e-06,
+      "loss": 0.0833,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7383297644539615,
+      "grad_norm": 2.2006679705376806,
+      "learning_rate": 1.6971894768975794e-06,
+      "loss": 0.132,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7387580299785867,
+      "grad_norm": 2.717340276566099,
+      "learning_rate": 1.6919836964981113e-06,
+      "loss": 0.144,
+      "step": 1725
+    },
+    {
+      "epoch": 0.739186295503212,
+      "grad_norm": 2.2222698417797946,
+      "learning_rate": 1.6867842857325462e-06,
+      "loss": 0.1141,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7396145610278373,
+      "grad_norm": 1.7377672561387636,
+      "learning_rate": 1.681591254612428e-06,
+      "loss": 0.13,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7400428265524626,
+      "grad_norm": 2.271636227677074,
+      "learning_rate": 1.6764046131370142e-06,
+      "loss": 0.1473,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7404710920770878,
+      "grad_norm": 2.006703403280961,
+      "learning_rate": 1.671224371293259e-06,
+      "loss": 0.1112,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7408993576017131,
+      "grad_norm": 2.004140140998794,
+      "learning_rate": 1.6660505390557958e-06,
+      "loss": 0.127,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7413276231263384,
+      "grad_norm": 1.8269739315620814,
+      "learning_rate": 1.6608831263869114e-06,
+      "loss": 0.1295,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7417558886509636,
+      "grad_norm": 2.1950940871536835,
+      "learning_rate": 1.6557221432365355e-06,
+      "loss": 0.0964,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7421841541755889,
+      "grad_norm": 1.8438517535104932,
+      "learning_rate": 1.6505675995422178e-06,
+      "loss": 0.1019,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7426124197002142,
+      "grad_norm": 2.189156545934545,
+      "learning_rate": 1.6454195052291066e-06,
+      "loss": 0.1186,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7430406852248393,
+      "grad_norm": 1.8877889387697562,
+      "learning_rate": 1.6402778702099358e-06,
+      "loss": 0.1217,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7434689507494646,
+      "grad_norm": 2.2143401282478448,
+      "learning_rate": 1.6351427043849955e-06,
+      "loss": 0.1598,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7438972162740899,
+      "grad_norm": 2.2341853684461657,
+      "learning_rate": 1.6300140176421242e-06,
+      "loss": 0.0984,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7443254817987152,
+      "grad_norm": 1.8664449347593035,
+      "learning_rate": 1.6248918198566832e-06,
+      "loss": 0.1273,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7447537473233404,
+      "grad_norm": 2.166458767846021,
+      "learning_rate": 1.6197761208915413e-06,
+      "loss": 0.1248,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7451820128479657,
+      "grad_norm": 1.9955047603317966,
+      "learning_rate": 1.6146669305970493e-06,
+      "loss": 0.1328,
+      "step": 1740
+    },
+    {
+      "epoch": 0.745610278372591,
+      "grad_norm": 2.352051880437306,
+      "learning_rate": 1.609564258811029e-06,
+      "loss": 0.1552,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7460385438972162,
+      "grad_norm": 1.8283677076586435,
+      "learning_rate": 1.6044681153587493e-06,
+      "loss": 0.1136,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7464668094218415,
+      "grad_norm": 1.8435761098680266,
+      "learning_rate": 1.5993785100529098e-06,
+      "loss": 0.0916,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7468950749464668,
+      "grad_norm": 2.051866592192121,
+      "learning_rate": 1.5942954526936217e-06,
+      "loss": 0.1338,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7473233404710921,
+      "grad_norm": 1.721339478515566,
+      "learning_rate": 1.5892189530683833e-06,
+      "loss": 0.0845,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7477516059957173,
+      "grad_norm": 1.8101149485628358,
+      "learning_rate": 1.5841490209520705e-06,
+      "loss": 0.1117,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7481798715203426,
+      "grad_norm": 2.2752672863655925,
+      "learning_rate": 1.579085666106912e-06,
+      "loss": 0.1301,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7486081370449679,
+      "grad_norm": 1.8632670516518857,
+      "learning_rate": 1.574028898282472e-06,
+      "loss": 0.0965,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7490364025695931,
+      "grad_norm": 1.930925382877709,
+      "learning_rate": 1.568978727215632e-06,
+      "loss": 0.1419,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7494646680942184,
+      "grad_norm": 2.486978577709939,
+      "learning_rate": 1.5639351626305682e-06,
+      "loss": 0.138,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7498929336188437,
+      "grad_norm": 2.635567915529565,
+      "learning_rate": 1.55889821423874e-06,
+      "loss": 0.1049,
+      "step": 1751
+    },
+    {
+      "epoch": 0.750321199143469,
+      "grad_norm": 1.8717202173073386,
+      "learning_rate": 1.5538678917388638e-06,
+      "loss": 0.0876,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7507494646680942,
+      "grad_norm": 1.8652227144634033,
+      "learning_rate": 1.5488442048169022e-06,
+      "loss": 0.1,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7511777301927195,
+      "grad_norm": 2.4602313839235306,
+      "learning_rate": 1.5438271631460366e-06,
+      "loss": 0.0893,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7516059957173448,
+      "grad_norm": 1.7681910401272023,
+      "learning_rate": 1.5388167763866518e-06,
+      "loss": 0.0851,
+      "step": 1755
+    },
+    {
+      "epoch": 0.75203426124197,
+      "grad_norm": 2.003924197374373,
+      "learning_rate": 1.5338130541863233e-06,
+      "loss": 0.1113,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7524625267665953,
+      "grad_norm": 2.250763815734399,
+      "learning_rate": 1.5288160061797907e-06,
+      "loss": 0.1024,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7528907922912206,
+      "grad_norm": 2.256567995878781,
+      "learning_rate": 1.5238256419889447e-06,
+      "loss": 0.1776,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7533190578158458,
+      "grad_norm": 2.2383463361666154,
+      "learning_rate": 1.5188419712228053e-06,
+      "loss": 0.0866,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7537473233404711,
+      "grad_norm": 2.6892950344387465,
+      "learning_rate": 1.5138650034775004e-06,
+      "loss": 0.1262,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7541755888650964,
+      "grad_norm": 2.3951768682204446,
+      "learning_rate": 1.5088947483362559e-06,
+      "loss": 0.1199,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7546038543897217,
+      "grad_norm": 2.4850054411974907,
+      "learning_rate": 1.503931215369372e-06,
+      "loss": 0.1511,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7550321199143469,
+      "grad_norm": 2.21750798993055,
+      "learning_rate": 1.4989744141342027e-06,
+      "loss": 0.1093,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7554603854389722,
+      "grad_norm": 1.9529231275466559,
+      "learning_rate": 1.4940243541751449e-06,
+      "loss": 0.09,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7558886509635975,
+      "grad_norm": 1.6023687978024914,
+      "learning_rate": 1.4890810450236064e-06,
+      "loss": 0.0536,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7563169164882227,
+      "grad_norm": 2.4872490085114323,
+      "learning_rate": 1.4841444961980045e-06,
+      "loss": 0.1194,
+      "step": 1766
+    },
+    {
+      "epoch": 0.756745182012848,
+      "grad_norm": 2.010014949461149,
+      "learning_rate": 1.4792147172037363e-06,
+      "loss": 0.094,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7571734475374733,
+      "grad_norm": 2.3184561821883896,
+      "learning_rate": 1.4742917175331644e-06,
+      "loss": 0.11,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7576017130620984,
+      "grad_norm": 2.2783848941760554,
+      "learning_rate": 1.469375506665594e-06,
+      "loss": 0.1183,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7580299785867237,
+      "grad_norm": 1.763052498545839,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.1025,
+      "step": 1770
+    },
+    {
+      "epoch": 0.758458244111349,
+      "grad_norm": 1.9842816196732285,
+      "learning_rate": 1.4595634891913168e-06,
+      "loss": 0.0725,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7588865096359743,
+      "grad_norm": 2.0447480015172723,
+      "learning_rate": 1.4546677014777938e-06,
+      "loss": 0.0958,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7593147751605995,
+      "grad_norm": 2.262806582270925,
+      "learning_rate": 1.4497787403536062e-06,
+      "loss": 0.1239,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7597430406852248,
+      "grad_norm": 1.8933151331053244,
+      "learning_rate": 1.4448966152325171e-06,
+      "loss": 0.0653,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7601713062098501,
+      "grad_norm": 2.0146708515656337,
+      "learning_rate": 1.4400213355151322e-06,
+      "loss": 0.0952,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7605995717344753,
+      "grad_norm": 2.013628258479325,
+      "learning_rate": 1.4351529105888735e-06,
+      "loss": 0.0949,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7610278372591006,
+      "grad_norm": 1.9543213747933397,
+      "learning_rate": 1.4302913498279658e-06,
+      "loss": 0.0794,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7614561027837259,
+      "grad_norm": 1.9641205817760612,
+      "learning_rate": 1.425436662593413e-06,
+      "loss": 0.1516,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7618843683083512,
+      "grad_norm": 2.0650269375740224,
+      "learning_rate": 1.4205888582329885e-06,
+      "loss": 0.0848,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7623126338329764,
+      "grad_norm": 2.255423463507378,
+      "learning_rate": 1.41574794608121e-06,
+      "loss": 0.1216,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7627408993576017,
+      "grad_norm": 2.0548811642508293,
+      "learning_rate": 1.4109139354593254e-06,
+      "loss": 0.0972,
+      "step": 1781
+    },
+    {
+      "epoch": 0.763169164882227,
+      "grad_norm": 2.3670839338918173,
+      "learning_rate": 1.4060868356752938e-06,
+      "loss": 0.151,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7635974304068522,
+      "grad_norm": 1.8956455315298428,
+      "learning_rate": 1.4012666560237648e-06,
+      "loss": 0.1107,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7640256959314775,
+      "grad_norm": 2.468969069898418,
+      "learning_rate": 1.3964534057860652e-06,
+      "loss": 0.116,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7644539614561028,
+      "grad_norm": 2.1578263898549066,
+      "learning_rate": 1.3916470942301796e-06,
+      "loss": 0.1435,
+      "step": 1785
+    },
+    {
+      "epoch": 0.764882226980728,
+      "grad_norm": 1.9541264611393145,
+      "learning_rate": 1.386847730610732e-06,
+      "loss": 0.0959,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7653104925053533,
+      "grad_norm": 1.7846971093702078,
+      "learning_rate": 1.3820553241689676e-06,
+      "loss": 0.0878,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7657387580299786,
+      "grad_norm": 2.094909679814358,
+      "learning_rate": 1.3772698841327347e-06,
+      "loss": 0.1028,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7661670235546039,
+      "grad_norm": 2.296731513632592,
+      "learning_rate": 1.3724914197164663e-06,
+      "loss": 0.0901,
+      "step": 1789
+    },
+    {
+      "epoch": 0.7665952890792291,
+      "grad_norm": 1.6033581370107242,
+      "learning_rate": 1.3677199401211672e-06,
+      "loss": 0.122,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7670235546038544,
+      "grad_norm": 1.8782834749396633,
+      "learning_rate": 1.3629554545343921e-06,
+      "loss": 0.0777,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7674518201284797,
+      "grad_norm": 1.8934125271472486,
+      "learning_rate": 1.3581979721302286e-06,
+      "loss": 0.0951,
+      "step": 1792
+    },
+    {
+      "epoch": 0.767880085653105,
+      "grad_norm": 2.4076333979495033,
+      "learning_rate": 1.353447502069276e-06,
+      "loss": 0.1486,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7683083511777302,
+      "grad_norm": 1.767505137201018,
+      "learning_rate": 1.3487040534986357e-06,
+      "loss": 0.0729,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7687366167023555,
+      "grad_norm": 1.7914545786146243,
+      "learning_rate": 1.343967635551887e-06,
+      "loss": 0.085,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7691648822269808,
+      "grad_norm": 2.5785569531127126,
+      "learning_rate": 1.339238257349073e-06,
+      "loss": 0.1229,
+      "step": 1796
+    },
+    {
+      "epoch": 0.769593147751606,
+      "grad_norm": 1.8750331010335577,
+      "learning_rate": 1.3345159279966818e-06,
+      "loss": 0.098,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7700214132762313,
+      "grad_norm": 1.9105900786650818,
+      "learning_rate": 1.3298006565876249e-06,
+      "loss": 0.1009,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7704496788008566,
+      "grad_norm": 1.9447894468501432,
+      "learning_rate": 1.325092452201228e-06,
+      "loss": 0.1178,
+      "step": 1799
+    },
+    {
+      "epoch": 0.7708779443254818,
+      "grad_norm": 2.5318126054021057,
+      "learning_rate": 1.3203913239032074e-06,
+      "loss": 0.1146,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7713062098501071,
+      "grad_norm": 1.8833297529866002,
+      "learning_rate": 1.3156972807456552e-06,
+      "loss": 0.0869,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7717344753747324,
+      "grad_norm": 2.283601727260495,
+      "learning_rate": 1.3110103317670215e-06,
+      "loss": 0.1099,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7721627408993575,
+      "grad_norm": 2.648404382367474,
+      "learning_rate": 1.3063304859920916e-06,
+      "loss": 0.1745,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7725910064239828,
+      "grad_norm": 1.9174319458420492,
+      "learning_rate": 1.30165775243198e-06,
+      "loss": 0.1253,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7730192719486081,
+      "grad_norm": 2.4934396112146855,
+      "learning_rate": 1.2969921400841029e-06,
+      "loss": 0.1727,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7734475374732334,
+      "grad_norm": 1.8798329812876906,
+      "learning_rate": 1.292333657932167e-06,
+      "loss": 0.0751,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7738758029978586,
+      "grad_norm": 2.819012380564757,
+      "learning_rate": 1.2876823149461453e-06,
+      "loss": 0.1231,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7743040685224839,
+      "grad_norm": 1.9578530920160313,
+      "learning_rate": 1.283038120082268e-06,
+      "loss": 0.1215,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7747323340471092,
+      "grad_norm": 1.7190605112399926,
+      "learning_rate": 1.278401082283e-06,
+      "loss": 0.1052,
+      "step": 1809
+    },
+    {
+      "epoch": 0.7751605995717344,
+      "grad_norm": 2.0236591689857404,
+      "learning_rate": 1.2737712104770267e-06,
+      "loss": 0.09,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7755888650963597,
+      "grad_norm": 2.094163105104976,
+      "learning_rate": 1.2691485135792341e-06,
+      "loss": 0.1204,
+      "step": 1811
+    },
+    {
+      "epoch": 0.776017130620985,
+      "grad_norm": 2.6259548311998846,
+      "learning_rate": 1.2645330004906919e-06,
+      "loss": 0.1052,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7764453961456103,
+      "grad_norm": 1.7554928791494102,
+      "learning_rate": 1.2599246800986382e-06,
+      "loss": 0.0789,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7768736616702355,
+      "grad_norm": 1.8111625390635668,
+      "learning_rate": 1.255323561276462e-06,
+      "loss": 0.1159,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7773019271948608,
+      "grad_norm": 1.8891161677536685,
+      "learning_rate": 1.2507296528836848e-06,
+      "loss": 0.0915,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7777301927194861,
+      "grad_norm": 2.108688887017527,
+      "learning_rate": 1.2461429637659466e-06,
+      "loss": 0.0904,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7781584582441113,
+      "grad_norm": 2.0505858329541065,
+      "learning_rate": 1.2415635027549817e-06,
+      "loss": 0.114,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7785867237687366,
+      "grad_norm": 1.7846660940667833,
+      "learning_rate": 1.23699127866861e-06,
+      "loss": 0.1338,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7790149892933619,
+      "grad_norm": 1.7118563703615908,
+      "learning_rate": 1.2324263003107162e-06,
+      "loss": 0.119,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7794432548179872,
+      "grad_norm": 1.9556237332098654,
+      "learning_rate": 1.2278685764712356e-06,
+      "loss": 0.1465,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7798715203426124,
+      "grad_norm": 2.247066734334015,
+      "learning_rate": 1.2233181159261282e-06,
+      "loss": 0.1183,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7802997858672377,
+      "grad_norm": 1.6912311690013284,
+      "learning_rate": 1.2187749274373744e-06,
+      "loss": 0.063,
+      "step": 1822
+    },
+    {
+      "epoch": 0.780728051391863,
+      "grad_norm": 2.0307382635726063,
+      "learning_rate": 1.2142390197529508e-06,
+      "loss": 0.0853,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7811563169164882,
+      "grad_norm": 2.0550548651355807,
+      "learning_rate": 1.2097104016068146e-06,
+      "loss": 0.124,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7815845824411135,
+      "grad_norm": 2.0652802972379134,
+      "learning_rate": 1.2051890817188843e-06,
+      "loss": 0.1068,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7820128479657388,
+      "grad_norm": 1.600910605247112,
+      "learning_rate": 1.2006750687950302e-06,
+      "loss": 0.081,
+      "step": 1826
+    },
+    {
+      "epoch": 0.782441113490364,
+      "grad_norm": 2.0525618338856466,
+      "learning_rate": 1.1961683715270478e-06,
+      "loss": 0.1153,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7828693790149893,
+      "grad_norm": 1.7092545702679034,
+      "learning_rate": 1.1916689985926494e-06,
+      "loss": 0.0906,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7832976445396146,
+      "grad_norm": 2.06346770008882,
+      "learning_rate": 1.187176958655445e-06,
+      "loss": 0.1142,
+      "step": 1829
+    },
+    {
+      "epoch": 0.7837259100642399,
+      "grad_norm": 2.058522832327674,
+      "learning_rate": 1.1826922603649221e-06,
+      "loss": 0.1438,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7841541755888651,
+      "grad_norm": 2.0927470184971213,
+      "learning_rate": 1.1782149123564352e-06,
+      "loss": 0.0979,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7845824411134904,
+      "grad_norm": 2.155967013451903,
+      "learning_rate": 1.1737449232511799e-06,
+      "loss": 0.1113,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7850107066381157,
+      "grad_norm": 1.9782844425899269,
+      "learning_rate": 1.1692823016561882e-06,
+      "loss": 0.1067,
+      "step": 1833
+    },
+    {
+      "epoch": 0.785438972162741,
+      "grad_norm": 2.0120273473544055,
+      "learning_rate": 1.1648270561643027e-06,
+      "loss": 0.0813,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7858672376873662,
+      "grad_norm": 2.419479897513306,
+      "learning_rate": 1.1603791953541654e-06,
+      "loss": 0.091,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7862955032119915,
+      "grad_norm": 2.3913353658943035,
+      "learning_rate": 1.1559387277901958e-06,
+      "loss": 0.1026,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7867237687366168,
+      "grad_norm": 2.0879558880659577,
+      "learning_rate": 1.151505662022579e-06,
+      "loss": 0.1008,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7871520342612419,
+      "grad_norm": 1.8565560279397697,
+      "learning_rate": 1.1470800065872489e-06,
+      "loss": 0.0811,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7875802997858672,
+      "grad_norm": 2.64402884528623,
+      "learning_rate": 1.14266177000587e-06,
+      "loss": 0.1332,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7880085653104925,
+      "grad_norm": 2.142953592819258,
+      "learning_rate": 1.1382509607858233e-06,
+      "loss": 0.1011,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7884368308351177,
+      "grad_norm": 2.0328024460899092,
+      "learning_rate": 1.1338475874201838e-06,
+      "loss": 0.1115,
+      "step": 1841
+    },
+    {
+      "epoch": 0.788865096359743,
+      "grad_norm": 1.7720190521213297,
+      "learning_rate": 1.1294516583877125e-06,
+      "loss": 0.0876,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7892933618843683,
+      "grad_norm": 1.8622923440072736,
+      "learning_rate": 1.1250631821528351e-06,
+      "loss": 0.0675,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7897216274089935,
+      "grad_norm": 1.88678614948603,
+      "learning_rate": 1.1206821671656277e-06,
+      "loss": 0.1389,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7901498929336188,
+      "grad_norm": 2.0096608127950892,
+      "learning_rate": 1.1163086218617997e-06,
+      "loss": 0.0933,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7905781584582441,
+      "grad_norm": 1.6611244358760333,
+      "learning_rate": 1.1119425546626738e-06,
+      "loss": 0.0825,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7910064239828694,
+      "grad_norm": 2.011919648223488,
+      "learning_rate": 1.1075839739751782e-06,
+      "loss": 0.1056,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7914346895074946,
+      "grad_norm": 2.222412999453805,
+      "learning_rate": 1.1032328881918237e-06,
+      "loss": 0.0882,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7918629550321199,
+      "grad_norm": 2.121801008077988,
+      "learning_rate": 1.0988893056906912e-06,
+      "loss": 0.104,
+      "step": 1849
+    },
+    {
+      "epoch": 0.7922912205567452,
+      "grad_norm": 1.8827937967594641,
+      "learning_rate": 1.0945532348354104e-06,
+      "loss": 0.1229,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7927194860813704,
+      "grad_norm": 2.0502986070003533,
+      "learning_rate": 1.0902246839751497e-06,
+      "loss": 0.1118,
+      "step": 1851
+    },
+    {
+      "epoch": 0.7931477516059957,
+      "grad_norm": 2.069046166847525,
+      "learning_rate": 1.0859036614445977e-06,
+      "loss": 0.098,
+      "step": 1852
+    },
+    {
+      "epoch": 0.793576017130621,
+      "grad_norm": 2.4765848775075185,
+      "learning_rate": 1.081590175563947e-06,
+      "loss": 0.1244,
+      "step": 1853
+    },
+    {
+      "epoch": 0.7940042826552463,
+      "grad_norm": 1.8034344768137773,
+      "learning_rate": 1.0772842346388784e-06,
+      "loss": 0.1039,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7944325481798715,
+      "grad_norm": 2.2138144906073167,
+      "learning_rate": 1.0729858469605426e-06,
+      "loss": 0.1074,
+      "step": 1855
+    },
+    {
+      "epoch": 0.7948608137044968,
+      "grad_norm": 2.2384560464807555,
+      "learning_rate": 1.0686950208055486e-06,
+      "loss": 0.1464,
+      "step": 1856
+    },
+    {
+      "epoch": 0.7952890792291221,
+      "grad_norm": 2.014586629825217,
+      "learning_rate": 1.0644117644359452e-06,
+      "loss": 0.0965,
+      "step": 1857
+    },
+    {
+      "epoch": 0.7957173447537473,
+      "grad_norm": 2.01239910302297,
+      "learning_rate": 1.060136086099206e-06,
+      "loss": 0.0865,
+      "step": 1858
+    },
+    {
+      "epoch": 0.7961456102783726,
+      "grad_norm": 2.5265962293711848,
+      "learning_rate": 1.0558679940282135e-06,
+      "loss": 0.1125,
+      "step": 1859
+    },
+    {
+      "epoch": 0.7965738758029979,
+      "grad_norm": 2.0499483468647277,
+      "learning_rate": 1.05160749644124e-06,
+      "loss": 0.1162,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7970021413276231,
+      "grad_norm": 1.9762013277461732,
+      "learning_rate": 1.0473546015419345e-06,
+      "loss": 0.1015,
+      "step": 1861
+    },
+    {
+      "epoch": 0.7974304068522484,
+      "grad_norm": 1.5696627595665038,
+      "learning_rate": 1.0431093175193102e-06,
+      "loss": 0.0543,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7978586723768737,
+      "grad_norm": 1.9689647662038137,
+      "learning_rate": 1.038871652547724e-06,
+      "loss": 0.1007,
+      "step": 1863
+    },
+    {
+      "epoch": 0.798286937901499,
+      "grad_norm": 1.8092806530889731,
+      "learning_rate": 1.034641614786862e-06,
+      "loss": 0.0882,
+      "step": 1864
+    },
+    {
+      "epoch": 0.7987152034261242,
+      "grad_norm": 2.115704368463057,
+      "learning_rate": 1.030419212381723e-06,
+      "loss": 0.0865,
+      "step": 1865
+    },
+    {
+      "epoch": 0.7991434689507495,
+      "grad_norm": 1.9790739607925858,
+      "learning_rate": 1.0262044534626053e-06,
+      "loss": 0.1348,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7995717344753748,
+      "grad_norm": 1.9095083834545175,
+      "learning_rate": 1.021997346145089e-06,
+      "loss": 0.1011,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1820242501957243,
+      "learning_rate": 1.0177978985300203e-06,
+      "loss": 0.0995,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8004282655246253,
+      "grad_norm": 2.023002449095124,
+      "learning_rate": 1.0136061187034995e-06,
+      "loss": 0.1277,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8008565310492506,
+      "grad_norm": 1.9219686882657994,
+      "learning_rate": 1.009422014736857e-06,
+      "loss": 0.1048,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8012847965738759,
+      "grad_norm": 2.2544770963846847,
+      "learning_rate": 1.0052455946866485e-06,
+      "loss": 0.0958,
+      "step": 1871
+    },
+    {
+      "epoch": 0.801713062098501,
+      "grad_norm": 1.8964107790958145,
+      "learning_rate": 1.0010768665946309e-06,
+      "loss": 0.0937,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8021413276231263,
+      "grad_norm": 1.8450646982543988,
+      "learning_rate": 9.969158384877514e-07,
+      "loss": 0.0996,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8025695931477516,
+      "grad_norm": 1.736468273207795,
+      "learning_rate": 9.927625183781331e-07,
+      "loss": 0.0656,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8029978586723768,
+      "grad_norm": 1.707548209247148,
+      "learning_rate": 9.886169142630502e-07,
+      "loss": 0.0922,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8034261241970021,
+      "grad_norm": 2.0101444493792853,
+      "learning_rate": 9.844790341249276e-07,
+      "loss": 0.1026,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8038543897216274,
+      "grad_norm": 1.6293606099279072,
+      "learning_rate": 9.80348885931313e-07,
+      "loss": 0.0639,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8042826552462526,
+      "grad_norm": 2.1737251116168554,
+      "learning_rate": 9.76226477634869e-07,
+      "loss": 0.0915,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8047109207708779,
+      "grad_norm": 1.7504081085656953,
+      "learning_rate": 9.721118171733501e-07,
+      "loss": 0.0778,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8051391862955032,
+      "grad_norm": 1.8937427682243944,
+      "learning_rate": 9.680049124695973e-07,
+      "loss": 0.1117,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8055674518201285,
+      "grad_norm": 2.08616472868548,
+      "learning_rate": 9.63905771431516e-07,
+      "loss": 0.1098,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8059957173447537,
+      "grad_norm": 2.5461324119276614,
+      "learning_rate": 9.598144019520612e-07,
+      "loss": 0.119,
+      "step": 1882
+    },
+    {
+      "epoch": 0.806423982869379,
+      "grad_norm": 2.048394642808668,
+      "learning_rate": 9.557308119092273e-07,
+      "loss": 0.0996,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8068522483940043,
+      "grad_norm": 1.836268000691859,
+      "learning_rate": 9.516550091660237e-07,
+      "loss": 0.1008,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8072805139186295,
+      "grad_norm": 1.9857820910974286,
+      "learning_rate": 9.475870015704702e-07,
+      "loss": 0.0846,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8077087794432548,
+      "grad_norm": 1.816660437977218,
+      "learning_rate": 9.435267969555745e-07,
+      "loss": 0.0899,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8081370449678801,
+      "grad_norm": 2.4874888547965277,
+      "learning_rate": 9.394744031393199e-07,
+      "loss": 0.1119,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8085653104925054,
+      "grad_norm": 2.1662911855228493,
+      "learning_rate": 9.35429827924652e-07,
+      "loss": 0.1734,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8089935760171306,
+      "grad_norm": 2.257394738013523,
+      "learning_rate": 9.313930790994574e-07,
+      "loss": 0.0889,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8094218415417559,
+      "grad_norm": 1.9219203012722352,
+      "learning_rate": 9.273641644365555e-07,
+      "loss": 0.0746,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8098501070663812,
+      "grad_norm": 2.215101139707443,
+      "learning_rate": 9.233430916936809e-07,
+      "loss": 0.0746,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8102783725910064,
+      "grad_norm": 1.9670051245059956,
+      "learning_rate": 9.193298686134699e-07,
+      "loss": 0.1325,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8107066381156317,
+      "grad_norm": 2.0626924580548804,
+      "learning_rate": 9.153245029234392e-07,
+      "loss": 0.115,
+      "step": 1893
+    },
+    {
+      "epoch": 0.811134903640257,
+      "grad_norm": 1.8966997723076204,
+      "learning_rate": 9.113270023359827e-07,
+      "loss": 0.08,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8115631691648822,
+      "grad_norm": 1.9641484591927383,
+      "learning_rate": 9.073373745483427e-07,
+      "loss": 0.0843,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8119914346895075,
+      "grad_norm": 1.99412673529411,
+      "learning_rate": 9.033556272426075e-07,
+      "loss": 0.1017,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8124197002141328,
+      "grad_norm": 2.171509091172803,
+      "learning_rate": 8.993817680856909e-07,
+      "loss": 0.1135,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8128479657387581,
+      "grad_norm": 1.8978924822469174,
+      "learning_rate": 8.954158047293172e-07,
+      "loss": 0.079,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8132762312633833,
+      "grad_norm": 2.019341829955686,
+      "learning_rate": 8.914577448100043e-07,
+      "loss": 0.1179,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8137044967880086,
+      "grad_norm": 2.1183371844494703,
+      "learning_rate": 8.87507595949057e-07,
+      "loss": 0.0993,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8141327623126339,
+      "grad_norm": 2.123017959132474,
+      "learning_rate": 8.835653657525439e-07,
+      "loss": 0.0807,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8145610278372591,
+      "grad_norm": 2.846661211749426,
+      "learning_rate": 8.79631061811288e-07,
+      "loss": 0.1687,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8149892933618844,
+      "grad_norm": 1.9749421709164268,
+      "learning_rate": 8.757046917008494e-07,
+      "loss": 0.0728,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8154175588865097,
+      "grad_norm": 2.0437323188817316,
+      "learning_rate": 8.717862629815099e-07,
+      "loss": 0.0902,
+      "step": 1904
+    },
+    {
+      "epoch": 0.815845824411135,
+      "grad_norm": 2.1825165349040065,
+      "learning_rate": 8.678757831982615e-07,
+      "loss": 0.1065,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8162740899357601,
+      "grad_norm": 1.9109502566238636,
+      "learning_rate": 8.639732598807915e-07,
+      "loss": 0.104,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8167023554603854,
+      "grad_norm": 2.193842720755467,
+      "learning_rate": 8.600787005434663e-07,
+      "loss": 0.0814,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8171306209850107,
+      "grad_norm": 2.08508527032698,
+      "learning_rate": 8.561921126853151e-07,
+      "loss": 0.1119,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8175588865096359,
+      "grad_norm": 1.8992076815816814,
+      "learning_rate": 8.523135037900205e-07,
+      "loss": 0.0759,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8179871520342612,
+      "grad_norm": 2.4234358112189116,
+      "learning_rate": 8.48442881325901e-07,
+      "loss": 0.0926,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8184154175588865,
+      "grad_norm": 2.114700254572066,
+      "learning_rate": 8.445802527458969e-07,
+      "loss": 0.1328,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8188436830835117,
+      "grad_norm": 1.694382184487371,
+      "learning_rate": 8.407256254875573e-07,
+      "loss": 0.1084,
+      "step": 1912
+    },
+    {
+      "epoch": 0.819271948608137,
+      "grad_norm": 2.219486414199205,
+      "learning_rate": 8.368790069730221e-07,
+      "loss": 0.1712,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8197002141327623,
+      "grad_norm": 1.7275442144819615,
+      "learning_rate": 8.330404046090112e-07,
+      "loss": 0.0903,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8201284796573876,
+      "grad_norm": 1.9721616823471808,
+      "learning_rate": 8.292098257868109e-07,
+      "loss": 0.1037,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8205567451820128,
+      "grad_norm": 1.9252496561882384,
+      "learning_rate": 8.253872778822564e-07,
+      "loss": 0.1021,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8209850107066381,
+      "grad_norm": 1.9723532501893495,
+      "learning_rate": 8.215727682557212e-07,
+      "loss": 0.1027,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8214132762312634,
+      "grad_norm": 2.282480242384323,
+      "learning_rate": 8.177663042520972e-07,
+      "loss": 0.0841,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8218415417558886,
+      "grad_norm": 4.298702946299436,
+      "learning_rate": 8.139678932007877e-07,
+      "loss": 0.146,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8222698072805139,
+      "grad_norm": 2.19640029046654,
+      "learning_rate": 8.101775424156888e-07,
+      "loss": 0.1353,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8226980728051392,
+      "grad_norm": 1.949703746099148,
+      "learning_rate": 8.063952591951773e-07,
+      "loss": 0.1089,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8231263383297645,
+      "grad_norm": 2.129973325828584,
+      "learning_rate": 8.026210508220938e-07,
+      "loss": 0.0766,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8235546038543897,
+      "grad_norm": 1.7825459111583906,
+      "learning_rate": 7.988549245637322e-07,
+      "loss": 0.048,
+      "step": 1923
+    },
+    {
+      "epoch": 0.823982869379015,
+      "grad_norm": 2.1920427625434007,
+      "learning_rate": 7.95096887671824e-07,
+      "loss": 0.1362,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8244111349036403,
+      "grad_norm": 1.6327350358753876,
+      "learning_rate": 7.913469473825247e-07,
+      "loss": 0.0971,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8248394004282655,
+      "grad_norm": 1.7117946692752706,
+      "learning_rate": 7.876051109163995e-07,
+      "loss": 0.0851,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8252676659528908,
+      "grad_norm": 1.684965924202361,
+      "learning_rate": 7.83871385478408e-07,
+      "loss": 0.0605,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8256959314775161,
+      "grad_norm": 1.7709900781103327,
+      "learning_rate": 7.801457782578947e-07,
+      "loss": 0.0803,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8261241970021413,
+      "grad_norm": 2.127378516918009,
+      "learning_rate": 7.764282964285697e-07,
+      "loss": 0.1069,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8265524625267666,
+      "grad_norm": 2.2790486145009146,
+      "learning_rate": 7.727189471485013e-07,
+      "loss": 0.1016,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8269807280513919,
+      "grad_norm": 1.835719352009276,
+      "learning_rate": 7.690177375600926e-07,
+      "loss": 0.1414,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8274089935760172,
+      "grad_norm": 2.1570530487403983,
+      "learning_rate": 7.653246747900794e-07,
+      "loss": 0.1696,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8278372591006424,
+      "grad_norm": 2.187980518601719,
+      "learning_rate": 7.616397659495068e-07,
+      "loss": 0.0954,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8282655246252677,
+      "grad_norm": 2.0100988038246874,
+      "learning_rate": 7.579630181337205e-07,
+      "loss": 0.1007,
+      "step": 1934
+    },
+    {
+      "epoch": 0.828693790149893,
+      "grad_norm": 2.2211207432922264,
+      "learning_rate": 7.542944384223539e-07,
+      "loss": 0.0898,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8291220556745182,
+      "grad_norm": 2.212920810754019,
+      "learning_rate": 7.506340338793111e-07,
+      "loss": 0.1285,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8295503211991435,
+      "grad_norm": 1.7723207851568428,
+      "learning_rate": 7.469818115527527e-07,
+      "loss": 0.1121,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8299785867237688,
+      "grad_norm": 2.034060601007009,
+      "learning_rate": 7.433377784750878e-07,
+      "loss": 0.1425,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8304068522483941,
+      "grad_norm": 1.5209190464795752,
+      "learning_rate": 7.397019416629553e-07,
+      "loss": 0.0857,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8308351177730193,
+      "grad_norm": 2.5199025788418434,
+      "learning_rate": 7.360743081172122e-07,
+      "loss": 0.1242,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8312633832976445,
+      "grad_norm": 1.966030374764833,
+      "learning_rate": 7.324548848229213e-07,
+      "loss": 0.0986,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8316916488222698,
+      "grad_norm": 1.7865225364012225,
+      "learning_rate": 7.28843678749333e-07,
+      "loss": 0.0669,
+      "step": 1942
+    },
+    {
+      "epoch": 0.832119914346895,
+      "grad_norm": 2.0289644584903117,
+      "learning_rate": 7.252406968498788e-07,
+      "loss": 0.1068,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8325481798715203,
+      "grad_norm": 1.939397957108772,
+      "learning_rate": 7.216459460621528e-07,
+      "loss": 0.0995,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8329764453961456,
+      "grad_norm": 2.287472829544063,
+      "learning_rate": 7.180594333079005e-07,
+      "loss": 0.1398,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8334047109207708,
+      "grad_norm": 1.8379209275514834,
+      "learning_rate": 7.144811654930067e-07,
+      "loss": 0.1153,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8338329764453961,
+      "grad_norm": 2.464208763412115,
+      "learning_rate": 7.10911149507475e-07,
+      "loss": 0.0983,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8342612419700214,
+      "grad_norm": 2.0272951387226255,
+      "learning_rate": 7.073493922254254e-07,
+      "loss": 0.1058,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8346895074946467,
+      "grad_norm": 2.06829398179105,
+      "learning_rate": 7.037959005050743e-07,
+      "loss": 0.1143,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8351177730192719,
+      "grad_norm": 1.946880833708518,
+      "learning_rate": 7.002506811887222e-07,
+      "loss": 0.0973,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8355460385438972,
+      "grad_norm": 1.9748205355409012,
+      "learning_rate": 6.967137411027392e-07,
+      "loss": 0.0851,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8359743040685225,
+      "grad_norm": 2.083871104740711,
+      "learning_rate": 6.931850870575563e-07,
+      "loss": 0.0894,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8364025695931477,
+      "grad_norm": 1.9763734237843116,
+      "learning_rate": 6.896647258476485e-07,
+      "loss": 0.0721,
+      "step": 1953
+    },
+    {
+      "epoch": 0.836830835117773,
+      "grad_norm": 1.832221330425259,
+      "learning_rate": 6.861526642515232e-07,
+      "loss": 0.1082,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8372591006423983,
+      "grad_norm": 2.2860739213255097,
+      "learning_rate": 6.826489090317073e-07,
+      "loss": 0.1102,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8376873661670236,
+      "grad_norm": 1.9830749656988804,
+      "learning_rate": 6.791534669347311e-07,
+      "loss": 0.126,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8381156316916488,
+      "grad_norm": 2.061630391200355,
+      "learning_rate": 6.7566634469112e-07,
+      "loss": 0.0873,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8385438972162741,
+      "grad_norm": 1.7558476540711567,
+      "learning_rate": 6.721875490153795e-07,
+      "loss": 0.1282,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8389721627408994,
+      "grad_norm": 1.8241831622850095,
+      "learning_rate": 6.687170866059822e-07,
+      "loss": 0.0786,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8394004282655246,
+      "grad_norm": 1.8502615044983146,
+      "learning_rate": 6.652549641453543e-07,
+      "loss": 0.0975,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8398286937901499,
+      "grad_norm": 2.2244165047576954,
+      "learning_rate": 6.618011882998621e-07,
+      "loss": 0.1204,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8402569593147752,
+      "grad_norm": 2.548223632354938,
+      "learning_rate": 6.583557657198025e-07,
+      "loss": 0.098,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8406852248394004,
+      "grad_norm": 2.026271114516728,
+      "learning_rate": 6.549187030393872e-07,
+      "loss": 0.096,
+      "step": 1963
+    },
+    {
+      "epoch": 0.8411134903640257,
+      "grad_norm": 1.7206676006564678,
+      "learning_rate": 6.514900068767316e-07,
+      "loss": 0.0666,
+      "step": 1964
+    },
+    {
+      "epoch": 0.841541755888651,
+      "grad_norm": 1.9822799608876212,
+      "learning_rate": 6.480696838338396e-07,
+      "loss": 0.1172,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8419700214132763,
+      "grad_norm": 1.7923469214608037,
+      "learning_rate": 6.446577404965926e-07,
+      "loss": 0.0752,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8423982869379015,
+      "grad_norm": 1.8891195203991555,
+      "learning_rate": 6.41254183434738e-07,
+      "loss": 0.0721,
+      "step": 1967
+    },
+    {
+      "epoch": 0.8428265524625268,
+      "grad_norm": 1.7772833224716837,
+      "learning_rate": 6.378590192018752e-07,
+      "loss": 0.0902,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8432548179871521,
+      "grad_norm": 1.877016430812966,
+      "learning_rate": 6.344722543354426e-07,
+      "loss": 0.0928,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8436830835117773,
+      "grad_norm": 2.0838886627205424,
+      "learning_rate": 6.310938953567069e-07,
+      "loss": 0.0787,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8441113490364026,
+      "grad_norm": 1.6997765105488383,
+      "learning_rate": 6.277239487707453e-07,
+      "loss": 0.1143,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8445396145610279,
+      "grad_norm": 1.937396099546873,
+      "learning_rate": 6.243624210664406e-07,
+      "loss": 0.1143,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8449678800856532,
+      "grad_norm": 2.0508020025396982,
+      "learning_rate": 6.21009318716464e-07,
+      "loss": 0.1276,
+      "step": 1973
+    },
+    {
+      "epoch": 0.8453961456102784,
+      "grad_norm": 2.1396827644870653,
+      "learning_rate": 6.17664648177263e-07,
+      "loss": 0.1024,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8458244111349036,
+      "grad_norm": 2.078583242762166,
+      "learning_rate": 6.143284158890511e-07,
+      "loss": 0.0939,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8462526766595289,
+      "grad_norm": 2.1349109069391536,
+      "learning_rate": 6.110006282757897e-07,
+      "loss": 0.0802,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8466809421841541,
+      "grad_norm": 1.827916302765162,
+      "learning_rate": 6.076812917451847e-07,
+      "loss": 0.0804,
+      "step": 1977
+    },
+    {
+      "epoch": 0.8471092077087794,
+      "grad_norm": 2.1276624337084127,
+      "learning_rate": 6.043704126886663e-07,
+      "loss": 0.1311,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8475374732334047,
+      "grad_norm": 2.1510203730913013,
+      "learning_rate": 6.010679974813821e-07,
+      "loss": 0.1207,
+      "step": 1979
+    },
+    {
+      "epoch": 0.8479657387580299,
+      "grad_norm": 2.0459360795537167,
+      "learning_rate": 5.977740524821796e-07,
+      "loss": 0.1035,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8483940042826552,
+      "grad_norm": 1.4233172491163002,
+      "learning_rate": 5.944885840335978e-07,
+      "loss": 0.0875,
+      "step": 1981
+    },
+    {
+      "epoch": 0.8488222698072805,
+      "grad_norm": 1.8346847877646772,
+      "learning_rate": 5.912115984618555e-07,
+      "loss": 0.0763,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8492505353319058,
+      "grad_norm": 2.031887359806279,
+      "learning_rate": 5.879431020768367e-07,
+      "loss": 0.086,
+      "step": 1983
+    },
+    {
+      "epoch": 0.849678800856531,
+      "grad_norm": 1.8839931362570257,
+      "learning_rate": 5.846831011720789e-07,
+      "loss": 0.1417,
+      "step": 1984
+    },
+    {
+      "epoch": 0.8501070663811563,
+      "grad_norm": 1.8420056397650515,
+      "learning_rate": 5.814316020247607e-07,
+      "loss": 0.0598,
+      "step": 1985
+    },
+    {
+      "epoch": 0.8505353319057816,
+      "grad_norm": 1.8101329917813898,
+      "learning_rate": 5.781886108956919e-07,
+      "loss": 0.066,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8509635974304068,
+      "grad_norm": 1.7967854657813915,
+      "learning_rate": 5.749541340292997e-07,
+      "loss": 0.1047,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8513918629550321,
+      "grad_norm": 1.8532172946512087,
+      "learning_rate": 5.717281776536166e-07,
+      "loss": 0.1038,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8518201284796574,
+      "grad_norm": 1.6370855427016244,
+      "learning_rate": 5.685107479802704e-07,
+      "loss": 0.0809,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8522483940042827,
+      "grad_norm": 1.9054256227394013,
+      "learning_rate": 5.653018512044661e-07,
+      "loss": 0.0749,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8526766595289079,
+      "grad_norm": 1.8406669794438646,
+      "learning_rate": 5.621014935049835e-07,
+      "loss": 0.0777,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8531049250535332,
+      "grad_norm": 2.2606457876096377,
+      "learning_rate": 5.589096810441574e-07,
+      "loss": 0.1111,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8535331905781585,
+      "grad_norm": 2.077670456318267,
+      "learning_rate": 5.557264199678713e-07,
+      "loss": 0.0875,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8539614561027837,
+      "grad_norm": 1.8708565131756,
+      "learning_rate": 5.525517164055388e-07,
+      "loss": 0.1124,
+      "step": 1994
+    },
+    {
+      "epoch": 0.854389721627409,
+      "grad_norm": 2.393587267110176,
+      "learning_rate": 5.493855764700989e-07,
+      "loss": 0.0876,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8548179871520343,
+      "grad_norm": 1.8106038861375993,
+      "learning_rate": 5.462280062580011e-07,
+      "loss": 0.094,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8552462526766595,
+      "grad_norm": 1.8407355289194944,
+      "learning_rate": 5.430790118491924e-07,
+      "loss": 0.1093,
+      "step": 1997
+    },
+    {
+      "epoch": 0.8556745182012848,
+      "grad_norm": 1.9382613524637973,
+      "learning_rate": 5.399385993071093e-07,
+      "loss": 0.1094,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8561027837259101,
+      "grad_norm": 1.7411763091868417,
+      "learning_rate": 5.368067746786598e-07,
+      "loss": 0.0816,
+      "step": 1999
+    },
+    {
+      "epoch": 0.8565310492505354,
+      "grad_norm": 1.8958129914064883,
+      "learning_rate": 5.3368354399422e-07,
+      "loss": 0.0934,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8569593147751606,
+      "grad_norm": 1.9980272821363083,
+      "learning_rate": 5.305689132676157e-07,
+      "loss": 0.0883,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8573875802997859,
+      "grad_norm": 2.0973386670832035,
+      "learning_rate": 5.27462888496113e-07,
+      "loss": 0.1056,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8578158458244112,
+      "grad_norm": 2.146955449693084,
+      "learning_rate": 5.243654756604099e-07,
+      "loss": 0.066,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8582441113490364,
+      "grad_norm": 1.9986276436166284,
+      "learning_rate": 5.212766807246206e-07,
+      "loss": 0.0866,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8586723768736617,
+      "grad_norm": 2.183961245353881,
+      "learning_rate": 5.181965096362624e-07,
+      "loss": 0.0987,
+      "step": 2005
+    },
+    {
+      "epoch": 0.859100642398287,
+      "grad_norm": 2.548661660338589,
+      "learning_rate": 5.151249683262516e-07,
+      "loss": 0.0981,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8595289079229123,
+      "grad_norm": 1.6860186728546704,
+      "learning_rate": 5.120620627088863e-07,
+      "loss": 0.0808,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8599571734475375,
+      "grad_norm": 2.078452825831932,
+      "learning_rate": 5.090077986818365e-07,
+      "loss": 0.0812,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8603854389721627,
+      "grad_norm": 2.09742759637865,
+      "learning_rate": 5.059621821261313e-07,
+      "loss": 0.0957,
+      "step": 2009
+    },
+    {
+      "epoch": 0.860813704496788,
+      "grad_norm": 1.9348007694312541,
+      "learning_rate": 5.029252189061496e-07,
+      "loss": 0.107,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8612419700214132,
+      "grad_norm": 1.6895081399852836,
+      "learning_rate": 4.998969148696098e-07,
+      "loss": 0.0943,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8616702355460385,
+      "grad_norm": 2.219552524451849,
+      "learning_rate": 4.968772758475554e-07,
+      "loss": 0.1011,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8620985010706638,
+      "grad_norm": 2.215781256873524,
+      "learning_rate": 4.938663076543465e-07,
+      "loss": 0.1011,
+      "step": 2013
+    },
+    {
+      "epoch": 0.862526766595289,
+      "grad_norm": 1.9005478596046508,
+      "learning_rate": 4.908640160876449e-07,
+      "loss": 0.0923,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8629550321199143,
+      "grad_norm": 2.0733095447316257,
+      "learning_rate": 4.87870406928408e-07,
+      "loss": 0.0788,
+      "step": 2015
+    },
+    {
+      "epoch": 0.8633832976445396,
+      "grad_norm": 1.6832647020304528,
+      "learning_rate": 4.848854859408731e-07,
+      "loss": 0.0749,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8638115631691649,
+      "grad_norm": 1.6965576529952893,
+      "learning_rate": 4.819092588725505e-07,
+      "loss": 0.0558,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8642398286937901,
+      "grad_norm": 2.133303822940741,
+      "learning_rate": 4.789417314542099e-07,
+      "loss": 0.09,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8646680942184154,
+      "grad_norm": 1.9909533013253011,
+      "learning_rate": 4.7598290939986633e-07,
+      "loss": 0.0789,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8650963597430407,
+      "grad_norm": 2.071691098234863,
+      "learning_rate": 4.7303279840677675e-07,
+      "loss": 0.1311,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8655246252676659,
+      "grad_norm": 2.0137487430883683,
+      "learning_rate": 4.700914041554222e-07,
+      "loss": 0.0798,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8659528907922912,
+      "grad_norm": 1.9586335039388223,
+      "learning_rate": 4.6715873230950225e-07,
+      "loss": 0.0905,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8663811563169165,
+      "grad_norm": 1.6403489127477124,
+      "learning_rate": 4.6423478851591763e-07,
+      "loss": 0.0893,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8668094218415417,
+      "grad_norm": 1.8165373137701237,
+      "learning_rate": 4.613195784047653e-07,
+      "loss": 0.0926,
+      "step": 2024
+    },
+    {
+      "epoch": 0.867237687366167,
+      "grad_norm": 1.8844646256062234,
+      "learning_rate": 4.5841310758932554e-07,
+      "loss": 0.1032,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8676659528907923,
+      "grad_norm": 2.2198107017224795,
+      "learning_rate": 4.555153816660507e-07,
+      "loss": 0.1022,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8680942184154176,
+      "grad_norm": 1.641840021766596,
+      "learning_rate": 4.5262640621455523e-07,
+      "loss": 0.0897,
+      "step": 2027
+    },
+    {
+      "epoch": 0.8685224839400428,
+      "grad_norm": 1.92157561877198,
+      "learning_rate": 4.4974618679760164e-07,
+      "loss": 0.0929,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8689507494646681,
+      "grad_norm": 2.0420624730789423,
+      "learning_rate": 4.4687472896109586e-07,
+      "loss": 0.0805,
+      "step": 2029
+    },
+    {
+      "epoch": 0.8693790149892934,
+      "grad_norm": 1.8606414538364795,
+      "learning_rate": 4.4401203823407236e-07,
+      "loss": 0.0858,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8698072805139186,
+      "grad_norm": 1.874216076449617,
+      "learning_rate": 4.4115812012868296e-07,
+      "loss": 0.1233,
+      "step": 2031
+    },
+    {
+      "epoch": 0.8702355460385439,
+      "grad_norm": 1.582583162096466,
+      "learning_rate": 4.3831298014019144e-07,
+      "loss": 0.0885,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8706638115631692,
+      "grad_norm": 1.7959587520894216,
+      "learning_rate": 4.354766237469532e-07,
+      "loss": 0.0813,
+      "step": 2033
+    },
+    {
+      "epoch": 0.8710920770877945,
+      "grad_norm": 1.8487477465977122,
+      "learning_rate": 4.3264905641041575e-07,
+      "loss": 0.1231,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8715203426124197,
+      "grad_norm": 1.8195282323746182,
+      "learning_rate": 4.298302835751017e-07,
+      "loss": 0.077,
+      "step": 2035
+    },
+    {
+      "epoch": 0.871948608137045,
+      "grad_norm": 2.007478716695311,
+      "learning_rate": 4.2702031066859993e-07,
+      "loss": 0.1087,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8723768736616703,
+      "grad_norm": 1.7046064189181802,
+      "learning_rate": 4.242191431015535e-07,
+      "loss": 0.102,
+      "step": 2037
+    },
+    {
+      "epoch": 0.8728051391862955,
+      "grad_norm": 2.2157170974673495,
+      "learning_rate": 4.214267862676508e-07,
+      "loss": 0.1245,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8732334047109208,
+      "grad_norm": 2.255868146145337,
+      "learning_rate": 4.1864324554361734e-07,
+      "loss": 0.0908,
+      "step": 2039
+    },
+    {
+      "epoch": 0.8736616702355461,
+      "grad_norm": 1.7188642742432156,
+      "learning_rate": 4.1586852628920095e-07,
+      "loss": 0.1006,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8740899357601714,
+      "grad_norm": 2.072245341452134,
+      "learning_rate": 4.1310263384716385e-07,
+      "loss": 0.111,
+      "step": 2041
+    },
+    {
+      "epoch": 0.8745182012847966,
+      "grad_norm": 2.0017984830250284,
+      "learning_rate": 4.1034557354327486e-07,
+      "loss": 0.0981,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8749464668094219,
+      "grad_norm": 1.8358972790099584,
+      "learning_rate": 4.075973506862907e-07,
+      "loss": 0.0735,
+      "step": 2043
+    },
+    {
+      "epoch": 0.8753747323340471,
+      "grad_norm": 1.882155166730242,
+      "learning_rate": 4.0485797056795675e-07,
+      "loss": 0.0946,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8758029978586723,
+      "grad_norm": 2.2608179576613203,
+      "learning_rate": 4.021274384629892e-07,
+      "loss": 0.0894,
+      "step": 2045
+    },
+    {
+      "epoch": 0.8762312633832976,
+      "grad_norm": 1.9924092653294023,
+      "learning_rate": 3.994057596290679e-07,
+      "loss": 0.1026,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8766595289079229,
+      "grad_norm": 1.8073533177815335,
+      "learning_rate": 3.966929393068253e-07,
+      "loss": 0.126,
+      "step": 2047
+    },
+    {
+      "epoch": 0.8770877944325481,
+      "grad_norm": 1.7112988326542204,
+      "learning_rate": 3.939889827198362e-07,
+      "loss": 0.0842,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8775160599571734,
+      "grad_norm": 2.178226640718448,
+      "learning_rate": 3.9129389507460926e-07,
+      "loss": 0.1141,
+      "step": 2049
+    },
+    {
+      "epoch": 0.8779443254817987,
+      "grad_norm": 1.8332129404121462,
+      "learning_rate": 3.886076815605744e-07,
+      "loss": 0.105,
+      "step": 2050
+    },
+    {
+      "epoch": 0.878372591006424,
+      "grad_norm": 1.7078580064386433,
+      "learning_rate": 3.85930347350077e-07,
+      "loss": 0.0633,
+      "step": 2051
+    },
+    {
+      "epoch": 0.8788008565310492,
+      "grad_norm": 2.0508293814320115,
+      "learning_rate": 3.8326189759836097e-07,
+      "loss": 0.1104,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8792291220556745,
+      "grad_norm": 2.105152240241882,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.0936,
+      "step": 2053
+    },
+    {
+      "epoch": 0.8796573875802998,
+      "grad_norm": 1.961177010495811,
+      "learning_rate": 3.7795167200671554e-07,
+      "loss": 0.0861,
+      "step": 2054
+    },
+    {
+      "epoch": 0.880085653104925,
+      "grad_norm": 1.4840235375454818,
+      "learning_rate": 3.7530990639170427e-07,
+      "loss": 0.0705,
+      "step": 2055
+    },
+    {
+      "epoch": 0.8805139186295503,
+      "grad_norm": 1.6142660448715787,
+      "learning_rate": 3.7267704568529015e-07,
+      "loss": 0.0745,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8809421841541756,
+      "grad_norm": 1.8535743895383656,
+      "learning_rate": 3.700530949570852e-07,
+      "loss": 0.0727,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8813704496788008,
+      "grad_norm": 2.272950329321654,
+      "learning_rate": 3.6743805925954446e-07,
+      "loss": 0.1288,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8817987152034261,
+      "grad_norm": 1.8529129016591002,
+      "learning_rate": 3.648319436279585e-07,
+      "loss": 0.1113,
+      "step": 2059
+    },
+    {
+      "epoch": 0.8822269807280514,
+      "grad_norm": 1.8347826480623295,
+      "learning_rate": 3.622347530804415e-07,
+      "loss": 0.0774,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8826552462526767,
+      "grad_norm": 1.6713851025085391,
+      "learning_rate": 3.596464926179205e-07,
+      "loss": 0.0793,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8830835117773019,
+      "grad_norm": 1.9688226009013277,
+      "learning_rate": 3.5706716722412995e-07,
+      "loss": 0.129,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8835117773019272,
+      "grad_norm": 1.9152152541922312,
+      "learning_rate": 3.544967818655981e-07,
+      "loss": 0.0879,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8839400428265525,
+      "grad_norm": 2.0077135310573673,
+      "learning_rate": 3.519353414916404e-07,
+      "loss": 0.0924,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8843683083511777,
+      "grad_norm": 1.8349193881477412,
+      "learning_rate": 3.4938285103434776e-07,
+      "loss": 0.089,
+      "step": 2065
+    },
+    {
+      "epoch": 0.884796573875803,
+      "grad_norm": 1.7973059252971162,
+      "learning_rate": 3.468393154085775e-07,
+      "loss": 0.0799,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8852248394004283,
+      "grad_norm": 2.4472211909604336,
+      "learning_rate": 3.4430473951194477e-07,
+      "loss": 0.0928,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8856531049250536,
+      "grad_norm": 2.2819949754900044,
+      "learning_rate": 3.4177912822481286e-07,
+      "loss": 0.1294,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8860813704496788,
+      "grad_norm": 2.0765435649504225,
+      "learning_rate": 3.392624864102834e-07,
+      "loss": 0.1101,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8865096359743041,
+      "grad_norm": 2.0649771851571845,
+      "learning_rate": 3.3675481891418784e-07,
+      "loss": 0.1091,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8869379014989294,
+      "grad_norm": 2.171492488931347,
+      "learning_rate": 3.34256130565076e-07,
+      "loss": 0.1029,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8873661670235546,
+      "grad_norm": 1.710976497080942,
+      "learning_rate": 3.3176642617420817e-07,
+      "loss": 0.0795,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8877944325481799,
+      "grad_norm": 2.0875695700233567,
+      "learning_rate": 3.292857105355474e-07,
+      "loss": 0.1017,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8882226980728052,
+      "grad_norm": 2.213716997139907,
+      "learning_rate": 3.268139884257476e-07,
+      "loss": 0.1132,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8886509635974305,
+      "grad_norm": 1.864673497417138,
+      "learning_rate": 3.2435126460414746e-07,
+      "loss": 0.0686,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8890792291220557,
+      "grad_norm": 1.8390145305604613,
+      "learning_rate": 3.218975438127558e-07,
+      "loss": 0.1424,
+      "step": 2076
+    },
+    {
+      "epoch": 0.889507494646681,
+      "grad_norm": 2.4136684592502244,
+      "learning_rate": 3.19452830776249e-07,
+      "loss": 0.1114,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8899357601713062,
+      "grad_norm": 2.262458719137513,
+      "learning_rate": 3.1701713020195757e-07,
+      "loss": 0.0945,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8903640256959314,
+      "grad_norm": 1.6345493288849522,
+      "learning_rate": 3.1459044677985883e-07,
+      "loss": 0.0589,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8907922912205567,
+      "grad_norm": 2.0819279896076845,
+      "learning_rate": 3.1217278518256844e-07,
+      "loss": 0.0984,
+      "step": 2080
+    },
+    {
+      "epoch": 0.891220556745182,
+      "grad_norm": 1.6694625726320587,
+      "learning_rate": 3.09764150065327e-07,
+      "loss": 0.0562,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8916488222698072,
+      "grad_norm": 1.8547397444479317,
+      "learning_rate": 3.07364546065998e-07,
+      "loss": 0.0741,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8920770877944325,
+      "grad_norm": 5.1146084696456136,
+      "learning_rate": 3.0497397780505424e-07,
+      "loss": 0.0849,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8925053533190578,
+      "grad_norm": 2.170409448145711,
+      "learning_rate": 3.0259244988556977e-07,
+      "loss": 0.1285,
+      "step": 2084
+    },
+    {
+      "epoch": 0.892933618843683,
+      "grad_norm": 1.6235718358334519,
+      "learning_rate": 3.002199668932121e-07,
+      "loss": 0.052,
+      "step": 2085
+    },
+    {
+      "epoch": 0.8933618843683083,
+      "grad_norm": 1.8625590442939792,
+      "learning_rate": 2.9785653339623075e-07,
+      "loss": 0.0884,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8937901498929336,
+      "grad_norm": 2.25019113317266,
+      "learning_rate": 2.9550215394545277e-07,
+      "loss": 0.0849,
+      "step": 2087
+    },
+    {
+      "epoch": 0.8942184154175589,
+      "grad_norm": 2.397584208447054,
+      "learning_rate": 2.93156833074269e-07,
+      "loss": 0.0866,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8946466809421841,
+      "grad_norm": 1.8105825857220579,
+      "learning_rate": 2.908205752986315e-07,
+      "loss": 0.0827,
+      "step": 2089
+    },
+    {
+      "epoch": 0.8950749464668094,
+      "grad_norm": 1.524161543215394,
+      "learning_rate": 2.884933851170357e-07,
+      "loss": 0.0747,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8955032119914347,
+      "grad_norm": 2.7672765584114964,
+      "learning_rate": 2.8617526701052277e-07,
+      "loss": 0.1215,
+      "step": 2091
+    },
+    {
+      "epoch": 0.89593147751606,
+      "grad_norm": 2.6035646394887206,
+      "learning_rate": 2.8386622544266273e-07,
+      "loss": 0.1157,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8963597430406852,
+      "grad_norm": 1.9285122691266525,
+      "learning_rate": 2.8156626485954796e-07,
+      "loss": 0.0792,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8967880085653105,
+      "grad_norm": 1.8596863420233114,
+      "learning_rate": 2.7927538968978816e-07,
+      "loss": 0.096,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8972162740899358,
+      "grad_norm": 1.9215425221672913,
+      "learning_rate": 2.769936043444954e-07,
+      "loss": 0.1086,
+      "step": 2095
+    },
+    {
+      "epoch": 0.897644539614561,
+      "grad_norm": 2.0725220057443794,
+      "learning_rate": 2.7472091321728067e-07,
+      "loss": 0.1106,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8980728051391863,
+      "grad_norm": 1.5267823693891664,
+      "learning_rate": 2.7245732068424567e-07,
+      "loss": 0.0637,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8985010706638116,
+      "grad_norm": 1.9721064630149536,
+      "learning_rate": 2.702028311039695e-07,
+      "loss": 0.0872,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8989293361884368,
+      "grad_norm": 1.729438748713927,
+      "learning_rate": 2.679574488175063e-07,
+      "loss": 0.0978,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8993576017130621,
+      "grad_norm": 1.7585645079915428,
+      "learning_rate": 2.6572117814837096e-07,
+      "loss": 0.0893,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8997858672376874,
+      "grad_norm": 1.9261318323659733,
+      "learning_rate": 2.634940234025368e-07,
+      "loss": 0.0805,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9002141327623127,
+      "grad_norm": 1.7034514906040545,
+      "learning_rate": 2.612759888684219e-07,
+      "loss": 0.0778,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9006423982869379,
+      "grad_norm": 1.6390722670476559,
+      "learning_rate": 2.5906707881688587e-07,
+      "loss": 0.0914,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9010706638115632,
+      "grad_norm": 2.053204052918339,
+      "learning_rate": 2.568672975012154e-07,
+      "loss": 0.1044,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9014989293361885,
+      "grad_norm": 1.7051884823011207,
+      "learning_rate": 2.546766491571229e-07,
+      "loss": 0.0937,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9019271948608137,
+      "grad_norm": 1.7543377574670076,
+      "learning_rate": 2.524951380027335e-07,
+      "loss": 0.0893,
+      "step": 2106
+    },
+    {
+      "epoch": 0.902355460385439,
+      "grad_norm": 2.365836333683227,
+      "learning_rate": 2.5032276823858003e-07,
+      "loss": 0.119,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9027837259100643,
+      "grad_norm": 2.0145582873798995,
+      "learning_rate": 2.4815954404759034e-07,
+      "loss": 0.1257,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9032119914346896,
+      "grad_norm": 1.8237118226732931,
+      "learning_rate": 2.4600546959508585e-07,
+      "loss": 0.1138,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9036402569593148,
+      "grad_norm": 2.3319755733358294,
+      "learning_rate": 2.4386054902876767e-07,
+      "loss": 0.0972,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9040685224839401,
+      "grad_norm": 2.0449395506806574,
+      "learning_rate": 2.41724786478712e-07,
+      "loss": 0.0935,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9044967880085653,
+      "grad_norm": 2.9083101066317654,
+      "learning_rate": 2.3959818605736095e-07,
+      "loss": 0.1092,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9049250535331905,
+      "grad_norm": 2.031298889424573,
+      "learning_rate": 2.3748075185951515e-07,
+      "loss": 0.0803,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9053533190578158,
+      "grad_norm": 1.7488958392412313,
+      "learning_rate": 2.3537248796232537e-07,
+      "loss": 0.1278,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9057815845824411,
+      "grad_norm": 1.9496078873664138,
+      "learning_rate": 2.3327339842528385e-07,
+      "loss": 0.0947,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9062098501070663,
+      "grad_norm": 2.222708382964901,
+      "learning_rate": 2.3118348729021856e-07,
+      "loss": 0.129,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9066381156316916,
+      "grad_norm": 2.4783669080550803,
+      "learning_rate": 2.2910275858128439e-07,
+      "loss": 0.101,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9070663811563169,
+      "grad_norm": 2.0132866064849364,
+      "learning_rate": 2.270312163049554e-07,
+      "loss": 0.1318,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9074946466809422,
+      "grad_norm": 1.6336165063117767,
+      "learning_rate": 2.249688644500153e-07,
+      "loss": 0.057,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9079229122055674,
+      "grad_norm": 1.5176292977752488,
+      "learning_rate": 2.229157069875537e-07,
+      "loss": 0.0913,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9083511777301927,
+      "grad_norm": 1.876026299884056,
+      "learning_rate": 2.2087174787095488e-07,
+      "loss": 0.0878,
+      "step": 2121
+    },
+    {
+      "epoch": 0.908779443254818,
+      "grad_norm": 1.6346029517592429,
+      "learning_rate": 2.1883699103589285e-07,
+      "loss": 0.0842,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9092077087794432,
+      "grad_norm": 1.727251511970511,
+      "learning_rate": 2.1681144040032188e-07,
+      "loss": 0.074,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9096359743040685,
+      "grad_norm": 1.9223893471700446,
+      "learning_rate": 2.1479509986446822e-07,
+      "loss": 0.1243,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9100642398286938,
+      "grad_norm": 1.9581979615810354,
+      "learning_rate": 2.1278797331082567e-07,
+      "loss": 0.064,
+      "step": 2125
+    },
+    {
+      "epoch": 0.910492505353319,
+      "grad_norm": 2.2510036329493035,
+      "learning_rate": 2.1079006460414608e-07,
+      "loss": 0.0904,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9109207708779443,
+      "grad_norm": 1.7098770410794524,
+      "learning_rate": 2.0880137759143216e-07,
+      "loss": 0.0656,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9113490364025696,
+      "grad_norm": 1.8739877736594273,
+      "learning_rate": 2.068219161019297e-07,
+      "loss": 0.0727,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9117773019271949,
+      "grad_norm": 1.7533771553103317,
+      "learning_rate": 2.048516839471204e-07,
+      "loss": 0.0802,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9122055674518201,
+      "grad_norm": 1.7804075535663664,
+      "learning_rate": 2.0289068492071574e-07,
+      "loss": 0.0838,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9126338329764454,
+      "grad_norm": 1.8504852841655546,
+      "learning_rate": 2.0093892279864802e-07,
+      "loss": 0.0948,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9130620985010707,
+      "grad_norm": 2.075162760097523,
+      "learning_rate": 1.9899640133906384e-07,
+      "loss": 0.107,
+      "step": 2132
+    },
+    {
+      "epoch": 0.913490364025696,
+      "grad_norm": 1.6326873135151379,
+      "learning_rate": 1.970631242823162e-07,
+      "loss": 0.1165,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9139186295503212,
+      "grad_norm": 2.185820347123858,
+      "learning_rate": 1.9513909535095853e-07,
+      "loss": 0.1097,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9143468950749465,
+      "grad_norm": 2.063229659285653,
+      "learning_rate": 1.9322431824973731e-07,
+      "loss": 0.094,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9147751605995718,
+      "grad_norm": 1.705872497325175,
+      "learning_rate": 1.9131879666558385e-07,
+      "loss": 0.0497,
+      "step": 2136
+    },
+    {
+      "epoch": 0.915203426124197,
+      "grad_norm": 1.5460248930523435,
+      "learning_rate": 1.8942253426760814e-07,
+      "loss": 0.069,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9156316916488223,
+      "grad_norm": 2.3255407612995906,
+      "learning_rate": 1.8753553470709063e-07,
+      "loss": 0.1436,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9160599571734476,
+      "grad_norm": 2.1441170031423393,
+      "learning_rate": 1.8565780161747648e-07,
+      "loss": 0.0845,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9164882226980728,
+      "grad_norm": 2.3267955510728204,
+      "learning_rate": 1.8378933861436855e-07,
+      "loss": 0.1239,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9169164882226981,
+      "grad_norm": 1.8007978326440754,
+      "learning_rate": 1.8193014929552e-07,
+      "loss": 0.0958,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9173447537473234,
+      "grad_norm": 2.1899851706514752,
+      "learning_rate": 1.800802372408278e-07,
+      "loss": 0.0992,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9177730192719487,
+      "grad_norm": 1.9587448309508733,
+      "learning_rate": 1.782396060123237e-07,
+      "loss": 0.09,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9182012847965739,
+      "grad_norm": 2.2804196541114643,
+      "learning_rate": 1.7640825915416994e-07,
+      "loss": 0.1133,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9186295503211992,
+      "grad_norm": 1.6127218153719036,
+      "learning_rate": 1.7458620019265183e-07,
+      "loss": 0.0984,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9190578158458245,
+      "grad_norm": 1.8968866698859284,
+      "learning_rate": 1.727734326361713e-07,
+      "loss": 0.0714,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9194860813704496,
+      "grad_norm": 2.080668239080616,
+      "learning_rate": 1.7096995997523902e-07,
+      "loss": 0.1454,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9199143468950749,
+      "grad_norm": 2.554820574714056,
+      "learning_rate": 1.6917578568246717e-07,
+      "loss": 0.1319,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9203426124197002,
+      "grad_norm": 2.0241204681936122,
+      "learning_rate": 1.673909132125645e-07,
+      "loss": 0.0821,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9207708779443254,
+      "grad_norm": 1.884302399272295,
+      "learning_rate": 1.656153460023302e-07,
+      "loss": 0.102,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9211991434689507,
+      "grad_norm": 2.1791001934071255,
+      "learning_rate": 1.6384908747064443e-07,
+      "loss": 0.081,
+      "step": 2151
+    },
+    {
+      "epoch": 0.921627408993576,
+      "grad_norm": 2.208903816620388,
+      "learning_rate": 1.6209214101846394e-07,
+      "loss": 0.1052,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9220556745182013,
+      "grad_norm": 1.875722747708676,
+      "learning_rate": 1.6034451002881478e-07,
+      "loss": 0.0963,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9224839400428265,
+      "grad_norm": 2.0137968897091,
+      "learning_rate": 1.5860619786678632e-07,
+      "loss": 0.0881,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9229122055674518,
+      "grad_norm": 1.8263463424059274,
+      "learning_rate": 1.5687720787952332e-07,
+      "loss": 0.0641,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9233404710920771,
+      "grad_norm": 2.0141103510719662,
+      "learning_rate": 1.5515754339622214e-07,
+      "loss": 0.0897,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9237687366167023,
+      "grad_norm": 1.8183366286514095,
+      "learning_rate": 1.5344720772812294e-07,
+      "loss": 0.0921,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9241970021413276,
+      "grad_norm": 2.036279406098828,
+      "learning_rate": 1.5174620416849972e-07,
+      "loss": 0.1144,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9246252676659529,
+      "grad_norm": 2.057341571537349,
+      "learning_rate": 1.500545359926614e-07,
+      "loss": 0.1074,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9250535331905781,
+      "grad_norm": 1.5560703276970007,
+      "learning_rate": 1.4837220645793905e-07,
+      "loss": 0.0885,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9254817987152034,
+      "grad_norm": 1.7598687321429338,
+      "learning_rate": 1.4669921880368422e-07,
+      "loss": 0.0928,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9259100642398287,
+      "grad_norm": 1.7330808055898828,
+      "learning_rate": 1.4503557625125787e-07,
+      "loss": 0.0573,
+      "step": 2162
+    },
+    {
+      "epoch": 0.926338329764454,
+      "grad_norm": 2.235830845063436,
+      "learning_rate": 1.4338128200402867e-07,
+      "loss": 0.076,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9267665952890792,
+      "grad_norm": 1.5969657493763263,
+      "learning_rate": 1.4173633924736364e-07,
+      "loss": 0.0555,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9271948608137045,
+      "grad_norm": 2.0321971422792258,
+      "learning_rate": 1.401007511486252e-07,
+      "loss": 0.0913,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9276231263383298,
+      "grad_norm": 2.3586765307652158,
+      "learning_rate": 1.3847452085716252e-07,
+      "loss": 0.1315,
+      "step": 2166
+    },
+    {
+      "epoch": 0.928051391862955,
+      "grad_norm": 1.514883966086733,
+      "learning_rate": 1.3685765150430462e-07,
+      "loss": 0.0904,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9284796573875803,
+      "grad_norm": 2.35842942546834,
+      "learning_rate": 1.3525014620335786e-07,
+      "loss": 0.0874,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9289079229122056,
+      "grad_norm": 1.819561457852712,
+      "learning_rate": 1.3365200804959731e-07,
+      "loss": 0.1031,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9293361884368309,
+      "grad_norm": 2.218630999737059,
+      "learning_rate": 1.3206324012026095e-07,
+      "loss": 0.0977,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9297644539614561,
+      "grad_norm": 1.9005852049610779,
+      "learning_rate": 1.304838454745455e-07,
+      "loss": 0.0731,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9301927194860814,
+      "grad_norm": 2.149906648638233,
+      "learning_rate": 1.289138271535978e-07,
+      "loss": 0.0827,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9306209850107067,
+      "grad_norm": 2.127444934055037,
+      "learning_rate": 1.2735318818051123e-07,
+      "loss": 0.0823,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9310492505353319,
+      "grad_norm": 1.6210918512549735,
+      "learning_rate": 1.2580193156031927e-07,
+      "loss": 0.1051,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9314775160599572,
+      "grad_norm": 2.1489038516958243,
+      "learning_rate": 1.2426006027998982e-07,
+      "loss": 0.0753,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9319057815845825,
+      "grad_norm": 1.8479358715216865,
+      "learning_rate": 1.2272757730841744e-07,
+      "loss": 0.1242,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9323340471092078,
+      "grad_norm": 2.0490385956245194,
+      "learning_rate": 1.2120448559642228e-07,
+      "loss": 0.103,
+      "step": 2177
+    },
+    {
+      "epoch": 0.932762312633833,
+      "grad_norm": 2.0508324457458977,
+      "learning_rate": 1.1969078807673728e-07,
+      "loss": 0.1155,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9331905781584583,
+      "grad_norm": 2.079486216040738,
+      "learning_rate": 1.1818648766401097e-07,
+      "loss": 0.0814,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9336188436830836,
+      "grad_norm": 1.8535737836322963,
+      "learning_rate": 1.1669158725479579e-07,
+      "loss": 0.0845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9340471092077087,
+      "grad_norm": 1.6593447618955748,
+      "learning_rate": 1.1520608972754422e-07,
+      "loss": 0.0886,
+      "step": 2181
+    },
+    {
+      "epoch": 0.934475374732334,
+      "grad_norm": 2.043977517227241,
+      "learning_rate": 1.137299979426032e-07,
+      "loss": 0.1353,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9349036402569593,
+      "grad_norm": 1.761925708760175,
+      "learning_rate": 1.122633147422092e-07,
+      "loss": 0.1005,
+      "step": 2183
+    },
+    {
+      "epoch": 0.9353319057815845,
+      "grad_norm": 1.8536756832089298,
+      "learning_rate": 1.1080604295048203e-07,
+      "loss": 0.0673,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9357601713062098,
+      "grad_norm": 1.9084781378134363,
+      "learning_rate": 1.0935818537342103e-07,
+      "loss": 0.0916,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9361884368308351,
+      "grad_norm": 1.7927939493883385,
+      "learning_rate": 1.079197447988961e-07,
+      "loss": 0.0991,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9366167023554604,
+      "grad_norm": 1.8705253072404289,
+      "learning_rate": 1.0649072399664562e-07,
+      "loss": 0.1326,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9370449678800856,
+      "grad_norm": 1.5904308490762913,
+      "learning_rate": 1.0507112571827072e-07,
+      "loss": 0.0757,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9374732334047109,
+      "grad_norm": 2.164119547470118,
+      "learning_rate": 1.036609526972282e-07,
+      "loss": 0.0819,
+      "step": 2189
+    },
+    {
+      "epoch": 0.9379014989293362,
+      "grad_norm": 2.2880037516785245,
+      "learning_rate": 1.0226020764882771e-07,
+      "loss": 0.0922,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9383297644539614,
+      "grad_norm": 1.7604268338637419,
+      "learning_rate": 1.0086889327022343e-07,
+      "loss": 0.0943,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9387580299785867,
+      "grad_norm": 2.182158856582203,
+      "learning_rate": 9.948701224041124e-08,
+      "loss": 0.1254,
+      "step": 2192
+    },
+    {
+      "epoch": 0.939186295503212,
+      "grad_norm": 2.079802327711691,
+      "learning_rate": 9.811456722022495e-08,
+      "loss": 0.1123,
+      "step": 2193
+    },
+    {
+      "epoch": 0.9396145610278372,
+      "grad_norm": 1.7501029127261278,
+      "learning_rate": 9.675156085232617e-08,
+      "loss": 0.0682,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9400428265524625,
+      "grad_norm": 2.0644879695165783,
+      "learning_rate": 9.539799576120445e-08,
+      "loss": 0.1109,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9404710920770878,
+      "grad_norm": 1.5891202327047582,
+      "learning_rate": 9.405387455316884e-08,
+      "loss": 0.079,
+      "step": 2196
+    },
+    {
+      "epoch": 0.9408993576017131,
+      "grad_norm": 1.7924970238925495,
+      "learning_rate": 9.271919981634459e-08,
+      "loss": 0.0871,
+      "step": 2197
+    },
+    {
+      "epoch": 0.9413276231263383,
+      "grad_norm": 1.7976532399423706,
+      "learning_rate": 9.139397412066709e-08,
+      "loss": 0.075,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9417558886509636,
+      "grad_norm": 1.8254822018188919,
+      "learning_rate": 9.007820001787848e-08,
+      "loss": 0.0884,
+      "step": 2199
+    },
+    {
+      "epoch": 0.9421841541755889,
+      "grad_norm": 1.758251583159856,
+      "learning_rate": 8.877188004152104e-08,
+      "loss": 0.0965,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9426124197002141,
+      "grad_norm": 1.469273607335507,
+      "learning_rate": 8.747501670693215e-08,
+      "loss": 0.0553,
+      "step": 2201
+    },
+    {
+      "epoch": 0.9430406852248394,
+      "grad_norm": 2.2052977706664247,
+      "learning_rate": 8.618761251124207e-08,
+      "loss": 0.0996,
+      "step": 2202
+    },
+    {
+      "epoch": 0.9434689507494647,
+      "grad_norm": 2.1161346775684566,
+      "learning_rate": 8.490966993336625e-08,
+      "loss": 0.0897,
+      "step": 2203
+    },
+    {
+      "epoch": 0.94389721627409,
+      "grad_norm": 2.143647782952967,
+      "learning_rate": 8.364119143400185e-08,
+      "loss": 0.1316,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9443254817987152,
+      "grad_norm": 2.137778782740592,
+      "learning_rate": 8.238217945562288e-08,
+      "loss": 0.1095,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9447537473233405,
+      "grad_norm": 2.139238974768021,
+      "learning_rate": 8.113263642247516e-08,
+      "loss": 0.0954,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9451820128479658,
+      "grad_norm": 1.5101254511314042,
+      "learning_rate": 7.989256474057183e-08,
+      "loss": 0.0734,
+      "step": 2207
+    },
+    {
+      "epoch": 0.945610278372591,
+      "grad_norm": 1.9120963093097392,
+      "learning_rate": 7.866196679768956e-08,
+      "loss": 0.1096,
+      "step": 2208
+    },
+    {
+      "epoch": 0.9460385438972163,
+      "grad_norm": 2.0198750775157452,
+      "learning_rate": 7.744084496336346e-08,
+      "loss": 0.1185,
+      "step": 2209
+    },
+    {
+      "epoch": 0.9464668094218416,
+      "grad_norm": 2.1983077575393657,
+      "learning_rate": 7.622920158888048e-08,
+      "loss": 0.1058,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9468950749464669,
+      "grad_norm": 1.910246765754372,
+      "learning_rate": 7.50270390072777e-08,
+      "loss": 0.0768,
+      "step": 2211
+    },
+    {
+      "epoch": 0.9473233404710921,
+      "grad_norm": 1.9844083328928481,
+      "learning_rate": 7.383435953333684e-08,
+      "loss": 0.0791,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9477516059957174,
+      "grad_norm": 2.4009357283358987,
+      "learning_rate": 7.265116546358086e-08,
+      "loss": 0.0842,
+      "step": 2213
+    },
+    {
+      "epoch": 0.9481798715203427,
+      "grad_norm": 1.7623128910598351,
+      "learning_rate": 7.147745907626569e-08,
+      "loss": 0.0938,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9486081370449678,
+      "grad_norm": 1.8214891756490985,
+      "learning_rate": 7.031324263138129e-08,
+      "loss": 0.0703,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9490364025695931,
+      "grad_norm": 2.0799484640790853,
+      "learning_rate": 6.91585183706428e-08,
+      "loss": 0.0932,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9494646680942184,
+      "grad_norm": 1.9060018769550933,
+      "learning_rate": 6.801328851748889e-08,
+      "loss": 0.0651,
+      "step": 2217
+    },
+    {
+      "epoch": 0.9498929336188436,
+      "grad_norm": 1.771851875002198,
+      "learning_rate": 6.687755527707673e-08,
+      "loss": 0.0818,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9503211991434689,
+      "grad_norm": 1.9396478378261757,
+      "learning_rate": 6.575132083627755e-08,
+      "loss": 0.0963,
+      "step": 2219
+    },
+    {
+      "epoch": 0.9507494646680942,
+      "grad_norm": 2.287061152756386,
+      "learning_rate": 6.463458736367111e-08,
+      "loss": 0.1091,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9511777301927195,
+      "grad_norm": 1.775595587153788,
+      "learning_rate": 6.352735700954571e-08,
+      "loss": 0.0794,
+      "step": 2221
+    },
+    {
+      "epoch": 0.9516059957173447,
+      "grad_norm": 2.076705370009214,
+      "learning_rate": 6.242963190588813e-08,
+      "loss": 0.092,
+      "step": 2222
+    },
+    {
+      "epoch": 0.95203426124197,
+      "grad_norm": 1.9357460384369356,
+      "learning_rate": 6.134141416638484e-08,
+      "loss": 0.072,
+      "step": 2223
+    },
+    {
+      "epoch": 0.9524625267665953,
+      "grad_norm": 1.7459544896788788,
+      "learning_rate": 6.02627058864158e-08,
+      "loss": 0.0796,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9528907922912205,
+      "grad_norm": 1.836770654548759,
+      "learning_rate": 5.9193509143048446e-08,
+      "loss": 0.0592,
+      "step": 2225
+    },
+    {
+      "epoch": 0.9533190578158458,
+      "grad_norm": 1.8914141395885018,
+      "learning_rate": 5.813382599503703e-08,
+      "loss": 0.0705,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9537473233404711,
+      "grad_norm": 1.5078008096069424,
+      "learning_rate": 5.7083658482817165e-08,
+      "loss": 0.0651,
+      "step": 2227
+    },
+    {
+      "epoch": 0.9541755888650963,
+      "grad_norm": 1.8708658871665158,
+      "learning_rate": 5.604300862850187e-08,
+      "loss": 0.0721,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9546038543897216,
+      "grad_norm": 1.7719682375896253,
+      "learning_rate": 5.501187843587774e-08,
+      "loss": 0.0991,
+      "step": 2229
+    },
+    {
+      "epoch": 0.9550321199143469,
+      "grad_norm": 1.691075578684216,
+      "learning_rate": 5.399026989040101e-08,
+      "loss": 0.0788,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9554603854389722,
+      "grad_norm": 1.792308882526915,
+      "learning_rate": 5.2978184959193714e-08,
+      "loss": 0.0748,
+      "step": 2231
+    },
+    {
+      "epoch": 0.9558886509635974,
+      "grad_norm": 2.1475261772586998,
+      "learning_rate": 5.19756255910403e-08,
+      "loss": 0.1042,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9563169164882227,
+      "grad_norm": 2.2081277918854836,
+      "learning_rate": 5.098259371638437e-08,
+      "loss": 0.0696,
+      "step": 2233
+    },
+    {
+      "epoch": 0.956745182012848,
+      "grad_norm": 1.6197862952788515,
+      "learning_rate": 4.999909124732194e-08,
+      "loss": 0.0656,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9571734475374732,
+      "grad_norm": 1.9941261555479501,
+      "learning_rate": 4.902512007760207e-08,
+      "loss": 0.0646,
+      "step": 2235
+    },
+    {
+      "epoch": 0.9576017130620985,
+      "grad_norm": 1.836558692425522,
+      "learning_rate": 4.806068208262071e-08,
+      "loss": 0.113,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9580299785867238,
+      "grad_norm": 1.778015688021829,
+      "learning_rate": 4.710577911941738e-08,
+      "loss": 0.0681,
+      "step": 2237
+    },
+    {
+      "epoch": 0.9584582441113491,
+      "grad_norm": 1.8603500022691677,
+      "learning_rate": 4.6160413026671844e-08,
+      "loss": 0.0612,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9588865096359743,
+      "grad_norm": 2.1359891459644884,
+      "learning_rate": 4.522458562469967e-08,
+      "loss": 0.0973,
+      "step": 2239
+    },
+    {
+      "epoch": 0.9593147751605996,
+      "grad_norm": 2.0411966133162003,
+      "learning_rate": 4.429829871545055e-08,
+      "loss": 0.1171,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9597430406852249,
+      "grad_norm": 2.233874380047149,
+      "learning_rate": 4.338155408250388e-08,
+      "loss": 0.1023,
+      "step": 2241
+    },
+    {
+      "epoch": 0.9601713062098501,
+      "grad_norm": 1.910316885530526,
+      "learning_rate": 4.247435349106488e-08,
+      "loss": 0.0887,
+      "step": 2242
+    },
+    {
+      "epoch": 0.9605995717344754,
+      "grad_norm": 1.8288802129891473,
+      "learning_rate": 4.1576698687961766e-08,
+      "loss": 0.0886,
+      "step": 2243
+    },
+    {
+      "epoch": 0.9610278372591007,
+      "grad_norm": 1.875607645632209,
+      "learning_rate": 4.068859140164083e-08,
+      "loss": 0.1275,
+      "step": 2244
+    },
+    {
+      "epoch": 0.961456102783726,
+      "grad_norm": 2.0533638730836246,
+      "learning_rate": 3.981003334216693e-08,
+      "loss": 0.1133,
+      "step": 2245
+    },
+    {
+      "epoch": 0.9618843683083512,
+      "grad_norm": 1.9459260505014058,
+      "learning_rate": 3.894102620121576e-08,
+      "loss": 0.0819,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9623126338329765,
+      "grad_norm": 2.074237578678291,
+      "learning_rate": 3.808157165207438e-08,
+      "loss": 0.0877,
+      "step": 2247
+    },
+    {
+      "epoch": 0.9627408993576018,
+      "grad_norm": 1.9358040709577922,
+      "learning_rate": 3.7231671349634015e-08,
+      "loss": 0.0814,
+      "step": 2248
+    },
+    {
+      "epoch": 0.963169164882227,
+      "grad_norm": 1.9926800737203083,
+      "learning_rate": 3.6391326930390045e-08,
+      "loss": 0.1027,
+      "step": 2249
+    },
+    {
+      "epoch": 0.9635974304068522,
+      "grad_norm": 2.1909552161420507,
+      "learning_rate": 3.5560540012438694e-08,
+      "loss": 0.0968,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9640256959314775,
+      "grad_norm": 1.4259474739503946,
+      "learning_rate": 3.4739312195471995e-08,
+      "loss": 0.0668,
+      "step": 2251
+    },
+    {
+      "epoch": 0.9644539614561027,
+      "grad_norm": 1.71476307954407,
+      "learning_rate": 3.3927645060776725e-08,
+      "loss": 0.0925,
+      "step": 2252
+    },
+    {
+      "epoch": 0.964882226980728,
+      "grad_norm": 1.5093887112939293,
+      "learning_rate": 3.312554017122882e-08,
+      "loss": 0.088,
+      "step": 2253
+    },
+    {
+      "epoch": 0.9653104925053533,
+      "grad_norm": 1.73263368713124,
+      "learning_rate": 3.233299907129395e-08,
+      "loss": 0.0762,
+      "step": 2254
+    },
+    {
+      "epoch": 0.9657387580299786,
+      "grad_norm": 1.813158013589331,
+      "learning_rate": 3.1550023287021394e-08,
+      "loss": 0.0966,
+      "step": 2255
+    },
+    {
+      "epoch": 0.9661670235546038,
+      "grad_norm": 2.073802613038506,
+      "learning_rate": 3.077661432604184e-08,
+      "loss": 0.1265,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9665952890792291,
+      "grad_norm": 2.1721805333459927,
+      "learning_rate": 3.0012773677566274e-08,
+      "loss": 0.0821,
+      "step": 2257
+    },
+    {
+      "epoch": 0.9670235546038544,
+      "grad_norm": 1.6525714494656556,
+      "learning_rate": 2.925850281238152e-08,
+      "loss": 0.0718,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9674518201284796,
+      "grad_norm": 1.710438124605696,
+      "learning_rate": 2.85138031828458e-08,
+      "loss": 0.0774,
+      "step": 2259
+    },
+    {
+      "epoch": 0.9678800856531049,
+      "grad_norm": 1.6251419626388464,
+      "learning_rate": 2.7778676222890433e-08,
+      "loss": 0.0831,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9683083511777302,
+      "grad_norm": 1.9441530979751092,
+      "learning_rate": 2.7053123348012023e-08,
+      "loss": 0.0806,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9687366167023554,
+      "grad_norm": 2.4086550687452384,
+      "learning_rate": 2.63371459552747e-08,
+      "loss": 0.1057,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9691648822269807,
+      "grad_norm": 2.657891161730371,
+      "learning_rate": 2.563074542330235e-08,
+      "loss": 0.1131,
+      "step": 2263
+    },
+    {
+      "epoch": 0.969593147751606,
+      "grad_norm": 1.6529635041496884,
+      "learning_rate": 2.4933923112279712e-08,
+      "loss": 0.1355,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9700214132762313,
+      "grad_norm": 2.138354075350018,
+      "learning_rate": 2.4246680363948504e-08,
+      "loss": 0.1085,
+      "step": 2265
+    },
+    {
+      "epoch": 0.9704496788008565,
+      "grad_norm": 1.6540067727449508,
+      "learning_rate": 2.3569018501605202e-08,
+      "loss": 0.0722,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9708779443254818,
+      "grad_norm": 1.6566570854640978,
+      "learning_rate": 2.290093883009714e-08,
+      "loss": 0.0783,
+      "step": 2267
+    },
+    {
+      "epoch": 0.9713062098501071,
+      "grad_norm": 1.9734291496976233,
+      "learning_rate": 2.224244263582087e-08,
+      "loss": 0.1053,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9717344753747323,
+      "grad_norm": 2.3218798061592683,
+      "learning_rate": 2.1593531186721585e-08,
+      "loss": 0.1066,
+      "step": 2269
+    },
+    {
+      "epoch": 0.9721627408993576,
+      "grad_norm": 2.3538216038994833,
+      "learning_rate": 2.095420573228757e-08,
+      "loss": 0.0838,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9725910064239829,
+      "grad_norm": 1.9562462596830206,
+      "learning_rate": 2.0324467503549107e-08,
+      "loss": 0.0787,
+      "step": 2271
+    },
+    {
+      "epoch": 0.9730192719486082,
+      "grad_norm": 1.805341008897864,
+      "learning_rate": 1.9704317713076236e-08,
+      "loss": 0.0808,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9734475374732334,
+      "grad_norm": 1.7872738857097992,
+      "learning_rate": 1.909375755497711e-08,
+      "loss": 0.0997,
+      "step": 2273
+    },
+    {
+      "epoch": 0.9738758029978587,
+      "grad_norm": 1.6259299182122677,
+      "learning_rate": 1.8492788204893532e-08,
+      "loss": 0.103,
+      "step": 2274
+    },
+    {
+      "epoch": 0.974304068522484,
+      "grad_norm": 1.7493212280320063,
+      "learning_rate": 1.7901410820000964e-08,
+      "loss": 0.0805,
+      "step": 2275
+    },
+    {
+      "epoch": 0.9747323340471092,
+      "grad_norm": 1.96996229145216,
+      "learning_rate": 1.7319626539005762e-08,
+      "loss": 0.0811,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9751605995717345,
+      "grad_norm": 2.014613215428129,
+      "learning_rate": 1.6747436482141276e-08,
+      "loss": 0.0831,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9755888650963598,
+      "grad_norm": 1.720886105024453,
+      "learning_rate": 1.6184841751168966e-08,
+      "loss": 0.0841,
+      "step": 2278
+    },
+    {
+      "epoch": 0.976017130620985,
+      "grad_norm": 1.8099263483951549,
+      "learning_rate": 1.563184342937174e-08,
+      "loss": 0.0788,
+      "step": 2279
+    },
+    {
+      "epoch": 0.9764453961456103,
+      "grad_norm": 2.008600014374065,
+      "learning_rate": 1.508844258155728e-08,
+      "loss": 0.106,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9768736616702356,
+      "grad_norm": 1.7994453406003306,
+      "learning_rate": 1.4554640254051955e-08,
+      "loss": 0.0756,
+      "step": 2281
+    },
+    {
+      "epoch": 0.9773019271948609,
+      "grad_norm": 1.7606468320624635,
+      "learning_rate": 1.4030437474698566e-08,
+      "loss": 0.0773,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9777301927194861,
+      "grad_norm": 2.001597295323409,
+      "learning_rate": 1.35158352528586e-08,
+      "loss": 0.116,
+      "step": 2283
+    },
+    {
+      "epoch": 0.9781584582441113,
+      "grad_norm": 1.9096447139375108,
+      "learning_rate": 1.3010834579405552e-08,
+      "loss": 0.0904,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9785867237687366,
+      "grad_norm": 1.6602585290234597,
+      "learning_rate": 1.2515436426725481e-08,
+      "loss": 0.0606,
+      "step": 2285
+    },
+    {
+      "epoch": 0.9790149892933618,
+      "grad_norm": 1.9487214545517622,
+      "learning_rate": 1.2029641748715349e-08,
+      "loss": 0.0866,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9794432548179871,
+      "grad_norm": 1.6416729014286375,
+      "learning_rate": 1.1553451480779687e-08,
+      "loss": 0.0753,
+      "step": 2287
+    },
+    {
+      "epoch": 0.9798715203426124,
+      "grad_norm": 1.8197526811911202,
+      "learning_rate": 1.1086866539830044e-08,
+      "loss": 0.0824,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9802997858672376,
+      "grad_norm": 1.8420407575901185,
+      "learning_rate": 1.0629887824282203e-08,
+      "loss": 0.0862,
+      "step": 2289
+    },
+    {
+      "epoch": 0.9807280513918629,
+      "grad_norm": 2.0679211309715395,
+      "learning_rate": 1.0182516214055083e-08,
+      "loss": 0.1256,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9811563169164882,
+      "grad_norm": 1.8039995077032147,
+      "learning_rate": 9.74475257057017e-09,
+      "loss": 0.0984,
+      "step": 2291
+    },
+    {
+      "epoch": 0.9815845824411135,
+      "grad_norm": 2.11358201171608,
+      "learning_rate": 9.316597736747091e-09,
+      "loss": 0.0943,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9820128479657387,
+      "grad_norm": 1.7018457605302033,
+      "learning_rate": 8.898052537004709e-09,
+      "loss": 0.0999,
+      "step": 2293
+    },
+    {
+      "epoch": 0.982441113490364,
+      "grad_norm": 2.140404705545362,
+      "learning_rate": 8.489117777257804e-09,
+      "loss": 0.1489,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9828693790149893,
+      "grad_norm": 1.7102431982822932,
+      "learning_rate": 8.08979424491596e-09,
+      "loss": 0.0614,
+      "step": 2295
+    },
+    {
+      "epoch": 0.9832976445396145,
+      "grad_norm": 2.1226175271901795,
+      "learning_rate": 7.700082708883006e-09,
+      "loss": 0.0991,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9837259100642398,
+      "grad_norm": 1.6642794085709158,
+      "learning_rate": 7.319983919553686e-09,
+      "loss": 0.0949,
+      "step": 2297
+    },
+    {
+      "epoch": 0.9841541755888651,
+      "grad_norm": 2.0887931163757636,
+      "learning_rate": 6.94949860881422e-09,
+      "loss": 0.0959,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9845824411134904,
+      "grad_norm": 2.035765233336901,
+      "learning_rate": 6.588627490038968e-09,
+      "loss": 0.0694,
+      "step": 2299
+    },
+    {
+      "epoch": 0.9850107066381156,
+      "grad_norm": 2.0333443061532708,
+      "learning_rate": 6.237371258090985e-09,
+      "loss": 0.0965,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9854389721627409,
+      "grad_norm": 1.904176095270097,
+      "learning_rate": 5.895730589319248e-09,
+      "loss": 0.0956,
+      "step": 2301
+    },
+    {
+      "epoch": 0.9858672376873662,
+      "grad_norm": 1.9875775405100682,
+      "learning_rate": 5.563706141558101e-09,
+      "loss": 0.0692,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9862955032119914,
+      "grad_norm": 1.9514387473966723,
+      "learning_rate": 5.241298554125029e-09,
+      "loss": 0.1055,
+      "step": 2303
+    },
+    {
+      "epoch": 0.9867237687366167,
+      "grad_norm": 2.29696524281879,
+      "learning_rate": 4.928508447821223e-09,
+      "loss": 0.0834,
+      "step": 2304
+    },
+    {
+      "epoch": 0.987152034261242,
+      "grad_norm": 1.8357185558755715,
+      "learning_rate": 4.625336424927685e-09,
+      "loss": 0.1093,
+      "step": 2305
+    },
+    {
+      "epoch": 0.9875802997858673,
+      "grad_norm": 2.058551100549323,
+      "learning_rate": 4.331783069207451e-09,
+      "loss": 0.0961,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9880085653104925,
+      "grad_norm": 2.238984910159028,
+      "learning_rate": 4.047848945901156e-09,
+      "loss": 0.0817,
+      "step": 2307
+    },
+    {
+      "epoch": 0.9884368308351178,
+      "grad_norm": 2.2269826663778027,
+      "learning_rate": 3.77353460172869e-09,
+      "loss": 0.1192,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9888650963597431,
+      "grad_norm": 1.8787394903693637,
+      "learning_rate": 3.5088405648858735e-09,
+      "loss": 0.0716,
+      "step": 2309
+    },
+    {
+      "epoch": 0.9892933618843683,
+      "grad_norm": 1.6226282143574633,
+      "learning_rate": 3.2537673450450115e-09,
+      "loss": 0.0604,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9897216274089936,
+      "grad_norm": 1.743215664730554,
+      "learning_rate": 3.0083154333537814e-09,
+      "loss": 0.069,
+      "step": 2311
+    },
+    {
+      "epoch": 0.9901498929336189,
+      "grad_norm": 1.9440042440337575,
+      "learning_rate": 2.7724853024324594e-09,
+      "loss": 0.112,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9905781584582442,
+      "grad_norm": 1.7183285850620549,
+      "learning_rate": 2.546277406376696e-09,
+      "loss": 0.083,
+      "step": 2313
+    },
+    {
+      "epoch": 0.9910064239828694,
+      "grad_norm": 1.6908378463323896,
+      "learning_rate": 2.3296921807525187e-09,
+      "loss": 0.0761,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9914346895074947,
+      "grad_norm": 2.647317077027036,
+      "learning_rate": 2.122730042597998e-09,
+      "loss": 0.1129,
+      "step": 2315
+    },
+    {
+      "epoch": 0.99186295503212,
+      "grad_norm": 1.9560346846913992,
+      "learning_rate": 1.925391390421583e-09,
+      "loss": 0.0719,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9922912205567452,
+      "grad_norm": 2.0060637561382837,
+      "learning_rate": 1.7376766042015437e-09,
+      "loss": 0.0785,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9927194860813704,
+      "grad_norm": 1.8576963432457074,
+      "learning_rate": 1.5595860453859745e-09,
+      "loss": 0.0888,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9931477516059957,
+      "grad_norm": 2.2718368949031267,
+      "learning_rate": 1.3911200568905713e-09,
+      "loss": 0.1108,
+      "step": 2319
+    },
+    {
+      "epoch": 0.9935760171306209,
+      "grad_norm": 1.8191785875241426,
+      "learning_rate": 1.2322789630997422e-09,
+      "loss": 0.0742,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9940042826552462,
+      "grad_norm": 1.7229480201322784,
+      "learning_rate": 1.0830630698632772e-09,
+      "loss": 0.0777,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9944325481798715,
+      "grad_norm": 1.852549223952128,
+      "learning_rate": 9.434726644991231e-10,
+      "loss": 0.1367,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9948608137044967,
+      "grad_norm": 1.8754387771716792,
+      "learning_rate": 8.135080157906095e-10,
+      "loss": 0.1042,
+      "step": 2323
+    },
+    {
+      "epoch": 0.995289079229122,
+      "grad_norm": 1.937714020877812,
+      "learning_rate": 6.931693739864465e-10,
+      "loss": 0.0828,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9957173447537473,
+      "grad_norm": 2.249789374073427,
+      "learning_rate": 5.824569708001715e-10,
+      "loss": 0.0796,
+      "step": 2325
+    },
+    {
+      "epoch": 0.9961456102783726,
+      "grad_norm": 2.3064343577047715,
+      "learning_rate": 4.813710194112586e-10,
+      "loss": 0.1115,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9965738758029978,
+      "grad_norm": 1.8144698853245798,
+      "learning_rate": 3.89911714461233e-10,
+      "loss": 0.0823,
+      "step": 2327
+    },
+    {
+      "epoch": 0.9970021413276231,
+      "grad_norm": 1.767548046279923,
+      "learning_rate": 3.080792320564463e-10,
+      "loss": 0.1137,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9974304068522484,
+      "grad_norm": 1.8602068211337615,
+      "learning_rate": 2.3587372976641154e-10,
+      "loss": 0.1222,
+      "step": 2329
+    },
+    {
+      "epoch": 0.9978586723768736,
+      "grad_norm": 1.7705213812741822,
+      "learning_rate": 1.732953466243581e-10,
+      "loss": 0.0636,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9982869379014989,
+      "grad_norm": 1.4639216957290855,
+      "learning_rate": 1.2034420312556638e-10,
+      "loss": 0.0887,
+      "step": 2331
+    },
+    {
+      "epoch": 0.9987152034261242,
+      "grad_norm": 1.809462716094516,
+      "learning_rate": 7.702040122847809e-11,
+      "loss": 0.0692,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9991434689507495,
+      "grad_norm": 1.973461026771547,
+      "learning_rate": 4.3324024353030845e-11,
+      "loss": 0.1135,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9995717344753747,
+      "grad_norm": 1.660273888473188,
+      "learning_rate": 1.925513738287865e-11,
+      "loss": 0.0751,
+      "step": 2334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1480105062453925,
+      "learning_rate": 4.813786663171449e-12,
+      "loss": 0.107,
+      "step": 2335
+    },
+    {
+      "epoch": 1.0,
+      "step": 2335,
+      "total_flos": 2699499666669568.0,
+      "train_loss": 0.18601225457080425,
+      "train_runtime": 8883.3109,
+      "train_samples_per_second": 16.824,
+      "train_steps_per_second": 0.263
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2335,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2699499666669568.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}