diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7730 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9958960328317374,
+  "eval_steps": 250,
+  "global_step": 1095,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0027359781121751026,
+      "grad_norm": 0.24075667560100555,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.9078,
+      "step": 1
+    },
+    {
+      "epoch": 0.005471956224350205,
+      "grad_norm": 0.2584983706474304,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 2.1152,
+      "step": 2
+    },
+    {
+      "epoch": 0.008207934336525308,
+      "grad_norm": 0.24978317320346832,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 1.9649,
+      "step": 3
+    },
+    {
+      "epoch": 0.01094391244870041,
+      "grad_norm": 0.2477930635213852,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.9163,
+      "step": 4
+    },
+    {
+      "epoch": 0.013679890560875513,
+      "grad_norm": 0.2425757646560669,
+      "learning_rate": 1.1363636363636365e-05,
+      "loss": 2.1254,
+      "step": 5
+    },
+    {
+      "epoch": 0.016415868673050615,
+      "grad_norm": 0.251596599817276,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 2.0832,
+      "step": 6
+    },
+    {
+      "epoch": 0.019151846785225718,
+      "grad_norm": 0.2359585464000702,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 1.8204,
+      "step": 7
+    },
+    {
+      "epoch": 0.02188782489740082,
+      "grad_norm": 0.25108030438423157,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.9224,
+      "step": 8
+    },
+    {
+      "epoch": 0.024623803009575923,
+      "grad_norm": 0.25240400433540344,
+      "learning_rate": 2.0454545454545457e-05,
+      "loss": 1.8371,
+      "step": 9
+    },
+    {
+      "epoch": 0.027359781121751026,
+      "grad_norm": 0.22698192298412323,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 1.9962,
+      "step": 10
+    },
+    {
+      "epoch": 0.030095759233926128,
+      "grad_norm": 0.24213118851184845,
+      "learning_rate": 2.5e-05,
+      "loss": 1.9918,
+      "step": 11
+    },
+    {
+      "epoch": 0.03283173734610123,
+      "grad_norm": 0.20904198288917542,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 1.9463,
+      "step": 12
+    },
+    {
+      "epoch": 0.03556771545827633,
+      "grad_norm": 0.21374110877513885,
+      "learning_rate": 2.954545454545455e-05,
+      "loss": 1.8628,
+      "step": 13
+    },
+    {
+      "epoch": 0.038303693570451436,
+      "grad_norm": 0.27983948588371277,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 2.1597,
+      "step": 14
+    },
+    {
+      "epoch": 0.04103967168262654,
+      "grad_norm": 0.2309175282716751,
+      "learning_rate": 3.409090909090909e-05,
+      "loss": 2.0305,
+      "step": 15
+    },
+    {
+      "epoch": 0.04377564979480164,
+      "grad_norm": 0.21376170217990875,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 1.8866,
+      "step": 16
+    },
+    {
+      "epoch": 0.046511627906976744,
+      "grad_norm": 0.23382064700126648,
+      "learning_rate": 3.8636363636363636e-05,
+      "loss": 1.9912,
+      "step": 17
+    },
+    {
+      "epoch": 0.049247606019151846,
+      "grad_norm": 0.20445913076400757,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 1.9773,
+      "step": 18
+    },
+    {
+      "epoch": 0.05198358413132695,
+      "grad_norm": 0.21313798427581787,
+      "learning_rate": 4.318181818181819e-05,
+      "loss": 1.9965,
+      "step": 19
+    },
+    {
+      "epoch": 0.05471956224350205,
+      "grad_norm": 0.22150328755378723,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 1.9538,
+      "step": 20
+    },
+    {
+      "epoch": 0.057455540355677154,
+      "grad_norm": 0.17602485418319702,
+      "learning_rate": 4.772727272727273e-05,
+      "loss": 1.9113,
+      "step": 21
+    },
+    {
+      "epoch": 0.060191518467852256,
+      "grad_norm": 0.20136825740337372,
+      "learning_rate": 5e-05,
+      "loss": 1.9837,
+      "step": 22
+    },
+    {
+      "epoch": 0.06292749658002736,
+      "grad_norm": 0.1977069228887558,
+      "learning_rate": 4.999989284560115e-05,
+      "loss": 1.8121,
+      "step": 23
+    },
+    {
+      "epoch": 0.06566347469220246,
+      "grad_norm": 0.18849150836467743,
+      "learning_rate": 4.9999571383323136e-05,
+      "loss": 1.8013,
+      "step": 24
+    },
+    {
+      "epoch": 0.06839945280437756,
+      "grad_norm": 0.17495211958885193,
+      "learning_rate": 4.9999035615921664e-05,
+      "loss": 1.8375,
+      "step": 25
+    },
+    {
+      "epoch": 0.07113543091655267,
+      "grad_norm": 0.28524237871170044,
+      "learning_rate": 4.999828554798952e-05,
+      "loss": 2.026,
+      "step": 26
+    },
+    {
+      "epoch": 0.07387140902872777,
+      "grad_norm": 0.2135252058506012,
+      "learning_rate": 4.999732118595654e-05,
+      "loss": 1.7874,
+      "step": 27
+    },
+    {
+      "epoch": 0.07660738714090287,
+      "grad_norm": 0.213677778840065,
+      "learning_rate": 4.999614253808959e-05,
+      "loss": 1.7844,
+      "step": 28
+    },
+    {
+      "epoch": 0.07934336525307797,
+      "grad_norm": 0.20930789411067963,
+      "learning_rate": 4.999474961449243e-05,
+      "loss": 1.857,
+      "step": 29
+    },
+    {
+      "epoch": 0.08207934336525308,
+      "grad_norm": 0.18232876062393188,
+      "learning_rate": 4.999314242710572e-05,
+      "loss": 1.7145,
+      "step": 30
+    },
+    {
+      "epoch": 0.08481532147742818,
+      "grad_norm": 0.2061523050069809,
+      "learning_rate": 4.9991320989706816e-05,
+      "loss": 1.8137,
+      "step": 31
+    },
+    {
+      "epoch": 0.08755129958960328,
+      "grad_norm": 0.19810044765472412,
+      "learning_rate": 4.9989285317909725e-05,
+      "loss": 1.8154,
+      "step": 32
+    },
+    {
+      "epoch": 0.09028727770177838,
+      "grad_norm": 0.19258971512317657,
+      "learning_rate": 4.9987035429164954e-05,
+      "loss": 1.8612,
+      "step": 33
+    },
+    {
+      "epoch": 0.09302325581395349,
+      "grad_norm": 0.21273574233055115,
+      "learning_rate": 4.9984571342759326e-05,
+      "loss": 1.6282,
+      "step": 34
+    },
+    {
+      "epoch": 0.09575923392612859,
+      "grad_norm": 0.18505199253559113,
+      "learning_rate": 4.998189307981586e-05,
+      "loss": 1.7134,
+      "step": 35
+    },
+    {
+      "epoch": 0.09849521203830369,
+      "grad_norm": 0.20874658226966858,
+      "learning_rate": 4.997900066329357e-05,
+      "loss": 1.7648,
+      "step": 36
+    },
+    {
+      "epoch": 0.1012311901504788,
+      "grad_norm": 0.18423666059970856,
+      "learning_rate": 4.9975894117987265e-05,
+      "loss": 1.7784,
+      "step": 37
+    },
+    {
+      "epoch": 0.1039671682626539,
+      "grad_norm": 0.16870878636837006,
+      "learning_rate": 4.9972573470527354e-05,
+      "loss": 1.7352,
+      "step": 38
+    },
+    {
+      "epoch": 0.106703146374829,
+      "grad_norm": 0.20023159682750702,
+      "learning_rate": 4.996903874937959e-05,
+      "loss": 1.7784,
+      "step": 39
+    },
+    {
+      "epoch": 0.1094391244870041,
+      "grad_norm": 0.22696524858474731,
+      "learning_rate": 4.9965289984844846e-05,
+      "loss": 1.7722,
+      "step": 40
+    },
+    {
+      "epoch": 0.1121751025991792,
+      "grad_norm": 0.18610844016075134,
+      "learning_rate": 4.9961327209058844e-05,
+      "loss": 1.761,
+      "step": 41
+    },
+    {
+      "epoch": 0.11491108071135431,
+      "grad_norm": 0.17688068747520447,
+      "learning_rate": 4.99571504559919e-05,
+      "loss": 1.6316,
+      "step": 42
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 0.20090024173259735,
+      "learning_rate": 4.995275976144862e-05,
+      "loss": 1.8064,
+      "step": 43
+    },
+    {
+      "epoch": 0.12038303693570451,
+      "grad_norm": 0.17762398719787598,
+      "learning_rate": 4.9948155163067565e-05,
+      "loss": 1.8417,
+      "step": 44
+    },
+    {
+      "epoch": 0.12311901504787962,
+      "grad_norm": 0.1874111294746399,
+      "learning_rate": 4.9943336700320985e-05,
+      "loss": 1.7245,
+      "step": 45
+    },
+    {
+      "epoch": 0.12585499316005472,
+      "grad_norm": 0.16139264404773712,
+      "learning_rate": 4.993830441451444e-05,
+      "loss": 1.8852,
+      "step": 46
+    },
+    {
+      "epoch": 0.12859097127222982,
+      "grad_norm": 0.16358217597007751,
+      "learning_rate": 4.993305834878643e-05,
+      "loss": 1.7314,
+      "step": 47
+    },
+    {
+      "epoch": 0.13132694938440492,
+      "grad_norm": 0.16465303301811218,
+      "learning_rate": 4.992759854810811e-05,
+      "loss": 1.6362,
+      "step": 48
+    },
+    {
+      "epoch": 0.13406292749658003,
+      "grad_norm": 0.1597519814968109,
+      "learning_rate": 4.99219250592828e-05,
+      "loss": 1.5936,
+      "step": 49
+    },
+    {
+      "epoch": 0.13679890560875513,
+      "grad_norm": 0.16433416306972504,
+      "learning_rate": 4.991603793094564e-05,
+      "loss": 1.5236,
+      "step": 50
+    },
+    {
+      "epoch": 0.13953488372093023,
+      "grad_norm": 0.16490432620048523,
+      "learning_rate": 4.9909937213563165e-05,
+      "loss": 1.7191,
+      "step": 51
+    },
+    {
+      "epoch": 0.14227086183310533,
+      "grad_norm": 0.1576634645462036,
+      "learning_rate": 4.990362295943287e-05,
+      "loss": 1.5944,
+      "step": 52
+    },
+    {
+      "epoch": 0.14500683994528044,
+      "grad_norm": 0.16687043011188507,
+      "learning_rate": 4.989709522268278e-05,
+      "loss": 1.7204,
+      "step": 53
+    },
+    {
+      "epoch": 0.14774281805745554,
+      "grad_norm": 0.18680374324321747,
+      "learning_rate": 4.989035405927093e-05,
+      "loss": 1.8231,
+      "step": 54
+    },
+    {
+      "epoch": 0.15047879616963064,
+      "grad_norm": 0.16919435560703278,
+      "learning_rate": 4.9883399526984944e-05,
+      "loss": 1.7527,
+      "step": 55
+    },
+    {
+      "epoch": 0.15321477428180574,
+      "grad_norm": 0.16811755299568176,
+      "learning_rate": 4.9876231685441545e-05,
+      "loss": 1.6824,
+      "step": 56
+    },
+    {
+      "epoch": 0.15595075239398085,
+      "grad_norm": 0.15195395052433014,
+      "learning_rate": 4.986885059608597e-05,
+      "loss": 1.6515,
+      "step": 57
+    },
+    {
+      "epoch": 0.15868673050615595,
+      "grad_norm": 0.17854413390159607,
+      "learning_rate": 4.986125632219152e-05,
+      "loss": 1.613,
+      "step": 58
+    },
+    {
+      "epoch": 0.16142270861833105,
+      "grad_norm": 0.168907031416893,
+      "learning_rate": 4.985344892885899e-05,
+      "loss": 1.798,
+      "step": 59
+    },
+    {
+      "epoch": 0.16415868673050615,
+      "grad_norm": 0.173956036567688,
+      "learning_rate": 4.98454284830161e-05,
+      "loss": 1.4941,
+      "step": 60
+    },
+    {
+      "epoch": 0.16689466484268126,
+      "grad_norm": 0.19879266619682312,
+      "learning_rate": 4.983719505341693e-05,
+      "loss": 1.7941,
+      "step": 61
+    },
+    {
+      "epoch": 0.16963064295485636,
+      "grad_norm": 0.16308054327964783,
+      "learning_rate": 4.982874871064134e-05,
+      "loss": 1.6984,
+      "step": 62
+    },
+    {
+      "epoch": 0.17236662106703146,
+      "grad_norm": 0.16789625585079193,
+      "learning_rate": 4.982008952709435e-05,
+      "loss": 1.6375,
+      "step": 63
+    },
+    {
+      "epoch": 0.17510259917920656,
+      "grad_norm": 0.16414505243301392,
+      "learning_rate": 4.9811217577005533e-05,
+      "loss": 1.5667,
+      "step": 64
+    },
+    {
+      "epoch": 0.17783857729138167,
+      "grad_norm": 0.16586610674858093,
+      "learning_rate": 4.980213293642837e-05,
+      "loss": 1.6688,
+      "step": 65
+    },
+    {
+      "epoch": 0.18057455540355677,
+      "grad_norm": 0.21007002890110016,
+      "learning_rate": 4.979283568323958e-05,
+      "loss": 1.7249,
+      "step": 66
+    },
+    {
+      "epoch": 0.18331053351573187,
+      "grad_norm": 0.15806783735752106,
+      "learning_rate": 4.97833258971385e-05,
+      "loss": 1.6918,
+      "step": 67
+    },
+    {
+      "epoch": 0.18604651162790697,
+      "grad_norm": 0.17978119850158691,
+      "learning_rate": 4.977360365964637e-05,
+      "loss": 1.6815,
+      "step": 68
+    },
+    {
+      "epoch": 0.18878248974008208,
+      "grad_norm": 0.16976381838321686,
+      "learning_rate": 4.976366905410562e-05,
+      "loss": 1.7212,
+      "step": 69
+    },
+    {
+      "epoch": 0.19151846785225718,
+      "grad_norm": 0.18810401856899261,
+      "learning_rate": 4.9753522165679195e-05,
+      "loss": 1.707,
+      "step": 70
+    },
+    {
+      "epoch": 0.19425444596443228,
+      "grad_norm": 0.17066176235675812,
+      "learning_rate": 4.974316308134978e-05,
+      "loss": 1.7765,
+      "step": 71
+    },
+    {
+      "epoch": 0.19699042407660738,
+      "grad_norm": 0.17143549025058746,
+      "learning_rate": 4.9732591889919114e-05,
+      "loss": 1.7317,
+      "step": 72
+    },
+    {
+      "epoch": 0.1997264021887825,
+      "grad_norm": 0.15364663302898407,
+      "learning_rate": 4.972180868200714e-05,
+      "loss": 1.6406,
+      "step": 73
+    },
+    {
+      "epoch": 0.2024623803009576,
+      "grad_norm": 0.1750328093767166,
+      "learning_rate": 4.971081355005133e-05,
+      "loss": 1.8737,
+      "step": 74
+    },
+    {
+      "epoch": 0.2051983584131327,
+      "grad_norm": 0.17950010299682617,
+      "learning_rate": 4.969960658830582e-05,
+      "loss": 1.7941,
+      "step": 75
+    },
+    {
+      "epoch": 0.2079343365253078,
+      "grad_norm": 0.1876641809940338,
+      "learning_rate": 4.968818789284063e-05,
+      "loss": 1.5792,
+      "step": 76
+    },
+    {
+      "epoch": 0.2106703146374829,
+      "grad_norm": 0.17048972845077515,
+      "learning_rate": 4.9676557561540845e-05,
+      "loss": 1.607,
+      "step": 77
+    },
+    {
+      "epoch": 0.213406292749658,
+      "grad_norm": 0.18965713679790497,
+      "learning_rate": 4.966471569410574e-05,
+      "loss": 1.6142,
+      "step": 78
+    },
+    {
+      "epoch": 0.2161422708618331,
+      "grad_norm": 0.16774089634418488,
+      "learning_rate": 4.965266239204798e-05,
+      "loss": 1.6446,
+      "step": 79
+    },
+    {
+      "epoch": 0.2188782489740082,
+      "grad_norm": 0.16337555646896362,
+      "learning_rate": 4.9640397758692715e-05,
+      "loss": 1.6319,
+      "step": 80
+    },
+    {
+      "epoch": 0.2216142270861833,
+      "grad_norm": 0.17532555758953094,
+      "learning_rate": 4.9627921899176686e-05,
+      "loss": 1.554,
+      "step": 81
+    },
+    {
+      "epoch": 0.2243502051983584,
+      "grad_norm": 0.16369390487670898,
+      "learning_rate": 4.9615234920447365e-05,
+      "loss": 1.7499,
+      "step": 82
+    },
+    {
+      "epoch": 0.2270861833105335,
+      "grad_norm": 0.170010045170784,
+      "learning_rate": 4.9602336931261995e-05,
+      "loss": 1.7786,
+      "step": 83
+    },
+    {
+      "epoch": 0.22982216142270862,
+      "grad_norm": 0.18399612605571747,
+      "learning_rate": 4.958922804218667e-05,
+      "loss": 1.8338,
+      "step": 84
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "grad_norm": 0.16553553938865662,
+      "learning_rate": 4.957590836559541e-05,
+      "loss": 1.8186,
+      "step": 85
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 0.17777277529239655,
+      "learning_rate": 4.9562378015669166e-05,
+      "loss": 1.7326,
+      "step": 86
+    },
+    {
+      "epoch": 0.23803009575923392,
+      "grad_norm": 0.18170584738254547,
+      "learning_rate": 4.954863710839485e-05,
+      "loss": 1.7293,
+      "step": 87
+    },
+    {
+      "epoch": 0.24076607387140903,
+      "grad_norm": 0.17840476334095,
+      "learning_rate": 4.953468576156437e-05,
+      "loss": 1.6941,
+      "step": 88
+    },
+    {
+      "epoch": 0.24350205198358413,
+      "grad_norm": 0.16968044638633728,
+      "learning_rate": 4.9520524094773575e-05,
+      "loss": 1.7643,
+      "step": 89
+    },
+    {
+      "epoch": 0.24623803009575923,
+      "grad_norm": 0.17756614089012146,
+      "learning_rate": 4.950615222942125e-05,
+      "loss": 1.7599,
+      "step": 90
+    },
+    {
+      "epoch": 0.24897400820793433,
+      "grad_norm": 0.162063866853714,
+      "learning_rate": 4.94915702887081e-05,
+      "loss": 1.7365,
+      "step": 91
+    },
+    {
+      "epoch": 0.25170998632010944,
+      "grad_norm": 0.18559984862804413,
+      "learning_rate": 4.947677839763563e-05,
+      "loss": 1.4424,
+      "step": 92
+    },
+    {
+      "epoch": 0.25444596443228457,
+      "grad_norm": 0.17151783406734467,
+      "learning_rate": 4.946177668300515e-05,
+      "loss": 1.5231,
+      "step": 93
+    },
+    {
+      "epoch": 0.25718194254445964,
+      "grad_norm": 0.20623093843460083,
+      "learning_rate": 4.944656527341663e-05,
+      "loss": 1.8024,
+      "step": 94
+    },
+    {
+      "epoch": 0.25991792065663477,
+      "grad_norm": 0.19318287074565887,
+      "learning_rate": 4.943114429926763e-05,
+      "loss": 1.6366,
+      "step": 95
+    },
+    {
+      "epoch": 0.26265389876880985,
+      "grad_norm": 0.1877240240573883,
+      "learning_rate": 4.941551389275217e-05,
+      "loss": 1.5424,
+      "step": 96
+    },
+    {
+      "epoch": 0.265389876880985,
+      "grad_norm": 0.17889100313186646,
+      "learning_rate": 4.939967418785959e-05,
+      "loss": 1.6603,
+      "step": 97
+    },
+    {
+      "epoch": 0.26812585499316005,
+      "grad_norm": 0.18533948063850403,
+      "learning_rate": 4.938362532037341e-05,
+      "loss": 1.7546,
+      "step": 98
+    },
+    {
+      "epoch": 0.2708618331053352,
+      "grad_norm": 0.19442051649093628,
+      "learning_rate": 4.936736742787018e-05,
+      "loss": 1.6938,
+      "step": 99
+    },
+    {
+      "epoch": 0.27359781121751026,
+      "grad_norm": 0.20610813796520233,
+      "learning_rate": 4.935090064971827e-05,
+      "loss": 1.7104,
+      "step": 100
+    },
+    {
+      "epoch": 0.2763337893296854,
+      "grad_norm": 0.19741714000701904,
+      "learning_rate": 4.933422512707671e-05,
+      "loss": 1.6205,
+      "step": 101
+    },
+    {
+      "epoch": 0.27906976744186046,
+      "grad_norm": 0.20613451302051544,
+      "learning_rate": 4.931734100289393e-05,
+      "loss": 1.7549,
+      "step": 102
+    },
+    {
+      "epoch": 0.2818057455540356,
+      "grad_norm": 0.19265608489513397,
+      "learning_rate": 4.930024842190658e-05,
+      "loss": 1.8493,
+      "step": 103
+    },
+    {
+      "epoch": 0.28454172366621067,
+      "grad_norm": 0.1976325511932373,
+      "learning_rate": 4.9282947530638295e-05,
+      "loss": 1.8059,
+      "step": 104
+    },
+    {
+      "epoch": 0.2872777017783858,
+      "grad_norm": 0.20444822311401367,
+      "learning_rate": 4.926543847739841e-05,
+      "loss": 1.5852,
+      "step": 105
+    },
+    {
+      "epoch": 0.29001367989056087,
+      "grad_norm": 0.18967504799365997,
+      "learning_rate": 4.9247721412280667e-05,
+      "loss": 1.6282,
+      "step": 106
+    },
+    {
+      "epoch": 0.292749658002736,
+      "grad_norm": 0.206316277384758,
+      "learning_rate": 4.9229796487162e-05,
+      "loss": 1.6186,
+      "step": 107
+    },
+    {
+      "epoch": 0.2954856361149111,
+      "grad_norm": 0.17524588108062744,
+      "learning_rate": 4.9211663855701165e-05,
+      "loss": 1.6383,
+      "step": 108
+    },
+    {
+      "epoch": 0.2982216142270862,
+      "grad_norm": 0.2032768577337265,
+      "learning_rate": 4.9193323673337476e-05,
+      "loss": 1.785,
+      "step": 109
+    },
+    {
+      "epoch": 0.3009575923392613,
+      "grad_norm": 0.17396964132785797,
+      "learning_rate": 4.917477609728941e-05,
+      "loss": 1.4461,
+      "step": 110
+    },
+    {
+      "epoch": 0.3036935704514364,
+      "grad_norm": 0.19011397659778595,
+      "learning_rate": 4.915602128655333e-05,
+      "loss": 1.7313,
+      "step": 111
+    },
+    {
+      "epoch": 0.3064295485636115,
+      "grad_norm": 0.19239680469036102,
+      "learning_rate": 4.913705940190207e-05,
+      "loss": 1.6015,
+      "step": 112
+    },
+    {
+      "epoch": 0.3091655266757866,
+      "grad_norm": 0.19917477667331696,
+      "learning_rate": 4.9117890605883574e-05,
+      "loss": 1.6411,
+      "step": 113
+    },
+    {
+      "epoch": 0.3119015047879617,
+      "grad_norm": 0.19318480789661407,
+      "learning_rate": 4.909851506281952e-05,
+      "loss": 1.6737,
+      "step": 114
+    },
+    {
+      "epoch": 0.3146374829001368,
+      "grad_norm": 0.1994866579771042,
+      "learning_rate": 4.907893293880387e-05,
+      "loss": 1.518,
+      "step": 115
+    },
+    {
+      "epoch": 0.3173734610123119,
+      "grad_norm": 0.1941988468170166,
+      "learning_rate": 4.905914440170147e-05,
+      "loss": 1.7589,
+      "step": 116
+    },
+    {
+      "epoch": 0.320109439124487,
+      "grad_norm": 0.20112977921962738,
+      "learning_rate": 4.903914962114665e-05,
+      "loss": 1.6671,
+      "step": 117
+    },
+    {
+      "epoch": 0.3228454172366621,
+      "grad_norm": 0.21141856908798218,
+      "learning_rate": 4.9018948768541694e-05,
+      "loss": 1.6731,
+      "step": 118
+    },
+    {
+      "epoch": 0.32558139534883723,
+      "grad_norm": 0.18585005402565002,
+      "learning_rate": 4.899854201705542e-05,
+      "loss": 1.6276,
+      "step": 119
+    },
+    {
+      "epoch": 0.3283173734610123,
+      "grad_norm": 0.1975102424621582,
+      "learning_rate": 4.8977929541621676e-05,
+      "loss": 1.6103,
+      "step": 120
+    },
+    {
+      "epoch": 0.33105335157318744,
+      "grad_norm": 0.19365696609020233,
+      "learning_rate": 4.895711151893786e-05,
+      "loss": 1.5401,
+      "step": 121
+    },
+    {
+      "epoch": 0.3337893296853625,
+      "grad_norm": 0.20823922753334045,
+      "learning_rate": 4.8936088127463396e-05,
+      "loss": 1.5811,
+      "step": 122
+    },
+    {
+      "epoch": 0.33652530779753764,
+      "grad_norm": 0.2523277997970581,
+      "learning_rate": 4.8914859547418186e-05,
+      "loss": 1.5561,
+      "step": 123
+    },
+    {
+      "epoch": 0.3392612859097127,
+      "grad_norm": 0.19290074706077576,
+      "learning_rate": 4.889342596078109e-05,
+      "loss": 1.6078,
+      "step": 124
+    },
+    {
+      "epoch": 0.34199726402188785,
+      "grad_norm": 0.17994068562984467,
+      "learning_rate": 4.887178755128835e-05,
+      "loss": 1.7232,
+      "step": 125
+    },
+    {
+      "epoch": 0.3447332421340629,
+      "grad_norm": 0.18888920545578003,
+      "learning_rate": 4.8849944504432035e-05,
+      "loss": 1.5538,
+      "step": 126
+    },
+    {
+      "epoch": 0.34746922024623805,
+      "grad_norm": 0.18798014521598816,
+      "learning_rate": 4.882789700745843e-05,
+      "loss": 1.6081,
+      "step": 127
+    },
+    {
+      "epoch": 0.35020519835841313,
+      "grad_norm": 0.19080984592437744,
+      "learning_rate": 4.880564524936643e-05,
+      "loss": 1.4809,
+      "step": 128
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 0.19421504437923431,
+      "learning_rate": 4.878318942090594e-05,
+      "loss": 1.6523,
+      "step": 129
+    },
+    {
+      "epoch": 0.35567715458276333,
+      "grad_norm": 0.1982857584953308,
+      "learning_rate": 4.876052971457623e-05,
+      "loss": 1.6948,
+      "step": 130
+    },
+    {
+      "epoch": 0.35841313269493846,
+      "grad_norm": 0.20162495970726013,
+      "learning_rate": 4.8737666324624265e-05,
+      "loss": 1.7504,
+      "step": 131
+    },
+    {
+      "epoch": 0.36114911080711354,
+      "grad_norm": 0.20931164920330048,
+      "learning_rate": 4.8714599447043085e-05,
+      "loss": 1.4829,
+      "step": 132
+    },
+    {
+      "epoch": 0.36388508891928867,
+      "grad_norm": 0.21997657418251038,
+      "learning_rate": 4.869132927957007e-05,
+      "loss": 1.5546,
+      "step": 133
+    },
+    {
+      "epoch": 0.36662106703146374,
+      "grad_norm": 0.20648987591266632,
+      "learning_rate": 4.866785602168528e-05,
+      "loss": 1.6716,
+      "step": 134
+    },
+    {
+      "epoch": 0.3693570451436389,
+      "grad_norm": 0.20821620523929596,
+      "learning_rate": 4.864417987460975e-05,
+      "loss": 1.6197,
+      "step": 135
+    },
+    {
+      "epoch": 0.37209302325581395,
+      "grad_norm": 0.21084287762641907,
+      "learning_rate": 4.862030104130375e-05,
+      "loss": 1.6454,
+      "step": 136
+    },
+    {
+      "epoch": 0.3748290013679891,
+      "grad_norm": 0.21337808668613434,
+      "learning_rate": 4.859621972646503e-05,
+      "loss": 1.5733,
+      "step": 137
+    },
+    {
+      "epoch": 0.37756497948016415,
+      "grad_norm": 0.21093016862869263,
+      "learning_rate": 4.857193613652711e-05,
+      "loss": 1.6716,
+      "step": 138
+    },
+    {
+      "epoch": 0.3803009575923393,
+      "grad_norm": 0.20863549411296844,
+      "learning_rate": 4.854745047965745e-05,
+      "loss": 1.6117,
+      "step": 139
+    },
+    {
+      "epoch": 0.38303693570451436,
+      "grad_norm": 0.20271432399749756,
+      "learning_rate": 4.852276296575573e-05,
+      "loss": 1.5765,
+      "step": 140
+    },
+    {
+      "epoch": 0.3857729138166895,
+      "grad_norm": 0.20297355949878693,
+      "learning_rate": 4.849787380645201e-05,
+      "loss": 1.7296,
+      "step": 141
+    },
+    {
+      "epoch": 0.38850889192886456,
+      "grad_norm": 0.20244932174682617,
+      "learning_rate": 4.847278321510491e-05,
+      "loss": 1.7651,
+      "step": 142
+    },
+    {
+      "epoch": 0.3912448700410397,
+      "grad_norm": 0.20252294838428497,
+      "learning_rate": 4.844749140679982e-05,
+      "loss": 1.6427,
+      "step": 143
+    },
+    {
+      "epoch": 0.39398084815321477,
+      "grad_norm": 0.20163066685199738,
+      "learning_rate": 4.842199859834702e-05,
+      "loss": 1.6398,
+      "step": 144
+    },
+    {
+      "epoch": 0.3967168262653899,
+      "grad_norm": 0.21187250316143036,
+      "learning_rate": 4.839630500827982e-05,
+      "loss": 1.6173,
+      "step": 145
+    },
+    {
+      "epoch": 0.399452804377565,
+      "grad_norm": 0.18655481934547424,
+      "learning_rate": 4.837041085685273e-05,
+      "loss": 1.5184,
+      "step": 146
+    },
+    {
+      "epoch": 0.4021887824897401,
+      "grad_norm": 0.21056880056858063,
+      "learning_rate": 4.834431636603953e-05,
+      "loss": 1.5687,
+      "step": 147
+    },
+    {
+      "epoch": 0.4049247606019152,
+      "grad_norm": 0.2020365595817566,
+      "learning_rate": 4.831802175953138e-05,
+      "loss": 1.6445,
+      "step": 148
+    },
+    {
+      "epoch": 0.4076607387140903,
+      "grad_norm": 0.18952035903930664,
+      "learning_rate": 4.8291527262734886e-05,
+      "loss": 1.7945,
+      "step": 149
+    },
+    {
+      "epoch": 0.4103967168262654,
+      "grad_norm": 0.1896515041589737,
+      "learning_rate": 4.826483310277021e-05,
+      "loss": 1.546,
+      "step": 150
+    },
+    {
+      "epoch": 0.4131326949384405,
+      "grad_norm": 0.20328114926815033,
+      "learning_rate": 4.8237939508469075e-05,
+      "loss": 1.5621,
+      "step": 151
+    },
+    {
+      "epoch": 0.4158686730506156,
+      "grad_norm": 0.18790675699710846,
+      "learning_rate": 4.821084671037285e-05,
+      "loss": 1.6609,
+      "step": 152
+    },
+    {
+      "epoch": 0.4186046511627907,
+      "grad_norm": 0.20515787601470947,
+      "learning_rate": 4.8183554940730524e-05,
+      "loss": 1.671,
+      "step": 153
+    },
+    {
+      "epoch": 0.4213406292749658,
+      "grad_norm": 0.19960254430770874,
+      "learning_rate": 4.815606443349675e-05,
+      "loss": 1.4732,
+      "step": 154
+    },
+    {
+      "epoch": 0.4240766073871409,
+      "grad_norm": 0.21313494443893433,
+      "learning_rate": 4.812837542432984e-05,
+      "loss": 1.7094,
+      "step": 155
+    },
+    {
+      "epoch": 0.426812585499316,
+      "grad_norm": 0.1988086998462677,
+      "learning_rate": 4.8100488150589704e-05,
+      "loss": 1.6241,
+      "step": 156
+    },
+    {
+      "epoch": 0.42954856361149113,
+      "grad_norm": 0.1885240077972412,
+      "learning_rate": 4.80724028513359e-05,
+      "loss": 1.64,
+      "step": 157
+    },
+    {
+      "epoch": 0.4322845417236662,
+      "grad_norm": 0.19793474674224854,
+      "learning_rate": 4.8044119767325455e-05,
+      "loss": 1.6615,
+      "step": 158
+    },
+    {
+      "epoch": 0.43502051983584133,
+      "grad_norm": 0.2065788209438324,
+      "learning_rate": 4.801563914101096e-05,
+      "loss": 1.5414,
+      "step": 159
+    },
+    {
+      "epoch": 0.4377564979480164,
+      "grad_norm": 0.19855208694934845,
+      "learning_rate": 4.798696121653833e-05,
+      "loss": 1.7804,
+      "step": 160
+    },
+    {
+      "epoch": 0.44049247606019154,
+      "grad_norm": 0.21048115193843842,
+      "learning_rate": 4.795808623974485e-05,
+      "loss": 1.6356,
+      "step": 161
+    },
+    {
+      "epoch": 0.4432284541723666,
+      "grad_norm": 0.18900255858898163,
+      "learning_rate": 4.792901445815698e-05,
+      "loss": 1.7154,
+      "step": 162
+    },
+    {
+      "epoch": 0.44596443228454175,
+      "grad_norm": 0.21627336740493774,
+      "learning_rate": 4.7899746120988245e-05,
+      "loss": 1.5644,
+      "step": 163
+    },
+    {
+      "epoch": 0.4487004103967168,
+      "grad_norm": 0.22913017868995667,
+      "learning_rate": 4.7870281479137146e-05,
+      "loss": 1.5764,
+      "step": 164
+    },
+    {
+      "epoch": 0.45143638850889195,
+      "grad_norm": 0.1932600438594818,
+      "learning_rate": 4.784062078518496e-05,
+      "loss": 1.7484,
+      "step": 165
+    },
+    {
+      "epoch": 0.454172366621067,
+      "grad_norm": 0.2302398383617401,
+      "learning_rate": 4.781076429339359e-05,
+      "loss": 1.5941,
+      "step": 166
+    },
+    {
+      "epoch": 0.45690834473324216,
+      "grad_norm": 0.2317684292793274,
+      "learning_rate": 4.77807122597034e-05,
+      "loss": 1.6844,
+      "step": 167
+    },
+    {
+      "epoch": 0.45964432284541723,
+      "grad_norm": 0.2009151577949524,
+      "learning_rate": 4.775046494173098e-05,
+      "loss": 1.735,
+      "step": 168
+    },
+    {
+      "epoch": 0.46238030095759236,
+      "grad_norm": 0.23902402818202972,
+      "learning_rate": 4.7720022598767e-05,
+      "loss": 1.6606,
+      "step": 169
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "grad_norm": 0.22881416976451874,
+      "learning_rate": 4.768938549177393e-05,
+      "loss": 1.6456,
+      "step": 170
+    },
+    {
+      "epoch": 0.46785225718194257,
+      "grad_norm": 0.1795547604560852,
+      "learning_rate": 4.765855388338383e-05,
+      "loss": 1.6853,
+      "step": 171
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.1893799602985382,
+      "learning_rate": 4.7627528037896106e-05,
+      "loss": 1.5587,
+      "step": 172
+    },
+    {
+      "epoch": 0.47332421340629277,
+      "grad_norm": 0.21993231773376465,
+      "learning_rate": 4.759630822127522e-05,
+      "loss": 1.8,
+      "step": 173
+    },
+    {
+      "epoch": 0.47606019151846785,
+      "grad_norm": 0.2114606648683548,
+      "learning_rate": 4.756489470114842e-05,
+      "loss": 1.543,
+      "step": 174
+    },
+    {
+      "epoch": 0.478796169630643,
+      "grad_norm": 0.24682380259037018,
+      "learning_rate": 4.753328774680347e-05,
+      "loss": 1.6537,
+      "step": 175
+    },
+    {
+      "epoch": 0.48153214774281805,
+      "grad_norm": 0.21504569053649902,
+      "learning_rate": 4.750148762918629e-05,
+      "loss": 1.7841,
+      "step": 176
+    },
+    {
+      "epoch": 0.4842681258549932,
+      "grad_norm": 0.2285577356815338,
+      "learning_rate": 4.746949462089868e-05,
+      "loss": 1.5911,
+      "step": 177
+    },
+    {
+      "epoch": 0.48700410396716826,
+      "grad_norm": 0.20859524607658386,
+      "learning_rate": 4.743730899619598e-05,
+      "loss": 1.5816,
+      "step": 178
+    },
+    {
+      "epoch": 0.4897400820793434,
+      "grad_norm": 0.2234918773174286,
+      "learning_rate": 4.740493103098468e-05,
+      "loss": 1.5763,
+      "step": 179
+    },
+    {
+      "epoch": 0.49247606019151846,
+      "grad_norm": 0.20745696127414703,
+      "learning_rate": 4.7372361002820085e-05,
+      "loss": 1.4441,
+      "step": 180
+    },
+    {
+      "epoch": 0.4952120383036936,
+      "grad_norm": 0.18486277759075165,
+      "learning_rate": 4.733959919090396e-05,
+      "loss": 1.6772,
+      "step": 181
+    },
+    {
+      "epoch": 0.49794801641586867,
+      "grad_norm": 0.21360483765602112,
+      "learning_rate": 4.7306645876082066e-05,
+      "loss": 1.5955,
+      "step": 182
+    },
+    {
+      "epoch": 0.5006839945280438,
+      "grad_norm": 0.2167551964521408,
+      "learning_rate": 4.727350134084182e-05,
+      "loss": 1.6222,
+      "step": 183
+    },
+    {
+      "epoch": 0.5034199726402189,
+      "grad_norm": 0.2134450525045395,
+      "learning_rate": 4.7240165869309846e-05,
+      "loss": 1.6059,
+      "step": 184
+    },
+    {
+      "epoch": 0.506155950752394,
+      "grad_norm": 0.21351881325244904,
+      "learning_rate": 4.720663974724953e-05,
+      "loss": 1.7258,
+      "step": 185
+    },
+    {
+      "epoch": 0.5088919288645691,
+      "grad_norm": 0.1952323615550995,
+      "learning_rate": 4.71729232620586e-05,
+      "loss": 1.4591,
+      "step": 186
+    },
+    {
+      "epoch": 0.5116279069767442,
+      "grad_norm": 0.2462124228477478,
+      "learning_rate": 4.7139016702766615e-05,
+      "loss": 1.5708,
+      "step": 187
+    },
+    {
+      "epoch": 0.5143638850889193,
+      "grad_norm": 0.24391762912273407,
+      "learning_rate": 4.7104920360032545e-05,
+      "loss": 1.7813,
+      "step": 188
+    },
+    {
+      "epoch": 0.5170998632010944,
+      "grad_norm": 0.21117964386940002,
+      "learning_rate": 4.707063452614224e-05,
+      "loss": 1.6948,
+      "step": 189
+    },
+    {
+      "epoch": 0.5198358413132695,
+      "grad_norm": 0.20722931623458862,
+      "learning_rate": 4.703615949500593e-05,
+      "loss": 1.5925,
+      "step": 190
+    },
+    {
+      "epoch": 0.5225718194254446,
+      "grad_norm": 0.22132059931755066,
+      "learning_rate": 4.700149556215571e-05,
+      "loss": 1.6448,
+      "step": 191
+    },
+    {
+      "epoch": 0.5253077975376197,
+      "grad_norm": 0.22742822766304016,
+      "learning_rate": 4.696664302474302e-05,
+      "loss": 1.5866,
+      "step": 192
+    },
+    {
+      "epoch": 0.5280437756497948,
+      "grad_norm": 0.2043454349040985,
+      "learning_rate": 4.693160218153607e-05,
+      "loss": 1.704,
+      "step": 193
+    },
+    {
+      "epoch": 0.53077975376197,
+      "grad_norm": 0.21721500158309937,
+      "learning_rate": 4.68963733329173e-05,
+      "loss": 1.6768,
+      "step": 194
+    },
+    {
+      "epoch": 0.533515731874145,
+      "grad_norm": 0.2336382120847702,
+      "learning_rate": 4.6860956780880796e-05,
+      "loss": 1.5906,
+      "step": 195
+    },
+    {
+      "epoch": 0.5362517099863201,
+      "grad_norm": 0.2269752472639084,
+      "learning_rate": 4.6825352829029705e-05,
+      "loss": 1.5308,
+      "step": 196
+    },
+    {
+      "epoch": 0.5389876880984952,
+      "grad_norm": 0.21256719529628754,
+      "learning_rate": 4.6789561782573635e-05,
+      "loss": 1.4821,
+      "step": 197
+    },
+    {
+      "epoch": 0.5417236662106704,
+      "grad_norm": 0.21226170659065247,
+      "learning_rate": 4.675358394832603e-05,
+      "loss": 1.7818,
+      "step": 198
+    },
+    {
+      "epoch": 0.5444596443228454,
+      "grad_norm": 0.21833762526512146,
+      "learning_rate": 4.671741963470155e-05,
+      "loss": 1.6556,
+      "step": 199
+    },
+    {
+      "epoch": 0.5471956224350205,
+      "grad_norm": 0.21411390602588654,
+      "learning_rate": 4.668106915171341e-05,
+      "loss": 1.5892,
+      "step": 200
+    },
+    {
+      "epoch": 0.5499316005471956,
+      "grad_norm": 0.22407633066177368,
+      "learning_rate": 4.664453281097075e-05,
+      "loss": 1.6843,
+      "step": 201
+    },
+    {
+      "epoch": 0.5526675786593708,
+      "grad_norm": 0.20786328613758087,
+      "learning_rate": 4.660781092567593e-05,
+      "loss": 1.6561,
+      "step": 202
+    },
+    {
+      "epoch": 0.5554035567715458,
+      "grad_norm": 0.2268221080303192,
+      "learning_rate": 4.657090381062189e-05,
+      "loss": 1.585,
+      "step": 203
+    },
+    {
+      "epoch": 0.5581395348837209,
+      "grad_norm": 0.2261151373386383,
+      "learning_rate": 4.6533811782189385e-05,
+      "loss": 1.6033,
+      "step": 204
+    },
+    {
+      "epoch": 0.560875512995896,
+      "grad_norm": 0.19799037277698517,
+      "learning_rate": 4.649653515834436e-05,
+      "loss": 1.6294,
+      "step": 205
+    },
+    {
+      "epoch": 0.5636114911080712,
+      "grad_norm": 0.20902524888515472,
+      "learning_rate": 4.645907425863514e-05,
+      "loss": 1.7012,
+      "step": 206
+    },
+    {
+      "epoch": 0.5663474692202463,
+      "grad_norm": 0.2055850476026535,
+      "learning_rate": 4.642142940418973e-05,
+      "loss": 1.5587,
+      "step": 207
+    },
+    {
+      "epoch": 0.5690834473324213,
+      "grad_norm": 0.2205619066953659,
+      "learning_rate": 4.638360091771309e-05,
+      "loss": 1.7727,
+      "step": 208
+    },
+    {
+      "epoch": 0.5718194254445964,
+      "grad_norm": 0.25723960995674133,
+      "learning_rate": 4.6345589123484314e-05,
+      "loss": 1.6037,
+      "step": 209
+    },
+    {
+      "epoch": 0.5745554035567716,
+      "grad_norm": 0.24420645833015442,
+      "learning_rate": 4.630739434735387e-05,
+      "loss": 1.7479,
+      "step": 210
+    },
+    {
+      "epoch": 0.5772913816689467,
+      "grad_norm": 0.24823623895645142,
+      "learning_rate": 4.626901691674083e-05,
+      "loss": 1.7085,
+      "step": 211
+    },
+    {
+      "epoch": 0.5800273597811217,
+      "grad_norm": 0.1966015100479126,
+      "learning_rate": 4.623045716063002e-05,
+      "loss": 1.544,
+      "step": 212
+    },
+    {
+      "epoch": 0.5827633378932968,
+      "grad_norm": 0.22418932616710663,
+      "learning_rate": 4.6191715409569244e-05,
+      "loss": 1.6949,
+      "step": 213
+    },
+    {
+      "epoch": 0.585499316005472,
+      "grad_norm": 0.24894970655441284,
+      "learning_rate": 4.6152791995666445e-05,
+      "loss": 1.5838,
+      "step": 214
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.230922132730484,
+      "learning_rate": 4.61136872525868e-05,
+      "loss": 1.6267,
+      "step": 215
+    },
+    {
+      "epoch": 0.5909712722298222,
+      "grad_norm": 0.22752411663532257,
+      "learning_rate": 4.6074401515549934e-05,
+      "loss": 1.4752,
+      "step": 216
+    },
+    {
+      "epoch": 0.5937072503419972,
+      "grad_norm": 0.22325444221496582,
+      "learning_rate": 4.6034935121327025e-05,
+      "loss": 1.7513,
+      "step": 217
+    },
+    {
+      "epoch": 0.5964432284541724,
+      "grad_norm": 0.21255116164684296,
+      "learning_rate": 4.599528840823787e-05,
+      "loss": 1.642,
+      "step": 218
+    },
+    {
+      "epoch": 0.5991792065663475,
+      "grad_norm": 0.24686329066753387,
+      "learning_rate": 4.5955461716148065e-05,
+      "loss": 1.6138,
+      "step": 219
+    },
+    {
+      "epoch": 0.6019151846785226,
+      "grad_norm": 0.20993772149085999,
+      "learning_rate": 4.5915455386466014e-05,
+      "loss": 1.7554,
+      "step": 220
+    },
+    {
+      "epoch": 0.6046511627906976,
+      "grad_norm": 0.2285137176513672,
+      "learning_rate": 4.587526976214006e-05,
+      "loss": 1.6073,
+      "step": 221
+    },
+    {
+      "epoch": 0.6073871409028728,
+      "grad_norm": 0.22348423302173615,
+      "learning_rate": 4.5834905187655526e-05,
+      "loss": 1.6707,
+      "step": 222
+    },
+    {
+      "epoch": 0.6101231190150479,
+      "grad_norm": 0.23014281690120697,
+      "learning_rate": 4.5794362009031735e-05,
+      "loss": 1.4968,
+      "step": 223
+    },
+    {
+      "epoch": 0.612859097127223,
+      "grad_norm": 0.23900607228279114,
+      "learning_rate": 4.575364057381909e-05,
+      "loss": 1.6317,
+      "step": 224
+    },
+    {
+      "epoch": 0.615595075239398,
+      "grad_norm": 0.24715475738048553,
+      "learning_rate": 4.571274123109606e-05,
+      "loss": 1.5805,
+      "step": 225
+    },
+    {
+      "epoch": 0.6183310533515732,
+      "grad_norm": 0.22451570630073547,
+      "learning_rate": 4.5671664331466205e-05,
+      "loss": 1.5956,
+      "step": 226
+    },
+    {
+      "epoch": 0.6210670314637483,
+      "grad_norm": 0.22505944967269897,
+      "learning_rate": 4.5630410227055154e-05,
+      "loss": 1.64,
+      "step": 227
+    },
+    {
+      "epoch": 0.6238030095759234,
+      "grad_norm": 0.24184450507164001,
+      "learning_rate": 4.5588979271507625e-05,
+      "loss": 1.6055,
+      "step": 228
+    },
+    {
+      "epoch": 0.6265389876880985,
+      "grad_norm": 0.22158478200435638,
+      "learning_rate": 4.5547371819984344e-05,
+      "loss": 1.5691,
+      "step": 229
+    },
+    {
+      "epoch": 0.6292749658002736,
+      "grad_norm": 0.23034054040908813,
+      "learning_rate": 4.5505588229159025e-05,
+      "loss": 1.6157,
+      "step": 230
+    },
+    {
+      "epoch": 0.6320109439124487,
+      "grad_norm": 0.240454763174057,
+      "learning_rate": 4.5463628857215306e-05,
+      "loss": 1.6337,
+      "step": 231
+    },
+    {
+      "epoch": 0.6347469220246238,
+      "grad_norm": 0.2200327068567276,
+      "learning_rate": 4.5421494063843695e-05,
+      "loss": 1.4941,
+      "step": 232
+    },
+    {
+      "epoch": 0.6374829001367989,
+      "grad_norm": 0.23234665393829346,
+      "learning_rate": 4.537918421023848e-05,
+      "loss": 1.5992,
+      "step": 233
+    },
+    {
+      "epoch": 0.640218878248974,
+      "grad_norm": 0.23742978274822235,
+      "learning_rate": 4.53366996590946e-05,
+      "loss": 1.481,
+      "step": 234
+    },
+    {
+      "epoch": 0.6429548563611491,
+      "grad_norm": 0.22323279082775116,
+      "learning_rate": 4.5294040774604576e-05,
+      "loss": 1.5291,
+      "step": 235
+    },
+    {
+      "epoch": 0.6456908344733242,
+      "grad_norm": 0.21182994544506073,
+      "learning_rate": 4.525120792245538e-05,
+      "loss": 1.5151,
+      "step": 236
+    },
+    {
+      "epoch": 0.6484268125854993,
+      "grad_norm": 0.24394573271274567,
+      "learning_rate": 4.5208201469825304e-05,
+      "loss": 1.6732,
+      "step": 237
+    },
+    {
+      "epoch": 0.6511627906976745,
+      "grad_norm": 0.20500163733959198,
+      "learning_rate": 4.516502178538079e-05,
+      "loss": 1.7066,
+      "step": 238
+    },
+    {
+      "epoch": 0.6538987688098495,
+      "grad_norm": 0.2329857349395752,
+      "learning_rate": 4.512166923927329e-05,
+      "loss": 1.4977,
+      "step": 239
+    },
+    {
+      "epoch": 0.6566347469220246,
+      "grad_norm": 0.23349729180335999,
+      "learning_rate": 4.507814420313608e-05,
+      "loss": 1.5837,
+      "step": 240
+    },
+    {
+      "epoch": 0.6593707250341997,
+      "grad_norm": 0.24870243668556213,
+      "learning_rate": 4.503444705008107e-05,
+      "loss": 1.6073,
+      "step": 241
+    },
+    {
+      "epoch": 0.6621067031463749,
+      "grad_norm": 0.2594373822212219,
+      "learning_rate": 4.4990578154695676e-05,
+      "loss": 1.6752,
+      "step": 242
+    },
+    {
+      "epoch": 0.66484268125855,
+      "grad_norm": 0.2545240819454193,
+      "learning_rate": 4.4946537893039476e-05,
+      "loss": 1.5185,
+      "step": 243
+    },
+    {
+      "epoch": 0.667578659370725,
+      "grad_norm": 0.24223731458187103,
+      "learning_rate": 4.4902326642641095e-05,
+      "loss": 1.7599,
+      "step": 244
+    },
+    {
+      "epoch": 0.6703146374829001,
+      "grad_norm": 0.23293475806713104,
+      "learning_rate": 4.485794478249493e-05,
+      "loss": 1.6919,
+      "step": 245
+    },
+    {
+      "epoch": 0.6730506155950753,
+      "grad_norm": 0.2452557235956192,
+      "learning_rate": 4.481339269305792e-05,
+      "loss": 1.4946,
+      "step": 246
+    },
+    {
+      "epoch": 0.6757865937072504,
+      "grad_norm": 0.23280148208141327,
+      "learning_rate": 4.4768670756246237e-05,
+      "loss": 1.548,
+      "step": 247
+    },
+    {
+      "epoch": 0.6785225718194254,
+      "grad_norm": 0.254730224609375,
+      "learning_rate": 4.4723779355432056e-05,
+      "loss": 1.6227,
+      "step": 248
+    },
+    {
+      "epoch": 0.6812585499316005,
+      "grad_norm": 0.25781765580177307,
+      "learning_rate": 4.4678718875440276e-05,
+      "loss": 1.6403,
+      "step": 249
+    },
+    {
+      "epoch": 0.6839945280437757,
+      "grad_norm": 0.2513294816017151,
+      "learning_rate": 4.463348970254518e-05,
+      "loss": 1.5416,
+      "step": 250
+    },
+    {
+      "epoch": 0.6839945280437757,
+      "eval_loss": 1.6065112352371216,
+      "eval_runtime": 13.3947,
+      "eval_samples_per_second": 11.497,
+      "eval_steps_per_second": 11.497,
+      "step": 250
+    },
+    {
+      "epoch": 0.6867305061559508,
+      "grad_norm": 0.26413506269454956,
+      "learning_rate": 4.458809222446717e-05,
+      "loss": 1.4294,
+      "step": 251
+    },
+    {
+      "epoch": 0.6894664842681258,
+      "grad_norm": 0.24414490163326263,
+      "learning_rate": 4.454252683036939e-05,
+      "loss": 1.6371,
+      "step": 252
+    },
+    {
+      "epoch": 0.6922024623803009,
+      "grad_norm": 0.23543910682201385,
+      "learning_rate": 4.449679391085444e-05,
+      "loss": 1.6226,
+      "step": 253
+    },
+    {
+      "epoch": 0.6949384404924761,
+      "grad_norm": 0.24758242070674896,
+      "learning_rate": 4.445089385796099e-05,
+      "loss": 1.7229,
+      "step": 254
+    },
+    {
+      "epoch": 0.6976744186046512,
+      "grad_norm": 0.2550957500934601,
+      "learning_rate": 4.4404827065160455e-05,
+      "loss": 1.6636,
+      "step": 255
+    },
+    {
+      "epoch": 0.7004103967168263,
+      "grad_norm": 0.22393615543842316,
+      "learning_rate": 4.4358593927353585e-05,
+      "loss": 1.5795,
+      "step": 256
+    },
+    {
+      "epoch": 0.7031463748290013,
+      "grad_norm": 0.22478732466697693,
+      "learning_rate": 4.431219484086712e-05,
+      "loss": 1.6236,
+      "step": 257
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 0.21901032328605652,
+      "learning_rate": 4.4265630203450355e-05,
+      "loss": 1.5929,
+      "step": 258
+    },
+    {
+      "epoch": 0.7086183310533516,
+      "grad_norm": 0.2183765321969986,
+      "learning_rate": 4.421890041427174e-05,
+      "loss": 1.6372,
+      "step": 259
+    },
+    {
+      "epoch": 0.7113543091655267,
+      "grad_norm": 0.23780032992362976,
+      "learning_rate": 4.417200587391547e-05,
+      "loss": 1.4413,
+      "step": 260
+    },
+    {
+      "epoch": 0.7140902872777017,
+      "grad_norm": 0.22141356766223907,
+      "learning_rate": 4.412494698437806e-05,
+      "loss": 1.6594,
+      "step": 261
+    },
+    {
+      "epoch": 0.7168262653898769,
+      "grad_norm": 0.2270331084728241,
+      "learning_rate": 4.407772414906487e-05,
+      "loss": 1.6701,
+      "step": 262
+    },
+    {
+      "epoch": 0.719562243502052,
+      "grad_norm": 0.24480143189430237,
+      "learning_rate": 4.403033777278666e-05,
+      "loss": 1.4914,
+      "step": 263
+    },
+    {
+      "epoch": 0.7222982216142271,
+      "grad_norm": 0.23042482137680054,
+      "learning_rate": 4.3982788261756114e-05,
+      "loss": 1.6845,
+      "step": 264
+    },
+    {
+      "epoch": 0.7250341997264022,
+      "grad_norm": 0.23995645344257355,
+      "learning_rate": 4.393507602358438e-05,
+      "loss": 1.658,
+      "step": 265
+    },
+    {
+      "epoch": 0.7277701778385773,
+      "grad_norm": 0.25735434889793396,
+      "learning_rate": 4.388720146727756e-05,
+      "loss": 1.4613,
+      "step": 266
+    },
+    {
+      "epoch": 0.7305061559507524,
+      "grad_norm": 0.27036193013191223,
+      "learning_rate": 4.3839165003233195e-05,
+      "loss": 1.5341,
+      "step": 267
+    },
+    {
+      "epoch": 0.7332421340629275,
+      "grad_norm": 0.2285039722919464,
+      "learning_rate": 4.3790967043236754e-05,
+      "loss": 1.453,
+      "step": 268
+    },
+    {
+      "epoch": 0.7359781121751026,
+      "grad_norm": 0.2619383931159973,
+      "learning_rate": 4.374260800045812e-05,
+      "loss": 1.4931,
+      "step": 269
+    },
+    {
+      "epoch": 0.7387140902872777,
+      "grad_norm": 0.22855600714683533,
+      "learning_rate": 4.369408828944801e-05,
+      "loss": 1.7458,
+      "step": 270
+    },
+    {
+      "epoch": 0.7414500683994528,
+      "grad_norm": 0.23299537599086761,
+      "learning_rate": 4.364540832613449e-05,
+      "loss": 1.5653,
+      "step": 271
+    },
+    {
+      "epoch": 0.7441860465116279,
+      "grad_norm": 0.25914594531059265,
+      "learning_rate": 4.35965685278193e-05,
+      "loss": 1.6876,
+      "step": 272
+    },
+    {
+      "epoch": 0.746922024623803,
+      "grad_norm": 0.2430303692817688,
+      "learning_rate": 4.3547569313174404e-05,
+      "loss": 1.5761,
+      "step": 273
+    },
+    {
+      "epoch": 0.7496580027359782,
+      "grad_norm": 0.23968341946601868,
+      "learning_rate": 4.34984111022383e-05,
+      "loss": 1.5577,
+      "step": 274
+    },
+    {
+      "epoch": 0.7523939808481532,
+      "grad_norm": 0.25374993681907654,
+      "learning_rate": 4.3449094316412485e-05,
+      "loss": 1.6936,
+      "step": 275
+    },
+    {
+      "epoch": 0.7551299589603283,
+      "grad_norm": 0.25317367911338806,
+      "learning_rate": 4.339961937845779e-05,
+      "loss": 1.6957,
+      "step": 276
+    },
+    {
+      "epoch": 0.7578659370725034,
+      "grad_norm": 0.231050044298172,
+      "learning_rate": 4.3349986712490795e-05,
+      "loss": 1.5386,
+      "step": 277
+    },
+    {
+      "epoch": 0.7606019151846786,
+      "grad_norm": 0.23644736409187317,
+      "learning_rate": 4.330019674398019e-05,
+      "loss": 1.6485,
+      "step": 278
+    },
+    {
+      "epoch": 0.7633378932968536,
+      "grad_norm": 0.24971966445446014,
+      "learning_rate": 4.32502498997431e-05,
+      "loss": 1.6196,
+      "step": 279
+    },
+    {
+      "epoch": 0.7660738714090287,
+      "grad_norm": 0.22572147846221924,
+      "learning_rate": 4.3200146607941444e-05,
+      "loss": 1.6654,
+      "step": 280
+    },
+    {
+      "epoch": 0.7688098495212038,
+      "grad_norm": 0.256597101688385,
+      "learning_rate": 4.3149887298078276e-05,
+      "loss": 1.6552,
+      "step": 281
+    },
+    {
+      "epoch": 0.771545827633379,
+      "grad_norm": 0.226451575756073,
+      "learning_rate": 4.3099472400994084e-05,
+      "loss": 1.6164,
+      "step": 282
+    },
+    {
+      "epoch": 0.774281805745554,
+      "grad_norm": 0.2507132291793823,
+      "learning_rate": 4.3048902348863116e-05,
+      "loss": 1.555,
+      "step": 283
+    },
+    {
+      "epoch": 0.7770177838577291,
+      "grad_norm": 0.23009182512760162,
+      "learning_rate": 4.2998177575189644e-05,
+      "loss": 1.4741,
+      "step": 284
+    },
+    {
+      "epoch": 0.7797537619699042,
+      "grad_norm": 0.24451418220996857,
+      "learning_rate": 4.294729851480429e-05,
+      "loss": 1.5704,
+      "step": 285
+    },
+    {
+      "epoch": 0.7824897400820794,
+      "grad_norm": 0.2464141994714737,
+      "learning_rate": 4.2896265603860246e-05,
+      "loss": 1.6357,
+      "step": 286
+    },
+    {
+      "epoch": 0.7852257181942545,
+      "grad_norm": 0.24296295642852783,
+      "learning_rate": 4.28450792798296e-05,
+      "loss": 1.5492,
+      "step": 287
+    },
+    {
+      "epoch": 0.7879616963064295,
+      "grad_norm": 0.2681036591529846,
+      "learning_rate": 4.279373998149954e-05,
+      "loss": 1.6973,
+      "step": 288
+    },
+    {
+      "epoch": 0.7906976744186046,
+      "grad_norm": 0.23149123787879944,
+      "learning_rate": 4.2742248148968576e-05,
+      "loss": 1.6626,
+      "step": 289
+    },
+    {
+      "epoch": 0.7934336525307798,
+      "grad_norm": 0.22642168402671814,
+      "learning_rate": 4.269060422364284e-05,
+      "loss": 1.7018,
+      "step": 290
+    },
+    {
+      "epoch": 0.7961696306429549,
+      "grad_norm": 0.23424457013607025,
+      "learning_rate": 4.263880864823221e-05,
+      "loss": 1.6973,
+      "step": 291
+    },
+    {
+      "epoch": 0.79890560875513,
+      "grad_norm": 0.24797867238521576,
+      "learning_rate": 4.2586861866746606e-05,
+      "loss": 1.6631,
+      "step": 292
+    },
+    {
+      "epoch": 0.801641586867305,
+      "grad_norm": 0.24017377197742462,
+      "learning_rate": 4.2534764324492115e-05,
+      "loss": 1.692,
+      "step": 293
+    },
+    {
+      "epoch": 0.8043775649794802,
+      "grad_norm": 0.22952371835708618,
+      "learning_rate": 4.248251646806719e-05,
+      "loss": 1.3692,
+      "step": 294
+    },
+    {
+      "epoch": 0.8071135430916553,
+      "grad_norm": 0.2534586787223816,
+      "learning_rate": 4.243011874535886e-05,
+      "loss": 1.7223,
+      "step": 295
+    },
+    {
+      "epoch": 0.8098495212038304,
+      "grad_norm": 0.23179998993873596,
+      "learning_rate": 4.237757160553883e-05,
+      "loss": 1.6688,
+      "step": 296
+    },
+    {
+      "epoch": 0.8125854993160054,
+      "grad_norm": 0.21562454104423523,
+      "learning_rate": 4.2324875499059693e-05,
+      "loss": 1.5706,
+      "step": 297
+    },
+    {
+      "epoch": 0.8153214774281806,
+      "grad_norm": 0.30047473311424255,
+      "learning_rate": 4.227203087765099e-05,
+      "loss": 1.5275,
+      "step": 298
+    },
+    {
+      "epoch": 0.8180574555403557,
+      "grad_norm": 0.23342545330524445,
+      "learning_rate": 4.221903819431543e-05,
+      "loss": 1.6569,
+      "step": 299
+    },
+    {
+      "epoch": 0.8207934336525308,
+      "grad_norm": 0.2696247100830078,
+      "learning_rate": 4.216589790332495e-05,
+      "loss": 1.5803,
+      "step": 300
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 0.24961510300636292,
+      "learning_rate": 4.21126104602168e-05,
+      "loss": 1.4535,
+      "step": 301
+    },
+    {
+      "epoch": 0.826265389876881,
+      "grad_norm": 0.24784626066684723,
+      "learning_rate": 4.205917632178972e-05,
+      "loss": 1.6035,
+      "step": 302
+    },
+    {
+      "epoch": 0.8290013679890561,
+      "grad_norm": 0.22941957414150238,
+      "learning_rate": 4.200559594609994e-05,
+      "loss": 1.6502,
+      "step": 303
+    },
+    {
+      "epoch": 0.8317373461012312,
+      "grad_norm": 0.23505181074142456,
+      "learning_rate": 4.195186979245728e-05,
+      "loss": 1.7577,
+      "step": 304
+    },
+    {
+      "epoch": 0.8344733242134063,
+      "grad_norm": 0.2516261041164398,
+      "learning_rate": 4.189799832142126e-05,
+      "loss": 1.4947,
+      "step": 305
+    },
+    {
+      "epoch": 0.8372093023255814,
+      "grad_norm": 0.28713610768318176,
+      "learning_rate": 4.1843981994797075e-05,
+      "loss": 1.8651,
+      "step": 306
+    },
+    {
+      "epoch": 0.8399452804377565,
+      "grad_norm": 0.2537023425102234,
+      "learning_rate": 4.178982127563169e-05,
+      "loss": 1.3971,
+      "step": 307
+    },
+    {
+      "epoch": 0.8426812585499316,
+      "grad_norm": 0.2575843930244446,
+      "learning_rate": 4.173551662820985e-05,
+      "loss": 1.7023,
+      "step": 308
+    },
+    {
+      "epoch": 0.8454172366621067,
+      "grad_norm": 0.2560933828353882,
+      "learning_rate": 4.168106851805009e-05,
+      "loss": 1.5333,
+      "step": 309
+    },
+    {
+      "epoch": 0.8481532147742818,
+      "grad_norm": 0.22800202667713165,
+      "learning_rate": 4.16264774119008e-05,
+      "loss": 1.5548,
+      "step": 310
+    },
+    {
+      "epoch": 0.8508891928864569,
+      "grad_norm": 0.2175893485546112,
+      "learning_rate": 4.1571743777736116e-05,
+      "loss": 1.7173,
+      "step": 311
+    },
+    {
+      "epoch": 0.853625170998632,
+      "grad_norm": 0.23807293176651,
+      "learning_rate": 4.151686808475204e-05,
+      "loss": 1.5548,
+      "step": 312
+    },
+    {
+      "epoch": 0.8563611491108071,
+      "grad_norm": 0.26670193672180176,
+      "learning_rate": 4.1461850803362314e-05,
+      "loss": 1.5273,
+      "step": 313
+    },
+    {
+      "epoch": 0.8590971272229823,
+      "grad_norm": 0.2579667866230011,
+      "learning_rate": 4.140669240519442e-05,
+      "loss": 1.6407,
+      "step": 314
+    },
+    {
+      "epoch": 0.8618331053351573,
+      "grad_norm": 0.23136496543884277,
+      "learning_rate": 4.135139336308559e-05,
+      "loss": 1.5134,
+      "step": 315
+    },
+    {
+      "epoch": 0.8645690834473324,
+      "grad_norm": 0.2494489699602127,
+      "learning_rate": 4.129595415107864e-05,
+      "loss": 1.5429,
+      "step": 316
+    },
+    {
+      "epoch": 0.8673050615595075,
+      "grad_norm": 0.2552199363708496,
+      "learning_rate": 4.124037524441803e-05,
+      "loss": 1.5348,
+      "step": 317
+    },
+    {
+      "epoch": 0.8700410396716827,
+      "grad_norm": 0.2449236959218979,
+      "learning_rate": 4.118465711954569e-05,
+      "loss": 1.7689,
+      "step": 318
+    },
+    {
+      "epoch": 0.8727770177838577,
+      "grad_norm": 0.2517363429069519,
+      "learning_rate": 4.112880025409701e-05,
+      "loss": 1.6258,
+      "step": 319
+    },
+    {
+      "epoch": 0.8755129958960328,
+      "grad_norm": 0.27004796266555786,
+      "learning_rate": 4.107280512689668e-05,
+      "loss": 1.5855,
+      "step": 320
+    },
+    {
+      "epoch": 0.8782489740082079,
+      "grad_norm": 0.27461642026901245,
+      "learning_rate": 4.101667221795465e-05,
+      "loss": 1.57,
+      "step": 321
+    },
+    {
+      "epoch": 0.8809849521203831,
+      "grad_norm": 0.2513730525970459,
+      "learning_rate": 4.0960402008461955e-05,
+      "loss": 1.5509,
+      "step": 322
+    },
+    {
+      "epoch": 0.8837209302325582,
+      "grad_norm": 0.2727186977863312,
+      "learning_rate": 4.090399498078664e-05,
+      "loss": 1.5905,
+      "step": 323
+    },
+    {
+      "epoch": 0.8864569083447332,
+      "grad_norm": 0.24590528011322021,
+      "learning_rate": 4.084745161846961e-05,
+      "loss": 1.5459,
+      "step": 324
+    },
+    {
+      "epoch": 0.8891928864569083,
+      "grad_norm": 0.22626182436943054,
+      "learning_rate": 4.079077240622043e-05,
+      "loss": 1.6127,
+      "step": 325
+    },
+    {
+      "epoch": 0.8919288645690835,
+      "grad_norm": 0.24243368208408356,
+      "learning_rate": 4.0733957829913296e-05,
+      "loss": 1.5687,
+      "step": 326
+    },
+    {
+      "epoch": 0.8946648426812586,
+      "grad_norm": 0.2647876739501953,
+      "learning_rate": 4.0677008376582716e-05,
+      "loss": 1.6761,
+      "step": 327
+    },
+    {
+      "epoch": 0.8974008207934336,
+      "grad_norm": 0.2398100048303604,
+      "learning_rate": 4.061992453441946e-05,
+      "loss": 1.5666,
+      "step": 328
+    },
+    {
+      "epoch": 0.9001367989056087,
+      "grad_norm": 0.2463439702987671,
+      "learning_rate": 4.056270679276631e-05,
+      "loss": 1.7682,
+      "step": 329
+    },
+    {
+      "epoch": 0.9028727770177839,
+      "grad_norm": 0.23695454001426697,
+      "learning_rate": 4.050535564211388e-05,
+      "loss": 1.4589,
+      "step": 330
+    },
+    {
+      "epoch": 0.905608755129959,
+      "grad_norm": 0.24376732110977173,
+      "learning_rate": 4.044787157409642e-05,
+      "loss": 1.4797,
+      "step": 331
+    },
+    {
+      "epoch": 0.908344733242134,
+      "grad_norm": 0.24585182964801788,
+      "learning_rate": 4.039025508148758e-05,
+      "loss": 1.5843,
+      "step": 332
+    },
+    {
+      "epoch": 0.9110807113543091,
+      "grad_norm": 0.266337126493454,
+      "learning_rate": 4.0332506658196225e-05,
+      "loss": 1.7075,
+      "step": 333
+    },
+    {
+      "epoch": 0.9138166894664843,
+      "grad_norm": 0.2664279043674469,
+      "learning_rate": 4.027462679926215e-05,
+      "loss": 1.6063,
+      "step": 334
+    },
+    {
+      "epoch": 0.9165526675786594,
+      "grad_norm": 0.2796095013618469,
+      "learning_rate": 4.021661600085189e-05,
+      "loss": 1.7072,
+      "step": 335
+    },
+    {
+      "epoch": 0.9192886456908345,
+      "grad_norm": 0.26439887285232544,
+      "learning_rate": 4.01584747602544e-05,
+      "loss": 1.5579,
+      "step": 336
+    },
+    {
+      "epoch": 0.9220246238030095,
+      "grad_norm": 0.2440110445022583,
+      "learning_rate": 4.010020357587687e-05,
+      "loss": 1.5581,
+      "step": 337
+    },
+    {
+      "epoch": 0.9247606019151847,
+      "grad_norm": 0.2567618787288666,
+      "learning_rate": 4.00418029472404e-05,
+      "loss": 1.7241,
+      "step": 338
+    },
+    {
+      "epoch": 0.9274965800273598,
+      "grad_norm": 0.26454323530197144,
+      "learning_rate": 3.9983273374975726e-05,
+      "loss": 1.5839,
+      "step": 339
+    },
+    {
+      "epoch": 0.9302325581395349,
+      "grad_norm": 0.2620810866355896,
+      "learning_rate": 3.9924615360818934e-05,
+      "loss": 1.7599,
+      "step": 340
+    },
+    {
+      "epoch": 0.93296853625171,
+      "grad_norm": 0.2683962285518646,
+      "learning_rate": 3.986582940760717e-05,
+      "loss": 1.6423,
+      "step": 341
+    },
+    {
+      "epoch": 0.9357045143638851,
+      "grad_norm": 0.22277259826660156,
+      "learning_rate": 3.9806916019274306e-05,
+      "loss": 1.546,
+      "step": 342
+    },
+    {
+      "epoch": 0.9384404924760602,
+      "grad_norm": 0.26338738203048706,
+      "learning_rate": 3.9747875700846646e-05,
+      "loss": 1.5739,
+      "step": 343
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.27578940987586975,
+      "learning_rate": 3.968870895843858e-05,
+      "loss": 1.6281,
+      "step": 344
+    },
+    {
+      "epoch": 0.9439124487004104,
+      "grad_norm": 0.24065442383289337,
+      "learning_rate": 3.962941629924824e-05,
+      "loss": 1.4456,
+      "step": 345
+    },
+    {
+      "epoch": 0.9466484268125855,
+      "grad_norm": 0.25176844000816345,
+      "learning_rate": 3.956999823155315e-05,
+      "loss": 1.6577,
+      "step": 346
+    },
+    {
+      "epoch": 0.9493844049247606,
+      "grad_norm": 0.2668481767177582,
+      "learning_rate": 3.951045526470592e-05,
+      "loss": 1.5013,
+      "step": 347
+    },
+    {
+      "epoch": 0.9521203830369357,
+      "grad_norm": 0.27450481057167053,
+      "learning_rate": 3.94507879091298e-05,
+      "loss": 1.5217,
+      "step": 348
+    },
+    {
+      "epoch": 0.9548563611491108,
+      "grad_norm": 0.2651226818561554,
+      "learning_rate": 3.939099667631438e-05,
+      "loss": 1.5877,
+      "step": 349
+    },
+    {
+      "epoch": 0.957592339261286,
+      "grad_norm": 0.27575191855430603,
+      "learning_rate": 3.933108207881112e-05,
+      "loss": 1.5836,
+      "step": 350
+    },
+    {
+      "epoch": 0.960328317373461,
+      "grad_norm": 0.248700350522995,
+      "learning_rate": 3.927104463022906e-05,
+      "loss": 1.6258,
+      "step": 351
+    },
+    {
+      "epoch": 0.9630642954856361,
+      "grad_norm": 0.27306076884269714,
+      "learning_rate": 3.921088484523032e-05,
+      "loss": 1.5255,
+      "step": 352
+    },
+    {
+      "epoch": 0.9658002735978112,
+      "grad_norm": 0.2685610055923462,
+      "learning_rate": 3.9150603239525765e-05,
+      "loss": 1.607,
+      "step": 353
+    },
+    {
+      "epoch": 0.9685362517099864,
+      "grad_norm": 0.2934316396713257,
+      "learning_rate": 3.909020032987051e-05,
+      "loss": 1.5252,
+      "step": 354
+    },
+    {
+      "epoch": 0.9712722298221614,
+      "grad_norm": 0.2768978774547577,
+      "learning_rate": 3.902967663405956e-05,
+      "loss": 1.6291,
+      "step": 355
+    },
+    {
+      "epoch": 0.9740082079343365,
+      "grad_norm": 0.29954272508621216,
+      "learning_rate": 3.8969032670923346e-05,
+      "loss": 1.481,
+      "step": 356
+    },
+    {
+      "epoch": 0.9767441860465116,
+      "grad_norm": 0.26432037353515625,
+      "learning_rate": 3.890826896032326e-05,
+      "loss": 1.8236,
+      "step": 357
+    },
+    {
+      "epoch": 0.9794801641586868,
+      "grad_norm": 0.2719402611255646,
+      "learning_rate": 3.884738602314719e-05,
+      "loss": 1.626,
+      "step": 358
+    },
+    {
+      "epoch": 0.9822161422708618,
+      "grad_norm": 0.2573629915714264,
+      "learning_rate": 3.8786384381305123e-05,
+      "loss": 1.4863,
+      "step": 359
+    },
+    {
+      "epoch": 0.9849521203830369,
+      "grad_norm": 0.27962571382522583,
+      "learning_rate": 3.87252645577246e-05,
+      "loss": 1.567,
+      "step": 360
+    },
+    {
+      "epoch": 0.987688098495212,
+      "grad_norm": 0.3235297203063965,
+      "learning_rate": 3.866402707634624e-05,
+      "loss": 1.5815,
+      "step": 361
+    },
+    {
+      "epoch": 0.9904240766073872,
+      "grad_norm": 0.24766239523887634,
+      "learning_rate": 3.8602672462119294e-05,
+      "loss": 1.6327,
+      "step": 362
+    },
+    {
+      "epoch": 0.9931600547195623,
+      "grad_norm": 0.2470933198928833,
+      "learning_rate": 3.85412012409971e-05,
+      "loss": 1.626,
+      "step": 363
+    },
+    {
+      "epoch": 0.9958960328317373,
+      "grad_norm": 0.2527347207069397,
+      "learning_rate": 3.847961393993261e-05,
+      "loss": 1.5373,
+      "step": 364
+    },
+    {
+      "epoch": 0.9986320109439124,
+      "grad_norm": 0.24633722007274628,
+      "learning_rate": 3.8417911086873834e-05,
+      "loss": 1.5994,
+      "step": 365
+    },
+    {
+      "epoch": 1.0013679890560876,
+      "grad_norm": 0.2634814381599426,
+      "learning_rate": 3.835609321075934e-05,
+      "loss": 1.5801,
+      "step": 366
+    },
+    {
+      "epoch": 1.0041039671682626,
+      "grad_norm": 0.2554261386394501,
+      "learning_rate": 3.8294160841513715e-05,
+      "loss": 1.6,
+      "step": 367
+    },
+    {
+      "epoch": 1.0068399452804377,
+      "grad_norm": 0.23322933912277222,
+      "learning_rate": 3.823211451004304e-05,
+      "loss": 1.4199,
+      "step": 368
+    },
+    {
+      "epoch": 1.009575923392613,
+      "grad_norm": 0.23709669709205627,
+      "learning_rate": 3.816995474823028e-05,
+      "loss": 1.5409,
+      "step": 369
+    },
+    {
+      "epoch": 1.012311901504788,
+      "grad_norm": 0.2459351122379303,
+      "learning_rate": 3.8107682088930794e-05,
+      "loss": 1.5137,
+      "step": 370
+    },
+    {
+      "epoch": 1.015047879616963,
+      "grad_norm": 0.2492617666721344,
+      "learning_rate": 3.8045297065967744e-05,
+      "loss": 1.6054,
+      "step": 371
+    },
+    {
+      "epoch": 1.0177838577291383,
+      "grad_norm": 0.2558859884738922,
+      "learning_rate": 3.798280021412749e-05,
+      "loss": 1.6189,
+      "step": 372
+    },
+    {
+      "epoch": 1.0205198358413132,
+      "grad_norm": 0.23855186998844147,
+      "learning_rate": 3.792019206915504e-05,
+      "loss": 1.7069,
+      "step": 373
+    },
+    {
+      "epoch": 1.0232558139534884,
+      "grad_norm": 0.2665524482727051,
+      "learning_rate": 3.7857473167749435e-05,
+      "loss": 1.5505,
+      "step": 374
+    },
+    {
+      "epoch": 1.0259917920656634,
+      "grad_norm": 0.24079938232898712,
+      "learning_rate": 3.779464404755919e-05,
+      "loss": 1.4712,
+      "step": 375
+    },
+    {
+      "epoch": 1.0287277701778386,
+      "grad_norm": 0.2564024329185486,
+      "learning_rate": 3.773170524717763e-05,
+      "loss": 1.5464,
+      "step": 376
+    },
+    {
+      "epoch": 1.0314637482900137,
+      "grad_norm": 0.23460035026073456,
+      "learning_rate": 3.766865730613828e-05,
+      "loss": 1.6146,
+      "step": 377
+    },
+    {
+      "epoch": 1.0341997264021887,
+      "grad_norm": 0.24386173486709595,
+      "learning_rate": 3.760550076491031e-05,
+      "loss": 1.4695,
+      "step": 378
+    },
+    {
+      "epoch": 1.036935704514364,
+      "grad_norm": 0.29760661721229553,
+      "learning_rate": 3.754223616489379e-05,
+      "loss": 1.3848,
+      "step": 379
+    },
+    {
+      "epoch": 1.039671682626539,
+      "grad_norm": 0.243236243724823,
+      "learning_rate": 3.7478864048415136e-05,
+      "loss": 1.4215,
+      "step": 380
+    },
+    {
+      "epoch": 1.042407660738714,
+      "grad_norm": 0.259069561958313,
+      "learning_rate": 3.7415384958722445e-05,
+      "loss": 1.5352,
+      "step": 381
+    },
+    {
+      "epoch": 1.0451436388508892,
+      "grad_norm": 0.281246542930603,
+      "learning_rate": 3.735179943998081e-05,
+      "loss": 1.6266,
+      "step": 382
+    },
+    {
+      "epoch": 1.0478796169630642,
+      "grad_norm": 0.2552565634250641,
+      "learning_rate": 3.728810803726767e-05,
+      "loss": 1.6849,
+      "step": 383
+    },
+    {
+      "epoch": 1.0506155950752394,
+      "grad_norm": 0.25851351022720337,
+      "learning_rate": 3.7224311296568134e-05,
+      "loss": 1.4284,
+      "step": 384
+    },
+    {
+      "epoch": 1.0533515731874146,
+      "grad_norm": 0.2506335973739624,
+      "learning_rate": 3.716040976477033e-05,
+      "loss": 1.4498,
+      "step": 385
+    },
+    {
+      "epoch": 1.0560875512995895,
+      "grad_norm": 0.3301680088043213,
+      "learning_rate": 3.709640398966067e-05,
+      "loss": 1.5649,
+      "step": 386
+    },
+    {
+      "epoch": 1.0588235294117647,
+      "grad_norm": 0.2851655185222626,
+      "learning_rate": 3.703229451991918e-05,
+      "loss": 1.6324,
+      "step": 387
+    },
+    {
+      "epoch": 1.06155950752394,
+      "grad_norm": 0.27497461438179016,
+      "learning_rate": 3.69680819051148e-05,
+      "loss": 1.5628,
+      "step": 388
+    },
+    {
+      "epoch": 1.0642954856361149,
+      "grad_norm": 0.2582625448703766,
+      "learning_rate": 3.6903766695700656e-05,
+      "loss": 1.5979,
+      "step": 389
+    },
+    {
+      "epoch": 1.06703146374829,
+      "grad_norm": 0.2679811418056488,
+      "learning_rate": 3.6839349443009364e-05,
+      "loss": 1.51,
+      "step": 390
+    },
+    {
+      "epoch": 1.069767441860465,
+      "grad_norm": 0.2523844242095947,
+      "learning_rate": 3.677483069924827e-05,
+      "loss": 1.6217,
+      "step": 391
+    },
+    {
+      "epoch": 1.0725034199726402,
+      "grad_norm": 0.23916080594062805,
+      "learning_rate": 3.671021101749476e-05,
+      "loss": 1.6128,
+      "step": 392
+    },
+    {
+      "epoch": 1.0752393980848154,
+      "grad_norm": 0.2704644203186035,
+      "learning_rate": 3.664549095169148e-05,
+      "loss": 1.5524,
+      "step": 393
+    },
+    {
+      "epoch": 1.0779753761969904,
+      "grad_norm": 0.27870067954063416,
+      "learning_rate": 3.6580671056641616e-05,
+      "loss": 1.6693,
+      "step": 394
+    },
+    {
+      "epoch": 1.0807113543091655,
+      "grad_norm": 0.2483988106250763,
+      "learning_rate": 3.6515751888004113e-05,
+      "loss": 1.4771,
+      "step": 395
+    },
+    {
+      "epoch": 1.0834473324213407,
+      "grad_norm": 0.2584916949272156,
+      "learning_rate": 3.645073400228895e-05,
+      "loss": 1.5099,
+      "step": 396
+    },
+    {
+      "epoch": 1.0861833105335157,
+      "grad_norm": 0.24778081476688385,
+      "learning_rate": 3.6385617956852286e-05,
+      "loss": 1.6118,
+      "step": 397
+    },
+    {
+      "epoch": 1.0889192886456909,
+      "grad_norm": 0.2260492891073227,
+      "learning_rate": 3.632040430989181e-05,
+      "loss": 1.5512,
+      "step": 398
+    },
+    {
+      "epoch": 1.0916552667578658,
+      "grad_norm": 0.2751753330230713,
+      "learning_rate": 3.6255093620441834e-05,
+      "loss": 1.559,
+      "step": 399
+    },
+    {
+      "epoch": 1.094391244870041,
+      "grad_norm": 0.2648756206035614,
+      "learning_rate": 3.618968644836859e-05,
+      "loss": 1.5666,
+      "step": 400
+    },
+    {
+      "epoch": 1.0971272229822162,
+      "grad_norm": 0.24056974053382874,
+      "learning_rate": 3.612418335436536e-05,
+      "loss": 1.5169,
+      "step": 401
+    },
+    {
+      "epoch": 1.0998632010943912,
+      "grad_norm": 0.23621565103530884,
+      "learning_rate": 3.605858489994771e-05,
+      "loss": 1.7854,
+      "step": 402
+    },
+    {
+      "epoch": 1.1025991792065664,
+      "grad_norm": 0.25893279910087585,
+      "learning_rate": 3.5992891647448696e-05,
+      "loss": 1.6353,
+      "step": 403
+    },
+    {
+      "epoch": 1.1053351573187415,
+      "grad_norm": 0.2806542217731476,
+      "learning_rate": 3.592710416001398e-05,
+      "loss": 1.7623,
+      "step": 404
+    },
+    {
+      "epoch": 1.1080711354309165,
+      "grad_norm": 0.29038006067276,
+      "learning_rate": 3.586122300159707e-05,
+      "loss": 1.4457,
+      "step": 405
+    },
+    {
+      "epoch": 1.1108071135430917,
+      "grad_norm": 0.27859950065612793,
+      "learning_rate": 3.5795248736954426e-05,
+      "loss": 1.4407,
+      "step": 406
+    },
+    {
+      "epoch": 1.1135430916552667,
+      "grad_norm": 0.2583613991737366,
+      "learning_rate": 3.5729181931640674e-05,
+      "loss": 1.7205,
+      "step": 407
+    },
+    {
+      "epoch": 1.1162790697674418,
+      "grad_norm": 0.25556913018226624,
+      "learning_rate": 3.5663023152003705e-05,
+      "loss": 1.5662,
+      "step": 408
+    },
+    {
+      "epoch": 1.119015047879617,
+      "grad_norm": 0.26569557189941406,
+      "learning_rate": 3.559677296517987e-05,
+      "loss": 1.5788,
+      "step": 409
+    },
+    {
+      "epoch": 1.121751025991792,
+      "grad_norm": 0.2778777778148651,
+      "learning_rate": 3.5530431939089084e-05,
+      "loss": 1.747,
+      "step": 410
+    },
+    {
+      "epoch": 1.1244870041039672,
+      "grad_norm": 0.29031357169151306,
+      "learning_rate": 3.546400064242997e-05,
+      "loss": 1.6342,
+      "step": 411
+    },
+    {
+      "epoch": 1.1272229822161424,
+      "grad_norm": 0.26415035128593445,
+      "learning_rate": 3.5397479644674964e-05,
+      "loss": 1.5624,
+      "step": 412
+    },
+    {
+      "epoch": 1.1299589603283173,
+      "grad_norm": 0.2662270963191986,
+      "learning_rate": 3.533086951606549e-05,
+      "loss": 1.6745,
+      "step": 413
+    },
+    {
+      "epoch": 1.1326949384404925,
+      "grad_norm": 0.2567162811756134,
+      "learning_rate": 3.5264170827607004e-05,
+      "loss": 1.6727,
+      "step": 414
+    },
+    {
+      "epoch": 1.1354309165526675,
+      "grad_norm": 0.28177884221076965,
+      "learning_rate": 3.519738415106413e-05,
+      "loss": 1.4554,
+      "step": 415
+    },
+    {
+      "epoch": 1.1381668946648427,
+      "grad_norm": 0.277425616979599,
+      "learning_rate": 3.513051005895576e-05,
+      "loss": 1.5096,
+      "step": 416
+    },
+    {
+      "epoch": 1.1409028727770179,
+      "grad_norm": 0.2886553108692169,
+      "learning_rate": 3.506354912455016e-05,
+      "loss": 1.6723,
+      "step": 417
+    },
+    {
+      "epoch": 1.1436388508891928,
+      "grad_norm": 0.245221808552742,
+      "learning_rate": 3.499650192186001e-05,
+      "loss": 1.4508,
+      "step": 418
+    },
+    {
+      "epoch": 1.146374829001368,
+      "grad_norm": 0.31830281019210815,
+      "learning_rate": 3.492936902563754e-05,
+      "loss": 1.4235,
+      "step": 419
+    },
+    {
+      "epoch": 1.1491108071135432,
+      "grad_norm": 0.2477853149175644,
+      "learning_rate": 3.486215101136954e-05,
+      "loss": 1.4991,
+      "step": 420
+    },
+    {
+      "epoch": 1.1518467852257182,
+      "grad_norm": 0.2735631465911865,
+      "learning_rate": 3.47948484552725e-05,
+      "loss": 1.5141,
+      "step": 421
+    },
+    {
+      "epoch": 1.1545827633378933,
+      "grad_norm": 0.263138085603714,
+      "learning_rate": 3.47274619342876e-05,
+      "loss": 1.7006,
+      "step": 422
+    },
+    {
+      "epoch": 1.1573187414500685,
+      "grad_norm": 0.2716856598854065,
+      "learning_rate": 3.465999202607583e-05,
+      "loss": 1.6627,
+      "step": 423
+    },
+    {
+      "epoch": 1.1600547195622435,
+      "grad_norm": 0.26602792739868164,
+      "learning_rate": 3.459243930901297e-05,
+      "loss": 1.5501,
+      "step": 424
+    },
+    {
+      "epoch": 1.1627906976744187,
+      "grad_norm": 0.27663204073905945,
+      "learning_rate": 3.452480436218471e-05,
+      "loss": 1.7229,
+      "step": 425
+    },
+    {
+      "epoch": 1.1655266757865936,
+      "grad_norm": 0.3033023178577423,
+      "learning_rate": 3.4457087765381584e-05,
+      "loss": 1.5789,
+      "step": 426
+    },
+    {
+      "epoch": 1.1682626538987688,
+      "grad_norm": 0.2735162675380707,
+      "learning_rate": 3.43892900990941e-05,
+      "loss": 1.6559,
+      "step": 427
+    },
+    {
+      "epoch": 1.170998632010944,
+      "grad_norm": 0.26795685291290283,
+      "learning_rate": 3.432141194450772e-05,
+      "loss": 1.4425,
+      "step": 428
+    },
+    {
+      "epoch": 1.173734610123119,
+      "grad_norm": 0.2699303925037384,
+      "learning_rate": 3.425345388349786e-05,
+      "loss": 1.5429,
+      "step": 429
+    },
+    {
+      "epoch": 1.1764705882352942,
+      "grad_norm": 0.2614208459854126,
+      "learning_rate": 3.418541649862494e-05,
+      "loss": 1.5916,
+      "step": 430
+    },
+    {
+      "epoch": 1.1792065663474691,
+      "grad_norm": 0.30162835121154785,
+      "learning_rate": 3.4117300373129376e-05,
+      "loss": 1.4801,
+      "step": 431
+    },
+    {
+      "epoch": 1.1819425444596443,
+      "grad_norm": 0.2973942458629608,
+      "learning_rate": 3.404910609092655e-05,
+      "loss": 1.5476,
+      "step": 432
+    },
+    {
+      "epoch": 1.1846785225718195,
+      "grad_norm": 0.26698416471481323,
+      "learning_rate": 3.3980834236601853e-05,
+      "loss": 1.6382,
+      "step": 433
+    },
+    {
+      "epoch": 1.1874145006839945,
+      "grad_norm": 0.3085285723209381,
+      "learning_rate": 3.391248539540565e-05,
+      "loss": 1.5059,
+      "step": 434
+    },
+    {
+      "epoch": 1.1901504787961696,
+      "grad_norm": 0.25455278158187866,
+      "learning_rate": 3.384406015324826e-05,
+      "loss": 1.6612,
+      "step": 435
+    },
+    {
+      "epoch": 1.1928864569083448,
+      "grad_norm": 0.2725888192653656,
+      "learning_rate": 3.3775559096694933e-05,
+      "loss": 1.5598,
+      "step": 436
+    },
+    {
+      "epoch": 1.1956224350205198,
+      "grad_norm": 0.34251880645751953,
+      "learning_rate": 3.370698281296083e-05,
+      "loss": 1.5973,
+      "step": 437
+    },
+    {
+      "epoch": 1.198358413132695,
+      "grad_norm": 0.2710118889808655,
+      "learning_rate": 3.363833188990599e-05,
+      "loss": 1.4116,
+      "step": 438
+    },
+    {
+      "epoch": 1.2010943912448702,
+      "grad_norm": 0.2622835636138916,
+      "learning_rate": 3.3569606916030294e-05,
+      "loss": 1.4546,
+      "step": 439
+    },
+    {
+      "epoch": 1.2038303693570451,
+      "grad_norm": 0.2685026228427887,
+      "learning_rate": 3.350080848046839e-05,
+      "loss": 1.5789,
+      "step": 440
+    },
+    {
+      "epoch": 1.2065663474692203,
+      "grad_norm": 0.2792280912399292,
+      "learning_rate": 3.343193717298469e-05,
+      "loss": 1.5287,
+      "step": 441
+    },
+    {
+      "epoch": 1.2093023255813953,
+      "grad_norm": 0.3190619647502899,
+      "learning_rate": 3.3362993583968264e-05,
+      "loss": 1.5803,
+      "step": 442
+    },
+    {
+      "epoch": 1.2120383036935705,
+      "grad_norm": 0.27128350734710693,
+      "learning_rate": 3.329397830442784e-05,
+      "loss": 1.5371,
+      "step": 443
+    },
+    {
+      "epoch": 1.2147742818057456,
+      "grad_norm": 0.2853277325630188,
+      "learning_rate": 3.322489192598665e-05,
+      "loss": 1.5739,
+      "step": 444
+    },
+    {
+      "epoch": 1.2175102599179206,
+      "grad_norm": 0.2518289387226105,
+      "learning_rate": 3.3155735040877465e-05,
+      "loss": 1.6408,
+      "step": 445
+    },
+    {
+      "epoch": 1.2202462380300958,
+      "grad_norm": 0.297750324010849,
+      "learning_rate": 3.308650824193744e-05,
+      "loss": 1.6255,
+      "step": 446
+    },
+    {
+      "epoch": 1.2229822161422708,
+      "grad_norm": 0.2988694906234741,
+      "learning_rate": 3.301721212260306e-05,
+      "loss": 1.5215,
+      "step": 447
+    },
+    {
+      "epoch": 1.225718194254446,
+      "grad_norm": 0.2906099557876587,
+      "learning_rate": 3.294784727690503e-05,
+      "loss": 1.5124,
+      "step": 448
+    },
+    {
+      "epoch": 1.2284541723666211,
+      "grad_norm": 0.28081414103507996,
+      "learning_rate": 3.2878414299463225e-05,
+      "loss": 1.6691,
+      "step": 449
+    },
+    {
+      "epoch": 1.231190150478796,
+      "grad_norm": 0.26118987798690796,
+      "learning_rate": 3.280891378548156e-05,
+      "loss": 1.524,
+      "step": 450
+    },
+    {
+      "epoch": 1.2339261285909713,
+      "grad_norm": 0.2828025817871094,
+      "learning_rate": 3.273934633074291e-05,
+      "loss": 1.7315,
+      "step": 451
+    },
+    {
+      "epoch": 1.2366621067031465,
+      "grad_norm": 0.27713146805763245,
+      "learning_rate": 3.2669712531603966e-05,
+      "loss": 1.534,
+      "step": 452
+    },
+    {
+      "epoch": 1.2393980848153214,
+      "grad_norm": 0.30562645196914673,
+      "learning_rate": 3.2600012984990165e-05,
+      "loss": 1.4404,
+      "step": 453
+    },
+    {
+      "epoch": 1.2421340629274966,
+      "grad_norm": 0.30407342314720154,
+      "learning_rate": 3.2530248288390555e-05,
+      "loss": 1.5327,
+      "step": 454
+    },
+    {
+      "epoch": 1.2448700410396718,
+      "grad_norm": 0.3026122450828552,
+      "learning_rate": 3.246041903985264e-05,
+      "loss": 1.4441,
+      "step": 455
+    },
+    {
+      "epoch": 1.2476060191518468,
+      "grad_norm": 0.3250129520893097,
+      "learning_rate": 3.2390525837977334e-05,
+      "loss": 1.661,
+      "step": 456
+    },
+    {
+      "epoch": 1.250341997264022,
+      "grad_norm": 0.29971036314964294,
+      "learning_rate": 3.232056928191376e-05,
+      "loss": 1.5989,
+      "step": 457
+    },
+    {
+      "epoch": 1.253077975376197,
+      "grad_norm": 0.296903133392334,
+      "learning_rate": 3.225054997135413e-05,
+      "loss": 1.5724,
+      "step": 458
+    },
+    {
+      "epoch": 1.255813953488372,
+      "grad_norm": 0.31808045506477356,
+      "learning_rate": 3.218046850652862e-05,
+      "loss": 1.6791,
+      "step": 459
+    },
+    {
+      "epoch": 1.2585499316005473,
+      "grad_norm": 0.2797161340713501,
+      "learning_rate": 3.211032548820019e-05,
+      "loss": 1.4345,
+      "step": 460
+    },
+    {
+      "epoch": 1.2612859097127223,
+      "grad_norm": 0.3660871386528015,
+      "learning_rate": 3.20401215176595e-05,
+      "loss": 1.5669,
+      "step": 461
+    },
+    {
+      "epoch": 1.2640218878248974,
+      "grad_norm": 0.30877435207366943,
+      "learning_rate": 3.196985719671968e-05,
+      "loss": 1.5406,
+      "step": 462
+    },
+    {
+      "epoch": 1.2667578659370724,
+      "grad_norm": 0.29524242877960205,
+      "learning_rate": 3.1899533127711215e-05,
+      "loss": 1.3333,
+      "step": 463
+    },
+    {
+      "epoch": 1.2694938440492476,
+      "grad_norm": 0.3187881112098694,
+      "learning_rate": 3.182914991347677e-05,
+      "loss": 1.5237,
+      "step": 464
+    },
+    {
+      "epoch": 1.2722298221614228,
+      "grad_norm": 0.2810283303260803,
+      "learning_rate": 3.1758708157366036e-05,
+      "loss": 1.4334,
+      "step": 465
+    },
+    {
+      "epoch": 1.2749658002735977,
+      "grad_norm": 0.27640587091445923,
+      "learning_rate": 3.168820846323053e-05,
+      "loss": 1.5078,
+      "step": 466
+    },
+    {
+      "epoch": 1.277701778385773,
+      "grad_norm": 0.2678714692592621,
+      "learning_rate": 3.161765143541843e-05,
+      "loss": 1.439,
+      "step": 467
+    },
+    {
+      "epoch": 1.2804377564979479,
+      "grad_norm": 0.315186470746994,
+      "learning_rate": 3.154703767876942e-05,
+      "loss": 1.4901,
+      "step": 468
+    },
+    {
+      "epoch": 1.283173734610123,
+      "grad_norm": 0.32562458515167236,
+      "learning_rate": 3.1476367798609475e-05,
+      "loss": 1.4822,
+      "step": 469
+    },
+    {
+      "epoch": 1.2859097127222983,
+      "grad_norm": 0.31798624992370605,
+      "learning_rate": 3.1405642400745664e-05,
+      "loss": 1.5145,
+      "step": 470
+    },
+    {
+      "epoch": 1.2886456908344734,
+      "grad_norm": 0.29922667145729065,
+      "learning_rate": 3.133486209146099e-05,
+      "loss": 1.4984,
+      "step": 471
+    },
+    {
+      "epoch": 1.2913816689466484,
+      "grad_norm": 0.3259302079677582,
+      "learning_rate": 3.12640274775092e-05,
+      "loss": 1.5048,
+      "step": 472
+    },
+    {
+      "epoch": 1.2941176470588236,
+      "grad_norm": 0.2870180010795593,
+      "learning_rate": 3.119313916610948e-05,
+      "loss": 1.52,
+      "step": 473
+    },
+    {
+      "epoch": 1.2968536251709986,
+      "grad_norm": 0.3135444223880768,
+      "learning_rate": 3.112219776494142e-05,
+      "loss": 1.494,
+      "step": 474
+    },
+    {
+      "epoch": 1.2995896032831737,
+      "grad_norm": 0.29159173369407654,
+      "learning_rate": 3.105120388213966e-05,
+      "loss": 1.7019,
+      "step": 475
+    },
+    {
+      "epoch": 1.302325581395349,
+      "grad_norm": 0.31548362970352173,
+      "learning_rate": 3.098015812628875e-05,
+      "loss": 1.6465,
+      "step": 476
+    },
+    {
+      "epoch": 1.305061559507524,
+      "grad_norm": 0.2820277214050293,
+      "learning_rate": 3.090906110641791e-05,
+      "loss": 1.5705,
+      "step": 477
+    },
+    {
+      "epoch": 1.307797537619699,
+      "grad_norm": 0.2773056924343109,
+      "learning_rate": 3.083791343199582e-05,
+      "loss": 1.785,
+      "step": 478
+    },
+    {
+      "epoch": 1.310533515731874,
+      "grad_norm": 0.2650894522666931,
+      "learning_rate": 3.0766715712925384e-05,
+      "loss": 1.6026,
+      "step": 479
+    },
+    {
+      "epoch": 1.3132694938440492,
+      "grad_norm": 0.2996635138988495,
+      "learning_rate": 3.06954685595385e-05,
+      "loss": 1.489,
+      "step": 480
+    },
+    {
+      "epoch": 1.3160054719562244,
+      "grad_norm": 0.3386262059211731,
+      "learning_rate": 3.062417258259084e-05,
+      "loss": 1.538,
+      "step": 481
+    },
+    {
+      "epoch": 1.3187414500683994,
+      "grad_norm": 0.30783751606941223,
+      "learning_rate": 3.055282839325661e-05,
+      "loss": 1.42,
+      "step": 482
+    },
+    {
+      "epoch": 1.3214774281805746,
+      "grad_norm": 0.28174683451652527,
+      "learning_rate": 3.0481436603123292e-05,
+      "loss": 1.5068,
+      "step": 483
+    },
+    {
+      "epoch": 1.3242134062927495,
+      "grad_norm": 0.2959563732147217,
+      "learning_rate": 3.0409997824186453e-05,
+      "loss": 1.4343,
+      "step": 484
+    },
+    {
+      "epoch": 1.3269493844049247,
+      "grad_norm": 0.24625587463378906,
+      "learning_rate": 3.0338512668844443e-05,
+      "loss": 1.5942,
+      "step": 485
+    },
+    {
+      "epoch": 1.3296853625171,
+      "grad_norm": 0.3001210689544678,
+      "learning_rate": 3.0266981749893157e-05,
+      "loss": 1.5715,
+      "step": 486
+    },
+    {
+      "epoch": 1.332421340629275,
+      "grad_norm": 0.2976323962211609,
+      "learning_rate": 3.0195405680520828e-05,
+      "loss": 1.4471,
+      "step": 487
+    },
+    {
+      "epoch": 1.33515731874145,
+      "grad_norm": 0.3303026854991913,
+      "learning_rate": 3.012378507430269e-05,
+      "loss": 1.6357,
+      "step": 488
+    },
+    {
+      "epoch": 1.3378932968536252,
+      "grad_norm": 0.308095246553421,
+      "learning_rate": 3.005212054519579e-05,
+      "loss": 1.6527,
+      "step": 489
+    },
+    {
+      "epoch": 1.3406292749658002,
+      "grad_norm": 0.2990988790988922,
+      "learning_rate": 2.99804127075337e-05,
+      "loss": 1.5795,
+      "step": 490
+    },
+    {
+      "epoch": 1.3433652530779754,
+      "grad_norm": 0.29910552501678467,
+      "learning_rate": 2.9908662176021225e-05,
+      "loss": 1.6597,
+      "step": 491
+    },
+    {
+      "epoch": 1.3461012311901506,
+      "grad_norm": 0.28705233335494995,
+      "learning_rate": 2.9836869565729176e-05,
+      "loss": 1.4978,
+      "step": 492
+    },
+    {
+      "epoch": 1.3488372093023255,
+      "grad_norm": 0.33388686180114746,
+      "learning_rate": 2.9765035492089072e-05,
+      "loss": 1.4049,
+      "step": 493
+    },
+    {
+      "epoch": 1.3515731874145007,
+      "grad_norm": 0.28467613458633423,
+      "learning_rate": 2.9693160570887873e-05,
+      "loss": 1.4809,
+      "step": 494
+    },
+    {
+      "epoch": 1.3543091655266757,
+      "grad_norm": 0.2524206340312958,
+      "learning_rate": 2.9621245418262694e-05,
+      "loss": 1.5561,
+      "step": 495
+    },
+    {
+      "epoch": 1.3570451436388509,
+      "grad_norm": 0.27135151624679565,
+      "learning_rate": 2.954929065069554e-05,
+      "loss": 1.5686,
+      "step": 496
+    },
+    {
+      "epoch": 1.359781121751026,
+      "grad_norm": 0.31753405928611755,
+      "learning_rate": 2.9477296885007988e-05,
+      "loss": 1.4994,
+      "step": 497
+    },
+    {
+      "epoch": 1.362517099863201,
+      "grad_norm": 0.274442583322525,
+      "learning_rate": 2.9405264738355946e-05,
+      "loss": 1.5835,
+      "step": 498
+    },
+    {
+      "epoch": 1.3652530779753762,
+      "grad_norm": 0.3164355754852295,
+      "learning_rate": 2.9333194828224316e-05,
+      "loss": 1.4077,
+      "step": 499
+    },
+    {
+      "epoch": 1.3679890560875512,
+      "grad_norm": 0.3447605073451996,
+      "learning_rate": 2.926108777242172e-05,
+      "loss": 1.4768,
+      "step": 500
+    },
+    {
+      "epoch": 1.3679890560875512,
+      "eval_loss": 1.573856234550476,
+      "eval_runtime": 13.7722,
+      "eval_samples_per_second": 11.182,
+      "eval_steps_per_second": 11.182,
+      "step": 500
+    },
+    {
+      "epoch": 1.3707250341997264,
+      "grad_norm": 0.31595364212989807,
+      "learning_rate": 2.9188944189075235e-05,
+      "loss": 1.6381,
+      "step": 501
+    },
+    {
+      "epoch": 1.3734610123119015,
+      "grad_norm": 0.28893882036209106,
+      "learning_rate": 2.9116764696625033e-05,
+      "loss": 1.6677,
+      "step": 502
+    },
+    {
+      "epoch": 1.3761969904240767,
+      "grad_norm": 0.27313435077667236,
+      "learning_rate": 2.9044549913819124e-05,
+      "loss": 1.5746,
+      "step": 503
+    },
+    {
+      "epoch": 1.3789329685362517,
+      "grad_norm": 0.2701267600059509,
+      "learning_rate": 2.897230045970804e-05,
+      "loss": 1.2618,
+      "step": 504
+    },
+    {
+      "epoch": 1.3816689466484269,
+      "grad_norm": 0.30766165256500244,
+      "learning_rate": 2.890001695363953e-05,
+      "loss": 1.4354,
+      "step": 505
+    },
+    {
+      "epoch": 1.3844049247606018,
+      "grad_norm": 0.34688156843185425,
+      "learning_rate": 2.8827700015253246e-05,
+      "loss": 1.4171,
+      "step": 506
+    },
+    {
+      "epoch": 1.387140902872777,
+      "grad_norm": 0.28550222516059875,
+      "learning_rate": 2.875535026447543e-05,
+      "loss": 1.6484,
+      "step": 507
+    },
+    {
+      "epoch": 1.3898768809849522,
+      "grad_norm": 0.2944300174713135,
+      "learning_rate": 2.868296832151361e-05,
+      "loss": 1.6701,
+      "step": 508
+    },
+    {
+      "epoch": 1.3926128590971272,
+      "grad_norm": 0.2670783996582031,
+      "learning_rate": 2.8610554806851264e-05,
+      "loss": 1.5159,
+      "step": 509
+    },
+    {
+      "epoch": 1.3953488372093024,
+      "grad_norm": 0.2730109989643097,
+      "learning_rate": 2.853811034124253e-05,
+      "loss": 1.463,
+      "step": 510
+    },
+    {
+      "epoch": 1.3980848153214773,
+      "grad_norm": 0.2924908995628357,
+      "learning_rate": 2.8465635545706858e-05,
+      "loss": 1.6673,
+      "step": 511
+    },
+    {
+      "epoch": 1.4008207934336525,
+      "grad_norm": 0.2832358777523041,
+      "learning_rate": 2.8393131041523702e-05,
+      "loss": 1.4224,
+      "step": 512
+    },
+    {
+      "epoch": 1.4035567715458277,
+      "grad_norm": 0.3116619884967804,
+      "learning_rate": 2.8320597450227186e-05,
+      "loss": 1.4651,
+      "step": 513
+    },
+    {
+      "epoch": 1.4062927496580027,
+      "grad_norm": 0.2656640410423279,
+      "learning_rate": 2.824803539360078e-05,
+      "loss": 1.5117,
+      "step": 514
+    },
+    {
+      "epoch": 1.4090287277701778,
+      "grad_norm": 0.321814626455307,
+      "learning_rate": 2.8175445493671972e-05,
+      "loss": 1.5859,
+      "step": 515
+    },
+    {
+      "epoch": 1.4117647058823528,
+      "grad_norm": 0.3401627540588379,
+      "learning_rate": 2.8102828372706926e-05,
+      "loss": 1.6612,
+      "step": 516
+    },
+    {
+      "epoch": 1.414500683994528,
+      "grad_norm": 0.3001568019390106,
+      "learning_rate": 2.803018465320515e-05,
+      "loss": 1.655,
+      "step": 517
+    },
+    {
+      "epoch": 1.4172366621067032,
+      "grad_norm": 0.27258241176605225,
+      "learning_rate": 2.795751495789418e-05,
+      "loss": 1.6612,
+      "step": 518
+    },
+    {
+      "epoch": 1.4199726402188784,
+      "grad_norm": 0.2985587418079376,
+      "learning_rate": 2.7884819909724224e-05,
+      "loss": 1.452,
+      "step": 519
+    },
+    {
+      "epoch": 1.4227086183310533,
+      "grad_norm": 0.2565891146659851,
+      "learning_rate": 2.78121001318628e-05,
+      "loss": 1.6304,
+      "step": 520
+    },
+    {
+      "epoch": 1.4254445964432285,
+      "grad_norm": 0.29893383383750916,
+      "learning_rate": 2.7739356247689446e-05,
+      "loss": 1.4655,
+      "step": 521
+    },
+    {
+      "epoch": 1.4281805745554035,
+      "grad_norm": 0.28050941228866577,
+      "learning_rate": 2.7666588880790335e-05,
+      "loss": 1.5134,
+      "step": 522
+    },
+    {
+      "epoch": 1.4309165526675787,
+      "grad_norm": 0.29267653822898865,
+      "learning_rate": 2.759379865495294e-05,
+      "loss": 1.4355,
+      "step": 523
+    },
+    {
+      "epoch": 1.4336525307797539,
+      "grad_norm": 0.279461145401001,
+      "learning_rate": 2.752098619416069e-05,
+      "loss": 1.5606,
+      "step": 524
+    },
+    {
+      "epoch": 1.4363885088919288,
+      "grad_norm": 0.260345995426178,
+      "learning_rate": 2.7448152122587634e-05,
+      "loss": 1.702,
+      "step": 525
+    },
+    {
+      "epoch": 1.439124487004104,
+      "grad_norm": 0.2932511270046234,
+      "learning_rate": 2.7375297064593063e-05,
+      "loss": 1.3368,
+      "step": 526
+    },
+    {
+      "epoch": 1.441860465116279,
+      "grad_norm": 0.3288111686706543,
+      "learning_rate": 2.730242164471616e-05,
+      "loss": 1.4919,
+      "step": 527
+    },
+    {
+      "epoch": 1.4445964432284542,
+      "grad_norm": 0.27260783314704895,
+      "learning_rate": 2.7229526487670676e-05,
+      "loss": 1.5306,
+      "step": 528
+    },
+    {
+      "epoch": 1.4473324213406293,
+      "grad_norm": 0.2748279273509979,
+      "learning_rate": 2.7156612218339544e-05,
+      "loss": 1.5238,
+      "step": 529
+    },
+    {
+      "epoch": 1.4500683994528043,
+      "grad_norm": 0.26095664501190186,
+      "learning_rate": 2.708367946176956e-05,
+      "loss": 1.5456,
+      "step": 530
+    },
+    {
+      "epoch": 1.4528043775649795,
+      "grad_norm": 0.3169284164905548,
+      "learning_rate": 2.701072884316595e-05,
+      "loss": 1.6298,
+      "step": 531
+    },
+    {
+      "epoch": 1.4555403556771545,
+      "grad_norm": 0.2977512776851654,
+      "learning_rate": 2.6937760987887112e-05,
+      "loss": 1.4394,
+      "step": 532
+    },
+    {
+      "epoch": 1.4582763337893296,
+      "grad_norm": 0.2777084708213806,
+      "learning_rate": 2.6864776521439166e-05,
+      "loss": 1.5062,
+      "step": 533
+    },
+    {
+      "epoch": 1.4610123119015048,
+      "grad_norm": 0.29830804467201233,
+      "learning_rate": 2.6791776069470658e-05,
+      "loss": 1.5056,
+      "step": 534
+    },
+    {
+      "epoch": 1.46374829001368,
+      "grad_norm": 0.35224923491477966,
+      "learning_rate": 2.6718760257767135e-05,
+      "loss": 1.4127,
+      "step": 535
+    },
+    {
+      "epoch": 1.466484268125855,
+      "grad_norm": 0.31385019421577454,
+      "learning_rate": 2.6645729712245832e-05,
+      "loss": 1.6469,
+      "step": 536
+    },
+    {
+      "epoch": 1.4692202462380302,
+      "grad_norm": 0.3666173815727234,
+      "learning_rate": 2.6572685058950298e-05,
+      "loss": 1.6698,
+      "step": 537
+    },
+    {
+      "epoch": 1.4719562243502051,
+      "grad_norm": 0.29228919744491577,
+      "learning_rate": 2.649962692404499e-05,
+      "loss": 1.5971,
+      "step": 538
+    },
+    {
+      "epoch": 1.4746922024623803,
+      "grad_norm": 0.2931617796421051,
+      "learning_rate": 2.6426555933809954e-05,
+      "loss": 1.4914,
+      "step": 539
+    },
+    {
+      "epoch": 1.4774281805745555,
+      "grad_norm": 0.2743561863899231,
+      "learning_rate": 2.635347271463544e-05,
+      "loss": 1.5475,
+      "step": 540
+    },
+    {
+      "epoch": 1.4801641586867305,
+      "grad_norm": 0.3290240466594696,
+      "learning_rate": 2.6280377893016516e-05,
+      "loss": 1.4155,
+      "step": 541
+    },
+    {
+      "epoch": 1.4829001367989056,
+      "grad_norm": 0.30161550641059875,
+      "learning_rate": 2.6207272095547718e-05,
+      "loss": 1.4564,
+      "step": 542
+    },
+    {
+      "epoch": 1.4856361149110806,
+      "grad_norm": 0.3175734877586365,
+      "learning_rate": 2.613415594891767e-05,
+      "loss": 1.5312,
+      "step": 543
+    },
+    {
+      "epoch": 1.4883720930232558,
+      "grad_norm": 0.3023424744606018,
+      "learning_rate": 2.606103007990371e-05,
+      "loss": 1.4688,
+      "step": 544
+    },
+    {
+      "epoch": 1.491108071135431,
+      "grad_norm": 0.3181714415550232,
+      "learning_rate": 2.5987895115366516e-05,
+      "loss": 1.738,
+      "step": 545
+    },
+    {
+      "epoch": 1.493844049247606,
+      "grad_norm": 0.2678375244140625,
+      "learning_rate": 2.5914751682244748e-05,
+      "loss": 1.5491,
+      "step": 546
+    },
+    {
+      "epoch": 1.4965800273597811,
+      "grad_norm": 0.3041532039642334,
+      "learning_rate": 2.5841600407549642e-05,
+      "loss": 1.6654,
+      "step": 547
+    },
+    {
+      "epoch": 1.499316005471956,
+      "grad_norm": 0.30127188563346863,
+      "learning_rate": 2.5768441918359692e-05,
+      "loss": 1.5671,
+      "step": 548
+    },
+    {
+      "epoch": 1.5020519835841313,
+      "grad_norm": 0.34145042300224304,
+      "learning_rate": 2.5695276841815186e-05,
+      "loss": 1.6131,
+      "step": 549
+    },
+    {
+      "epoch": 1.5047879616963065,
+      "grad_norm": 0.2628389000892639,
+      "learning_rate": 2.562210580511291e-05,
+      "loss": 1.7636,
+      "step": 550
+    },
+    {
+      "epoch": 1.5075239398084817,
+      "grad_norm": 0.2784722149372101,
+      "learning_rate": 2.5548929435500758e-05,
+      "loss": 1.4967,
+      "step": 551
+    },
+    {
+      "epoch": 1.5102599179206566,
+      "grad_norm": 0.29418709874153137,
+      "learning_rate": 2.547574836027231e-05,
+      "loss": 1.5771,
+      "step": 552
+    },
+    {
+      "epoch": 1.5129958960328316,
+      "grad_norm": 0.29850900173187256,
+      "learning_rate": 2.54025632067615e-05,
+      "loss": 1.4134,
+      "step": 553
+    },
+    {
+      "epoch": 1.5157318741450068,
+      "grad_norm": 0.29349079728126526,
+      "learning_rate": 2.5329374602337215e-05,
+      "loss": 1.4498,
+      "step": 554
+    },
+    {
+      "epoch": 1.518467852257182,
+      "grad_norm": 0.27603679895401,
+      "learning_rate": 2.525618317439793e-05,
+      "loss": 1.5103,
+      "step": 555
+    },
+    {
+      "epoch": 1.5212038303693571,
+      "grad_norm": 0.29642152786254883,
+      "learning_rate": 2.518298955036632e-05,
+      "loss": 1.5597,
+      "step": 556
+    },
+    {
+      "epoch": 1.523939808481532,
+      "grad_norm": 0.3032752573490143,
+      "learning_rate": 2.5109794357683885e-05,
+      "loss": 1.5697,
+      "step": 557
+    },
+    {
+      "epoch": 1.5266757865937073,
+      "grad_norm": 0.3295765519142151,
+      "learning_rate": 2.503659822380558e-05,
+      "loss": 1.6343,
+      "step": 558
+    },
+    {
+      "epoch": 1.5294117647058822,
+      "grad_norm": 0.28324607014656067,
+      "learning_rate": 2.496340177619442e-05,
+      "loss": 1.551,
+      "step": 559
+    },
+    {
+      "epoch": 1.5321477428180574,
+      "grad_norm": 0.2895617187023163,
+      "learning_rate": 2.489020564231612e-05,
+      "loss": 1.6563,
+      "step": 560
+    },
+    {
+      "epoch": 1.5348837209302326,
+      "grad_norm": 0.3268161714076996,
+      "learning_rate": 2.4817010449633688e-05,
+      "loss": 1.6045,
+      "step": 561
+    },
+    {
+      "epoch": 1.5376196990424078,
+      "grad_norm": 0.27146685123443604,
+      "learning_rate": 2.474381682560208e-05,
+      "loss": 1.5727,
+      "step": 562
+    },
+    {
+      "epoch": 1.5403556771545828,
+      "grad_norm": 0.27629193663597107,
+      "learning_rate": 2.4670625397662787e-05,
+      "loss": 1.4352,
+      "step": 563
+    },
+    {
+      "epoch": 1.5430916552667577,
+      "grad_norm": 0.27017953991889954,
+      "learning_rate": 2.4597436793238506e-05,
+      "loss": 1.646,
+      "step": 564
+    },
+    {
+      "epoch": 1.545827633378933,
+      "grad_norm": 0.289787232875824,
+      "learning_rate": 2.45242516397277e-05,
+      "loss": 1.7107,
+      "step": 565
+    },
+    {
+      "epoch": 1.548563611491108,
+      "grad_norm": 0.2762577533721924,
+      "learning_rate": 2.4451070564499245e-05,
+      "loss": 1.5694,
+      "step": 566
+    },
+    {
+      "epoch": 1.5512995896032833,
+      "grad_norm": 0.2959858179092407,
+      "learning_rate": 2.4377894194887095e-05,
+      "loss": 1.639,
+      "step": 567
+    },
+    {
+      "epoch": 1.5540355677154583,
+      "grad_norm": 0.28595441579818726,
+      "learning_rate": 2.4304723158184827e-05,
+      "loss": 1.3091,
+      "step": 568
+    },
+    {
+      "epoch": 1.5567715458276332,
+      "grad_norm": 0.3006284236907959,
+      "learning_rate": 2.4231558081640314e-05,
+      "loss": 1.569,
+      "step": 569
+    },
+    {
+      "epoch": 1.5595075239398084,
+      "grad_norm": 0.2800249457359314,
+      "learning_rate": 2.415839959245036e-05,
+      "loss": 1.5124,
+      "step": 570
+    },
+    {
+      "epoch": 1.5622435020519836,
+      "grad_norm": 0.3035907745361328,
+      "learning_rate": 2.4085248317755254e-05,
+      "loss": 1.5785,
+      "step": 571
+    },
+    {
+      "epoch": 1.5649794801641588,
+      "grad_norm": 0.3111293315887451,
+      "learning_rate": 2.4012104884633486e-05,
+      "loss": 1.5544,
+      "step": 572
+    },
+    {
+      "epoch": 1.5677154582763337,
+      "grad_norm": 0.31877052783966064,
+      "learning_rate": 2.39389699200963e-05,
+      "loss": 1.482,
+      "step": 573
+    },
+    {
+      "epoch": 1.570451436388509,
+      "grad_norm": 0.3056713342666626,
+      "learning_rate": 2.386584405108233e-05,
+      "loss": 1.5234,
+      "step": 574
+    },
+    {
+      "epoch": 1.573187414500684,
+      "grad_norm": 0.289650022983551,
+      "learning_rate": 2.3792727904452285e-05,
+      "loss": 1.398,
+      "step": 575
+    },
+    {
+      "epoch": 1.575923392612859,
+      "grad_norm": 0.28617268800735474,
+      "learning_rate": 2.3719622106983486e-05,
+      "loss": 1.5721,
+      "step": 576
+    },
+    {
+      "epoch": 1.5786593707250343,
+      "grad_norm": 0.29906001687049866,
+      "learning_rate": 2.3646527285364565e-05,
+      "loss": 1.3543,
+      "step": 577
+    },
+    {
+      "epoch": 1.5813953488372094,
+      "grad_norm": 0.2876405417919159,
+      "learning_rate": 2.3573444066190052e-05,
+      "loss": 1.5365,
+      "step": 578
+    },
+    {
+      "epoch": 1.5841313269493844,
+      "grad_norm": 0.3173394799232483,
+      "learning_rate": 2.3500373075955022e-05,
+      "loss": 1.4437,
+      "step": 579
+    },
+    {
+      "epoch": 1.5868673050615594,
+      "grad_norm": 0.2820620834827423,
+      "learning_rate": 2.342731494104971e-05,
+      "loss": 1.5992,
+      "step": 580
+    },
+    {
+      "epoch": 1.5896032831737346,
+      "grad_norm": 0.3694467842578888,
+      "learning_rate": 2.3354270287754174e-05,
+      "loss": 1.5443,
+      "step": 581
+    },
+    {
+      "epoch": 1.5923392612859097,
+      "grad_norm": 0.2960888147354126,
+      "learning_rate": 2.328123974223288e-05,
+      "loss": 1.5173,
+      "step": 582
+    },
+    {
+      "epoch": 1.595075239398085,
+      "grad_norm": 0.3080427348613739,
+      "learning_rate": 2.3208223930529347e-05,
+      "loss": 1.749,
+      "step": 583
+    },
+    {
+      "epoch": 1.59781121751026,
+      "grad_norm": 0.29521963000297546,
+      "learning_rate": 2.313522347856084e-05,
+      "loss": 1.522,
+      "step": 584
+    },
+    {
+      "epoch": 1.6005471956224349,
+      "grad_norm": 0.32432520389556885,
+      "learning_rate": 2.306223901211289e-05,
+      "loss": 1.6394,
+      "step": 585
+    },
+    {
+      "epoch": 1.60328317373461,
+      "grad_norm": 0.326926052570343,
+      "learning_rate": 2.2989271156834057e-05,
+      "loss": 1.5092,
+      "step": 586
+    },
+    {
+      "epoch": 1.6060191518467852,
+      "grad_norm": 0.2793523967266083,
+      "learning_rate": 2.291632053823045e-05,
+      "loss": 1.5266,
+      "step": 587
+    },
+    {
+      "epoch": 1.6087551299589604,
+      "grad_norm": 0.3044006824493408,
+      "learning_rate": 2.2843387781660452e-05,
+      "loss": 1.7137,
+      "step": 588
+    },
+    {
+      "epoch": 1.6114911080711354,
+      "grad_norm": 0.33133581280708313,
+      "learning_rate": 2.2770473512329333e-05,
+      "loss": 1.5494,
+      "step": 589
+    },
+    {
+      "epoch": 1.6142270861833106,
+      "grad_norm": 0.30066078901290894,
+      "learning_rate": 2.269757835528385e-05,
+      "loss": 1.387,
+      "step": 590
+    },
+    {
+      "epoch": 1.6169630642954855,
+      "grad_norm": 0.29674389958381653,
+      "learning_rate": 2.2624702935406943e-05,
+      "loss": 1.6985,
+      "step": 591
+    },
+    {
+      "epoch": 1.6196990424076607,
+      "grad_norm": 0.2950797975063324,
+      "learning_rate": 2.255184787741237e-05,
+      "loss": 1.558,
+      "step": 592
+    },
+    {
+      "epoch": 1.622435020519836,
+      "grad_norm": 0.32414302229881287,
+      "learning_rate": 2.2479013805839318e-05,
+      "loss": 1.4547,
+      "step": 593
+    },
+    {
+      "epoch": 1.625170998632011,
+      "grad_norm": 0.2893849015235901,
+      "learning_rate": 2.240620134504707e-05,
+      "loss": 1.6205,
+      "step": 594
+    },
+    {
+      "epoch": 1.627906976744186,
+      "grad_norm": 0.30795803666114807,
+      "learning_rate": 2.233341111920967e-05,
+      "loss": 1.4725,
+      "step": 595
+    },
+    {
+      "epoch": 1.630642954856361,
+      "grad_norm": 0.28829425573349,
+      "learning_rate": 2.226064375231056e-05,
+      "loss": 1.5676,
+      "step": 596
+    },
+    {
+      "epoch": 1.6333789329685362,
+      "grad_norm": 0.32631853222846985,
+      "learning_rate": 2.2187899868137206e-05,
+      "loss": 1.447,
+      "step": 597
+    },
+    {
+      "epoch": 1.6361149110807114,
+      "grad_norm": 0.2999882996082306,
+      "learning_rate": 2.211518009027579e-05,
+      "loss": 1.4184,
+      "step": 598
+    },
+    {
+      "epoch": 1.6388508891928866,
+      "grad_norm": 0.302869975566864,
+      "learning_rate": 2.204248504210582e-05,
+      "loss": 1.5356,
+      "step": 599
+    },
+    {
+      "epoch": 1.6415868673050615,
+      "grad_norm": 0.32066816091537476,
+      "learning_rate": 2.1969815346794857e-05,
+      "loss": 1.507,
+      "step": 600
+    },
+    {
+      "epoch": 1.6443228454172365,
+      "grad_norm": 0.26394781470298767,
+      "learning_rate": 2.189717162729309e-05,
+      "loss": 1.519,
+      "step": 601
+    },
+    {
+      "epoch": 1.6470588235294117,
+      "grad_norm": 0.28778427839279175,
+      "learning_rate": 2.182455450632803e-05,
+      "loss": 1.6243,
+      "step": 602
+    },
+    {
+      "epoch": 1.6497948016415869,
+      "grad_norm": 0.28757721185684204,
+      "learning_rate": 2.1751964606399224e-05,
+      "loss": 1.6119,
+      "step": 603
+    },
+    {
+      "epoch": 1.652530779753762,
+      "grad_norm": 0.32660529017448425,
+      "learning_rate": 2.167940254977282e-05,
+      "loss": 1.481,
+      "step": 604
+    },
+    {
+      "epoch": 1.655266757865937,
+      "grad_norm": 0.2857491075992584,
+      "learning_rate": 2.1606868958476304e-05,
+      "loss": 1.4287,
+      "step": 605
+    },
+    {
+      "epoch": 1.6580027359781122,
+      "grad_norm": 0.3346758782863617,
+      "learning_rate": 2.1534364454293148e-05,
+      "loss": 1.6915,
+      "step": 606
+    },
+    {
+      "epoch": 1.6607387140902872,
+      "grad_norm": 0.2858891189098358,
+      "learning_rate": 2.146188965875747e-05,
+      "loss": 1.4857,
+      "step": 607
+    },
+    {
+      "epoch": 1.6634746922024624,
+      "grad_norm": 0.2926919758319855,
+      "learning_rate": 2.1389445193148742e-05,
+      "loss": 1.5348,
+      "step": 608
+    },
+    {
+      "epoch": 1.6662106703146375,
+      "grad_norm": 0.3735974133014679,
+      "learning_rate": 2.1317031678486402e-05,
+      "loss": 1.697,
+      "step": 609
+    },
+    {
+      "epoch": 1.6689466484268127,
+      "grad_norm": 0.3130210340023041,
+      "learning_rate": 2.124464973552457e-05,
+      "loss": 1.6426,
+      "step": 610
+    },
+    {
+      "epoch": 1.6716826265389877,
+      "grad_norm": 0.36122313141822815,
+      "learning_rate": 2.117229998474676e-05,
+      "loss": 1.7516,
+      "step": 611
+    },
+    {
+      "epoch": 1.6744186046511627,
+      "grad_norm": 0.2955617904663086,
+      "learning_rate": 2.109998304636048e-05,
+      "loss": 1.4681,
+      "step": 612
+    },
+    {
+      "epoch": 1.6771545827633378,
+      "grad_norm": 0.3331190347671509,
+      "learning_rate": 2.1027699540291965e-05,
+      "loss": 1.5334,
+      "step": 613
+    },
+    {
+      "epoch": 1.679890560875513,
+      "grad_norm": 0.39551764726638794,
+      "learning_rate": 2.0955450086180882e-05,
+      "loss": 1.497,
+      "step": 614
+    },
+    {
+      "epoch": 1.6826265389876882,
+      "grad_norm": 0.2910989224910736,
+      "learning_rate": 2.088323530337498e-05,
+      "loss": 1.4593,
+      "step": 615
+    },
+    {
+      "epoch": 1.6853625170998632,
+      "grad_norm": 0.2927062511444092,
+      "learning_rate": 2.0811055810924768e-05,
+      "loss": 1.5019,
+      "step": 616
+    },
+    {
+      "epoch": 1.6880984952120381,
+      "grad_norm": 0.2986646592617035,
+      "learning_rate": 2.0738912227578283e-05,
+      "loss": 1.5167,
+      "step": 617
+    },
+    {
+      "epoch": 1.6908344733242133,
+      "grad_norm": 0.2710409462451935,
+      "learning_rate": 2.0666805171775687e-05,
+      "loss": 1.5491,
+      "step": 618
+    },
+    {
+      "epoch": 1.6935704514363885,
+      "grad_norm": 0.32555902004241943,
+      "learning_rate": 2.0594735261644056e-05,
+      "loss": 1.5945,
+      "step": 619
+    },
+    {
+      "epoch": 1.6963064295485637,
+      "grad_norm": 0.3129582703113556,
+      "learning_rate": 2.0522703114992018e-05,
+      "loss": 1.5388,
+      "step": 620
+    },
+    {
+      "epoch": 1.6990424076607387,
+      "grad_norm": 0.29060521721839905,
+      "learning_rate": 2.0450709349304463e-05,
+      "loss": 1.5335,
+      "step": 621
+    },
+    {
+      "epoch": 1.7017783857729138,
+      "grad_norm": 0.2991231381893158,
+      "learning_rate": 2.0378754581737308e-05,
+      "loss": 1.5857,
+      "step": 622
+    },
+    {
+      "epoch": 1.7045143638850888,
+      "grad_norm": 0.3089081346988678,
+      "learning_rate": 2.0306839429112136e-05,
+      "loss": 1.509,
+      "step": 623
+    },
+    {
+      "epoch": 1.707250341997264,
+      "grad_norm": 0.30709123611450195,
+      "learning_rate": 2.023496450791093e-05,
+      "loss": 1.5935,
+      "step": 624
+    },
+    {
+      "epoch": 1.7099863201094392,
+      "grad_norm": 0.29340359568595886,
+      "learning_rate": 2.0163130434270833e-05,
+      "loss": 1.547,
+      "step": 625
+    },
+    {
+      "epoch": 1.7127222982216144,
+      "grad_norm": 0.3518681824207306,
+      "learning_rate": 2.009133782397879e-05,
+      "loss": 1.513,
+      "step": 626
+    },
+    {
+      "epoch": 1.7154582763337893,
+      "grad_norm": 0.31759360432624817,
+      "learning_rate": 2.0019587292466306e-05,
+      "loss": 1.5252,
+      "step": 627
+    },
+    {
+      "epoch": 1.7181942544459643,
+      "grad_norm": 0.2931315302848816,
+      "learning_rate": 1.9947879454804216e-05,
+      "loss": 1.7602,
+      "step": 628
+    },
+    {
+      "epoch": 1.7209302325581395,
+      "grad_norm": 0.2938210368156433,
+      "learning_rate": 1.9876214925697323e-05,
+      "loss": 1.5154,
+      "step": 629
+    },
+    {
+      "epoch": 1.7236662106703147,
+      "grad_norm": 0.28195831179618835,
+      "learning_rate": 1.980459431947918e-05,
+      "loss": 1.4129,
+      "step": 630
+    },
+    {
+      "epoch": 1.7264021887824899,
+      "grad_norm": 0.31024137139320374,
+      "learning_rate": 1.973301825010685e-05,
+      "loss": 1.5008,
+      "step": 631
+    },
+    {
+      "epoch": 1.7291381668946648,
+      "grad_norm": 0.3060285747051239,
+      "learning_rate": 1.9661487331155563e-05,
+      "loss": 1.5469,
+      "step": 632
+    },
+    {
+      "epoch": 1.7318741450068398,
+      "grad_norm": 0.38896477222442627,
+      "learning_rate": 1.9590002175813553e-05,
+      "loss": 1.6018,
+      "step": 633
+    },
+    {
+      "epoch": 1.734610123119015,
+      "grad_norm": 0.32051512598991394,
+      "learning_rate": 1.9518563396876717e-05,
+      "loss": 1.5078,
+      "step": 634
+    },
+    {
+      "epoch": 1.7373461012311902,
+      "grad_norm": 0.2726178467273712,
+      "learning_rate": 1.9447171606743398e-05,
+      "loss": 1.4586,
+      "step": 635
+    },
+    {
+      "epoch": 1.7400820793433653,
+      "grad_norm": 0.2772533595561981,
+      "learning_rate": 1.9375827417409165e-05,
+      "loss": 1.5022,
+      "step": 636
+    },
+    {
+      "epoch": 1.7428180574555403,
+      "grad_norm": 0.30202561616897583,
+      "learning_rate": 1.9304531440461506e-05,
+      "loss": 1.5036,
+      "step": 637
+    },
+    {
+      "epoch": 1.7455540355677155,
+      "grad_norm": 0.2932077944278717,
+      "learning_rate": 1.923328428707461e-05,
+      "loss": 1.5541,
+      "step": 638
+    },
+    {
+      "epoch": 1.7482900136798905,
+      "grad_norm": 0.3096188008785248,
+      "learning_rate": 1.916208656800418e-05,
+      "loss": 1.3744,
+      "step": 639
+    },
+    {
+      "epoch": 1.7510259917920656,
+      "grad_norm": 0.282226026058197,
+      "learning_rate": 1.9090938893582088e-05,
+      "loss": 1.6306,
+      "step": 640
+    },
+    {
+      "epoch": 1.7537619699042408,
+      "grad_norm": 0.27213621139526367,
+      "learning_rate": 1.9019841873711255e-05,
+      "loss": 1.4953,
+      "step": 641
+    },
+    {
+      "epoch": 1.756497948016416,
+      "grad_norm": 0.30166521668434143,
+      "learning_rate": 1.8948796117860348e-05,
+      "loss": 1.5158,
+      "step": 642
+    },
+    {
+      "epoch": 1.759233926128591,
+      "grad_norm": 0.27089768648147583,
+      "learning_rate": 1.8877802235058585e-05,
+      "loss": 1.4954,
+      "step": 643
+    },
+    {
+      "epoch": 1.761969904240766,
+      "grad_norm": 0.31477153301239014,
+      "learning_rate": 1.8806860833890528e-05,
+      "loss": 1.5449,
+      "step": 644
+    },
+    {
+      "epoch": 1.7647058823529411,
+      "grad_norm": 0.2903262674808502,
+      "learning_rate": 1.8735972522490818e-05,
+      "loss": 1.4129,
+      "step": 645
+    },
+    {
+      "epoch": 1.7674418604651163,
+      "grad_norm": 0.3140583038330078,
+      "learning_rate": 1.8665137908539004e-05,
+      "loss": 1.6374,
+      "step": 646
+    },
+    {
+      "epoch": 1.7701778385772915,
+      "grad_norm": 0.29897943139076233,
+      "learning_rate": 1.859435759925434e-05,
+      "loss": 1.605,
+      "step": 647
+    },
+    {
+      "epoch": 1.7729138166894665,
+      "grad_norm": 0.2993355989456177,
+      "learning_rate": 1.8523632201390537e-05,
+      "loss": 1.495,
+      "step": 648
+    },
+    {
+      "epoch": 1.7756497948016414,
+      "grad_norm": 0.3301627039909363,
+      "learning_rate": 1.845296232123058e-05,
+      "loss": 1.6064,
+      "step": 649
+    },
+    {
+      "epoch": 1.7783857729138166,
+      "grad_norm": 0.30024030804634094,
+      "learning_rate": 1.8382348564581574e-05,
+      "loss": 1.3258,
+      "step": 650
+    },
+    {
+      "epoch": 1.7811217510259918,
+      "grad_norm": 0.3279036283493042,
+      "learning_rate": 1.8311791536769483e-05,
+      "loss": 1.4348,
+      "step": 651
+    },
+    {
+      "epoch": 1.783857729138167,
+      "grad_norm": 0.3231453001499176,
+      "learning_rate": 1.8241291842633966e-05,
+      "loss": 1.4327,
+      "step": 652
+    },
+    {
+      "epoch": 1.786593707250342,
+      "grad_norm": 0.3033272624015808,
+      "learning_rate": 1.817085008652324e-05,
+      "loss": 1.4668,
+      "step": 653
+    },
+    {
+      "epoch": 1.7893296853625171,
+      "grad_norm": 0.28523799777030945,
+      "learning_rate": 1.810046687228879e-05,
+      "loss": 1.5058,
+      "step": 654
+    },
+    {
+      "epoch": 1.792065663474692,
+      "grad_norm": 0.35666465759277344,
+      "learning_rate": 1.803014280328033e-05,
+      "loss": 1.4551,
+      "step": 655
+    },
+    {
+      "epoch": 1.7948016415868673,
+      "grad_norm": 0.33736082911491394,
+      "learning_rate": 1.795987848234051e-05,
+      "loss": 1.2991,
+      "step": 656
+    },
+    {
+      "epoch": 1.7975376196990425,
+      "grad_norm": 0.3194032311439514,
+      "learning_rate": 1.7889674511799812e-05,
+      "loss": 1.4764,
+      "step": 657
+    },
+    {
+      "epoch": 1.8002735978112177,
+      "grad_norm": 0.28460192680358887,
+      "learning_rate": 1.7819531493471392e-05,
+      "loss": 1.4962,
+      "step": 658
+    },
+    {
+      "epoch": 1.8030095759233926,
+      "grad_norm": 0.3380921483039856,
+      "learning_rate": 1.7749450028645875e-05,
+      "loss": 1.4772,
+      "step": 659
+    },
+    {
+      "epoch": 1.8057455540355676,
+      "grad_norm": 0.32514598965644836,
+      "learning_rate": 1.7679430718086243e-05,
+      "loss": 1.6421,
+      "step": 660
+    },
+    {
+      "epoch": 1.8084815321477428,
+      "grad_norm": 0.293314129114151,
+      "learning_rate": 1.7609474162022665e-05,
+      "loss": 1.5151,
+      "step": 661
+    },
+    {
+      "epoch": 1.811217510259918,
+      "grad_norm": 0.3001095950603485,
+      "learning_rate": 1.753958096014737e-05,
+      "loss": 1.4422,
+      "step": 662
+    },
+    {
+      "epoch": 1.8139534883720931,
+      "grad_norm": 0.3403547704219818,
+      "learning_rate": 1.7469751711609454e-05,
+      "loss": 1.6122,
+      "step": 663
+    },
+    {
+      "epoch": 1.816689466484268,
+      "grad_norm": 0.3033615052700043,
+      "learning_rate": 1.739998701500984e-05,
+      "loss": 1.596,
+      "step": 664
+    },
+    {
+      "epoch": 1.819425444596443,
+      "grad_norm": 0.34662926197052,
+      "learning_rate": 1.7330287468396033e-05,
+      "loss": 1.5458,
+      "step": 665
+    },
+    {
+      "epoch": 1.8221614227086183,
+      "grad_norm": 0.32165685296058655,
+      "learning_rate": 1.7260653669257093e-05,
+      "loss": 1.5433,
+      "step": 666
+    },
+    {
+      "epoch": 1.8248974008207934,
+      "grad_norm": 0.32157987356185913,
+      "learning_rate": 1.7191086214518447e-05,
+      "loss": 1.6236,
+      "step": 667
+    },
+    {
+      "epoch": 1.8276333789329686,
+      "grad_norm": 0.3016352355480194,
+      "learning_rate": 1.712158570053678e-05,
+      "loss": 1.3893,
+      "step": 668
+    },
+    {
+      "epoch": 1.8303693570451436,
+      "grad_norm": 0.40459707379341125,
+      "learning_rate": 1.7052152723094976e-05,
+      "loss": 1.3767,
+      "step": 669
+    },
+    {
+      "epoch": 1.8331053351573188,
+      "grad_norm": 0.34724652767181396,
+      "learning_rate": 1.698278787739695e-05,
+      "loss": 1.6067,
+      "step": 670
+    },
+    {
+      "epoch": 1.8358413132694937,
+      "grad_norm": 0.3703734874725342,
+      "learning_rate": 1.6913491758062557e-05,
+      "loss": 1.4538,
+      "step": 671
+    },
+    {
+      "epoch": 1.838577291381669,
+      "grad_norm": 0.31886354088783264,
+      "learning_rate": 1.6844264959122534e-05,
+      "loss": 1.6776,
+      "step": 672
+    },
+    {
+      "epoch": 1.841313269493844,
+      "grad_norm": 0.32411274313926697,
+      "learning_rate": 1.6775108074013356e-05,
+      "loss": 1.5405,
+      "step": 673
+    },
+    {
+      "epoch": 1.8440492476060193,
+      "grad_norm": 0.32398101687431335,
+      "learning_rate": 1.670602169557217e-05,
+      "loss": 1.5429,
+      "step": 674
+    },
+    {
+      "epoch": 1.8467852257181943,
+      "grad_norm": 0.3220042288303375,
+      "learning_rate": 1.663700641603174e-05,
+      "loss": 1.5339,
+      "step": 675
+    },
+    {
+      "epoch": 1.8495212038303692,
+      "grad_norm": 0.2694465219974518,
+      "learning_rate": 1.6568062827015317e-05,
+      "loss": 1.5641,
+      "step": 676
+    },
+    {
+      "epoch": 1.8522571819425444,
+      "grad_norm": 0.3089315593242645,
+      "learning_rate": 1.6499191519531614e-05,
+      "loss": 1.3319,
+      "step": 677
+    },
+    {
+      "epoch": 1.8549931600547196,
+      "grad_norm": 0.3302357792854309,
+      "learning_rate": 1.643039308396971e-05,
+      "loss": 1.478,
+      "step": 678
+    },
+    {
+      "epoch": 1.8577291381668948,
+      "grad_norm": 0.30908071994781494,
+      "learning_rate": 1.6361668110094007e-05,
+      "loss": 1.5557,
+      "step": 679
+    },
+    {
+      "epoch": 1.8604651162790697,
+      "grad_norm": 0.2804611027240753,
+      "learning_rate": 1.6293017187039174e-05,
+      "loss": 1.6756,
+      "step": 680
+    },
+    {
+      "epoch": 1.8632010943912447,
+      "grad_norm": 0.3693443238735199,
+      "learning_rate": 1.6224440903305076e-05,
+      "loss": 1.5825,
+      "step": 681
+    },
+    {
+      "epoch": 1.86593707250342,
+      "grad_norm": 0.33380305767059326,
+      "learning_rate": 1.615593984675174e-05,
+      "loss": 1.5312,
+      "step": 682
+    },
+    {
+      "epoch": 1.868673050615595,
+      "grad_norm": 0.3250941038131714,
+      "learning_rate": 1.6087514604594353e-05,
+      "loss": 1.5642,
+      "step": 683
+    },
+    {
+      "epoch": 1.8714090287277703,
+      "grad_norm": 0.3749421238899231,
+      "learning_rate": 1.6019165763398152e-05,
+      "loss": 1.6141,
+      "step": 684
+    },
+    {
+      "epoch": 1.8741450068399452,
+      "grad_norm": 0.32209932804107666,
+      "learning_rate": 1.5950893909073453e-05,
+      "loss": 1.4791,
+      "step": 685
+    },
+    {
+      "epoch": 1.8768809849521204,
+      "grad_norm": 0.34071722626686096,
+      "learning_rate": 1.5882699626870633e-05,
+      "loss": 1.6126,
+      "step": 686
+    },
+    {
+      "epoch": 1.8796169630642954,
+      "grad_norm": 0.2984413802623749,
+      "learning_rate": 1.5814583501375064e-05,
+      "loss": 1.5191,
+      "step": 687
+    },
+    {
+      "epoch": 1.8823529411764706,
+      "grad_norm": 0.33248278498649597,
+      "learning_rate": 1.574654611650214e-05,
+      "loss": 1.6346,
+      "step": 688
+    },
+    {
+      "epoch": 1.8850889192886457,
+      "grad_norm": 0.308098703622818,
+      "learning_rate": 1.567858805549229e-05,
+      "loss": 1.5521,
+      "step": 689
+    },
+    {
+      "epoch": 1.887824897400821,
+      "grad_norm": 0.2899650037288666,
+      "learning_rate": 1.56107099009059e-05,
+      "loss": 1.3675,
+      "step": 690
+    },
+    {
+      "epoch": 1.890560875512996,
+      "grad_norm": 0.33524468541145325,
+      "learning_rate": 1.5542912234618422e-05,
+      "loss": 1.6501,
+      "step": 691
+    },
+    {
+      "epoch": 1.8932968536251709,
+      "grad_norm": 0.3770296275615692,
+      "learning_rate": 1.54751956378153e-05,
+      "loss": 1.5518,
+      "step": 692
+    },
+    {
+      "epoch": 1.896032831737346,
+      "grad_norm": 0.3348696529865265,
+      "learning_rate": 1.540756069098702e-05,
+      "loss": 1.4388,
+      "step": 693
+    },
+    {
+      "epoch": 1.8987688098495212,
+      "grad_norm": 0.37193214893341064,
+      "learning_rate": 1.5340007973924176e-05,
+      "loss": 1.499,
+      "step": 694
+    },
+    {
+      "epoch": 1.9015047879616964,
+      "grad_norm": 0.33301231265068054,
+      "learning_rate": 1.5272538065712403e-05,
+      "loss": 1.6687,
+      "step": 695
+    },
+    {
+      "epoch": 1.9042407660738714,
+      "grad_norm": 0.33597227931022644,
+      "learning_rate": 1.5205151544727509e-05,
+      "loss": 1.589,
+      "step": 696
+    },
+    {
+      "epoch": 1.9069767441860463,
+      "grad_norm": 0.3002086281776428,
+      "learning_rate": 1.5137848988630465e-05,
+      "loss": 1.7462,
+      "step": 697
+    },
+    {
+      "epoch": 1.9097127222982215,
+      "grad_norm": 0.30558449029922485,
+      "learning_rate": 1.5070630974362473e-05,
+      "loss": 1.6179,
+      "step": 698
+    },
+    {
+      "epoch": 1.9124487004103967,
+      "grad_norm": 0.314813494682312,
+      "learning_rate": 1.5003498078139988e-05,
+      "loss": 1.6029,
+      "step": 699
+    },
+    {
+      "epoch": 1.915184678522572,
+      "grad_norm": 0.31106236577033997,
+      "learning_rate": 1.4936450875449845e-05,
+      "loss": 1.5462,
+      "step": 700
+    },
+    {
+      "epoch": 1.9179206566347469,
+      "grad_norm": 0.32419320940971375,
+      "learning_rate": 1.4869489941044235e-05,
+      "loss": 1.632,
+      "step": 701
+    },
+    {
+      "epoch": 1.920656634746922,
+      "grad_norm": 0.28728801012039185,
+      "learning_rate": 1.4802615848935875e-05,
+      "loss": 1.5625,
+      "step": 702
+    },
+    {
+      "epoch": 1.923392612859097,
+      "grad_norm": 0.35819682478904724,
+      "learning_rate": 1.4735829172393007e-05,
+      "loss": 1.416,
+      "step": 703
+    },
+    {
+      "epoch": 1.9261285909712722,
+      "grad_norm": 0.3097285032272339,
+      "learning_rate": 1.4669130483934512e-05,
+      "loss": 1.4333,
+      "step": 704
+    },
+    {
+      "epoch": 1.9288645690834474,
+      "grad_norm": 0.31058430671691895,
+      "learning_rate": 1.4602520355325039e-05,
+      "loss": 1.5175,
+      "step": 705
+    },
+    {
+      "epoch": 1.9316005471956226,
+      "grad_norm": 0.3093342185020447,
+      "learning_rate": 1.4535999357570046e-05,
+      "loss": 1.4105,
+      "step": 706
+    },
+    {
+      "epoch": 1.9343365253077975,
+      "grad_norm": 0.3064606189727783,
+      "learning_rate": 1.4469568060910915e-05,
+      "loss": 1.6593,
+      "step": 707
+    },
+    {
+      "epoch": 1.9370725034199725,
+      "grad_norm": 0.3192756772041321,
+      "learning_rate": 1.4403227034820139e-05,
+      "loss": 1.5919,
+      "step": 708
+    },
+    {
+      "epoch": 1.9398084815321477,
+      "grad_norm": 0.30623388290405273,
+      "learning_rate": 1.4336976847996303e-05,
+      "loss": 1.5053,
+      "step": 709
+    },
+    {
+      "epoch": 1.9425444596443229,
+      "grad_norm": 0.3014221787452698,
+      "learning_rate": 1.4270818068359336e-05,
+      "loss": 1.4876,
+      "step": 710
+    },
+    {
+      "epoch": 1.945280437756498,
+      "grad_norm": 0.34191763401031494,
+      "learning_rate": 1.420475126304558e-05,
+      "loss": 1.435,
+      "step": 711
+    },
+    {
+      "epoch": 1.948016415868673,
+      "grad_norm": 0.33173930644989014,
+      "learning_rate": 1.4138776998402927e-05,
+      "loss": 1.5467,
+      "step": 712
+    },
+    {
+      "epoch": 1.950752393980848,
+      "grad_norm": 0.3360328674316406,
+      "learning_rate": 1.4072895839986023e-05,
+      "loss": 1.4525,
+      "step": 713
+    },
+    {
+      "epoch": 1.9534883720930232,
+      "grad_norm": 0.2906378209590912,
+      "learning_rate": 1.4007108352551313e-05,
+      "loss": 1.5788,
+      "step": 714
+    },
+    {
+      "epoch": 1.9562243502051984,
+      "grad_norm": 0.3077907860279083,
+      "learning_rate": 1.3941415100052293e-05,
+      "loss": 1.4087,
+      "step": 715
+    },
+    {
+      "epoch": 1.9589603283173735,
+      "grad_norm": 0.2951776385307312,
+      "learning_rate": 1.387581664563465e-05,
+      "loss": 1.4699,
+      "step": 716
+    },
+    {
+      "epoch": 1.9616963064295485,
+      "grad_norm": 0.29453444480895996,
+      "learning_rate": 1.3810313551631426e-05,
+      "loss": 1.6203,
+      "step": 717
+    },
+    {
+      "epoch": 1.9644322845417237,
+      "grad_norm": 0.3803625702857971,
+      "learning_rate": 1.3744906379558165e-05,
+      "loss": 1.6514,
+      "step": 718
+    },
+    {
+      "epoch": 1.9671682626538987,
+      "grad_norm": 0.3245275318622589,
+      "learning_rate": 1.3679595690108193e-05,
+      "loss": 1.5904,
+      "step": 719
+    },
+    {
+      "epoch": 1.9699042407660738,
+      "grad_norm": 0.3239131271839142,
+      "learning_rate": 1.3614382043147725e-05,
+      "loss": 1.5117,
+      "step": 720
+    },
+    {
+      "epoch": 1.972640218878249,
+      "grad_norm": 0.3001211881637573,
+      "learning_rate": 1.3549265997711057e-05,
+      "loss": 1.5985,
+      "step": 721
+    },
+    {
+      "epoch": 1.9753761969904242,
+      "grad_norm": 0.29681622982025146,
+      "learning_rate": 1.3484248111995892e-05,
+      "loss": 1.3762,
+      "step": 722
+    },
+    {
+      "epoch": 1.9781121751025992,
+      "grad_norm": 0.38813284039497375,
+      "learning_rate": 1.3419328943358392e-05,
+      "loss": 1.548,
+      "step": 723
+    },
+    {
+      "epoch": 1.9808481532147741,
+      "grad_norm": 0.29337620735168457,
+      "learning_rate": 1.3354509048308527e-05,
+      "loss": 1.4466,
+      "step": 724
+    },
+    {
+      "epoch": 1.9835841313269493,
+      "grad_norm": 0.3134080171585083,
+      "learning_rate": 1.328978898250525e-05,
+      "loss": 1.523,
+      "step": 725
+    },
+    {
+      "epoch": 1.9863201094391245,
+      "grad_norm": 0.32457372546195984,
+      "learning_rate": 1.3225169300751738e-05,
+      "loss": 1.4018,
+      "step": 726
+    },
+    {
+      "epoch": 1.9890560875512997,
+      "grad_norm": 0.30359819531440735,
+      "learning_rate": 1.3160650556990644e-05,
+      "loss": 1.4209,
+      "step": 727
+    },
+    {
+      "epoch": 1.9917920656634747,
+      "grad_norm": 0.3376854360103607,
+      "learning_rate": 1.3096233304299346e-05,
+      "loss": 1.4914,
+      "step": 728
+    },
+    {
+      "epoch": 1.9945280437756496,
+      "grad_norm": 0.3038989305496216,
+      "learning_rate": 1.30319180948852e-05,
+      "loss": 1.4995,
+      "step": 729
+    },
+    {
+      "epoch": 1.9972640218878248,
+      "grad_norm": 0.2987303137779236,
+      "learning_rate": 1.2967705480080819e-05,
+      "loss": 1.5415,
+      "step": 730
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.36076483130455017,
+      "learning_rate": 1.2903596010339338e-05,
+      "loss": 1.3103,
+      "step": 731
+    },
+    {
+      "epoch": 2.002735978112175,
+      "grad_norm": 0.329522043466568,
+      "learning_rate": 1.2839590235229668e-05,
+      "loss": 1.7937,
+      "step": 732
+    },
+    {
+      "epoch": 2.0054719562243504,
+      "grad_norm": 0.33667078614234924,
+      "learning_rate": 1.2775688703431871e-05,
+      "loss": 1.333,
+      "step": 733
+    },
+    {
+      "epoch": 2.008207934336525,
+      "grad_norm": 0.30493631958961487,
+      "learning_rate": 1.2711891962732342e-05,
+      "loss": 1.3868,
+      "step": 734
+    },
+    {
+      "epoch": 2.0109439124487003,
+      "grad_norm": 0.3139844536781311,
+      "learning_rate": 1.26482005600192e-05,
+      "loss": 1.4361,
+      "step": 735
+    },
+    {
+      "epoch": 2.0136798905608755,
+      "grad_norm": 0.30529195070266724,
+      "learning_rate": 1.258461504127756e-05,
+      "loss": 1.5679,
+      "step": 736
+    },
+    {
+      "epoch": 2.0164158686730507,
+      "grad_norm": 0.3701570928096771,
+      "learning_rate": 1.252113595158487e-05,
+      "loss": 1.7087,
+      "step": 737
+    },
+    {
+      "epoch": 2.019151846785226,
+      "grad_norm": 0.3016231060028076,
+      "learning_rate": 1.245776383510622e-05,
+      "loss": 1.5592,
+      "step": 738
+    },
+    {
+      "epoch": 2.0218878248974006,
+      "grad_norm": 0.29768821597099304,
+      "learning_rate": 1.2394499235089699e-05,
+      "loss": 1.4926,
+      "step": 739
+    },
+    {
+      "epoch": 2.024623803009576,
+      "grad_norm": 0.3342791497707367,
+      "learning_rate": 1.2331342693861716e-05,
+      "loss": 1.4944,
+      "step": 740
+    },
+    {
+      "epoch": 2.027359781121751,
+      "grad_norm": 0.30851346254348755,
+      "learning_rate": 1.2268294752822376e-05,
+      "loss": 1.7049,
+      "step": 741
+    },
+    {
+      "epoch": 2.030095759233926,
+      "grad_norm": 0.3023451864719391,
+      "learning_rate": 1.2205355952440817e-05,
+      "loss": 1.5705,
+      "step": 742
+    },
+    {
+      "epoch": 2.0328317373461013,
+      "grad_norm": 0.3253107964992523,
+      "learning_rate": 1.2142526832250561e-05,
+      "loss": 1.4404,
+      "step": 743
+    },
+    {
+      "epoch": 2.0355677154582765,
+      "grad_norm": 0.33139538764953613,
+      "learning_rate": 1.2079807930844977e-05,
+      "loss": 1.4355,
+      "step": 744
+    },
+    {
+      "epoch": 2.0383036935704513,
+      "grad_norm": 0.3191182613372803,
+      "learning_rate": 1.2017199785872523e-05,
+      "loss": 1.428,
+      "step": 745
+    },
+    {
+      "epoch": 2.0410396716826265,
+      "grad_norm": 0.33677810430526733,
+      "learning_rate": 1.1954702934032267e-05,
+      "loss": 1.4214,
+      "step": 746
+    },
+    {
+      "epoch": 2.0437756497948016,
+      "grad_norm": 0.31797662377357483,
+      "learning_rate": 1.1892317911069212e-05,
+      "loss": 1.5024,
+      "step": 747
+    },
+    {
+      "epoch": 2.046511627906977,
+      "grad_norm": 0.30894604325294495,
+      "learning_rate": 1.183004525176973e-05,
+      "loss": 1.5084,
+      "step": 748
+    },
+    {
+      "epoch": 2.049247606019152,
+      "grad_norm": 0.30327004194259644,
+      "learning_rate": 1.176788548995697e-05,
+      "loss": 1.5586,
+      "step": 749
+    },
+    {
+      "epoch": 2.0519835841313268,
+      "grad_norm": 0.31576216220855713,
+      "learning_rate": 1.1705839158486284e-05,
+      "loss": 1.5315,
+      "step": 750
+    },
+    {
+      "epoch": 2.0519835841313268,
+      "eval_loss": 1.5589618682861328,
+      "eval_runtime": 12.9752,
+      "eval_samples_per_second": 11.869,
+      "eval_steps_per_second": 11.869,
+      "step": 750
+    },
+    {
+      "epoch": 2.054719562243502,
+      "grad_norm": 0.3079953193664551,
+      "learning_rate": 1.1643906789240664e-05,
+      "loss": 1.6127,
+      "step": 751
+    },
+    {
+      "epoch": 2.057455540355677,
+      "grad_norm": 0.30269429087638855,
+      "learning_rate": 1.1582088913126172e-05,
+      "loss": 1.5077,
+      "step": 752
+    },
+    {
+      "epoch": 2.0601915184678523,
+      "grad_norm": 0.34395742416381836,
+      "learning_rate": 1.1520386060067401e-05,
+      "loss": 1.4675,
+      "step": 753
+    },
+    {
+      "epoch": 2.0629274965800275,
+      "grad_norm": 0.3478207588195801,
+      "learning_rate": 1.1458798759002897e-05,
+      "loss": 1.5474,
+      "step": 754
+    },
+    {
+      "epoch": 2.0656634746922027,
+      "grad_norm": 0.3087264895439148,
+      "learning_rate": 1.139732753788072e-05,
+      "loss": 1.4078,
+      "step": 755
+    },
+    {
+      "epoch": 2.0683994528043774,
+      "grad_norm": 0.2971005141735077,
+      "learning_rate": 1.1335972923653774e-05,
+      "loss": 1.6057,
+      "step": 756
+    },
+    {
+      "epoch": 2.0711354309165526,
+      "grad_norm": 0.30974239110946655,
+      "learning_rate": 1.1274735442275402e-05,
+      "loss": 1.4899,
+      "step": 757
+    },
+    {
+      "epoch": 2.073871409028728,
+      "grad_norm": 0.3504452705383301,
+      "learning_rate": 1.121361561869488e-05,
+      "loss": 1.5413,
+      "step": 758
+    },
+    {
+      "epoch": 2.076607387140903,
+      "grad_norm": 0.3584803640842438,
+      "learning_rate": 1.1152613976852804e-05,
+      "loss": 1.5634,
+      "step": 759
+    },
+    {
+      "epoch": 2.079343365253078,
+      "grad_norm": 0.3279834985733032,
+      "learning_rate": 1.1091731039676754e-05,
+      "loss": 1.4428,
+      "step": 760
+    },
+    {
+      "epoch": 2.082079343365253,
+      "grad_norm": 0.2989148199558258,
+      "learning_rate": 1.1030967329076658e-05,
+      "loss": 1.3392,
+      "step": 761
+    },
+    {
+      "epoch": 2.084815321477428,
+      "grad_norm": 0.3270657956600189,
+      "learning_rate": 1.0970323365940444e-05,
+      "loss": 1.4114,
+      "step": 762
+    },
+    {
+      "epoch": 2.0875512995896033,
+      "grad_norm": 0.31774088740348816,
+      "learning_rate": 1.0909799670129497e-05,
+      "loss": 1.4251,
+      "step": 763
+    },
+    {
+      "epoch": 2.0902872777017785,
+      "grad_norm": 0.3060950040817261,
+      "learning_rate": 1.0849396760474246e-05,
+      "loss": 1.4517,
+      "step": 764
+    },
+    {
+      "epoch": 2.0930232558139537,
+      "grad_norm": 0.2948378026485443,
+      "learning_rate": 1.078911515476968e-05,
+      "loss": 1.4872,
+      "step": 765
+    },
+    {
+      "epoch": 2.0957592339261284,
+      "grad_norm": 0.2836145758628845,
+      "learning_rate": 1.0728955369770941e-05,
+      "loss": 1.4987,
+      "step": 766
+    },
+    {
+      "epoch": 2.0984952120383036,
+      "grad_norm": 0.33373308181762695,
+      "learning_rate": 1.0668917921188885e-05,
+      "loss": 1.4806,
+      "step": 767
+    },
+    {
+      "epoch": 2.1012311901504788,
+      "grad_norm": 0.29727426171302795,
+      "learning_rate": 1.060900332368562e-05,
+      "loss": 1.506,
+      "step": 768
+    },
+    {
+      "epoch": 2.103967168262654,
+      "grad_norm": 0.3088816702365875,
+      "learning_rate": 1.0549212090870203e-05,
+      "loss": 1.7262,
+      "step": 769
+    },
+    {
+      "epoch": 2.106703146374829,
+      "grad_norm": 0.37396734952926636,
+      "learning_rate": 1.0489544735294088e-05,
+      "loss": 1.4311,
+      "step": 770
+    },
+    {
+      "epoch": 2.109439124487004,
+      "grad_norm": 0.3265590965747833,
+      "learning_rate": 1.0430001768446856e-05,
+      "loss": 1.3106,
+      "step": 771
+    },
+    {
+      "epoch": 2.112175102599179,
+      "grad_norm": 0.3493903577327728,
+      "learning_rate": 1.0370583700751774e-05,
+      "loss": 1.6644,
+      "step": 772
+    },
+    {
+      "epoch": 2.1149110807113543,
+      "grad_norm": 0.32931625843048096,
+      "learning_rate": 1.0311291041561428e-05,
+      "loss": 1.448,
+      "step": 773
+    },
+    {
+      "epoch": 2.1176470588235294,
+      "grad_norm": 0.3126463294029236,
+      "learning_rate": 1.0252124299153353e-05,
+      "loss": 1.5439,
+      "step": 774
+    },
+    {
+      "epoch": 2.1203830369357046,
+      "grad_norm": 0.3198404014110565,
+      "learning_rate": 1.0193083980725696e-05,
+      "loss": 1.4068,
+      "step": 775
+    },
+    {
+      "epoch": 2.12311901504788,
+      "grad_norm": 0.3446153998374939,
+      "learning_rate": 1.0134170592392836e-05,
+      "loss": 1.4597,
+      "step": 776
+    },
+    {
+      "epoch": 2.1258549931600546,
+      "grad_norm": 0.34225985407829285,
+      "learning_rate": 1.007538463918107e-05,
+      "loss": 1.5849,
+      "step": 777
+    },
+    {
+      "epoch": 2.1285909712722297,
+      "grad_norm": 0.3094581663608551,
+      "learning_rate": 1.0016726625024287e-05,
+      "loss": 1.5853,
+      "step": 778
+    },
+    {
+      "epoch": 2.131326949384405,
+      "grad_norm": 0.27853038907051086,
+      "learning_rate": 9.9581970527596e-06,
+      "loss": 1.5799,
+      "step": 779
+    },
+    {
+      "epoch": 2.13406292749658,
+      "grad_norm": 0.33034032583236694,
+      "learning_rate": 9.899796424123136e-06,
+      "loss": 1.456,
+      "step": 780
+    },
+    {
+      "epoch": 2.1367989056087553,
+      "grad_norm": 0.3268153965473175,
+      "learning_rate": 9.841525239745605e-06,
+      "loss": 1.3883,
+      "step": 781
+    },
+    {
+      "epoch": 2.13953488372093,
+      "grad_norm": 0.32485464215278625,
+      "learning_rate": 9.783383999148118e-06,
+      "loss": 1.4159,
+      "step": 782
+    },
+    {
+      "epoch": 2.1422708618331052,
+      "grad_norm": 0.29737138748168945,
+      "learning_rate": 9.72537320073785e-06,
+      "loss": 1.5254,
+      "step": 783
+    },
+    {
+      "epoch": 2.1450068399452804,
+      "grad_norm": 0.2851496934890747,
+      "learning_rate": 9.667493341803777e-06,
+      "loss": 1.3903,
+      "step": 784
+    },
+    {
+      "epoch": 2.1477428180574556,
+      "grad_norm": 0.32795658707618713,
+      "learning_rate": 9.60974491851242e-06,
+      "loss": 1.5147,
+      "step": 785
+    },
+    {
+      "epoch": 2.150478796169631,
+      "grad_norm": 0.33503639698028564,
+      "learning_rate": 9.552128425903586e-06,
+      "loss": 1.4495,
+      "step": 786
+    },
+    {
+      "epoch": 2.153214774281806,
+      "grad_norm": 0.31696316599845886,
+      "learning_rate": 9.494644357886124e-06,
+      "loss": 1.5787,
+      "step": 787
+    },
+    {
+      "epoch": 2.1559507523939807,
+      "grad_norm": 0.321809321641922,
+      "learning_rate": 9.437293207233695e-06,
+      "loss": 1.5817,
+      "step": 788
+    },
+    {
+      "epoch": 2.158686730506156,
+      "grad_norm": 0.29343757033348083,
+      "learning_rate": 9.380075465580552e-06,
+      "loss": 1.5604,
+      "step": 789
+    },
+    {
+      "epoch": 2.161422708618331,
+      "grad_norm": 0.4285193383693695,
+      "learning_rate": 9.322991623417285e-06,
+      "loss": 1.4387,
+      "step": 790
+    },
+    {
+      "epoch": 2.1641586867305063,
+      "grad_norm": 0.29187867045402527,
+      "learning_rate": 9.266042170086717e-06,
+      "loss": 1.4871,
+      "step": 791
+    },
+    {
+      "epoch": 2.1668946648426815,
+      "grad_norm": 0.2987869679927826,
+      "learning_rate": 9.209227593779573e-06,
+      "loss": 1.3972,
+      "step": 792
+    },
+    {
+      "epoch": 2.169630642954856,
+      "grad_norm": 0.29556939005851746,
+      "learning_rate": 9.152548381530405e-06,
+      "loss": 1.4057,
+      "step": 793
+    },
+    {
+      "epoch": 2.1723666210670314,
+      "grad_norm": 0.32454240322113037,
+      "learning_rate": 9.096005019213363e-06,
+      "loss": 1.5509,
+      "step": 794
+    },
+    {
+      "epoch": 2.1751025991792066,
+      "grad_norm": 0.30358558893203735,
+      "learning_rate": 9.039597991538043e-06,
+      "loss": 1.5466,
+      "step": 795
+    },
+    {
+      "epoch": 2.1778385772913817,
+      "grad_norm": 0.3079793155193329,
+      "learning_rate": 8.983327782045359e-06,
+      "loss": 1.8167,
+      "step": 796
+    },
+    {
+      "epoch": 2.180574555403557,
+      "grad_norm": 0.3283805847167969,
+      "learning_rate": 8.927194873103322e-06,
+      "loss": 1.5464,
+      "step": 797
+    },
+    {
+      "epoch": 2.1833105335157317,
+      "grad_norm": 0.3089551031589508,
+      "learning_rate": 8.871199745902997e-06,
+      "loss": 1.4067,
+      "step": 798
+    },
+    {
+      "epoch": 2.186046511627907,
+      "grad_norm": 0.3624851405620575,
+      "learning_rate": 8.815342880454311e-06,
+      "loss": 1.4652,
+      "step": 799
+    },
+    {
+      "epoch": 2.188782489740082,
+      "grad_norm": 0.31375181674957275,
+      "learning_rate": 8.75962475558198e-06,
+      "loss": 1.4584,
+      "step": 800
+    },
+    {
+      "epoch": 2.1915184678522572,
+      "grad_norm": 0.36298835277557373,
+      "learning_rate": 8.704045848921358e-06,
+      "loss": 1.6755,
+      "step": 801
+    },
+    {
+      "epoch": 2.1942544459644324,
+      "grad_norm": 0.3472815155982971,
+      "learning_rate": 8.648606636914416e-06,
+      "loss": 1.4749,
+      "step": 802
+    },
+    {
+      "epoch": 2.196990424076607,
+      "grad_norm": 0.30575889348983765,
+      "learning_rate": 8.593307594805586e-06,
+      "loss": 1.616,
+      "step": 803
+    },
+    {
+      "epoch": 2.1997264021887823,
+      "grad_norm": 0.33051928877830505,
+      "learning_rate": 8.53814919663769e-06,
+      "loss": 1.4458,
+      "step": 804
+    },
+    {
+      "epoch": 2.2024623803009575,
+      "grad_norm": 0.31675076484680176,
+      "learning_rate": 8.483131915247968e-06,
+      "loss": 1.4981,
+      "step": 805
+    },
+    {
+      "epoch": 2.2051983584131327,
+      "grad_norm": 0.33154037594795227,
+      "learning_rate": 8.428256222263888e-06,
+      "loss": 1.5108,
+      "step": 806
+    },
+    {
+      "epoch": 2.207934336525308,
+      "grad_norm": 0.3389985263347626,
+      "learning_rate": 8.373522588099211e-06,
+      "loss": 1.4522,
+      "step": 807
+    },
+    {
+      "epoch": 2.210670314637483,
+      "grad_norm": 0.35865381360054016,
+      "learning_rate": 8.318931481949906e-06,
+      "loss": 1.5059,
+      "step": 808
+    },
+    {
+      "epoch": 2.213406292749658,
+      "grad_norm": 0.3100070357322693,
+      "learning_rate": 8.264483371790156e-06,
+      "loss": 1.441,
+      "step": 809
+    },
+    {
+      "epoch": 2.216142270861833,
+      "grad_norm": 0.3514600694179535,
+      "learning_rate": 8.210178724368312e-06,
+      "loss": 1.6657,
+      "step": 810
+    },
+    {
+      "epoch": 2.218878248974008,
+      "grad_norm": 0.2982495427131653,
+      "learning_rate": 8.156018005202926e-06,
+      "loss": 1.5945,
+      "step": 811
+    },
+    {
+      "epoch": 2.2216142270861834,
+      "grad_norm": 0.2904994487762451,
+      "learning_rate": 8.10200167857874e-06,
+      "loss": 1.4298,
+      "step": 812
+    },
+    {
+      "epoch": 2.2243502051983586,
+      "grad_norm": 0.3352917432785034,
+      "learning_rate": 8.04813020754272e-06,
+      "loss": 1.5512,
+      "step": 813
+    },
+    {
+      "epoch": 2.2270861833105333,
+      "grad_norm": 0.31548169255256653,
+      "learning_rate": 7.994404053900076e-06,
+      "loss": 1.5735,
+      "step": 814
+    },
+    {
+      "epoch": 2.2298221614227085,
+      "grad_norm": 0.31222277879714966,
+      "learning_rate": 7.940823678210282e-06,
+      "loss": 1.4862,
+      "step": 815
+    },
+    {
+      "epoch": 2.2325581395348837,
+      "grad_norm": 0.3492494225502014,
+      "learning_rate": 7.887389539783208e-06,
+      "loss": 1.6087,
+      "step": 816
+    },
+    {
+      "epoch": 2.235294117647059,
+      "grad_norm": 0.32157906889915466,
+      "learning_rate": 7.834102096675064e-06,
+      "loss": 1.5771,
+      "step": 817
+    },
+    {
+      "epoch": 2.238030095759234,
+      "grad_norm": 0.3391493260860443,
+      "learning_rate": 7.780961805684575e-06,
+      "loss": 1.5289,
+      "step": 818
+    },
+    {
+      "epoch": 2.2407660738714092,
+      "grad_norm": 0.32275089621543884,
+      "learning_rate": 7.727969122349017e-06,
+      "loss": 1.407,
+      "step": 819
+    },
+    {
+      "epoch": 2.243502051983584,
+      "grad_norm": 0.34808090329170227,
+      "learning_rate": 7.675124500940317e-06,
+      "loss": 1.4914,
+      "step": 820
+    },
+    {
+      "epoch": 2.246238030095759,
+      "grad_norm": 0.35853490233421326,
+      "learning_rate": 7.62242839446117e-06,
+      "loss": 1.4587,
+      "step": 821
+    },
+    {
+      "epoch": 2.2489740082079344,
+      "grad_norm": 0.298554390668869,
+      "learning_rate": 7.569881254641142e-06,
+      "loss": 1.52,
+      "step": 822
+    },
+    {
+      "epoch": 2.2517099863201095,
+      "grad_norm": 0.3177081048488617,
+      "learning_rate": 7.517483531932809e-06,
+      "loss": 1.3685,
+      "step": 823
+    },
+    {
+      "epoch": 2.2544459644322847,
+      "grad_norm": 0.30928516387939453,
+      "learning_rate": 7.4652356755078895e-06,
+      "loss": 1.4636,
+      "step": 824
+    },
+    {
+      "epoch": 2.2571819425444595,
+      "grad_norm": 0.3332304060459137,
+      "learning_rate": 7.4131381332534e-06,
+      "loss": 1.4894,
+      "step": 825
+    },
+    {
+      "epoch": 2.2599179206566347,
+      "grad_norm": 0.331177681684494,
+      "learning_rate": 7.3611913517677875e-06,
+      "loss": 1.5236,
+      "step": 826
+    },
+    {
+      "epoch": 2.26265389876881,
+      "grad_norm": 0.3152174949645996,
+      "learning_rate": 7.30939577635717e-06,
+      "loss": 1.4508,
+      "step": 827
+    },
+    {
+      "epoch": 2.265389876880985,
+      "grad_norm": 0.33167538046836853,
+      "learning_rate": 7.257751851031428e-06,
+      "loss": 1.4507,
+      "step": 828
+    },
+    {
+      "epoch": 2.26812585499316,
+      "grad_norm": 0.26186057925224304,
+      "learning_rate": 7.206260018500469e-06,
+      "loss": 1.581,
+      "step": 829
+    },
+    {
+      "epoch": 2.270861833105335,
+      "grad_norm": 0.31686386466026306,
+      "learning_rate": 7.154920720170399e-06,
+      "loss": 1.5677,
+      "step": 830
+    },
+    {
+      "epoch": 2.27359781121751,
+      "grad_norm": 0.32669878005981445,
+      "learning_rate": 7.1037343961397495e-06,
+      "loss": 1.5512,
+      "step": 831
+    },
+    {
+      "epoch": 2.2763337893296853,
+      "grad_norm": 0.3603561520576477,
+      "learning_rate": 7.05270148519572e-06,
+      "loss": 1.6478,
+      "step": 832
+    },
+    {
+      "epoch": 2.2790697674418605,
+      "grad_norm": 0.3097962737083435,
+      "learning_rate": 7.001822424810359e-06,
+      "loss": 1.5819,
+      "step": 833
+    },
+    {
+      "epoch": 2.2818057455540357,
+      "grad_norm": 0.32818931341171265,
+      "learning_rate": 6.951097651136889e-06,
+      "loss": 1.5053,
+      "step": 834
+    },
+    {
+      "epoch": 2.2845417236662104,
+      "grad_norm": 0.41288748383522034,
+      "learning_rate": 6.900527599005918e-06,
+      "loss": 1.5391,
+      "step": 835
+    },
+    {
+      "epoch": 2.2872777017783856,
+      "grad_norm": 0.3428272008895874,
+      "learning_rate": 6.8501127019217346e-06,
+      "loss": 1.4097,
+      "step": 836
+    },
+    {
+      "epoch": 2.290013679890561,
+      "grad_norm": 0.34903377294540405,
+      "learning_rate": 6.799853392058561e-06,
+      "loss": 1.3627,
+      "step": 837
+    },
+    {
+      "epoch": 2.292749658002736,
+      "grad_norm": 0.3197937309741974,
+      "learning_rate": 6.749750100256902e-06,
+      "loss": 1.5362,
+      "step": 838
+    },
+    {
+      "epoch": 2.295485636114911,
+      "grad_norm": 0.3300299644470215,
+      "learning_rate": 6.699803256019815e-06,
+      "loss": 1.6251,
+      "step": 839
+    },
+    {
+      "epoch": 2.2982216142270864,
+      "grad_norm": 0.33497533202171326,
+      "learning_rate": 6.650013287509199e-06,
+      "loss": 1.5388,
+      "step": 840
+    },
+    {
+      "epoch": 2.300957592339261,
+      "grad_norm": 0.36092567443847656,
+      "learning_rate": 6.600380621542216e-06,
+      "loss": 1.4969,
+      "step": 841
+    },
+    {
+      "epoch": 2.3036935704514363,
+      "grad_norm": 0.3416774868965149,
+      "learning_rate": 6.550905683587513e-06,
+      "loss": 1.5842,
+      "step": 842
+    },
+    {
+      "epoch": 2.3064295485636115,
+      "grad_norm": 0.30258703231811523,
+      "learning_rate": 6.5015888977617016e-06,
+      "loss": 1.3653,
+      "step": 843
+    },
+    {
+      "epoch": 2.3091655266757867,
+      "grad_norm": 0.2986971139907837,
+      "learning_rate": 6.452430686825603e-06,
+      "loss": 1.5223,
+      "step": 844
+    },
+    {
+      "epoch": 2.311901504787962,
+      "grad_norm": 0.32431551814079285,
+      "learning_rate": 6.403431472180707e-06,
+      "loss": 1.4931,
+      "step": 845
+    },
+    {
+      "epoch": 2.314637482900137,
+      "grad_norm": 0.32962799072265625,
+      "learning_rate": 6.354591673865523e-06,
+      "loss": 1.4833,
+      "step": 846
+    },
+    {
+      "epoch": 2.317373461012312,
+      "grad_norm": 0.36730074882507324,
+      "learning_rate": 6.30591171055199e-06,
+      "loss": 1.4419,
+      "step": 847
+    },
+    {
+      "epoch": 2.320109439124487,
+      "grad_norm": 0.3103657066822052,
+      "learning_rate": 6.257391999541887e-06,
+      "loss": 1.4914,
+      "step": 848
+    },
+    {
+      "epoch": 2.322845417236662,
+      "grad_norm": 0.3909322917461395,
+      "learning_rate": 6.209032956763247e-06,
+      "loss": 1.4929,
+      "step": 849
+    },
+    {
+      "epoch": 2.3255813953488373,
+      "grad_norm": 0.346003919839859,
+      "learning_rate": 6.160834996766815e-06,
+      "loss": 1.3601,
+      "step": 850
+    },
+    {
+      "epoch": 2.3283173734610125,
+      "grad_norm": 0.3200609087944031,
+      "learning_rate": 6.112798532722438e-06,
+      "loss": 1.6096,
+      "step": 851
+    },
+    {
+      "epoch": 2.3310533515731873,
+      "grad_norm": 0.31736400723457336,
+      "learning_rate": 6.064923976415626e-06,
+      "loss": 1.4462,
+      "step": 852
+    },
+    {
+      "epoch": 2.3337893296853625,
+      "grad_norm": 0.2982243299484253,
+      "learning_rate": 6.0172117382438944e-06,
+      "loss": 1.6742,
+      "step": 853
+    },
+    {
+      "epoch": 2.3365253077975376,
+      "grad_norm": 0.36678165197372437,
+      "learning_rate": 5.969662227213352e-06,
+      "loss": 1.429,
+      "step": 854
+    },
+    {
+      "epoch": 2.339261285909713,
+      "grad_norm": 0.29900214076042175,
+      "learning_rate": 5.922275850935136e-06,
+      "loss": 1.5513,
+      "step": 855
+    },
+    {
+      "epoch": 2.341997264021888,
+      "grad_norm": 0.3283601403236389,
+      "learning_rate": 5.875053015621943e-06,
+      "loss": 1.7077,
+      "step": 856
+    },
+    {
+      "epoch": 2.3447332421340628,
+      "grad_norm": 0.33298107981681824,
+      "learning_rate": 5.827994126084532e-06,
+      "loss": 1.5916,
+      "step": 857
+    },
+    {
+      "epoch": 2.347469220246238,
+      "grad_norm": 0.3546065390110016,
+      "learning_rate": 5.781099585728267e-06,
+      "loss": 1.4382,
+      "step": 858
+    },
+    {
+      "epoch": 2.350205198358413,
+      "grad_norm": 0.30142465233802795,
+      "learning_rate": 5.734369796549652e-06,
+      "loss": 1.4578,
+      "step": 859
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 0.3024732768535614,
+      "learning_rate": 5.687805159132881e-06,
+      "loss": 1.4102,
+      "step": 860
+    },
+    {
+      "epoch": 2.3556771545827635,
+      "grad_norm": 0.3300219774246216,
+      "learning_rate": 5.641406072646418e-06,
+      "loss": 1.5036,
+      "step": 861
+    },
+    {
+      "epoch": 2.3584131326949382,
+      "grad_norm": 0.37552839517593384,
+      "learning_rate": 5.595172934839546e-06,
+      "loss": 1.4401,
+      "step": 862
+    },
+    {
+      "epoch": 2.3611491108071134,
+      "grad_norm": 0.3124082386493683,
+      "learning_rate": 5.549106142039018e-06,
+      "loss": 1.3749,
+      "step": 863
+    },
+    {
+      "epoch": 2.3638850889192886,
+      "grad_norm": 0.34462398290634155,
+      "learning_rate": 5.503206089145568e-06,
+      "loss": 1.578,
+      "step": 864
+    },
+    {
+      "epoch": 2.366621067031464,
+      "grad_norm": 0.3301493227481842,
+      "learning_rate": 5.4574731696306146e-06,
+      "loss": 1.7064,
+      "step": 865
+    },
+    {
+      "epoch": 2.369357045143639,
+      "grad_norm": 0.33227112889289856,
+      "learning_rate": 5.411907775532832e-06,
+      "loss": 1.4069,
+      "step": 866
+    },
+    {
+      "epoch": 2.3720930232558137,
+      "grad_norm": 0.3299727439880371,
+      "learning_rate": 5.366510297454816e-06,
+      "loss": 1.4481,
+      "step": 867
+    },
+    {
+      "epoch": 2.374829001367989,
+      "grad_norm": 0.3437407314777374,
+      "learning_rate": 5.321281124559727e-06,
+      "loss": 1.3406,
+      "step": 868
+    },
+    {
+      "epoch": 2.377564979480164,
+      "grad_norm": 0.3354208469390869,
+      "learning_rate": 5.276220644567948e-06,
+      "loss": 1.3835,
+      "step": 869
+    },
+    {
+      "epoch": 2.3803009575923393,
+      "grad_norm": 0.29964345693588257,
+      "learning_rate": 5.231329243753772e-06,
+      "loss": 1.3945,
+      "step": 870
+    },
+    {
+      "epoch": 2.3830369357045145,
+      "grad_norm": 0.3642790615558624,
+      "learning_rate": 5.186607306942085e-06,
+      "loss": 1.3748,
+      "step": 871
+    },
+    {
+      "epoch": 2.3857729138166897,
+      "grad_norm": 0.29215675592422485,
+      "learning_rate": 5.142055217505074e-06,
+      "loss": 1.5209,
+      "step": 872
+    },
+    {
+      "epoch": 2.3885088919288644,
+      "grad_norm": 0.35326606035232544,
+      "learning_rate": 5.097673357358907e-06,
+      "loss": 1.5554,
+      "step": 873
+    },
+    {
+      "epoch": 2.3912448700410396,
+      "grad_norm": 0.3523205816745758,
+      "learning_rate": 5.053462106960532e-06,
+      "loss": 1.5203,
+      "step": 874
+    },
+    {
+      "epoch": 2.3939808481532148,
+      "grad_norm": 0.3213190734386444,
+      "learning_rate": 5.00942184530433e-06,
+      "loss": 1.4171,
+      "step": 875
+    },
+    {
+      "epoch": 2.39671682626539,
+      "grad_norm": 0.3184296488761902,
+      "learning_rate": 4.96555294991892e-06,
+      "loss": 1.3045,
+      "step": 876
+    },
+    {
+      "epoch": 2.399452804377565,
+      "grad_norm": 0.3270394802093506,
+      "learning_rate": 4.921855796863933e-06,
+      "loss": 1.5365,
+      "step": 877
+    },
+    {
+      "epoch": 2.4021887824897403,
+      "grad_norm": 0.36887189745903015,
+      "learning_rate": 4.878330760726713e-06,
+      "loss": 1.2823,
+      "step": 878
+    },
+    {
+      "epoch": 2.404924760601915,
+      "grad_norm": 0.3077680766582489,
+      "learning_rate": 4.834978214619215e-06,
+      "loss": 1.5222,
+      "step": 879
+    },
+    {
+      "epoch": 2.4076607387140903,
+      "grad_norm": 0.3177037835121155,
+      "learning_rate": 4.791798530174699e-06,
+      "loss": 1.4489,
+      "step": 880
+    },
+    {
+      "epoch": 2.4103967168262654,
+      "grad_norm": 0.3479926586151123,
+      "learning_rate": 4.748792077544623e-06,
+      "loss": 1.5376,
+      "step": 881
+    },
+    {
+      "epoch": 2.4131326949384406,
+      "grad_norm": 0.32515963912010193,
+      "learning_rate": 4.7059592253954315e-06,
+      "loss": 1.5803,
+      "step": 882
+    },
+    {
+      "epoch": 2.415868673050616,
+      "grad_norm": 0.33568742871284485,
+      "learning_rate": 4.66330034090541e-06,
+      "loss": 1.5066,
+      "step": 883
+    },
+    {
+      "epoch": 2.4186046511627906,
+      "grad_norm": 0.328945130109787,
+      "learning_rate": 4.620815789761526e-06,
+      "loss": 1.4279,
+      "step": 884
+    },
+    {
+      "epoch": 2.4213406292749657,
+      "grad_norm": 0.3277706503868103,
+      "learning_rate": 4.578505936156302e-06,
+      "loss": 1.5832,
+      "step": 885
+    },
+    {
+      "epoch": 2.424076607387141,
+      "grad_norm": 0.35384321212768555,
+      "learning_rate": 4.5363711427847015e-06,
+      "loss": 1.4504,
+      "step": 886
+    },
+    {
+      "epoch": 2.426812585499316,
+      "grad_norm": 0.3523906469345093,
+      "learning_rate": 4.494411770840978e-06,
+      "loss": 1.5453,
+      "step": 887
+    },
+    {
+      "epoch": 2.4295485636114913,
+      "grad_norm": 0.30951249599456787,
+      "learning_rate": 4.452628180015664e-06,
+      "loss": 1.5488,
+      "step": 888
+    },
+    {
+      "epoch": 2.432284541723666,
+      "grad_norm": 0.3258694112300873,
+      "learning_rate": 4.411020728492374e-06,
+      "loss": 1.4878,
+      "step": 889
+    },
+    {
+      "epoch": 2.4350205198358412,
+      "grad_norm": 0.33756595849990845,
+      "learning_rate": 4.3695897729448485e-06,
+      "loss": 1.4415,
+      "step": 890
+    },
+    {
+      "epoch": 2.4377564979480164,
+      "grad_norm": 0.3134118318557739,
+      "learning_rate": 4.328335668533806e-06,
+      "loss": 1.3717,
+      "step": 891
+    },
+    {
+      "epoch": 2.4404924760601916,
+      "grad_norm": 0.2975947856903076,
+      "learning_rate": 4.2872587689039484e-06,
+      "loss": 1.5299,
+      "step": 892
+    },
+    {
+      "epoch": 2.443228454172367,
+      "grad_norm": 0.3447892963886261,
+      "learning_rate": 4.246359426180918e-06,
+      "loss": 1.4769,
+      "step": 893
+    },
+    {
+      "epoch": 2.4459644322845415,
+      "grad_norm": 0.32196173071861267,
+      "learning_rate": 4.20563799096827e-06,
+      "loss": 1.4507,
+      "step": 894
+    },
+    {
+      "epoch": 2.4487004103967167,
+      "grad_norm": 0.30560824275016785,
+      "learning_rate": 4.165094812344478e-06,
+      "loss": 1.6074,
+      "step": 895
+    },
+    {
+      "epoch": 2.451436388508892,
+      "grad_norm": 0.3304429352283478,
+      "learning_rate": 4.124730237859939e-06,
+      "loss": 1.5368,
+      "step": 896
+    },
+    {
+      "epoch": 2.454172366621067,
+      "grad_norm": 0.33104217052459717,
+      "learning_rate": 4.0845446135339945e-06,
+      "loss": 1.4162,
+      "step": 897
+    },
+    {
+      "epoch": 2.4569083447332423,
+      "grad_norm": 0.3179687261581421,
+      "learning_rate": 4.0445382838519365e-06,
+      "loss": 1.6278,
+      "step": 898
+    },
+    {
+      "epoch": 2.459644322845417,
+      "grad_norm": 0.3108776807785034,
+      "learning_rate": 4.004711591762133e-06,
+      "loss": 1.6218,
+      "step": 899
+    },
+    {
+      "epoch": 2.462380300957592,
+      "grad_norm": 0.3062944710254669,
+      "learning_rate": 3.965064878672983e-06,
+      "loss": 1.4816,
+      "step": 900
+    },
+    {
+      "epoch": 2.4651162790697674,
+      "grad_norm": 0.3266476094722748,
+      "learning_rate": 3.925598484450066e-06,
+      "loss": 1.846,
+      "step": 901
+    },
+    {
+      "epoch": 2.4678522571819426,
+      "grad_norm": 0.344200998544693,
+      "learning_rate": 3.886312747413204e-06,
+      "loss": 1.3894,
+      "step": 902
+    },
+    {
+      "epoch": 2.4705882352941178,
+      "grad_norm": 0.315949410200119,
+      "learning_rate": 3.847208004333561e-06,
+      "loss": 1.4174,
+      "step": 903
+    },
+    {
+      "epoch": 2.473324213406293,
+      "grad_norm": 0.3364209234714508,
+      "learning_rate": 3.8082845904307525e-06,
+      "loss": 1.4093,
+      "step": 904
+    },
+    {
+      "epoch": 2.4760601915184677,
+      "grad_norm": 0.3172418475151062,
+      "learning_rate": 3.7695428393699854e-06,
+      "loss": 1.3446,
+      "step": 905
+    },
+    {
+      "epoch": 2.478796169630643,
+      "grad_norm": 0.3134661316871643,
+      "learning_rate": 3.730983083259179e-06,
+      "loss": 1.507,
+      "step": 906
+    },
+    {
+      "epoch": 2.481532147742818,
+      "grad_norm": 0.3210635185241699,
+      "learning_rate": 3.6926056526461334e-06,
+      "loss": 1.3344,
+      "step": 907
+    },
+    {
+      "epoch": 2.4842681258549932,
+      "grad_norm": 0.32122674584388733,
+      "learning_rate": 3.6544108765156933e-06,
+      "loss": 1.5514,
+      "step": 908
+    },
+    {
+      "epoch": 2.4870041039671684,
+      "grad_norm": 0.36177971959114075,
+      "learning_rate": 3.6163990822869088e-06,
+      "loss": 1.5981,
+      "step": 909
+    },
+    {
+      "epoch": 2.4897400820793436,
+      "grad_norm": 0.3207845389842987,
+      "learning_rate": 3.578570595810274e-06,
+      "loss": 1.5989,
+      "step": 910
+    },
+    {
+      "epoch": 2.4924760601915183,
+      "grad_norm": 0.32466524839401245,
+      "learning_rate": 3.540925741364873e-06,
+      "loss": 1.483,
+      "step": 911
+    },
+    {
+      "epoch": 2.4952120383036935,
+      "grad_norm": 0.31727364659309387,
+      "learning_rate": 3.5034648416556486e-06,
+      "loss": 1.5308,
+      "step": 912
+    },
+    {
+      "epoch": 2.4979480164158687,
+      "grad_norm": 0.3401485085487366,
+      "learning_rate": 3.4661882178106176e-06,
+      "loss": 1.6316,
+      "step": 913
+    },
+    {
+      "epoch": 2.500683994528044,
+      "grad_norm": 0.3355841338634491,
+      "learning_rate": 3.429096189378114e-06,
+      "loss": 1.5952,
+      "step": 914
+    },
+    {
+      "epoch": 2.503419972640219,
+      "grad_norm": 0.3085014522075653,
+      "learning_rate": 3.392189074324073e-06,
+      "loss": 1.6774,
+      "step": 915
+    },
+    {
+      "epoch": 2.506155950752394,
+      "grad_norm": 0.3420393466949463,
+      "learning_rate": 3.355467189029257e-06,
+      "loss": 1.4805,
+      "step": 916
+    },
+    {
+      "epoch": 2.508891928864569,
+      "grad_norm": 0.3142971098423004,
+      "learning_rate": 3.3189308482865917e-06,
+      "loss": 1.4958,
+      "step": 917
+    },
+    {
+      "epoch": 2.511627906976744,
+      "grad_norm": 0.37307271361351013,
+      "learning_rate": 3.2825803652984516e-06,
+      "loss": 1.4541,
+      "step": 918
+    },
+    {
+      "epoch": 2.5143638850889194,
+      "grad_norm": 0.33256828784942627,
+      "learning_rate": 3.2464160516739755e-06,
+      "loss": 1.4565,
+      "step": 919
+    },
+    {
+      "epoch": 2.5170998632010946,
+      "grad_norm": 0.29588747024536133,
+      "learning_rate": 3.210438217426365e-06,
+      "loss": 1.4852,
+      "step": 920
+    },
+    {
+      "epoch": 2.5198358413132693,
+      "grad_norm": 0.314485102891922,
+      "learning_rate": 3.1746471709702964e-06,
+      "loss": 1.3962,
+      "step": 921
+    },
+    {
+      "epoch": 2.5225718194254445,
+      "grad_norm": 0.30730655789375305,
+      "learning_rate": 3.1390432191192115e-06,
+      "loss": 1.5221,
+      "step": 922
+    },
+    {
+      "epoch": 2.5253077975376197,
+      "grad_norm": 0.34515947103500366,
+      "learning_rate": 3.1036266670827014e-06,
+      "loss": 1.4745,
+      "step": 923
+    },
+    {
+      "epoch": 2.528043775649795,
+      "grad_norm": 0.3266810476779938,
+      "learning_rate": 3.068397818463936e-06,
+      "loss": 1.4541,
+      "step": 924
+    },
+    {
+      "epoch": 2.53077975376197,
+      "grad_norm": 0.2986440062522888,
+      "learning_rate": 3.033356975256979e-06,
+      "loss": 1.5795,
+      "step": 925
+    },
+    {
+      "epoch": 2.533515731874145,
+      "grad_norm": 0.337341845035553,
+      "learning_rate": 2.9985044378442933e-06,
+      "loss": 1.5815,
+      "step": 926
+    },
+    {
+      "epoch": 2.53625170998632,
+      "grad_norm": 0.3632700741291046,
+      "learning_rate": 2.963840504994075e-06,
+      "loss": 1.54,
+      "step": 927
+    },
+    {
+      "epoch": 2.538987688098495,
+      "grad_norm": 0.3714698553085327,
+      "learning_rate": 2.9293654738577647e-06,
+      "loss": 1.5019,
+      "step": 928
+    },
+    {
+      "epoch": 2.5417236662106704,
+      "grad_norm": 0.3200991153717041,
+      "learning_rate": 2.8950796399674573e-06,
+      "loss": 1.4587,
+      "step": 929
+    },
+    {
+      "epoch": 2.5444596443228455,
+      "grad_norm": 0.3523624539375305,
+      "learning_rate": 2.860983297233388e-06,
+      "loss": 1.4111,
+      "step": 930
+    },
+    {
+      "epoch": 2.5471956224350203,
+      "grad_norm": 0.3558140993118286,
+      "learning_rate": 2.8270767379414087e-06,
+      "loss": 1.3241,
+      "step": 931
+    },
+    {
+      "epoch": 2.5499316005471955,
+      "grad_norm": 0.3104929029941559,
+      "learning_rate": 2.7933602527504738e-06,
+      "loss": 1.6253,
+      "step": 932
+    },
+    {
+      "epoch": 2.5526675786593707,
+      "grad_norm": 0.33675897121429443,
+      "learning_rate": 2.7598341306901643e-06,
+      "loss": 1.5365,
+      "step": 933
+    },
+    {
+      "epoch": 2.555403556771546,
+      "grad_norm": 0.39364495873451233,
+      "learning_rate": 2.726498659158183e-06,
+      "loss": 1.5522,
+      "step": 934
+    },
+    {
+      "epoch": 2.558139534883721,
+      "grad_norm": 0.31920477747917175,
+      "learning_rate": 2.693354123917943e-06,
+      "loss": 1.5633,
+      "step": 935
+    },
+    {
+      "epoch": 2.5608755129958958,
+      "grad_norm": 0.37531107664108276,
+      "learning_rate": 2.660400809096045e-06,
+      "loss": 1.3603,
+      "step": 936
+    },
+    {
+      "epoch": 2.5636114911080714,
+      "grad_norm": 0.31020471453666687,
+      "learning_rate": 2.6276389971799153e-06,
+      "loss": 1.3862,
+      "step": 937
+    },
+    {
+      "epoch": 2.566347469220246,
+      "grad_norm": 0.3524358570575714,
+      "learning_rate": 2.595068969015327e-06,
+      "loss": 1.5923,
+      "step": 938
+    },
+    {
+      "epoch": 2.5690834473324213,
+      "grad_norm": 0.3156653940677643,
+      "learning_rate": 2.562691003804024e-06,
+      "loss": 1.6537,
+      "step": 939
+    },
+    {
+      "epoch": 2.5718194254445965,
+      "grad_norm": 0.3592538833618164,
+      "learning_rate": 2.5305053791013194e-06,
+      "loss": 1.3821,
+      "step": 940
+    },
+    {
+      "epoch": 2.5745554035567717,
+      "grad_norm": 0.3512822091579437,
+      "learning_rate": 2.498512370813716e-06,
+      "loss": 1.5147,
+      "step": 941
+    },
+    {
+      "epoch": 2.577291381668947,
+      "grad_norm": 0.31111329793930054,
+      "learning_rate": 2.466712253196535e-06,
+      "loss": 1.443,
+      "step": 942
+    },
+    {
+      "epoch": 2.5800273597811216,
+      "grad_norm": 0.31163808703422546,
+      "learning_rate": 2.4351052988515783e-06,
+      "loss": 1.5938,
+      "step": 943
+    },
+    {
+      "epoch": 2.582763337893297,
+      "grad_norm": 0.34021854400634766,
+      "learning_rate": 2.403691778724787e-06,
+      "loss": 1.4879,
+      "step": 944
+    },
+    {
+      "epoch": 2.585499316005472,
+      "grad_norm": 0.30090874433517456,
+      "learning_rate": 2.3724719621038923e-06,
+      "loss": 1.5188,
+      "step": 945
+    },
+    {
+      "epoch": 2.588235294117647,
+      "grad_norm": 0.2972038984298706,
+      "learning_rate": 2.341446116616172e-06,
+      "loss": 1.4422,
+      "step": 946
+    },
+    {
+      "epoch": 2.5909712722298224,
+      "grad_norm": 0.360759973526001,
+      "learning_rate": 2.310614508226078e-06,
+      "loss": 1.4662,
+      "step": 947
+    },
+    {
+      "epoch": 2.593707250341997,
+      "grad_norm": 0.33597201108932495,
+      "learning_rate": 2.2799774012330076e-06,
+      "loss": 1.4008,
+      "step": 948
+    },
+    {
+      "epoch": 2.5964432284541723,
+      "grad_norm": 0.3385523855686188,
+      "learning_rate": 2.2495350582690254e-06,
+      "loss": 1.4652,
+      "step": 949
+    },
+    {
+      "epoch": 2.5991792065663475,
+      "grad_norm": 0.3732840418815613,
+      "learning_rate": 2.219287740296605e-06,
+      "loss": 1.3399,
+      "step": 950
+    },
+    {
+      "epoch": 2.6019151846785227,
+      "grad_norm": 0.37210366129875183,
+      "learning_rate": 2.1892357066064128e-06,
+      "loss": 1.3226,
+      "step": 951
+    },
+    {
+      "epoch": 2.604651162790698,
+      "grad_norm": 0.33239394426345825,
+      "learning_rate": 2.1593792148150437e-06,
+      "loss": 1.3563,
+      "step": 952
+    },
+    {
+      "epoch": 2.6073871409028726,
+      "grad_norm": 0.3127799332141876,
+      "learning_rate": 2.1297185208628585e-06,
+      "loss": 1.5005,
+      "step": 953
+    },
+    {
+      "epoch": 2.610123119015048,
+      "grad_norm": 0.3832361698150635,
+      "learning_rate": 2.10025387901176e-06,
+      "loss": 1.3859,
+      "step": 954
+    },
+    {
+      "epoch": 2.612859097127223,
+      "grad_norm": 0.3399605453014374,
+      "learning_rate": 2.0709855418430317e-06,
+      "loss": 1.559,
+      "step": 955
+    },
+    {
+      "epoch": 2.615595075239398,
+      "grad_norm": 0.36757224798202515,
+      "learning_rate": 2.0419137602551516e-06,
+      "loss": 1.475,
+      "step": 956
+    },
+    {
+      "epoch": 2.6183310533515733,
+      "grad_norm": 0.3125869929790497,
+      "learning_rate": 2.013038783461674e-06,
+      "loss": 1.4586,
+      "step": 957
+    },
+    {
+      "epoch": 2.621067031463748,
+      "grad_norm": 0.31474483013153076,
+      "learning_rate": 1.9843608589890513e-06,
+      "loss": 1.4757,
+      "step": 958
+    },
+    {
+      "epoch": 2.6238030095759233,
+      "grad_norm": 0.3160763382911682,
+      "learning_rate": 1.9558802326745424e-06,
+      "loss": 1.5655,
+      "step": 959
+    },
+    {
+      "epoch": 2.6265389876880985,
+      "grad_norm": 0.41808021068573,
+      "learning_rate": 1.927597148664112e-06,
+      "loss": 1.4522,
+      "step": 960
+    },
+    {
+      "epoch": 2.6292749658002736,
+      "grad_norm": 0.38478416204452515,
+      "learning_rate": 1.8995118494102903e-06,
+      "loss": 1.5056,
+      "step": 961
+    },
+    {
+      "epoch": 2.632010943912449,
+      "grad_norm": 0.33798423409461975,
+      "learning_rate": 1.8716245756701694e-06,
+      "loss": 1.5478,
+      "step": 962
+    },
+    {
+      "epoch": 2.6347469220246236,
+      "grad_norm": 0.34396031498908997,
+      "learning_rate": 1.843935566503252e-06,
+      "loss": 1.438,
+      "step": 963
+    },
+    {
+      "epoch": 2.6374829001367988,
+      "grad_norm": 0.317111998796463,
+      "learning_rate": 1.816445059269481e-06,
+      "loss": 1.4886,
+      "step": 964
+    },
+    {
+      "epoch": 2.640218878248974,
+      "grad_norm": 0.314860075712204,
+      "learning_rate": 1.7891532896271547e-06,
+      "loss": 1.5923,
+      "step": 965
+    },
+    {
+      "epoch": 2.642954856361149,
+      "grad_norm": 0.29904037714004517,
+      "learning_rate": 1.7620604915309257e-06,
+      "loss": 1.3853,
+      "step": 966
+    },
+    {
+      "epoch": 2.6456908344733243,
+      "grad_norm": 0.34991443157196045,
+      "learning_rate": 1.7351668972297924e-06,
+      "loss": 1.5588,
+      "step": 967
+    },
+    {
+      "epoch": 2.648426812585499,
+      "grad_norm": 0.3276471197605133,
+      "learning_rate": 1.7084727372651155e-06,
+      "loss": 1.4885,
+      "step": 968
+    },
+    {
+      "epoch": 2.6511627906976747,
+      "grad_norm": 0.3197418749332428,
+      "learning_rate": 1.6819782404686263e-06,
+      "loss": 1.5063,
+      "step": 969
+    },
+    {
+      "epoch": 2.6538987688098494,
+      "grad_norm": 0.3408433198928833,
+      "learning_rate": 1.655683633960467e-06,
+      "loss": 1.4564,
+      "step": 970
+    },
+    {
+      "epoch": 2.6566347469220246,
+      "grad_norm": 0.33302435278892517,
+      "learning_rate": 1.62958914314727e-06,
+      "loss": 1.3688,
+      "step": 971
+    },
+    {
+      "epoch": 2.6593707250342,
+      "grad_norm": 0.34161651134490967,
+      "learning_rate": 1.6036949917201783e-06,
+      "loss": 1.5104,
+      "step": 972
+    },
+    {
+      "epoch": 2.662106703146375,
+      "grad_norm": 0.33324992656707764,
+      "learning_rate": 1.578001401652987e-06,
+      "loss": 1.4552,
+      "step": 973
+    },
+    {
+      "epoch": 2.66484268125855,
+      "grad_norm": 0.30797359347343445,
+      "learning_rate": 1.5525085932001832e-06,
+      "loss": 1.5157,
+      "step": 974
+    },
+    {
+      "epoch": 2.667578659370725,
+      "grad_norm": 0.3368433117866516,
+      "learning_rate": 1.5272167848950913e-06,
+      "loss": 1.435,
+      "step": 975
+    },
+    {
+      "epoch": 2.6703146374829,
+      "grad_norm": 0.3299857974052429,
+      "learning_rate": 1.5021261935479924e-06,
+      "loss": 1.5031,
+      "step": 976
+    },
+    {
+      "epoch": 2.6730506155950753,
+      "grad_norm": 0.29794323444366455,
+      "learning_rate": 1.4772370342442694e-06,
+      "loss": 1.4679,
+      "step": 977
+    },
+    {
+      "epoch": 2.6757865937072505,
+      "grad_norm": 0.31903737783432007,
+      "learning_rate": 1.4525495203425498e-06,
+      "loss": 1.6308,
+      "step": 978
+    },
+    {
+      "epoch": 2.6785225718194257,
+      "grad_norm": 0.3376205861568451,
+      "learning_rate": 1.428063863472895e-06,
+      "loss": 1.5566,
+      "step": 979
+    },
+    {
+      "epoch": 2.6812585499316004,
+      "grad_norm": 0.3032618761062622,
+      "learning_rate": 1.4037802735349743e-06,
+      "loss": 1.4632,
+      "step": 980
+    },
+    {
+      "epoch": 2.6839945280437756,
+      "grad_norm": 0.33116862177848816,
+      "learning_rate": 1.3796989586962544e-06,
+      "loss": 1.5811,
+      "step": 981
+    },
+    {
+      "epoch": 2.6867305061559508,
+      "grad_norm": 0.34618982672691345,
+      "learning_rate": 1.3558201253902531e-06,
+      "loss": 1.3734,
+      "step": 982
+    },
+    {
+      "epoch": 2.689466484268126,
+      "grad_norm": 0.3325488567352295,
+      "learning_rate": 1.3321439783147233e-06,
+      "loss": 1.4946,
+      "step": 983
+    },
+    {
+      "epoch": 2.692202462380301,
+      "grad_norm": 0.32722416520118713,
+      "learning_rate": 1.3086707204299414e-06,
+      "loss": 1.5425,
+      "step": 984
+    },
+    {
+      "epoch": 2.694938440492476,
+      "grad_norm": 0.3376059830188751,
+      "learning_rate": 1.2854005529569224e-06,
+      "loss": 1.4031,
+      "step": 985
+    },
+    {
+      "epoch": 2.697674418604651,
+      "grad_norm": 0.3256185054779053,
+      "learning_rate": 1.2623336753757348e-06,
+      "loss": 1.493,
+      "step": 986
+    },
+    {
+      "epoch": 2.7004103967168263,
+      "grad_norm": 0.3338906168937683,
+      "learning_rate": 1.2394702854237744e-06,
+      "loss": 1.5289,
+      "step": 987
+    },
+    {
+      "epoch": 2.7031463748290014,
+      "grad_norm": 0.31372472643852234,
+      "learning_rate": 1.2168105790940603e-06,
+      "loss": 1.5584,
+      "step": 988
+    },
+    {
+      "epoch": 2.7058823529411766,
+      "grad_norm": 0.3929591774940491,
+      "learning_rate": 1.1943547506335723e-06,
+      "loss": 1.482,
+      "step": 989
+    },
+    {
+      "epoch": 2.7086183310533514,
+      "grad_norm": 0.3206530809402466,
+      "learning_rate": 1.1721029925415738e-06,
+      "loss": 1.5406,
+      "step": 990
+    },
+    {
+      "epoch": 2.7113543091655266,
+      "grad_norm": 0.32885223627090454,
+      "learning_rate": 1.15005549556797e-06,
+      "loss": 1.3643,
+      "step": 991
+    },
+    {
+      "epoch": 2.7140902872777017,
+      "grad_norm": 0.3618469536304474,
+      "learning_rate": 1.1282124487116518e-06,
+      "loss": 1.5104,
+      "step": 992
+    },
+    {
+      "epoch": 2.716826265389877,
+      "grad_norm": 0.31617462635040283,
+      "learning_rate": 1.106574039218919e-06,
+      "loss": 1.4661,
+      "step": 993
+    },
+    {
+      "epoch": 2.719562243502052,
+      "grad_norm": 0.31909239292144775,
+      "learning_rate": 1.0851404525818204e-06,
+      "loss": 1.5708,
+      "step": 994
+    },
+    {
+      "epoch": 2.722298221614227,
+      "grad_norm": 0.3482041358947754,
+      "learning_rate": 1.0639118725366049e-06,
+      "loss": 1.4782,
+      "step": 995
+    },
+    {
+      "epoch": 2.725034199726402,
+      "grad_norm": 0.3372204601764679,
+      "learning_rate": 1.0428884810621403e-06,
+      "loss": 1.4833,
+      "step": 996
+    },
+    {
+      "epoch": 2.7277701778385772,
+      "grad_norm": 0.32383978366851807,
+      "learning_rate": 1.0220704583783242e-06,
+      "loss": 1.5989,
+      "step": 997
+    },
+    {
+      "epoch": 2.7305061559507524,
+      "grad_norm": 0.3657885491847992,
+      "learning_rate": 1.0014579829445842e-06,
+      "loss": 1.6592,
+      "step": 998
+    },
+    {
+      "epoch": 2.7332421340629276,
+      "grad_norm": 0.3399549722671509,
+      "learning_rate": 9.810512314583075e-07,
+      "loss": 1.3902,
+      "step": 999
+    },
+    {
+      "epoch": 2.7359781121751023,
+      "grad_norm": 0.3272568881511688,
+      "learning_rate": 9.60850378853348e-07,
+      "loss": 1.3684,
+      "step": 1000
+    },
+    {
+      "epoch": 2.7359781121751023,
+      "eval_loss": 1.5569868087768555,
+      "eval_runtime": 13.8259,
+      "eval_samples_per_second": 11.138,
+      "eval_steps_per_second": 11.138,
+      "step": 1000
+    },
+    {
+      "epoch": 2.738714090287278,
+      "grad_norm": 0.3006269037723541,
+      "learning_rate": 9.408555982985262e-07,
+      "loss": 1.4081,
+      "step": 1001
+    },
+    {
+      "epoch": 2.7414500683994527,
+      "grad_norm": 0.32424142956733704,
+      "learning_rate": 9.210670611961375e-07,
+      "loss": 1.47,
+      "step": 1002
+    },
+    {
+      "epoch": 2.744186046511628,
+      "grad_norm": 0.3788749873638153,
+      "learning_rate": 9.014849371804829e-07,
+      "loss": 1.4683,
+      "step": 1003
+    },
+    {
+      "epoch": 2.746922024623803,
+      "grad_norm": 0.3551234006881714,
+      "learning_rate": 8.821093941164233e-07,
+      "loss": 1.6791,
+      "step": 1004
+    },
+    {
+      "epoch": 2.7496580027359783,
+      "grad_norm": 0.38576632738113403,
+      "learning_rate": 8.629405980979332e-07,
+      "loss": 1.49,
+      "step": 1005
+    },
+    {
+      "epoch": 2.7523939808481535,
+      "grad_norm": 0.3681055009365082,
+      "learning_rate": 8.439787134466692e-07,
+      "loss": 1.4691,
+      "step": 1006
+    },
+    {
+      "epoch": 2.755129958960328,
+      "grad_norm": 0.3119199872016907,
+      "learning_rate": 8.252239027105924e-07,
+      "loss": 1.5351,
+      "step": 1007
+    },
+    {
+      "epoch": 2.7578659370725034,
+      "grad_norm": 0.31998804211616516,
+      "learning_rate": 8.066763266625282e-07,
+      "loss": 1.468,
+      "step": 1008
+    },
+    {
+      "epoch": 2.7606019151846786,
+      "grad_norm": 0.3693805932998657,
+      "learning_rate": 7.883361442988374e-07,
+      "loss": 1.5278,
+      "step": 1009
+    },
+    {
+      "epoch": 2.7633378932968538,
+      "grad_norm": 0.3538669943809509,
+      "learning_rate": 7.702035128380053e-07,
+      "loss": 1.5411,
+      "step": 1010
+    },
+    {
+      "epoch": 2.766073871409029,
+      "grad_norm": 0.3684946596622467,
+      "learning_rate": 7.522785877193378e-07,
+      "loss": 1.4696,
+      "step": 1011
+    },
+    {
+      "epoch": 2.7688098495212037,
+      "grad_norm": 0.28718748688697815,
+      "learning_rate": 7.345615226016011e-07,
+      "loss": 1.5301,
+      "step": 1012
+    },
+    {
+      "epoch": 2.771545827633379,
+      "grad_norm": 0.33110928535461426,
+      "learning_rate": 7.170524693617064e-07,
+      "loss": 1.6021,
+      "step": 1013
+    },
+    {
+      "epoch": 2.774281805745554,
+      "grad_norm": 0.31040212512016296,
+      "learning_rate": 6.997515780934244e-07,
+      "loss": 1.4569,
+      "step": 1014
+    },
+    {
+      "epoch": 2.7770177838577292,
+      "grad_norm": 0.3069436848163605,
+      "learning_rate": 6.826589971060837e-07,
+      "loss": 1.4963,
+      "step": 1015
+    },
+    {
+      "epoch": 2.7797537619699044,
+      "grad_norm": 0.3085753321647644,
+      "learning_rate": 6.657748729233026e-07,
+      "loss": 1.6278,
+      "step": 1016
+    },
+    {
+      "epoch": 2.782489740082079,
+      "grad_norm": 0.31628352403640747,
+      "learning_rate": 6.490993502817289e-07,
+      "loss": 1.4166,
+      "step": 1017
+    },
+    {
+      "epoch": 2.7852257181942544,
+      "grad_norm": 0.32792040705680847,
+      "learning_rate": 6.32632572129821e-07,
+      "loss": 1.4757,
+      "step": 1018
+    },
+    {
+      "epoch": 2.7879616963064295,
+      "grad_norm": 0.34112995862960815,
+      "learning_rate": 6.163746796265885e-07,
+      "loss": 1.4505,
+      "step": 1019
+    },
+    {
+      "epoch": 2.7906976744186047,
+      "grad_norm": 0.3526972830295563,
+      "learning_rate": 6.003258121404148e-07,
+      "loss": 1.4417,
+      "step": 1020
+    },
+    {
+      "epoch": 2.79343365253078,
+      "grad_norm": 0.3196633458137512,
+      "learning_rate": 5.844861072478336e-07,
+      "loss": 1.5024,
+      "step": 1021
+    },
+    {
+      "epoch": 2.7961696306429547,
+      "grad_norm": 0.3368687033653259,
+      "learning_rate": 5.688557007323708e-07,
+      "loss": 1.6036,
+      "step": 1022
+    },
+    {
+      "epoch": 2.79890560875513,
+      "grad_norm": 0.33089619874954224,
+      "learning_rate": 5.534347265833711e-07,
+      "loss": 1.5716,
+      "step": 1023
+    },
+    {
+      "epoch": 2.801641586867305,
+      "grad_norm": 0.35482943058013916,
+      "learning_rate": 5.382233169948541e-07,
+      "loss": 1.5263,
+      "step": 1024
+    },
+    {
+      "epoch": 2.80437756497948,
+      "grad_norm": 0.3278489112854004,
+      "learning_rate": 5.232216023643711e-07,
+      "loss": 1.4273,
+      "step": 1025
+    },
+    {
+      "epoch": 2.8071135430916554,
+      "grad_norm": 0.3070161044597626,
+      "learning_rate": 5.084297112919051e-07,
+      "loss": 1.7237,
+      "step": 1026
+    },
+    {
+      "epoch": 2.80984952120383,
+      "grad_norm": 0.3581262528896332,
+      "learning_rate": 4.938477705787508e-07,
+      "loss": 1.4228,
+      "step": 1027
+    },
+    {
+      "epoch": 2.8125854993160053,
+      "grad_norm": 0.3500869572162628,
+      "learning_rate": 4.794759052264259e-07,
+      "loss": 1.6744,
+      "step": 1028
+    },
+    {
+      "epoch": 2.8153214774281805,
+      "grad_norm": 0.3327549397945404,
+      "learning_rate": 4.653142384356324e-07,
+      "loss": 1.3299,
+      "step": 1029
+    },
+    {
+      "epoch": 2.8180574555403557,
+      "grad_norm": 0.32055315375328064,
+      "learning_rate": 4.513628916051532e-07,
+      "loss": 1.5822,
+      "step": 1030
+    },
+    {
+      "epoch": 2.820793433652531,
+      "grad_norm": 0.35483717918395996,
+      "learning_rate": 4.376219843308438e-07,
+      "loss": 1.4478,
+      "step": 1031
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 0.3212404251098633,
+      "learning_rate": 4.2409163440459697e-07,
+      "loss": 1.6226,
+      "step": 1032
+    },
+    {
+      "epoch": 2.8262653898768813,
+      "grad_norm": 0.3329022526741028,
+      "learning_rate": 4.107719578133301e-07,
+      "loss": 1.5192,
+      "step": 1033
+    },
+    {
+      "epoch": 2.829001367989056,
+      "grad_norm": 0.3651573061943054,
+      "learning_rate": 3.9766306873801064e-07,
+      "loss": 1.5169,
+      "step": 1034
+    },
+    {
+      "epoch": 2.831737346101231,
+      "grad_norm": 0.3816221356391907,
+      "learning_rate": 3.8476507955263475e-07,
+      "loss": 1.39,
+      "step": 1035
+    },
+    {
+      "epoch": 2.8344733242134064,
+      "grad_norm": 0.35796451568603516,
+      "learning_rate": 3.720781008233143e-07,
+      "loss": 1.4573,
+      "step": 1036
+    },
+    {
+      "epoch": 2.8372093023255816,
+      "grad_norm": 0.3537936210632324,
+      "learning_rate": 3.5960224130728857e-07,
+      "loss": 1.5344,
+      "step": 1037
+    },
+    {
+      "epoch": 2.8399452804377567,
+      "grad_norm": 0.35659074783325195,
+      "learning_rate": 3.473376079520224e-07,
+      "loss": 1.3814,
+      "step": 1038
+    },
+    {
+      "epoch": 2.8426812585499315,
+      "grad_norm": 0.31420376896858215,
+      "learning_rate": 3.352843058942623e-07,
+      "loss": 1.4956,
+      "step": 1039
+    },
+    {
+      "epoch": 2.8454172366621067,
+      "grad_norm": 0.31859591603279114,
+      "learning_rate": 3.2344243845915945e-07,
+      "loss": 1.6171,
+      "step": 1040
+    },
+    {
+      "epoch": 2.848153214774282,
+      "grad_norm": 0.30240893363952637,
+      "learning_rate": 3.118121071593677e-07,
+      "loss": 1.4719,
+      "step": 1041
+    },
+    {
+      "epoch": 2.850889192886457,
+      "grad_norm": 0.3180094361305237,
+      "learning_rate": 3.003934116941776e-07,
+      "loss": 1.4596,
+      "step": 1042
+    },
+    {
+      "epoch": 2.853625170998632,
+      "grad_norm": 0.3070809841156006,
+      "learning_rate": 2.891864499486724e-07,
+      "loss": 1.5858,
+      "step": 1043
+    },
+    {
+      "epoch": 2.856361149110807,
+      "grad_norm": 0.3325824737548828,
+      "learning_rate": 2.7819131799285967e-07,
+      "loss": 1.5769,
+      "step": 1044
+    },
+    {
+      "epoch": 2.859097127222982,
+      "grad_norm": 0.3298075199127197,
+      "learning_rate": 2.6740811008089393e-07,
+      "loss": 1.5655,
+      "step": 1045
+    },
+    {
+      "epoch": 2.8618331053351573,
+      "grad_norm": 0.35839757323265076,
+      "learning_rate": 2.568369186502162e-07,
+      "loss": 1.3463,
+      "step": 1046
+    },
+    {
+      "epoch": 2.8645690834473325,
+      "grad_norm": 0.3779064118862152,
+      "learning_rate": 2.464778343208074e-07,
+      "loss": 1.4612,
+      "step": 1047
+    },
+    {
+      "epoch": 2.8673050615595077,
+      "grad_norm": 0.32933345437049866,
+      "learning_rate": 2.36330945894378e-07,
+      "loss": 1.4131,
+      "step": 1048
+    },
+    {
+      "epoch": 2.8700410396716824,
+      "grad_norm": 0.3202325999736786,
+      "learning_rate": 2.2639634035363234e-07,
+      "loss": 1.5189,
+      "step": 1049
+    },
+    {
+      "epoch": 2.8727770177838576,
+      "grad_norm": 0.3435043394565582,
+      "learning_rate": 2.1667410286149992e-07,
+      "loss": 1.4026,
+      "step": 1050
+    },
+    {
+      "epoch": 2.875512995896033,
+      "grad_norm": 0.33565640449523926,
+      "learning_rate": 2.071643167604248e-07,
+      "loss": 1.4982,
+      "step": 1051
+    },
+    {
+      "epoch": 2.878248974008208,
+      "grad_norm": 0.31919005513191223,
+      "learning_rate": 1.9786706357163842e-07,
+      "loss": 1.6241,
+      "step": 1052
+    },
+    {
+      "epoch": 2.880984952120383,
+      "grad_norm": 0.33048737049102783,
+      "learning_rate": 1.8878242299446848e-07,
+      "loss": 1.586,
+      "step": 1053
+    },
+    {
+      "epoch": 2.883720930232558,
+      "grad_norm": 0.31295284628868103,
+      "learning_rate": 1.7991047290565056e-07,
+      "loss": 1.5296,
+      "step": 1054
+    },
+    {
+      "epoch": 2.886456908344733,
+      "grad_norm": 0.30597612261772156,
+      "learning_rate": 1.712512893586593e-07,
+      "loss": 1.4865,
+      "step": 1055
+    },
+    {
+      "epoch": 2.8891928864569083,
+      "grad_norm": 0.3064374327659607,
+      "learning_rate": 1.6280494658307e-07,
+      "loss": 1.5211,
+      "step": 1056
+    },
+    {
+      "epoch": 2.8919288645690835,
+      "grad_norm": 0.3949945867061615,
+      "learning_rate": 1.5457151698390347e-07,
+      "loss": 1.5316,
+      "step": 1057
+    },
+    {
+      "epoch": 2.8946648426812587,
+      "grad_norm": 0.4148292541503906,
+      "learning_rate": 1.4655107114101007e-07,
+      "loss": 1.3986,
+      "step": 1058
+    },
+    {
+      "epoch": 2.8974008207934334,
+      "grad_norm": 0.32118478417396545,
+      "learning_rate": 1.387436778084783e-07,
+      "loss": 1.7486,
+      "step": 1059
+    },
+    {
+      "epoch": 2.9001367989056086,
+      "grad_norm": 0.34837549924850464,
+      "learning_rate": 1.3114940391403263e-07,
+      "loss": 1.3545,
+      "step": 1060
+    },
+    {
+      "epoch": 2.902872777017784,
+      "grad_norm": 0.331442266702652,
+      "learning_rate": 1.2376831455845893e-07,
+      "loss": 1.5355,
+      "step": 1061
+    },
+    {
+      "epoch": 2.905608755129959,
+      "grad_norm": 0.3617335855960846,
+      "learning_rate": 1.1660047301505495e-07,
+      "loss": 1.5907,
+      "step": 1062
+    },
+    {
+      "epoch": 2.908344733242134,
+      "grad_norm": 0.32971206307411194,
+      "learning_rate": 1.0964594072907786e-07,
+      "loss": 1.6133,
+      "step": 1063
+    },
+    {
+      "epoch": 2.911080711354309,
+      "grad_norm": 0.34654051065444946,
+      "learning_rate": 1.0290477731722814e-07,
+      "loss": 1.7007,
+      "step": 1064
+    },
+    {
+      "epoch": 2.9138166894664845,
+      "grad_norm": 0.3612649738788605,
+      "learning_rate": 9.637704056713049e-08,
+      "loss": 1.2852,
+      "step": 1065
+    },
+    {
+      "epoch": 2.9165526675786593,
+      "grad_norm": 0.2997490465641022,
+      "learning_rate": 9.006278643683696e-08,
+      "loss": 1.5689,
+      "step": 1066
+    },
+    {
+      "epoch": 2.9192886456908345,
+      "grad_norm": 0.3305412828922272,
+      "learning_rate": 8.396206905436355e-08,
+      "loss": 1.5476,
+      "step": 1067
+    },
+    {
+      "epoch": 2.9220246238030096,
+      "grad_norm": 0.36031877994537354,
+      "learning_rate": 7.807494071720156e-08,
+      "loss": 1.6777,
+      "step": 1068
+    },
+    {
+      "epoch": 2.924760601915185,
+      "grad_norm": 0.3095710873603821,
+      "learning_rate": 7.24014518918903e-08,
+      "loss": 1.589,
+      "step": 1069
+    },
+    {
+      "epoch": 2.92749658002736,
+      "grad_norm": 0.30806005001068115,
+      "learning_rate": 6.694165121356732e-08,
+      "loss": 1.4103,
+      "step": 1070
+    },
+    {
+      "epoch": 2.9302325581395348,
+      "grad_norm": 0.36173051595687866,
+      "learning_rate": 6.169558548556887e-08,
+      "loss": 1.3676,
+      "step": 1071
+    },
+    {
+      "epoch": 2.93296853625171,
+      "grad_norm": 0.2915642559528351,
+      "learning_rate": 5.6663299679018955e-08,
+      "loss": 1.4523,
+      "step": 1072
+    },
+    {
+      "epoch": 2.935704514363885,
+      "grad_norm": 0.3511941730976105,
+      "learning_rate": 5.1844836932438134e-08,
+      "loss": 1.5194,
+      "step": 1073
+    },
+    {
+      "epoch": 2.9384404924760603,
+      "grad_norm": 0.3125559389591217,
+      "learning_rate": 4.7240238551385376e-08,
+      "loss": 1.4604,
+      "step": 1074
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 0.3312918245792389,
+      "learning_rate": 4.284954400810004e-08,
+      "loss": 1.4807,
+      "step": 1075
+    },
+    {
+      "epoch": 2.9439124487004102,
+      "grad_norm": 0.3593026101589203,
+      "learning_rate": 3.86727909411605e-08,
+      "loss": 1.7197,
+      "step": 1076
+    },
+    {
+      "epoch": 2.9466484268125854,
+      "grad_norm": 0.3443048298358917,
+      "learning_rate": 3.471001515516215e-08,
+      "loss": 1.6237,
+      "step": 1077
+    },
+    {
+      "epoch": 2.9493844049247606,
+      "grad_norm": 0.32331955432891846,
+      "learning_rate": 3.096125062041488e-08,
+      "loss": 1.5434,
+      "step": 1078
+    },
+    {
+      "epoch": 2.952120383036936,
+      "grad_norm": 0.3328414261341095,
+      "learning_rate": 2.7426529472648898e-08,
+      "loss": 1.6007,
+      "step": 1079
+    },
+    {
+      "epoch": 2.954856361149111,
+      "grad_norm": 0.31277886033058167,
+      "learning_rate": 2.4105882012734336e-08,
+      "loss": 1.6117,
+      "step": 1080
+    },
+    {
+      "epoch": 2.9575923392612857,
+      "grad_norm": 0.32368844747543335,
+      "learning_rate": 2.0999336706434282e-08,
+      "loss": 1.4795,
+      "step": 1081
+    },
+    {
+      "epoch": 2.960328317373461,
+      "grad_norm": 0.3160129487514496,
+      "learning_rate": 1.810692018414384e-08,
+      "loss": 1.5325,
+      "step": 1082
+    },
+    {
+      "epoch": 2.963064295485636,
+      "grad_norm": 0.3517950773239136,
+      "learning_rate": 1.5428657240676437e-08,
+      "loss": 1.4103,
+      "step": 1083
+    },
+    {
+      "epoch": 2.9658002735978113,
+      "grad_norm": 0.36211562156677246,
+      "learning_rate": 1.2964570835047318e-08,
+      "loss": 1.6729,
+      "step": 1084
+    },
+    {
+      "epoch": 2.9685362517099865,
+      "grad_norm": 0.3121832609176636,
+      "learning_rate": 1.0714682090270933e-08,
+      "loss": 1.3272,
+      "step": 1085
+    },
+    {
+      "epoch": 2.971272229822161,
+      "grad_norm": 0.33123496174812317,
+      "learning_rate": 8.679010293183298e-09,
+      "loss": 1.506,
+      "step": 1086
+    },
+    {
+      "epoch": 2.9740082079343364,
+      "grad_norm": 0.32456517219543457,
+      "learning_rate": 6.857572894281017e-09,
+      "loss": 1.5943,
+      "step": 1087
+    },
+    {
+      "epoch": 2.9767441860465116,
+      "grad_norm": 0.3651670813560486,
+      "learning_rate": 5.250385507568623e-09,
+      "loss": 1.4488,
+      "step": 1088
+    },
+    {
+      "epoch": 2.9794801641586868,
+      "grad_norm": 0.3096288740634918,
+      "learning_rate": 3.8574619104170265e-09,
+      "loss": 1.4775,
+      "step": 1089
+    },
+    {
+      "epoch": 2.982216142270862,
+      "grad_norm": 0.33430325984954834,
+      "learning_rate": 2.67881404346082e-09,
+      "loss": 1.5276,
+      "step": 1090
+    },
+    {
+      "epoch": 2.9849521203830367,
+      "grad_norm": 0.2988739609718323,
+      "learning_rate": 1.7144520104844798e-09,
+      "loss": 1.4893,
+      "step": 1091
+    },
+    {
+      "epoch": 2.987688098495212,
+      "grad_norm": 0.35142961144447327,
+      "learning_rate": 9.643840783363224e-10,
+      "loss": 1.4864,
+      "step": 1092
+    },
+    {
+      "epoch": 2.990424076607387,
+      "grad_norm": 0.2971309721469879,
+      "learning_rate": 4.2861667686744377e-10,
+      "loss": 1.6171,
+      "step": 1093
+    },
+    {
+      "epoch": 2.9931600547195623,
+      "grad_norm": 0.36940792202949524,
+      "learning_rate": 1.0715439885677913e-10,
+      "loss": 1.5264,
+      "step": 1094
+    },
+    {
+      "epoch": 2.9958960328317374,
+      "grad_norm": 0.31959155201911926,
+      "learning_rate": 0.0,
+      "loss": 1.585,
+      "step": 1095
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1095,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 250,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0313245522312704e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}