diff --git "a/checkpoint-900/trainer_state.json" "b/checkpoint-900/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-900/trainer_state.json"
@@ -0,0 +1,6342 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.622514265951928,
+  "eval_steps": 500,
+  "global_step": 900,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006916825177243646,
+      "grad_norm": 0.2954893708229065,
+      "learning_rate": 0.0,
+      "loss": 2.3812,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013833650354487291,
+      "grad_norm": 0.3372870683670044,
+      "learning_rate": 1.3698630136986302e-06,
+      "loss": 2.6562,
+      "step": 2
+    },
+    {
+      "epoch": 0.0020750475531730937,
+      "grad_norm": 0.41708794236183167,
+      "learning_rate": 2.7397260273972604e-06,
+      "loss": 2.779,
+      "step": 3
+    },
+    {
+      "epoch": 0.0027667300708974583,
+      "grad_norm": 0.48284393548965454,
+      "learning_rate": 4.10958904109589e-06,
+      "loss": 2.83,
+      "step": 4
+    },
+    {
+      "epoch": 0.0034584125886218224,
+      "grad_norm": 0.328521728515625,
+      "learning_rate": 5.479452054794521e-06,
+      "loss": 2.3607,
+      "step": 5
+    },
+    {
+      "epoch": 0.004150095106346187,
+      "grad_norm": 0.3413754403591156,
+      "learning_rate": 6.849315068493151e-06,
+      "loss": 1.9744,
+      "step": 6
+    },
+    {
+      "epoch": 0.0048417776240705515,
+      "grad_norm": 0.43425995111465454,
+      "learning_rate": 8.21917808219178e-06,
+      "loss": 1.915,
+      "step": 7
+    },
+    {
+      "epoch": 0.0055334601417949165,
+      "grad_norm": 0.45184195041656494,
+      "learning_rate": 9.589041095890411e-06,
+      "loss": 2.0002,
+      "step": 8
+    },
+    {
+      "epoch": 0.006225142659519281,
+      "grad_norm": 0.3451146185398102,
+      "learning_rate": 1.0958904109589042e-05,
+      "loss": 2.5739,
+      "step": 9
+    },
+    {
+      "epoch": 0.006916825177243645,
+      "grad_norm": 0.39400503039360046,
+      "learning_rate": 1.2328767123287671e-05,
+      "loss": 2.6139,
+      "step": 10
+    },
+    {
+      "epoch": 0.00760850769496801,
+      "grad_norm": 0.4555888772010803,
+      "learning_rate": 1.3698630136986302e-05,
+      "loss": 2.4346,
+      "step": 11
+    },
+    {
+      "epoch": 0.008300190212692375,
+      "grad_norm": 0.5475619435310364,
+      "learning_rate": 1.5068493150684931e-05,
+      "loss": 1.485,
+      "step": 12
+    },
+    {
+      "epoch": 0.008991872730416739,
+      "grad_norm": 1.219076156616211,
+      "learning_rate": 1.643835616438356e-05,
+      "loss": 2.0555,
+      "step": 13
+    },
+    {
+      "epoch": 0.009683555248141103,
+      "grad_norm": 0.3271433711051941,
+      "learning_rate": 1.780821917808219e-05,
+      "loss": 2.8418,
+      "step": 14
+    },
+    {
+      "epoch": 0.010375237765865467,
+      "grad_norm": 0.3261394500732422,
+      "learning_rate": 1.9178082191780822e-05,
+      "loss": 2.287,
+      "step": 15
+    },
+    {
+      "epoch": 0.011066920283589833,
+      "grad_norm": 0.39713796973228455,
+      "learning_rate": 2.0547945205479453e-05,
+      "loss": 2.7331,
+      "step": 16
+    },
+    {
+      "epoch": 0.011758602801314197,
+      "grad_norm": 0.3226444125175476,
+      "learning_rate": 2.1917808219178083e-05,
+      "loss": 2.7307,
+      "step": 17
+    },
+    {
+      "epoch": 0.012450285319038561,
+      "grad_norm": 0.4153122007846832,
+      "learning_rate": 2.328767123287671e-05,
+      "loss": 2.3229,
+      "step": 18
+    },
+    {
+      "epoch": 0.013141967836762926,
+      "grad_norm": 0.29302462935447693,
+      "learning_rate": 2.4657534246575342e-05,
+      "loss": 2.2093,
+      "step": 19
+    },
+    {
+      "epoch": 0.01383365035448729,
+      "grad_norm": 0.42535701394081116,
+      "learning_rate": 2.6027397260273973e-05,
+      "loss": 2.3698,
+      "step": 20
+    },
+    {
+      "epoch": 0.014525332872211656,
+      "grad_norm": 0.33567023277282715,
+      "learning_rate": 2.7397260273972603e-05,
+      "loss": 2.6456,
+      "step": 21
+    },
+    {
+      "epoch": 0.01521701538993602,
+      "grad_norm": 0.37375620007514954,
+      "learning_rate": 2.8767123287671234e-05,
+      "loss": 2.4148,
+      "step": 22
+    },
+    {
+      "epoch": 0.015908697907660384,
+      "grad_norm": 0.2704203128814697,
+      "learning_rate": 3.0136986301369862e-05,
+      "loss": 1.6722,
+      "step": 23
+    },
+    {
+      "epoch": 0.01660038042538475,
+      "grad_norm": 0.3946782946586609,
+      "learning_rate": 3.1506849315068496e-05,
+      "loss": 2.6759,
+      "step": 24
+    },
+    {
+      "epoch": 0.017292062943109112,
+      "grad_norm": 0.368335098028183,
+      "learning_rate": 3.287671232876712e-05,
+      "loss": 2.6882,
+      "step": 25
+    },
+    {
+      "epoch": 0.017983745460833478,
+      "grad_norm": 0.38029783964157104,
+      "learning_rate": 3.424657534246575e-05,
+      "loss": 2.6915,
+      "step": 26
+    },
+    {
+      "epoch": 0.01867542797855784,
+      "grad_norm": 0.36253222823143005,
+      "learning_rate": 3.561643835616438e-05,
+      "loss": 2.1426,
+      "step": 27
+    },
+    {
+      "epoch": 0.019367110496282206,
+      "grad_norm": 0.35769322514533997,
+      "learning_rate": 3.698630136986301e-05,
+      "loss": 2.6634,
+      "step": 28
+    },
+    {
+      "epoch": 0.020058793014006572,
+      "grad_norm": 0.44577229022979736,
+      "learning_rate": 3.8356164383561644e-05,
+      "loss": 2.5996,
+      "step": 29
+    },
+    {
+      "epoch": 0.020750475531730934,
+      "grad_norm": 0.5225628614425659,
+      "learning_rate": 3.9726027397260274e-05,
+      "loss": 2.5993,
+      "step": 30
+    },
+    {
+      "epoch": 0.0214421580494553,
+      "grad_norm": 0.45649707317352295,
+      "learning_rate": 4.1095890410958905e-05,
+      "loss": 2.3344,
+      "step": 31
+    },
+    {
+      "epoch": 0.022133840567179666,
+      "grad_norm": 0.31408172845840454,
+      "learning_rate": 4.2465753424657536e-05,
+      "loss": 2.343,
+      "step": 32
+    },
+    {
+      "epoch": 0.02282552308490403,
+      "grad_norm": 0.3498693108558655,
+      "learning_rate": 4.383561643835617e-05,
+      "loss": 1.7196,
+      "step": 33
+    },
+    {
+      "epoch": 0.023517205602628394,
+      "grad_norm": 0.5255292654037476,
+      "learning_rate": 4.520547945205479e-05,
+      "loss": 1.5354,
+      "step": 34
+    },
+    {
+      "epoch": 0.024208888120352757,
+      "grad_norm": 0.3460487425327301,
+      "learning_rate": 4.657534246575342e-05,
+      "loss": 2.1435,
+      "step": 35
+    },
+    {
+      "epoch": 0.024900570638077123,
+      "grad_norm": 0.47466233372688293,
+      "learning_rate": 4.794520547945205e-05,
+      "loss": 2.7025,
+      "step": 36
+    },
+    {
+      "epoch": 0.02559225315580149,
+      "grad_norm": 0.39667466282844543,
+      "learning_rate": 4.9315068493150684e-05,
+      "loss": 2.2736,
+      "step": 37
+    },
+    {
+      "epoch": 0.02628393567352585,
+      "grad_norm": 0.40798133611679077,
+      "learning_rate": 5.068493150684932e-05,
+      "loss": 1.4074,
+      "step": 38
+    },
+    {
+      "epoch": 0.026975618191250217,
+      "grad_norm": 0.39603471755981445,
+      "learning_rate": 5.2054794520547945e-05,
+      "loss": 2.4354,
+      "step": 39
+    },
+    {
+      "epoch": 0.02766730070897458,
+      "grad_norm": 0.5650593042373657,
+      "learning_rate": 5.342465753424658e-05,
+      "loss": 1.7307,
+      "step": 40
+    },
+    {
+      "epoch": 0.028358983226698945,
+      "grad_norm": 0.7711919546127319,
+      "learning_rate": 5.479452054794521e-05,
+      "loss": 1.2067,
+      "step": 41
+    },
+    {
+      "epoch": 0.02905066574442331,
+      "grad_norm": 0.46603119373321533,
+      "learning_rate": 5.616438356164384e-05,
+      "loss": 1.0539,
+      "step": 42
+    },
+    {
+      "epoch": 0.029742348262147673,
+      "grad_norm": 0.2997550666332245,
+      "learning_rate": 5.753424657534247e-05,
+      "loss": 1.9997,
+      "step": 43
+    },
+    {
+      "epoch": 0.03043403077987204,
+      "grad_norm": 0.4058617651462555,
+      "learning_rate": 5.89041095890411e-05,
+      "loss": 2.3323,
+      "step": 44
+    },
+    {
+      "epoch": 0.0311257132975964,
+      "grad_norm": 0.3763635754585266,
+      "learning_rate": 6.0273972602739724e-05,
+      "loss": 1.8749,
+      "step": 45
+    },
+    {
+      "epoch": 0.03181739581532077,
+      "grad_norm": 0.3507993221282959,
+      "learning_rate": 6.164383561643835e-05,
+      "loss": 2.3828,
+      "step": 46
+    },
+    {
+      "epoch": 0.03250907833304513,
+      "grad_norm": 0.34240859746932983,
+      "learning_rate": 6.301369863013699e-05,
+      "loss": 2.3809,
+      "step": 47
+    },
+    {
+      "epoch": 0.0332007608507695,
+      "grad_norm": 0.4183844029903412,
+      "learning_rate": 6.438356164383562e-05,
+      "loss": 2.2065,
+      "step": 48
+    },
+    {
+      "epoch": 0.03389244336849386,
+      "grad_norm": 0.5209120512008667,
+      "learning_rate": 6.575342465753424e-05,
+      "loss": 2.061,
+      "step": 49
+    },
+    {
+      "epoch": 0.034584125886218224,
+      "grad_norm": 0.6568111181259155,
+      "learning_rate": 6.712328767123288e-05,
+      "loss": 2.1576,
+      "step": 50
+    },
+    {
+      "epoch": 0.03527580840394259,
+      "grad_norm": 0.5327390432357788,
+      "learning_rate": 6.84931506849315e-05,
+      "loss": 1.5562,
+      "step": 51
+    },
+    {
+      "epoch": 0.035967490921666956,
+      "grad_norm": 0.6389997005462646,
+      "learning_rate": 6.986301369863014e-05,
+      "loss": 2.3966,
+      "step": 52
+    },
+    {
+      "epoch": 0.03665917343939132,
+      "grad_norm": 0.6374388933181763,
+      "learning_rate": 7.123287671232876e-05,
+      "loss": 2.4013,
+      "step": 53
+    },
+    {
+      "epoch": 0.03735085595711568,
+      "grad_norm": 0.4383719563484192,
+      "learning_rate": 7.26027397260274e-05,
+      "loss": 2.061,
+      "step": 54
+    },
+    {
+      "epoch": 0.03804253847484005,
+      "grad_norm": 0.4306814670562744,
+      "learning_rate": 7.397260273972603e-05,
+      "loss": 2.4931,
+      "step": 55
+    },
+    {
+      "epoch": 0.03873422099256441,
+      "grad_norm": 0.5693032145500183,
+      "learning_rate": 7.534246575342466e-05,
+      "loss": 2.2842,
+      "step": 56
+    },
+    {
+      "epoch": 0.039425903510288775,
+      "grad_norm": 0.7792916297912598,
+      "learning_rate": 7.671232876712329e-05,
+      "loss": 1.8452,
+      "step": 57
+    },
+    {
+      "epoch": 0.040117586028013144,
+      "grad_norm": 0.4936494827270508,
+      "learning_rate": 7.808219178082192e-05,
+      "loss": 1.8668,
+      "step": 58
+    },
+    {
+      "epoch": 0.040809268545737507,
+      "grad_norm": 0.49626269936561584,
+      "learning_rate": 7.945205479452055e-05,
+      "loss": 1.9564,
+      "step": 59
+    },
+    {
+      "epoch": 0.04150095106346187,
+      "grad_norm": 0.4702766537666321,
+      "learning_rate": 8.082191780821919e-05,
+      "loss": 2.419,
+      "step": 60
+    },
+    {
+      "epoch": 0.04219263358118624,
+      "grad_norm": 0.6959900259971619,
+      "learning_rate": 8.219178082191781e-05,
+      "loss": 1.5835,
+      "step": 61
+    },
+    {
+      "epoch": 0.0428843160989106,
+      "grad_norm": 0.5963976979255676,
+      "learning_rate": 8.356164383561645e-05,
+      "loss": 2.2932,
+      "step": 62
+    },
+    {
+      "epoch": 0.04357599861663496,
+      "grad_norm": 0.4605785310268402,
+      "learning_rate": 8.493150684931507e-05,
+      "loss": 2.397,
+      "step": 63
+    },
+    {
+      "epoch": 0.04426768113435933,
+      "grad_norm": 0.4479796588420868,
+      "learning_rate": 8.630136986301371e-05,
+      "loss": 2.0257,
+      "step": 64
+    },
+    {
+      "epoch": 0.044959363652083695,
+      "grad_norm": 0.4342099726200104,
+      "learning_rate": 8.767123287671233e-05,
+      "loss": 1.9476,
+      "step": 65
+    },
+    {
+      "epoch": 0.04565104616980806,
+      "grad_norm": 0.44643279910087585,
+      "learning_rate": 8.904109589041096e-05,
+      "loss": 2.4669,
+      "step": 66
+    },
+    {
+      "epoch": 0.04634272868753242,
+      "grad_norm": 0.513254702091217,
+      "learning_rate": 9.041095890410958e-05,
+      "loss": 2.2581,
+      "step": 67
+    },
+    {
+      "epoch": 0.04703441120525679,
+      "grad_norm": 0.6621958613395691,
+      "learning_rate": 9.178082191780822e-05,
+      "loss": 1.3013,
+      "step": 68
+    },
+    {
+      "epoch": 0.04772609372298115,
+      "grad_norm": 0.43388229608535767,
+      "learning_rate": 9.315068493150684e-05,
+      "loss": 2.5325,
+      "step": 69
+    },
+    {
+      "epoch": 0.048417776240705514,
+      "grad_norm": 0.539757490158081,
+      "learning_rate": 9.452054794520548e-05,
+      "loss": 1.76,
+      "step": 70
+    },
+    {
+      "epoch": 0.04910945875842988,
+      "grad_norm": 0.33598029613494873,
+      "learning_rate": 9.58904109589041e-05,
+      "loss": 1.6931,
+      "step": 71
+    },
+    {
+      "epoch": 0.049801141276154245,
+      "grad_norm": 0.8661749958992004,
+      "learning_rate": 9.726027397260274e-05,
+      "loss": 2.1326,
+      "step": 72
+    },
+    {
+      "epoch": 0.05049282379387861,
+      "grad_norm": 0.5956202149391174,
+      "learning_rate": 9.863013698630137e-05,
+      "loss": 2.0554,
+      "step": 73
+    },
+    {
+      "epoch": 0.05118450631160298,
+      "grad_norm": 0.7723128199577332,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 74
+    },
+    {
+      "epoch": 0.05187618882932734,
+      "grad_norm": 0.5357879996299744,
+      "learning_rate": 9.992716678805537e-05,
+      "loss": 2.1688,
+      "step": 75
+    },
+    {
+      "epoch": 0.0525678713470517,
+      "grad_norm": 0.48623108863830566,
+      "learning_rate": 9.985433357611071e-05,
+      "loss": 2.1567,
+      "step": 76
+    },
+    {
+      "epoch": 0.053259553864776064,
+      "grad_norm": 0.4244164228439331,
+      "learning_rate": 9.978150036416607e-05,
+      "loss": 1.8235,
+      "step": 77
+    },
+    {
+      "epoch": 0.053951236382500434,
+      "grad_norm": 0.6056337356567383,
+      "learning_rate": 9.970866715222141e-05,
+      "loss": 2.0154,
+      "step": 78
+    },
+    {
+      "epoch": 0.054642918900224796,
+      "grad_norm": 0.5280382037162781,
+      "learning_rate": 9.963583394027677e-05,
+      "loss": 1.9959,
+      "step": 79
+    },
+    {
+      "epoch": 0.05533460141794916,
+      "grad_norm": 0.44745901226997375,
+      "learning_rate": 9.956300072833212e-05,
+      "loss": 1.2171,
+      "step": 80
+    },
+    {
+      "epoch": 0.05602628393567353,
+      "grad_norm": 0.7180289030075073,
+      "learning_rate": 9.949016751638748e-05,
+      "loss": 1.7036,
+      "step": 81
+    },
+    {
+      "epoch": 0.05671796645339789,
+      "grad_norm": 0.6465222239494324,
+      "learning_rate": 9.941733430444284e-05,
+      "loss": 1.8992,
+      "step": 82
+    },
+    {
+      "epoch": 0.05740964897112225,
+      "grad_norm": 0.4792153239250183,
+      "learning_rate": 9.934450109249819e-05,
+      "loss": 2.0712,
+      "step": 83
+    },
+    {
+      "epoch": 0.05810133148884662,
+      "grad_norm": 0.47839808464050293,
+      "learning_rate": 9.927166788055353e-05,
+      "loss": 1.7548,
+      "step": 84
+    },
+    {
+      "epoch": 0.058793014006570984,
+      "grad_norm": 0.7118557095527649,
+      "learning_rate": 9.919883466860888e-05,
+      "loss": 1.5835,
+      "step": 85
+    },
+    {
+      "epoch": 0.05948469652429535,
+      "grad_norm": 0.5179592370986938,
+      "learning_rate": 9.912600145666424e-05,
+      "loss": 1.621,
+      "step": 86
+    },
+    {
+      "epoch": 0.060176379042019716,
+      "grad_norm": 0.9235703349113464,
+      "learning_rate": 9.905316824471959e-05,
+      "loss": 0.8581,
+      "step": 87
+    },
+    {
+      "epoch": 0.06086806155974408,
+      "grad_norm": 0.8734163045883179,
+      "learning_rate": 9.898033503277495e-05,
+      "loss": 2.2878,
+      "step": 88
+    },
+    {
+      "epoch": 0.06155974407746844,
+      "grad_norm": 0.45049625635147095,
+      "learning_rate": 9.890750182083031e-05,
+      "loss": 2.2186,
+      "step": 89
+    },
+    {
+      "epoch": 0.0622514265951928,
+      "grad_norm": 0.6119429469108582,
+      "learning_rate": 9.883466860888566e-05,
+      "loss": 2.2461,
+      "step": 90
+    },
+    {
+      "epoch": 0.06294310911291717,
+      "grad_norm": 0.5095340013504028,
+      "learning_rate": 9.876183539694101e-05,
+      "loss": 2.2954,
+      "step": 91
+    },
+    {
+      "epoch": 0.06363479163064154,
+      "grad_norm": 0.6094918251037598,
+      "learning_rate": 9.868900218499635e-05,
+      "loss": 2.3478,
+      "step": 92
+    },
+    {
+      "epoch": 0.0643264741483659,
+      "grad_norm": 1.8420301675796509,
+      "learning_rate": 9.861616897305172e-05,
+      "loss": 2.213,
+      "step": 93
+    },
+    {
+      "epoch": 0.06501815666609026,
+      "grad_norm": 0.6151532530784607,
+      "learning_rate": 9.854333576110706e-05,
+      "loss": 1.8737,
+      "step": 94
+    },
+    {
+      "epoch": 0.06570983918381462,
+      "grad_norm": 0.5210903286933899,
+      "learning_rate": 9.847050254916242e-05,
+      "loss": 2.019,
+      "step": 95
+    },
+    {
+      "epoch": 0.066401521701539,
+      "grad_norm": 0.535746693611145,
+      "learning_rate": 9.839766933721779e-05,
+      "loss": 2.399,
+      "step": 96
+    },
+    {
+      "epoch": 0.06709320421926336,
+      "grad_norm": 0.8154505491256714,
+      "learning_rate": 9.832483612527313e-05,
+      "loss": 1.9442,
+      "step": 97
+    },
+    {
+      "epoch": 0.06778488673698772,
+      "grad_norm": 0.521247923374176,
+      "learning_rate": 9.825200291332848e-05,
+      "loss": 2.5264,
+      "step": 98
+    },
+    {
+      "epoch": 0.06847656925471209,
+      "grad_norm": 0.41544124484062195,
+      "learning_rate": 9.817916970138383e-05,
+      "loss": 1.8267,
+      "step": 99
+    },
+    {
+      "epoch": 0.06916825177243645,
+      "grad_norm": 0.538253128528595,
+      "learning_rate": 9.810633648943919e-05,
+      "loss": 1.6463,
+      "step": 100
+    },
+    {
+      "epoch": 0.06985993429016081,
+      "grad_norm": 0.8509485125541687,
+      "learning_rate": 9.803350327749454e-05,
+      "loss": 1.5877,
+      "step": 101
+    },
+    {
+      "epoch": 0.07055161680788519,
+      "grad_norm": 0.47714293003082275,
+      "learning_rate": 9.79606700655499e-05,
+      "loss": 1.5043,
+      "step": 102
+    },
+    {
+      "epoch": 0.07124329932560955,
+      "grad_norm": 0.5433914065361023,
+      "learning_rate": 9.788783685360526e-05,
+      "loss": 1.6454,
+      "step": 103
+    },
+    {
+      "epoch": 0.07193498184333391,
+      "grad_norm": 0.6400073766708374,
+      "learning_rate": 9.781500364166059e-05,
+      "loss": 1.6994,
+      "step": 104
+    },
+    {
+      "epoch": 0.07262666436105827,
+      "grad_norm": 0.6944612860679626,
+      "learning_rate": 9.774217042971595e-05,
+      "loss": 2.2063,
+      "step": 105
+    },
+    {
+      "epoch": 0.07331834687878264,
+      "grad_norm": 0.6701168417930603,
+      "learning_rate": 9.76693372177713e-05,
+      "loss": 1.2487,
+      "step": 106
+    },
+    {
+      "epoch": 0.074010029396507,
+      "grad_norm": 0.7888538837432861,
+      "learning_rate": 9.759650400582666e-05,
+      "loss": 1.8784,
+      "step": 107
+    },
+    {
+      "epoch": 0.07470171191423136,
+      "grad_norm": 0.7106865048408508,
+      "learning_rate": 9.752367079388201e-05,
+      "loss": 2.3673,
+      "step": 108
+    },
+    {
+      "epoch": 0.07539339443195574,
+      "grad_norm": 0.49588626623153687,
+      "learning_rate": 9.745083758193737e-05,
+      "loss": 1.8638,
+      "step": 109
+    },
+    {
+      "epoch": 0.0760850769496801,
+      "grad_norm": 0.48046427965164185,
+      "learning_rate": 9.737800436999273e-05,
+      "loss": 2.333,
+      "step": 110
+    },
+    {
+      "epoch": 0.07677675946740446,
+      "grad_norm": 0.6239060163497925,
+      "learning_rate": 9.730517115804807e-05,
+      "loss": 1.8378,
+      "step": 111
+    },
+    {
+      "epoch": 0.07746844198512882,
+      "grad_norm": 0.6083034873008728,
+      "learning_rate": 9.723233794610343e-05,
+      "loss": 1.9826,
+      "step": 112
+    },
+    {
+      "epoch": 0.07816012450285319,
+      "grad_norm": 0.8679235577583313,
+      "learning_rate": 9.715950473415877e-05,
+      "loss": 1.3375,
+      "step": 113
+    },
+    {
+      "epoch": 0.07885180702057755,
+      "grad_norm": 0.542757511138916,
+      "learning_rate": 9.708667152221414e-05,
+      "loss": 1.6349,
+      "step": 114
+    },
+    {
+      "epoch": 0.07954348953830193,
+      "grad_norm": 0.5968992114067078,
+      "learning_rate": 9.701383831026948e-05,
+      "loss": 1.3598,
+      "step": 115
+    },
+    {
+      "epoch": 0.08023517205602629,
+      "grad_norm": 0.5693714022636414,
+      "learning_rate": 9.694100509832484e-05,
+      "loss": 2.1527,
+      "step": 116
+    },
+    {
+      "epoch": 0.08092685457375065,
+      "grad_norm": 0.6307690143585205,
+      "learning_rate": 9.68681718863802e-05,
+      "loss": 1.7178,
+      "step": 117
+    },
+    {
+      "epoch": 0.08161853709147501,
+      "grad_norm": 0.5192516446113586,
+      "learning_rate": 9.679533867443554e-05,
+      "loss": 1.8115,
+      "step": 118
+    },
+    {
+      "epoch": 0.08231021960919938,
+      "grad_norm": 1.036036729812622,
+      "learning_rate": 9.67225054624909e-05,
+      "loss": 2.2423,
+      "step": 119
+    },
+    {
+      "epoch": 0.08300190212692374,
+      "grad_norm": 0.6450679302215576,
+      "learning_rate": 9.664967225054625e-05,
+      "loss": 1.9298,
+      "step": 120
+    },
+    {
+      "epoch": 0.0836935846446481,
+      "grad_norm": 0.8683953881263733,
+      "learning_rate": 9.657683903860161e-05,
+      "loss": 2.0791,
+      "step": 121
+    },
+    {
+      "epoch": 0.08438526716237248,
+      "grad_norm": 0.7013575434684753,
+      "learning_rate": 9.650400582665696e-05,
+      "loss": 2.0814,
+      "step": 122
+    },
+    {
+      "epoch": 0.08507694968009684,
+      "grad_norm": 0.6849325299263,
+      "learning_rate": 9.643117261471232e-05,
+      "loss": 1.83,
+      "step": 123
+    },
+    {
+      "epoch": 0.0857686321978212,
+      "grad_norm": 0.9820392727851868,
+      "learning_rate": 9.635833940276767e-05,
+      "loss": 1.5665,
+      "step": 124
+    },
+    {
+      "epoch": 0.08646031471554556,
+      "grad_norm": 0.5546866655349731,
+      "learning_rate": 9.628550619082301e-05,
+      "loss": 1.7215,
+      "step": 125
+    },
+    {
+      "epoch": 0.08715199723326993,
+      "grad_norm": 0.6748363971710205,
+      "learning_rate": 9.621267297887837e-05,
+      "loss": 1.6532,
+      "step": 126
+    },
+    {
+      "epoch": 0.08784367975099429,
+      "grad_norm": 0.4760434627532959,
+      "learning_rate": 9.613983976693372e-05,
+      "loss": 1.3476,
+      "step": 127
+    },
+    {
+      "epoch": 0.08853536226871866,
+      "grad_norm": 0.5710906386375427,
+      "learning_rate": 9.606700655498908e-05,
+      "loss": 2.025,
+      "step": 128
+    },
+    {
+      "epoch": 0.08922704478644303,
+      "grad_norm": 0.5950302481651306,
+      "learning_rate": 9.599417334304443e-05,
+      "loss": 2.6324,
+      "step": 129
+    },
+    {
+      "epoch": 0.08991872730416739,
+      "grad_norm": 0.4994860887527466,
+      "learning_rate": 9.592134013109979e-05,
+      "loss": 2.0315,
+      "step": 130
+    },
+    {
+      "epoch": 0.09061040982189175,
+      "grad_norm": 0.6400249004364014,
+      "learning_rate": 9.584850691915514e-05,
+      "loss": 2.3628,
+      "step": 131
+    },
+    {
+      "epoch": 0.09130209233961611,
+      "grad_norm": 1.0873314142227173,
+      "learning_rate": 9.577567370721049e-05,
+      "loss": 1.2664,
+      "step": 132
+    },
+    {
+      "epoch": 0.09199377485734048,
+      "grad_norm": 0.5430288314819336,
+      "learning_rate": 9.570284049526585e-05,
+      "loss": 1.7676,
+      "step": 133
+    },
+    {
+      "epoch": 0.09268545737506484,
+      "grad_norm": 0.5969283580780029,
+      "learning_rate": 9.56300072833212e-05,
+      "loss": 2.0824,
+      "step": 134
+    },
+    {
+      "epoch": 0.09337713989278922,
+      "grad_norm": 0.6228020191192627,
+      "learning_rate": 9.555717407137656e-05,
+      "loss": 2.0687,
+      "step": 135
+    },
+    {
+      "epoch": 0.09406882241051358,
+      "grad_norm": 0.6905536651611328,
+      "learning_rate": 9.54843408594319e-05,
+      "loss": 2.2948,
+      "step": 136
+    },
+    {
+      "epoch": 0.09476050492823794,
+      "grad_norm": 0.6397948861122131,
+      "learning_rate": 9.541150764748726e-05,
+      "loss": 1.4147,
+      "step": 137
+    },
+    {
+      "epoch": 0.0954521874459623,
+      "grad_norm": 0.5489908456802368,
+      "learning_rate": 9.533867443554261e-05,
+      "loss": 2.0582,
+      "step": 138
+    },
+    {
+      "epoch": 0.09614386996368667,
+      "grad_norm": 0.5713220834732056,
+      "learning_rate": 9.526584122359796e-05,
+      "loss": 2.1708,
+      "step": 139
+    },
+    {
+      "epoch": 0.09683555248141103,
+      "grad_norm": 0.6736558079719543,
+      "learning_rate": 9.519300801165332e-05,
+      "loss": 1.6808,
+      "step": 140
+    },
+    {
+      "epoch": 0.0975272349991354,
+      "grad_norm": 1.1222553253173828,
+      "learning_rate": 9.512017479970867e-05,
+      "loss": 1.563,
+      "step": 141
+    },
+    {
+      "epoch": 0.09821891751685977,
+      "grad_norm": 0.6231582760810852,
+      "learning_rate": 9.504734158776403e-05,
+      "loss": 1.6458,
+      "step": 142
+    },
+    {
+      "epoch": 0.09891060003458413,
+      "grad_norm": 0.716089129447937,
+      "learning_rate": 9.497450837581938e-05,
+      "loss": 1.7055,
+      "step": 143
+    },
+    {
+      "epoch": 0.09960228255230849,
+      "grad_norm": 0.49629953503608704,
+      "learning_rate": 9.490167516387472e-05,
+      "loss": 1.2985,
+      "step": 144
+    },
+    {
+      "epoch": 0.10029396507003285,
+      "grad_norm": 0.8354843258857727,
+      "learning_rate": 9.482884195193008e-05,
+      "loss": 1.8557,
+      "step": 145
+    },
+    {
+      "epoch": 0.10098564758775722,
+      "grad_norm": 0.7507709860801697,
+      "learning_rate": 9.475600873998543e-05,
+      "loss": 1.1816,
+      "step": 146
+    },
+    {
+      "epoch": 0.10167733010548158,
+      "grad_norm": 0.8973211050033569,
+      "learning_rate": 9.46831755280408e-05,
+      "loss": 1.4495,
+      "step": 147
+    },
+    {
+      "epoch": 0.10236901262320595,
+      "grad_norm": 0.7720927000045776,
+      "learning_rate": 9.461034231609614e-05,
+      "loss": 2.1743,
+      "step": 148
+    },
+    {
+      "epoch": 0.10306069514093032,
+      "grad_norm": 0.511075496673584,
+      "learning_rate": 9.45375091041515e-05,
+      "loss": 1.8267,
+      "step": 149
+    },
+    {
+      "epoch": 0.10375237765865468,
+      "grad_norm": 0.6523205041885376,
+      "learning_rate": 9.446467589220685e-05,
+      "loss": 1.9682,
+      "step": 150
+    },
+    {
+      "epoch": 0.10444406017637904,
+      "grad_norm": 0.703565239906311,
+      "learning_rate": 9.43918426802622e-05,
+      "loss": 2.3272,
+      "step": 151
+    },
+    {
+      "epoch": 0.1051357426941034,
+      "grad_norm": 0.5773376226425171,
+      "learning_rate": 9.431900946831756e-05,
+      "loss": 1.9753,
+      "step": 152
+    },
+    {
+      "epoch": 0.10582742521182777,
+      "grad_norm": 0.6234054565429688,
+      "learning_rate": 9.42461762563729e-05,
+      "loss": 2.5085,
+      "step": 153
+    },
+    {
+      "epoch": 0.10651910772955213,
+      "grad_norm": 0.7076170444488525,
+      "learning_rate": 9.417334304442827e-05,
+      "loss": 1.9139,
+      "step": 154
+    },
+    {
+      "epoch": 0.1072107902472765,
+      "grad_norm": 0.5035070776939392,
+      "learning_rate": 9.410050983248361e-05,
+      "loss": 2.0164,
+      "step": 155
+    },
+    {
+      "epoch": 0.10790247276500087,
+      "grad_norm": 1.046075701713562,
+      "learning_rate": 9.402767662053898e-05,
+      "loss": 1.2478,
+      "step": 156
+    },
+    {
+      "epoch": 0.10859415528272523,
+      "grad_norm": 0.5620921850204468,
+      "learning_rate": 9.395484340859432e-05,
+      "loss": 1.565,
+      "step": 157
+    },
+    {
+      "epoch": 0.10928583780044959,
+      "grad_norm": 0.8284108638763428,
+      "learning_rate": 9.388201019664967e-05,
+      "loss": 2.0823,
+      "step": 158
+    },
+    {
+      "epoch": 0.10997752031817395,
+      "grad_norm": 0.584373950958252,
+      "learning_rate": 9.380917698470503e-05,
+      "loss": 1.6352,
+      "step": 159
+    },
+    {
+      "epoch": 0.11066920283589832,
+      "grad_norm": 0.7779399752616882,
+      "learning_rate": 9.373634377276038e-05,
+      "loss": 1.83,
+      "step": 160
+    },
+    {
+      "epoch": 0.1113608853536227,
+      "grad_norm": 0.6635437607765198,
+      "learning_rate": 9.366351056081574e-05,
+      "loss": 1.8935,
+      "step": 161
+    },
+    {
+      "epoch": 0.11205256787134706,
+      "grad_norm": 0.6441919207572937,
+      "learning_rate": 9.359067734887109e-05,
+      "loss": 1.9566,
+      "step": 162
+    },
+    {
+      "epoch": 0.11274425038907142,
+      "grad_norm": 1.20408296585083,
+      "learning_rate": 9.351784413692645e-05,
+      "loss": 1.2529,
+      "step": 163
+    },
+    {
+      "epoch": 0.11343593290679578,
+      "grad_norm": 0.6320855021476746,
+      "learning_rate": 9.34450109249818e-05,
+      "loss": 1.8071,
+      "step": 164
+    },
+    {
+      "epoch": 0.11412761542452014,
+      "grad_norm": 0.5197457075119019,
+      "learning_rate": 9.337217771303714e-05,
+      "loss": 1.7303,
+      "step": 165
+    },
+    {
+      "epoch": 0.1148192979422445,
+      "grad_norm": 0.6230719089508057,
+      "learning_rate": 9.32993445010925e-05,
+      "loss": 2.0685,
+      "step": 166
+    },
+    {
+      "epoch": 0.11551098045996887,
+      "grad_norm": 0.6185227036476135,
+      "learning_rate": 9.322651128914785e-05,
+      "loss": 1.9485,
+      "step": 167
+    },
+    {
+      "epoch": 0.11620266297769324,
+      "grad_norm": 0.7290086150169373,
+      "learning_rate": 9.315367807720321e-05,
+      "loss": 1.6923,
+      "step": 168
+    },
+    {
+      "epoch": 0.1168943454954176,
+      "grad_norm": 0.9122477769851685,
+      "learning_rate": 9.308084486525856e-05,
+      "loss": 1.3395,
+      "step": 169
+    },
+    {
+      "epoch": 0.11758602801314197,
+      "grad_norm": 0.5755578875541687,
+      "learning_rate": 9.300801165331392e-05,
+      "loss": 1.0775,
+      "step": 170
+    },
+    {
+      "epoch": 0.11827771053086633,
+      "grad_norm": 0.6516758799552917,
+      "learning_rate": 9.293517844136927e-05,
+      "loss": 1.6174,
+      "step": 171
+    },
+    {
+      "epoch": 0.1189693930485907,
+      "grad_norm": 0.5685858726501465,
+      "learning_rate": 9.286234522942462e-05,
+      "loss": 1.851,
+      "step": 172
+    },
+    {
+      "epoch": 0.11966107556631506,
+      "grad_norm": 0.5784198641777039,
+      "learning_rate": 9.278951201747998e-05,
+      "loss": 1.712,
+      "step": 173
+    },
+    {
+      "epoch": 0.12035275808403943,
+      "grad_norm": 0.7357056736946106,
+      "learning_rate": 9.271667880553533e-05,
+      "loss": 2.2286,
+      "step": 174
+    },
+    {
+      "epoch": 0.1210444406017638,
+      "grad_norm": 0.5122838616371155,
+      "learning_rate": 9.264384559359069e-05,
+      "loss": 1.5581,
+      "step": 175
+    },
+    {
+      "epoch": 0.12173612311948816,
+      "grad_norm": 0.6864873170852661,
+      "learning_rate": 9.257101238164603e-05,
+      "loss": 1.2737,
+      "step": 176
+    },
+    {
+      "epoch": 0.12242780563721252,
+      "grad_norm": 0.441771000623703,
+      "learning_rate": 9.24981791697014e-05,
+      "loss": 1.2291,
+      "step": 177
+    },
+    {
+      "epoch": 0.12311948815493688,
+      "grad_norm": 0.634203314781189,
+      "learning_rate": 9.242534595775674e-05,
+      "loss": 2.1797,
+      "step": 178
+    },
+    {
+      "epoch": 0.12381117067266124,
+      "grad_norm": 0.6232409477233887,
+      "learning_rate": 9.235251274581209e-05,
+      "loss": 1.5954,
+      "step": 179
+    },
+    {
+      "epoch": 0.1245028531903856,
+      "grad_norm": 0.7475627660751343,
+      "learning_rate": 9.227967953386745e-05,
+      "loss": 2.3014,
+      "step": 180
+    },
+    {
+      "epoch": 0.12519453570810998,
+      "grad_norm": 0.7436168789863586,
+      "learning_rate": 9.22068463219228e-05,
+      "loss": 2.1053,
+      "step": 181
+    },
+    {
+      "epoch": 0.12588621822583435,
+      "grad_norm": 0.754830002784729,
+      "learning_rate": 9.213401310997816e-05,
+      "loss": 2.2968,
+      "step": 182
+    },
+    {
+      "epoch": 0.1265779007435587,
+      "grad_norm": 0.7968791723251343,
+      "learning_rate": 9.206117989803351e-05,
+      "loss": 1.9433,
+      "step": 183
+    },
+    {
+      "epoch": 0.12726958326128307,
+      "grad_norm": 0.7031124234199524,
+      "learning_rate": 9.198834668608885e-05,
+      "loss": 1.6162,
+      "step": 184
+    },
+    {
+      "epoch": 0.12796126577900743,
+      "grad_norm": 1.5919851064682007,
+      "learning_rate": 9.191551347414422e-05,
+      "loss": 2.3995,
+      "step": 185
+    },
+    {
+      "epoch": 0.1286529482967318,
+      "grad_norm": 0.8990558385848999,
+      "learning_rate": 9.184268026219956e-05,
+      "loss": 2.1597,
+      "step": 186
+    },
+    {
+      "epoch": 0.12934463081445616,
+      "grad_norm": 0.6514426469802856,
+      "learning_rate": 9.176984705025492e-05,
+      "loss": 2.2172,
+      "step": 187
+    },
+    {
+      "epoch": 0.13003631333218052,
+      "grad_norm": 0.5041804313659668,
+      "learning_rate": 9.169701383831027e-05,
+      "loss": 1.6619,
+      "step": 188
+    },
+    {
+      "epoch": 0.13072799584990488,
+      "grad_norm": 0.5292882323265076,
+      "learning_rate": 9.162418062636563e-05,
+      "loss": 1.3204,
+      "step": 189
+    },
+    {
+      "epoch": 0.13141967836762924,
+      "grad_norm": 0.9279568791389465,
+      "learning_rate": 9.155134741442098e-05,
+      "loss": 1.2023,
+      "step": 190
+    },
+    {
+      "epoch": 0.13211136088535363,
+      "grad_norm": 0.8143125772476196,
+      "learning_rate": 9.147851420247633e-05,
+      "loss": 1.5768,
+      "step": 191
+    },
+    {
+      "epoch": 0.132803043403078,
+      "grad_norm": 0.9918676614761353,
+      "learning_rate": 9.140568099053169e-05,
+      "loss": 1.3813,
+      "step": 192
+    },
+    {
+      "epoch": 0.13349472592080236,
+      "grad_norm": 0.6452721357345581,
+      "learning_rate": 9.133284777858704e-05,
+      "loss": 2.4017,
+      "step": 193
+    },
+    {
+      "epoch": 0.13418640843852672,
+      "grad_norm": 0.7932286858558655,
+      "learning_rate": 9.12600145666424e-05,
+      "loss": 1.1203,
+      "step": 194
+    },
+    {
+      "epoch": 0.13487809095625108,
+      "grad_norm": 0.5378614068031311,
+      "learning_rate": 9.118718135469774e-05,
+      "loss": 1.9103,
+      "step": 195
+    },
+    {
+      "epoch": 0.13556977347397545,
+      "grad_norm": 0.6193660497665405,
+      "learning_rate": 9.11143481427531e-05,
+      "loss": 2.0729,
+      "step": 196
+    },
+    {
+      "epoch": 0.1362614559916998,
+      "grad_norm": 0.8437771797180176,
+      "learning_rate": 9.104151493080845e-05,
+      "loss": 2.4214,
+      "step": 197
+    },
+    {
+      "epoch": 0.13695313850942417,
+      "grad_norm": 0.8080397248268127,
+      "learning_rate": 9.09686817188638e-05,
+      "loss": 1.2323,
+      "step": 198
+    },
+    {
+      "epoch": 0.13764482102714853,
+      "grad_norm": 0.5907486081123352,
+      "learning_rate": 9.089584850691916e-05,
+      "loss": 1.6866,
+      "step": 199
+    },
+    {
+      "epoch": 0.1383365035448729,
+      "grad_norm": 0.6559591293334961,
+      "learning_rate": 9.082301529497451e-05,
+      "loss": 2.1268,
+      "step": 200
+    },
+    {
+      "epoch": 0.13902818606259726,
+      "grad_norm": 0.6248623728752136,
+      "learning_rate": 9.075018208302987e-05,
+      "loss": 1.2631,
+      "step": 201
+    },
+    {
+      "epoch": 0.13971986858032162,
+      "grad_norm": 0.5625057220458984,
+      "learning_rate": 9.067734887108522e-05,
+      "loss": 1.1348,
+      "step": 202
+    },
+    {
+      "epoch": 0.14041155109804598,
+      "grad_norm": 0.5977293252944946,
+      "learning_rate": 9.060451565914058e-05,
+      "loss": 1.136,
+      "step": 203
+    },
+    {
+      "epoch": 0.14110323361577037,
+      "grad_norm": 0.6973084211349487,
+      "learning_rate": 9.053168244719593e-05,
+      "loss": 2.0764,
+      "step": 204
+    },
+    {
+      "epoch": 0.14179491613349474,
+      "grad_norm": 0.761034369468689,
+      "learning_rate": 9.045884923525127e-05,
+      "loss": 2.4017,
+      "step": 205
+    },
+    {
+      "epoch": 0.1424865986512191,
+      "grad_norm": 0.7098332047462463,
+      "learning_rate": 9.038601602330664e-05,
+      "loss": 1.6705,
+      "step": 206
+    },
+    {
+      "epoch": 0.14317828116894346,
+      "grad_norm": 0.6458803415298462,
+      "learning_rate": 9.031318281136198e-05,
+      "loss": 1.2114,
+      "step": 207
+    },
+    {
+      "epoch": 0.14386996368666782,
+      "grad_norm": 0.6271463632583618,
+      "learning_rate": 9.024034959941734e-05,
+      "loss": 1.6756,
+      "step": 208
+    },
+    {
+      "epoch": 0.14456164620439219,
+      "grad_norm": 0.6152657866477966,
+      "learning_rate": 9.016751638747269e-05,
+      "loss": 1.5943,
+      "step": 209
+    },
+    {
+      "epoch": 0.14525332872211655,
+      "grad_norm": 0.6465914845466614,
+      "learning_rate": 9.009468317552805e-05,
+      "loss": 1.8369,
+      "step": 210
+    },
+    {
+      "epoch": 0.1459450112398409,
+      "grad_norm": 2.220533847808838,
+      "learning_rate": 9.00218499635834e-05,
+      "loss": 2.7842,
+      "step": 211
+    },
+    {
+      "epoch": 0.14663669375756527,
+      "grad_norm": 0.562271773815155,
+      "learning_rate": 8.994901675163875e-05,
+      "loss": 1.9363,
+      "step": 212
+    },
+    {
+      "epoch": 0.14732837627528964,
+      "grad_norm": 0.646512508392334,
+      "learning_rate": 8.987618353969411e-05,
+      "loss": 1.9429,
+      "step": 213
+    },
+    {
+      "epoch": 0.148020058793014,
+      "grad_norm": 0.8589336276054382,
+      "learning_rate": 8.980335032774946e-05,
+      "loss": 1.6978,
+      "step": 214
+    },
+    {
+      "epoch": 0.14871174131073836,
+      "grad_norm": 0.7746891975402832,
+      "learning_rate": 8.973051711580482e-05,
+      "loss": 1.1804,
+      "step": 215
+    },
+    {
+      "epoch": 0.14940342382846272,
+      "grad_norm": 0.809123158454895,
+      "learning_rate": 8.965768390386016e-05,
+      "loss": 1.5176,
+      "step": 216
+    },
+    {
+      "epoch": 0.1500951063461871,
+      "grad_norm": 0.6705849170684814,
+      "learning_rate": 8.958485069191553e-05,
+      "loss": 2.1671,
+      "step": 217
+    },
+    {
+      "epoch": 0.15078678886391148,
+      "grad_norm": 0.5576826333999634,
+      "learning_rate": 8.951201747997087e-05,
+      "loss": 1.2096,
+      "step": 218
+    },
+    {
+      "epoch": 0.15147847138163584,
+      "grad_norm": 0.8005768060684204,
+      "learning_rate": 8.943918426802622e-05,
+      "loss": 2.3528,
+      "step": 219
+    },
+    {
+      "epoch": 0.1521701538993602,
+      "grad_norm": 0.7752550840377808,
+      "learning_rate": 8.936635105608158e-05,
+      "loss": 1.7417,
+      "step": 220
+    },
+    {
+      "epoch": 0.15286183641708456,
+      "grad_norm": 0.819672703742981,
+      "learning_rate": 8.929351784413693e-05,
+      "loss": 2.0707,
+      "step": 221
+    },
+    {
+      "epoch": 0.15355351893480892,
+      "grad_norm": 1.488770842552185,
+      "learning_rate": 8.922068463219229e-05,
+      "loss": 1.7984,
+      "step": 222
+    },
+    {
+      "epoch": 0.1542452014525333,
+      "grad_norm": 0.7640134692192078,
+      "learning_rate": 8.914785142024764e-05,
+      "loss": 1.858,
+      "step": 223
+    },
+    {
+      "epoch": 0.15493688397025765,
+      "grad_norm": 0.8346788883209229,
+      "learning_rate": 8.907501820830299e-05,
+      "loss": 1.7544,
+      "step": 224
+    },
+    {
+      "epoch": 0.155628566487982,
+      "grad_norm": 0.5825446844100952,
+      "learning_rate": 8.900218499635835e-05,
+      "loss": 1.8188,
+      "step": 225
+    },
+    {
+      "epoch": 0.15632024900570637,
+      "grad_norm": 0.7010117173194885,
+      "learning_rate": 8.89293517844137e-05,
+      "loss": 1.9475,
+      "step": 226
+    },
+    {
+      "epoch": 0.15701193152343074,
+      "grad_norm": 0.5123314261436462,
+      "learning_rate": 8.885651857246906e-05,
+      "loss": 2.1235,
+      "step": 227
+    },
+    {
+      "epoch": 0.1577036140411551,
+      "grad_norm": 0.62752366065979,
+      "learning_rate": 8.87836853605244e-05,
+      "loss": 1.872,
+      "step": 228
+    },
+    {
+      "epoch": 0.15839529655887946,
+      "grad_norm": 0.6518980264663696,
+      "learning_rate": 8.871085214857976e-05,
+      "loss": 1.4556,
+      "step": 229
+    },
+    {
+      "epoch": 0.15908697907660385,
+      "grad_norm": 0.7556332349777222,
+      "learning_rate": 8.863801893663511e-05,
+      "loss": 2.0122,
+      "step": 230
+    },
+    {
+      "epoch": 0.15977866159432821,
+      "grad_norm": 0.6503921747207642,
+      "learning_rate": 8.856518572469046e-05,
+      "loss": 2.0475,
+      "step": 231
+    },
+    {
+      "epoch": 0.16047034411205258,
+      "grad_norm": 0.7834317088127136,
+      "learning_rate": 8.849235251274582e-05,
+      "loss": 1.8671,
+      "step": 232
+    },
+    {
+      "epoch": 0.16116202662977694,
+      "grad_norm": 2.123568534851074,
+      "learning_rate": 8.841951930080117e-05,
+      "loss": 1.4212,
+      "step": 233
+    },
+    {
+      "epoch": 0.1618537091475013,
+      "grad_norm": 0.5151258111000061,
+      "learning_rate": 8.834668608885653e-05,
+      "loss": 1.1815,
+      "step": 234
+    },
+    {
+      "epoch": 0.16254539166522566,
+      "grad_norm": 0.6445286273956299,
+      "learning_rate": 8.827385287691188e-05,
+      "loss": 2.1009,
+      "step": 235
+    },
+    {
+      "epoch": 0.16323707418295003,
+      "grad_norm": 0.9389359951019287,
+      "learning_rate": 8.820101966496724e-05,
+      "loss": 2.3638,
+      "step": 236
+    },
+    {
+      "epoch": 0.1639287567006744,
+      "grad_norm": 0.7288535833358765,
+      "learning_rate": 8.812818645302258e-05,
+      "loss": 1.5602,
+      "step": 237
+    },
+    {
+      "epoch": 0.16462043921839875,
+      "grad_norm": 0.5802240967750549,
+      "learning_rate": 8.805535324107793e-05,
+      "loss": 1.8588,
+      "step": 238
+    },
+    {
+      "epoch": 0.1653121217361231,
+      "grad_norm": 0.6751568913459778,
+      "learning_rate": 8.798252002913329e-05,
+      "loss": 2.0206,
+      "step": 239
+    },
+    {
+      "epoch": 0.16600380425384748,
+      "grad_norm": 0.6931467056274414,
+      "learning_rate": 8.790968681718864e-05,
+      "loss": 1.9663,
+      "step": 240
+    },
+    {
+      "epoch": 0.16669548677157184,
+      "grad_norm": 0.6931905150413513,
+      "learning_rate": 8.7836853605244e-05,
+      "loss": 1.5422,
+      "step": 241
+    },
+    {
+      "epoch": 0.1673871692892962,
+      "grad_norm": 0.7654033899307251,
+      "learning_rate": 8.776402039329935e-05,
+      "loss": 2.2165,
+      "step": 242
+    },
+    {
+      "epoch": 0.1680788518070206,
+      "grad_norm": 0.8461145162582397,
+      "learning_rate": 8.769118718135471e-05,
+      "loss": 2.1441,
+      "step": 243
+    },
+    {
+      "epoch": 0.16877053432474495,
+      "grad_norm": 0.6745240092277527,
+      "learning_rate": 8.761835396941004e-05,
+      "loss": 2.0226,
+      "step": 244
+    },
+    {
+      "epoch": 0.16946221684246932,
+      "grad_norm": 0.9059496521949768,
+      "learning_rate": 8.75455207574654e-05,
+      "loss": 1.3178,
+      "step": 245
+    },
+    {
+      "epoch": 0.17015389936019368,
+      "grad_norm": 0.7393327951431274,
+      "learning_rate": 8.747268754552075e-05,
+      "loss": 1.7787,
+      "step": 246
+    },
+    {
+      "epoch": 0.17084558187791804,
+      "grad_norm": 1.1192548274993896,
+      "learning_rate": 8.739985433357611e-05,
+      "loss": 1.2677,
+      "step": 247
+    },
+    {
+      "epoch": 0.1715372643956424,
+      "grad_norm": 0.500165581703186,
+      "learning_rate": 8.732702112163147e-05,
+      "loss": 1.6749,
+      "step": 248
+    },
+    {
+      "epoch": 0.17222894691336676,
+      "grad_norm": 0.7118367552757263,
+      "learning_rate": 8.725418790968682e-05,
+      "loss": 1.6768,
+      "step": 249
+    },
+    {
+      "epoch": 0.17292062943109113,
+      "grad_norm": 0.6697540283203125,
+      "learning_rate": 8.718135469774218e-05,
+      "loss": 1.7236,
+      "step": 250
+    },
+    {
+      "epoch": 0.1736123119488155,
+      "grad_norm": 0.621541440486908,
+      "learning_rate": 8.710852148579752e-05,
+      "loss": 1.1761,
+      "step": 251
+    },
+    {
+      "epoch": 0.17430399446653985,
+      "grad_norm": 1.05433189868927,
+      "learning_rate": 8.703568827385288e-05,
+      "loss": 1.5783,
+      "step": 252
+    },
+    {
+      "epoch": 0.17499567698426421,
+      "grad_norm": 0.7111071944236755,
+      "learning_rate": 8.696285506190823e-05,
+      "loss": 1.9291,
+      "step": 253
+    },
+    {
+      "epoch": 0.17568735950198858,
+      "grad_norm": 0.7538565397262573,
+      "learning_rate": 8.689002184996359e-05,
+      "loss": 1.5598,
+      "step": 254
+    },
+    {
+      "epoch": 0.17637904201971294,
+      "grad_norm": 0.7115334868431091,
+      "learning_rate": 8.681718863801895e-05,
+      "loss": 1.7721,
+      "step": 255
+    },
+    {
+      "epoch": 0.17707072453743733,
+      "grad_norm": 0.8842628002166748,
+      "learning_rate": 8.67443554260743e-05,
+      "loss": 1.8147,
+      "step": 256
+    },
+    {
+      "epoch": 0.1777624070551617,
+      "grad_norm": 0.6481782793998718,
+      "learning_rate": 8.667152221412966e-05,
+      "loss": 1.9606,
+      "step": 257
+    },
+    {
+      "epoch": 0.17845408957288605,
+      "grad_norm": 0.719637393951416,
+      "learning_rate": 8.659868900218499e-05,
+      "loss": 1.2452,
+      "step": 258
+    },
+    {
+      "epoch": 0.17914577209061042,
+      "grad_norm": 0.6171206831932068,
+      "learning_rate": 8.652585579024035e-05,
+      "loss": 1.367,
+      "step": 259
+    },
+    {
+      "epoch": 0.17983745460833478,
+      "grad_norm": 0.618354320526123,
+      "learning_rate": 8.64530225782957e-05,
+      "loss": 1.9576,
+      "step": 260
+    },
+    {
+      "epoch": 0.18052913712605914,
+      "grad_norm": 0.6816132664680481,
+      "learning_rate": 8.638018936635106e-05,
+      "loss": 1.1272,
+      "step": 261
+    },
+    {
+      "epoch": 0.1812208196437835,
+      "grad_norm": 0.6745633482933044,
+      "learning_rate": 8.630735615440642e-05,
+      "loss": 2.1625,
+      "step": 262
+    },
+    {
+      "epoch": 0.18191250216150787,
+      "grad_norm": 0.8869469165802002,
+      "learning_rate": 8.623452294246177e-05,
+      "loss": 2.0639,
+      "step": 263
+    },
+    {
+      "epoch": 0.18260418467923223,
+      "grad_norm": 0.9394845366477966,
+      "learning_rate": 8.616168973051712e-05,
+      "loss": 1.6026,
+      "step": 264
+    },
+    {
+      "epoch": 0.1832958671969566,
+      "grad_norm": 0.6131375432014465,
+      "learning_rate": 8.608885651857246e-05,
+      "loss": 2.0913,
+      "step": 265
+    },
+    {
+      "epoch": 0.18398754971468095,
+      "grad_norm": 0.6758245229721069,
+      "learning_rate": 8.601602330662782e-05,
+      "loss": 1.2526,
+      "step": 266
+    },
+    {
+      "epoch": 0.18467923223240532,
+      "grad_norm": 0.7595621943473816,
+      "learning_rate": 8.594319009468317e-05,
+      "loss": 2.1043,
+      "step": 267
+    },
+    {
+      "epoch": 0.18537091475012968,
+      "grad_norm": 0.8051510453224182,
+      "learning_rate": 8.587035688273853e-05,
+      "loss": 1.5743,
+      "step": 268
+    },
+    {
+      "epoch": 0.18606259726785407,
+      "grad_norm": 0.6870526671409607,
+      "learning_rate": 8.57975236707939e-05,
+      "loss": 1.475,
+      "step": 269
+    },
+    {
+      "epoch": 0.18675427978557843,
+      "grad_norm": 0.6781574487686157,
+      "learning_rate": 8.572469045884924e-05,
+      "loss": 2.1719,
+      "step": 270
+    },
+    {
+      "epoch": 0.1874459623033028,
+      "grad_norm": 0.7067494988441467,
+      "learning_rate": 8.565185724690459e-05,
+      "loss": 2.3096,
+      "step": 271
+    },
+    {
+      "epoch": 0.18813764482102716,
+      "grad_norm": 0.8800467848777771,
+      "learning_rate": 8.557902403495994e-05,
+      "loss": 1.9574,
+      "step": 272
+    },
+    {
+      "epoch": 0.18882932733875152,
+      "grad_norm": 0.5349634885787964,
+      "learning_rate": 8.55061908230153e-05,
+      "loss": 1.4169,
+      "step": 273
+    },
+    {
+      "epoch": 0.18952100985647588,
+      "grad_norm": 0.6334154009819031,
+      "learning_rate": 8.543335761107065e-05,
+      "loss": 1.9563,
+      "step": 274
+    },
+    {
+      "epoch": 0.19021269237420024,
+      "grad_norm": 0.8906685709953308,
+      "learning_rate": 8.536052439912601e-05,
+      "loss": 2.1072,
+      "step": 275
+    },
+    {
+      "epoch": 0.1909043748919246,
+      "grad_norm": 0.7288912534713745,
+      "learning_rate": 8.528769118718137e-05,
+      "loss": 1.9452,
+      "step": 276
+    },
+    {
+      "epoch": 0.19159605740964897,
+      "grad_norm": 1.1331645250320435,
+      "learning_rate": 8.521485797523672e-05,
+      "loss": 1.9652,
+      "step": 277
+    },
+    {
+      "epoch": 0.19228773992737333,
+      "grad_norm": 0.6032379269599915,
+      "learning_rate": 8.514202476329206e-05,
+      "loss": 1.9178,
+      "step": 278
+    },
+    {
+      "epoch": 0.1929794224450977,
+      "grad_norm": 0.6424663662910461,
+      "learning_rate": 8.506919155134741e-05,
+      "loss": 1.9931,
+      "step": 279
+    },
+    {
+      "epoch": 0.19367110496282205,
+      "grad_norm": 0.5931838750839233,
+      "learning_rate": 8.499635833940277e-05,
+      "loss": 1.2334,
+      "step": 280
+    },
+    {
+      "epoch": 0.19436278748054642,
+      "grad_norm": 0.7451290488243103,
+      "learning_rate": 8.492352512745812e-05,
+      "loss": 1.9906,
+      "step": 281
+    },
+    {
+      "epoch": 0.1950544699982708,
+      "grad_norm": 0.7306643128395081,
+      "learning_rate": 8.485069191551348e-05,
+      "loss": 2.1793,
+      "step": 282
+    },
+    {
+      "epoch": 0.19574615251599517,
+      "grad_norm": 0.6340318918228149,
+      "learning_rate": 8.477785870356884e-05,
+      "loss": 1.3639,
+      "step": 283
+    },
+    {
+      "epoch": 0.19643783503371953,
+      "grad_norm": 0.9806922674179077,
+      "learning_rate": 8.470502549162418e-05,
+      "loss": 1.5363,
+      "step": 284
+    },
+    {
+      "epoch": 0.1971295175514439,
+      "grad_norm": 0.6847662925720215,
+      "learning_rate": 8.463219227967954e-05,
+      "loss": 1.4774,
+      "step": 285
+    },
+    {
+      "epoch": 0.19782120006916826,
+      "grad_norm": 0.5814263224601746,
+      "learning_rate": 8.455935906773488e-05,
+      "loss": 1.4903,
+      "step": 286
+    },
+    {
+      "epoch": 0.19851288258689262,
+      "grad_norm": 0.582711935043335,
+      "learning_rate": 8.448652585579024e-05,
+      "loss": 1.479,
+      "step": 287
+    },
+    {
+      "epoch": 0.19920456510461698,
+      "grad_norm": 0.8541072607040405,
+      "learning_rate": 8.441369264384559e-05,
+      "loss": 1.6464,
+      "step": 288
+    },
+    {
+      "epoch": 0.19989624762234134,
+      "grad_norm": 0.6419352889060974,
+      "learning_rate": 8.434085943190095e-05,
+      "loss": 1.9266,
+      "step": 289
+    },
+    {
+      "epoch": 0.2005879301400657,
+      "grad_norm": 0.694685161113739,
+      "learning_rate": 8.426802621995631e-05,
+      "loss": 1.5357,
+      "step": 290
+    },
+    {
+      "epoch": 0.20127961265779007,
+      "grad_norm": 0.7046255469322205,
+      "learning_rate": 8.419519300801165e-05,
+      "loss": 1.4103,
+      "step": 291
+    },
+    {
+      "epoch": 0.20197129517551443,
+      "grad_norm": 1.0252553224563599,
+      "learning_rate": 8.412235979606701e-05,
+      "loss": 1.8172,
+      "step": 292
+    },
+    {
+      "epoch": 0.2026629776932388,
+      "grad_norm": 0.7682600021362305,
+      "learning_rate": 8.404952658412236e-05,
+      "loss": 1.455,
+      "step": 293
+    },
+    {
+      "epoch": 0.20335466021096316,
+      "grad_norm": 0.9454272389411926,
+      "learning_rate": 8.397669337217772e-05,
+      "loss": 1.5882,
+      "step": 294
+    },
+    {
+      "epoch": 0.20404634272868752,
+      "grad_norm": 1.333929419517517,
+      "learning_rate": 8.390386016023307e-05,
+      "loss": 1.6718,
+      "step": 295
+    },
+    {
+      "epoch": 0.2047380252464119,
+      "grad_norm": 0.9410814046859741,
+      "learning_rate": 8.383102694828843e-05,
+      "loss": 2.0226,
+      "step": 296
+    },
+    {
+      "epoch": 0.20542970776413627,
+      "grad_norm": 0.7768377065658569,
+      "learning_rate": 8.375819373634379e-05,
+      "loss": 2.0137,
+      "step": 297
+    },
+    {
+      "epoch": 0.20612139028186063,
+      "grad_norm": 1.2727937698364258,
+      "learning_rate": 8.368536052439912e-05,
+      "loss": 1.9412,
+      "step": 298
+    },
+    {
+      "epoch": 0.206813072799585,
+      "grad_norm": 0.9015520811080933,
+      "learning_rate": 8.361252731245448e-05,
+      "loss": 1.6939,
+      "step": 299
+    },
+    {
+      "epoch": 0.20750475531730936,
+      "grad_norm": 0.5938074588775635,
+      "learning_rate": 8.353969410050983e-05,
+      "loss": 1.3246,
+      "step": 300
+    },
+    {
+      "epoch": 0.20819643783503372,
+      "grad_norm": 0.7636774182319641,
+      "learning_rate": 8.346686088856519e-05,
+      "loss": 1.687,
+      "step": 301
+    },
+    {
+      "epoch": 0.20888812035275808,
+      "grad_norm": 0.8219266533851624,
+      "learning_rate": 8.339402767662054e-05,
+      "loss": 1.9761,
+      "step": 302
+    },
+    {
+      "epoch": 0.20957980287048245,
+      "grad_norm": 0.8183760643005371,
+      "learning_rate": 8.33211944646759e-05,
+      "loss": 2.2887,
+      "step": 303
+    },
+    {
+      "epoch": 0.2102714853882068,
+      "grad_norm": 0.8675097227096558,
+      "learning_rate": 8.324836125273125e-05,
+      "loss": 2.0024,
+      "step": 304
+    },
+    {
+      "epoch": 0.21096316790593117,
+      "grad_norm": 0.5965743660926819,
+      "learning_rate": 8.31755280407866e-05,
+      "loss": 1.4569,
+      "step": 305
+    },
+    {
+      "epoch": 0.21165485042365553,
+      "grad_norm": 0.8379791378974915,
+      "learning_rate": 8.310269482884196e-05,
+      "loss": 1.9685,
+      "step": 306
+    },
+    {
+      "epoch": 0.2123465329413799,
+      "grad_norm": 0.5711637735366821,
+      "learning_rate": 8.30298616168973e-05,
+      "loss": 1.8149,
+      "step": 307
+    },
+    {
+      "epoch": 0.21303821545910426,
+      "grad_norm": 0.8288681507110596,
+      "learning_rate": 8.295702840495266e-05,
+      "loss": 2.0087,
+      "step": 308
+    },
+    {
+      "epoch": 0.21372989797682865,
+      "grad_norm": 0.7129418849945068,
+      "learning_rate": 8.288419519300801e-05,
+      "loss": 1.7503,
+      "step": 309
+    },
+    {
+      "epoch": 0.214421580494553,
+      "grad_norm": 0.7114739418029785,
+      "learning_rate": 8.281136198106337e-05,
+      "loss": 1.6055,
+      "step": 310
+    },
+    {
+      "epoch": 0.21511326301227737,
+      "grad_norm": 0.821076512336731,
+      "learning_rate": 8.273852876911872e-05,
+      "loss": 1.6553,
+      "step": 311
+    },
+    {
+      "epoch": 0.21580494553000173,
+      "grad_norm": 0.7488536238670349,
+      "learning_rate": 8.266569555717407e-05,
+      "loss": 1.9126,
+      "step": 312
+    },
+    {
+      "epoch": 0.2164966280477261,
+      "grad_norm": 0.822170615196228,
+      "learning_rate": 8.259286234522943e-05,
+      "loss": 2.0362,
+      "step": 313
+    },
+    {
+      "epoch": 0.21718831056545046,
+      "grad_norm": 0.7749188542366028,
+      "learning_rate": 8.252002913328478e-05,
+      "loss": 1.6924,
+      "step": 314
+    },
+    {
+      "epoch": 0.21787999308317482,
+      "grad_norm": 0.7008628845214844,
+      "learning_rate": 8.244719592134014e-05,
+      "loss": 2.1542,
+      "step": 315
+    },
+    {
+      "epoch": 0.21857167560089918,
+      "grad_norm": 0.8683955669403076,
+      "learning_rate": 8.237436270939549e-05,
+      "loss": 1.6697,
+      "step": 316
+    },
+    {
+      "epoch": 0.21926335811862355,
+      "grad_norm": 0.7530885338783264,
+      "learning_rate": 8.230152949745085e-05,
+      "loss": 1.2485,
+      "step": 317
+    },
+    {
+      "epoch": 0.2199550406363479,
+      "grad_norm": 0.787668764591217,
+      "learning_rate": 8.22286962855062e-05,
+      "loss": 1.1665,
+      "step": 318
+    },
+    {
+      "epoch": 0.22064672315407227,
+      "grad_norm": 0.7304548025131226,
+      "learning_rate": 8.215586307356154e-05,
+      "loss": 2.0012,
+      "step": 319
+    },
+    {
+      "epoch": 0.22133840567179663,
+      "grad_norm": 0.6933902502059937,
+      "learning_rate": 8.20830298616169e-05,
+      "loss": 1.8554,
+      "step": 320
+    },
+    {
+      "epoch": 0.222030088189521,
+      "grad_norm": 1.5794916152954102,
+      "learning_rate": 8.201019664967225e-05,
+      "loss": 1.6497,
+      "step": 321
+    },
+    {
+      "epoch": 0.2227217707072454,
+      "grad_norm": 0.788654625415802,
+      "learning_rate": 8.193736343772761e-05,
+      "loss": 1.9427,
+      "step": 322
+    },
+    {
+      "epoch": 0.22341345322496975,
+      "grad_norm": 0.6310513019561768,
+      "learning_rate": 8.186453022578296e-05,
+      "loss": 1.1892,
+      "step": 323
+    },
+    {
+      "epoch": 0.2241051357426941,
+      "grad_norm": 0.7976911664009094,
+      "learning_rate": 8.17916970138383e-05,
+      "loss": 1.8037,
+      "step": 324
+    },
+    {
+      "epoch": 0.22479681826041847,
+      "grad_norm": 0.8486120104789734,
+      "learning_rate": 8.171886380189367e-05,
+      "loss": 0.8769,
+      "step": 325
+    },
+    {
+      "epoch": 0.22548850077814284,
+      "grad_norm": 0.6910994648933411,
+      "learning_rate": 8.164603058994901e-05,
+      "loss": 1.2035,
+      "step": 326
+    },
+    {
+      "epoch": 0.2261801832958672,
+      "grad_norm": 0.7402215003967285,
+      "learning_rate": 8.157319737800438e-05,
+      "loss": 1.8359,
+      "step": 327
+    },
+    {
+      "epoch": 0.22687186581359156,
+      "grad_norm": 0.8741387724876404,
+      "learning_rate": 8.150036416605972e-05,
+      "loss": 2.1804,
+      "step": 328
+    },
+    {
+      "epoch": 0.22756354833131592,
+      "grad_norm": 0.7987048625946045,
+      "learning_rate": 8.142753095411508e-05,
+      "loss": 1.8705,
+      "step": 329
+    },
+    {
+      "epoch": 0.22825523084904029,
+      "grad_norm": 0.9527040123939514,
+      "learning_rate": 8.135469774217043e-05,
+      "loss": 1.9079,
+      "step": 330
+    },
+    {
+      "epoch": 0.22894691336676465,
+      "grad_norm": 0.7261909246444702,
+      "learning_rate": 8.128186453022578e-05,
+      "loss": 1.9903,
+      "step": 331
+    },
+    {
+      "epoch": 0.229638595884489,
+      "grad_norm": 0.7130205035209656,
+      "learning_rate": 8.120903131828114e-05,
+      "loss": 1.9952,
+      "step": 332
+    },
+    {
+      "epoch": 0.23033027840221337,
+      "grad_norm": 0.6477673649787903,
+      "learning_rate": 8.113619810633649e-05,
+      "loss": 1.4532,
+      "step": 333
+    },
+    {
+      "epoch": 0.23102196091993774,
+      "grad_norm": 0.7342897653579712,
+      "learning_rate": 8.106336489439185e-05,
+      "loss": 1.9345,
+      "step": 334
+    },
+    {
+      "epoch": 0.23171364343766213,
+      "grad_norm": 0.9584718346595764,
+      "learning_rate": 8.09905316824472e-05,
+      "loss": 2.1087,
+      "step": 335
+    },
+    {
+      "epoch": 0.2324053259553865,
+      "grad_norm": 0.7027872204780579,
+      "learning_rate": 8.091769847050256e-05,
+      "loss": 1.9372,
+      "step": 336
+    },
+    {
+      "epoch": 0.23309700847311085,
+      "grad_norm": 0.5855789184570312,
+      "learning_rate": 8.08448652585579e-05,
+      "loss": 1.7773,
+      "step": 337
+    },
+    {
+      "epoch": 0.2337886909908352,
+      "grad_norm": 0.6886003017425537,
+      "learning_rate": 8.077203204661325e-05,
+      "loss": 1.7705,
+      "step": 338
+    },
+    {
+      "epoch": 0.23448037350855958,
+      "grad_norm": 0.6997604966163635,
+      "learning_rate": 8.069919883466861e-05,
+      "loss": 1.3358,
+      "step": 339
+    },
+    {
+      "epoch": 0.23517205602628394,
+      "grad_norm": 0.6345038414001465,
+      "learning_rate": 8.062636562272396e-05,
+      "loss": 1.2833,
+      "step": 340
+    },
+    {
+      "epoch": 0.2358637385440083,
+      "grad_norm": 0.6772000789642334,
+      "learning_rate": 8.055353241077932e-05,
+      "loss": 1.5009,
+      "step": 341
+    },
+    {
+      "epoch": 0.23655542106173266,
+      "grad_norm": 0.7856854200363159,
+      "learning_rate": 8.048069919883467e-05,
+      "loss": 1.8073,
+      "step": 342
+    },
+    {
+      "epoch": 0.23724710357945702,
+      "grad_norm": 0.729583740234375,
+      "learning_rate": 8.040786598689003e-05,
+      "loss": 2.3742,
+      "step": 343
+    },
+    {
+      "epoch": 0.2379387860971814,
+      "grad_norm": 0.6015090346336365,
+      "learning_rate": 8.033503277494538e-05,
+      "loss": 1.0806,
+      "step": 344
+    },
+    {
+      "epoch": 0.23863046861490575,
+      "grad_norm": 0.6671364307403564,
+      "learning_rate": 8.026219956300073e-05,
+      "loss": 1.3656,
+      "step": 345
+    },
+    {
+      "epoch": 0.2393221511326301,
+      "grad_norm": 0.7358818054199219,
+      "learning_rate": 8.018936635105609e-05,
+      "loss": 1.769,
+      "step": 346
+    },
+    {
+      "epoch": 0.24001383365035447,
+      "grad_norm": 0.697853147983551,
+      "learning_rate": 8.011653313911143e-05,
+      "loss": 1.596,
+      "step": 347
+    },
+    {
+      "epoch": 0.24070551616807886,
+      "grad_norm": 0.6559710502624512,
+      "learning_rate": 8.00436999271668e-05,
+      "loss": 1.4413,
+      "step": 348
+    },
+    {
+      "epoch": 0.24139719868580323,
+      "grad_norm": 0.770010232925415,
+      "learning_rate": 7.997086671522214e-05,
+      "loss": 2.2503,
+      "step": 349
+    },
+    {
+      "epoch": 0.2420888812035276,
+      "grad_norm": 0.6468133926391602,
+      "learning_rate": 7.98980335032775e-05,
+      "loss": 2.0731,
+      "step": 350
+    },
+    {
+      "epoch": 0.24278056372125195,
+      "grad_norm": 0.8330799341201782,
+      "learning_rate": 7.982520029133285e-05,
+      "loss": 1.5362,
+      "step": 351
+    },
+    {
+      "epoch": 0.24347224623897631,
+      "grad_norm": 1.0678739547729492,
+      "learning_rate": 7.97523670793882e-05,
+      "loss": 1.814,
+      "step": 352
+    },
+    {
+      "epoch": 0.24416392875670068,
+      "grad_norm": 0.8693723678588867,
+      "learning_rate": 7.967953386744356e-05,
+      "loss": 2.3669,
+      "step": 353
+    },
+    {
+      "epoch": 0.24485561127442504,
+      "grad_norm": 0.7095674872398376,
+      "learning_rate": 7.960670065549891e-05,
+      "loss": 1.2215,
+      "step": 354
+    },
+    {
+      "epoch": 0.2455472937921494,
+      "grad_norm": 0.7264676094055176,
+      "learning_rate": 7.953386744355427e-05,
+      "loss": 1.7455,
+      "step": 355
+    },
+    {
+      "epoch": 0.24623897630987376,
+      "grad_norm": 0.7226927280426025,
+      "learning_rate": 7.946103423160962e-05,
+      "loss": 1.722,
+      "step": 356
+    },
+    {
+      "epoch": 0.24693065882759813,
+      "grad_norm": 0.8582272529602051,
+      "learning_rate": 7.938820101966498e-05,
+      "loss": 2.2391,
+      "step": 357
+    },
+    {
+      "epoch": 0.2476223413453225,
+      "grad_norm": 0.633518636226654,
+      "learning_rate": 7.931536780772032e-05,
+      "loss": 1.8529,
+      "step": 358
+    },
+    {
+      "epoch": 0.24831402386304685,
+      "grad_norm": 0.9155897498130798,
+      "learning_rate": 7.924253459577567e-05,
+      "loss": 1.5049,
+      "step": 359
+    },
+    {
+      "epoch": 0.2490057063807712,
+      "grad_norm": 0.8335781097412109,
+      "learning_rate": 7.916970138383103e-05,
+      "loss": 2.0309,
+      "step": 360
+    },
+    {
+      "epoch": 0.2496973888984956,
+      "grad_norm": 0.7219622731208801,
+      "learning_rate": 7.909686817188638e-05,
+      "loss": 1.1584,
+      "step": 361
+    },
+    {
+      "epoch": 0.25038907141621997,
+      "grad_norm": 0.8952975869178772,
+      "learning_rate": 7.902403495994174e-05,
+      "loss": 2.1489,
+      "step": 362
+    },
+    {
+      "epoch": 0.2510807539339443,
+      "grad_norm": 0.7409587502479553,
+      "learning_rate": 7.895120174799709e-05,
+      "loss": 2.0858,
+      "step": 363
+    },
+    {
+      "epoch": 0.2517724364516687,
+      "grad_norm": 0.5998899936676025,
+      "learning_rate": 7.887836853605244e-05,
+      "loss": 1.0058,
+      "step": 364
+    },
+    {
+      "epoch": 0.252464118969393,
+      "grad_norm": 0.7379968762397766,
+      "learning_rate": 7.88055353241078e-05,
+      "loss": 1.2178,
+      "step": 365
+    },
+    {
+      "epoch": 0.2531558014871174,
+      "grad_norm": 1.2425731420516968,
+      "learning_rate": 7.873270211216315e-05,
+      "loss": 1.5815,
+      "step": 366
+    },
+    {
+      "epoch": 0.25384748400484175,
+      "grad_norm": 0.7375080585479736,
+      "learning_rate": 7.86598689002185e-05,
+      "loss": 2.2113,
+      "step": 367
+    },
+    {
+      "epoch": 0.25453916652256614,
+      "grad_norm": 1.1793482303619385,
+      "learning_rate": 7.858703568827385e-05,
+      "loss": 1.7387,
+      "step": 368
+    },
+    {
+      "epoch": 0.25523084904029053,
+      "grad_norm": 1.397199273109436,
+      "learning_rate": 7.851420247632922e-05,
+      "loss": 1.288,
+      "step": 369
+    },
+    {
+      "epoch": 0.25592253155801487,
+      "grad_norm": 0.7143409252166748,
+      "learning_rate": 7.844136926438456e-05,
+      "loss": 1.6329,
+      "step": 370
+    },
+    {
+      "epoch": 0.25661421407573926,
+      "grad_norm": 0.6984686255455017,
+      "learning_rate": 7.836853605243991e-05,
+      "loss": 1.7279,
+      "step": 371
+    },
+    {
+      "epoch": 0.2573058965934636,
+      "grad_norm": 0.7117382287979126,
+      "learning_rate": 7.829570284049527e-05,
+      "loss": 1.8195,
+      "step": 372
+    },
+    {
+      "epoch": 0.257997579111188,
+      "grad_norm": 0.7039769887924194,
+      "learning_rate": 7.822286962855062e-05,
+      "loss": 1.8619,
+      "step": 373
+    },
+    {
+      "epoch": 0.2586892616289123,
+      "grad_norm": 1.0917625427246094,
+      "learning_rate": 7.815003641660598e-05,
+      "loss": 1.8889,
+      "step": 374
+    },
+    {
+      "epoch": 0.2593809441466367,
+      "grad_norm": 0.6653533577919006,
+      "learning_rate": 7.807720320466133e-05,
+      "loss": 1.8691,
+      "step": 375
+    },
+    {
+      "epoch": 0.26007262666436104,
+      "grad_norm": 0.7204722166061401,
+      "learning_rate": 7.800436999271669e-05,
+      "loss": 1.4544,
+      "step": 376
+    },
+    {
+      "epoch": 0.26076430918208543,
+      "grad_norm": 0.523810863494873,
+      "learning_rate": 7.793153678077204e-05,
+      "loss": 0.9303,
+      "step": 377
+    },
+    {
+      "epoch": 0.26145599169980976,
+      "grad_norm": 0.6838625073432922,
+      "learning_rate": 7.785870356882738e-05,
+      "loss": 1.6911,
+      "step": 378
+    },
+    {
+      "epoch": 0.26214767421753415,
+      "grad_norm": 0.7531947493553162,
+      "learning_rate": 7.778587035688274e-05,
+      "loss": 1.8134,
+      "step": 379
+    },
+    {
+      "epoch": 0.2628393567352585,
+      "grad_norm": 0.7090226411819458,
+      "learning_rate": 7.771303714493809e-05,
+      "loss": 2.1142,
+      "step": 380
+    },
+    {
+      "epoch": 0.2635310392529829,
+      "grad_norm": 0.7452750205993652,
+      "learning_rate": 7.764020393299345e-05,
+      "loss": 1.3778,
+      "step": 381
+    },
+    {
+      "epoch": 0.26422272177070727,
+      "grad_norm": 0.709818422794342,
+      "learning_rate": 7.75673707210488e-05,
+      "loss": 1.0387,
+      "step": 382
+    },
+    {
+      "epoch": 0.2649144042884316,
+      "grad_norm": 0.7260245680809021,
+      "learning_rate": 7.749453750910416e-05,
+      "loss": 1.5823,
+      "step": 383
+    },
+    {
+      "epoch": 0.265606086806156,
+      "grad_norm": 1.0123919248580933,
+      "learning_rate": 7.742170429715951e-05,
+      "loss": 1.7737,
+      "step": 384
+    },
+    {
+      "epoch": 0.26629776932388033,
+      "grad_norm": 0.92912757396698,
+      "learning_rate": 7.734887108521486e-05,
+      "loss": 1.0985,
+      "step": 385
+    },
+    {
+      "epoch": 0.2669894518416047,
+      "grad_norm": 0.7285180687904358,
+      "learning_rate": 7.727603787327022e-05,
+      "loss": 1.6727,
+      "step": 386
+    },
+    {
+      "epoch": 0.26768113435932905,
+      "grad_norm": 0.9794962406158447,
+      "learning_rate": 7.720320466132557e-05,
+      "loss": 2.2499,
+      "step": 387
+    },
+    {
+      "epoch": 0.26837281687705344,
+      "grad_norm": 0.8266851902008057,
+      "learning_rate": 7.713037144938093e-05,
+      "loss": 1.7979,
+      "step": 388
+    },
+    {
+      "epoch": 0.2690644993947778,
+      "grad_norm": 0.8111055493354797,
+      "learning_rate": 7.705753823743627e-05,
+      "loss": 1.6586,
+      "step": 389
+    },
+    {
+      "epoch": 0.26975618191250217,
+      "grad_norm": 0.76368248462677,
+      "learning_rate": 7.698470502549163e-05,
+      "loss": 1.6407,
+      "step": 390
+    },
+    {
+      "epoch": 0.2704478644302265,
+      "grad_norm": 0.7116588950157166,
+      "learning_rate": 7.691187181354698e-05,
+      "loss": 2.0222,
+      "step": 391
+    },
+    {
+      "epoch": 0.2711395469479509,
+      "grad_norm": 0.7411038875579834,
+      "learning_rate": 7.683903860160233e-05,
+      "loss": 1.9575,
+      "step": 392
+    },
+    {
+      "epoch": 0.27183122946567523,
+      "grad_norm": 0.6649416089057922,
+      "learning_rate": 7.676620538965769e-05,
+      "loss": 1.0947,
+      "step": 393
+    },
+    {
+      "epoch": 0.2725229119833996,
+      "grad_norm": 0.8183586597442627,
+      "learning_rate": 7.669337217771304e-05,
+      "loss": 1.5777,
+      "step": 394
+    },
+    {
+      "epoch": 0.273214594501124,
+      "grad_norm": 0.7400631904602051,
+      "learning_rate": 7.66205389657684e-05,
+      "loss": 1.9312,
+      "step": 395
+    },
+    {
+      "epoch": 0.27390627701884834,
+      "grad_norm": 0.6919555068016052,
+      "learning_rate": 7.654770575382375e-05,
+      "loss": 1.8406,
+      "step": 396
+    },
+    {
+      "epoch": 0.27459795953657273,
+      "grad_norm": 1.2769502401351929,
+      "learning_rate": 7.647487254187911e-05,
+      "loss": 2.2262,
+      "step": 397
+    },
+    {
+      "epoch": 0.27528964205429707,
+      "grad_norm": 0.9513121247291565,
+      "learning_rate": 7.640203932993446e-05,
+      "loss": 2.0865,
+      "step": 398
+    },
+    {
+      "epoch": 0.27598132457202146,
+      "grad_norm": 0.7086082100868225,
+      "learning_rate": 7.63292061179898e-05,
+      "loss": 1.8785,
+      "step": 399
+    },
+    {
+      "epoch": 0.2766730070897458,
+      "grad_norm": 0.7725284695625305,
+      "learning_rate": 7.625637290604516e-05,
+      "loss": 1.6341,
+      "step": 400
+    },
+    {
+      "epoch": 0.2773646896074702,
+      "grad_norm": 0.819273829460144,
+      "learning_rate": 7.618353969410051e-05,
+      "loss": 1.6705,
+      "step": 401
+    },
+    {
+      "epoch": 0.2780563721251945,
+      "grad_norm": 0.7302514910697937,
+      "learning_rate": 7.611070648215587e-05,
+      "loss": 2.2798,
+      "step": 402
+    },
+    {
+      "epoch": 0.2787480546429189,
+      "grad_norm": 0.7020559906959534,
+      "learning_rate": 7.603787327021122e-05,
+      "loss": 1.549,
+      "step": 403
+    },
+    {
+      "epoch": 0.27943973716064324,
+      "grad_norm": 0.6759452819824219,
+      "learning_rate": 7.596504005826657e-05,
+      "loss": 0.9358,
+      "step": 404
+    },
+    {
+      "epoch": 0.28013141967836763,
+      "grad_norm": 0.8225228190422058,
+      "learning_rate": 7.589220684632193e-05,
+      "loss": 0.764,
+      "step": 405
+    },
+    {
+      "epoch": 0.28082310219609197,
+      "grad_norm": 0.6742099523544312,
+      "learning_rate": 7.581937363437728e-05,
+      "loss": 1.9551,
+      "step": 406
+    },
+    {
+      "epoch": 0.28151478471381636,
+      "grad_norm": 0.6638647317886353,
+      "learning_rate": 7.574654042243264e-05,
+      "loss": 1.1788,
+      "step": 407
+    },
+    {
+      "epoch": 0.28220646723154075,
+      "grad_norm": 0.8848326802253723,
+      "learning_rate": 7.567370721048798e-05,
+      "loss": 1.2803,
+      "step": 408
+    },
+    {
+      "epoch": 0.2828981497492651,
+      "grad_norm": 0.8638424873352051,
+      "learning_rate": 7.560087399854335e-05,
+      "loss": 2.0904,
+      "step": 409
+    },
+    {
+      "epoch": 0.28358983226698947,
+      "grad_norm": 0.8653358221054077,
+      "learning_rate": 7.55280407865987e-05,
+      "loss": 1.5278,
+      "step": 410
+    },
+    {
+      "epoch": 0.2842815147847138,
+      "grad_norm": 0.9951801896095276,
+      "learning_rate": 7.545520757465404e-05,
+      "loss": 1.7283,
+      "step": 411
+    },
+    {
+      "epoch": 0.2849731973024382,
+      "grad_norm": 0.6709397435188293,
+      "learning_rate": 7.53823743627094e-05,
+      "loss": 1.7356,
+      "step": 412
+    },
+    {
+      "epoch": 0.28566487982016253,
+      "grad_norm": 0.6480190753936768,
+      "learning_rate": 7.530954115076475e-05,
+      "loss": 1.4348,
+      "step": 413
+    },
+    {
+      "epoch": 0.2863565623378869,
+      "grad_norm": 0.9461554288864136,
+      "learning_rate": 7.523670793882011e-05,
+      "loss": 1.7292,
+      "step": 414
+    },
+    {
+      "epoch": 0.28704824485561126,
+      "grad_norm": 0.632386326789856,
+      "learning_rate": 7.516387472687546e-05,
+      "loss": 1.7186,
+      "step": 415
+    },
+    {
+      "epoch": 0.28773992737333565,
+      "grad_norm": 0.702663004398346,
+      "learning_rate": 7.509104151493082e-05,
+      "loss": 1.9939,
+      "step": 416
+    },
+    {
+      "epoch": 0.28843160989106,
+      "grad_norm": 0.8615918755531311,
+      "learning_rate": 7.501820830298617e-05,
+      "loss": 1.195,
+      "step": 417
+    },
+    {
+      "epoch": 0.28912329240878437,
+      "grad_norm": 0.9229804277420044,
+      "learning_rate": 7.494537509104151e-05,
+      "loss": 1.155,
+      "step": 418
+    },
+    {
+      "epoch": 0.2898149749265087,
+      "grad_norm": 0.8323528170585632,
+      "learning_rate": 7.487254187909688e-05,
+      "loss": 1.9316,
+      "step": 419
+    },
+    {
+      "epoch": 0.2905066574442331,
+      "grad_norm": 1.3782216310501099,
+      "learning_rate": 7.479970866715222e-05,
+      "loss": 1.0946,
+      "step": 420
+    },
+    {
+      "epoch": 0.2911983399619575,
+      "grad_norm": 0.6939939856529236,
+      "learning_rate": 7.472687545520758e-05,
+      "loss": 0.9669,
+      "step": 421
+    },
+    {
+      "epoch": 0.2918900224796818,
+      "grad_norm": 0.8428712487220764,
+      "learning_rate": 7.465404224326293e-05,
+      "loss": 1.0747,
+      "step": 422
+    },
+    {
+      "epoch": 0.2925817049974062,
+      "grad_norm": 0.8275536298751831,
+      "learning_rate": 7.458120903131829e-05,
+      "loss": 1.5087,
+      "step": 423
+    },
+    {
+      "epoch": 0.29327338751513055,
+      "grad_norm": 0.6948786973953247,
+      "learning_rate": 7.450837581937363e-05,
+      "loss": 1.604,
+      "step": 424
+    },
+    {
+      "epoch": 0.29396507003285494,
+      "grad_norm": 0.807960569858551,
+      "learning_rate": 7.443554260742899e-05,
+      "loss": 2.1098,
+      "step": 425
+    },
+    {
+      "epoch": 0.29465675255057927,
+      "grad_norm": 0.7478271722793579,
+      "learning_rate": 7.436270939548435e-05,
+      "loss": 2.1171,
+      "step": 426
+    },
+    {
+      "epoch": 0.29534843506830366,
+      "grad_norm": 0.6956362128257751,
+      "learning_rate": 7.42898761835397e-05,
+      "loss": 1.8568,
+      "step": 427
+    },
+    {
+      "epoch": 0.296040117586028,
+      "grad_norm": 1.215596079826355,
+      "learning_rate": 7.421704297159506e-05,
+      "loss": 1.6761,
+      "step": 428
+    },
+    {
+      "epoch": 0.2967318001037524,
+      "grad_norm": 0.623933732509613,
+      "learning_rate": 7.41442097596504e-05,
+      "loss": 1.807,
+      "step": 429
+    },
+    {
+      "epoch": 0.2974234826214767,
+      "grad_norm": 0.9105553030967712,
+      "learning_rate": 7.407137654770577e-05,
+      "loss": 1.8098,
+      "step": 430
+    },
+    {
+      "epoch": 0.2981151651392011,
+      "grad_norm": 0.7315935492515564,
+      "learning_rate": 7.39985433357611e-05,
+      "loss": 1.6137,
+      "step": 431
+    },
+    {
+      "epoch": 0.29880684765692545,
+      "grad_norm": 0.6291669011116028,
+      "learning_rate": 7.392571012381646e-05,
+      "loss": 1.3534,
+      "step": 432
+    },
+    {
+      "epoch": 0.29949853017464984,
+      "grad_norm": 0.8852145671844482,
+      "learning_rate": 7.385287691187182e-05,
+      "loss": 1.8691,
+      "step": 433
+    },
+    {
+      "epoch": 0.3001902126923742,
+      "grad_norm": 0.8524838089942932,
+      "learning_rate": 7.378004369992717e-05,
+      "loss": 1.8757,
+      "step": 434
+    },
+    {
+      "epoch": 0.30088189521009856,
+      "grad_norm": 0.6716812252998352,
+      "learning_rate": 7.370721048798253e-05,
+      "loss": 1.6939,
+      "step": 435
+    },
+    {
+      "epoch": 0.30157357772782295,
+      "grad_norm": 0.9623708128929138,
+      "learning_rate": 7.363437727603788e-05,
+      "loss": 1.969,
+      "step": 436
+    },
+    {
+      "epoch": 0.3022652602455473,
+      "grad_norm": 0.8178302049636841,
+      "learning_rate": 7.356154406409324e-05,
+      "loss": 1.9667,
+      "step": 437
+    },
+    {
+      "epoch": 0.3029569427632717,
+      "grad_norm": 0.8692168593406677,
+      "learning_rate": 7.348871085214857e-05,
+      "loss": 1.8676,
+      "step": 438
+    },
+    {
+      "epoch": 0.303648625280996,
+      "grad_norm": 0.7447508573532104,
+      "learning_rate": 7.341587764020393e-05,
+      "loss": 1.1265,
+      "step": 439
+    },
+    {
+      "epoch": 0.3043403077987204,
+      "grad_norm": 1.3428897857666016,
+      "learning_rate": 7.33430444282593e-05,
+      "loss": 1.7997,
+      "step": 440
+    },
+    {
+      "epoch": 0.30503199031644473,
+      "grad_norm": 0.813104510307312,
+      "learning_rate": 7.327021121631464e-05,
+      "loss": 2.0517,
+      "step": 441
+    },
+    {
+      "epoch": 0.3057236728341691,
+      "grad_norm": 0.6551740169525146,
+      "learning_rate": 7.319737800437e-05,
+      "loss": 1.533,
+      "step": 442
+    },
+    {
+      "epoch": 0.30641535535189346,
+      "grad_norm": 0.7581228613853455,
+      "learning_rate": 7.312454479242535e-05,
+      "loss": 1.0951,
+      "step": 443
+    },
+    {
+      "epoch": 0.30710703786961785,
+      "grad_norm": 0.8519687652587891,
+      "learning_rate": 7.30517115804807e-05,
+      "loss": 1.8467,
+      "step": 444
+    },
+    {
+      "epoch": 0.3077987203873422,
+      "grad_norm": 0.5666621923446655,
+      "learning_rate": 7.297887836853605e-05,
+      "loss": 1.2154,
+      "step": 445
+    },
+    {
+      "epoch": 0.3084904029050666,
+      "grad_norm": 0.9918830990791321,
+      "learning_rate": 7.290604515659141e-05,
+      "loss": 1.3048,
+      "step": 446
+    },
+    {
+      "epoch": 0.30918208542279096,
+      "grad_norm": 0.6438243985176086,
+      "learning_rate": 7.283321194464677e-05,
+      "loss": 1.5674,
+      "step": 447
+    },
+    {
+      "epoch": 0.3098737679405153,
+      "grad_norm": 0.9330650568008423,
+      "learning_rate": 7.276037873270212e-05,
+      "loss": 1.4974,
+      "step": 448
+    },
+    {
+      "epoch": 0.3105654504582397,
+      "grad_norm": 0.6821238398551941,
+      "learning_rate": 7.268754552075748e-05,
+      "loss": 1.651,
+      "step": 449
+    },
+    {
+      "epoch": 0.311257132975964,
+      "grad_norm": 0.7768619656562805,
+      "learning_rate": 7.261471230881282e-05,
+      "loss": 2.1974,
+      "step": 450
+    },
+    {
+      "epoch": 0.3119488154936884,
+      "grad_norm": 0.8299420475959778,
+      "learning_rate": 7.254187909686817e-05,
+      "loss": 2.0038,
+      "step": 451
+    },
+    {
+      "epoch": 0.31264049801141275,
+      "grad_norm": 0.7438697814941406,
+      "learning_rate": 7.246904588492352e-05,
+      "loss": 1.4635,
+      "step": 452
+    },
+    {
+      "epoch": 0.31333218052913714,
+      "grad_norm": 1.1031484603881836,
+      "learning_rate": 7.239621267297888e-05,
+      "loss": 1.6887,
+      "step": 453
+    },
+    {
+      "epoch": 0.3140238630468615,
+      "grad_norm": 0.7802501916885376,
+      "learning_rate": 7.232337946103424e-05,
+      "loss": 1.9786,
+      "step": 454
+    },
+    {
+      "epoch": 0.31471554556458586,
+      "grad_norm": 0.6793453693389893,
+      "learning_rate": 7.225054624908959e-05,
+      "loss": 1.7345,
+      "step": 455
+    },
+    {
+      "epoch": 0.3154072280823102,
+      "grad_norm": 0.9189644455909729,
+      "learning_rate": 7.217771303714495e-05,
+      "loss": 1.7558,
+      "step": 456
+    },
+    {
+      "epoch": 0.3160989106000346,
+      "grad_norm": 0.8324854373931885,
+      "learning_rate": 7.21048798252003e-05,
+      "loss": 1.5798,
+      "step": 457
+    },
+    {
+      "epoch": 0.3167905931177589,
+      "grad_norm": 0.8758043050765991,
+      "learning_rate": 7.203204661325565e-05,
+      "loss": 1.2295,
+      "step": 458
+    },
+    {
+      "epoch": 0.3174822756354833,
+      "grad_norm": 0.9270045757293701,
+      "learning_rate": 7.195921340131099e-05,
+      "loss": 1.7192,
+      "step": 459
+    },
+    {
+      "epoch": 0.3181739581532077,
+      "grad_norm": 0.7827816605567932,
+      "learning_rate": 7.188638018936635e-05,
+      "loss": 1.2111,
+      "step": 460
+    },
+    {
+      "epoch": 0.31886564067093204,
+      "grad_norm": 0.8331306576728821,
+      "learning_rate": 7.181354697742171e-05,
+      "loss": 1.177,
+      "step": 461
+    },
+    {
+      "epoch": 0.31955732318865643,
+      "grad_norm": 0.8764622807502747,
+      "learning_rate": 7.174071376547706e-05,
+      "loss": 1.9659,
+      "step": 462
+    },
+    {
+      "epoch": 0.32024900570638076,
+      "grad_norm": 1.100085735321045,
+      "learning_rate": 7.166788055353242e-05,
+      "loss": 1.1225,
+      "step": 463
+    },
+    {
+      "epoch": 0.32094068822410515,
+      "grad_norm": 0.7537538409233093,
+      "learning_rate": 7.159504734158776e-05,
+      "loss": 1.9381,
+      "step": 464
+    },
+    {
+      "epoch": 0.3216323707418295,
+      "grad_norm": 0.7370760440826416,
+      "learning_rate": 7.152221412964312e-05,
+      "loss": 2.0892,
+      "step": 465
+    },
+    {
+      "epoch": 0.3223240532595539,
+      "grad_norm": 0.6847136616706848,
+      "learning_rate": 7.144938091769847e-05,
+      "loss": 2.1787,
+      "step": 466
+    },
+    {
+      "epoch": 0.3230157357772782,
+      "grad_norm": 0.644542396068573,
+      "learning_rate": 7.137654770575383e-05,
+      "loss": 1.0054,
+      "step": 467
+    },
+    {
+      "epoch": 0.3237074182950026,
+      "grad_norm": 0.7460236549377441,
+      "learning_rate": 7.130371449380919e-05,
+      "loss": 1.3674,
+      "step": 468
+    },
+    {
+      "epoch": 0.32439910081272694,
+      "grad_norm": 0.7521992325782776,
+      "learning_rate": 7.123088128186454e-05,
+      "loss": 1.5823,
+      "step": 469
+    },
+    {
+      "epoch": 0.3250907833304513,
+      "grad_norm": 0.6718364357948303,
+      "learning_rate": 7.11580480699199e-05,
+      "loss": 1.7935,
+      "step": 470
+    },
+    {
+      "epoch": 0.32578246584817566,
+      "grad_norm": 0.888697624206543,
+      "learning_rate": 7.108521485797523e-05,
+      "loss": 0.3041,
+      "step": 471
+    },
+    {
+      "epoch": 0.32647414836590005,
+      "grad_norm": 1.017770767211914,
+      "learning_rate": 7.101238164603059e-05,
+      "loss": 2.0955,
+      "step": 472
+    },
+    {
+      "epoch": 0.32716583088362444,
+      "grad_norm": 0.7191439867019653,
+      "learning_rate": 7.093954843408594e-05,
+      "loss": 2.1291,
+      "step": 473
+    },
+    {
+      "epoch": 0.3278575134013488,
+      "grad_norm": 0.6852924227714539,
+      "learning_rate": 7.08667152221413e-05,
+      "loss": 1.9207,
+      "step": 474
+    },
+    {
+      "epoch": 0.32854919591907317,
+      "grad_norm": 0.8421714901924133,
+      "learning_rate": 7.079388201019666e-05,
+      "loss": 1.9089,
+      "step": 475
+    },
+    {
+      "epoch": 0.3292408784367975,
+      "grad_norm": 0.8098534941673279,
+      "learning_rate": 7.072104879825201e-05,
+      "loss": 1.9783,
+      "step": 476
+    },
+    {
+      "epoch": 0.3299325609545219,
+      "grad_norm": 0.8071694374084473,
+      "learning_rate": 7.064821558630737e-05,
+      "loss": 2.2857,
+      "step": 477
+    },
+    {
+      "epoch": 0.3306242434722462,
+      "grad_norm": 0.7348088622093201,
+      "learning_rate": 7.05753823743627e-05,
+      "loss": 1.8574,
+      "step": 478
+    },
+    {
+      "epoch": 0.3313159259899706,
+      "grad_norm": 0.7593449950218201,
+      "learning_rate": 7.050254916241806e-05,
+      "loss": 1.8705,
+      "step": 479
+    },
+    {
+      "epoch": 0.33200760850769495,
+      "grad_norm": 0.9493029117584229,
+      "learning_rate": 7.042971595047341e-05,
+      "loss": 1.3423,
+      "step": 480
+    },
+    {
+      "epoch": 0.33269929102541934,
+      "grad_norm": 0.7410417199134827,
+      "learning_rate": 7.035688273852877e-05,
+      "loss": 1.5451,
+      "step": 481
+    },
+    {
+      "epoch": 0.3333909735431437,
+      "grad_norm": 0.7440536618232727,
+      "learning_rate": 7.028404952658413e-05,
+      "loss": 1.1216,
+      "step": 482
+    },
+    {
+      "epoch": 0.33408265606086807,
+      "grad_norm": 1.203395128250122,
+      "learning_rate": 7.021121631463948e-05,
+      "loss": 1.5371,
+      "step": 483
+    },
+    {
+      "epoch": 0.3347743385785924,
+      "grad_norm": 0.8062857985496521,
+      "learning_rate": 7.013838310269483e-05,
+      "loss": 1.9374,
+      "step": 484
+    },
+    {
+      "epoch": 0.3354660210963168,
+      "grad_norm": 1.429983139038086,
+      "learning_rate": 7.006554989075018e-05,
+      "loss": 2.1104,
+      "step": 485
+    },
+    {
+      "epoch": 0.3361577036140412,
+      "grad_norm": 1.058760404586792,
+      "learning_rate": 6.999271667880554e-05,
+      "loss": 1.6387,
+      "step": 486
+    },
+    {
+      "epoch": 0.3368493861317655,
+      "grad_norm": 0.8884686231613159,
+      "learning_rate": 6.991988346686089e-05,
+      "loss": 2.1725,
+      "step": 487
+    },
+    {
+      "epoch": 0.3375410686494899,
+      "grad_norm": 0.7855187058448792,
+      "learning_rate": 6.984705025491625e-05,
+      "loss": 1.4707,
+      "step": 488
+    },
+    {
+      "epoch": 0.33823275116721424,
+      "grad_norm": 0.8604965209960938,
+      "learning_rate": 6.977421704297161e-05,
+      "loss": 2.0402,
+      "step": 489
+    },
+    {
+      "epoch": 0.33892443368493863,
+      "grad_norm": 1.5640790462493896,
+      "learning_rate": 6.970138383102696e-05,
+      "loss": 1.1966,
+      "step": 490
+    },
+    {
+      "epoch": 0.33961611620266297,
+      "grad_norm": 0.8234649896621704,
+      "learning_rate": 6.96285506190823e-05,
+      "loss": 1.4036,
+      "step": 491
+    },
+    {
+      "epoch": 0.34030779872038736,
+      "grad_norm": 0.6905968189239502,
+      "learning_rate": 6.955571740713765e-05,
+      "loss": 1.1985,
+      "step": 492
+    },
+    {
+      "epoch": 0.3409994812381117,
+      "grad_norm": 0.6831912398338318,
+      "learning_rate": 6.948288419519301e-05,
+      "loss": 1.941,
+      "step": 493
+    },
+    {
+      "epoch": 0.3416911637558361,
+      "grad_norm": 1.2796226739883423,
+      "learning_rate": 6.941005098324836e-05,
+      "loss": 1.7139,
+      "step": 494
+    },
+    {
+      "epoch": 0.3423828462735604,
+      "grad_norm": 0.7864359021186829,
+      "learning_rate": 6.933721777130372e-05,
+      "loss": 1.6839,
+      "step": 495
+    },
+    {
+      "epoch": 0.3430745287912848,
+      "grad_norm": 0.7287671566009521,
+      "learning_rate": 6.926438455935908e-05,
+      "loss": 1.8232,
+      "step": 496
+    },
+    {
+      "epoch": 0.34376621130900914,
+      "grad_norm": 0.8624921441078186,
+      "learning_rate": 6.919155134741443e-05,
+      "loss": 1.993,
+      "step": 497
+    },
+    {
+      "epoch": 0.34445789382673353,
+      "grad_norm": 0.7365380525588989,
+      "learning_rate": 6.911871813546978e-05,
+      "loss": 1.6436,
+      "step": 498
+    },
+    {
+      "epoch": 0.3451495763444579,
+      "grad_norm": 0.7367970943450928,
+      "learning_rate": 6.904588492352512e-05,
+      "loss": 2.151,
+      "step": 499
+    },
+    {
+      "epoch": 0.34584125886218225,
+      "grad_norm": 0.7151849865913391,
+      "learning_rate": 6.897305171158048e-05,
+      "loss": 1.1134,
+      "step": 500
+    },
+    {
+      "epoch": 0.34584125886218225,
+      "eval_loss": 1.5177690982818604,
+      "eval_runtime": 853.3383,
+      "eval_samples_per_second": 1.506,
+      "eval_steps_per_second": 0.754,
+      "step": 500
+    },
+    {
+      "epoch": 0.34653294137990664,
+      "grad_norm": 0.8107964992523193,
+      "learning_rate": 6.890021849963583e-05,
+      "loss": 2.1715,
+      "step": 501
+    },
+    {
+      "epoch": 0.347224623897631,
+      "grad_norm": 0.7223472595214844,
+      "learning_rate": 6.882738528769119e-05,
+      "loss": 1.8998,
+      "step": 502
+    },
+    {
+      "epoch": 0.34791630641535537,
+      "grad_norm": 0.7550078630447388,
+      "learning_rate": 6.875455207574655e-05,
+      "loss": 1.8598,
+      "step": 503
+    },
+    {
+      "epoch": 0.3486079889330797,
+      "grad_norm": 0.809738039970398,
+      "learning_rate": 6.868171886380189e-05,
+      "loss": 1.2291,
+      "step": 504
+    },
+    {
+      "epoch": 0.3492996714508041,
+      "grad_norm": 0.7400919795036316,
+      "learning_rate": 6.860888565185725e-05,
+      "loss": 1.6067,
+      "step": 505
+    },
+    {
+      "epoch": 0.34999135396852843,
+      "grad_norm": 0.6700140833854675,
+      "learning_rate": 6.85360524399126e-05,
+      "loss": 1.7842,
+      "step": 506
+    },
+    {
+      "epoch": 0.3506830364862528,
+      "grad_norm": 0.7716772556304932,
+      "learning_rate": 6.846321922796796e-05,
+      "loss": 1.2516,
+      "step": 507
+    },
+    {
+      "epoch": 0.35137471900397715,
+      "grad_norm": 0.6294199824333191,
+      "learning_rate": 6.83903860160233e-05,
+      "loss": 1.2296,
+      "step": 508
+    },
+    {
+      "epoch": 0.35206640152170154,
+      "grad_norm": 0.7753986716270447,
+      "learning_rate": 6.831755280407867e-05,
+      "loss": 1.4373,
+      "step": 509
+    },
+    {
+      "epoch": 0.3527580840394259,
+      "grad_norm": 0.7429059147834778,
+      "learning_rate": 6.824471959213403e-05,
+      "loss": 0.8905,
+      "step": 510
+    },
+    {
+      "epoch": 0.35344976655715027,
+      "grad_norm": 0.6748328804969788,
+      "learning_rate": 6.817188638018936e-05,
+      "loss": 1.4384,
+      "step": 511
+    },
+    {
+      "epoch": 0.35414144907487466,
+      "grad_norm": 1.1439710855484009,
+      "learning_rate": 6.809905316824472e-05,
+      "loss": 1.5329,
+      "step": 512
+    },
+    {
+      "epoch": 0.354833131592599,
+      "grad_norm": 0.7162743806838989,
+      "learning_rate": 6.802621995630007e-05,
+      "loss": 2.1022,
+      "step": 513
+    },
+    {
+      "epoch": 0.3555248141103234,
+      "grad_norm": 0.7660753726959229,
+      "learning_rate": 6.795338674435543e-05,
+      "loss": 2.1194,
+      "step": 514
+    },
+    {
+      "epoch": 0.3562164966280477,
+      "grad_norm": 0.7181832194328308,
+      "learning_rate": 6.788055353241078e-05,
+      "loss": 2.1352,
+      "step": 515
+    },
+    {
+      "epoch": 0.3569081791457721,
+      "grad_norm": 2.123023748397827,
+      "learning_rate": 6.780772032046614e-05,
+      "loss": 2.1801,
+      "step": 516
+    },
+    {
+      "epoch": 0.35759986166349644,
+      "grad_norm": 0.776202380657196,
+      "learning_rate": 6.77348871085215e-05,
+      "loss": 1.4564,
+      "step": 517
+    },
+    {
+      "epoch": 0.35829154418122083,
+      "grad_norm": 0.7404360175132751,
+      "learning_rate": 6.766205389657683e-05,
+      "loss": 1.6543,
+      "step": 518
+    },
+    {
+      "epoch": 0.35898322669894517,
+      "grad_norm": 0.807712197303772,
+      "learning_rate": 6.75892206846322e-05,
+      "loss": 2.0634,
+      "step": 519
+    },
+    {
+      "epoch": 0.35967490921666956,
+      "grad_norm": 0.8026816248893738,
+      "learning_rate": 6.751638747268754e-05,
+      "loss": 1.3129,
+      "step": 520
+    },
+    {
+      "epoch": 0.3603665917343939,
+      "grad_norm": 0.724224328994751,
+      "learning_rate": 6.74435542607429e-05,
+      "loss": 1.5048,
+      "step": 521
+    },
+    {
+      "epoch": 0.3610582742521183,
+      "grad_norm": 0.7268713116645813,
+      "learning_rate": 6.737072104879825e-05,
+      "loss": 1.3965,
+      "step": 522
+    },
+    {
+      "epoch": 0.3617499567698426,
+      "grad_norm": 0.6612991690635681,
+      "learning_rate": 6.729788783685361e-05,
+      "loss": 1.9812,
+      "step": 523
+    },
+    {
+      "epoch": 0.362441639287567,
+      "grad_norm": 1.0223232507705688,
+      "learning_rate": 6.722505462490896e-05,
+      "loss": 1.3578,
+      "step": 524
+    },
+    {
+      "epoch": 0.3631333218052914,
+      "grad_norm": 0.9744395613670349,
+      "learning_rate": 6.715222141296431e-05,
+      "loss": 1.6046,
+      "step": 525
+    },
+    {
+      "epoch": 0.36382500432301573,
+      "grad_norm": 1.1930111646652222,
+      "learning_rate": 6.707938820101967e-05,
+      "loss": 2.1776,
+      "step": 526
+    },
+    {
+      "epoch": 0.3645166868407401,
+      "grad_norm": 0.8520130515098572,
+      "learning_rate": 6.700655498907502e-05,
+      "loss": 1.9713,
+      "step": 527
+    },
+    {
+      "epoch": 0.36520836935846446,
+      "grad_norm": 0.7934679985046387,
+      "learning_rate": 6.693372177713038e-05,
+      "loss": 1.5202,
+      "step": 528
+    },
+    {
+      "epoch": 0.36590005187618885,
+      "grad_norm": 0.8228335976600647,
+      "learning_rate": 6.686088856518573e-05,
+      "loss": 1.8218,
+      "step": 529
+    },
+    {
+      "epoch": 0.3665917343939132,
+      "grad_norm": 0.6831315755844116,
+      "learning_rate": 6.678805535324109e-05,
+      "loss": 1.3568,
+      "step": 530
+    },
+    {
+      "epoch": 0.3672834169116376,
+      "grad_norm": 0.84079509973526,
+      "learning_rate": 6.671522214129643e-05,
+      "loss": 1.5994,
+      "step": 531
+    },
+    {
+      "epoch": 0.3679750994293619,
+      "grad_norm": 0.7911645770072937,
+      "learning_rate": 6.664238892935178e-05,
+      "loss": 1.0892,
+      "step": 532
+    },
+    {
+      "epoch": 0.3686667819470863,
+      "grad_norm": 0.7923005819320679,
+      "learning_rate": 6.656955571740714e-05,
+      "loss": 1.3006,
+      "step": 533
+    },
+    {
+      "epoch": 0.36935846446481063,
+      "grad_norm": 0.7611230611801147,
+      "learning_rate": 6.649672250546249e-05,
+      "loss": 1.2649,
+      "step": 534
+    },
+    {
+      "epoch": 0.370050146982535,
+      "grad_norm": 0.6713980436325073,
+      "learning_rate": 6.642388929351785e-05,
+      "loss": 2.1305,
+      "step": 535
+    },
+    {
+      "epoch": 0.37074182950025936,
+      "grad_norm": 1.0868254899978638,
+      "learning_rate": 6.63510560815732e-05,
+      "loss": 1.6372,
+      "step": 536
+    },
+    {
+      "epoch": 0.37143351201798375,
+      "grad_norm": 0.7361617684364319,
+      "learning_rate": 6.627822286962856e-05,
+      "loss": 1.8807,
+      "step": 537
+    },
+    {
+      "epoch": 0.37212519453570814,
+      "grad_norm": 0.7903793454170227,
+      "learning_rate": 6.620538965768391e-05,
+      "loss": 2.0083,
+      "step": 538
+    },
+    {
+      "epoch": 0.37281687705343247,
+      "grad_norm": 0.8564360737800598,
+      "learning_rate": 6.613255644573925e-05,
+      "loss": 2.1507,
+      "step": 539
+    },
+    {
+      "epoch": 0.37350855957115686,
+      "grad_norm": 0.6386767029762268,
+      "learning_rate": 6.605972323379462e-05,
+      "loss": 1.4482,
+      "step": 540
+    },
+    {
+      "epoch": 0.3742002420888812,
+      "grad_norm": 0.937019944190979,
+      "learning_rate": 6.598689002184996e-05,
+      "loss": 2.3786,
+      "step": 541
+    },
+    {
+      "epoch": 0.3748919246066056,
+      "grad_norm": 1.052297830581665,
+      "learning_rate": 6.591405680990532e-05,
+      "loss": 2.1223,
+      "step": 542
+    },
+    {
+      "epoch": 0.3755836071243299,
+      "grad_norm": 0.8783723711967468,
+      "learning_rate": 6.584122359796067e-05,
+      "loss": 1.7049,
+      "step": 543
+    },
+    {
+      "epoch": 0.3762752896420543,
+      "grad_norm": 0.9165799021720886,
+      "learning_rate": 6.576839038601602e-05,
+      "loss": 2.062,
+      "step": 544
+    },
+    {
+      "epoch": 0.37696697215977865,
+      "grad_norm": 0.722274661064148,
+      "learning_rate": 6.569555717407138e-05,
+      "loss": 2.0566,
+      "step": 545
+    },
+    {
+      "epoch": 0.37765865467750304,
+      "grad_norm": 0.6692977547645569,
+      "learning_rate": 6.562272396212673e-05,
+      "loss": 1.8474,
+      "step": 546
+    },
+    {
+      "epoch": 0.37835033719522737,
+      "grad_norm": 0.8514979481697083,
+      "learning_rate": 6.554989075018209e-05,
+      "loss": 2.1704,
+      "step": 547
+    },
+    {
+      "epoch": 0.37904201971295176,
+      "grad_norm": 0.8156755566596985,
+      "learning_rate": 6.547705753823744e-05,
+      "loss": 2.0282,
+      "step": 548
+    },
+    {
+      "epoch": 0.3797337022306761,
+      "grad_norm": 0.728892982006073,
+      "learning_rate": 6.54042243262928e-05,
+      "loss": 1.6018,
+      "step": 549
+    },
+    {
+      "epoch": 0.3804253847484005,
+      "grad_norm": 0.6585319638252258,
+      "learning_rate": 6.533139111434814e-05,
+      "loss": 1.2368,
+      "step": 550
+    },
+    {
+      "epoch": 0.3811170672661249,
+      "grad_norm": 0.998591423034668,
+      "learning_rate": 6.525855790240349e-05,
+      "loss": 2.1364,
+      "step": 551
+    },
+    {
+      "epoch": 0.3818087497838492,
+      "grad_norm": 0.7700340151786804,
+      "learning_rate": 6.518572469045885e-05,
+      "loss": 1.602,
+      "step": 552
+    },
+    {
+      "epoch": 0.3825004323015736,
+      "grad_norm": 0.8582996129989624,
+      "learning_rate": 6.51128914785142e-05,
+      "loss": 1.9084,
+      "step": 553
+    },
+    {
+      "epoch": 0.38319211481929794,
+      "grad_norm": 0.6509499549865723,
+      "learning_rate": 6.504005826656956e-05,
+      "loss": 1.4457,
+      "step": 554
+    },
+    {
+      "epoch": 0.3838837973370223,
+      "grad_norm": 0.8587490320205688,
+      "learning_rate": 6.496722505462491e-05,
+      "loss": 1.5043,
+      "step": 555
+    },
+    {
+      "epoch": 0.38457547985474666,
+      "grad_norm": 0.8849571943283081,
+      "learning_rate": 6.489439184268027e-05,
+      "loss": 1.801,
+      "step": 556
+    },
+    {
+      "epoch": 0.38526716237247105,
+      "grad_norm": 0.8862190246582031,
+      "learning_rate": 6.482155863073562e-05,
+      "loss": 2.1286,
+      "step": 557
+    },
+    {
+      "epoch": 0.3859588448901954,
+      "grad_norm": 1.023720383644104,
+      "learning_rate": 6.474872541879097e-05,
+      "loss": 1.2497,
+      "step": 558
+    },
+    {
+      "epoch": 0.3866505274079198,
+      "grad_norm": 0.6962110996246338,
+      "learning_rate": 6.467589220684633e-05,
+      "loss": 1.8769,
+      "step": 559
+    },
+    {
+      "epoch": 0.3873422099256441,
+      "grad_norm": 0.8252090215682983,
+      "learning_rate": 6.460305899490167e-05,
+      "loss": 1.8998,
+      "step": 560
+    },
+    {
+      "epoch": 0.3880338924433685,
+      "grad_norm": 0.7156630754470825,
+      "learning_rate": 6.453022578295704e-05,
+      "loss": 0.8263,
+      "step": 561
+    },
+    {
+      "epoch": 0.38872557496109283,
+      "grad_norm": 0.685234785079956,
+      "learning_rate": 6.445739257101238e-05,
+      "loss": 0.7216,
+      "step": 562
+    },
+    {
+      "epoch": 0.3894172574788172,
+      "grad_norm": 0.7868284583091736,
+      "learning_rate": 6.438455935906774e-05,
+      "loss": 2.0959,
+      "step": 563
+    },
+    {
+      "epoch": 0.3901089399965416,
+      "grad_norm": 0.8539921045303345,
+      "learning_rate": 6.431172614712309e-05,
+      "loss": 1.3863,
+      "step": 564
+    },
+    {
+      "epoch": 0.39080062251426595,
+      "grad_norm": 1.0316025018692017,
+      "learning_rate": 6.423889293517844e-05,
+      "loss": 1.0292,
+      "step": 565
+    },
+    {
+      "epoch": 0.39149230503199034,
+      "grad_norm": 0.7655407786369324,
+      "learning_rate": 6.41660597232338e-05,
+      "loss": 1.8794,
+      "step": 566
+    },
+    {
+      "epoch": 0.3921839875497147,
+      "grad_norm": 0.6936174035072327,
+      "learning_rate": 6.409322651128915e-05,
+      "loss": 1.3902,
+      "step": 567
+    },
+    {
+      "epoch": 0.39287567006743906,
+      "grad_norm": 0.7873200178146362,
+      "learning_rate": 6.402039329934451e-05,
+      "loss": 1.5278,
+      "step": 568
+    },
+    {
+      "epoch": 0.3935673525851634,
+      "grad_norm": 1.3935117721557617,
+      "learning_rate": 6.394756008739986e-05,
+      "loss": 0.7081,
+      "step": 569
+    },
+    {
+      "epoch": 0.3942590351028878,
+      "grad_norm": 0.7773651480674744,
+      "learning_rate": 6.387472687545522e-05,
+      "loss": 1.548,
+      "step": 570
+    },
+    {
+      "epoch": 0.3949507176206121,
+      "grad_norm": 0.8644070625305176,
+      "learning_rate": 6.380189366351056e-05,
+      "loss": 1.4528,
+      "step": 571
+    },
+    {
+      "epoch": 0.3956424001383365,
+      "grad_norm": 0.7735791206359863,
+      "learning_rate": 6.372906045156591e-05,
+      "loss": 1.8554,
+      "step": 572
+    },
+    {
+      "epoch": 0.39633408265606085,
+      "grad_norm": 0.7515410780906677,
+      "learning_rate": 6.365622723962127e-05,
+      "loss": 1.786,
+      "step": 573
+    },
+    {
+      "epoch": 0.39702576517378524,
+      "grad_norm": 1.0612385272979736,
+      "learning_rate": 6.358339402767662e-05,
+      "loss": 1.3569,
+      "step": 574
+    },
+    {
+      "epoch": 0.3977174476915096,
+      "grad_norm": 0.919940173625946,
+      "learning_rate": 6.351056081573198e-05,
+      "loss": 2.0286,
+      "step": 575
+    },
+    {
+      "epoch": 0.39840913020923396,
+      "grad_norm": 0.8676942586898804,
+      "learning_rate": 6.343772760378733e-05,
+      "loss": 1.2646,
+      "step": 576
+    },
+    {
+      "epoch": 0.39910081272695835,
+      "grad_norm": 0.6343767642974854,
+      "learning_rate": 6.336489439184269e-05,
+      "loss": 0.9905,
+      "step": 577
+    },
+    {
+      "epoch": 0.3997924952446827,
+      "grad_norm": 1.3755874633789062,
+      "learning_rate": 6.329206117989804e-05,
+      "loss": 1.4657,
+      "step": 578
+    },
+    {
+      "epoch": 0.4004841777624071,
+      "grad_norm": 0.9563671946525574,
+      "learning_rate": 6.321922796795339e-05,
+      "loss": 1.6394,
+      "step": 579
+    },
+    {
+      "epoch": 0.4011758602801314,
+      "grad_norm": 1.3659626245498657,
+      "learning_rate": 6.314639475600875e-05,
+      "loss": 1.3927,
+      "step": 580
+    },
+    {
+      "epoch": 0.4018675427978558,
+      "grad_norm": 0.8047080039978027,
+      "learning_rate": 6.30735615440641e-05,
+      "loss": 1.9877,
+      "step": 581
+    },
+    {
+      "epoch": 0.40255922531558014,
+      "grad_norm": 0.872704267501831,
+      "learning_rate": 6.300072833211945e-05,
+      "loss": 1.7843,
+      "step": 582
+    },
+    {
+      "epoch": 0.40325090783330453,
+      "grad_norm": 0.5949119329452515,
+      "learning_rate": 6.29278951201748e-05,
+      "loss": 1.041,
+      "step": 583
+    },
+    {
+      "epoch": 0.40394259035102886,
+      "grad_norm": 0.8101840615272522,
+      "learning_rate": 6.285506190823015e-05,
+      "loss": 1.4963,
+      "step": 584
+    },
+    {
+      "epoch": 0.40463427286875325,
+      "grad_norm": 0.8406373858451843,
+      "learning_rate": 6.278222869628551e-05,
+      "loss": 1.0127,
+      "step": 585
+    },
+    {
+      "epoch": 0.4053259553864776,
+      "grad_norm": 0.9587801098823547,
+      "learning_rate": 6.270939548434086e-05,
+      "loss": 2.0066,
+      "step": 586
+    },
+    {
+      "epoch": 0.406017637904202,
+      "grad_norm": 0.9393382668495178,
+      "learning_rate": 6.263656227239622e-05,
+      "loss": 1.3456,
+      "step": 587
+    },
+    {
+      "epoch": 0.4067093204219263,
+      "grad_norm": 0.7072182297706604,
+      "learning_rate": 6.256372906045157e-05,
+      "loss": 1.381,
+      "step": 588
+    },
+    {
+      "epoch": 0.4074010029396507,
+      "grad_norm": 0.7435779571533203,
+      "learning_rate": 6.249089584850693e-05,
+      "loss": 1.3972,
+      "step": 589
+    },
+    {
+      "epoch": 0.40809268545737504,
+      "grad_norm": 0.8602951169013977,
+      "learning_rate": 6.241806263656228e-05,
+      "loss": 1.581,
+      "step": 590
+    },
+    {
+      "epoch": 0.4087843679750994,
+      "grad_norm": 0.975732684135437,
+      "learning_rate": 6.234522942461762e-05,
+      "loss": 2.0073,
+      "step": 591
+    },
+    {
+      "epoch": 0.4094760504928238,
+      "grad_norm": 0.8923630118370056,
+      "learning_rate": 6.227239621267298e-05,
+      "loss": 2.0375,
+      "step": 592
+    },
+    {
+      "epoch": 0.41016773301054815,
+      "grad_norm": 0.6812586188316345,
+      "learning_rate": 6.219956300072833e-05,
+      "loss": 1.3661,
+      "step": 593
+    },
+    {
+      "epoch": 0.41085941552827254,
+      "grad_norm": 1.0283271074295044,
+      "learning_rate": 6.212672978878369e-05,
+      "loss": 1.7544,
+      "step": 594
+    },
+    {
+      "epoch": 0.4115510980459969,
+      "grad_norm": 0.8602912425994873,
+      "learning_rate": 6.205389657683904e-05,
+      "loss": 1.9679,
+      "step": 595
+    },
+    {
+      "epoch": 0.41224278056372127,
+      "grad_norm": 0.7977223992347717,
+      "learning_rate": 6.19810633648944e-05,
+      "loss": 1.7853,
+      "step": 596
+    },
+    {
+      "epoch": 0.4129344630814456,
+      "grad_norm": 0.8174147605895996,
+      "learning_rate": 6.190823015294975e-05,
+      "loss": 1.8254,
+      "step": 597
+    },
+    {
+      "epoch": 0.41362614559917,
+      "grad_norm": 1.1011167764663696,
+      "learning_rate": 6.18353969410051e-05,
+      "loss": 1.6991,
+      "step": 598
+    },
+    {
+      "epoch": 0.4143178281168943,
+      "grad_norm": 0.8485236167907715,
+      "learning_rate": 6.176256372906046e-05,
+      "loss": 1.709,
+      "step": 599
+    },
+    {
+      "epoch": 0.4150095106346187,
+      "grad_norm": 0.8971522450447083,
+      "learning_rate": 6.16897305171158e-05,
+      "loss": 1.7905,
+      "step": 600
+    },
+    {
+      "epoch": 0.41570119315234305,
+      "grad_norm": 0.8392210602760315,
+      "learning_rate": 6.161689730517117e-05,
+      "loss": 1.5102,
+      "step": 601
+    },
+    {
+      "epoch": 0.41639287567006744,
+      "grad_norm": 0.7831077575683594,
+      "learning_rate": 6.154406409322651e-05,
+      "loss": 1.1469,
+      "step": 602
+    },
+    {
+      "epoch": 0.4170845581877918,
+      "grad_norm": 0.7571394443511963,
+      "learning_rate": 6.147123088128187e-05,
+      "loss": 1.1778,
+      "step": 603
+    },
+    {
+      "epoch": 0.41777624070551617,
+      "grad_norm": 0.7868969440460205,
+      "learning_rate": 6.139839766933721e-05,
+      "loss": 1.8391,
+      "step": 604
+    },
+    {
+      "epoch": 0.41846792322324056,
+      "grad_norm": 0.967659592628479,
+      "learning_rate": 6.132556445739257e-05,
+      "loss": 1.9437,
+      "step": 605
+    },
+    {
+      "epoch": 0.4191596057409649,
+      "grad_norm": 1.1932754516601562,
+      "learning_rate": 6.125273124544793e-05,
+      "loss": 2.0343,
+      "step": 606
+    },
+    {
+      "epoch": 0.4198512882586893,
+      "grad_norm": 0.692387044429779,
+      "learning_rate": 6.117989803350328e-05,
+      "loss": 1.2014,
+      "step": 607
+    },
+    {
+      "epoch": 0.4205429707764136,
+      "grad_norm": 0.9984387755393982,
+      "learning_rate": 6.110706482155864e-05,
+      "loss": 1.1817,
+      "step": 608
+    },
+    {
+      "epoch": 0.421234653294138,
+      "grad_norm": 1.1018468141555786,
+      "learning_rate": 6.103423160961399e-05,
+      "loss": 1.6788,
+      "step": 609
+    },
+    {
+      "epoch": 0.42192633581186234,
+      "grad_norm": 0.9406405091285706,
+      "learning_rate": 6.096139839766935e-05,
+      "loss": 1.5549,
+      "step": 610
+    },
+    {
+      "epoch": 0.42261801832958673,
+      "grad_norm": 0.867577075958252,
+      "learning_rate": 6.088856518572469e-05,
+      "loss": 2.139,
+      "step": 611
+    },
+    {
+      "epoch": 0.42330970084731107,
+      "grad_norm": 0.7860278487205505,
+      "learning_rate": 6.081573197378004e-05,
+      "loss": 1.6854,
+      "step": 612
+    },
+    {
+      "epoch": 0.42400138336503546,
+      "grad_norm": 0.884482741355896,
+      "learning_rate": 6.07428987618354e-05,
+      "loss": 1.304,
+      "step": 613
+    },
+    {
+      "epoch": 0.4246930658827598,
+      "grad_norm": 0.8831844329833984,
+      "learning_rate": 6.067006554989075e-05,
+      "loss": 1.9461,
+      "step": 614
+    },
+    {
+      "epoch": 0.4253847484004842,
+      "grad_norm": 0.831235945224762,
+      "learning_rate": 6.0597232337946106e-05,
+      "loss": 1.9683,
+      "step": 615
+    },
+    {
+      "epoch": 0.4260764309182085,
+      "grad_norm": 0.634988009929657,
+      "learning_rate": 6.052439912600146e-05,
+      "loss": 1.6396,
+      "step": 616
+    },
+    {
+      "epoch": 0.4267681134359329,
+      "grad_norm": 0.6646054983139038,
+      "learning_rate": 6.045156591405682e-05,
+      "loss": 1.1917,
+      "step": 617
+    },
+    {
+      "epoch": 0.4274597959536573,
+      "grad_norm": 0.9487079977989197,
+      "learning_rate": 6.037873270211216e-05,
+      "loss": 1.9417,
+      "step": 618
+    },
+    {
+      "epoch": 0.42815147847138163,
+      "grad_norm": 0.6744237542152405,
+      "learning_rate": 6.0305899490167516e-05,
+      "loss": 0.8873,
+      "step": 619
+    },
+    {
+      "epoch": 0.428843160989106,
+      "grad_norm": 0.8524720668792725,
+      "learning_rate": 6.023306627822287e-05,
+      "loss": 1.8525,
+      "step": 620
+    },
+    {
+      "epoch": 0.42953484350683035,
+      "grad_norm": 0.8825080394744873,
+      "learning_rate": 6.0160233066278225e-05,
+      "loss": 1.3985,
+      "step": 621
+    },
+    {
+      "epoch": 0.43022652602455475,
+      "grad_norm": 1.0640051364898682,
+      "learning_rate": 6.008739985433358e-05,
+      "loss": 2.0577,
+      "step": 622
+    },
+    {
+      "epoch": 0.4309182085422791,
+      "grad_norm": 1.0512269735336304,
+      "learning_rate": 6.001456664238893e-05,
+      "loss": 1.744,
+      "step": 623
+    },
+    {
+      "epoch": 0.43160989106000347,
+      "grad_norm": 0.6345813870429993,
+      "learning_rate": 5.994173343044428e-05,
+      "loss": 1.7586,
+      "step": 624
+    },
+    {
+      "epoch": 0.4323015735777278,
+      "grad_norm": 1.1025394201278687,
+      "learning_rate": 5.9868900218499635e-05,
+      "loss": 1.7011,
+      "step": 625
+    },
+    {
+      "epoch": 0.4329932560954522,
+      "grad_norm": 1.0276751518249512,
+      "learning_rate": 5.979606700655499e-05,
+      "loss": 1.8624,
+      "step": 626
+    },
+    {
+      "epoch": 0.43368493861317653,
+      "grad_norm": 0.9713005423545837,
+      "learning_rate": 5.9723233794610344e-05,
+      "loss": 1.5493,
+      "step": 627
+    },
+    {
+      "epoch": 0.4343766211309009,
+      "grad_norm": 0.7826513051986694,
+      "learning_rate": 5.96504005826657e-05,
+      "loss": 1.6086,
+      "step": 628
+    },
+    {
+      "epoch": 0.43506830364862525,
+      "grad_norm": 0.7218225002288818,
+      "learning_rate": 5.957756737072105e-05,
+      "loss": 1.5445,
+      "step": 629
+    },
+    {
+      "epoch": 0.43575998616634964,
+      "grad_norm": 0.8310878872871399,
+      "learning_rate": 5.950473415877641e-05,
+      "loss": 1.9013,
+      "step": 630
+    },
+    {
+      "epoch": 0.43645166868407403,
+      "grad_norm": 1.0086404085159302,
+      "learning_rate": 5.9431900946831754e-05,
+      "loss": 2.1482,
+      "step": 631
+    },
+    {
+      "epoch": 0.43714335120179837,
+      "grad_norm": 0.7284911870956421,
+      "learning_rate": 5.935906773488711e-05,
+      "loss": 1.5093,
+      "step": 632
+    },
+    {
+      "epoch": 0.43783503371952276,
+      "grad_norm": 0.8275542259216309,
+      "learning_rate": 5.928623452294246e-05,
+      "loss": 1.5309,
+      "step": 633
+    },
+    {
+      "epoch": 0.4385267162372471,
+      "grad_norm": 0.9007952809333801,
+      "learning_rate": 5.921340131099782e-05,
+      "loss": 2.0425,
+      "step": 634
+    },
+    {
+      "epoch": 0.4392183987549715,
+      "grad_norm": 0.6324974298477173,
+      "learning_rate": 5.914056809905317e-05,
+      "loss": 1.1334,
+      "step": 635
+    },
+    {
+      "epoch": 0.4399100812726958,
+      "grad_norm": 0.8025960326194763,
+      "learning_rate": 5.9067734887108526e-05,
+      "loss": 1.381,
+      "step": 636
+    },
+    {
+      "epoch": 0.4406017637904202,
+      "grad_norm": 1.008012294769287,
+      "learning_rate": 5.899490167516388e-05,
+      "loss": 1.7118,
+      "step": 637
+    },
+    {
+      "epoch": 0.44129344630814454,
+      "grad_norm": 0.9379268288612366,
+      "learning_rate": 5.892206846321923e-05,
+      "loss": 1.7564,
+      "step": 638
+    },
+    {
+      "epoch": 0.44198512882586893,
+      "grad_norm": 0.9095461964607239,
+      "learning_rate": 5.884923525127458e-05,
+      "loss": 1.7896,
+      "step": 639
+    },
+    {
+      "epoch": 0.44267681134359327,
+      "grad_norm": 0.8308787941932678,
+      "learning_rate": 5.8776402039329936e-05,
+      "loss": 1.7003,
+      "step": 640
+    },
+    {
+      "epoch": 0.44336849386131766,
+      "grad_norm": 0.7289214730262756,
+      "learning_rate": 5.870356882738529e-05,
+      "loss": 1.8206,
+      "step": 641
+    },
+    {
+      "epoch": 0.444060176379042,
+      "grad_norm": 0.6933354139328003,
+      "learning_rate": 5.8630735615440645e-05,
+      "loss": 1.4336,
+      "step": 642
+    },
+    {
+      "epoch": 0.4447518588967664,
+      "grad_norm": 0.6738266944885254,
+      "learning_rate": 5.8557902403496e-05,
+      "loss": 1.3322,
+      "step": 643
+    },
+    {
+      "epoch": 0.4454435414144908,
+      "grad_norm": 0.6887166500091553,
+      "learning_rate": 5.8485069191551346e-05,
+      "loss": 1.2408,
+      "step": 644
+    },
+    {
+      "epoch": 0.4461352239322151,
+      "grad_norm": 0.8172757029533386,
+      "learning_rate": 5.84122359796067e-05,
+      "loss": 1.0118,
+      "step": 645
+    },
+    {
+      "epoch": 0.4468269064499395,
+      "grad_norm": 0.8521438241004944,
+      "learning_rate": 5.8339402767662055e-05,
+      "loss": 1.9106,
+      "step": 646
+    },
+    {
+      "epoch": 0.44751858896766383,
+      "grad_norm": 0.6485393047332764,
+      "learning_rate": 5.826656955571741e-05,
+      "loss": 1.1009,
+      "step": 647
+    },
+    {
+      "epoch": 0.4482102714853882,
+      "grad_norm": 0.6865597367286682,
+      "learning_rate": 5.8193736343772763e-05,
+      "loss": 1.3842,
+      "step": 648
+    },
+    {
+      "epoch": 0.44890195400311256,
+      "grad_norm": 0.8116838335990906,
+      "learning_rate": 5.812090313182812e-05,
+      "loss": 1.61,
+      "step": 649
+    },
+    {
+      "epoch": 0.44959363652083695,
+      "grad_norm": 0.6286530494689941,
+      "learning_rate": 5.804806991988347e-05,
+      "loss": 1.6078,
+      "step": 650
+    },
+    {
+      "epoch": 0.4502853190385613,
+      "grad_norm": 1.3772015571594238,
+      "learning_rate": 5.797523670793882e-05,
+      "loss": 1.3162,
+      "step": 651
+    },
+    {
+      "epoch": 0.4509770015562857,
+      "grad_norm": 1.0899200439453125,
+      "learning_rate": 5.7902403495994174e-05,
+      "loss": 1.7445,
+      "step": 652
+    },
+    {
+      "epoch": 0.45166868407401,
+      "grad_norm": 0.8981252908706665,
+      "learning_rate": 5.782957028404953e-05,
+      "loss": 1.4101,
+      "step": 653
+    },
+    {
+      "epoch": 0.4523603665917344,
+      "grad_norm": 0.8054170608520508,
+      "learning_rate": 5.775673707210488e-05,
+      "loss": 1.5026,
+      "step": 654
+    },
+    {
+      "epoch": 0.45305204910945873,
+      "grad_norm": 0.7385475635528564,
+      "learning_rate": 5.768390386016024e-05,
+      "loss": 1.1069,
+      "step": 655
+    },
+    {
+      "epoch": 0.4537437316271831,
+      "grad_norm": 0.7658798098564148,
+      "learning_rate": 5.761107064821559e-05,
+      "loss": 1.4528,
+      "step": 656
+    },
+    {
+      "epoch": 0.4544354141449075,
+      "grad_norm": 0.8078415989875793,
+      "learning_rate": 5.7538237436270945e-05,
+      "loss": 1.9071,
+      "step": 657
+    },
+    {
+      "epoch": 0.45512709666263185,
+      "grad_norm": 0.8280815482139587,
+      "learning_rate": 5.746540422432629e-05,
+      "loss": 1.7216,
+      "step": 658
+    },
+    {
+      "epoch": 0.45581877918035624,
+      "grad_norm": 0.8316612243652344,
+      "learning_rate": 5.739257101238165e-05,
+      "loss": 1.8904,
+      "step": 659
+    },
+    {
+      "epoch": 0.45651046169808057,
+      "grad_norm": 0.6992987394332886,
+      "learning_rate": 5.7319737800437e-05,
+      "loss": 2.0018,
+      "step": 660
+    },
+    {
+      "epoch": 0.45720214421580496,
+      "grad_norm": 1.0973151922225952,
+      "learning_rate": 5.7246904588492356e-05,
+      "loss": 1.2894,
+      "step": 661
+    },
+    {
+      "epoch": 0.4578938267335293,
+      "grad_norm": 1.1129229068756104,
+      "learning_rate": 5.717407137654771e-05,
+      "loss": 2.0258,
+      "step": 662
+    },
+    {
+      "epoch": 0.4585855092512537,
+      "grad_norm": 0.8779314756393433,
+      "learning_rate": 5.7101238164603064e-05,
+      "loss": 1.887,
+      "step": 663
+    },
+    {
+      "epoch": 0.459277191768978,
+      "grad_norm": 0.7320972681045532,
+      "learning_rate": 5.702840495265841e-05,
+      "loss": 0.7374,
+      "step": 664
+    },
+    {
+      "epoch": 0.4599688742867024,
+      "grad_norm": 0.6837896704673767,
+      "learning_rate": 5.6955571740713766e-05,
+      "loss": 1.4664,
+      "step": 665
+    },
+    {
+      "epoch": 0.46066055680442675,
+      "grad_norm": 0.9255065321922302,
+      "learning_rate": 5.688273852876912e-05,
+      "loss": 1.4429,
+      "step": 666
+    },
+    {
+      "epoch": 0.46135223932215114,
+      "grad_norm": 0.7060451507568359,
+      "learning_rate": 5.6809905316824475e-05,
+      "loss": 1.8314,
+      "step": 667
+    },
+    {
+      "epoch": 0.46204392183987547,
+      "grad_norm": 0.6706486344337463,
+      "learning_rate": 5.673707210487983e-05,
+      "loss": 1.2471,
+      "step": 668
+    },
+    {
+      "epoch": 0.46273560435759986,
+      "grad_norm": 0.6796824336051941,
+      "learning_rate": 5.666423889293518e-05,
+      "loss": 1.9348,
+      "step": 669
+    },
+    {
+      "epoch": 0.46342728687532425,
+      "grad_norm": 0.8009390830993652,
+      "learning_rate": 5.659140568099054e-05,
+      "loss": 1.0403,
+      "step": 670
+    },
+    {
+      "epoch": 0.4641189693930486,
+      "grad_norm": 1.052715539932251,
+      "learning_rate": 5.6518572469045885e-05,
+      "loss": 1.4263,
+      "step": 671
+    },
+    {
+      "epoch": 0.464810651910773,
+      "grad_norm": 0.8654411435127258,
+      "learning_rate": 5.644573925710124e-05,
+      "loss": 1.5326,
+      "step": 672
+    },
+    {
+      "epoch": 0.4655023344284973,
+      "grad_norm": 0.8049682378768921,
+      "learning_rate": 5.6372906045156594e-05,
+      "loss": 1.8094,
+      "step": 673
+    },
+    {
+      "epoch": 0.4661940169462217,
+      "grad_norm": 0.6251804828643799,
+      "learning_rate": 5.630007283321195e-05,
+      "loss": 1.1813,
+      "step": 674
+    },
+    {
+      "epoch": 0.46688569946394604,
+      "grad_norm": 0.7386276125907898,
+      "learning_rate": 5.62272396212673e-05,
+      "loss": 1.3354,
+      "step": 675
+    },
+    {
+      "epoch": 0.4675773819816704,
+      "grad_norm": 0.8355931043624878,
+      "learning_rate": 5.6154406409322656e-05,
+      "loss": 1.515,
+      "step": 676
+    },
+    {
+      "epoch": 0.46826906449939476,
+      "grad_norm": 0.8459252715110779,
+      "learning_rate": 5.608157319737801e-05,
+      "loss": 2.2214,
+      "step": 677
+    },
+    {
+      "epoch": 0.46896074701711915,
+      "grad_norm": 0.8006874322891235,
+      "learning_rate": 5.600873998543336e-05,
+      "loss": 1.8108,
+      "step": 678
+    },
+    {
+      "epoch": 0.4696524295348435,
+      "grad_norm": 1.2057729959487915,
+      "learning_rate": 5.593590677348871e-05,
+      "loss": 1.2744,
+      "step": 679
+    },
+    {
+      "epoch": 0.4703441120525679,
+      "grad_norm": 0.8938283324241638,
+      "learning_rate": 5.586307356154407e-05,
+      "loss": 2.0349,
+      "step": 680
+    },
+    {
+      "epoch": 0.4710357945702922,
+      "grad_norm": 0.8696619272232056,
+      "learning_rate": 5.579024034959942e-05,
+      "loss": 1.552,
+      "step": 681
+    },
+    {
+      "epoch": 0.4717274770880166,
+      "grad_norm": 0.8940863013267517,
+      "learning_rate": 5.5717407137654775e-05,
+      "loss": 1.5838,
+      "step": 682
+    },
+    {
+      "epoch": 0.472419159605741,
+      "grad_norm": 0.6525213122367859,
+      "learning_rate": 5.564457392571013e-05,
+      "loss": 1.4496,
+      "step": 683
+    },
+    {
+      "epoch": 0.4731108421234653,
+      "grad_norm": 0.8376979827880859,
+      "learning_rate": 5.557174071376548e-05,
+      "loss": 1.8819,
+      "step": 684
+    },
+    {
+      "epoch": 0.4738025246411897,
+      "grad_norm": 1.6936135292053223,
+      "learning_rate": 5.549890750182083e-05,
+      "loss": 1.5663,
+      "step": 685
+    },
+    {
+      "epoch": 0.47449420715891405,
+      "grad_norm": 0.7069249153137207,
+      "learning_rate": 5.5426074289876186e-05,
+      "loss": 1.9715,
+      "step": 686
+    },
+    {
+      "epoch": 0.47518588967663844,
+      "grad_norm": 0.7965700626373291,
+      "learning_rate": 5.535324107793154e-05,
+      "loss": 1.6389,
+      "step": 687
+    },
+    {
+      "epoch": 0.4758775721943628,
+      "grad_norm": 1.397172451019287,
+      "learning_rate": 5.5280407865986894e-05,
+      "loss": 1.5286,
+      "step": 688
+    },
+    {
+      "epoch": 0.47656925471208716,
+      "grad_norm": 0.7060932517051697,
+      "learning_rate": 5.520757465404225e-05,
+      "loss": 1.1942,
+      "step": 689
+    },
+    {
+      "epoch": 0.4772609372298115,
+      "grad_norm": 0.9016676545143127,
+      "learning_rate": 5.51347414420976e-05,
+      "loss": 1.5552,
+      "step": 690
+    },
+    {
+      "epoch": 0.4779526197475359,
+      "grad_norm": 0.7569277882575989,
+      "learning_rate": 5.506190823015295e-05,
+      "loss": 2.1902,
+      "step": 691
+    },
+    {
+      "epoch": 0.4786443022652602,
+      "grad_norm": 0.8146937489509583,
+      "learning_rate": 5.4989075018208305e-05,
+      "loss": 1.3199,
+      "step": 692
+    },
+    {
+      "epoch": 0.4793359847829846,
+      "grad_norm": 0.8721641898155212,
+      "learning_rate": 5.491624180626366e-05,
+      "loss": 2.0197,
+      "step": 693
+    },
+    {
+      "epoch": 0.48002766730070895,
+      "grad_norm": 0.7318241596221924,
+      "learning_rate": 5.484340859431901e-05,
+      "loss": 1.8798,
+      "step": 694
+    },
+    {
+      "epoch": 0.48071934981843334,
+      "grad_norm": 0.9212726354598999,
+      "learning_rate": 5.477057538237437e-05,
+      "loss": 1.1606,
+      "step": 695
+    },
+    {
+      "epoch": 0.48141103233615773,
+      "grad_norm": 0.8294724822044373,
+      "learning_rate": 5.469774217042972e-05,
+      "loss": 1.0334,
+      "step": 696
+    },
+    {
+      "epoch": 0.48210271485388206,
+      "grad_norm": 0.9914029240608215,
+      "learning_rate": 5.4624908958485076e-05,
+      "loss": 1.9953,
+      "step": 697
+    },
+    {
+      "epoch": 0.48279439737160645,
+      "grad_norm": 0.6344811916351318,
+      "learning_rate": 5.4552075746540424e-05,
+      "loss": 0.8681,
+      "step": 698
+    },
+    {
+      "epoch": 0.4834860798893308,
+      "grad_norm": 0.864594578742981,
+      "learning_rate": 5.447924253459578e-05,
+      "loss": 1.5566,
+      "step": 699
+    },
+    {
+      "epoch": 0.4841777624070552,
+      "grad_norm": 0.6789159178733826,
+      "learning_rate": 5.440640932265113e-05,
+      "loss": 0.8052,
+      "step": 700
+    },
+    {
+      "epoch": 0.4848694449247795,
+      "grad_norm": 0.7401624321937561,
+      "learning_rate": 5.4333576110706487e-05,
+      "loss": 1.3816,
+      "step": 701
+    },
+    {
+      "epoch": 0.4855611274425039,
+      "grad_norm": 0.8231687545776367,
+      "learning_rate": 5.426074289876184e-05,
+      "loss": 1.7608,
+      "step": 702
+    },
+    {
+      "epoch": 0.48625280996022824,
+      "grad_norm": 0.7512969374656677,
+      "learning_rate": 5.4187909686817195e-05,
+      "loss": 0.9139,
+      "step": 703
+    },
+    {
+      "epoch": 0.48694449247795263,
+      "grad_norm": 2.862189531326294,
+      "learning_rate": 5.4115076474872536e-05,
+      "loss": 1.4689,
+      "step": 704
+    },
+    {
+      "epoch": 0.48763617499567696,
+      "grad_norm": 0.8271031975746155,
+      "learning_rate": 5.40422432629279e-05,
+      "loss": 2.0143,
+      "step": 705
+    },
+    {
+      "epoch": 0.48832785751340135,
+      "grad_norm": 0.8950322270393372,
+      "learning_rate": 5.396941005098325e-05,
+      "loss": 1.9843,
+      "step": 706
+    },
+    {
+      "epoch": 0.4890195400311257,
+      "grad_norm": 0.6882688999176025,
+      "learning_rate": 5.3896576839038606e-05,
+      "loss": 1.2655,
+      "step": 707
+    },
+    {
+      "epoch": 0.4897112225488501,
+      "grad_norm": 0.990772545337677,
+      "learning_rate": 5.382374362709396e-05,
+      "loss": 2.0287,
+      "step": 708
+    },
+    {
+      "epoch": 0.49040290506657447,
+      "grad_norm": 0.8727272152900696,
+      "learning_rate": 5.3750910415149314e-05,
+      "loss": 1.5164,
+      "step": 709
+    },
+    {
+      "epoch": 0.4910945875842988,
+      "grad_norm": 0.7968826293945312,
+      "learning_rate": 5.367807720320467e-05,
+      "loss": 1.6738,
+      "step": 710
+    },
+    {
+      "epoch": 0.4917862701020232,
+      "grad_norm": 0.8723095655441284,
+      "learning_rate": 5.360524399126001e-05,
+      "loss": 1.7006,
+      "step": 711
+    },
+    {
+      "epoch": 0.4924779526197475,
+      "grad_norm": 0.7931122779846191,
+      "learning_rate": 5.353241077931537e-05,
+      "loss": 2.2392,
+      "step": 712
+    },
+    {
+      "epoch": 0.4931696351374719,
+      "grad_norm": 0.7815384268760681,
+      "learning_rate": 5.3459577567370724e-05,
+      "loss": 1.5135,
+      "step": 713
+    },
+    {
+      "epoch": 0.49386131765519625,
+      "grad_norm": 0.6223137974739075,
+      "learning_rate": 5.338674435542608e-05,
+      "loss": 0.9653,
+      "step": 714
+    },
+    {
+      "epoch": 0.49455300017292064,
+      "grad_norm": 0.8133209347724915,
+      "learning_rate": 5.331391114348143e-05,
+      "loss": 1.7424,
+      "step": 715
+    },
+    {
+      "epoch": 0.495244682690645,
+      "grad_norm": 0.9321537017822266,
+      "learning_rate": 5.324107793153679e-05,
+      "loss": 2.057,
+      "step": 716
+    },
+    {
+      "epoch": 0.49593636520836937,
+      "grad_norm": 0.6840877532958984,
+      "learning_rate": 5.316824471959214e-05,
+      "loss": 1.814,
+      "step": 717
+    },
+    {
+      "epoch": 0.4966280477260937,
+      "grad_norm": 0.833908200263977,
+      "learning_rate": 5.309541150764748e-05,
+      "loss": 1.9942,
+      "step": 718
+    },
+    {
+      "epoch": 0.4973197302438181,
+      "grad_norm": 0.9567387104034424,
+      "learning_rate": 5.3022578295702843e-05,
+      "loss": 2.1152,
+      "step": 719
+    },
+    {
+      "epoch": 0.4980114127615424,
+      "grad_norm": 0.7711800336837769,
+      "learning_rate": 5.29497450837582e-05,
+      "loss": 2.0442,
+      "step": 720
+    },
+    {
+      "epoch": 0.4987030952792668,
+      "grad_norm": 1.2333884239196777,
+      "learning_rate": 5.287691187181355e-05,
+      "loss": 1.5561,
+      "step": 721
+    },
+    {
+      "epoch": 0.4993947777969912,
+      "grad_norm": 0.7317585945129395,
+      "learning_rate": 5.2804078659868906e-05,
+      "loss": 1.7834,
+      "step": 722
+    },
+    {
+      "epoch": 0.5000864603147156,
+      "grad_norm": 0.8020793795585632,
+      "learning_rate": 5.273124544792426e-05,
+      "loss": 1.0252,
+      "step": 723
+    },
+    {
+      "epoch": 0.5007781428324399,
+      "grad_norm": 0.9253045916557312,
+      "learning_rate": 5.26584122359796e-05,
+      "loss": 2.0242,
+      "step": 724
+    },
+    {
+      "epoch": 0.5014698253501643,
+      "grad_norm": 0.8840431571006775,
+      "learning_rate": 5.2585579024034956e-05,
+      "loss": 1.1137,
+      "step": 725
+    },
+    {
+      "epoch": 0.5021615078678886,
+      "grad_norm": 0.9559575319290161,
+      "learning_rate": 5.251274581209032e-05,
+      "loss": 1.7501,
+      "step": 726
+    },
+    {
+      "epoch": 0.502853190385613,
+      "grad_norm": 1.340638518333435,
+      "learning_rate": 5.243991260014567e-05,
+      "loss": 1.9166,
+      "step": 727
+    },
+    {
+      "epoch": 0.5035448729033374,
+      "grad_norm": 0.8229387998580933,
+      "learning_rate": 5.2367079388201025e-05,
+      "loss": 1.9784,
+      "step": 728
+    },
+    {
+      "epoch": 0.5042365554210617,
+      "grad_norm": 0.9030795693397522,
+      "learning_rate": 5.229424617625638e-05,
+      "loss": 1.5277,
+      "step": 729
+    },
+    {
+      "epoch": 0.504928237938786,
+      "grad_norm": 1.2007991075515747,
+      "learning_rate": 5.2221412964311734e-05,
+      "loss": 1.552,
+      "step": 730
+    },
+    {
+      "epoch": 0.5056199204565105,
+      "grad_norm": 0.7516019940376282,
+      "learning_rate": 5.2148579752367075e-05,
+      "loss": 1.7162,
+      "step": 731
+    },
+    {
+      "epoch": 0.5063116029742348,
+      "grad_norm": 0.8056371212005615,
+      "learning_rate": 5.207574654042243e-05,
+      "loss": 1.8927,
+      "step": 732
+    },
+    {
+      "epoch": 0.5070032854919592,
+      "grad_norm": 0.839518666267395,
+      "learning_rate": 5.200291332847779e-05,
+      "loss": 1.828,
+      "step": 733
+    },
+    {
+      "epoch": 0.5076949680096835,
+      "grad_norm": 1.1328129768371582,
+      "learning_rate": 5.1930080116533144e-05,
+      "loss": 1.7335,
+      "step": 734
+    },
+    {
+      "epoch": 0.508386650527408,
+      "grad_norm": 0.9398472905158997,
+      "learning_rate": 5.18572469045885e-05,
+      "loss": 1.602,
+      "step": 735
+    },
+    {
+      "epoch": 0.5090783330451323,
+      "grad_norm": 0.9798534512519836,
+      "learning_rate": 5.178441369264385e-05,
+      "loss": 2.0447,
+      "step": 736
+    },
+    {
+      "epoch": 0.5097700155628566,
+      "grad_norm": 0.7347403168678284,
+      "learning_rate": 5.171158048069921e-05,
+      "loss": 1.8348,
+      "step": 737
+    },
+    {
+      "epoch": 0.5104616980805811,
+      "grad_norm": 1.5911897420883179,
+      "learning_rate": 5.163874726875455e-05,
+      "loss": 2.3152,
+      "step": 738
+    },
+    {
+      "epoch": 0.5111533805983054,
+      "grad_norm": 0.8013065457344055,
+      "learning_rate": 5.15659140568099e-05,
+      "loss": 1.6716,
+      "step": 739
+    },
+    {
+      "epoch": 0.5118450631160297,
+      "grad_norm": 0.7294222712516785,
+      "learning_rate": 5.149308084486526e-05,
+      "loss": 0.9302,
+      "step": 740
+    },
+    {
+      "epoch": 0.5125367456337541,
+      "grad_norm": 0.8843238949775696,
+      "learning_rate": 5.142024763292062e-05,
+      "loss": 2.0707,
+      "step": 741
+    },
+    {
+      "epoch": 0.5132284281514785,
+      "grad_norm": 0.8716109395027161,
+      "learning_rate": 5.134741442097597e-05,
+      "loss": 2.0515,
+      "step": 742
+    },
+    {
+      "epoch": 0.5139201106692028,
+      "grad_norm": 0.8159763813018799,
+      "learning_rate": 5.1274581209031326e-05,
+      "loss": 1.6694,
+      "step": 743
+    },
+    {
+      "epoch": 0.5146117931869272,
+      "grad_norm": 0.8231744766235352,
+      "learning_rate": 5.120174799708667e-05,
+      "loss": 1.0262,
+      "step": 744
+    },
+    {
+      "epoch": 0.5153034757046515,
+      "grad_norm": 0.8191300630569458,
+      "learning_rate": 5.112891478514202e-05,
+      "loss": 1.2072,
+      "step": 745
+    },
+    {
+      "epoch": 0.515995158222376,
+      "grad_norm": 0.775730550289154,
+      "learning_rate": 5.1056081573197375e-05,
+      "loss": 2.1546,
+      "step": 746
+    },
+    {
+      "epoch": 0.5166868407401003,
+      "grad_norm": 0.7432054877281189,
+      "learning_rate": 5.0983248361252736e-05,
+      "loss": 1.9489,
+      "step": 747
+    },
+    {
+      "epoch": 0.5173785232578246,
+      "grad_norm": 0.8312445282936096,
+      "learning_rate": 5.091041514930809e-05,
+      "loss": 1.3698,
+      "step": 748
+    },
+    {
+      "epoch": 0.5180702057755491,
+      "grad_norm": 0.631904125213623,
+      "learning_rate": 5.0837581937363445e-05,
+      "loss": 1.8463,
+      "step": 749
+    },
+    {
+      "epoch": 0.5187618882932734,
+      "grad_norm": 0.8167743682861328,
+      "learning_rate": 5.07647487254188e-05,
+      "loss": 1.5307,
+      "step": 750
+    },
+    {
+      "epoch": 0.5194535708109977,
+      "grad_norm": 0.8714820742607117,
+      "learning_rate": 5.069191551347414e-05,
+      "loss": 1.6496,
+      "step": 751
+    },
+    {
+      "epoch": 0.5201452533287221,
+      "grad_norm": 0.7490879893302917,
+      "learning_rate": 5.0619082301529494e-05,
+      "loss": 1.4393,
+      "step": 752
+    },
+    {
+      "epoch": 0.5208369358464465,
+      "grad_norm": 0.8953008651733398,
+      "learning_rate": 5.054624908958485e-05,
+      "loss": 1.3666,
+      "step": 753
+    },
+    {
+      "epoch": 0.5215286183641709,
+      "grad_norm": 0.8200691938400269,
+      "learning_rate": 5.047341587764021e-05,
+      "loss": 0.9515,
+      "step": 754
+    },
+    {
+      "epoch": 0.5222203008818952,
+      "grad_norm": 0.7947381734848022,
+      "learning_rate": 5.0400582665695564e-05,
+      "loss": 1.4832,
+      "step": 755
+    },
+    {
+      "epoch": 0.5229119833996195,
+      "grad_norm": 0.8643359541893005,
+      "learning_rate": 5.032774945375092e-05,
+      "loss": 1.7467,
+      "step": 756
+    },
+    {
+      "epoch": 0.523603665917344,
+      "grad_norm": 0.8458116054534912,
+      "learning_rate": 5.025491624180627e-05,
+      "loss": 1.789,
+      "step": 757
+    },
+    {
+      "epoch": 0.5242953484350683,
+      "grad_norm": 0.9124221205711365,
+      "learning_rate": 5.018208302986161e-05,
+      "loss": 1.892,
+      "step": 758
+    },
+    {
+      "epoch": 0.5249870309527926,
+      "grad_norm": 0.9957555532455444,
+      "learning_rate": 5.010924981791697e-05,
+      "loss": 1.4827,
+      "step": 759
+    },
+    {
+      "epoch": 0.525678713470517,
+      "grad_norm": 1.0438709259033203,
+      "learning_rate": 5.003641660597232e-05,
+      "loss": 2.1047,
+      "step": 760
+    },
+    {
+      "epoch": 0.5263703959882414,
+      "grad_norm": 0.8911094069480896,
+      "learning_rate": 4.996358339402768e-05,
+      "loss": 2.0326,
+      "step": 761
+    },
+    {
+      "epoch": 0.5270620785059658,
+      "grad_norm": 0.8831300735473633,
+      "learning_rate": 4.989075018208304e-05,
+      "loss": 1.3775,
+      "step": 762
+    },
+    {
+      "epoch": 0.5277537610236901,
+      "grad_norm": 0.6428950428962708,
+      "learning_rate": 4.9817916970138385e-05,
+      "loss": 1.1614,
+      "step": 763
+    },
+    {
+      "epoch": 0.5284454435414145,
+      "grad_norm": 0.7880016565322876,
+      "learning_rate": 4.974508375819374e-05,
+      "loss": 1.9676,
+      "step": 764
+    },
+    {
+      "epoch": 0.5291371260591389,
+      "grad_norm": 0.8804534077644348,
+      "learning_rate": 4.967225054624909e-05,
+      "loss": 1.6124,
+      "step": 765
+    },
+    {
+      "epoch": 0.5298288085768632,
+      "grad_norm": 1.2754981517791748,
+      "learning_rate": 4.959941733430444e-05,
+      "loss": 1.8053,
+      "step": 766
+    },
+    {
+      "epoch": 0.5305204910945875,
+      "grad_norm": 0.8418866395950317,
+      "learning_rate": 4.9526584122359795e-05,
+      "loss": 1.8303,
+      "step": 767
+    },
+    {
+      "epoch": 0.531212173612312,
+      "grad_norm": 0.8615307807922363,
+      "learning_rate": 4.9453750910415156e-05,
+      "loss": 1.456,
+      "step": 768
+    },
+    {
+      "epoch": 0.5319038561300363,
+      "grad_norm": 1.1805399656295776,
+      "learning_rate": 4.9380917698470504e-05,
+      "loss": 1.6149,
+      "step": 769
+    },
+    {
+      "epoch": 0.5325955386477607,
+      "grad_norm": 0.7090519666671753,
+      "learning_rate": 4.930808448652586e-05,
+      "loss": 1.297,
+      "step": 770
+    },
+    {
+      "epoch": 0.533287221165485,
+      "grad_norm": 1.0220928192138672,
+      "learning_rate": 4.923525127458121e-05,
+      "loss": 1.1299,
+      "step": 771
+    },
+    {
+      "epoch": 0.5339789036832094,
+      "grad_norm": 0.7434409856796265,
+      "learning_rate": 4.9162418062636567e-05,
+      "loss": 1.5863,
+      "step": 772
+    },
+    {
+      "epoch": 0.5346705862009338,
+      "grad_norm": 0.733049213886261,
+      "learning_rate": 4.9089584850691914e-05,
+      "loss": 1.4169,
+      "step": 773
+    },
+    {
+      "epoch": 0.5353622687186581,
+      "grad_norm": 0.8031397461891174,
+      "learning_rate": 4.901675163874727e-05,
+      "loss": 1.8983,
+      "step": 774
+    },
+    {
+      "epoch": 0.5360539512363826,
+      "grad_norm": 0.9674841165542603,
+      "learning_rate": 4.894391842680263e-05,
+      "loss": 1.8049,
+      "step": 775
+    },
+    {
+      "epoch": 0.5367456337541069,
+      "grad_norm": 0.676889181137085,
+      "learning_rate": 4.887108521485798e-05,
+      "loss": 1.3293,
+      "step": 776
+    },
+    {
+      "epoch": 0.5374373162718312,
+      "grad_norm": 1.0977071523666382,
+      "learning_rate": 4.879825200291333e-05,
+      "loss": 1.6395,
+      "step": 777
+    },
+    {
+      "epoch": 0.5381289987895556,
+      "grad_norm": 1.1417105197906494,
+      "learning_rate": 4.8725418790968686e-05,
+      "loss": 1.4429,
+      "step": 778
+    },
+    {
+      "epoch": 0.53882068130728,
+      "grad_norm": 0.7437918186187744,
+      "learning_rate": 4.865258557902403e-05,
+      "loss": 1.1679,
+      "step": 779
+    },
+    {
+      "epoch": 0.5395123638250043,
+      "grad_norm": 0.770984947681427,
+      "learning_rate": 4.857975236707939e-05,
+      "loss": 1.8322,
+      "step": 780
+    },
+    {
+      "epoch": 0.5402040463427287,
+      "grad_norm": 0.6398515701293945,
+      "learning_rate": 4.850691915513474e-05,
+      "loss": 1.3439,
+      "step": 781
+    },
+    {
+      "epoch": 0.540895728860453,
+      "grad_norm": 0.8483062982559204,
+      "learning_rate": 4.84340859431901e-05,
+      "loss": 1.7972,
+      "step": 782
+    },
+    {
+      "epoch": 0.5415874113781775,
+      "grad_norm": 1.1934760808944702,
+      "learning_rate": 4.836125273124545e-05,
+      "loss": 2.0156,
+      "step": 783
+    },
+    {
+      "epoch": 0.5422790938959018,
+      "grad_norm": 0.8203370571136475,
+      "learning_rate": 4.8288419519300804e-05,
+      "loss": 1.9463,
+      "step": 784
+    },
+    {
+      "epoch": 0.5429707764136261,
+      "grad_norm": 0.952383279800415,
+      "learning_rate": 4.821558630735616e-05,
+      "loss": 2.1658,
+      "step": 785
+    },
+    {
+      "epoch": 0.5436624589313505,
+      "grad_norm": 0.9334859251976013,
+      "learning_rate": 4.8142753095411506e-05,
+      "loss": 2.037,
+      "step": 786
+    },
+    {
+      "epoch": 0.5443541414490749,
+      "grad_norm": 0.762305736541748,
+      "learning_rate": 4.806991988346686e-05,
+      "loss": 1.4987,
+      "step": 787
+    },
+    {
+      "epoch": 0.5450458239667992,
+      "grad_norm": 0.8765081763267517,
+      "learning_rate": 4.7997086671522215e-05,
+      "loss": 1.5731,
+      "step": 788
+    },
+    {
+      "epoch": 0.5457375064845236,
+      "grad_norm": 0.8487034440040588,
+      "learning_rate": 4.792425345957757e-05,
+      "loss": 1.1847,
+      "step": 789
+    },
+    {
+      "epoch": 0.546429189002248,
+      "grad_norm": 1.0878641605377197,
+      "learning_rate": 4.7851420247632923e-05,
+      "loss": 1.6723,
+      "step": 790
+    },
+    {
+      "epoch": 0.5471208715199724,
+      "grad_norm": 1.0606861114501953,
+      "learning_rate": 4.777858703568828e-05,
+      "loss": 1.8039,
+      "step": 791
+    },
+    {
+      "epoch": 0.5478125540376967,
+      "grad_norm": 0.7105616927146912,
+      "learning_rate": 4.770575382374363e-05,
+      "loss": 1.4779,
+      "step": 792
+    },
+    {
+      "epoch": 0.548504236555421,
+      "grad_norm": 1.0841084718704224,
+      "learning_rate": 4.763292061179898e-05,
+      "loss": 1.3641,
+      "step": 793
+    },
+    {
+      "epoch": 0.5491959190731455,
+      "grad_norm": 1.6941927671432495,
+      "learning_rate": 4.7560087399854334e-05,
+      "loss": 1.3551,
+      "step": 794
+    },
+    {
+      "epoch": 0.5498876015908698,
+      "grad_norm": 0.5613378882408142,
+      "learning_rate": 4.748725418790969e-05,
+      "loss": 0.9728,
+      "step": 795
+    },
+    {
+      "epoch": 0.5505792841085941,
+      "grad_norm": 0.949262261390686,
+      "learning_rate": 4.741442097596504e-05,
+      "loss": 1.5284,
+      "step": 796
+    },
+    {
+      "epoch": 0.5512709666263185,
+      "grad_norm": 0.7887808680534363,
+      "learning_rate": 4.73415877640204e-05,
+      "loss": 2.0379,
+      "step": 797
+    },
+    {
+      "epoch": 0.5519626491440429,
+      "grad_norm": 1.312240481376648,
+      "learning_rate": 4.726875455207575e-05,
+      "loss": 1.3929,
+      "step": 798
+    },
+    {
+      "epoch": 0.5526543316617673,
+      "grad_norm": 0.9849413633346558,
+      "learning_rate": 4.71959213401311e-05,
+      "loss": 1.8282,
+      "step": 799
+    },
+    {
+      "epoch": 0.5533460141794916,
+      "grad_norm": 0.7647308111190796,
+      "learning_rate": 4.712308812818645e-05,
+      "loss": 2.0537,
+      "step": 800
+    },
+    {
+      "epoch": 0.554037696697216,
+      "grad_norm": 0.8727704882621765,
+      "learning_rate": 4.705025491624181e-05,
+      "loss": 1.9806,
+      "step": 801
+    },
+    {
+      "epoch": 0.5547293792149404,
+      "grad_norm": 0.8957968950271606,
+      "learning_rate": 4.697742170429716e-05,
+      "loss": 1.8797,
+      "step": 802
+    },
+    {
+      "epoch": 0.5554210617326647,
+      "grad_norm": 0.7387208938598633,
+      "learning_rate": 4.6904588492352516e-05,
+      "loss": 1.9765,
+      "step": 803
+    },
+    {
+      "epoch": 0.556112744250389,
+      "grad_norm": 0.7798844575881958,
+      "learning_rate": 4.683175528040787e-05,
+      "loss": 1.6741,
+      "step": 804
+    },
+    {
+      "epoch": 0.5568044267681135,
+      "grad_norm": 0.668756365776062,
+      "learning_rate": 4.6758922068463224e-05,
+      "loss": 1.4564,
+      "step": 805
+    },
+    {
+      "epoch": 0.5574961092858378,
+      "grad_norm": 0.7988947629928589,
+      "learning_rate": 4.668608885651857e-05,
+      "loss": 1.9911,
+      "step": 806
+    },
+    {
+      "epoch": 0.5581877918035621,
+      "grad_norm": 0.8549099564552307,
+      "learning_rate": 4.6613255644573926e-05,
+      "loss": 1.2702,
+      "step": 807
+    },
+    {
+      "epoch": 0.5588794743212865,
+      "grad_norm": 0.6838178634643555,
+      "learning_rate": 4.654042243262928e-05,
+      "loss": 1.438,
+      "step": 808
+    },
+    {
+      "epoch": 0.5595711568390109,
+      "grad_norm": 0.7729970812797546,
+      "learning_rate": 4.6467589220684635e-05,
+      "loss": 1.691,
+      "step": 809
+    },
+    {
+      "epoch": 0.5602628393567353,
+      "grad_norm": 0.6941683888435364,
+      "learning_rate": 4.639475600873999e-05,
+      "loss": 1.9519,
+      "step": 810
+    },
+    {
+      "epoch": 0.5609545218744596,
+      "grad_norm": 0.5844329595565796,
+      "learning_rate": 4.632192279679534e-05,
+      "loss": 1.2985,
+      "step": 811
+    },
+    {
+      "epoch": 0.5616462043921839,
+      "grad_norm": 0.8213440775871277,
+      "learning_rate": 4.62490895848507e-05,
+      "loss": 2.0151,
+      "step": 812
+    },
+    {
+      "epoch": 0.5623378869099084,
+      "grad_norm": 0.9131556749343872,
+      "learning_rate": 4.6176256372906045e-05,
+      "loss": 2.1183,
+      "step": 813
+    },
+    {
+      "epoch": 0.5630295694276327,
+      "grad_norm": 0.8296758532524109,
+      "learning_rate": 4.61034231609614e-05,
+      "loss": 1.5903,
+      "step": 814
+    },
+    {
+      "epoch": 0.563721251945357,
+      "grad_norm": 0.7798346877098083,
+      "learning_rate": 4.6030589949016754e-05,
+      "loss": 1.4633,
+      "step": 815
+    },
+    {
+      "epoch": 0.5644129344630815,
+      "grad_norm": 1.0995211601257324,
+      "learning_rate": 4.595775673707211e-05,
+      "loss": 1.5795,
+      "step": 816
+    },
+    {
+      "epoch": 0.5651046169808058,
+      "grad_norm": 0.8087084889411926,
+      "learning_rate": 4.588492352512746e-05,
+      "loss": 1.8313,
+      "step": 817
+    },
+    {
+      "epoch": 0.5657962994985302,
+      "grad_norm": 0.8411030173301697,
+      "learning_rate": 4.5812090313182816e-05,
+      "loss": 1.9952,
+      "step": 818
+    },
+    {
+      "epoch": 0.5664879820162545,
+      "grad_norm": 0.6816720962524414,
+      "learning_rate": 4.5739257101238164e-05,
+      "loss": 1.5931,
+      "step": 819
+    },
+    {
+      "epoch": 0.5671796645339789,
+      "grad_norm": 0.8486337661743164,
+      "learning_rate": 4.566642388929352e-05,
+      "loss": 1.7394,
+      "step": 820
+    },
+    {
+      "epoch": 0.5678713470517033,
+      "grad_norm": 0.863264262676239,
+      "learning_rate": 4.559359067734887e-05,
+      "loss": 1.8192,
+      "step": 821
+    },
+    {
+      "epoch": 0.5685630295694276,
+      "grad_norm": 0.5527456402778625,
+      "learning_rate": 4.552075746540423e-05,
+      "loss": 1.1761,
+      "step": 822
+    },
+    {
+      "epoch": 0.569254712087152,
+      "grad_norm": 0.894025981426239,
+      "learning_rate": 4.544792425345958e-05,
+      "loss": 2.0718,
+      "step": 823
+    },
+    {
+      "epoch": 0.5699463946048764,
+      "grad_norm": 0.9638246893882751,
+      "learning_rate": 4.5375091041514935e-05,
+      "loss": 2.0523,
+      "step": 824
+    },
+    {
+      "epoch": 0.5706380771226007,
+      "grad_norm": 0.6295531988143921,
+      "learning_rate": 4.530225782957029e-05,
+      "loss": 1.1311,
+      "step": 825
+    },
+    {
+      "epoch": 0.5713297596403251,
+      "grad_norm": 0.8926659822463989,
+      "learning_rate": 4.522942461762564e-05,
+      "loss": 1.8242,
+      "step": 826
+    },
+    {
+      "epoch": 0.5720214421580495,
+      "grad_norm": 0.9920228719711304,
+      "learning_rate": 4.515659140568099e-05,
+      "loss": 1.4683,
+      "step": 827
+    },
+    {
+      "epoch": 0.5727131246757738,
+      "grad_norm": 0.9386698007583618,
+      "learning_rate": 4.5083758193736346e-05,
+      "loss": 1.7646,
+      "step": 828
+    },
+    {
+      "epoch": 0.5734048071934982,
+      "grad_norm": 0.8631352186203003,
+      "learning_rate": 4.50109249817917e-05,
+      "loss": 1.5608,
+      "step": 829
+    },
+    {
+      "epoch": 0.5740964897112225,
+      "grad_norm": 0.8916756510734558,
+      "learning_rate": 4.4938091769847054e-05,
+      "loss": 1.5689,
+      "step": 830
+    },
+    {
+      "epoch": 0.574788172228947,
+      "grad_norm": 0.7972738146781921,
+      "learning_rate": 4.486525855790241e-05,
+      "loss": 1.4542,
+      "step": 831
+    },
+    {
+      "epoch": 0.5754798547466713,
+      "grad_norm": 0.8566955924034119,
+      "learning_rate": 4.479242534595776e-05,
+      "loss": 1.5176,
+      "step": 832
+    },
+    {
+      "epoch": 0.5761715372643956,
+      "grad_norm": 1.0372542142868042,
+      "learning_rate": 4.471959213401311e-05,
+      "loss": 1.4787,
+      "step": 833
+    },
+    {
+      "epoch": 0.57686321978212,
+      "grad_norm": 0.7778179049491882,
+      "learning_rate": 4.4646758922068465e-05,
+      "loss": 1.5001,
+      "step": 834
+    },
+    {
+      "epoch": 0.5775549022998444,
+      "grad_norm": 1.0114778280258179,
+      "learning_rate": 4.457392571012382e-05,
+      "loss": 1.7393,
+      "step": 835
+    },
+    {
+      "epoch": 0.5782465848175687,
+      "grad_norm": 0.8515650629997253,
+      "learning_rate": 4.450109249817917e-05,
+      "loss": 1.8581,
+      "step": 836
+    },
+    {
+      "epoch": 0.5789382673352931,
+      "grad_norm": 1.318204641342163,
+      "learning_rate": 4.442825928623453e-05,
+      "loss": 1.5188,
+      "step": 837
+    },
+    {
+      "epoch": 0.5796299498530174,
+      "grad_norm": 1.0010000467300415,
+      "learning_rate": 4.435542607428988e-05,
+      "loss": 1.6703,
+      "step": 838
+    },
+    {
+      "epoch": 0.5803216323707419,
+      "grad_norm": 0.7508212327957153,
+      "learning_rate": 4.428259286234523e-05,
+      "loss": 1.8278,
+      "step": 839
+    },
+    {
+      "epoch": 0.5810133148884662,
+      "grad_norm": 0.7435583472251892,
+      "learning_rate": 4.4209759650400584e-05,
+      "loss": 1.5339,
+      "step": 840
+    },
+    {
+      "epoch": 0.5817049974061905,
+      "grad_norm": 0.6677370667457581,
+      "learning_rate": 4.413692643845594e-05,
+      "loss": 1.7071,
+      "step": 841
+    },
+    {
+      "epoch": 0.582396679923915,
+      "grad_norm": 0.6046009659767151,
+      "learning_rate": 4.406409322651129e-05,
+      "loss": 1.8311,
+      "step": 842
+    },
+    {
+      "epoch": 0.5830883624416393,
+      "grad_norm": 0.8315126895904541,
+      "learning_rate": 4.3991260014566647e-05,
+      "loss": 1.9365,
+      "step": 843
+    },
+    {
+      "epoch": 0.5837800449593636,
+      "grad_norm": 0.9006065726280212,
+      "learning_rate": 4.3918426802622e-05,
+      "loss": 1.4826,
+      "step": 844
+    },
+    {
+      "epoch": 0.584471727477088,
+      "grad_norm": 1.6106517314910889,
+      "learning_rate": 4.3845593590677355e-05,
+      "loss": 1.3065,
+      "step": 845
+    },
+    {
+      "epoch": 0.5851634099948124,
+      "grad_norm": 0.6510477662086487,
+      "learning_rate": 4.37727603787327e-05,
+      "loss": 1.291,
+      "step": 846
+    },
+    {
+      "epoch": 0.5858550925125368,
+      "grad_norm": 1.2163658142089844,
+      "learning_rate": 4.369992716678806e-05,
+      "loss": 1.65,
+      "step": 847
+    },
+    {
+      "epoch": 0.5865467750302611,
+      "grad_norm": 0.7138450145721436,
+      "learning_rate": 4.362709395484341e-05,
+      "loss": 0.9941,
+      "step": 848
+    },
+    {
+      "epoch": 0.5872384575479854,
+      "grad_norm": 1.4511529207229614,
+      "learning_rate": 4.355426074289876e-05,
+      "loss": 1.1884,
+      "step": 849
+    },
+    {
+      "epoch": 0.5879301400657099,
+      "grad_norm": 0.806222677230835,
+      "learning_rate": 4.348142753095411e-05,
+      "loss": 1.5618,
+      "step": 850
+    },
+    {
+      "epoch": 0.5886218225834342,
+      "grad_norm": 0.9857581257820129,
+      "learning_rate": 4.3408594319009474e-05,
+      "loss": 1.1434,
+      "step": 851
+    },
+    {
+      "epoch": 0.5893135051011585,
+      "grad_norm": 0.8316003084182739,
+      "learning_rate": 4.333576110706483e-05,
+      "loss": 1.9074,
+      "step": 852
+    },
+    {
+      "epoch": 0.590005187618883,
+      "grad_norm": 0.6793386936187744,
+      "learning_rate": 4.3262927895120176e-05,
+      "loss": 2.0039,
+      "step": 853
+    },
+    {
+      "epoch": 0.5906968701366073,
+      "grad_norm": 0.7505836486816406,
+      "learning_rate": 4.319009468317553e-05,
+      "loss": 1.6837,
+      "step": 854
+    },
+    {
+      "epoch": 0.5913885526543317,
+      "grad_norm": 1.0526964664459229,
+      "learning_rate": 4.3117261471230884e-05,
+      "loss": 2.0755,
+      "step": 855
+    },
+    {
+      "epoch": 0.592080235172056,
+      "grad_norm": 0.9285861849784851,
+      "learning_rate": 4.304442825928623e-05,
+      "loss": 2.0415,
+      "step": 856
+    },
+    {
+      "epoch": 0.5927719176897804,
+      "grad_norm": 0.7729842066764832,
+      "learning_rate": 4.2971595047341586e-05,
+      "loss": 1.7624,
+      "step": 857
+    },
+    {
+      "epoch": 0.5934636002075048,
+      "grad_norm": 0.857053279876709,
+      "learning_rate": 4.289876183539695e-05,
+      "loss": 1.7886,
+      "step": 858
+    },
+    {
+      "epoch": 0.5941552827252291,
+      "grad_norm": 1.1205097436904907,
+      "learning_rate": 4.2825928623452295e-05,
+      "loss": 1.1998,
+      "step": 859
+    },
+    {
+      "epoch": 0.5948469652429534,
+      "grad_norm": 0.8811049461364746,
+      "learning_rate": 4.275309541150765e-05,
+      "loss": 1.7367,
+      "step": 860
+    },
+    {
+      "epoch": 0.5955386477606779,
+      "grad_norm": 0.7999034523963928,
+      "learning_rate": 4.2680262199563003e-05,
+      "loss": 1.6502,
+      "step": 861
+    },
+    {
+      "epoch": 0.5962303302784022,
+      "grad_norm": 0.9138179421424866,
+      "learning_rate": 4.260742898761836e-05,
+      "loss": 1.7061,
+      "step": 862
+    },
+    {
+      "epoch": 0.5969220127961266,
+      "grad_norm": 0.943294882774353,
+      "learning_rate": 4.2534595775673705e-05,
+      "loss": 1.1506,
+      "step": 863
+    },
+    {
+      "epoch": 0.5976136953138509,
+      "grad_norm": 0.9944688677787781,
+      "learning_rate": 4.246176256372906e-05,
+      "loss": 2.025,
+      "step": 864
+    },
+    {
+      "epoch": 0.5983053778315753,
+      "grad_norm": 0.9417821764945984,
+      "learning_rate": 4.238892935178442e-05,
+      "loss": 2.0722,
+      "step": 865
+    },
+    {
+      "epoch": 0.5989970603492997,
+      "grad_norm": 1.3148717880249023,
+      "learning_rate": 4.231609613983977e-05,
+      "loss": 1.8245,
+      "step": 866
+    },
+    {
+      "epoch": 0.599688742867024,
+      "grad_norm": 0.8092090487480164,
+      "learning_rate": 4.224326292789512e-05,
+      "loss": 1.9595,
+      "step": 867
+    },
+    {
+      "epoch": 0.6003804253847485,
+      "grad_norm": 0.7386168837547302,
+      "learning_rate": 4.217042971595048e-05,
+      "loss": 1.4617,
+      "step": 868
+    },
+    {
+      "epoch": 0.6010721079024728,
+      "grad_norm": 0.7619878649711609,
+      "learning_rate": 4.2097596504005824e-05,
+      "loss": 1.3067,
+      "step": 869
+    },
+    {
+      "epoch": 0.6017637904201971,
+      "grad_norm": 0.8451841473579407,
+      "learning_rate": 4.202476329206118e-05,
+      "loss": 2.0969,
+      "step": 870
+    },
+    {
+      "epoch": 0.6024554729379215,
+      "grad_norm": 0.7799137830734253,
+      "learning_rate": 4.195193008011653e-05,
+      "loss": 1.9536,
+      "step": 871
+    },
+    {
+      "epoch": 0.6031471554556459,
+      "grad_norm": 0.6715651154518127,
+      "learning_rate": 4.1879096868171894e-05,
+      "loss": 1.1611,
+      "step": 872
+    },
+    {
+      "epoch": 0.6038388379733702,
+      "grad_norm": 0.8742201328277588,
+      "learning_rate": 4.180626365622724e-05,
+      "loss": 1.7178,
+      "step": 873
+    },
+    {
+      "epoch": 0.6045305204910946,
+      "grad_norm": 0.9098167419433594,
+      "learning_rate": 4.1733430444282596e-05,
+      "loss": 1.4663,
+      "step": 874
+    },
+    {
+      "epoch": 0.6052222030088189,
+      "grad_norm": 1.1566745042800903,
+      "learning_rate": 4.166059723233795e-05,
+      "loss": 1.5812,
+      "step": 875
+    },
+    {
+      "epoch": 0.6059138855265433,
+      "grad_norm": 0.7112817764282227,
+      "learning_rate": 4.15877640203933e-05,
+      "loss": 1.4423,
+      "step": 876
+    },
+    {
+      "epoch": 0.6066055680442677,
+      "grad_norm": 0.773082971572876,
+      "learning_rate": 4.151493080844865e-05,
+      "loss": 1.2592,
+      "step": 877
+    },
+    {
+      "epoch": 0.607297250561992,
+      "grad_norm": 0.8665560483932495,
+      "learning_rate": 4.1442097596504006e-05,
+      "loss": 1.5569,
+      "step": 878
+    },
+    {
+      "epoch": 0.6079889330797165,
+      "grad_norm": 0.7002996206283569,
+      "learning_rate": 4.136926438455936e-05,
+      "loss": 1.7474,
+      "step": 879
+    },
+    {
+      "epoch": 0.6086806155974408,
+      "grad_norm": 0.780541718006134,
+      "learning_rate": 4.1296431172614715e-05,
+      "loss": 1.8956,
+      "step": 880
+    },
+    {
+      "epoch": 0.6093722981151651,
+      "grad_norm": 0.9581499695777893,
+      "learning_rate": 4.122359796067007e-05,
+      "loss": 1.5679,
+      "step": 881
+    },
+    {
+      "epoch": 0.6100639806328895,
+      "grad_norm": 1.0939525365829468,
+      "learning_rate": 4.115076474872542e-05,
+      "loss": 1.8858,
+      "step": 882
+    },
+    {
+      "epoch": 0.6107556631506139,
+      "grad_norm": 0.7007079720497131,
+      "learning_rate": 4.107793153678077e-05,
+      "loss": 1.4574,
+      "step": 883
+    },
+    {
+      "epoch": 0.6114473456683382,
+      "grad_norm": 0.7020848989486694,
+      "learning_rate": 4.1005098324836125e-05,
+      "loss": 1.0492,
+      "step": 884
+    },
+    {
+      "epoch": 0.6121390281860626,
+      "grad_norm": 0.7588708996772766,
+      "learning_rate": 4.093226511289148e-05,
+      "loss": 1.5429,
+      "step": 885
+    },
+    {
+      "epoch": 0.6128307107037869,
+      "grad_norm": 0.841346025466919,
+      "learning_rate": 4.0859431900946834e-05,
+      "loss": 1.3996,
+      "step": 886
+    },
+    {
+      "epoch": 0.6135223932215114,
+      "grad_norm": 1.0934274196624756,
+      "learning_rate": 4.078659868900219e-05,
+      "loss": 1.7602,
+      "step": 887
+    },
+    {
+      "epoch": 0.6142140757392357,
+      "grad_norm": 0.7109699249267578,
+      "learning_rate": 4.071376547705754e-05,
+      "loss": 1.6866,
+      "step": 888
+    },
+    {
+      "epoch": 0.61490575825696,
+      "grad_norm": 0.9574000835418701,
+      "learning_rate": 4.064093226511289e-05,
+      "loss": 1.389,
+      "step": 889
+    },
+    {
+      "epoch": 0.6155974407746844,
+      "grad_norm": 0.6910289525985718,
+      "learning_rate": 4.0568099053168244e-05,
+      "loss": 1.4255,
+      "step": 890
+    },
+    {
+      "epoch": 0.6162891232924088,
+      "grad_norm": 1.038681149482727,
+      "learning_rate": 4.04952658412236e-05,
+      "loss": 1.8823,
+      "step": 891
+    },
+    {
+      "epoch": 0.6169808058101331,
+      "grad_norm": 0.7626045942306519,
+      "learning_rate": 4.042243262927895e-05,
+      "loss": 1.682,
+      "step": 892
+    },
+    {
+      "epoch": 0.6176724883278575,
+      "grad_norm": 0.8613514304161072,
+      "learning_rate": 4.034959941733431e-05,
+      "loss": 0.9454,
+      "step": 893
+    },
+    {
+      "epoch": 0.6183641708455819,
+      "grad_norm": 0.7835326790809631,
+      "learning_rate": 4.027676620538966e-05,
+      "loss": 1.8522,
+      "step": 894
+    },
+    {
+      "epoch": 0.6190558533633063,
+      "grad_norm": 1.4106545448303223,
+      "learning_rate": 4.0203932993445015e-05,
+      "loss": 1.4148,
+      "step": 895
+    },
+    {
+      "epoch": 0.6197475358810306,
+      "grad_norm": 0.7465313673019409,
+      "learning_rate": 4.013109978150036e-05,
+      "loss": 1.4045,
+      "step": 896
+    },
+    {
+      "epoch": 0.6204392183987549,
+      "grad_norm": 0.7496219873428345,
+      "learning_rate": 4.005826656955572e-05,
+      "loss": 2.0982,
+      "step": 897
+    },
+    {
+      "epoch": 0.6211309009164794,
+      "grad_norm": 0.8977241516113281,
+      "learning_rate": 3.998543335761107e-05,
+      "loss": 1.1532,
+      "step": 898
+    },
+    {
+      "epoch": 0.6218225834342037,
+      "grad_norm": 0.787335991859436,
+      "learning_rate": 3.9912600145666426e-05,
+      "loss": 2.1213,
+      "step": 899
+    },
+    {
+      "epoch": 0.622514265951928,
+      "grad_norm": 1.1078623533248901,
+      "learning_rate": 3.983976693372178e-05,
+      "loss": 1.5371,
+      "step": 900
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1446,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.855213829690163e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}