diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4889 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9989126495106923,
+  "eval_steps": 173,
+  "global_step": 689,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014498006524102935,
+      "grad_norm": 22.47921084493209,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 1.8971,
+      "step": 1
+    },
+    {
+      "epoch": 0.002899601304820587,
+      "grad_norm": 29.212530103395608,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 1.9568,
+      "step": 2
+    },
+    {
+      "epoch": 0.004349401957230881,
+      "grad_norm": 27.791010901767706,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 2.0294,
+      "step": 3
+    },
+    {
+      "epoch": 0.005799202609641174,
+      "grad_norm": 17.72267553468782,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 1.8542,
+      "step": 4
+    },
+    {
+      "epoch": 0.007249003262051468,
+      "grad_norm": 11.322906547281054,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 1.7099,
+      "step": 5
+    },
+    {
+      "epoch": 0.008698803914461761,
+      "grad_norm": 7.552826114767235,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.5584,
+      "step": 6
+    },
+    {
+      "epoch": 0.010148604566872055,
+      "grad_norm": 6.426028629972506,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.5593,
+      "step": 7
+    },
+    {
+      "epoch": 0.011598405219282348,
+      "grad_norm": 8.388751884163005,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 1.4325,
+      "step": 8
+    },
+    {
+      "epoch": 0.013048205871692642,
+      "grad_norm": 4.006153803982815,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.4088,
+      "step": 9
+    },
+    {
+      "epoch": 0.014498006524102935,
+      "grad_norm": 4.620580116636459,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 1.3702,
+      "step": 10
+    },
+    {
+      "epoch": 0.01594780717651323,
+      "grad_norm": 3.114603192845655,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 1.4002,
+      "step": 11
+    },
+    {
+      "epoch": 0.017397607828923523,
+      "grad_norm": 4.90901693649213,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.3606,
+      "step": 12
+    },
+    {
+      "epoch": 0.018847408481333816,
+      "grad_norm": 3.3810740001714326,
+      "learning_rate": 1.2380952380952383e-05,
+      "loss": 1.2568,
+      "step": 13
+    },
+    {
+      "epoch": 0.02029720913374411,
+      "grad_norm": 3.249493396935377,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.3677,
+      "step": 14
+    },
+    {
+      "epoch": 0.021747009786154403,
+      "grad_norm": 3.093405223429388,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 1.2785,
+      "step": 15
+    },
+    {
+      "epoch": 0.023196810438564697,
+      "grad_norm": 2.8636742279984593,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 1.2201,
+      "step": 16
+    },
+    {
+      "epoch": 0.02464661109097499,
+      "grad_norm": 2.8113081821202615,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 1.2688,
+      "step": 17
+    },
+    {
+      "epoch": 0.026096411743385284,
+      "grad_norm": 2.326431557985723,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 1.1451,
+      "step": 18
+    },
+    {
+      "epoch": 0.027546212395795577,
+      "grad_norm": 2.1890845574639286,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 1.2595,
+      "step": 19
+    },
+    {
+      "epoch": 0.02899601304820587,
+      "grad_norm": 2.0316484616392154,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 1.1634,
+      "step": 20
+    },
+    {
+      "epoch": 0.030445813700616164,
+      "grad_norm": 2.1032536357160354,
+      "learning_rate": 2e-05,
+      "loss": 1.2486,
+      "step": 21
+    },
+    {
+      "epoch": 0.03189561435302646,
+      "grad_norm": 2.153088817641949,
+      "learning_rate": 1.999988940995767e-05,
+      "loss": 1.2253,
+      "step": 22
+    },
+    {
+      "epoch": 0.03334541500543675,
+      "grad_norm": 2.049206731914148,
+      "learning_rate": 1.9999557642276715e-05,
+      "loss": 1.2147,
+      "step": 23
+    },
+    {
+      "epoch": 0.034795215657847045,
+      "grad_norm": 1.8533242148282187,
+      "learning_rate": 1.9999004704295167e-05,
+      "loss": 1.2775,
+      "step": 24
+    },
+    {
+      "epoch": 0.03624501631025734,
+      "grad_norm": 1.7199213866686218,
+      "learning_rate": 1.9998230608242917e-05,
+      "loss": 1.1722,
+      "step": 25
+    },
+    {
+      "epoch": 0.03769481696266763,
+      "grad_norm": 1.89257124283304,
+      "learning_rate": 1.999723537124143e-05,
+      "loss": 1.2739,
+      "step": 26
+    },
+    {
+      "epoch": 0.039144617615077926,
+      "grad_norm": 1.9192296454139428,
+      "learning_rate": 1.9996019015303366e-05,
+      "loss": 1.3087,
+      "step": 27
+    },
+    {
+      "epoch": 0.04059441826748822,
+      "grad_norm": 1.849426642636138,
+      "learning_rate": 1.9994581567332095e-05,
+      "loss": 1.2227,
+      "step": 28
+    },
+    {
+      "epoch": 0.04204421891989851,
+      "grad_norm": 1.7662341724804647,
+      "learning_rate": 1.9992923059121107e-05,
+      "loss": 1.2358,
+      "step": 29
+    },
+    {
+      "epoch": 0.043494019572308806,
+      "grad_norm": 1.8663942485219291,
+      "learning_rate": 1.999104352735329e-05,
+      "loss": 1.1676,
+      "step": 30
+    },
+    {
+      "epoch": 0.0449438202247191,
+      "grad_norm": 1.8313769504549111,
+      "learning_rate": 1.9988943013600156e-05,
+      "loss": 1.1458,
+      "step": 31
+    },
+    {
+      "epoch": 0.04639362087712939,
+      "grad_norm": 1.8549702164486843,
+      "learning_rate": 1.9986621564320878e-05,
+      "loss": 1.2451,
+      "step": 32
+    },
+    {
+      "epoch": 0.04784342152953969,
+      "grad_norm": 1.7721110361194345,
+      "learning_rate": 1.9984079230861296e-05,
+      "loss": 1.1446,
+      "step": 33
+    },
+    {
+      "epoch": 0.04929322218194998,
+      "grad_norm": 1.79315683421693,
+      "learning_rate": 1.998131606945276e-05,
+      "loss": 1.2002,
+      "step": 34
+    },
+    {
+      "epoch": 0.050743022834360274,
+      "grad_norm": 1.7619297929194548,
+      "learning_rate": 1.9978332141210894e-05,
+      "loss": 1.16,
+      "step": 35
+    },
+    {
+      "epoch": 0.05219282348677057,
+      "grad_norm": 1.7557344100126386,
+      "learning_rate": 1.997512751213426e-05,
+      "loss": 1.2576,
+      "step": 36
+    },
+    {
+      "epoch": 0.05364262413918086,
+      "grad_norm": 1.6682715425247785,
+      "learning_rate": 1.9971702253102856e-05,
+      "loss": 1.2412,
+      "step": 37
+    },
+    {
+      "epoch": 0.055092424791591155,
+      "grad_norm": 1.5054000837293153,
+      "learning_rate": 1.99680564398766e-05,
+      "loss": 1.1546,
+      "step": 38
+    },
+    {
+      "epoch": 0.05654222544400145,
+      "grad_norm": 1.6305996348324838,
+      "learning_rate": 1.9964190153093613e-05,
+      "loss": 1.1383,
+      "step": 39
+    },
+    {
+      "epoch": 0.05799202609641174,
+      "grad_norm": 1.8452558784097721,
+      "learning_rate": 1.996010347826847e-05,
+      "loss": 1.1433,
+      "step": 40
+    },
+    {
+      "epoch": 0.059441826748822035,
+      "grad_norm": 1.6222012672663748,
+      "learning_rate": 1.9955796505790266e-05,
+      "loss": 1.1716,
+      "step": 41
+    },
+    {
+      "epoch": 0.06089162740123233,
+      "grad_norm": 1.6637995619824535,
+      "learning_rate": 1.9951269330920667e-05,
+      "loss": 1.195,
+      "step": 42
+    },
+    {
+      "epoch": 0.06234142805364262,
+      "grad_norm": 1.7573421138658332,
+      "learning_rate": 1.9946522053791755e-05,
+      "loss": 1.1977,
+      "step": 43
+    },
+    {
+      "epoch": 0.06379122870605292,
+      "grad_norm": 1.620251131694878,
+      "learning_rate": 1.9941554779403854e-05,
+      "loss": 1.1223,
+      "step": 44
+    },
+    {
+      "epoch": 0.06524102935846321,
+      "grad_norm": 1.7364676604536926,
+      "learning_rate": 1.993636761762317e-05,
+      "loss": 1.0842,
+      "step": 45
+    },
+    {
+      "epoch": 0.0666908300108735,
+      "grad_norm": 1.5967650991319868,
+      "learning_rate": 1.9930960683179404e-05,
+      "loss": 1.146,
+      "step": 46
+    },
+    {
+      "epoch": 0.0681406306632838,
+      "grad_norm": 2.1462130649903495,
+      "learning_rate": 1.992533409566317e-05,
+      "loss": 1.2389,
+      "step": 47
+    },
+    {
+      "epoch": 0.06959043131569409,
+      "grad_norm": 1.5461340443066385,
+      "learning_rate": 1.991948797952338e-05,
+      "loss": 1.1727,
+      "step": 48
+    },
+    {
+      "epoch": 0.07104023196810438,
+      "grad_norm": 1.53452383630284,
+      "learning_rate": 1.991342246406448e-05,
+      "loss": 1.1794,
+      "step": 49
+    },
+    {
+      "epoch": 0.07249003262051468,
+      "grad_norm": 1.823231005832227,
+      "learning_rate": 1.9907137683443593e-05,
+      "loss": 1.1269,
+      "step": 50
+    },
+    {
+      "epoch": 0.07393983327292497,
+      "grad_norm": 1.4161334829183205,
+      "learning_rate": 1.990063377666755e-05,
+      "loss": 1.1762,
+      "step": 51
+    },
+    {
+      "epoch": 0.07538963392533526,
+      "grad_norm": 1.4110618176312044,
+      "learning_rate": 1.9893910887589817e-05,
+      "loss": 1.1451,
+      "step": 52
+    },
+    {
+      "epoch": 0.07683943457774556,
+      "grad_norm": 1.4772820370731468,
+      "learning_rate": 1.9886969164907306e-05,
+      "loss": 1.2367,
+      "step": 53
+    },
+    {
+      "epoch": 0.07828923523015585,
+      "grad_norm": 1.6854195963021186,
+      "learning_rate": 1.9879808762157106e-05,
+      "loss": 1.1267,
+      "step": 54
+    },
+    {
+      "epoch": 0.07973903588256614,
+      "grad_norm": 1.7555308648784804,
+      "learning_rate": 1.987242983771306e-05,
+      "loss": 1.1527,
+      "step": 55
+    },
+    {
+      "epoch": 0.08118883653497644,
+      "grad_norm": 1.6176407306562521,
+      "learning_rate": 1.986483255478229e-05,
+      "loss": 1.1832,
+      "step": 56
+    },
+    {
+      "epoch": 0.08263863718738673,
+      "grad_norm": 1.4791882296730032,
+      "learning_rate": 1.985701708140155e-05,
+      "loss": 1.2262,
+      "step": 57
+    },
+    {
+      "epoch": 0.08408843783979703,
+      "grad_norm": 1.5293989025600263,
+      "learning_rate": 1.9848983590433555e-05,
+      "loss": 1.1988,
+      "step": 58
+    },
+    {
+      "epoch": 0.08553823849220732,
+      "grad_norm": 1.7461881138563102,
+      "learning_rate": 1.984073225956313e-05,
+      "loss": 1.0908,
+      "step": 59
+    },
+    {
+      "epoch": 0.08698803914461761,
+      "grad_norm": 1.505126937999859,
+      "learning_rate": 1.983226327129327e-05,
+      "loss": 1.1955,
+      "step": 60
+    },
+    {
+      "epoch": 0.0884378397970279,
+      "grad_norm": 1.5517411887913295,
+      "learning_rate": 1.982357681294114e-05,
+      "loss": 1.0568,
+      "step": 61
+    },
+    {
+      "epoch": 0.0898876404494382,
+      "grad_norm": 1.6010613456853238,
+      "learning_rate": 1.981467307663389e-05,
+      "loss": 1.095,
+      "step": 62
+    },
+    {
+      "epoch": 0.0913374411018485,
+      "grad_norm": 1.5984617535884107,
+      "learning_rate": 1.9805552259304447e-05,
+      "loss": 1.1318,
+      "step": 63
+    },
+    {
+      "epoch": 0.09278724175425879,
+      "grad_norm": 1.6758994462265333,
+      "learning_rate": 1.9796214562687114e-05,
+      "loss": 1.1377,
+      "step": 64
+    },
+    {
+      "epoch": 0.09423704240666908,
+      "grad_norm": 1.4132970911261034,
+      "learning_rate": 1.978666019331315e-05,
+      "loss": 1.1289,
+      "step": 65
+    },
+    {
+      "epoch": 0.09568684305907937,
+      "grad_norm": 1.476648748614766,
+      "learning_rate": 1.9776889362506175e-05,
+      "loss": 1.0684,
+      "step": 66
+    },
+    {
+      "epoch": 0.09713664371148967,
+      "grad_norm": 1.5190130660494991,
+      "learning_rate": 1.9766902286377507e-05,
+      "loss": 1.114,
+      "step": 67
+    },
+    {
+      "epoch": 0.09858644436389996,
+      "grad_norm": 1.531845221805904,
+      "learning_rate": 1.9756699185821387e-05,
+      "loss": 1.2046,
+      "step": 68
+    },
+    {
+      "epoch": 0.10003624501631025,
+      "grad_norm": 1.5527420846893765,
+      "learning_rate": 1.974628028651007e-05,
+      "loss": 1.1269,
+      "step": 69
+    },
+    {
+      "epoch": 0.10148604566872055,
+      "grad_norm": 1.4941200713833973,
+      "learning_rate": 1.973564581888887e-05,
+      "loss": 1.118,
+      "step": 70
+    },
+    {
+      "epoch": 0.10293584632113084,
+      "grad_norm": 1.5175836270306755,
+      "learning_rate": 1.9724796018171022e-05,
+      "loss": 1.191,
+      "step": 71
+    },
+    {
+      "epoch": 0.10438564697354114,
+      "grad_norm": 1.367597652901422,
+      "learning_rate": 1.9713731124332513e-05,
+      "loss": 1.0716,
+      "step": 72
+    },
+    {
+      "epoch": 0.10583544762595143,
+      "grad_norm": 1.5678222370782957,
+      "learning_rate": 1.9702451382106763e-05,
+      "loss": 1.146,
+      "step": 73
+    },
+    {
+      "epoch": 0.10728524827836172,
+      "grad_norm": 1.44024988905656,
+      "learning_rate": 1.96909570409792e-05,
+      "loss": 1.165,
+      "step": 74
+    },
+    {
+      "epoch": 0.10873504893077202,
+      "grad_norm": 1.4857188479739345,
+      "learning_rate": 1.967924835518176e-05,
+      "loss": 1.2136,
+      "step": 75
+    },
+    {
+      "epoch": 0.11018484958318231,
+      "grad_norm": 1.4816821516883913,
+      "learning_rate": 1.9667325583687255e-05,
+      "loss": 1.1995,
+      "step": 76
+    },
+    {
+      "epoch": 0.1116346502355926,
+      "grad_norm": 1.471928264176159,
+      "learning_rate": 1.9655188990203648e-05,
+      "loss": 1.1189,
+      "step": 77
+    },
+    {
+      "epoch": 0.1130844508880029,
+      "grad_norm": 1.3598477529860473,
+      "learning_rate": 1.9642838843168215e-05,
+      "loss": 1.0926,
+      "step": 78
+    },
+    {
+      "epoch": 0.11453425154041319,
+      "grad_norm": 1.3918604182664005,
+      "learning_rate": 1.963027541574161e-05,
+      "loss": 1.2357,
+      "step": 79
+    },
+    {
+      "epoch": 0.11598405219282348,
+      "grad_norm": 1.4445325343534552,
+      "learning_rate": 1.9617498985801834e-05,
+      "loss": 1.0361,
+      "step": 80
+    },
+    {
+      "epoch": 0.11743385284523378,
+      "grad_norm": 1.5844508424253305,
+      "learning_rate": 1.9604509835938066e-05,
+      "loss": 1.1391,
+      "step": 81
+    },
+    {
+      "epoch": 0.11888365349764407,
+      "grad_norm": 1.3947715223935726,
+      "learning_rate": 1.9591308253444437e-05,
+      "loss": 1.1731,
+      "step": 82
+    },
+    {
+      "epoch": 0.12033345415005436,
+      "grad_norm": 1.429704481047618,
+      "learning_rate": 1.9577894530313655e-05,
+      "loss": 1.1635,
+      "step": 83
+    },
+    {
+      "epoch": 0.12178325480246466,
+      "grad_norm": 1.6942790137385033,
+      "learning_rate": 1.9564268963230564e-05,
+      "loss": 1.1596,
+      "step": 84
+    },
+    {
+      "epoch": 0.12323305545487495,
+      "grad_norm": 1.5559590346219652,
+      "learning_rate": 1.9550431853565577e-05,
+      "loss": 1.1828,
+      "step": 85
+    },
+    {
+      "epoch": 0.12468285610728524,
+      "grad_norm": 1.5722198588317637,
+      "learning_rate": 1.9536383507367994e-05,
+      "loss": 1.122,
+      "step": 86
+    },
+    {
+      "epoch": 0.12613265675969554,
+      "grad_norm": 1.4071413684659504,
+      "learning_rate": 1.9522124235359265e-05,
+      "loss": 1.0916,
+      "step": 87
+    },
+    {
+      "epoch": 0.12758245741210583,
+      "grad_norm": 1.539766939300241,
+      "learning_rate": 1.950765435292608e-05,
+      "loss": 1.1938,
+      "step": 88
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 1.721939314737648,
+      "learning_rate": 1.9492974180113425e-05,
+      "loss": 1.175,
+      "step": 89
+    },
+    {
+      "epoch": 0.13048205871692642,
+      "grad_norm": 1.4364509556358267,
+      "learning_rate": 1.9478084041617492e-05,
+      "loss": 1.1205,
+      "step": 90
+    },
+    {
+      "epoch": 0.1319318593693367,
+      "grad_norm": 1.400216410067683,
+      "learning_rate": 1.946298426677848e-05,
+      "loss": 1.1568,
+      "step": 91
+    },
+    {
+      "epoch": 0.133381660021747,
+      "grad_norm": 1.5735590237326518,
+      "learning_rate": 1.944767518957334e-05,
+      "loss": 1.158,
+      "step": 92
+    },
+    {
+      "epoch": 0.1348314606741573,
+      "grad_norm": 1.4382370297928824,
+      "learning_rate": 1.943215714860838e-05,
+      "loss": 1.1029,
+      "step": 93
+    },
+    {
+      "epoch": 0.1362812613265676,
+      "grad_norm": 1.2693278649325574,
+      "learning_rate": 1.941643048711175e-05,
+      "loss": 1.1579,
+      "step": 94
+    },
+    {
+      "epoch": 0.1377310619789779,
+      "grad_norm": 1.375575060015626,
+      "learning_rate": 1.9400495552925887e-05,
+      "loss": 1.0929,
+      "step": 95
+    },
+    {
+      "epoch": 0.13918086263138818,
+      "grad_norm": 1.5103601938936249,
+      "learning_rate": 1.93843526984998e-05,
+      "loss": 1.1741,
+      "step": 96
+    },
+    {
+      "epoch": 0.14063066328379847,
+      "grad_norm": 1.4096419501236,
+      "learning_rate": 1.936800228088128e-05,
+      "loss": 1.1674,
+      "step": 97
+    },
+    {
+      "epoch": 0.14208046393620877,
+      "grad_norm": 1.4535501080466646,
+      "learning_rate": 1.9351444661709e-05,
+      "loss": 1.1773,
+      "step": 98
+    },
+    {
+      "epoch": 0.14353026458861906,
+      "grad_norm": 1.482698015976603,
+      "learning_rate": 1.9334680207204525e-05,
+      "loss": 1.1046,
+      "step": 99
+    },
+    {
+      "epoch": 0.14498006524102935,
+      "grad_norm": 1.392779032425837,
+      "learning_rate": 1.9317709288164198e-05,
+      "loss": 1.0075,
+      "step": 100
+    },
+    {
+      "epoch": 0.14642986589343965,
+      "grad_norm": 1.5050842471289037,
+      "learning_rate": 1.930053227995095e-05,
+      "loss": 1.1393,
+      "step": 101
+    },
+    {
+      "epoch": 0.14787966654584994,
+      "grad_norm": 1.3514262892953859,
+      "learning_rate": 1.9283149562485997e-05,
+      "loss": 1.1103,
+      "step": 102
+    },
+    {
+      "epoch": 0.14932946719826024,
+      "grad_norm": 1.6027106173383818,
+      "learning_rate": 1.9265561520240434e-05,
+      "loss": 1.1564,
+      "step": 103
+    },
+    {
+      "epoch": 0.15077926785067053,
+      "grad_norm": 1.4523181501957745,
+      "learning_rate": 1.924776854222672e-05,
+      "loss": 1.1451,
+      "step": 104
+    },
+    {
+      "epoch": 0.15222906850308082,
+      "grad_norm": 1.468929444746814,
+      "learning_rate": 1.9229771021990094e-05,
+      "loss": 1.1046,
+      "step": 105
+    },
+    {
+      "epoch": 0.15367886915549112,
+      "grad_norm": 1.571669347997552,
+      "learning_rate": 1.9211569357599868e-05,
+      "loss": 1.1092,
+      "step": 106
+    },
+    {
+      "epoch": 0.1551286698079014,
+      "grad_norm": 1.466957885253298,
+      "learning_rate": 1.9193163951640608e-05,
+      "loss": 1.2021,
+      "step": 107
+    },
+    {
+      "epoch": 0.1565784704603117,
+      "grad_norm": 1.5514297656947849,
+      "learning_rate": 1.9174555211203235e-05,
+      "loss": 1.2321,
+      "step": 108
+    },
+    {
+      "epoch": 0.158028271112722,
+      "grad_norm": 1.662361900190696,
+      "learning_rate": 1.9155743547876026e-05,
+      "loss": 1.1215,
+      "step": 109
+    },
+    {
+      "epoch": 0.1594780717651323,
+      "grad_norm": 1.4055683816860303,
+      "learning_rate": 1.9136729377735514e-05,
+      "loss": 1.073,
+      "step": 110
+    },
+    {
+      "epoch": 0.16092787241754258,
+      "grad_norm": 1.562213366778017,
+      "learning_rate": 1.9117513121337273e-05,
+      "loss": 1.1551,
+      "step": 111
+    },
+    {
+      "epoch": 0.16237767306995288,
+      "grad_norm": 1.6060489844165238,
+      "learning_rate": 1.9098095203706625e-05,
+      "loss": 1.1706,
+      "step": 112
+    },
+    {
+      "epoch": 0.16382747372236317,
+      "grad_norm": 1.8556202603195604,
+      "learning_rate": 1.9078476054329237e-05,
+      "loss": 1.1171,
+      "step": 113
+    },
+    {
+      "epoch": 0.16527727437477346,
+      "grad_norm": 1.7090320630358615,
+      "learning_rate": 1.905865610714162e-05,
+      "loss": 1.1617,
+      "step": 114
+    },
+    {
+      "epoch": 0.16672707502718376,
+      "grad_norm": 1.4643321344000653,
+      "learning_rate": 1.9038635800521538e-05,
+      "loss": 1.1308,
+      "step": 115
+    },
+    {
+      "epoch": 0.16817687567959405,
+      "grad_norm": 1.5018160051324643,
+      "learning_rate": 1.9018415577278295e-05,
+      "loss": 1.1615,
+      "step": 116
+    },
+    {
+      "epoch": 0.16962667633200434,
+      "grad_norm": 1.5357408934097792,
+      "learning_rate": 1.8997995884642968e-05,
+      "loss": 1.1358,
+      "step": 117
+    },
+    {
+      "epoch": 0.17107647698441464,
+      "grad_norm": 1.4894643819708282,
+      "learning_rate": 1.8977377174258485e-05,
+      "loss": 1.2152,
+      "step": 118
+    },
+    {
+      "epoch": 0.17252627763682493,
+      "grad_norm": 1.4622909727617928,
+      "learning_rate": 1.895655990216966e-05,
+      "loss": 1.1477,
+      "step": 119
+    },
+    {
+      "epoch": 0.17397607828923523,
+      "grad_norm": 1.3843622144924201,
+      "learning_rate": 1.893554452881309e-05,
+      "loss": 1.0963,
+      "step": 120
+    },
+    {
+      "epoch": 0.17542587894164552,
+      "grad_norm": 1.3802773797790167,
+      "learning_rate": 1.8914331519006985e-05,
+      "loss": 1.087,
+      "step": 121
+    },
+    {
+      "epoch": 0.1768756795940558,
+      "grad_norm": 1.3860519519347008,
+      "learning_rate": 1.8892921341940873e-05,
+      "loss": 1.1253,
+      "step": 122
+    },
+    {
+      "epoch": 0.1783254802464661,
+      "grad_norm": 1.402888618675544,
+      "learning_rate": 1.8871314471165233e-05,
+      "loss": 1.1347,
+      "step": 123
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 1.4853426212807326,
+      "learning_rate": 1.8849511384581015e-05,
+      "loss": 1.111,
+      "step": 124
+    },
+    {
+      "epoch": 0.1812250815512867,
+      "grad_norm": 1.305335577546698,
+      "learning_rate": 1.8827512564429076e-05,
+      "loss": 1.1121,
+      "step": 125
+    },
+    {
+      "epoch": 0.182674882203697,
+      "grad_norm": 1.2403899478271456,
+      "learning_rate": 1.8805318497279498e-05,
+      "loss": 1.0564,
+      "step": 126
+    },
+    {
+      "epoch": 0.18412468285610728,
+      "grad_norm": 1.4975304356267847,
+      "learning_rate": 1.8782929674020856e-05,
+      "loss": 1.1015,
+      "step": 127
+    },
+    {
+      "epoch": 0.18557448350851757,
+      "grad_norm": 1.5576431905392216,
+      "learning_rate": 1.8760346589849323e-05,
+      "loss": 1.1226,
+      "step": 128
+    },
+    {
+      "epoch": 0.18702428416092787,
+      "grad_norm": 1.488403623529011,
+      "learning_rate": 1.8737569744257754e-05,
+      "loss": 1.092,
+      "step": 129
+    },
+    {
+      "epoch": 0.18847408481333816,
+      "grad_norm": 1.3283937976790703,
+      "learning_rate": 1.871459964102461e-05,
+      "loss": 1.0942,
+      "step": 130
+    },
+    {
+      "epoch": 0.18992388546574845,
+      "grad_norm": 1.6187906615159156,
+      "learning_rate": 1.8691436788202823e-05,
+      "loss": 1.1802,
+      "step": 131
+    },
+    {
+      "epoch": 0.19137368611815875,
+      "grad_norm": 1.683184520335961,
+      "learning_rate": 1.8668081698108576e-05,
+      "loss": 1.2293,
+      "step": 132
+    },
+    {
+      "epoch": 0.19282348677056904,
+      "grad_norm": 1.4826566875019622,
+      "learning_rate": 1.8644534887309947e-05,
+      "loss": 1.1086,
+      "step": 133
+    },
+    {
+      "epoch": 0.19427328742297933,
+      "grad_norm": 1.5863759229146668,
+      "learning_rate": 1.8620796876615494e-05,
+      "loss": 1.1327,
+      "step": 134
+    },
+    {
+      "epoch": 0.19572308807538963,
+      "grad_norm": 1.6304679377684546,
+      "learning_rate": 1.859686819106274e-05,
+      "loss": 1.1653,
+      "step": 135
+    },
+    {
+      "epoch": 0.19717288872779992,
+      "grad_norm": 1.3217063308417187,
+      "learning_rate": 1.8572749359906556e-05,
+      "loss": 1.1362,
+      "step": 136
+    },
+    {
+      "epoch": 0.19862268938021022,
+      "grad_norm": 1.3561432883735813,
+      "learning_rate": 1.8548440916607454e-05,
+      "loss": 1.1004,
+      "step": 137
+    },
+    {
+      "epoch": 0.2000724900326205,
+      "grad_norm": 1.4131206062623582,
+      "learning_rate": 1.8523943398819786e-05,
+      "loss": 1.1307,
+      "step": 138
+    },
+    {
+      "epoch": 0.2015222906850308,
+      "grad_norm": 1.471537720055082,
+      "learning_rate": 1.8499257348379857e-05,
+      "loss": 1.127,
+      "step": 139
+    },
+    {
+      "epoch": 0.2029720913374411,
+      "grad_norm": 1.3915619265636519,
+      "learning_rate": 1.8474383311293943e-05,
+      "loss": 1.1377,
+      "step": 140
+    },
+    {
+      "epoch": 0.2044218919898514,
+      "grad_norm": 1.4213857540378092,
+      "learning_rate": 1.8449321837726207e-05,
+      "loss": 1.1931,
+      "step": 141
+    },
+    {
+      "epoch": 0.20587169264226168,
+      "grad_norm": 1.3866002340392658,
+      "learning_rate": 1.842407348198653e-05,
+      "loss": 1.0383,
+      "step": 142
+    },
+    {
+      "epoch": 0.20732149329467198,
+      "grad_norm": 1.4933458445109034,
+      "learning_rate": 1.8398638802518263e-05,
+      "loss": 1.0664,
+      "step": 143
+    },
+    {
+      "epoch": 0.20877129394708227,
+      "grad_norm": 1.583104805277793,
+      "learning_rate": 1.8373018361885856e-05,
+      "loss": 1.069,
+      "step": 144
+    },
+    {
+      "epoch": 0.21022109459949256,
+      "grad_norm": 1.4714275313905285,
+      "learning_rate": 1.8347212726762436e-05,
+      "loss": 1.087,
+      "step": 145
+    },
+    {
+      "epoch": 0.21167089525190286,
+      "grad_norm": 1.4664672377784298,
+      "learning_rate": 1.832122246791726e-05,
+      "loss": 1.1051,
+      "step": 146
+    },
+    {
+      "epoch": 0.21312069590431315,
+      "grad_norm": 1.4431063549437655,
+      "learning_rate": 1.8295048160203087e-05,
+      "loss": 1.1654,
+      "step": 147
+    },
+    {
+      "epoch": 0.21457049655672344,
+      "grad_norm": 1.3900812349680303,
+      "learning_rate": 1.8268690382543484e-05,
+      "loss": 1.14,
+      "step": 148
+    },
+    {
+      "epoch": 0.21602029720913374,
+      "grad_norm": 1.266664129893108,
+      "learning_rate": 1.8242149717919993e-05,
+      "loss": 1.0561,
+      "step": 149
+    },
+    {
+      "epoch": 0.21747009786154403,
+      "grad_norm": 1.2928373328736114,
+      "learning_rate": 1.8215426753359268e-05,
+      "loss": 1.0394,
+      "step": 150
+    },
+    {
+      "epoch": 0.21891989851395433,
+      "grad_norm": 1.3795412647189715,
+      "learning_rate": 1.8188522079920056e-05,
+      "loss": 1.1646,
+      "step": 151
+    },
+    {
+      "epoch": 0.22036969916636462,
+      "grad_norm": 1.2484572985848708,
+      "learning_rate": 1.816143629268016e-05,
+      "loss": 1.0706,
+      "step": 152
+    },
+    {
+      "epoch": 0.2218194998187749,
+      "grad_norm": 1.343244551848235,
+      "learning_rate": 1.8134169990723246e-05,
+      "loss": 1.0825,
+      "step": 153
+    },
+    {
+      "epoch": 0.2232693004711852,
+      "grad_norm": 1.491624139238055,
+      "learning_rate": 1.8106723777125614e-05,
+      "loss": 1.1422,
+      "step": 154
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 1.5441831111656004,
+      "learning_rate": 1.8079098258942846e-05,
+      "loss": 1.1942,
+      "step": 155
+    },
+    {
+      "epoch": 0.2261689017760058,
+      "grad_norm": 1.4993573645842802,
+      "learning_rate": 1.8051294047196393e-05,
+      "loss": 1.175,
+      "step": 156
+    },
+    {
+      "epoch": 0.2276187024284161,
+      "grad_norm": 1.4599943153308388,
+      "learning_rate": 1.8023311756860036e-05,
+      "loss": 1.1265,
+      "step": 157
+    },
+    {
+      "epoch": 0.22906850308082638,
+      "grad_norm": 1.4592962242329999,
+      "learning_rate": 1.799515200684632e-05,
+      "loss": 1.1112,
+      "step": 158
+    },
+    {
+      "epoch": 0.23051830373323667,
+      "grad_norm": 1.4977501199677632,
+      "learning_rate": 1.796681541999283e-05,
+      "loss": 1.129,
+      "step": 159
+    },
+    {
+      "epoch": 0.23196810438564697,
+      "grad_norm": 1.3420077981513285,
+      "learning_rate": 1.793830262304843e-05,
+      "loss": 1.1235,
+      "step": 160
+    },
+    {
+      "epoch": 0.23341790503805726,
+      "grad_norm": 1.3384853724058963,
+      "learning_rate": 1.7909614246659406e-05,
+      "loss": 1.1785,
+      "step": 161
+    },
+    {
+      "epoch": 0.23486770569046755,
+      "grad_norm": 1.348282211605465,
+      "learning_rate": 1.7880750925355516e-05,
+      "loss": 1.1024,
+      "step": 162
+    },
+    {
+      "epoch": 0.23631750634287785,
+      "grad_norm": 1.509179228758862,
+      "learning_rate": 1.785171329753594e-05,
+      "loss": 1.0923,
+      "step": 163
+    },
+    {
+      "epoch": 0.23776730699528814,
+      "grad_norm": 1.499945819442489,
+      "learning_rate": 1.7822502005455176e-05,
+      "loss": 1.0614,
+      "step": 164
+    },
+    {
+      "epoch": 0.23921710764769843,
+      "grad_norm": 1.3848322033850378,
+      "learning_rate": 1.7793117695208828e-05,
+      "loss": 1.1082,
+      "step": 165
+    },
+    {
+      "epoch": 0.24066690830010873,
+      "grad_norm": 1.4837729503310633,
+      "learning_rate": 1.7763561016719323e-05,
+      "loss": 1.0345,
+      "step": 166
+    },
+    {
+      "epoch": 0.24211670895251902,
+      "grad_norm": 1.4000163516055224,
+      "learning_rate": 1.773383262372152e-05,
+      "loss": 1.1252,
+      "step": 167
+    },
+    {
+      "epoch": 0.24356650960492932,
+      "grad_norm": 1.3582651357456224,
+      "learning_rate": 1.7703933173748276e-05,
+      "loss": 1.1638,
+      "step": 168
+    },
+    {
+      "epoch": 0.2450163102573396,
+      "grad_norm": 1.4411384457116327,
+      "learning_rate": 1.767386332811587e-05,
+      "loss": 1.0803,
+      "step": 169
+    },
+    {
+      "epoch": 0.2464661109097499,
+      "grad_norm": 1.7615098689709985,
+      "learning_rate": 1.764362375190941e-05,
+      "loss": 1.1198,
+      "step": 170
+    },
+    {
+      "epoch": 0.2479159115621602,
+      "grad_norm": 1.4181183402041688,
+      "learning_rate": 1.761321511396809e-05,
+      "loss": 1.1036,
+      "step": 171
+    },
+    {
+      "epoch": 0.2493657122145705,
+      "grad_norm": 1.4910783816986892,
+      "learning_rate": 1.7582638086870426e-05,
+      "loss": 1.1397,
+      "step": 172
+    },
+    {
+      "epoch": 0.2508155128669808,
+      "grad_norm": 1.4706104271724008,
+      "learning_rate": 1.7551893346919363e-05,
+      "loss": 1.1333,
+      "step": 173
+    },
+    {
+      "epoch": 0.2508155128669808,
+      "eval_loss": 1.8410959243774414,
+      "eval_runtime": 353.3748,
+      "eval_samples_per_second": 49.958,
+      "eval_steps_per_second": 0.781,
+      "step": 173
+    },
+    {
+      "epoch": 0.2522653135193911,
+      "grad_norm": 1.4186603818499712,
+      "learning_rate": 1.752098157412732e-05,
+      "loss": 1.0633,
+      "step": 174
+    },
+    {
+      "epoch": 0.25371511417180137,
+      "grad_norm": 1.4415892238131922,
+      "learning_rate": 1.7489903452201144e-05,
+      "loss": 1.0758,
+      "step": 175
+    },
+    {
+      "epoch": 0.25516491482421166,
+      "grad_norm": 1.267760040471212,
+      "learning_rate": 1.7458659668527005e-05,
+      "loss": 1.0985,
+      "step": 176
+    },
+    {
+      "epoch": 0.25661471547662196,
+      "grad_norm": 1.40014380323161,
+      "learning_rate": 1.742725091415517e-05,
+      "loss": 1.0933,
+      "step": 177
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 1.5319507604029738,
+      "learning_rate": 1.7395677883784742e-05,
+      "loss": 1.104,
+      "step": 178
+    },
+    {
+      "epoch": 0.25951431678144254,
+      "grad_norm": 1.389192022293072,
+      "learning_rate": 1.7363941275748263e-05,
+      "loss": 1.1126,
+      "step": 179
+    },
+    {
+      "epoch": 0.26096411743385284,
+      "grad_norm": 1.6605141655333846,
+      "learning_rate": 1.733204179199631e-05,
+      "loss": 1.1607,
+      "step": 180
+    },
+    {
+      "epoch": 0.26241391808626313,
+      "grad_norm": 1.4884464648131481,
+      "learning_rate": 1.729998013808192e-05,
+      "loss": 1.1933,
+      "step": 181
+    },
+    {
+      "epoch": 0.2638637187386734,
+      "grad_norm": 1.3165875567539376,
+      "learning_rate": 1.726775702314504e-05,
+      "loss": 1.0377,
+      "step": 182
+    },
+    {
+      "epoch": 0.2653135193910837,
+      "grad_norm": 1.3006144534458735,
+      "learning_rate": 1.7235373159896792e-05,
+      "loss": 1.0688,
+      "step": 183
+    },
+    {
+      "epoch": 0.266763320043494,
+      "grad_norm": 1.3866267078678514,
+      "learning_rate": 1.720282926460374e-05,
+      "loss": 1.1689,
+      "step": 184
+    },
+    {
+      "epoch": 0.2682131206959043,
+      "grad_norm": 1.4401920187839905,
+      "learning_rate": 1.717012605707203e-05,
+      "loss": 1.112,
+      "step": 185
+    },
+    {
+      "epoch": 0.2696629213483146,
+      "grad_norm": 1.2184124246267176,
+      "learning_rate": 1.7137264260631493e-05,
+      "loss": 1.0678,
+      "step": 186
+    },
+    {
+      "epoch": 0.2711127220007249,
+      "grad_norm": 1.3807993935064806,
+      "learning_rate": 1.7104244602119613e-05,
+      "loss": 1.1464,
+      "step": 187
+    },
+    {
+      "epoch": 0.2725625226531352,
+      "grad_norm": 1.3033433008096598,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.1598,
+      "step": 188
+    },
+    {
+      "epoch": 0.2740123233055455,
+      "grad_norm": 1.4039851108939905,
+      "learning_rate": 1.7037734623673616e-05,
+      "loss": 1.1113,
+      "step": 189
+    },
+    {
+      "epoch": 0.2754621239579558,
+      "grad_norm": 1.328439617000884,
+      "learning_rate": 1.7004245774807764e-05,
+      "loss": 1.0514,
+      "step": 190
+    },
+    {
+      "epoch": 0.27691192461036607,
+      "grad_norm": 1.5801690394540369,
+      "learning_rate": 1.697060200597457e-05,
+      "loss": 1.1803,
+      "step": 191
+    },
+    {
+      "epoch": 0.27836172526277636,
+      "grad_norm": 1.4593708782496257,
+      "learning_rate": 1.6936804061307196e-05,
+      "loss": 1.0019,
+      "step": 192
+    },
+    {
+      "epoch": 0.27981152591518665,
+      "grad_norm": 1.5969777513858072,
+      "learning_rate": 1.6902852688348866e-05,
+      "loss": 1.1398,
+      "step": 193
+    },
+    {
+      "epoch": 0.28126132656759695,
+      "grad_norm": 1.333517362517829,
+      "learning_rate": 1.686874863803633e-05,
+      "loss": 1.0863,
+      "step": 194
+    },
+    {
+      "epoch": 0.28271112722000724,
+      "grad_norm": 1.3417123989496027,
+      "learning_rate": 1.6834492664683275e-05,
+      "loss": 1.1305,
+      "step": 195
+    },
+    {
+      "epoch": 0.28416092787241753,
+      "grad_norm": 1.447846262988497,
+      "learning_rate": 1.68000855259636e-05,
+      "loss": 1.1368,
+      "step": 196
+    },
+    {
+      "epoch": 0.28561072852482783,
+      "grad_norm": 1.4040038067546314,
+      "learning_rate": 1.676552798289469e-05,
+      "loss": 1.12,
+      "step": 197
+    },
+    {
+      "epoch": 0.2870605291772381,
+      "grad_norm": 1.4550545349269826,
+      "learning_rate": 1.6730820799820575e-05,
+      "loss": 1.0146,
+      "step": 198
+    },
+    {
+      "epoch": 0.2885103298296484,
+      "grad_norm": 1.3852545950579782,
+      "learning_rate": 1.669596474439503e-05,
+      "loss": 1.0622,
+      "step": 199
+    },
+    {
+      "epoch": 0.2899601304820587,
+      "grad_norm": 1.5296657364269963,
+      "learning_rate": 1.666096058756458e-05,
+      "loss": 1.1951,
+      "step": 200
+    },
+    {
+      "epoch": 0.291409931134469,
+      "grad_norm": 1.415775201154584,
+      "learning_rate": 1.6625809103551462e-05,
+      "loss": 1.0387,
+      "step": 201
+    },
+    {
+      "epoch": 0.2928597317868793,
+      "grad_norm": 1.4821555303643048,
+      "learning_rate": 1.6590511069836498e-05,
+      "loss": 1.1276,
+      "step": 202
+    },
+    {
+      "epoch": 0.2943095324392896,
+      "grad_norm": 1.4067945666222472,
+      "learning_rate": 1.6555067267141896e-05,
+      "loss": 1.092,
+      "step": 203
+    },
+    {
+      "epoch": 0.2957593330916999,
+      "grad_norm": 1.5349351461123342,
+      "learning_rate": 1.651947847941398e-05,
+      "loss": 1.1588,
+      "step": 204
+    },
+    {
+      "epoch": 0.2972091337441102,
+      "grad_norm": 1.5405852117625323,
+      "learning_rate": 1.648374549380587e-05,
+      "loss": 1.0485,
+      "step": 205
+    },
+    {
+      "epoch": 0.29865893439652047,
+      "grad_norm": 1.4113203255730782,
+      "learning_rate": 1.644786910066003e-05,
+      "loss": 1.08,
+      "step": 206
+    },
+    {
+      "epoch": 0.30010873504893076,
+      "grad_norm": 1.3976553320596006,
+      "learning_rate": 1.641185009349084e-05,
+      "loss": 1.0498,
+      "step": 207
+    },
+    {
+      "epoch": 0.30155853570134106,
+      "grad_norm": 1.427267184720538,
+      "learning_rate": 1.637568926896699e-05,
+      "loss": 1.0735,
+      "step": 208
+    },
+    {
+      "epoch": 0.30300833635375135,
+      "grad_norm": 1.4045347476533137,
+      "learning_rate": 1.633938742689392e-05,
+      "loss": 1.0753,
+      "step": 209
+    },
+    {
+      "epoch": 0.30445813700616164,
+      "grad_norm": 1.3716558705276194,
+      "learning_rate": 1.6302945370196074e-05,
+      "loss": 1.1086,
+      "step": 210
+    },
+    {
+      "epoch": 0.30590793765857194,
+      "grad_norm": 1.390291098238023,
+      "learning_rate": 1.626636390489916e-05,
+      "loss": 1.1076,
+      "step": 211
+    },
+    {
+      "epoch": 0.30735773831098223,
+      "grad_norm": 1.490370890156314,
+      "learning_rate": 1.6229643840112355e-05,
+      "loss": 1.1179,
+      "step": 212
+    },
+    {
+      "epoch": 0.3088075389633925,
+      "grad_norm": 1.5052855129764682,
+      "learning_rate": 1.6192785988010354e-05,
+      "loss": 1.1205,
+      "step": 213
+    },
+    {
+      "epoch": 0.3102573396158028,
+      "grad_norm": 1.3812509764847685,
+      "learning_rate": 1.6155791163815438e-05,
+      "loss": 1.0227,
+      "step": 214
+    },
+    {
+      "epoch": 0.3117071402682131,
+      "grad_norm": 1.417234079590203,
+      "learning_rate": 1.6118660185779448e-05,
+      "loss": 1.0496,
+      "step": 215
+    },
+    {
+      "epoch": 0.3131569409206234,
+      "grad_norm": 1.3678381950724512,
+      "learning_rate": 1.6081393875165665e-05,
+      "loss": 1.0668,
+      "step": 216
+    },
+    {
+      "epoch": 0.3146067415730337,
+      "grad_norm": 1.3948780406911645,
+      "learning_rate": 1.604399305623067e-05,
+      "loss": 1.1048,
+      "step": 217
+    },
+    {
+      "epoch": 0.316056542225444,
+      "grad_norm": 1.4227521715408016,
+      "learning_rate": 1.6006458556206087e-05,
+      "loss": 1.0648,
+      "step": 218
+    },
+    {
+      "epoch": 0.3175063428778543,
+      "grad_norm": 1.3417118248283553,
+      "learning_rate": 1.5968791205280306e-05,
+      "loss": 1.0903,
+      "step": 219
+    },
+    {
+      "epoch": 0.3189561435302646,
+      "grad_norm": 1.2920869246864828,
+      "learning_rate": 1.593099183658012e-05,
+      "loss": 1.0687,
+      "step": 220
+    },
+    {
+      "epoch": 0.3204059441826749,
+      "grad_norm": 1.2838376644336658,
+      "learning_rate": 1.5893061286152275e-05,
+      "loss": 1.1088,
+      "step": 221
+    },
+    {
+      "epoch": 0.32185574483508517,
+      "grad_norm": 1.3319412300815356,
+      "learning_rate": 1.585500039294501e-05,
+      "loss": 1.077,
+      "step": 222
+    },
+    {
+      "epoch": 0.32330554548749546,
+      "grad_norm": 1.3244489886752098,
+      "learning_rate": 1.5816809998789493e-05,
+      "loss": 1.0608,
+      "step": 223
+    },
+    {
+      "epoch": 0.32475534613990575,
+      "grad_norm": 1.3159278855429084,
+      "learning_rate": 1.5778490948381175e-05,
+      "loss": 1.1671,
+      "step": 224
+    },
+    {
+      "epoch": 0.32620514679231605,
+      "grad_norm": 1.3721010691162927,
+      "learning_rate": 1.5740044089261137e-05,
+      "loss": 1.1425,
+      "step": 225
+    },
+    {
+      "epoch": 0.32765494744472634,
+      "grad_norm": 1.3809836600711134,
+      "learning_rate": 1.570147027179734e-05,
+      "loss": 0.9912,
+      "step": 226
+    },
+    {
+      "epoch": 0.32910474809713663,
+      "grad_norm": 1.3575066656674317,
+      "learning_rate": 1.5662770349165802e-05,
+      "loss": 1.0461,
+      "step": 227
+    },
+    {
+      "epoch": 0.33055454874954693,
+      "grad_norm": 1.6300239390939768,
+      "learning_rate": 1.562394517733174e-05,
+      "loss": 1.1063,
+      "step": 228
+    },
+    {
+      "epoch": 0.3320043494019572,
+      "grad_norm": 1.261139069504718,
+      "learning_rate": 1.5584995615030634e-05,
+      "loss": 1.0587,
+      "step": 229
+    },
+    {
+      "epoch": 0.3334541500543675,
+      "grad_norm": 1.2938249752112447,
+      "learning_rate": 1.554592252374923e-05,
+      "loss": 1.1205,
+      "step": 230
+    },
+    {
+      "epoch": 0.3349039507067778,
+      "grad_norm": 1.4297940821365565,
+      "learning_rate": 1.55067267677065e-05,
+      "loss": 1.1033,
+      "step": 231
+    },
+    {
+      "epoch": 0.3363537513591881,
+      "grad_norm": 1.364601620445359,
+      "learning_rate": 1.5467409213834496e-05,
+      "loss": 1.0401,
+      "step": 232
+    },
+    {
+      "epoch": 0.3378035520115984,
+      "grad_norm": 1.3778776511917616,
+      "learning_rate": 1.5427970731759214e-05,
+      "loss": 1.1111,
+      "step": 233
+    },
+    {
+      "epoch": 0.3392533526640087,
+      "grad_norm": 1.3273900227776951,
+      "learning_rate": 1.5388412193781334e-05,
+      "loss": 1.0791,
+      "step": 234
+    },
+    {
+      "epoch": 0.340703153316419,
+      "grad_norm": 1.3428607308677851,
+      "learning_rate": 1.5348734474856934e-05,
+      "loss": 1.0846,
+      "step": 235
+    },
+    {
+      "epoch": 0.3421529539688293,
+      "grad_norm": 1.325060473516598,
+      "learning_rate": 1.5308938452578136e-05,
+      "loss": 1.1043,
+      "step": 236
+    },
+    {
+      "epoch": 0.34360275462123957,
+      "grad_norm": 1.384221471184493,
+      "learning_rate": 1.52690250071537e-05,
+      "loss": 1.1038,
+      "step": 237
+    },
+    {
+      "epoch": 0.34505255527364986,
+      "grad_norm": 1.2891773035979663,
+      "learning_rate": 1.5228995021389548e-05,
+      "loss": 1.103,
+      "step": 238
+    },
+    {
+      "epoch": 0.34650235592606016,
+      "grad_norm": 1.3444797515207645,
+      "learning_rate": 1.5188849380669241e-05,
+      "loss": 1.0294,
+      "step": 239
+    },
+    {
+      "epoch": 0.34795215657847045,
+      "grad_norm": 1.4131296703751735,
+      "learning_rate": 1.5148588972934408e-05,
+      "loss": 1.1404,
+      "step": 240
+    },
+    {
+      "epoch": 0.34940195723088074,
+      "grad_norm": 1.3469899260203577,
+      "learning_rate": 1.510821468866508e-05,
+      "loss": 1.1288,
+      "step": 241
+    },
+    {
+      "epoch": 0.35085175788329104,
+      "grad_norm": 1.253623538250413,
+      "learning_rate": 1.5067727420860023e-05,
+      "loss": 1.0128,
+      "step": 242
+    },
+    {
+      "epoch": 0.35230155853570133,
+      "grad_norm": 1.301853583888586,
+      "learning_rate": 1.5027128065016969e-05,
+      "loss": 1.042,
+      "step": 243
+    },
+    {
+      "epoch": 0.3537513591881116,
+      "grad_norm": 1.3628627538362457,
+      "learning_rate": 1.4986417519112813e-05,
+      "loss": 1.0207,
+      "step": 244
+    },
+    {
+      "epoch": 0.3552011598405219,
+      "grad_norm": 1.2832135111549552,
+      "learning_rate": 1.4945596683583754e-05,
+      "loss": 1.0602,
+      "step": 245
+    },
+    {
+      "epoch": 0.3566509604929322,
+      "grad_norm": 1.2696832146247121,
+      "learning_rate": 1.4904666461305377e-05,
+      "loss": 0.9848,
+      "step": 246
+    },
+    {
+      "epoch": 0.3581007611453425,
+      "grad_norm": 1.3731887101569602,
+      "learning_rate": 1.4863627757572687e-05,
+      "loss": 1.0441,
+      "step": 247
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 1.6851171026867449,
+      "learning_rate": 1.4822481480080079e-05,
+      "loss": 1.0833,
+      "step": 248
+    },
+    {
+      "epoch": 0.3610003624501631,
+      "grad_norm": 1.449272413580042,
+      "learning_rate": 1.4781228538901267e-05,
+      "loss": 1.1162,
+      "step": 249
+    },
+    {
+      "epoch": 0.3624501631025734,
+      "grad_norm": 1.4030680298845881,
+      "learning_rate": 1.4739869846469155e-05,
+      "loss": 1.1526,
+      "step": 250
+    },
+    {
+      "epoch": 0.3638999637549837,
+      "grad_norm": 1.278030297665338,
+      "learning_rate": 1.4698406317555652e-05,
+      "loss": 1.0345,
+      "step": 251
+    },
+    {
+      "epoch": 0.365349764407394,
+      "grad_norm": 1.3621191400058899,
+      "learning_rate": 1.465683886925144e-05,
+      "loss": 1.0898,
+      "step": 252
+    },
+    {
+      "epoch": 0.36679956505980427,
+      "grad_norm": 1.3664447386742236,
+      "learning_rate": 1.4615168420945689e-05,
+      "loss": 1.1098,
+      "step": 253
+    },
+    {
+      "epoch": 0.36824936571221456,
+      "grad_norm": 1.3795417613418475,
+      "learning_rate": 1.4573395894305736e-05,
+      "loss": 0.9992,
+      "step": 254
+    },
+    {
+      "epoch": 0.36969916636462485,
+      "grad_norm": 1.411957955986415,
+      "learning_rate": 1.4531522213256675e-05,
+      "loss": 1.0981,
+      "step": 255
+    },
+    {
+      "epoch": 0.37114896701703515,
+      "grad_norm": 1.3405326453225355,
+      "learning_rate": 1.4489548303960931e-05,
+      "loss": 1.03,
+      "step": 256
+    },
+    {
+      "epoch": 0.37259876766944544,
+      "grad_norm": 1.3510436716891028,
+      "learning_rate": 1.4447475094797796e-05,
+      "loss": 1.1063,
+      "step": 257
+    },
+    {
+      "epoch": 0.37404856832185573,
+      "grad_norm": 1.243592704325701,
+      "learning_rate": 1.440530351634286e-05,
+      "loss": 1.1048,
+      "step": 258
+    },
+    {
+      "epoch": 0.37549836897426603,
+      "grad_norm": 1.3889607739727103,
+      "learning_rate": 1.4363034501347457e-05,
+      "loss": 1.0723,
+      "step": 259
+    },
+    {
+      "epoch": 0.3769481696266763,
+      "grad_norm": 1.4745378035989436,
+      "learning_rate": 1.4320668984718013e-05,
+      "loss": 1.0354,
+      "step": 260
+    },
+    {
+      "epoch": 0.3783979702790866,
+      "grad_norm": 1.3193383091133537,
+      "learning_rate": 1.4278207903495388e-05,
+      "loss": 1.0104,
+      "step": 261
+    },
+    {
+      "epoch": 0.3798477709314969,
+      "grad_norm": 1.136432114330158,
+      "learning_rate": 1.4235652196834135e-05,
+      "loss": 1.0372,
+      "step": 262
+    },
+    {
+      "epoch": 0.3812975715839072,
+      "grad_norm": 1.3669025210454944,
+      "learning_rate": 1.4193002805981731e-05,
+      "loss": 1.0218,
+      "step": 263
+    },
+    {
+      "epoch": 0.3827473722363175,
+      "grad_norm": 1.3584552175457463,
+      "learning_rate": 1.4150260674257767e-05,
+      "loss": 1.1254,
+      "step": 264
+    },
+    {
+      "epoch": 0.3841971728887278,
+      "grad_norm": 1.4308621844855682,
+      "learning_rate": 1.4107426747033075e-05,
+      "loss": 1.0708,
+      "step": 265
+    },
+    {
+      "epoch": 0.3856469735411381,
+      "grad_norm": 1.329912139265403,
+      "learning_rate": 1.406450197170882e-05,
+      "loss": 1.0869,
+      "step": 266
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 1.4823511693828864,
+      "learning_rate": 1.4021487297695539e-05,
+      "loss": 1.1072,
+      "step": 267
+    },
+    {
+      "epoch": 0.38854657484595867,
+      "grad_norm": 1.3262884290128323,
+      "learning_rate": 1.3978383676392166e-05,
+      "loss": 1.0828,
+      "step": 268
+    },
+    {
+      "epoch": 0.38999637549836896,
+      "grad_norm": 1.3048792835359517,
+      "learning_rate": 1.3935192061164957e-05,
+      "loss": 1.1389,
+      "step": 269
+    },
+    {
+      "epoch": 0.39144617615077926,
+      "grad_norm": 1.2446188359006123,
+      "learning_rate": 1.3891913407326424e-05,
+      "loss": 1.085,
+      "step": 270
+    },
+    {
+      "epoch": 0.39289597680318955,
+      "grad_norm": 1.2976830208668324,
+      "learning_rate": 1.3848548672114197e-05,
+      "loss": 1.1056,
+      "step": 271
+    },
+    {
+      "epoch": 0.39434577745559984,
+      "grad_norm": 1.584549312644218,
+      "learning_rate": 1.3805098814669857e-05,
+      "loss": 1.1194,
+      "step": 272
+    },
+    {
+      "epoch": 0.39579557810801014,
+      "grad_norm": 1.391880419727745,
+      "learning_rate": 1.3761564796017724e-05,
+      "loss": 1.1039,
+      "step": 273
+    },
+    {
+      "epoch": 0.39724537876042043,
+      "grad_norm": 1.3853285789532175,
+      "learning_rate": 1.3717947579043584e-05,
+      "loss": 1.0969,
+      "step": 274
+    },
+    {
+      "epoch": 0.3986951794128307,
+      "grad_norm": 1.2673655032293596,
+      "learning_rate": 1.3674248128473417e-05,
+      "loss": 1.0619,
+      "step": 275
+    },
+    {
+      "epoch": 0.400144980065241,
+      "grad_norm": 1.3043981298851446,
+      "learning_rate": 1.3630467410852038e-05,
+      "loss": 1.0515,
+      "step": 276
+    },
+    {
+      "epoch": 0.4015947807176513,
+      "grad_norm": 1.3600432920865784,
+      "learning_rate": 1.3586606394521729e-05,
+      "loss": 1.0022,
+      "step": 277
+    },
+    {
+      "epoch": 0.4030445813700616,
+      "grad_norm": 1.531153825103627,
+      "learning_rate": 1.3542666049600822e-05,
+      "loss": 0.9828,
+      "step": 278
+    },
+    {
+      "epoch": 0.4044943820224719,
+      "grad_norm": 1.3547716661330707,
+      "learning_rate": 1.349864734796224e-05,
+      "loss": 1.1231,
+      "step": 279
+    },
+    {
+      "epoch": 0.4059441826748822,
+      "grad_norm": 1.369992377905573,
+      "learning_rate": 1.3454551263211993e-05,
+      "loss": 1.0193,
+      "step": 280
+    },
+    {
+      "epoch": 0.4073939833272925,
+      "grad_norm": 1.425401650352753,
+      "learning_rate": 1.341037877066766e-05,
+      "loss": 1.0688,
+      "step": 281
+    },
+    {
+      "epoch": 0.4088437839797028,
+      "grad_norm": 1.439167458665046,
+      "learning_rate": 1.3366130847336806e-05,
+      "loss": 1.0426,
+      "step": 282
+    },
+    {
+      "epoch": 0.4102935846321131,
+      "grad_norm": 1.3763670515198667,
+      "learning_rate": 1.3321808471895374e-05,
+      "loss": 1.0489,
+      "step": 283
+    },
+    {
+      "epoch": 0.41174338528452337,
+      "grad_norm": 1.2620839776062673,
+      "learning_rate": 1.3277412624666037e-05,
+      "loss": 1.0333,
+      "step": 284
+    },
+    {
+      "epoch": 0.41319318593693366,
+      "grad_norm": 1.316749378731664,
+      "learning_rate": 1.323294428759652e-05,
+      "loss": 1.0245,
+      "step": 285
+    },
+    {
+      "epoch": 0.41464298658934395,
+      "grad_norm": 1.3581220384401822,
+      "learning_rate": 1.3188404444237883e-05,
+      "loss": 1.0587,
+      "step": 286
+    },
+    {
+      "epoch": 0.41609278724175425,
+      "grad_norm": 1.262717615373415,
+      "learning_rate": 1.3143794079722754e-05,
+      "loss": 1.0567,
+      "step": 287
+    },
+    {
+      "epoch": 0.41754258789416454,
+      "grad_norm": 1.2905632909373939,
+      "learning_rate": 1.309911418074355e-05,
+      "loss": 1.1642,
+      "step": 288
+    },
+    {
+      "epoch": 0.41899238854657483,
+      "grad_norm": 1.3568417356919367,
+      "learning_rate": 1.3054365735530666e-05,
+      "loss": 0.9295,
+      "step": 289
+    },
+    {
+      "epoch": 0.42044218919898513,
+      "grad_norm": 1.5062806024352924,
+      "learning_rate": 1.3009549733830582e-05,
+      "loss": 1.1359,
+      "step": 290
+    },
+    {
+      "epoch": 0.4218919898513954,
+      "grad_norm": 1.3148073803108749,
+      "learning_rate": 1.2964667166884005e-05,
+      "loss": 1.0285,
+      "step": 291
+    },
+    {
+      "epoch": 0.4233417905038057,
+      "grad_norm": 1.3664141791369366,
+      "learning_rate": 1.2919719027403935e-05,
+      "loss": 1.0945,
+      "step": 292
+    },
+    {
+      "epoch": 0.424791591156216,
+      "grad_norm": 1.3623861718048031,
+      "learning_rate": 1.2874706309553697e-05,
+      "loss": 1.1435,
+      "step": 293
+    },
+    {
+      "epoch": 0.4262413918086263,
+      "grad_norm": 1.5089869603916821,
+      "learning_rate": 1.2829630008924966e-05,
+      "loss": 1.1331,
+      "step": 294
+    },
+    {
+      "epoch": 0.4276911924610366,
+      "grad_norm": 1.4337412266617349,
+      "learning_rate": 1.2784491122515743e-05,
+      "loss": 1.0074,
+      "step": 295
+    },
+    {
+      "epoch": 0.4291409931134469,
+      "grad_norm": 1.3856703534273527,
+      "learning_rate": 1.2739290648708298e-05,
+      "loss": 1.0423,
+      "step": 296
+    },
+    {
+      "epoch": 0.4305907937658572,
+      "grad_norm": 1.2957355692305799,
+      "learning_rate": 1.2694029587247097e-05,
+      "loss": 1.0455,
+      "step": 297
+    },
+    {
+      "epoch": 0.4320405944182675,
+      "grad_norm": 1.3273638964130718,
+      "learning_rate": 1.2648708939216671e-05,
+      "loss": 1.0392,
+      "step": 298
+    },
+    {
+      "epoch": 0.43349039507067777,
+      "grad_norm": 1.3353775796295373,
+      "learning_rate": 1.260332970701951e-05,
+      "loss": 1.098,
+      "step": 299
+    },
+    {
+      "epoch": 0.43494019572308806,
+      "grad_norm": 1.3923420528186754,
+      "learning_rate": 1.2557892894353848e-05,
+      "loss": 1.0816,
+      "step": 300
+    },
+    {
+      "epoch": 0.43638999637549836,
+      "grad_norm": 1.295717729844177,
+      "learning_rate": 1.251239950619149e-05,
+      "loss": 1.1171,
+      "step": 301
+    },
+    {
+      "epoch": 0.43783979702790865,
+      "grad_norm": 1.5410442780345024,
+      "learning_rate": 1.2466850548755586e-05,
+      "loss": 1.029,
+      "step": 302
+    },
+    {
+      "epoch": 0.43928959768031894,
+      "grad_norm": 1.4875151003850813,
+      "learning_rate": 1.242124702949836e-05,
+      "loss": 1.0339,
+      "step": 303
+    },
+    {
+      "epoch": 0.44073939833272924,
+      "grad_norm": 1.3539538663690127,
+      "learning_rate": 1.237558995707884e-05,
+      "loss": 1.0096,
+      "step": 304
+    },
+    {
+      "epoch": 0.44218919898513953,
+      "grad_norm": 1.4167885844016397,
+      "learning_rate": 1.2329880341340533e-05,
+      "loss": 1.0544,
+      "step": 305
+    },
+    {
+      "epoch": 0.4436389996375498,
+      "grad_norm": 1.3665806009146937,
+      "learning_rate": 1.2284119193289114e-05,
+      "loss": 1.0517,
+      "step": 306
+    },
+    {
+      "epoch": 0.4450888002899601,
+      "grad_norm": 1.3906583520168478,
+      "learning_rate": 1.223830752507004e-05,
+      "loss": 1.0786,
+      "step": 307
+    },
+    {
+      "epoch": 0.4465386009423704,
+      "grad_norm": 1.4145142571256235,
+      "learning_rate": 1.2192446349946181e-05,
+      "loss": 1.0063,
+      "step": 308
+    },
+    {
+      "epoch": 0.4479884015947807,
+      "grad_norm": 1.3760850981686388,
+      "learning_rate": 1.2146536682275388e-05,
+      "loss": 1.0491,
+      "step": 309
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 1.4448528854740226,
+      "learning_rate": 1.2100579537488084e-05,
+      "loss": 0.9365,
+      "step": 310
+    },
+    {
+      "epoch": 0.4508880028996013,
+      "grad_norm": 1.5113803236004408,
+      "learning_rate": 1.2054575932064787e-05,
+      "loss": 1.1046,
+      "step": 311
+    },
+    {
+      "epoch": 0.4523378035520116,
+      "grad_norm": 1.2736460404311223,
+      "learning_rate": 1.200852688351363e-05,
+      "loss": 1.0819,
+      "step": 312
+    },
+    {
+      "epoch": 0.4537876042044219,
+      "grad_norm": 1.414714249817259,
+      "learning_rate": 1.196243341034786e-05,
+      "loss": 1.0603,
+      "step": 313
+    },
+    {
+      "epoch": 0.4552374048568322,
+      "grad_norm": 1.398538630461494,
+      "learning_rate": 1.1916296532063308e-05,
+      "loss": 1.097,
+      "step": 314
+    },
+    {
+      "epoch": 0.45668720550924247,
+      "grad_norm": 1.3291137395378012,
+      "learning_rate": 1.1870117269115833e-05,
+      "loss": 1.121,
+      "step": 315
+    },
+    {
+      "epoch": 0.45813700616165276,
+      "grad_norm": 1.3241885081894262,
+      "learning_rate": 1.1823896642898769e-05,
+      "loss": 1.0995,
+      "step": 316
+    },
+    {
+      "epoch": 0.45958680681406305,
+      "grad_norm": 1.3294989059763207,
+      "learning_rate": 1.1777635675720313e-05,
+      "loss": 1.0288,
+      "step": 317
+    },
+    {
+      "epoch": 0.46103660746647335,
+      "grad_norm": 1.4556611440168887,
+      "learning_rate": 1.1731335390780935e-05,
+      "loss": 1.0542,
+      "step": 318
+    },
+    {
+      "epoch": 0.46248640811888364,
+      "grad_norm": 1.5094674955010488,
+      "learning_rate": 1.1684996812150726e-05,
+      "loss": 1.098,
+      "step": 319
+    },
+    {
+      "epoch": 0.46393620877129393,
+      "grad_norm": 1.4287115077156827,
+      "learning_rate": 1.1638620964746761e-05,
+      "loss": 1.0988,
+      "step": 320
+    },
+    {
+      "epoch": 0.4653860094237042,
+      "grad_norm": 1.3690573378202604,
+      "learning_rate": 1.1592208874310426e-05,
+      "loss": 1.0515,
+      "step": 321
+    },
+    {
+      "epoch": 0.4668358100761145,
+      "grad_norm": 1.2792451481134444,
+      "learning_rate": 1.1545761567384725e-05,
+      "loss": 0.9956,
+      "step": 322
+    },
+    {
+      "epoch": 0.4682856107285248,
+      "grad_norm": 1.3308204615991506,
+      "learning_rate": 1.1499280071291596e-05,
+      "loss": 1.0173,
+      "step": 323
+    },
+    {
+      "epoch": 0.4697354113809351,
+      "grad_norm": 1.2185429537197658,
+      "learning_rate": 1.1452765414109155e-05,
+      "loss": 1.1213,
+      "step": 324
+    },
+    {
+      "epoch": 0.4711852120333454,
+      "grad_norm": 1.277184576403942,
+      "learning_rate": 1.1406218624648986e-05,
+      "loss": 1.0507,
+      "step": 325
+    },
+    {
+      "epoch": 0.4726350126857557,
+      "grad_norm": 1.4310484869488227,
+      "learning_rate": 1.1359640732433369e-05,
+      "loss": 0.9771,
+      "step": 326
+    },
+    {
+      "epoch": 0.474084813338166,
+      "grad_norm": 1.1950443794527996,
+      "learning_rate": 1.1313032767672525e-05,
+      "loss": 1.0998,
+      "step": 327
+    },
+    {
+      "epoch": 0.4755346139905763,
+      "grad_norm": 1.245928965518204,
+      "learning_rate": 1.1266395761241809e-05,
+      "loss": 1.061,
+      "step": 328
+    },
+    {
+      "epoch": 0.4769844146429866,
+      "grad_norm": 1.2273640505796817,
+      "learning_rate": 1.1219730744658921e-05,
+      "loss": 0.9998,
+      "step": 329
+    },
+    {
+      "epoch": 0.47843421529539687,
+      "grad_norm": 1.224320523959283,
+      "learning_rate": 1.1173038750061099e-05,
+      "loss": 0.9796,
+      "step": 330
+    },
+    {
+      "epoch": 0.47988401594780716,
+      "grad_norm": 1.4066002839506198,
+      "learning_rate": 1.1126320810182271e-05,
+      "loss": 1.056,
+      "step": 331
+    },
+    {
+      "epoch": 0.48133381660021746,
+      "grad_norm": 1.380462021021754,
+      "learning_rate": 1.107957795833023e-05,
+      "loss": 1.0188,
+      "step": 332
+    },
+    {
+      "epoch": 0.48278361725262775,
+      "grad_norm": 1.2611059126132427,
+      "learning_rate": 1.1032811228363766e-05,
+      "loss": 0.964,
+      "step": 333
+    },
+    {
+      "epoch": 0.48423341790503804,
+      "grad_norm": 1.3213161598546017,
+      "learning_rate": 1.0986021654669807e-05,
+      "loss": 1.0228,
+      "step": 334
+    },
+    {
+      "epoch": 0.48568321855744834,
+      "grad_norm": 1.3244687898493506,
+      "learning_rate": 1.0939210272140545e-05,
+      "loss": 1.1905,
+      "step": 335
+    },
+    {
+      "epoch": 0.48713301920985863,
+      "grad_norm": 1.2657091226561747,
+      "learning_rate": 1.089237811615053e-05,
+      "loss": 1.0406,
+      "step": 336
+    },
+    {
+      "epoch": 0.4885828198622689,
+      "grad_norm": 1.401377067365228,
+      "learning_rate": 1.0845526222533791e-05,
+      "loss": 1.012,
+      "step": 337
+    },
+    {
+      "epoch": 0.4900326205146792,
+      "grad_norm": 1.3802477728078786,
+      "learning_rate": 1.07986556275609e-05,
+      "loss": 0.9791,
+      "step": 338
+    },
+    {
+      "epoch": 0.4914824211670895,
+      "grad_norm": 1.4280870612361602,
+      "learning_rate": 1.0751767367916081e-05,
+      "loss": 1.0664,
+      "step": 339
+    },
+    {
+      "epoch": 0.4929322218194998,
+      "grad_norm": 1.419969378682578,
+      "learning_rate": 1.0704862480674251e-05,
+      "loss": 1.0491,
+      "step": 340
+    },
+    {
+      "epoch": 0.4943820224719101,
+      "grad_norm": 1.3105839725915047,
+      "learning_rate": 1.0657942003278107e-05,
+      "loss": 1.0062,
+      "step": 341
+    },
+    {
+      "epoch": 0.4958318231243204,
+      "grad_norm": 1.1983146147450168,
+      "learning_rate": 1.0611006973515165e-05,
+      "loss": 1.0111,
+      "step": 342
+    },
+    {
+      "epoch": 0.4972816237767307,
+      "grad_norm": 1.2738558388342,
+      "learning_rate": 1.0564058429494808e-05,
+      "loss": 1.0729,
+      "step": 343
+    },
+    {
+      "epoch": 0.498731424429141,
+      "grad_norm": 1.33934602814329,
+      "learning_rate": 1.0517097409625335e-05,
+      "loss": 1.1268,
+      "step": 344
+    },
+    {
+      "epoch": 0.5001812250815513,
+      "grad_norm": 1.55583317754474,
+      "learning_rate": 1.0470124952590976e-05,
+      "loss": 1.0271,
+      "step": 345
+    },
+    {
+      "epoch": 0.5016310257339616,
+      "grad_norm": 1.3148573597296123,
+      "learning_rate": 1.0423142097328938e-05,
+      "loss": 1.0349,
+      "step": 346
+    },
+    {
+      "epoch": 0.5016310257339616,
+      "eval_loss": 1.8077555894851685,
+      "eval_runtime": 336.9133,
+      "eval_samples_per_second": 52.399,
+      "eval_steps_per_second": 0.819,
+      "step": 346
+    },
+    {
+      "epoch": 0.5030808263863719,
+      "grad_norm": 1.4420727552751345,
+      "learning_rate": 1.0376149883006406e-05,
+      "loss": 1.065,
+      "step": 347
+    },
+    {
+      "epoch": 0.5045306270387822,
+      "grad_norm": 1.3554631021911403,
+      "learning_rate": 1.0329149348997576e-05,
+      "loss": 1.061,
+      "step": 348
+    },
+    {
+      "epoch": 0.5059804276911924,
+      "grad_norm": 1.2520888233774288,
+      "learning_rate": 1.0282141534860662e-05,
+      "loss": 1.0037,
+      "step": 349
+    },
+    {
+      "epoch": 0.5074302283436027,
+      "grad_norm": 1.331415995629297,
+      "learning_rate": 1.0235127480314888e-05,
+      "loss": 1.0856,
+      "step": 350
+    },
+    {
+      "epoch": 0.508880028996013,
+      "grad_norm": 1.2376496795375593,
+      "learning_rate": 1.0188108225217516e-05,
+      "loss": 1.06,
+      "step": 351
+    },
+    {
+      "epoch": 0.5103298296484233,
+      "grad_norm": 1.4683795988238608,
+      "learning_rate": 1.0141084809540826e-05,
+      "loss": 1.1097,
+      "step": 352
+    },
+    {
+      "epoch": 0.5117796303008336,
+      "grad_norm": 1.2758662790468451,
+      "learning_rate": 1.0094058273349125e-05,
+      "loss": 0.999,
+      "step": 353
+    },
+    {
+      "epoch": 0.5132294309532439,
+      "grad_norm": 1.4234236066999306,
+      "learning_rate": 1.0047029656775735e-05,
+      "loss": 1.0407,
+      "step": 354
+    },
+    {
+      "epoch": 0.5146792316056542,
+      "grad_norm": 1.2837571061936146,
+      "learning_rate": 1e-05,
+      "loss": 1.0489,
+      "step": 355
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 1.3811452244201645,
+      "learning_rate": 9.952970343224268e-06,
+      "loss": 1.0266,
+      "step": 356
+    },
+    {
+      "epoch": 0.5175788329104748,
+      "grad_norm": 1.3706746656509412,
+      "learning_rate": 9.90594172665088e-06,
+      "loss": 1.0079,
+      "step": 357
+    },
+    {
+      "epoch": 0.5190286335628851,
+      "grad_norm": 1.3611467941537612,
+      "learning_rate": 9.858915190459179e-06,
+      "loss": 0.9805,
+      "step": 358
+    },
+    {
+      "epoch": 0.5204784342152954,
+      "grad_norm": 1.3014493335149024,
+      "learning_rate": 9.811891774782484e-06,
+      "loss": 1.0645,
+      "step": 359
+    },
+    {
+      "epoch": 0.5219282348677057,
+      "grad_norm": 1.3032059031448693,
+      "learning_rate": 9.764872519685112e-06,
+      "loss": 0.9783,
+      "step": 360
+    },
+    {
+      "epoch": 0.523378035520116,
+      "grad_norm": 1.4070425505827868,
+      "learning_rate": 9.717858465139341e-06,
+      "loss": 1.1419,
+      "step": 361
+    },
+    {
+      "epoch": 0.5248278361725263,
+      "grad_norm": 1.3033399285027985,
+      "learning_rate": 9.670850651002427e-06,
+      "loss": 1.0435,
+      "step": 362
+    },
+    {
+      "epoch": 0.5262776368249366,
+      "grad_norm": 1.1432394869236526,
+      "learning_rate": 9.623850116993598e-06,
+      "loss": 0.9418,
+      "step": 363
+    },
+    {
+      "epoch": 0.5277274374773469,
+      "grad_norm": 1.3277004963085377,
+      "learning_rate": 9.576857902671067e-06,
+      "loss": 1.0301,
+      "step": 364
+    },
+    {
+      "epoch": 0.5291772381297571,
+      "grad_norm": 1.421084621129887,
+      "learning_rate": 9.529875047409027e-06,
+      "loss": 1.0226,
+      "step": 365
+    },
+    {
+      "epoch": 0.5306270387821674,
+      "grad_norm": 1.319851633356501,
+      "learning_rate": 9.482902590374665e-06,
+      "loss": 1.0536,
+      "step": 366
+    },
+    {
+      "epoch": 0.5320768394345777,
+      "grad_norm": 1.4423863953216414,
+      "learning_rate": 9.435941570505192e-06,
+      "loss": 1.0893,
+      "step": 367
+    },
+    {
+      "epoch": 0.533526640086988,
+      "grad_norm": 1.3246949110488309,
+      "learning_rate": 9.388993026484837e-06,
+      "loss": 1.0841,
+      "step": 368
+    },
+    {
+      "epoch": 0.5349764407393983,
+      "grad_norm": 1.3854695000429353,
+      "learning_rate": 9.342057996721894e-06,
+      "loss": 1.1334,
+      "step": 369
+    },
+    {
+      "epoch": 0.5364262413918086,
+      "grad_norm": 1.2888442788606653,
+      "learning_rate": 9.295137519325752e-06,
+      "loss": 1.1002,
+      "step": 370
+    },
+    {
+      "epoch": 0.5378760420442189,
+      "grad_norm": 1.2938379820040973,
+      "learning_rate": 9.248232632083924e-06,
+      "loss": 0.9898,
+      "step": 371
+    },
+    {
+      "epoch": 0.5393258426966292,
+      "grad_norm": 1.2857643228124016,
+      "learning_rate": 9.201344372439103e-06,
+      "loss": 1.0287,
+      "step": 372
+    },
+    {
+      "epoch": 0.5407756433490395,
+      "grad_norm": 1.3161835381883626,
+      "learning_rate": 9.154473777466209e-06,
+      "loss": 1.0638,
+      "step": 373
+    },
+    {
+      "epoch": 0.5422254440014498,
+      "grad_norm": 1.2637441555203863,
+      "learning_rate": 9.10762188384947e-06,
+      "loss": 1.0539,
+      "step": 374
+    },
+    {
+      "epoch": 0.5436752446538601,
+      "grad_norm": 1.385080112544529,
+      "learning_rate": 9.060789727859457e-06,
+      "loss": 1.0409,
+      "step": 375
+    },
+    {
+      "epoch": 0.5451250453062704,
+      "grad_norm": 1.2336081724917547,
+      "learning_rate": 9.013978345330195e-06,
+      "loss": 1.0041,
+      "step": 376
+    },
+    {
+      "epoch": 0.5465748459586807,
+      "grad_norm": 1.2772783658027675,
+      "learning_rate": 8.967188771636237e-06,
+      "loss": 1.0687,
+      "step": 377
+    },
+    {
+      "epoch": 0.548024646611091,
+      "grad_norm": 1.5360228274930754,
+      "learning_rate": 8.920422041669774e-06,
+      "loss": 1.1388,
+      "step": 378
+    },
+    {
+      "epoch": 0.5494744472635013,
+      "grad_norm": 1.3616854601791177,
+      "learning_rate": 8.873679189817732e-06,
+      "loss": 1.0475,
+      "step": 379
+    },
+    {
+      "epoch": 0.5509242479159115,
+      "grad_norm": 1.411219093599177,
+      "learning_rate": 8.826961249938903e-06,
+      "loss": 1.0124,
+      "step": 380
+    },
+    {
+      "epoch": 0.5523740485683218,
+      "grad_norm": 1.526783750017392,
+      "learning_rate": 8.78026925534108e-06,
+      "loss": 1.0708,
+      "step": 381
+    },
+    {
+      "epoch": 0.5538238492207321,
+      "grad_norm": 1.414663897048244,
+      "learning_rate": 8.733604238758196e-06,
+      "loss": 1.0688,
+      "step": 382
+    },
+    {
+      "epoch": 0.5552736498731424,
+      "grad_norm": 1.383560239131,
+      "learning_rate": 8.68696723232748e-06,
+      "loss": 1.0006,
+      "step": 383
+    },
+    {
+      "epoch": 0.5567234505255527,
+      "grad_norm": 1.3339859174485575,
+      "learning_rate": 8.640359267566633e-06,
+      "loss": 1.0343,
+      "step": 384
+    },
+    {
+      "epoch": 0.558173251177963,
+      "grad_norm": 1.3213252031457365,
+      "learning_rate": 8.59378137535102e-06,
+      "loss": 1.0621,
+      "step": 385
+    },
+    {
+      "epoch": 0.5596230518303733,
+      "grad_norm": 1.3905146202204568,
+      "learning_rate": 8.54723458589085e-06,
+      "loss": 1.1255,
+      "step": 386
+    },
+    {
+      "epoch": 0.5610728524827836,
+      "grad_norm": 1.2307848968985167,
+      "learning_rate": 8.500719928708406e-06,
+      "loss": 0.9469,
+      "step": 387
+    },
+    {
+      "epoch": 0.5625226531351939,
+      "grad_norm": 1.3532274888219538,
+      "learning_rate": 8.454238432615275e-06,
+      "loss": 1.0236,
+      "step": 388
+    },
+    {
+      "epoch": 0.5639724537876042,
+      "grad_norm": 1.347137778188515,
+      "learning_rate": 8.407791125689577e-06,
+      "loss": 0.9966,
+      "step": 389
+    },
+    {
+      "epoch": 0.5654222544400145,
+      "grad_norm": 1.4652711818595845,
+      "learning_rate": 8.36137903525324e-06,
+      "loss": 1.0502,
+      "step": 390
+    },
+    {
+      "epoch": 0.5668720550924248,
+      "grad_norm": 1.3859992879047083,
+      "learning_rate": 8.315003187849276e-06,
+      "loss": 1.0871,
+      "step": 391
+    },
+    {
+      "epoch": 0.5683218557448351,
+      "grad_norm": 1.4891799803892296,
+      "learning_rate": 8.268664609219066e-06,
+      "loss": 1.0776,
+      "step": 392
+    },
+    {
+      "epoch": 0.5697716563972454,
+      "grad_norm": 1.3461111269934627,
+      "learning_rate": 8.222364324279689e-06,
+      "loss": 1.0634,
+      "step": 393
+    },
+    {
+      "epoch": 0.5712214570496557,
+      "grad_norm": 1.2888958997984548,
+      "learning_rate": 8.176103357101235e-06,
+      "loss": 1.0858,
+      "step": 394
+    },
+    {
+      "epoch": 0.572671257702066,
+      "grad_norm": 1.3769762527219886,
+      "learning_rate": 8.129882730884168e-06,
+      "loss": 1.0497,
+      "step": 395
+    },
+    {
+      "epoch": 0.5741210583544762,
+      "grad_norm": 1.5311916598990292,
+      "learning_rate": 8.083703467936696e-06,
+      "loss": 0.991,
+      "step": 396
+    },
+    {
+      "epoch": 0.5755708590068865,
+      "grad_norm": 1.349922244593328,
+      "learning_rate": 8.037566589652141e-06,
+      "loss": 1.0734,
+      "step": 397
+    },
+    {
+      "epoch": 0.5770206596592968,
+      "grad_norm": 1.3613957765937217,
+      "learning_rate": 7.991473116486372e-06,
+      "loss": 1.0152,
+      "step": 398
+    },
+    {
+      "epoch": 0.5784704603117071,
+      "grad_norm": 1.3864683299562783,
+      "learning_rate": 7.945424067935216e-06,
+      "loss": 1.0369,
+      "step": 399
+    },
+    {
+      "epoch": 0.5799202609641174,
+      "grad_norm": 1.33588586682402,
+      "learning_rate": 7.899420462511921e-06,
+      "loss": 1.0445,
+      "step": 400
+    },
+    {
+      "epoch": 0.5813700616165277,
+      "grad_norm": 1.3840089916166154,
+      "learning_rate": 7.853463317724614e-06,
+      "loss": 0.9492,
+      "step": 401
+    },
+    {
+      "epoch": 0.582819862268938,
+      "grad_norm": 1.6496515850287372,
+      "learning_rate": 7.807553650053824e-06,
+      "loss": 1.0063,
+      "step": 402
+    },
+    {
+      "epoch": 0.5842696629213483,
+      "grad_norm": 1.2683362072535584,
+      "learning_rate": 7.761692474929961e-06,
+      "loss": 0.9942,
+      "step": 403
+    },
+    {
+      "epoch": 0.5857194635737586,
+      "grad_norm": 1.476093798236584,
+      "learning_rate": 7.715880806710888e-06,
+      "loss": 1.0139,
+      "step": 404
+    },
+    {
+      "epoch": 0.5871692642261689,
+      "grad_norm": 1.3832527319586156,
+      "learning_rate": 7.670119658659469e-06,
+      "loss": 1.0288,
+      "step": 405
+    },
+    {
+      "epoch": 0.5886190648785792,
+      "grad_norm": 1.337205386873013,
+      "learning_rate": 7.6244100429211644e-06,
+      "loss": 1.111,
+      "step": 406
+    },
+    {
+      "epoch": 0.5900688655309895,
+      "grad_norm": 1.3007683016486016,
+      "learning_rate": 7.578752970501641e-06,
+      "loss": 1.0501,
+      "step": 407
+    },
+    {
+      "epoch": 0.5915186661833998,
+      "grad_norm": 1.4326315488897896,
+      "learning_rate": 7.533149451244416e-06,
+      "loss": 1.0034,
+      "step": 408
+    },
+    {
+      "epoch": 0.5929684668358101,
+      "grad_norm": 1.2900230806035275,
+      "learning_rate": 7.487600493808513e-06,
+      "loss": 1.0064,
+      "step": 409
+    },
+    {
+      "epoch": 0.5944182674882204,
+      "grad_norm": 1.2812198102736636,
+      "learning_rate": 7.442107105646156e-06,
+      "loss": 1.0651,
+      "step": 410
+    },
+    {
+      "epoch": 0.5958680681406306,
+      "grad_norm": 1.3023676156417663,
+      "learning_rate": 7.396670292980492e-06,
+      "loss": 1.0598,
+      "step": 411
+    },
+    {
+      "epoch": 0.5973178687930409,
+      "grad_norm": 1.4221868423370116,
+      "learning_rate": 7.35129106078333e-06,
+      "loss": 1.0526,
+      "step": 412
+    },
+    {
+      "epoch": 0.5987676694454512,
+      "grad_norm": 1.2363756368908436,
+      "learning_rate": 7.30597041275291e-06,
+      "loss": 1.0231,
+      "step": 413
+    },
+    {
+      "epoch": 0.6002174700978615,
+      "grad_norm": 1.3100743091337919,
+      "learning_rate": 7.260709351291706e-06,
+      "loss": 1.079,
+      "step": 414
+    },
+    {
+      "epoch": 0.6016672707502718,
+      "grad_norm": 1.3559391633912112,
+      "learning_rate": 7.215508877484259e-06,
+      "loss": 0.9683,
+      "step": 415
+    },
+    {
+      "epoch": 0.6031170714026821,
+      "grad_norm": 1.2391397762843972,
+      "learning_rate": 7.170369991075035e-06,
+      "loss": 1.0038,
+      "step": 416
+    },
+    {
+      "epoch": 0.6045668720550924,
+      "grad_norm": 1.2805033031465005,
+      "learning_rate": 7.125293690446307e-06,
+      "loss": 1.1191,
+      "step": 417
+    },
+    {
+      "epoch": 0.6060166727075027,
+      "grad_norm": 1.2590613584644856,
+      "learning_rate": 7.08028097259607e-06,
+      "loss": 0.9602,
+      "step": 418
+    },
+    {
+      "epoch": 0.607466473359913,
+      "grad_norm": 1.2467561330796926,
+      "learning_rate": 7.035332833115998e-06,
+      "loss": 1.0396,
+      "step": 419
+    },
+    {
+      "epoch": 0.6089162740123233,
+      "grad_norm": 1.356583060972714,
+      "learning_rate": 6.990450266169424e-06,
+      "loss": 0.9845,
+      "step": 420
+    },
+    {
+      "epoch": 0.6103660746647336,
+      "grad_norm": 1.4368300916568508,
+      "learning_rate": 6.945634264469338e-06,
+      "loss": 1.0536,
+      "step": 421
+    },
+    {
+      "epoch": 0.6118158753171439,
+      "grad_norm": 1.4719432988529924,
+      "learning_rate": 6.900885819256449e-06,
+      "loss": 1.0767,
+      "step": 422
+    },
+    {
+      "epoch": 0.6132656759695542,
+      "grad_norm": 1.5202820701276154,
+      "learning_rate": 6.85620592027725e-06,
+      "loss": 1.0381,
+      "step": 423
+    },
+    {
+      "epoch": 0.6147154766219645,
+      "grad_norm": 1.2725254628375995,
+      "learning_rate": 6.811595555762118e-06,
+      "loss": 0.9977,
+      "step": 424
+    },
+    {
+      "epoch": 0.6161652772743748,
+      "grad_norm": 1.26379238811477,
+      "learning_rate": 6.767055712403481e-06,
+      "loss": 0.9419,
+      "step": 425
+    },
+    {
+      "epoch": 0.617615077926785,
+      "grad_norm": 1.292865030162402,
+      "learning_rate": 6.722587375333967e-06,
+      "loss": 1.0333,
+      "step": 426
+    },
+    {
+      "epoch": 0.6190648785791953,
+      "grad_norm": 1.3408000558059614,
+      "learning_rate": 6.678191528104629e-06,
+      "loss": 1.0425,
+      "step": 427
+    },
+    {
+      "epoch": 0.6205146792316056,
+      "grad_norm": 1.5043339237538254,
+      "learning_rate": 6.633869152663197e-06,
+      "loss": 1.0759,
+      "step": 428
+    },
+    {
+      "epoch": 0.6219644798840159,
+      "grad_norm": 1.2406316576092251,
+      "learning_rate": 6.58962122933234e-06,
+      "loss": 0.9408,
+      "step": 429
+    },
+    {
+      "epoch": 0.6234142805364262,
+      "grad_norm": 1.3059983607773458,
+      "learning_rate": 6.545448736788009e-06,
+      "loss": 1.0824,
+      "step": 430
+    },
+    {
+      "epoch": 0.6248640811888365,
+      "grad_norm": 1.1905714141479788,
+      "learning_rate": 6.501352652037765e-06,
+      "loss": 0.9091,
+      "step": 431
+    },
+    {
+      "epoch": 0.6263138818412468,
+      "grad_norm": 1.3706236408019423,
+      "learning_rate": 6.457333950399181e-06,
+      "loss": 0.9951,
+      "step": 432
+    },
+    {
+      "epoch": 0.6277636824936571,
+      "grad_norm": 1.3464812813758535,
+      "learning_rate": 6.413393605478275e-06,
+      "loss": 1.0083,
+      "step": 433
+    },
+    {
+      "epoch": 0.6292134831460674,
+      "grad_norm": 1.5482620161549068,
+      "learning_rate": 6.369532589147967e-06,
+      "loss": 1.0843,
+      "step": 434
+    },
+    {
+      "epoch": 0.6306632837984777,
+      "grad_norm": 1.2642215211811383,
+      "learning_rate": 6.325751871526589e-06,
+      "loss": 0.9379,
+      "step": 435
+    },
+    {
+      "epoch": 0.632113084450888,
+      "grad_norm": 1.4177584029745869,
+      "learning_rate": 6.282052420956417e-06,
+      "loss": 0.9872,
+      "step": 436
+    },
+    {
+      "epoch": 0.6335628851032983,
+      "grad_norm": 1.4271400868991369,
+      "learning_rate": 6.238435203982278e-06,
+      "loss": 1.0785,
+      "step": 437
+    },
+    {
+      "epoch": 0.6350126857557086,
+      "grad_norm": 1.2604541563479463,
+      "learning_rate": 6.194901185330145e-06,
+      "loss": 1.0321,
+      "step": 438
+    },
+    {
+      "epoch": 0.6364624864081189,
+      "grad_norm": 1.4292504789000968,
+      "learning_rate": 6.1514513278858055e-06,
+      "loss": 0.9982,
+      "step": 439
+    },
+    {
+      "epoch": 0.6379122870605292,
+      "grad_norm": 1.3655368575149138,
+      "learning_rate": 6.1080865926735814e-06,
+      "loss": 1.0321,
+      "step": 440
+    },
+    {
+      "epoch": 0.6393620877129395,
+      "grad_norm": 1.2547190564362218,
+      "learning_rate": 6.064807938835046e-06,
+      "loss": 1.076,
+      "step": 441
+    },
+    {
+      "epoch": 0.6408118883653497,
+      "grad_norm": 1.3637441454912735,
+      "learning_rate": 6.021616323607834e-06,
+      "loss": 1.0778,
+      "step": 442
+    },
+    {
+      "epoch": 0.64226168901776,
+      "grad_norm": 1.3191737075842687,
+      "learning_rate": 5.978512702304461e-06,
+      "loss": 1.0044,
+      "step": 443
+    },
+    {
+      "epoch": 0.6437114896701703,
+      "grad_norm": 1.4358907567418602,
+      "learning_rate": 5.935498028291183e-06,
+      "loss": 1.0427,
+      "step": 444
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 1.3799701766605277,
+      "learning_rate": 5.892573252966926e-06,
+      "loss": 1.0372,
+      "step": 445
+    },
+    {
+      "epoch": 0.6466110909749909,
+      "grad_norm": 1.3882909807578712,
+      "learning_rate": 5.849739325742234e-06,
+      "loss": 1.0585,
+      "step": 446
+    },
+    {
+      "epoch": 0.6480608916274012,
+      "grad_norm": 1.243421259508741,
+      "learning_rate": 5.806997194018272e-06,
+      "loss": 1.0131,
+      "step": 447
+    },
+    {
+      "epoch": 0.6495106922798115,
+      "grad_norm": 1.384906012737019,
+      "learning_rate": 5.7643478031658676e-06,
+      "loss": 0.982,
+      "step": 448
+    },
+    {
+      "epoch": 0.6509604929322218,
+      "grad_norm": 1.2861293995098508,
+      "learning_rate": 5.7217920965046115e-06,
+      "loss": 1.0076,
+      "step": 449
+    },
+    {
+      "epoch": 0.6524102935846321,
+      "grad_norm": 1.2342057170927183,
+      "learning_rate": 5.679331015281984e-06,
+      "loss": 0.959,
+      "step": 450
+    },
+    {
+      "epoch": 0.6538600942370424,
+      "grad_norm": 1.2769596100810268,
+      "learning_rate": 5.636965498652544e-06,
+      "loss": 0.9909,
+      "step": 451
+    },
+    {
+      "epoch": 0.6553098948894527,
+      "grad_norm": 1.284999585086877,
+      "learning_rate": 5.594696483657138e-06,
+      "loss": 1.0122,
+      "step": 452
+    },
+    {
+      "epoch": 0.656759695541863,
+      "grad_norm": 1.3775476328168168,
+      "learning_rate": 5.552524905202208e-06,
+      "loss": 1.0395,
+      "step": 453
+    },
+    {
+      "epoch": 0.6582094961942733,
+      "grad_norm": 1.3958141983305359,
+      "learning_rate": 5.5104516960390695e-06,
+      "loss": 1.1147,
+      "step": 454
+    },
+    {
+      "epoch": 0.6596592968466836,
+      "grad_norm": 1.1897636206633544,
+      "learning_rate": 5.468477786743334e-06,
+      "loss": 1.0003,
+      "step": 455
+    },
+    {
+      "epoch": 0.6611090974990939,
+      "grad_norm": 1.401760802014749,
+      "learning_rate": 5.4266041056942644e-06,
+      "loss": 0.9985,
+      "step": 456
+    },
+    {
+      "epoch": 0.6625588981515041,
+      "grad_norm": 1.318245423393006,
+      "learning_rate": 5.384831579054313e-06,
+      "loss": 1.0835,
+      "step": 457
+    },
+    {
+      "epoch": 0.6640086988039144,
+      "grad_norm": 1.2156706519215021,
+      "learning_rate": 5.343161130748564e-06,
+      "loss": 1.0706,
+      "step": 458
+    },
+    {
+      "epoch": 0.6654584994563247,
+      "grad_norm": 1.2667934969094834,
+      "learning_rate": 5.3015936824443526e-06,
+      "loss": 0.988,
+      "step": 459
+    },
+    {
+      "epoch": 0.666908300108735,
+      "grad_norm": 1.448796527195574,
+      "learning_rate": 5.260130153530847e-06,
+      "loss": 0.9986,
+      "step": 460
+    },
+    {
+      "epoch": 0.6683581007611453,
+      "grad_norm": 1.3009390751080618,
+      "learning_rate": 5.218771461098733e-06,
+      "loss": 0.9402,
+      "step": 461
+    },
+    {
+      "epoch": 0.6698079014135556,
+      "grad_norm": 1.379297214017358,
+      "learning_rate": 5.177518519919924e-06,
+      "loss": 1.0571,
+      "step": 462
+    },
+    {
+      "epoch": 0.6712577020659659,
+      "grad_norm": 1.2951802434264668,
+      "learning_rate": 5.136372242427316e-06,
+      "loss": 1.0454,
+      "step": 463
+    },
+    {
+      "epoch": 0.6727075027183762,
+      "grad_norm": 1.390214458419132,
+      "learning_rate": 5.095333538694624e-06,
+      "loss": 0.9877,
+      "step": 464
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 1.2799241537202823,
+      "learning_rate": 5.054403316416247e-06,
+      "loss": 1.0644,
+      "step": 465
+    },
+    {
+      "epoch": 0.6756071040231968,
+      "grad_norm": 1.375985405640331,
+      "learning_rate": 5.0135824808871905e-06,
+      "loss": 1.1034,
+      "step": 466
+    },
+    {
+      "epoch": 0.6770569046756071,
+      "grad_norm": 1.3332446902296693,
+      "learning_rate": 4.972871934983032e-06,
+      "loss": 0.9948,
+      "step": 467
+    },
+    {
+      "epoch": 0.6785067053280174,
+      "grad_norm": 1.2373968612880124,
+      "learning_rate": 4.932272579139981e-06,
+      "loss": 0.9613,
+      "step": 468
+    },
+    {
+      "epoch": 0.6799565059804277,
+      "grad_norm": 1.3897775317127234,
+      "learning_rate": 4.891785311334924e-06,
+      "loss": 1.0003,
+      "step": 469
+    },
+    {
+      "epoch": 0.681406306632838,
+      "grad_norm": 1.266653583429857,
+      "learning_rate": 4.851411027065596e-06,
+      "loss": 1.0514,
+      "step": 470
+    },
+    {
+      "epoch": 0.6828561072852483,
+      "grad_norm": 1.4055987085065904,
+      "learning_rate": 4.811150619330759e-06,
+      "loss": 1.0507,
+      "step": 471
+    },
+    {
+      "epoch": 0.6843059079376586,
+      "grad_norm": 1.330017466327231,
+      "learning_rate": 4.771004978610457e-06,
+      "loss": 0.9867,
+      "step": 472
+    },
+    {
+      "epoch": 0.6857557085900688,
+      "grad_norm": 1.3223920410879977,
+      "learning_rate": 4.730974992846304e-06,
+      "loss": 1.0002,
+      "step": 473
+    },
+    {
+      "epoch": 0.6872055092424791,
+      "grad_norm": 1.2799779984507171,
+      "learning_rate": 4.6910615474218675e-06,
+      "loss": 0.8818,
+      "step": 474
+    },
+    {
+      "epoch": 0.6886553098948894,
+      "grad_norm": 1.257688034516758,
+      "learning_rate": 4.651265525143069e-06,
+      "loss": 1.0294,
+      "step": 475
+    },
+    {
+      "epoch": 0.6901051105472997,
+      "grad_norm": 1.4355279487044768,
+      "learning_rate": 4.611587806218669e-06,
+      "loss": 0.9195,
+      "step": 476
+    },
+    {
+      "epoch": 0.69155491119971,
+      "grad_norm": 1.4245434972375823,
+      "learning_rate": 4.572029268240787e-06,
+      "loss": 1.0519,
+      "step": 477
+    },
+    {
+      "epoch": 0.6930047118521203,
+      "grad_norm": 1.3657576039812214,
+      "learning_rate": 4.532590786165509e-06,
+      "loss": 1.0458,
+      "step": 478
+    },
+    {
+      "epoch": 0.6944545125045306,
+      "grad_norm": 1.3246348434145814,
+      "learning_rate": 4.4932732322935055e-06,
+      "loss": 1.0111,
+      "step": 479
+    },
+    {
+      "epoch": 0.6959043131569409,
+      "grad_norm": 1.252663742103646,
+      "learning_rate": 4.45407747625077e-06,
+      "loss": 1.09,
+      "step": 480
+    },
+    {
+      "epoch": 0.6973541138093512,
+      "grad_norm": 1.2514480608107745,
+      "learning_rate": 4.41500438496937e-06,
+      "loss": 0.9988,
+      "step": 481
+    },
+    {
+      "epoch": 0.6988039144617615,
+      "grad_norm": 1.2601704153444035,
+      "learning_rate": 4.376054822668262e-06,
+      "loss": 0.9624,
+      "step": 482
+    },
+    {
+      "epoch": 0.7002537151141718,
+      "grad_norm": 1.3667040506827728,
+      "learning_rate": 4.3372296508342025e-06,
+      "loss": 1.0571,
+      "step": 483
+    },
+    {
+      "epoch": 0.7017035157665821,
+      "grad_norm": 1.313374023872567,
+      "learning_rate": 4.298529728202661e-06,
+      "loss": 1.0506,
+      "step": 484
+    },
+    {
+      "epoch": 0.7031533164189924,
+      "grad_norm": 1.223638770151155,
+      "learning_rate": 4.2599559107388645e-06,
+      "loss": 0.9996,
+      "step": 485
+    },
+    {
+      "epoch": 0.7046031170714027,
+      "grad_norm": 1.4013529447707769,
+      "learning_rate": 4.221509051618827e-06,
+      "loss": 1.0672,
+      "step": 486
+    },
+    {
+      "epoch": 0.706052917723813,
+      "grad_norm": 1.247779741094988,
+      "learning_rate": 4.183190001210511e-06,
+      "loss": 0.9134,
+      "step": 487
+    },
+    {
+      "epoch": 0.7075027183762232,
+      "grad_norm": 1.5522286912524503,
+      "learning_rate": 4.1449996070549895e-06,
+      "loss": 1.0435,
+      "step": 488
+    },
+    {
+      "epoch": 0.7089525190286335,
+      "grad_norm": 1.3425386992973771,
+      "learning_rate": 4.10693871384773e-06,
+      "loss": 0.9762,
+      "step": 489
+    },
+    {
+      "epoch": 0.7104023196810438,
+      "grad_norm": 1.3830220232567356,
+      "learning_rate": 4.0690081634198865e-06,
+      "loss": 0.9335,
+      "step": 490
+    },
+    {
+      "epoch": 0.7118521203334541,
+      "grad_norm": 1.374824944218665,
+      "learning_rate": 4.031208794719695e-06,
+      "loss": 1.0042,
+      "step": 491
+    },
+    {
+      "epoch": 0.7133019209858644,
+      "grad_norm": 1.1624671718010195,
+      "learning_rate": 3.993541443793914e-06,
+      "loss": 0.9736,
+      "step": 492
+    },
+    {
+      "epoch": 0.7147517216382747,
+      "grad_norm": 1.3430877502727783,
+      "learning_rate": 3.956006943769331e-06,
+      "loss": 0.983,
+      "step": 493
+    },
+    {
+      "epoch": 0.716201522290685,
+      "grad_norm": 1.2596547159354614,
+      "learning_rate": 3.918606124834337e-06,
+      "loss": 0.9817,
+      "step": 494
+    },
+    {
+      "epoch": 0.7176513229430953,
+      "grad_norm": 1.3308829575744618,
+      "learning_rate": 3.881339814220555e-06,
+      "loss": 1.0201,
+      "step": 495
+    },
+    {
+      "epoch": 0.7191011235955056,
+      "grad_norm": 1.3121710014551011,
+      "learning_rate": 3.844208836184567e-06,
+      "loss": 0.9669,
+      "step": 496
+    },
+    {
+      "epoch": 0.7205509242479159,
+      "grad_norm": 1.4659117928686862,
+      "learning_rate": 3.8072140119896507e-06,
+      "loss": 1.0443,
+      "step": 497
+    },
+    {
+      "epoch": 0.7220007249003262,
+      "grad_norm": 1.5373082462147176,
+      "learning_rate": 3.770356159887646e-06,
+      "loss": 1.0154,
+      "step": 498
+    },
+    {
+      "epoch": 0.7234505255527365,
+      "grad_norm": 1.3970901862628455,
+      "learning_rate": 3.733636095100838e-06,
+      "loss": 1.0873,
+      "step": 499
+    },
+    {
+      "epoch": 0.7249003262051468,
+      "grad_norm": 1.3696002886768786,
+      "learning_rate": 3.697054629803931e-06,
+      "loss": 1.0451,
+      "step": 500
+    },
+    {
+      "epoch": 0.7263501268575571,
+      "grad_norm": 1.3056126005463657,
+      "learning_rate": 3.6606125731060815e-06,
+      "loss": 1.0499,
+      "step": 501
+    },
+    {
+      "epoch": 0.7277999275099674,
+      "grad_norm": 1.2293357342690303,
+      "learning_rate": 3.6243107310330117e-06,
+      "loss": 0.9572,
+      "step": 502
+    },
+    {
+      "epoch": 0.7292497281623777,
+      "grad_norm": 1.441206235584573,
+      "learning_rate": 3.5881499065091663e-06,
+      "loss": 1.0562,
+      "step": 503
+    },
+    {
+      "epoch": 0.730699528814788,
+      "grad_norm": 1.2833018346954448,
+      "learning_rate": 3.552130899339975e-06,
+      "loss": 0.9856,
+      "step": 504
+    },
+    {
+      "epoch": 0.7321493294671982,
+      "grad_norm": 1.2739700454201517,
+      "learning_rate": 3.5162545061941332e-06,
+      "loss": 1.0224,
+      "step": 505
+    },
+    {
+      "epoch": 0.7335991301196085,
+      "grad_norm": 1.3699299458257053,
+      "learning_rate": 3.480521520586021e-06,
+      "loss": 1.0798,
+      "step": 506
+    },
+    {
+      "epoch": 0.7350489307720188,
+      "grad_norm": 1.2823717950220108,
+      "learning_rate": 3.4449327328581075e-06,
+      "loss": 1.0366,
+      "step": 507
+    },
+    {
+      "epoch": 0.7364987314244291,
+      "grad_norm": 1.2510889045326017,
+      "learning_rate": 3.4094889301635036e-06,
+      "loss": 0.9002,
+      "step": 508
+    },
+    {
+      "epoch": 0.7379485320768394,
+      "grad_norm": 1.4586520712838673,
+      "learning_rate": 3.3741908964485415e-06,
+      "loss": 0.9413,
+      "step": 509
+    },
+    {
+      "epoch": 0.7393983327292497,
+      "grad_norm": 1.3124464942492935,
+      "learning_rate": 3.339039412435422e-06,
+      "loss": 1.1183,
+      "step": 510
+    },
+    {
+      "epoch": 0.74084813338166,
+      "grad_norm": 1.2719144081764076,
+      "learning_rate": 3.304035255604974e-06,
+      "loss": 1.0401,
+      "step": 511
+    },
+    {
+      "epoch": 0.7422979340340703,
+      "grad_norm": 1.3134312904240224,
+      "learning_rate": 3.2691792001794244e-06,
+      "loss": 1.0443,
+      "step": 512
+    },
+    {
+      "epoch": 0.7437477346864806,
+      "grad_norm": 1.3604087276640688,
+      "learning_rate": 3.2344720171053134e-06,
+      "loss": 1.0927,
+      "step": 513
+    },
+    {
+      "epoch": 0.7451975353388909,
+      "grad_norm": 1.6030644255662119,
+      "learning_rate": 3.199914474036403e-06,
+      "loss": 0.9913,
+      "step": 514
+    },
+    {
+      "epoch": 0.7466473359913012,
+      "grad_norm": 1.4580431168352719,
+      "learning_rate": 3.1655073353167285e-06,
+      "loss": 1.0665,
+      "step": 515
+    },
+    {
+      "epoch": 0.7480971366437115,
+      "grad_norm": 1.385036578088552,
+      "learning_rate": 3.1312513619636686e-06,
+      "loss": 1.108,
+      "step": 516
+    },
+    {
+      "epoch": 0.7495469372961218,
+      "grad_norm": 1.549024762705719,
+      "learning_rate": 3.09714731165114e-06,
+      "loss": 1.0323,
+      "step": 517
+    },
+    {
+      "epoch": 0.7509967379485321,
+      "grad_norm": 1.2869471846142893,
+      "learning_rate": 3.0631959386928078e-06,
+      "loss": 1.0396,
+      "step": 518
+    },
+    {
+      "epoch": 0.7524465386009423,
+      "grad_norm": 1.3471019495239385,
+      "learning_rate": 3.0293979940254316e-06,
+      "loss": 1.0188,
+      "step": 519
+    },
+    {
+      "epoch": 0.7524465386009423,
+      "eval_loss": 1.756138563156128,
+      "eval_runtime": 337.0558,
+      "eval_samples_per_second": 52.377,
+      "eval_steps_per_second": 0.819,
+      "step": 519
+    },
+    {
+      "epoch": 0.7538963392533526,
+      "grad_norm": 1.2434901218449843,
+      "learning_rate": 2.9957542251922354e-06,
+      "loss": 0.9458,
+      "step": 520
+    },
+    {
+      "epoch": 0.7553461399057629,
+      "grad_norm": 1.302405319305409,
+      "learning_rate": 2.9622653763263874e-06,
+      "loss": 0.9885,
+      "step": 521
+    },
+    {
+      "epoch": 0.7567959405581732,
+      "grad_norm": 1.3745892644317177,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 1.065,
+      "step": 522
+    },
+    {
+      "epoch": 0.7582457412105835,
+      "grad_norm": 1.4247157677022666,
+      "learning_rate": 2.8957553978803897e-06,
+      "loss": 1.0509,
+      "step": 523
+    },
+    {
+      "epoch": 0.7596955418629938,
+      "grad_norm": 1.321095624689294,
+      "learning_rate": 2.8627357393685117e-06,
+      "loss": 0.9613,
+      "step": 524
+    },
+    {
+      "epoch": 0.7611453425154041,
+      "grad_norm": 1.4303547357465254,
+      "learning_rate": 2.829873942927971e-06,
+      "loss": 1.0577,
+      "step": 525
+    },
+    {
+      "epoch": 0.7625951431678144,
+      "grad_norm": 1.2759368225063656,
+      "learning_rate": 2.797170735396263e-06,
+      "loss": 1.0034,
+      "step": 526
+    },
+    {
+      "epoch": 0.7640449438202247,
+      "grad_norm": 1.4634167438447727,
+      "learning_rate": 2.7646268401032086e-06,
+      "loss": 1.1156,
+      "step": 527
+    },
+    {
+      "epoch": 0.765494744472635,
+      "grad_norm": 1.2277914303422794,
+      "learning_rate": 2.732242976854962e-06,
+      "loss": 0.9346,
+      "step": 528
+    },
+    {
+      "epoch": 0.7669445451250453,
+      "grad_norm": 1.3015618940232396,
+      "learning_rate": 2.7000198619180795e-06,
+      "loss": 1.0275,
+      "step": 529
+    },
+    {
+      "epoch": 0.7683943457774556,
+      "grad_norm": 1.3101369055090744,
+      "learning_rate": 2.6679582080036957e-06,
+      "loss": 1.023,
+      "step": 530
+    },
+    {
+      "epoch": 0.7698441464298659,
+      "grad_norm": 1.3723672953021895,
+      "learning_rate": 2.636058724251739e-06,
+      "loss": 1.0436,
+      "step": 531
+    },
+    {
+      "epoch": 0.7712939470822762,
+      "grad_norm": 1.2777023951329174,
+      "learning_rate": 2.6043221162152632e-06,
+      "loss": 0.9828,
+      "step": 532
+    },
+    {
+      "epoch": 0.7727437477346865,
+      "grad_norm": 1.30678765033226,
+      "learning_rate": 2.572749085844829e-06,
+      "loss": 1.0482,
+      "step": 533
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 1.1933072812672492,
+      "learning_rate": 2.541340331472998e-06,
+      "loss": 1.0748,
+      "step": 534
+    },
+    {
+      "epoch": 0.775643349039507,
+      "grad_norm": 1.3056150546952323,
+      "learning_rate": 2.510096547798857e-06,
+      "loss": 0.9439,
+      "step": 535
+    },
+    {
+      "epoch": 0.7770931496919173,
+      "grad_norm": 1.2965455242827506,
+      "learning_rate": 2.479018425872686e-06,
+      "loss": 0.9569,
+      "step": 536
+    },
+    {
+      "epoch": 0.7785429503443276,
+      "grad_norm": 1.3779199436730478,
+      "learning_rate": 2.44810665308064e-06,
+      "loss": 1.0236,
+      "step": 537
+    },
+    {
+      "epoch": 0.7799927509967379,
+      "grad_norm": 1.2589025278449242,
+      "learning_rate": 2.4173619131295754e-06,
+      "loss": 1.1007,
+      "step": 538
+    },
+    {
+      "epoch": 0.7814425516491482,
+      "grad_norm": 1.2988851125148146,
+      "learning_rate": 2.3867848860319143e-06,
+      "loss": 1.0602,
+      "step": 539
+    },
+    {
+      "epoch": 0.7828923523015585,
+      "grad_norm": 1.3858874427345413,
+      "learning_rate": 2.3563762480905917e-06,
+      "loss": 1.0725,
+      "step": 540
+    },
+    {
+      "epoch": 0.7843421529539688,
+      "grad_norm": 1.2427699307364515,
+      "learning_rate": 2.3261366718841307e-06,
+      "loss": 0.9729,
+      "step": 541
+    },
+    {
+      "epoch": 0.7857919536063791,
+      "grad_norm": 1.2832058288639472,
+      "learning_rate": 2.296066826251725e-06,
+      "loss": 0.973,
+      "step": 542
+    },
+    {
+      "epoch": 0.7872417542587894,
+      "grad_norm": 1.301632555150497,
+      "learning_rate": 2.266167376278481e-06,
+      "loss": 1.0431,
+      "step": 543
+    },
+    {
+      "epoch": 0.7886915549111997,
+      "grad_norm": 1.3702492722058395,
+      "learning_rate": 2.23643898328068e-06,
+      "loss": 1.0243,
+      "step": 544
+    },
+    {
+      "epoch": 0.79014135556361,
+      "grad_norm": 1.432548865693198,
+      "learning_rate": 2.206882304791176e-06,
+      "loss": 1.0265,
+      "step": 545
+    },
+    {
+      "epoch": 0.7915911562160203,
+      "grad_norm": 1.2167034618957844,
+      "learning_rate": 2.177497994544828e-06,
+      "loss": 1.0207,
+      "step": 546
+    },
+    {
+      "epoch": 0.7930409568684306,
+      "grad_norm": 1.3693734676283214,
+      "learning_rate": 2.148286702464063e-06,
+      "loss": 1.002,
+      "step": 547
+    },
+    {
+      "epoch": 0.7944907575208409,
+      "grad_norm": 1.277307549554258,
+      "learning_rate": 2.119249074644485e-06,
+      "loss": 0.9765,
+      "step": 548
+    },
+    {
+      "epoch": 0.7959405581732512,
+      "grad_norm": 1.457380572642945,
+      "learning_rate": 2.090385753340596e-06,
+      "loss": 0.9847,
+      "step": 549
+    },
+    {
+      "epoch": 0.7973903588256614,
+      "grad_norm": 1.2795058790071878,
+      "learning_rate": 2.0616973769515737e-06,
+      "loss": 1.0711,
+      "step": 550
+    },
+    {
+      "epoch": 0.7988401594780717,
+      "grad_norm": 1.2914650960408245,
+      "learning_rate": 2.033184580007177e-06,
+      "loss": 0.973,
+      "step": 551
+    },
+    {
+      "epoch": 0.800289960130482,
+      "grad_norm": 1.2849424512148861,
+      "learning_rate": 2.0048479931536833e-06,
+      "loss": 1.0246,
+      "step": 552
+    },
+    {
+      "epoch": 0.8017397607828923,
+      "grad_norm": 1.3196699477478921,
+      "learning_rate": 1.976688243139965e-06,
+      "loss": 0.9549,
+      "step": 553
+    },
+    {
+      "epoch": 0.8031895614353026,
+      "grad_norm": 1.27870727376407,
+      "learning_rate": 1.94870595280361e-06,
+      "loss": 0.9917,
+      "step": 554
+    },
+    {
+      "epoch": 0.8046393620877129,
+      "grad_norm": 1.4102598643358017,
+      "learning_rate": 1.920901741057153e-06,
+      "loss": 1.0362,
+      "step": 555
+    },
+    {
+      "epoch": 0.8060891627401232,
+      "grad_norm": 1.3428369390389148,
+      "learning_rate": 1.8932762228743884e-06,
+      "loss": 0.9882,
+      "step": 556
+    },
+    {
+      "epoch": 0.8075389633925335,
+      "grad_norm": 1.418893962357456,
+      "learning_rate": 1.8658300092767546e-06,
+      "loss": 1.0966,
+      "step": 557
+    },
+    {
+      "epoch": 0.8089887640449438,
+      "grad_norm": 1.2030360788347458,
+      "learning_rate": 1.8385637073198426e-06,
+      "loss": 1.0994,
+      "step": 558
+    },
+    {
+      "epoch": 0.8104385646973541,
+      "grad_norm": 1.4449586023944951,
+      "learning_rate": 1.8114779200799437e-06,
+      "loss": 1.0885,
+      "step": 559
+    },
+    {
+      "epoch": 0.8118883653497644,
+      "grad_norm": 1.2975883975195892,
+      "learning_rate": 1.7845732466407361e-06,
+      "loss": 1.0139,
+      "step": 560
+    },
+    {
+      "epoch": 0.8133381660021747,
+      "grad_norm": 1.2428100792582848,
+      "learning_rate": 1.7578502820800048e-06,
+      "loss": 0.9813,
+      "step": 561
+    },
+    {
+      "epoch": 0.814787966654585,
+      "grad_norm": 1.267996208189487,
+      "learning_rate": 1.731309617456518e-06,
+      "loss": 0.92,
+      "step": 562
+    },
+    {
+      "epoch": 0.8162377673069953,
+      "grad_norm": 1.2884634615906705,
+      "learning_rate": 1.7049518397969135e-06,
+      "loss": 1.0553,
+      "step": 563
+    },
+    {
+      "epoch": 0.8176875679594056,
+      "grad_norm": 1.236390738153001,
+      "learning_rate": 1.6787775320827438e-06,
+      "loss": 1.0866,
+      "step": 564
+    },
+    {
+      "epoch": 0.8191373686118159,
+      "grad_norm": 1.3359066714359038,
+      "learning_rate": 1.652787273237565e-06,
+      "loss": 0.971,
+      "step": 565
+    },
+    {
+      "epoch": 0.8205871692642261,
+      "grad_norm": 1.4555166307589265,
+      "learning_rate": 1.6269816381141446e-06,
+      "loss": 0.9125,
+      "step": 566
+    },
+    {
+      "epoch": 0.8220369699166364,
+      "grad_norm": 1.4321887021986563,
+      "learning_rate": 1.601361197481741e-06,
+      "loss": 1.0578,
+      "step": 567
+    },
+    {
+      "epoch": 0.8234867705690467,
+      "grad_norm": 1.4621089330523545,
+      "learning_rate": 1.5759265180134709e-06,
+      "loss": 1.0591,
+      "step": 568
+    },
+    {
+      "epoch": 0.824936571221457,
+      "grad_norm": 1.254702830875676,
+      "learning_rate": 1.5506781622737943e-06,
+      "loss": 0.9957,
+      "step": 569
+    },
+    {
+      "epoch": 0.8263863718738673,
+      "grad_norm": 1.318143563983686,
+      "learning_rate": 1.5256166887060575e-06,
+      "loss": 1.075,
+      "step": 570
+    },
+    {
+      "epoch": 0.8278361725262776,
+      "grad_norm": 1.299914528498022,
+      "learning_rate": 1.5007426516201451e-06,
+      "loss": 0.9639,
+      "step": 571
+    },
+    {
+      "epoch": 0.8292859731786879,
+      "grad_norm": 1.2655311960828108,
+      "learning_rate": 1.4760566011802168e-06,
+      "loss": 1.0704,
+      "step": 572
+    },
+    {
+      "epoch": 0.8307357738310982,
+      "grad_norm": 1.433242708326693,
+      "learning_rate": 1.4515590833925508e-06,
+      "loss": 1.0084,
+      "step": 573
+    },
+    {
+      "epoch": 0.8321855744835085,
+      "grad_norm": 1.3185324301785566,
+      "learning_rate": 1.4272506400934461e-06,
+      "loss": 0.9849,
+      "step": 574
+    },
+    {
+      "epoch": 0.8336353751359188,
+      "grad_norm": 1.2521237143004595,
+      "learning_rate": 1.4031318089372614e-06,
+      "loss": 0.9133,
+      "step": 575
+    },
+    {
+      "epoch": 0.8350851757883291,
+      "grad_norm": 1.3045917130841314,
+      "learning_rate": 1.379203123384507e-06,
+      "loss": 1.004,
+      "step": 576
+    },
+    {
+      "epoch": 0.8365349764407394,
+      "grad_norm": 1.263317030979965,
+      "learning_rate": 1.3554651126900564e-06,
+      "loss": 1.0522,
+      "step": 577
+    },
+    {
+      "epoch": 0.8379847770931497,
+      "grad_norm": 1.353258062692782,
+      "learning_rate": 1.331918301891425e-06,
+      "loss": 1.0913,
+      "step": 578
+    },
+    {
+      "epoch": 0.83943457774556,
+      "grad_norm": 1.2945115509902212,
+      "learning_rate": 1.3085632117971792e-06,
+      "loss": 0.9362,
+      "step": 579
+    },
+    {
+      "epoch": 0.8408843783979703,
+      "grad_norm": 1.3926445361876894,
+      "learning_rate": 1.285400358975395e-06,
+      "loss": 0.9946,
+      "step": 580
+    },
+    {
+      "epoch": 0.8423341790503805,
+      "grad_norm": 1.3484141207456959,
+      "learning_rate": 1.2624302557422475e-06,
+      "loss": 0.9937,
+      "step": 581
+    },
+    {
+      "epoch": 0.8437839797027908,
+      "grad_norm": 1.3405127727326804,
+      "learning_rate": 1.2396534101506774e-06,
+      "loss": 1.0272,
+      "step": 582
+    },
+    {
+      "epoch": 0.8452337803552011,
+      "grad_norm": 1.3121025637754966,
+      "learning_rate": 1.2170703259791472e-06,
+      "loss": 1.0032,
+      "step": 583
+    },
+    {
+      "epoch": 0.8466835810076114,
+      "grad_norm": 1.2543463725434825,
+      "learning_rate": 1.1946815027205027e-06,
+      "loss": 1.039,
+      "step": 584
+    },
+    {
+      "epoch": 0.8481333816600217,
+      "grad_norm": 1.2595696335622666,
+      "learning_rate": 1.1724874355709259e-06,
+      "loss": 0.9782,
+      "step": 585
+    },
+    {
+      "epoch": 0.849583182312432,
+      "grad_norm": 1.275912697740584,
+      "learning_rate": 1.1504886154189854e-06,
+      "loss": 0.9855,
+      "step": 586
+    },
+    {
+      "epoch": 0.8510329829648423,
+      "grad_norm": 1.1766968575089594,
+      "learning_rate": 1.1286855288347676e-06,
+      "loss": 1.0282,
+      "step": 587
+    },
+    {
+      "epoch": 0.8524827836172526,
+      "grad_norm": 1.1988171850592366,
+      "learning_rate": 1.1070786580591287e-06,
+      "loss": 1.0749,
+      "step": 588
+    },
+    {
+      "epoch": 0.8539325842696629,
+      "grad_norm": 1.2408454880671107,
+      "learning_rate": 1.085668480993015e-06,
+      "loss": 0.9625,
+      "step": 589
+    },
+    {
+      "epoch": 0.8553823849220732,
+      "grad_norm": 1.332405261038831,
+      "learning_rate": 1.0644554711869115e-06,
+      "loss": 1.0154,
+      "step": 590
+    },
+    {
+      "epoch": 0.8568321855744835,
+      "grad_norm": 1.3434695511965489,
+      "learning_rate": 1.0434400978303427e-06,
+      "loss": 0.9528,
+      "step": 591
+    },
+    {
+      "epoch": 0.8582819862268938,
+      "grad_norm": 1.402048435976599,
+      "learning_rate": 1.0226228257415183e-06,
+      "loss": 0.9226,
+      "step": 592
+    },
+    {
+      "epoch": 0.8597317868793041,
+      "grad_norm": 1.4491424704388471,
+      "learning_rate": 1.0020041153570349e-06,
+      "loss": 1.053,
+      "step": 593
+    },
+    {
+      "epoch": 0.8611815875317144,
+      "grad_norm": 1.2116795182521507,
+      "learning_rate": 9.815844227217065e-07,
+      "loss": 0.9152,
+      "step": 594
+    },
+    {
+      "epoch": 0.8626313881841247,
+      "grad_norm": 1.2578018154901727,
+      "learning_rate": 9.61364199478465e-07,
+      "loss": 0.9729,
+      "step": 595
+    },
+    {
+      "epoch": 0.864081188836535,
+      "grad_norm": 1.3338844051158256,
+      "learning_rate": 9.413438928583807e-07,
+      "loss": 1.0646,
+      "step": 596
+    },
+    {
+      "epoch": 0.8655309894889452,
+      "grad_norm": 1.3953544646730935,
+      "learning_rate": 9.215239456707636e-07,
+      "loss": 0.9662,
+      "step": 597
+    },
+    {
+      "epoch": 0.8669807901413555,
+      "grad_norm": 1.305945665578386,
+      "learning_rate": 9.01904796293378e-07,
+      "loss": 1.0057,
+      "step": 598
+    },
+    {
+      "epoch": 0.8684305907937658,
+      "grad_norm": 1.364640661531156,
+      "learning_rate": 8.824868786627305e-07,
+      "loss": 1.0059,
+      "step": 599
+    },
+    {
+      "epoch": 0.8698803914461761,
+      "grad_norm": 1.296996440744034,
+      "learning_rate": 8.632706222644893e-07,
+      "loss": 0.9793,
+      "step": 600
+    },
+    {
+      "epoch": 0.8713301920985864,
+      "grad_norm": 1.4527533040437381,
+      "learning_rate": 8.442564521239782e-07,
+      "loss": 0.9825,
+      "step": 601
+    },
+    {
+      "epoch": 0.8727799927509967,
+      "grad_norm": 1.2826359516418513,
+      "learning_rate": 8.254447887967698e-07,
+      "loss": 0.9364,
+      "step": 602
+    },
+    {
+      "epoch": 0.874229793403407,
+      "grad_norm": 1.4678346752178948,
+      "learning_rate": 8.06836048359394e-07,
+      "loss": 1.0496,
+      "step": 603
+    },
+    {
+      "epoch": 0.8756795940558173,
+      "grad_norm": 1.3575023245122086,
+      "learning_rate": 7.884306424001309e-07,
+      "loss": 0.9825,
+      "step": 604
+    },
+    {
+      "epoch": 0.8771293947082276,
+      "grad_norm": 1.3239755203112757,
+      "learning_rate": 7.70228978009907e-07,
+      "loss": 0.9783,
+      "step": 605
+    },
+    {
+      "epoch": 0.8785791953606379,
+      "grad_norm": 1.3715714186260306,
+      "learning_rate": 7.522314577732836e-07,
+      "loss": 0.9595,
+      "step": 606
+    },
+    {
+      "epoch": 0.8800289960130482,
+      "grad_norm": 1.3115042473016585,
+      "learning_rate": 7.344384797595717e-07,
+      "loss": 1.0478,
+      "step": 607
+    },
+    {
+      "epoch": 0.8814787966654585,
+      "grad_norm": 1.3289394919102624,
+      "learning_rate": 7.16850437514004e-07,
+      "loss": 0.9377,
+      "step": 608
+    },
+    {
+      "epoch": 0.8829285973178688,
+      "grad_norm": 1.2997392587334409,
+      "learning_rate": 6.994677200490508e-07,
+      "loss": 0.968,
+      "step": 609
+    },
+    {
+      "epoch": 0.8843783979702791,
+      "grad_norm": 1.3480679055642706,
+      "learning_rate": 6.822907118358047e-07,
+      "loss": 0.9519,
+      "step": 610
+    },
+    {
+      "epoch": 0.8858281986226894,
+      "grad_norm": 1.296876980493611,
+      "learning_rate": 6.65319792795479e-07,
+      "loss": 0.9967,
+      "step": 611
+    },
+    {
+      "epoch": 0.8872779992750996,
+      "grad_norm": 1.3735950155419208,
+      "learning_rate": 6.485553382910026e-07,
+      "loss": 1.0482,
+      "step": 612
+    },
+    {
+      "epoch": 0.8887277999275099,
+      "grad_norm": 1.4139081772434936,
+      "learning_rate": 6.319977191187232e-07,
+      "loss": 0.9599,
+      "step": 613
+    },
+    {
+      "epoch": 0.8901776005799202,
+      "grad_norm": 1.2881706709586878,
+      "learning_rate": 6.15647301500203e-07,
+      "loss": 0.9434,
+      "step": 614
+    },
+    {
+      "epoch": 0.8916274012323305,
+      "grad_norm": 1.3054337924445512,
+      "learning_rate": 5.995044470741151e-07,
+      "loss": 0.9279,
+      "step": 615
+    },
+    {
+      "epoch": 0.8930772018847408,
+      "grad_norm": 1.4436158237108208,
+      "learning_rate": 5.835695128882512e-07,
+      "loss": 1.061,
+      "step": 616
+    },
+    {
+      "epoch": 0.8945270025371511,
+      "grad_norm": 1.1582915726834864,
+      "learning_rate": 5.678428513916212e-07,
+      "loss": 0.9624,
+      "step": 617
+    },
+    {
+      "epoch": 0.8959768031895614,
+      "grad_norm": 1.4909187088715723,
+      "learning_rate": 5.523248104266577e-07,
+      "loss": 0.9922,
+      "step": 618
+    },
+    {
+      "epoch": 0.8974266038419717,
+      "grad_norm": 1.3832576562564656,
+      "learning_rate": 5.370157332215209e-07,
+      "loss": 1.0116,
+      "step": 619
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 1.399464193132112,
+      "learning_rate": 5.219159583825107e-07,
+      "loss": 0.9662,
+      "step": 620
+    },
+    {
+      "epoch": 0.9003262051467923,
+      "grad_norm": 1.2823463828780335,
+      "learning_rate": 5.070258198865741e-07,
+      "loss": 0.9686,
+      "step": 621
+    },
+    {
+      "epoch": 0.9017760057992026,
+      "grad_norm": 1.280299155323607,
+      "learning_rate": 4.923456470739219e-07,
+      "loss": 0.9688,
+      "step": 622
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 1.3563344480243746,
+      "learning_rate": 4.778757646407361e-07,
+      "loss": 0.994,
+      "step": 623
+    },
+    {
+      "epoch": 0.9046756071040232,
+      "grad_norm": 1.2866329100675669,
+      "learning_rate": 4.636164926320064e-07,
+      "loss": 1.0026,
+      "step": 624
+    },
+    {
+      "epoch": 0.9061254077564335,
+      "grad_norm": 1.2206905860446844,
+      "learning_rate": 4.495681464344259e-07,
+      "loss": 0.9708,
+      "step": 625
+    },
+    {
+      "epoch": 0.9075752084088438,
+      "grad_norm": 1.4398096118244226,
+      "learning_rate": 4.3573103676943784e-07,
+      "loss": 1.0514,
+      "step": 626
+    },
+    {
+      "epoch": 0.909025009061254,
+      "grad_norm": 1.2398987655164297,
+      "learning_rate": 4.2210546968634803e-07,
+      "loss": 1.0068,
+      "step": 627
+    },
+    {
+      "epoch": 0.9104748097136643,
+      "grad_norm": 1.405166483432629,
+      "learning_rate": 4.086917465555662e-07,
+      "loss": 1.0448,
+      "step": 628
+    },
+    {
+      "epoch": 0.9119246103660746,
+      "grad_norm": 1.3464885047336612,
+      "learning_rate": 3.9549016406193686e-07,
+      "loss": 1.0312,
+      "step": 629
+    },
+    {
+      "epoch": 0.9133744110184849,
+      "grad_norm": 1.351547595002901,
+      "learning_rate": 3.8250101419816777e-07,
+      "loss": 1.0232,
+      "step": 630
+    },
+    {
+      "epoch": 0.9148242116708952,
+      "grad_norm": 1.3935973264965327,
+      "learning_rate": 3.6972458425838963e-07,
+      "loss": 0.9885,
+      "step": 631
+    },
+    {
+      "epoch": 0.9162740123233055,
+      "grad_norm": 1.3769810798375854,
+      "learning_rate": 3.5716115683178563e-07,
+      "loss": 1.0722,
+      "step": 632
+    },
+    {
+      "epoch": 0.9177238129757158,
+      "grad_norm": 1.3568354564975023,
+      "learning_rate": 3.4481100979635304e-07,
+      "loss": 1.0439,
+      "step": 633
+    },
+    {
+      "epoch": 0.9191736136281261,
+      "grad_norm": 1.2323702124303846,
+      "learning_rate": 3.32674416312746e-07,
+      "loss": 1.0081,
+      "step": 634
+    },
+    {
+      "epoch": 0.9206234142805364,
+      "grad_norm": 1.2616025417378836,
+      "learning_rate": 3.207516448182435e-07,
+      "loss": 1.0347,
+      "step": 635
+    },
+    {
+      "epoch": 0.9220732149329467,
+      "grad_norm": 1.2966968431867678,
+      "learning_rate": 3.090429590208033e-07,
+      "loss": 0.9814,
+      "step": 636
+    },
+    {
+      "epoch": 0.923523015585357,
+      "grad_norm": 1.3229393017389983,
+      "learning_rate": 2.9754861789324074e-07,
+      "loss": 1.0697,
+      "step": 637
+    },
+    {
+      "epoch": 0.9249728162377673,
+      "grad_norm": 1.3344008771210325,
+      "learning_rate": 2.862688756674881e-07,
+      "loss": 0.9895,
+      "step": 638
+    },
+    {
+      "epoch": 0.9264226168901776,
+      "grad_norm": 1.2624318586528753,
+      "learning_rate": 2.752039818289809e-07,
+      "loss": 1.0022,
+      "step": 639
+    },
+    {
+      "epoch": 0.9278724175425879,
+      "grad_norm": 1.2871854632831545,
+      "learning_rate": 2.6435418111113277e-07,
+      "loss": 1.0663,
+      "step": 640
+    },
+    {
+      "epoch": 0.9293222181949982,
+      "grad_norm": 1.2619269233740271,
+      "learning_rate": 2.537197134899294e-07,
+      "loss": 0.9675,
+      "step": 641
+    },
+    {
+      "epoch": 0.9307720188474085,
+      "grad_norm": 1.2705652572486517,
+      "learning_rate": 2.4330081417861527e-07,
+      "loss": 1.0097,
+      "step": 642
+    },
+    {
+      "epoch": 0.9322218194998187,
+      "grad_norm": 1.3229337637036596,
+      "learning_rate": 2.3309771362249323e-07,
+      "loss": 1.0379,
+      "step": 643
+    },
+    {
+      "epoch": 0.933671620152229,
+      "grad_norm": 1.3291126382857967,
+      "learning_rate": 2.2311063749382743e-07,
+      "loss": 0.9518,
+      "step": 644
+    },
+    {
+      "epoch": 0.9351214208046393,
+      "grad_norm": 1.3652945261942868,
+      "learning_rate": 2.1333980668685416e-07,
+      "loss": 1.0468,
+      "step": 645
+    },
+    {
+      "epoch": 0.9365712214570496,
+      "grad_norm": 1.2592609968560355,
+      "learning_rate": 2.037854373128889e-07,
+      "loss": 1.0356,
+      "step": 646
+    },
+    {
+      "epoch": 0.9380210221094599,
+      "grad_norm": 1.3166306499332168,
+      "learning_rate": 1.9444774069555694e-07,
+      "loss": 1.0659,
+      "step": 647
+    },
+    {
+      "epoch": 0.9394708227618702,
+      "grad_norm": 1.3356856379094966,
+      "learning_rate": 1.8532692336611037e-07,
+      "loss": 1.0302,
+      "step": 648
+    },
+    {
+      "epoch": 0.9409206234142805,
+      "grad_norm": 1.2101970960464068,
+      "learning_rate": 1.7642318705886284e-07,
+      "loss": 1.0596,
+      "step": 649
+    },
+    {
+      "epoch": 0.9423704240666908,
+      "grad_norm": 1.3510969943577162,
+      "learning_rate": 1.677367287067322e-07,
+      "loss": 1.0345,
+      "step": 650
+    },
+    {
+      "epoch": 0.9438202247191011,
+      "grad_norm": 1.3126149433317704,
+      "learning_rate": 1.5926774043687366e-07,
+      "loss": 0.9589,
+      "step": 651
+    },
+    {
+      "epoch": 0.9452700253715114,
+      "grad_norm": 1.3337510610140673,
+      "learning_rate": 1.510164095664457e-07,
+      "loss": 1.0302,
+      "step": 652
+    },
+    {
+      "epoch": 0.9467198260239217,
+      "grad_norm": 1.3093648667745805,
+      "learning_rate": 1.4298291859845215e-07,
+      "loss": 1.0574,
+      "step": 653
+    },
+    {
+      "epoch": 0.948169626676332,
+      "grad_norm": 1.3559668682140107,
+      "learning_rate": 1.351674452177143e-07,
+      "loss": 1.0572,
+      "step": 654
+    },
+    {
+      "epoch": 0.9496194273287423,
+      "grad_norm": 1.4326848635245282,
+      "learning_rate": 1.2757016228693964e-07,
+      "loss": 1.0294,
+      "step": 655
+    },
+    {
+      "epoch": 0.9510692279811526,
+      "grad_norm": 1.2637647597051096,
+      "learning_rate": 1.2019123784289488e-07,
+      "loss": 1.013,
+      "step": 656
+    },
+    {
+      "epoch": 0.9525190286335629,
+      "grad_norm": 1.3520946289830953,
+      "learning_rate": 1.1303083509269452e-07,
+      "loss": 0.974,
+      "step": 657
+    },
+    {
+      "epoch": 0.9539688292859732,
+      "grad_norm": 1.1875589206750796,
+      "learning_rate": 1.0608911241018594e-07,
+      "loss": 0.9423,
+      "step": 658
+    },
+    {
+      "epoch": 0.9554186299383834,
+      "grad_norm": 1.3857200485808747,
+      "learning_rate": 9.936622333245105e-08,
+      "loss": 1.0628,
+      "step": 659
+    },
+    {
+      "epoch": 0.9568684305907937,
+      "grad_norm": 1.2676205708371784,
+      "learning_rate": 9.286231655640799e-08,
+      "loss": 1.0091,
+      "step": 660
+    },
+    {
+      "epoch": 0.958318231243204,
+      "grad_norm": 1.290649259272742,
+      "learning_rate": 8.657753593552142e-08,
+      "loss": 0.9096,
+      "step": 661
+    },
+    {
+      "epoch": 0.9597680318956143,
+      "grad_norm": 1.3482301177902623,
+      "learning_rate": 8.051202047662187e-08,
+      "loss": 1.0276,
+      "step": 662
+    },
+    {
+      "epoch": 0.9612178325480246,
+      "grad_norm": 1.3086694012137101,
+      "learning_rate": 7.466590433683252e-08,
+      "loss": 0.9633,
+      "step": 663
+    },
+    {
+      "epoch": 0.9626676332004349,
+      "grad_norm": 1.1925364513139989,
+      "learning_rate": 6.903931682059828e-08,
+      "loss": 0.9412,
+      "step": 664
+    },
+    {
+      "epoch": 0.9641174338528452,
+      "grad_norm": 1.2762038893581211,
+      "learning_rate": 6.363238237683034e-08,
+      "loss": 0.9233,
+      "step": 665
+    },
+    {
+      "epoch": 0.9655672345052555,
+      "grad_norm": 1.3822925348445716,
+      "learning_rate": 5.8445220596149434e-08,
+      "loss": 0.9634,
+      "step": 666
+    },
+    {
+      "epoch": 0.9670170351576658,
+      "grad_norm": 1.277114768288311,
+      "learning_rate": 5.347794620824576e-08,
+      "loss": 0.8826,
+      "step": 667
+    },
+    {
+      "epoch": 0.9684668358100761,
+      "grad_norm": 1.3938462508056708,
+      "learning_rate": 4.8730669079335434e-08,
+      "loss": 1.0585,
+      "step": 668
+    },
+    {
+      "epoch": 0.9699166364624864,
+      "grad_norm": 1.2277802175201395,
+      "learning_rate": 4.420349420973358e-08,
+      "loss": 0.9827,
+      "step": 669
+    },
+    {
+      "epoch": 0.9713664371148967,
+      "grad_norm": 1.6217975006770653,
+      "learning_rate": 3.98965217315328e-08,
+      "loss": 0.9548,
+      "step": 670
+    },
+    {
+      "epoch": 0.972816237767307,
+      "grad_norm": 1.2109387277954269,
+      "learning_rate": 3.580984690638611e-08,
+      "loss": 0.9497,
+      "step": 671
+    },
+    {
+      "epoch": 0.9742660384197173,
+      "grad_norm": 1.3386935356665675,
+      "learning_rate": 3.194356012340305e-08,
+      "loss": 0.9515,
+      "step": 672
+    },
+    {
+      "epoch": 0.9757158390721276,
+      "grad_norm": 1.4088394737758427,
+      "learning_rate": 2.8297746897146816e-08,
+      "loss": 0.9924,
+      "step": 673
+    },
+    {
+      "epoch": 0.9771656397245378,
+      "grad_norm": 1.3685399200371613,
+      "learning_rate": 2.4872487865744698e-08,
+      "loss": 1.0685,
+      "step": 674
+    },
+    {
+      "epoch": 0.9786154403769481,
+      "grad_norm": 1.2901496226578117,
+      "learning_rate": 2.1667858789105046e-08,
+      "loss": 0.9616,
+      "step": 675
+    },
+    {
+      "epoch": 0.9800652410293584,
+      "grad_norm": 1.3681492953692462,
+      "learning_rate": 1.8683930547243044e-08,
+      "loss": 1.0457,
+      "step": 676
+    },
+    {
+      "epoch": 0.9815150416817687,
+      "grad_norm": 1.3445299630727452,
+      "learning_rate": 1.592076913870644e-08,
+      "loss": 1.0001,
+      "step": 677
+    },
+    {
+      "epoch": 0.982964842334179,
+      "grad_norm": 1.2616793644470028,
+      "learning_rate": 1.3378435679122225e-08,
+      "loss": 0.9837,
+      "step": 678
+    },
+    {
+      "epoch": 0.9844146429865893,
+      "grad_norm": 1.2872826628783247,
+      "learning_rate": 1.1056986399845537e-08,
+      "loss": 0.9735,
+      "step": 679
+    },
+    {
+      "epoch": 0.9858644436389996,
+      "grad_norm": 1.3869839709058764,
+      "learning_rate": 8.956472646710624e-09,
+      "loss": 1.0459,
+      "step": 680
+    },
+    {
+      "epoch": 0.9873142442914099,
+      "grad_norm": 1.3610292722253483,
+      "learning_rate": 7.07694087889621e-09,
+      "loss": 1.0229,
+      "step": 681
+    },
+    {
+      "epoch": 0.9887640449438202,
+      "grad_norm": 1.3605085129139796,
+      "learning_rate": 5.418432667905204e-09,
+      "loss": 0.9873,
+      "step": 682
+    },
+    {
+      "epoch": 0.9902138455962305,
+      "grad_norm": 1.4637076983888284,
+      "learning_rate": 3.9809846966343265e-09,
+      "loss": 1.0355,
+      "step": 683
+    },
+    {
+      "epoch": 0.9916636462486408,
+      "grad_norm": 1.3033729621214818,
+      "learning_rate": 2.764628758570309e-09,
+      "loss": 0.9862,
+      "step": 684
+    },
+    {
+      "epoch": 0.9931134469010511,
+      "grad_norm": 1.272655389886507,
+      "learning_rate": 1.7693917570837938e-09,
+      "loss": 0.9795,
+      "step": 685
+    },
+    {
+      "epoch": 0.9945632475534614,
+      "grad_norm": 1.5329645413361683,
+      "learning_rate": 9.95295704835364e-10,
+      "loss": 0.9747,
+      "step": 686
+    },
+    {
+      "epoch": 0.9960130482058717,
+      "grad_norm": 1.1682022017992713,
+      "learning_rate": 4.4235772328815594e-10,
+      "loss": 0.9575,
+      "step": 687
+    },
+    {
+      "epoch": 0.997462848858282,
+      "grad_norm": 1.294658750709239,
+      "learning_rate": 1.1059004233038296e-10,
+      "loss": 0.9938,
+      "step": 688
+    },
+    {
+      "epoch": 0.9989126495106923,
+      "grad_norm": 1.3461895076729873,
+      "learning_rate": 0.0,
+      "loss": 1.0023,
+      "step": 689
+    },
+    {
+      "epoch": 0.9989126495106923,
+      "step": 689,
+      "total_flos": 6993087226183680.0,
+      "train_loss": 1.0724846332270452,
+      "train_runtime": 12824.8689,
+      "train_samples_per_second": 13.766,
+      "train_steps_per_second": 0.054
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 689,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 173,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6993087226183680.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}