diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12327 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 15.0,
+  "eval_steps": 500,
+  "global_step": 1755,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008547008547008548,
+      "grad_norm": 32.62020740979074,
+      "learning_rate": 4.999995994512315e-06,
+      "loss": 4.276,
+      "step": 1
+    },
+    {
+      "epoch": 0.017094017094017096,
+      "grad_norm": 29.072421384539055,
+      "learning_rate": 4.999983978062096e-06,
+      "loss": 2.8774,
+      "step": 2
+    },
+    {
+      "epoch": 0.02564102564102564,
+      "grad_norm": 30.135778249629343,
+      "learning_rate": 4.999963950687846e-06,
+      "loss": 3.7408,
+      "step": 3
+    },
+    {
+      "epoch": 0.03418803418803419,
+      "grad_norm": 24.189256487874317,
+      "learning_rate": 4.999935912453743e-06,
+      "loss": 3.1117,
+      "step": 4
+    },
+    {
+      "epoch": 0.042735042735042736,
+      "grad_norm": 18.836727485689817,
+      "learning_rate": 4.999899863449631e-06,
+      "loss": 3.1051,
+      "step": 5
+    },
+    {
+      "epoch": 0.05128205128205128,
+      "grad_norm": 12.501967391799338,
+      "learning_rate": 4.999855803791026e-06,
+      "loss": 2.6904,
+      "step": 6
+    },
+    {
+      "epoch": 0.05982905982905983,
+      "grad_norm": 9.637791567564221,
+      "learning_rate": 4.9998037336191115e-06,
+      "loss": 1.8663,
+      "step": 7
+    },
+    {
+      "epoch": 0.06837606837606838,
+      "grad_norm": 4.939579902739148,
+      "learning_rate": 4.9997436531007415e-06,
+      "loss": 1.7693,
+      "step": 8
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 2.4003696989542815,
+      "learning_rate": 4.999675562428437e-06,
+      "loss": 1.1684,
+      "step": 9
+    },
+    {
+      "epoch": 0.08547008547008547,
+      "grad_norm": 5.971593419685188,
+      "learning_rate": 4.999599461820387e-06,
+      "loss": 2.2486,
+      "step": 10
+    },
+    {
+      "epoch": 0.09401709401709402,
+      "grad_norm": 3.977027000496355,
+      "learning_rate": 4.999515351520447e-06,
+      "loss": 1.7208,
+      "step": 11
+    },
+    {
+      "epoch": 0.10256410256410256,
+      "grad_norm": 2.783958221119982,
+      "learning_rate": 4.9994232317981405e-06,
+      "loss": 1.6043,
+      "step": 12
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 4.538183754473571,
+      "learning_rate": 4.999323102948655e-06,
+      "loss": 1.9126,
+      "step": 13
+    },
+    {
+      "epoch": 0.11965811965811966,
+      "grad_norm": 2.8220664880830966,
+      "learning_rate": 4.999214965292841e-06,
+      "loss": 1.5208,
+      "step": 14
+    },
+    {
+      "epoch": 0.1282051282051282,
+      "grad_norm": 2.9402595332539736,
+      "learning_rate": 4.999098819177214e-06,
+      "loss": 1.0275,
+      "step": 15
+    },
+    {
+      "epoch": 0.13675213675213677,
+      "grad_norm": 36.901371585877804,
+      "learning_rate": 4.998974664973953e-06,
+      "loss": 1.081,
+      "step": 16
+    },
+    {
+      "epoch": 0.1452991452991453,
+      "grad_norm": 2.9647572686177552,
+      "learning_rate": 4.998842503080894e-06,
+      "loss": 1.6457,
+      "step": 17
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 2.1277618530844227,
+      "learning_rate": 4.998702333921538e-06,
+      "loss": 1.0708,
+      "step": 18
+    },
+    {
+      "epoch": 0.1623931623931624,
+      "grad_norm": 2.8870619886137434,
+      "learning_rate": 4.99855415794504e-06,
+      "loss": 1.6499,
+      "step": 19
+    },
+    {
+      "epoch": 0.17094017094017094,
+      "grad_norm": 2.0120452229027626,
+      "learning_rate": 4.998397975626213e-06,
+      "loss": 1.5059,
+      "step": 20
+    },
+    {
+      "epoch": 0.1794871794871795,
+      "grad_norm": 1.9920958843151744,
+      "learning_rate": 4.998233787465529e-06,
+      "loss": 1.0625,
+      "step": 21
+    },
+    {
+      "epoch": 0.18803418803418803,
+      "grad_norm": 2.260535248606528,
+      "learning_rate": 4.998061593989108e-06,
+      "loss": 1.3096,
+      "step": 22
+    },
+    {
+      "epoch": 0.19658119658119658,
+      "grad_norm": 1.8563891675327266,
+      "learning_rate": 4.997881395748727e-06,
+      "loss": 1.1027,
+      "step": 23
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 1.7562642717959271,
+      "learning_rate": 4.99769319332181e-06,
+      "loss": 1.1636,
+      "step": 24
+    },
+    {
+      "epoch": 0.21367521367521367,
+      "grad_norm": 1.809288311360476,
+      "learning_rate": 4.997496987311431e-06,
+      "loss": 1.3529,
+      "step": 25
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 2.8497055141007057,
+      "learning_rate": 4.997292778346312e-06,
+      "loss": 1.2778,
+      "step": 26
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 1.3305413686651983,
+      "learning_rate": 4.9970805670808174e-06,
+      "loss": 1.0894,
+      "step": 27
+    },
+    {
+      "epoch": 0.23931623931623933,
+      "grad_norm": 2.926773487016974,
+      "learning_rate": 4.996860354194954e-06,
+      "loss": 1.7158,
+      "step": 28
+    },
+    {
+      "epoch": 0.24786324786324787,
+      "grad_norm": 2.1198500878749846,
+      "learning_rate": 4.996632140394372e-06,
+      "loss": 1.5559,
+      "step": 29
+    },
+    {
+      "epoch": 0.2564102564102564,
+      "grad_norm": 1.184358048001977,
+      "learning_rate": 4.996395926410354e-06,
+      "loss": 0.9109,
+      "step": 30
+    },
+    {
+      "epoch": 0.26495726495726496,
+      "grad_norm": 1.646041152309647,
+      "learning_rate": 4.996151712999826e-06,
+      "loss": 1.1522,
+      "step": 31
+    },
+    {
+      "epoch": 0.27350427350427353,
+      "grad_norm": 1.368219183766303,
+      "learning_rate": 4.995899500945341e-06,
+      "loss": 0.8634,
+      "step": 32
+    },
+    {
+      "epoch": 0.28205128205128205,
+      "grad_norm": 1.577497651144209,
+      "learning_rate": 4.995639291055084e-06,
+      "loss": 1.1489,
+      "step": 33
+    },
+    {
+      "epoch": 0.2905982905982906,
+      "grad_norm": 1.691998808976061,
+      "learning_rate": 4.99537108416287e-06,
+      "loss": 1.2498,
+      "step": 34
+    },
+    {
+      "epoch": 0.29914529914529914,
+      "grad_norm": 1.4882087466641996,
+      "learning_rate": 4.995094881128138e-06,
+      "loss": 1.0567,
+      "step": 35
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.0271930703785612,
+      "learning_rate": 4.994810682835951e-06,
+      "loss": 0.8448,
+      "step": 36
+    },
+    {
+      "epoch": 0.3162393162393162,
+      "grad_norm": 1.1199239553430391,
+      "learning_rate": 4.99451849019699e-06,
+      "loss": 0.7009,
+      "step": 37
+    },
+    {
+      "epoch": 0.3247863247863248,
+      "grad_norm": 1.669850515052373,
+      "learning_rate": 4.994218304147556e-06,
+      "loss": 1.0606,
+      "step": 38
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.5674947016471914,
+      "learning_rate": 4.993910125649561e-06,
+      "loss": 1.2569,
+      "step": 39
+    },
+    {
+      "epoch": 0.3418803418803419,
+      "grad_norm": 1.0864716234200489,
+      "learning_rate": 4.993593955690529e-06,
+      "loss": 0.7918,
+      "step": 40
+    },
+    {
+      "epoch": 0.3504273504273504,
+      "grad_norm": 1.5801532642600238,
+      "learning_rate": 4.9932697952835925e-06,
+      "loss": 1.1788,
+      "step": 41
+    },
+    {
+      "epoch": 0.358974358974359,
+      "grad_norm": 1.5026314328402226,
+      "learning_rate": 4.992937645467489e-06,
+      "loss": 1.174,
+      "step": 42
+    },
+    {
+      "epoch": 0.36752136752136755,
+      "grad_norm": 1.5519979381594173,
+      "learning_rate": 4.992597507306552e-06,
+      "loss": 1.3143,
+      "step": 43
+    },
+    {
+      "epoch": 0.37606837606837606,
+      "grad_norm": 1.3591841494684331,
+      "learning_rate": 4.992249381890722e-06,
+      "loss": 1.1037,
+      "step": 44
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.4384122655785836,
+      "learning_rate": 4.991893270335526e-06,
+      "loss": 1.2459,
+      "step": 45
+    },
+    {
+      "epoch": 0.39316239316239315,
+      "grad_norm": 1.3895569778660726,
+      "learning_rate": 4.9915291737820836e-06,
+      "loss": 0.9972,
+      "step": 46
+    },
+    {
+      "epoch": 0.4017094017094017,
+      "grad_norm": 1.4589079891461676,
+      "learning_rate": 4.991157093397104e-06,
+      "loss": 1.0955,
+      "step": 47
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 1.077697915403046,
+      "learning_rate": 4.990777030372877e-06,
+      "loss": 0.8011,
+      "step": 48
+    },
+    {
+      "epoch": 0.4188034188034188,
+      "grad_norm": 1.2910938732132287,
+      "learning_rate": 4.990388985927273e-06,
+      "loss": 1.0665,
+      "step": 49
+    },
+    {
+      "epoch": 0.42735042735042733,
+      "grad_norm": 1.2714720712304037,
+      "learning_rate": 4.989992961303738e-06,
+      "loss": 1.0145,
+      "step": 50
+    },
+    {
+      "epoch": 0.4358974358974359,
+      "grad_norm": 1.2845842411221682,
+      "learning_rate": 4.989588957771289e-06,
+      "loss": 0.9378,
+      "step": 51
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.3419174977624397,
+      "learning_rate": 4.989176976624511e-06,
+      "loss": 0.8341,
+      "step": 52
+    },
+    {
+      "epoch": 0.452991452991453,
+      "grad_norm": 1.0231416966065818,
+      "learning_rate": 4.988757019183553e-06,
+      "loss": 0.7548,
+      "step": 53
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.1252549457024466,
+      "learning_rate": 4.988329086794122e-06,
+      "loss": 0.8745,
+      "step": 54
+    },
+    {
+      "epoch": 0.4700854700854701,
+      "grad_norm": 0.9536636897289945,
+      "learning_rate": 4.9878931808274796e-06,
+      "loss": 0.6516,
+      "step": 55
+    },
+    {
+      "epoch": 0.47863247863247865,
+      "grad_norm": 1.6209383242246733,
+      "learning_rate": 4.98744930268044e-06,
+      "loss": 0.9706,
+      "step": 56
+    },
+    {
+      "epoch": 0.48717948717948717,
+      "grad_norm": 1.0731934865207635,
+      "learning_rate": 4.986997453775361e-06,
+      "loss": 0.8936,
+      "step": 57
+    },
+    {
+      "epoch": 0.49572649572649574,
+      "grad_norm": 1.362773877359389,
+      "learning_rate": 4.986537635560144e-06,
+      "loss": 0.8172,
+      "step": 58
+    },
+    {
+      "epoch": 0.5042735042735043,
+      "grad_norm": 4.543126580901282,
+      "learning_rate": 4.986069849508223e-06,
+      "loss": 1.0017,
+      "step": 59
+    },
+    {
+      "epoch": 0.5128205128205128,
+      "grad_norm": 1.047286807290104,
+      "learning_rate": 4.9855940971185705e-06,
+      "loss": 0.9148,
+      "step": 60
+    },
+    {
+      "epoch": 0.5213675213675214,
+      "grad_norm": 0.9239545215151804,
+      "learning_rate": 4.985110379915681e-06,
+      "loss": 0.803,
+      "step": 61
+    },
+    {
+      "epoch": 0.5299145299145299,
+      "grad_norm": 1.014258532084436,
+      "learning_rate": 4.984618699449573e-06,
+      "loss": 0.8317,
+      "step": 62
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.1970397088754383,
+      "learning_rate": 4.984119057295783e-06,
+      "loss": 0.8198,
+      "step": 63
+    },
+    {
+      "epoch": 0.5470085470085471,
+      "grad_norm": 1.8467307385552865,
+      "learning_rate": 4.983611455055359e-06,
+      "loss": 0.7764,
+      "step": 64
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 1.0137011006880812,
+      "learning_rate": 4.983095894354858e-06,
+      "loss": 0.8778,
+      "step": 65
+    },
+    {
+      "epoch": 0.5641025641025641,
+      "grad_norm": 1.1254196981388234,
+      "learning_rate": 4.982572376846336e-06,
+      "loss": 0.7872,
+      "step": 66
+    },
+    {
+      "epoch": 0.5726495726495726,
+      "grad_norm": 1.0458612262024898,
+      "learning_rate": 4.982040904207348e-06,
+      "loss": 1.03,
+      "step": 67
+    },
+    {
+      "epoch": 0.5811965811965812,
+      "grad_norm": 1.0425261033892632,
+      "learning_rate": 4.98150147814094e-06,
+      "loss": 0.9989,
+      "step": 68
+    },
+    {
+      "epoch": 0.5897435897435898,
+      "grad_norm": 0.8940316221494978,
+      "learning_rate": 4.980954100375642e-06,
+      "loss": 0.6748,
+      "step": 69
+    },
+    {
+      "epoch": 0.5982905982905983,
+      "grad_norm": 1.1615256423288627,
+      "learning_rate": 4.980398772665468e-06,
+      "loss": 0.7753,
+      "step": 70
+    },
+    {
+      "epoch": 0.6068376068376068,
+      "grad_norm": 0.9363884553573214,
+      "learning_rate": 4.979835496789904e-06,
+      "loss": 0.8255,
+      "step": 71
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 1.0064167314698989,
+      "learning_rate": 4.979264274553906e-06,
+      "loss": 1.0582,
+      "step": 72
+    },
+    {
+      "epoch": 0.6239316239316239,
+      "grad_norm": 1.106482001780077,
+      "learning_rate": 4.97868510778789e-06,
+      "loss": 0.7719,
+      "step": 73
+    },
+    {
+      "epoch": 0.6324786324786325,
+      "grad_norm": 0.9957808978272576,
+      "learning_rate": 4.978097998347737e-06,
+      "loss": 0.8936,
+      "step": 74
+    },
+    {
+      "epoch": 0.6410256410256411,
+      "grad_norm": 0.9138226587744296,
+      "learning_rate": 4.977502948114772e-06,
+      "loss": 0.7896,
+      "step": 75
+    },
+    {
+      "epoch": 0.6495726495726496,
+      "grad_norm": 0.9641099744456224,
+      "learning_rate": 4.9768999589957675e-06,
+      "loss": 0.7238,
+      "step": 76
+    },
+    {
+      "epoch": 0.6581196581196581,
+      "grad_norm": 0.965423829764785,
+      "learning_rate": 4.976289032922937e-06,
+      "loss": 0.9367,
+      "step": 77
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.9153110347105345,
+      "learning_rate": 4.975670171853926e-06,
+      "loss": 0.9802,
+      "step": 78
+    },
+    {
+      "epoch": 0.6752136752136753,
+      "grad_norm": 1.0031899781143425,
+      "learning_rate": 4.975043377771806e-06,
+      "loss": 0.7416,
+      "step": 79
+    },
+    {
+      "epoch": 0.6837606837606838,
+      "grad_norm": 0.9678094486698757,
+      "learning_rate": 4.9744086526850724e-06,
+      "loss": 0.8766,
+      "step": 80
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 1.0103884039407407,
+      "learning_rate": 4.973765998627628e-06,
+      "loss": 1.0188,
+      "step": 81
+    },
+    {
+      "epoch": 0.7008547008547008,
+      "grad_norm": 1.0510378757673386,
+      "learning_rate": 4.97311541765879e-06,
+      "loss": 0.9294,
+      "step": 82
+    },
+    {
+      "epoch": 0.7094017094017094,
+      "grad_norm": 1.0911046658822772,
+      "learning_rate": 4.972456911863273e-06,
+      "loss": 0.9152,
+      "step": 83
+    },
+    {
+      "epoch": 0.717948717948718,
+      "grad_norm": 1.0560120280283998,
+      "learning_rate": 4.971790483351186e-06,
+      "loss": 0.7679,
+      "step": 84
+    },
+    {
+      "epoch": 0.7264957264957265,
+      "grad_norm": 0.9437311405916173,
+      "learning_rate": 4.971116134258026e-06,
+      "loss": 0.7396,
+      "step": 85
+    },
+    {
+      "epoch": 0.7350427350427351,
+      "grad_norm": 0.9132375746546699,
+      "learning_rate": 4.97043386674467e-06,
+      "loss": 0.9389,
+      "step": 86
+    },
+    {
+      "epoch": 0.7435897435897436,
+      "grad_norm": 1.081470401037969,
+      "learning_rate": 4.969743682997372e-06,
+      "loss": 0.7307,
+      "step": 87
+    },
+    {
+      "epoch": 0.7521367521367521,
+      "grad_norm": 0.7993087792851001,
+      "learning_rate": 4.969045585227747e-06,
+      "loss": 0.7206,
+      "step": 88
+    },
+    {
+      "epoch": 0.7606837606837606,
+      "grad_norm": 1.0592462938558866,
+      "learning_rate": 4.968339575672773e-06,
+      "loss": 0.8962,
+      "step": 89
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.1781322670806287,
+      "learning_rate": 4.967625656594782e-06,
+      "loss": 0.8418,
+      "step": 90
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.8851002604634136,
+      "learning_rate": 4.966903830281449e-06,
+      "loss": 0.7403,
+      "step": 91
+    },
+    {
+      "epoch": 0.7863247863247863,
+      "grad_norm": 1.1731634314905843,
+      "learning_rate": 4.966174099045784e-06,
+      "loss": 0.8497,
+      "step": 92
+    },
+    {
+      "epoch": 0.7948717948717948,
+      "grad_norm": 0.9637038553583984,
+      "learning_rate": 4.9654364652261345e-06,
+      "loss": 0.8345,
+      "step": 93
+    },
+    {
+      "epoch": 0.8034188034188035,
+      "grad_norm": 0.8910531904379403,
+      "learning_rate": 4.964690931186165e-06,
+      "loss": 0.8016,
+      "step": 94
+    },
+    {
+      "epoch": 0.811965811965812,
+      "grad_norm": 1.0055847088527494,
+      "learning_rate": 4.963937499314857e-06,
+      "loss": 1.0533,
+      "step": 95
+    },
+    {
+      "epoch": 0.8205128205128205,
+      "grad_norm": 1.0675718725688317,
+      "learning_rate": 4.963176172026501e-06,
+      "loss": 0.9226,
+      "step": 96
+    },
+    {
+      "epoch": 0.8290598290598291,
+      "grad_norm": 0.9609720971979216,
+      "learning_rate": 4.962406951760687e-06,
+      "loss": 0.7381,
+      "step": 97
+    },
+    {
+      "epoch": 0.8376068376068376,
+      "grad_norm": 0.9128632849930912,
+      "learning_rate": 4.961629840982296e-06,
+      "loss": 0.5482,
+      "step": 98
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.8600066418600247,
+      "learning_rate": 4.9608448421814944e-06,
+      "loss": 0.6207,
+      "step": 99
+    },
+    {
+      "epoch": 0.8547008547008547,
+      "grad_norm": 1.035150298649069,
+      "learning_rate": 4.960051957873726e-06,
+      "loss": 0.8249,
+      "step": 100
+    },
+    {
+      "epoch": 0.8632478632478633,
+      "grad_norm": 1.1083245611640722,
+      "learning_rate": 4.959251190599699e-06,
+      "loss": 1.0021,
+      "step": 101
+    },
+    {
+      "epoch": 0.8717948717948718,
+      "grad_norm": 0.7667808215315018,
+      "learning_rate": 4.958442542925385e-06,
+      "loss": 0.6136,
+      "step": 102
+    },
+    {
+      "epoch": 0.8803418803418803,
+      "grad_norm": 0.8877096581857563,
+      "learning_rate": 4.9576260174420085e-06,
+      "loss": 0.6767,
+      "step": 103
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.1246219207085053,
+      "learning_rate": 4.956801616766033e-06,
+      "loss": 0.783,
+      "step": 104
+    },
+    {
+      "epoch": 0.8974358974358975,
+      "grad_norm": 0.9568379203614392,
+      "learning_rate": 4.955969343539162e-06,
+      "loss": 0.8778,
+      "step": 105
+    },
+    {
+      "epoch": 0.905982905982906,
+      "grad_norm": 0.9605105213476728,
+      "learning_rate": 4.955129200428323e-06,
+      "loss": 0.6912,
+      "step": 106
+    },
+    {
+      "epoch": 0.9145299145299145,
+      "grad_norm": 1.1668396082908756,
+      "learning_rate": 4.9542811901256615e-06,
+      "loss": 0.7407,
+      "step": 107
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.8666294775123025,
+      "learning_rate": 4.953425315348534e-06,
+      "loss": 0.6811,
+      "step": 108
+    },
+    {
+      "epoch": 0.9316239316239316,
+      "grad_norm": 1.0107691793778373,
+      "learning_rate": 4.952561578839498e-06,
+      "loss": 0.7776,
+      "step": 109
+    },
+    {
+      "epoch": 0.9401709401709402,
+      "grad_norm": 1.0252997044098662,
+      "learning_rate": 4.9516899833663e-06,
+      "loss": 0.7653,
+      "step": 110
+    },
+    {
+      "epoch": 0.9487179487179487,
+      "grad_norm": 1.0043988910829953,
+      "learning_rate": 4.950810531721874e-06,
+      "loss": 0.7352,
+      "step": 111
+    },
+    {
+      "epoch": 0.9572649572649573,
+      "grad_norm": 0.9966732259935451,
+      "learning_rate": 4.949923226724325e-06,
+      "loss": 0.7956,
+      "step": 112
+    },
+    {
+      "epoch": 0.9658119658119658,
+      "grad_norm": 1.0002785418115747,
+      "learning_rate": 4.949028071216926e-06,
+      "loss": 0.6522,
+      "step": 113
+    },
+    {
+      "epoch": 0.9743589743589743,
+      "grad_norm": 0.8434616881462327,
+      "learning_rate": 4.948125068068102e-06,
+      "loss": 0.5084,
+      "step": 114
+    },
+    {
+      "epoch": 0.9829059829059829,
+      "grad_norm": 0.9643032365710266,
+      "learning_rate": 4.94721422017143e-06,
+      "loss": 0.966,
+      "step": 115
+    },
+    {
+      "epoch": 0.9914529914529915,
+      "grad_norm": 0.8101745858666015,
+      "learning_rate": 4.946295530445621e-06,
+      "loss": 0.7198,
+      "step": 116
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8729701781074534,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.6823,
+      "step": 117
+    },
+    {
+      "epoch": 1.0085470085470085,
+      "grad_norm": 0.842080592587243,
+      "learning_rate": 4.94443463730707e-06,
+      "loss": 0.6881,
+      "step": 118
+    },
+    {
+      "epoch": 1.017094017094017,
+      "grad_norm": 0.845463199764785,
+      "learning_rate": 4.943492439857357e-06,
+      "loss": 0.7705,
+      "step": 119
+    },
+    {
+      "epoch": 1.0256410256410255,
+      "grad_norm": 0.9180778501256365,
+      "learning_rate": 4.942542412504543e-06,
+      "loss": 0.804,
+      "step": 120
+    },
+    {
+      "epoch": 1.0341880341880343,
+      "grad_norm": 0.9068334850104435,
+      "learning_rate": 4.9415845582928866e-06,
+      "loss": 0.5952,
+      "step": 121
+    },
+    {
+      "epoch": 1.0427350427350428,
+      "grad_norm": 1.0711610169869845,
+      "learning_rate": 4.940618880291725e-06,
+      "loss": 1.0312,
+      "step": 122
+    },
+    {
+      "epoch": 1.0512820512820513,
+      "grad_norm": 0.8660635593426306,
+      "learning_rate": 4.9396453815954695e-06,
+      "loss": 0.6006,
+      "step": 123
+    },
+    {
+      "epoch": 1.0598290598290598,
+      "grad_norm": 0.7295836928618811,
+      "learning_rate": 4.938664065323588e-06,
+      "loss": 0.6308,
+      "step": 124
+    },
+    {
+      "epoch": 1.0683760683760684,
+      "grad_norm": 0.9418396522120512,
+      "learning_rate": 4.937674934620601e-06,
+      "loss": 0.7542,
+      "step": 125
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 1.0198957963558342,
+      "learning_rate": 4.9366779926560705e-06,
+      "loss": 0.7289,
+      "step": 126
+    },
+    {
+      "epoch": 1.0854700854700854,
+      "grad_norm": 0.751898046812773,
+      "learning_rate": 4.935673242624585e-06,
+      "loss": 0.5528,
+      "step": 127
+    },
+    {
+      "epoch": 1.0940170940170941,
+      "grad_norm": 0.792545272551657,
+      "learning_rate": 4.934660687745758e-06,
+      "loss": 0.6599,
+      "step": 128
+    },
+    {
+      "epoch": 1.1025641025641026,
+      "grad_norm": 0.904722485954194,
+      "learning_rate": 4.93364033126421e-06,
+      "loss": 0.6349,
+      "step": 129
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 1.0909141456525318,
+      "learning_rate": 4.93261217644956e-06,
+      "loss": 0.8165,
+      "step": 130
+    },
+    {
+      "epoch": 1.1196581196581197,
+      "grad_norm": 0.8711040377712898,
+      "learning_rate": 4.931576226596418e-06,
+      "loss": 0.8699,
+      "step": 131
+    },
+    {
+      "epoch": 1.1282051282051282,
+      "grad_norm": 0.8633774504272915,
+      "learning_rate": 4.930532485024372e-06,
+      "loss": 0.5695,
+      "step": 132
+    },
+    {
+      "epoch": 1.1367521367521367,
+      "grad_norm": 0.8889574637503499,
+      "learning_rate": 4.929480955077976e-06,
+      "loss": 0.5382,
+      "step": 133
+    },
+    {
+      "epoch": 1.1452991452991452,
+      "grad_norm": 0.9325352651456993,
+      "learning_rate": 4.928421640126742e-06,
+      "loss": 0.6109,
+      "step": 134
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.7740567991589352,
+      "learning_rate": 4.927354543565131e-06,
+      "loss": 0.6082,
+      "step": 135
+    },
+    {
+      "epoch": 1.1623931623931625,
+      "grad_norm": 0.8569473937370548,
+      "learning_rate": 4.926279668812533e-06,
+      "loss": 0.8508,
+      "step": 136
+    },
+    {
+      "epoch": 1.170940170940171,
+      "grad_norm": 0.9000831758374732,
+      "learning_rate": 4.925197019313269e-06,
+      "loss": 0.6447,
+      "step": 137
+    },
+    {
+      "epoch": 1.1794871794871795,
+      "grad_norm": 0.8268169038977321,
+      "learning_rate": 4.9241065985365695e-06,
+      "loss": 0.6781,
+      "step": 138
+    },
+    {
+      "epoch": 1.188034188034188,
+      "grad_norm": 0.9296228757358989,
+      "learning_rate": 4.923008409976568e-06,
+      "loss": 0.736,
+      "step": 139
+    },
+    {
+      "epoch": 1.1965811965811965,
+      "grad_norm": 0.921899150937168,
+      "learning_rate": 4.921902457152289e-06,
+      "loss": 0.7198,
+      "step": 140
+    },
+    {
+      "epoch": 1.205128205128205,
+      "grad_norm": 0.9602358855116797,
+      "learning_rate": 4.920788743607636e-06,
+      "loss": 0.8336,
+      "step": 141
+    },
+    {
+      "epoch": 1.2136752136752136,
+      "grad_norm": 0.8896085782433711,
+      "learning_rate": 4.919667272911383e-06,
+      "loss": 0.6959,
+      "step": 142
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.846293805538022,
+      "learning_rate": 4.91853804865716e-06,
+      "loss": 0.7555,
+      "step": 143
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.8603018280332412,
+      "learning_rate": 4.917401074463441e-06,
+      "loss": 0.7288,
+      "step": 144
+    },
+    {
+      "epoch": 1.2393162393162394,
+      "grad_norm": 0.9498713513811322,
+      "learning_rate": 4.916256353973535e-06,
+      "loss": 0.647,
+      "step": 145
+    },
+    {
+      "epoch": 1.2478632478632479,
+      "grad_norm": 0.7982026276207213,
+      "learning_rate": 4.915103890855574e-06,
+      "loss": 0.6042,
+      "step": 146
+    },
+    {
+      "epoch": 1.2564102564102564,
+      "grad_norm": 0.8589595835743361,
+      "learning_rate": 4.913943688802497e-06,
+      "loss": 0.6211,
+      "step": 147
+    },
+    {
+      "epoch": 1.264957264957265,
+      "grad_norm": 0.9630850664374468,
+      "learning_rate": 4.912775751532047e-06,
+      "loss": 0.7537,
+      "step": 148
+    },
+    {
+      "epoch": 1.2735042735042734,
+      "grad_norm": 1.084629843197026,
+      "learning_rate": 4.91160008278675e-06,
+      "loss": 0.651,
+      "step": 149
+    },
+    {
+      "epoch": 1.282051282051282,
+      "grad_norm": 1.077510994503243,
+      "learning_rate": 4.9104166863339065e-06,
+      "loss": 0.6962,
+      "step": 150
+    },
+    {
+      "epoch": 1.2905982905982907,
+      "grad_norm": 0.9523844356052744,
+      "learning_rate": 4.90922556596558e-06,
+      "loss": 0.9121,
+      "step": 151
+    },
+    {
+      "epoch": 1.2991452991452992,
+      "grad_norm": 0.8170982239546886,
+      "learning_rate": 4.908026725498586e-06,
+      "loss": 0.5877,
+      "step": 152
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.8321214392716167,
+      "learning_rate": 4.9068201687744774e-06,
+      "loss": 0.6241,
+      "step": 153
+    },
+    {
+      "epoch": 1.3162393162393162,
+      "grad_norm": 0.744635231568867,
+      "learning_rate": 4.905605899659532e-06,
+      "loss": 0.5604,
+      "step": 154
+    },
+    {
+      "epoch": 1.3247863247863247,
+      "grad_norm": 0.9902531868558738,
+      "learning_rate": 4.90438392204474e-06,
+      "loss": 0.7455,
+      "step": 155
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.871166691569676,
+      "learning_rate": 4.903154239845798e-06,
+      "loss": 0.7321,
+      "step": 156
+    },
+    {
+      "epoch": 1.341880341880342,
+      "grad_norm": 0.9899344615854253,
+      "learning_rate": 4.901916857003084e-06,
+      "loss": 0.9039,
+      "step": 157
+    },
+    {
+      "epoch": 1.3504273504273505,
+      "grad_norm": 0.771145237334227,
+      "learning_rate": 4.9006717774816585e-06,
+      "loss": 0.9487,
+      "step": 158
+    },
+    {
+      "epoch": 1.358974358974359,
+      "grad_norm": 0.8764015362376394,
+      "learning_rate": 4.8994190052712406e-06,
+      "loss": 0.4107,
+      "step": 159
+    },
+    {
+      "epoch": 1.3675213675213675,
+      "grad_norm": 1.0257151713300179,
+      "learning_rate": 4.898158544386201e-06,
+      "loss": 0.9209,
+      "step": 160
+    },
+    {
+      "epoch": 1.376068376068376,
+      "grad_norm": 0.6891258050787235,
+      "learning_rate": 4.896890398865548e-06,
+      "loss": 0.473,
+      "step": 161
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.8800514494711961,
+      "learning_rate": 4.895614572772916e-06,
+      "loss": 0.5578,
+      "step": 162
+    },
+    {
+      "epoch": 1.393162393162393,
+      "grad_norm": 1.0877581722845024,
+      "learning_rate": 4.894331070196548e-06,
+      "loss": 0.7935,
+      "step": 163
+    },
+    {
+      "epoch": 1.4017094017094016,
+      "grad_norm": 0.7380344836306842,
+      "learning_rate": 4.893039895249288e-06,
+      "loss": 0.5807,
+      "step": 164
+    },
+    {
+      "epoch": 1.4102564102564101,
+      "grad_norm": 0.9374611560377141,
+      "learning_rate": 4.8917410520685635e-06,
+      "loss": 0.7081,
+      "step": 165
+    },
+    {
+      "epoch": 1.4188034188034189,
+      "grad_norm": 0.9888064163493844,
+      "learning_rate": 4.890434544816375e-06,
+      "loss": 0.621,
+      "step": 166
+    },
+    {
+      "epoch": 1.4273504273504274,
+      "grad_norm": 3.3286973317835074,
+      "learning_rate": 4.889120377679282e-06,
+      "loss": 0.6302,
+      "step": 167
+    },
+    {
+      "epoch": 1.435897435897436,
+      "grad_norm": 0.8952054360752244,
+      "learning_rate": 4.887798554868388e-06,
+      "loss": 0.5884,
+      "step": 168
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.8659549188864638,
+      "learning_rate": 4.88646908061933e-06,
+      "loss": 0.5574,
+      "step": 169
+    },
+    {
+      "epoch": 1.452991452991453,
+      "grad_norm": 0.9342371205015247,
+      "learning_rate": 4.885131959192262e-06,
+      "loss": 0.5912,
+      "step": 170
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.9723104482463416,
+      "learning_rate": 4.883787194871841e-06,
+      "loss": 0.5887,
+      "step": 171
+    },
+    {
+      "epoch": 1.4700854700854702,
+      "grad_norm": 0.7578829988984429,
+      "learning_rate": 4.882434791967219e-06,
+      "loss": 0.6065,
+      "step": 172
+    },
+    {
+      "epoch": 1.4786324786324787,
+      "grad_norm": 0.8891827518266336,
+      "learning_rate": 4.881074754812021e-06,
+      "loss": 0.5384,
+      "step": 173
+    },
+    {
+      "epoch": 1.4871794871794872,
+      "grad_norm": 1.4201619598272113,
+      "learning_rate": 4.879707087764336e-06,
+      "loss": 0.6335,
+      "step": 174
+    },
+    {
+      "epoch": 1.4957264957264957,
+      "grad_norm": 0.7970843509207138,
+      "learning_rate": 4.878331795206705e-06,
+      "loss": 0.7274,
+      "step": 175
+    },
+    {
+      "epoch": 1.5042735042735043,
+      "grad_norm": 1.0616993712483198,
+      "learning_rate": 4.876948881546101e-06,
+      "loss": 0.6813,
+      "step": 176
+    },
+    {
+      "epoch": 1.5128205128205128,
+      "grad_norm": 0.8826451632766461,
+      "learning_rate": 4.875558351213918e-06,
+      "loss": 0.6883,
+      "step": 177
+    },
+    {
+      "epoch": 1.5213675213675213,
+      "grad_norm": 0.8241904159222645,
+      "learning_rate": 4.874160208665958e-06,
+      "loss": 0.4233,
+      "step": 178
+    },
+    {
+      "epoch": 1.5299145299145298,
+      "grad_norm": 1.0537700071390603,
+      "learning_rate": 4.872754458382416e-06,
+      "loss": 0.6817,
+      "step": 179
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.8301266822469117,
+      "learning_rate": 4.8713411048678635e-06,
+      "loss": 0.8945,
+      "step": 180
+    },
+    {
+      "epoch": 1.547008547008547,
+      "grad_norm": 0.8550186654323384,
+      "learning_rate": 4.869920152651239e-06,
+      "loss": 0.7814,
+      "step": 181
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 1.3037022506462685,
+      "learning_rate": 4.868491606285823e-06,
+      "loss": 0.7902,
+      "step": 182
+    },
+    {
+      "epoch": 1.564102564102564,
+      "grad_norm": 1.710455466821508,
+      "learning_rate": 4.86705547034924e-06,
+      "loss": 0.7026,
+      "step": 183
+    },
+    {
+      "epoch": 1.5726495726495726,
+      "grad_norm": 0.8504039369254447,
+      "learning_rate": 4.865611749443428e-06,
+      "loss": 0.6511,
+      "step": 184
+    },
+    {
+      "epoch": 1.5811965811965814,
+      "grad_norm": 0.8002095192240137,
+      "learning_rate": 4.864160448194632e-06,
+      "loss": 0.5932,
+      "step": 185
+    },
+    {
+      "epoch": 1.5897435897435899,
+      "grad_norm": 0.95226059536856,
+      "learning_rate": 4.862701571253387e-06,
+      "loss": 0.5688,
+      "step": 186
+    },
+    {
+      "epoch": 1.5982905982905984,
+      "grad_norm": 0.7962164201947765,
+      "learning_rate": 4.861235123294505e-06,
+      "loss": 0.5607,
+      "step": 187
+    },
+    {
+      "epoch": 1.606837606837607,
+      "grad_norm": 1.0060466476736434,
+      "learning_rate": 4.859761109017056e-06,
+      "loss": 0.5513,
+      "step": 188
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.7785587180515594,
+      "learning_rate": 4.858279533144358e-06,
+      "loss": 0.6487,
+      "step": 189
+    },
+    {
+      "epoch": 1.623931623931624,
+      "grad_norm": 0.8540481332902222,
+      "learning_rate": 4.856790400423958e-06,
+      "loss": 0.6159,
+      "step": 190
+    },
+    {
+      "epoch": 1.6324786324786325,
+      "grad_norm": 0.8289468924265531,
+      "learning_rate": 4.8552937156276185e-06,
+      "loss": 0.5114,
+      "step": 191
+    },
+    {
+      "epoch": 1.641025641025641,
+      "grad_norm": 0.9526151621277151,
+      "learning_rate": 4.8537894835513e-06,
+      "loss": 0.9854,
+      "step": 192
+    },
+    {
+      "epoch": 1.6495726495726495,
+      "grad_norm": 0.7910488490007835,
+      "learning_rate": 4.8522777090151505e-06,
+      "loss": 0.5659,
+      "step": 193
+    },
+    {
+      "epoch": 1.658119658119658,
+      "grad_norm": 0.9166695744046246,
+      "learning_rate": 4.8507583968634845e-06,
+      "loss": 0.6635,
+      "step": 194
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.8411571133758534,
+      "learning_rate": 4.849231551964771e-06,
+      "loss": 0.5762,
+      "step": 195
+    },
+    {
+      "epoch": 1.6752136752136753,
+      "grad_norm": 1.0037339289648717,
+      "learning_rate": 4.847697179211618e-06,
+      "loss": 0.4801,
+      "step": 196
+    },
+    {
+      "epoch": 1.6837606837606838,
+      "grad_norm": 0.9650823545296664,
+      "learning_rate": 4.8461552835207524e-06,
+      "loss": 0.5585,
+      "step": 197
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.7514863818600783,
+      "learning_rate": 4.844605869833011e-06,
+      "loss": 0.5708,
+      "step": 198
+    },
+    {
+      "epoch": 1.7008547008547008,
+      "grad_norm": 0.9806737566272027,
+      "learning_rate": 4.84304894311332e-06,
+      "loss": 0.8652,
+      "step": 199
+    },
+    {
+      "epoch": 1.7094017094017095,
+      "grad_norm": 0.8228043634560646,
+      "learning_rate": 4.841484508350679e-06,
+      "loss": 0.5162,
+      "step": 200
+    },
+    {
+      "epoch": 1.717948717948718,
+      "grad_norm": 0.8211227168659619,
+      "learning_rate": 4.839912570558148e-06,
+      "loss": 0.5694,
+      "step": 201
+    },
+    {
+      "epoch": 1.7264957264957266,
+      "grad_norm": 0.8195609975365399,
+      "learning_rate": 4.838333134772828e-06,
+      "loss": 0.4863,
+      "step": 202
+    },
+    {
+      "epoch": 1.735042735042735,
+      "grad_norm": 0.974131586369127,
+      "learning_rate": 4.836746206055849e-06,
+      "loss": 0.551,
+      "step": 203
+    },
+    {
+      "epoch": 1.7435897435897436,
+      "grad_norm": 0.9708650745354207,
+      "learning_rate": 4.835151789492348e-06,
+      "loss": 0.6683,
+      "step": 204
+    },
+    {
+      "epoch": 1.7521367521367521,
+      "grad_norm": 0.8306266939889401,
+      "learning_rate": 4.83354989019146e-06,
+      "loss": 0.5167,
+      "step": 205
+    },
+    {
+      "epoch": 1.7606837606837606,
+      "grad_norm": 0.8692823244900135,
+      "learning_rate": 4.831940513286293e-06,
+      "loss": 0.6876,
+      "step": 206
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.9425701508115182,
+      "learning_rate": 4.83032366393392e-06,
+      "loss": 0.6458,
+      "step": 207
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.9353325892585003,
+      "learning_rate": 4.828699347315357e-06,
+      "loss": 0.8378,
+      "step": 208
+    },
+    {
+      "epoch": 1.7863247863247862,
+      "grad_norm": 0.9162028835750752,
+      "learning_rate": 4.827067568635546e-06,
+      "loss": 0.5463,
+      "step": 209
+    },
+    {
+      "epoch": 1.7948717948717947,
+      "grad_norm": 1.047609340341694,
+      "learning_rate": 4.825428333123346e-06,
+      "loss": 0.7311,
+      "step": 210
+    },
+    {
+      "epoch": 1.8034188034188035,
+      "grad_norm": 1.0234012438481053,
+      "learning_rate": 4.823781646031505e-06,
+      "loss": 0.7,
+      "step": 211
+    },
+    {
+      "epoch": 1.811965811965812,
+      "grad_norm": 0.9046824788092985,
+      "learning_rate": 4.822127512636652e-06,
+      "loss": 0.3795,
+      "step": 212
+    },
+    {
+      "epoch": 1.8205128205128205,
+      "grad_norm": 0.9534823492778117,
+      "learning_rate": 4.820465938239274e-06,
+      "loss": 0.6623,
+      "step": 213
+    },
+    {
+      "epoch": 1.8290598290598292,
+      "grad_norm": 0.9956882984441957,
+      "learning_rate": 4.8187969281637054e-06,
+      "loss": 0.6584,
+      "step": 214
+    },
+    {
+      "epoch": 1.8376068376068377,
+      "grad_norm": 0.9077439371948645,
+      "learning_rate": 4.817120487758105e-06,
+      "loss": 0.6596,
+      "step": 215
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.781496567826663,
+      "learning_rate": 4.815436622394442e-06,
+      "loss": 0.5389,
+      "step": 216
+    },
+    {
+      "epoch": 1.8547008547008548,
+      "grad_norm": 0.9458504816016381,
+      "learning_rate": 4.813745337468478e-06,
+      "loss": 0.7911,
+      "step": 217
+    },
+    {
+      "epoch": 1.8632478632478633,
+      "grad_norm": 0.8741433154654207,
+      "learning_rate": 4.8120466383997486e-06,
+      "loss": 0.6354,
+      "step": 218
+    },
+    {
+      "epoch": 1.8717948717948718,
+      "grad_norm": 0.9740111296003311,
+      "learning_rate": 4.81034053063155e-06,
+      "loss": 0.6901,
+      "step": 219
+    },
+    {
+      "epoch": 1.8803418803418803,
+      "grad_norm": 0.8376294681303905,
+      "learning_rate": 4.8086270196309174e-06,
+      "loss": 0.8544,
+      "step": 220
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.9076663090188682,
+      "learning_rate": 4.806906110888606e-06,
+      "loss": 0.6304,
+      "step": 221
+    },
+    {
+      "epoch": 1.8974358974358974,
+      "grad_norm": 0.9713152148006099,
+      "learning_rate": 4.805177809919081e-06,
+      "loss": 0.7956,
+      "step": 222
+    },
+    {
+      "epoch": 1.9059829059829059,
+      "grad_norm": 1.0226062030376948,
+      "learning_rate": 4.803442122260494e-06,
+      "loss": 0.6901,
+      "step": 223
+    },
+    {
+      "epoch": 1.9145299145299144,
+      "grad_norm": 0.8119065258323606,
+      "learning_rate": 4.801699053474663e-06,
+      "loss": 0.5318,
+      "step": 224
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.9150490205225476,
+      "learning_rate": 4.799948609147061e-06,
+      "loss": 0.5784,
+      "step": 225
+    },
+    {
+      "epoch": 1.9316239316239316,
+      "grad_norm": 0.9035676881199971,
+      "learning_rate": 4.798190794886795e-06,
+      "loss": 0.6402,
+      "step": 226
+    },
+    {
+      "epoch": 1.9401709401709402,
+      "grad_norm": 0.7583311032119001,
+      "learning_rate": 4.796425616326588e-06,
+      "loss": 0.4664,
+      "step": 227
+    },
+    {
+      "epoch": 1.9487179487179487,
+      "grad_norm": 0.9669728608522254,
+      "learning_rate": 4.79465307912276e-06,
+      "loss": 0.5637,
+      "step": 228
+    },
+    {
+      "epoch": 1.9572649572649574,
+      "grad_norm": 0.8667261712455026,
+      "learning_rate": 4.792873188955213e-06,
+      "loss": 0.8446,
+      "step": 229
+    },
+    {
+      "epoch": 1.965811965811966,
+      "grad_norm": 0.9687898626902562,
+      "learning_rate": 4.791085951527408e-06,
+      "loss": 0.5547,
+      "step": 230
+    },
+    {
+      "epoch": 1.9743589743589745,
+      "grad_norm": 0.8157471679820112,
+      "learning_rate": 4.789291372566352e-06,
+      "loss": 0.675,
+      "step": 231
+    },
+    {
+      "epoch": 1.982905982905983,
+      "grad_norm": 0.8596776874736839,
+      "learning_rate": 4.787489457822576e-06,
+      "loss": 0.7064,
+      "step": 232
+    },
+    {
+      "epoch": 1.9914529914529915,
+      "grad_norm": 0.992779009649159,
+      "learning_rate": 4.785680213070117e-06,
+      "loss": 0.9077,
+      "step": 233
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9030091834967064,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.6325,
+      "step": 234
+    },
+    {
+      "epoch": 2.0085470085470085,
+      "grad_norm": 0.9423554920024213,
+      "learning_rate": 4.782039756752728e-06,
+      "loss": 0.5702,
+      "step": 235
+    },
+    {
+      "epoch": 2.017094017094017,
+      "grad_norm": 0.8379965111584239,
+      "learning_rate": 4.780208556853239e-06,
+      "loss": 0.4504,
+      "step": 236
+    },
+    {
+      "epoch": 2.0256410256410255,
+      "grad_norm": 0.7824169343393776,
+      "learning_rate": 4.7783700502759145e-06,
+      "loss": 0.5999,
+      "step": 237
+    },
+    {
+      "epoch": 2.034188034188034,
+      "grad_norm": 0.8558656070638038,
+      "learning_rate": 4.776524242912047e-06,
+      "loss": 0.7578,
+      "step": 238
+    },
+    {
+      "epoch": 2.0427350427350426,
+      "grad_norm": 0.8526003041858661,
+      "learning_rate": 4.774671140676325e-06,
+      "loss": 0.5964,
+      "step": 239
+    },
+    {
+      "epoch": 2.051282051282051,
+      "grad_norm": 0.9112032508436396,
+      "learning_rate": 4.77281074950681e-06,
+      "loss": 0.8572,
+      "step": 240
+    },
+    {
+      "epoch": 2.0598290598290596,
+      "grad_norm": 0.8581968072977434,
+      "learning_rate": 4.77094307536492e-06,
+      "loss": 0.5085,
+      "step": 241
+    },
+    {
+      "epoch": 2.0683760683760686,
+      "grad_norm": 0.7253249437985466,
+      "learning_rate": 4.769068124235413e-06,
+      "loss": 0.4323,
+      "step": 242
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 1.0234933231378351,
+      "learning_rate": 4.7671859021263635e-06,
+      "loss": 0.6929,
+      "step": 243
+    },
+    {
+      "epoch": 2.0854700854700856,
+      "grad_norm": 0.8536915333923158,
+      "learning_rate": 4.765296415069146e-06,
+      "loss": 0.5469,
+      "step": 244
+    },
+    {
+      "epoch": 2.094017094017094,
+      "grad_norm": 0.8417797765711249,
+      "learning_rate": 4.763399669118414e-06,
+      "loss": 0.7197,
+      "step": 245
+    },
+    {
+      "epoch": 2.1025641025641026,
+      "grad_norm": 0.7504923334053624,
+      "learning_rate": 4.761495670352081e-06,
+      "loss": 0.5067,
+      "step": 246
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 0.8825999832331881,
+      "learning_rate": 4.759584424871302e-06,
+      "loss": 0.5412,
+      "step": 247
+    },
+    {
+      "epoch": 2.1196581196581197,
+      "grad_norm": 0.7981758550893384,
+      "learning_rate": 4.757665938800453e-06,
+      "loss": 0.7171,
+      "step": 248
+    },
+    {
+      "epoch": 2.128205128205128,
+      "grad_norm": 0.7853772293068558,
+      "learning_rate": 4.755740218287113e-06,
+      "loss": 0.3035,
+      "step": 249
+    },
+    {
+      "epoch": 2.1367521367521367,
+      "grad_norm": 0.7971592537457343,
+      "learning_rate": 4.753807269502041e-06,
+      "loss": 0.6156,
+      "step": 250
+    },
+    {
+      "epoch": 2.1452991452991452,
+      "grad_norm": 0.7213055348629424,
+      "learning_rate": 4.7518670986391576e-06,
+      "loss": 0.4422,
+      "step": 251
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.8314372014471705,
+      "learning_rate": 4.749919711915531e-06,
+      "loss": 0.6491,
+      "step": 252
+    },
+    {
+      "epoch": 2.1623931623931623,
+      "grad_norm": 0.8906551957721053,
+      "learning_rate": 4.747965115571345e-06,
+      "loss": 0.5359,
+      "step": 253
+    },
+    {
+      "epoch": 2.1709401709401708,
+      "grad_norm": 0.8109146133188956,
+      "learning_rate": 4.746003315869889e-06,
+      "loss": 0.5431,
+      "step": 254
+    },
+    {
+      "epoch": 2.1794871794871793,
+      "grad_norm": 1.0629574061098976,
+      "learning_rate": 4.744034319097536e-06,
+      "loss": 0.5012,
+      "step": 255
+    },
+    {
+      "epoch": 2.1880341880341883,
+      "grad_norm": 0.8497678709421456,
+      "learning_rate": 4.742058131563718e-06,
+      "loss": 0.5794,
+      "step": 256
+    },
+    {
+      "epoch": 2.1965811965811968,
+      "grad_norm": 0.8307773403474242,
+      "learning_rate": 4.7400747596009125e-06,
+      "loss": 0.4953,
+      "step": 257
+    },
+    {
+      "epoch": 2.2051282051282053,
+      "grad_norm": 0.8059782529981341,
+      "learning_rate": 4.738084209564617e-06,
+      "loss": 0.4722,
+      "step": 258
+    },
+    {
+      "epoch": 2.213675213675214,
+      "grad_norm": 0.7633142931516772,
+      "learning_rate": 4.73608648783333e-06,
+      "loss": 0.535,
+      "step": 259
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.8445214601230614,
+      "learning_rate": 4.734081600808531e-06,
+      "loss": 0.6529,
+      "step": 260
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.7888800398644392,
+      "learning_rate": 4.73206955491466e-06,
+      "loss": 0.4053,
+      "step": 261
+    },
+    {
+      "epoch": 2.2393162393162394,
+      "grad_norm": 0.8353629817434199,
+      "learning_rate": 4.7300503565990985e-06,
+      "loss": 0.4727,
+      "step": 262
+    },
+    {
+      "epoch": 2.247863247863248,
+      "grad_norm": 0.7104386481467555,
+      "learning_rate": 4.728024012332145e-06,
+      "loss": 0.4199,
+      "step": 263
+    },
+    {
+      "epoch": 2.2564102564102564,
+      "grad_norm": 0.8725564755846308,
+      "learning_rate": 4.725990528606996e-06,
+      "loss": 0.4241,
+      "step": 264
+    },
+    {
+      "epoch": 2.264957264957265,
+      "grad_norm": 0.7957719847040631,
+      "learning_rate": 4.723949911939728e-06,
+      "loss": 0.5536,
+      "step": 265
+    },
+    {
+      "epoch": 2.2735042735042734,
+      "grad_norm": 1.082844473015555,
+      "learning_rate": 4.7219021688692725e-06,
+      "loss": 0.5833,
+      "step": 266
+    },
+    {
+      "epoch": 2.282051282051282,
+      "grad_norm": 0.9065232591964502,
+      "learning_rate": 4.719847305957398e-06,
+      "loss": 0.5619,
+      "step": 267
+    },
+    {
+      "epoch": 2.2905982905982905,
+      "grad_norm": 0.8547973656651287,
+      "learning_rate": 4.717785329788685e-06,
+      "loss": 0.6832,
+      "step": 268
+    },
+    {
+      "epoch": 2.299145299145299,
+      "grad_norm": 0.9296372212216126,
+      "learning_rate": 4.715716246970511e-06,
+      "loss": 0.7332,
+      "step": 269
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.7861517591016323,
+      "learning_rate": 4.7136400641330245e-06,
+      "loss": 0.5523,
+      "step": 270
+    },
+    {
+      "epoch": 2.316239316239316,
+      "grad_norm": 0.8432735931836609,
+      "learning_rate": 4.7115567879291265e-06,
+      "loss": 0.4783,
+      "step": 271
+    },
+    {
+      "epoch": 2.324786324786325,
+      "grad_norm": 0.810227255259576,
+      "learning_rate": 4.709466425034445e-06,
+      "loss": 0.6832,
+      "step": 272
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.8627425682827192,
+      "learning_rate": 4.707368982147318e-06,
+      "loss": 0.6009,
+      "step": 273
+    },
+    {
+      "epoch": 2.341880341880342,
+      "grad_norm": 0.8933833494570184,
+      "learning_rate": 4.705264465988771e-06,
+      "loss": 0.6507,
+      "step": 274
+    },
+    {
+      "epoch": 2.3504273504273505,
+      "grad_norm": 0.9536316043302439,
+      "learning_rate": 4.703152883302498e-06,
+      "loss": 0.4248,
+      "step": 275
+    },
+    {
+      "epoch": 2.358974358974359,
+      "grad_norm": 0.8128802583501903,
+      "learning_rate": 4.701034240854829e-06,
+      "loss": 0.5562,
+      "step": 276
+    },
+    {
+      "epoch": 2.3675213675213675,
+      "grad_norm": 0.858099919053899,
+      "learning_rate": 4.6989085454347236e-06,
+      "loss": 0.5479,
+      "step": 277
+    },
+    {
+      "epoch": 2.376068376068376,
+      "grad_norm": 0.8472803686361211,
+      "learning_rate": 4.696775803853739e-06,
+      "loss": 0.5566,
+      "step": 278
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.7851374043861304,
+      "learning_rate": 4.694636022946012e-06,
+      "loss": 0.5711,
+      "step": 279
+    },
+    {
+      "epoch": 2.393162393162393,
+      "grad_norm": 0.8317592786523537,
+      "learning_rate": 4.692489209568234e-06,
+      "loss": 0.7418,
+      "step": 280
+    },
+    {
+      "epoch": 2.4017094017094016,
+      "grad_norm": 0.8853581482172665,
+      "learning_rate": 4.690335370599633e-06,
+      "loss": 0.6775,
+      "step": 281
+    },
+    {
+      "epoch": 2.41025641025641,
+      "grad_norm": 0.8539229096616547,
+      "learning_rate": 4.68817451294195e-06,
+      "loss": 0.5026,
+      "step": 282
+    },
+    {
+      "epoch": 2.4188034188034186,
+      "grad_norm": 0.8208950002596763,
+      "learning_rate": 4.686006643519415e-06,
+      "loss": 0.6944,
+      "step": 283
+    },
+    {
+      "epoch": 2.427350427350427,
+      "grad_norm": 0.8841090381845832,
+      "learning_rate": 4.683831769278729e-06,
+      "loss": 0.5945,
+      "step": 284
+    },
+    {
+      "epoch": 2.435897435897436,
+      "grad_norm": 0.8232663064958912,
+      "learning_rate": 4.681649897189036e-06,
+      "loss": 0.556,
+      "step": 285
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.8148096515474825,
+      "learning_rate": 4.679461034241906e-06,
+      "loss": 0.5738,
+      "step": 286
+    },
+    {
+      "epoch": 2.452991452991453,
+      "grad_norm": 0.8082914935046664,
+      "learning_rate": 4.677265187451311e-06,
+      "loss": 0.4538,
+      "step": 287
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.9043659946127143,
+      "learning_rate": 4.675062363853599e-06,
+      "loss": 0.641,
+      "step": 288
+    },
+    {
+      "epoch": 2.47008547008547,
+      "grad_norm": 0.6764037851210951,
+      "learning_rate": 4.672852570507476e-06,
+      "loss": 0.3586,
+      "step": 289
+    },
+    {
+      "epoch": 2.4786324786324787,
+      "grad_norm": 0.7621893086172707,
+      "learning_rate": 4.670635814493985e-06,
+      "loss": 0.5006,
+      "step": 290
+    },
+    {
+      "epoch": 2.4871794871794872,
+      "grad_norm": 0.7696631077204455,
+      "learning_rate": 4.668412102916474e-06,
+      "loss": 0.4888,
+      "step": 291
+    },
+    {
+      "epoch": 2.4957264957264957,
+      "grad_norm": 0.8724597404888246,
+      "learning_rate": 4.666181442900584e-06,
+      "loss": 0.6726,
+      "step": 292
+    },
+    {
+      "epoch": 2.5042735042735043,
+      "grad_norm": 1.0178600439419474,
+      "learning_rate": 4.663943841594219e-06,
+      "loss": 0.8028,
+      "step": 293
+    },
+    {
+      "epoch": 2.5128205128205128,
+      "grad_norm": 0.821176529609733,
+      "learning_rate": 4.6616993061675275e-06,
+      "loss": 0.4614,
+      "step": 294
+    },
+    {
+      "epoch": 2.5213675213675213,
+      "grad_norm": 0.954374041308384,
+      "learning_rate": 4.659447843812876e-06,
+      "loss": 0.5892,
+      "step": 295
+    },
+    {
+      "epoch": 2.52991452991453,
+      "grad_norm": 0.8797608205693254,
+      "learning_rate": 4.657189461744829e-06,
+      "loss": 0.5927,
+      "step": 296
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.8384246808599966,
+      "learning_rate": 4.654924167200124e-06,
+      "loss": 0.5969,
+      "step": 297
+    },
+    {
+      "epoch": 2.547008547008547,
+      "grad_norm": 0.8469590124967289,
+      "learning_rate": 4.652651967437647e-06,
+      "loss": 0.4979,
+      "step": 298
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 0.7910351749484542,
+      "learning_rate": 4.650372869738415e-06,
+      "loss": 0.3562,
+      "step": 299
+    },
+    {
+      "epoch": 2.564102564102564,
+      "grad_norm": 0.9453743074559633,
+      "learning_rate": 4.648086881405542e-06,
+      "loss": 0.5599,
+      "step": 300
+    },
+    {
+      "epoch": 2.5726495726495724,
+      "grad_norm": 0.8425619907341163,
+      "learning_rate": 4.6457940097642315e-06,
+      "loss": 0.3276,
+      "step": 301
+    },
+    {
+      "epoch": 2.5811965811965814,
+      "grad_norm": 0.8428502499947056,
+      "learning_rate": 4.643494262161735e-06,
+      "loss": 0.4904,
+      "step": 302
+    },
+    {
+      "epoch": 2.58974358974359,
+      "grad_norm": 0.9186239857634485,
+      "learning_rate": 4.6411876459673435e-06,
+      "loss": 0.6082,
+      "step": 303
+    },
+    {
+      "epoch": 2.5982905982905984,
+      "grad_norm": 0.8347795234984768,
+      "learning_rate": 4.638874168572355e-06,
+      "loss": 0.5763,
+      "step": 304
+    },
+    {
+      "epoch": 2.606837606837607,
+      "grad_norm": 0.9453440342446296,
+      "learning_rate": 4.636553837390051e-06,
+      "loss": 0.7632,
+      "step": 305
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.9624505274395776,
+      "learning_rate": 4.634226659855681e-06,
+      "loss": 0.5405,
+      "step": 306
+    },
+    {
+      "epoch": 2.623931623931624,
+      "grad_norm": 0.739440573623984,
+      "learning_rate": 4.631892643426428e-06,
+      "loss": 0.6543,
+      "step": 307
+    },
+    {
+      "epoch": 2.6324786324786325,
+      "grad_norm": 0.8183326036817442,
+      "learning_rate": 4.629551795581393e-06,
+      "loss": 0.4819,
+      "step": 308
+    },
+    {
+      "epoch": 2.641025641025641,
+      "grad_norm": 1.0800027921184026,
+      "learning_rate": 4.627204123821563e-06,
+      "loss": 0.6374,
+      "step": 309
+    },
+    {
+      "epoch": 2.6495726495726495,
+      "grad_norm": 0.8644012525735127,
+      "learning_rate": 4.624849635669797e-06,
+      "loss": 0.676,
+      "step": 310
+    },
+    {
+      "epoch": 2.658119658119658,
+      "grad_norm": 0.9303352301220938,
+      "learning_rate": 4.622488338670792e-06,
+      "loss": 0.6964,
+      "step": 311
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.8250093622508519,
+      "learning_rate": 4.620120240391065e-06,
+      "loss": 0.4156,
+      "step": 312
+    },
+    {
+      "epoch": 2.6752136752136755,
+      "grad_norm": 0.8760662748622655,
+      "learning_rate": 4.617745348418928e-06,
+      "loss": 0.8247,
+      "step": 313
+    },
+    {
+      "epoch": 2.683760683760684,
+      "grad_norm": 0.7826247149013078,
+      "learning_rate": 4.61536367036446e-06,
+      "loss": 0.4056,
+      "step": 314
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.9719580059372056,
+      "learning_rate": 4.612975213859487e-06,
+      "loss": 0.6637,
+      "step": 315
+    },
+    {
+      "epoch": 2.700854700854701,
+      "grad_norm": 0.7519148907524309,
+      "learning_rate": 4.6105799865575565e-06,
+      "loss": 0.3255,
+      "step": 316
+    },
+    {
+      "epoch": 2.7094017094017095,
+      "grad_norm": 0.8302883777619188,
+      "learning_rate": 4.60817799613391e-06,
+      "loss": 0.5064,
+      "step": 317
+    },
+    {
+      "epoch": 2.717948717948718,
+      "grad_norm": 0.7963915256809042,
+      "learning_rate": 4.605769250285462e-06,
+      "loss": 0.5783,
+      "step": 318
+    },
+    {
+      "epoch": 2.7264957264957266,
+      "grad_norm": 0.8526922440352511,
+      "learning_rate": 4.603353756730775e-06,
+      "loss": 0.8161,
+      "step": 319
+    },
+    {
+      "epoch": 2.735042735042735,
+      "grad_norm": 0.7735568824902,
+      "learning_rate": 4.600931523210032e-06,
+      "loss": 0.4646,
+      "step": 320
+    },
+    {
+      "epoch": 2.7435897435897436,
+      "grad_norm": 0.7480599875712123,
+      "learning_rate": 4.598502557485015e-06,
+      "loss": 0.5126,
+      "step": 321
+    },
+    {
+      "epoch": 2.752136752136752,
+      "grad_norm": 0.978569107209946,
+      "learning_rate": 4.5960668673390776e-06,
+      "loss": 0.7456,
+      "step": 322
+    },
+    {
+      "epoch": 2.7606837606837606,
+      "grad_norm": 0.8165048476366037,
+      "learning_rate": 4.59362446057712e-06,
+      "loss": 0.5948,
+      "step": 323
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.8341162025089808,
+      "learning_rate": 4.591175345025567e-06,
+      "loss": 0.3868,
+      "step": 324
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 0.78033690790899,
+      "learning_rate": 4.588719528532342e-06,
+      "loss": 0.5951,
+      "step": 325
+    },
+    {
+      "epoch": 2.786324786324786,
+      "grad_norm": 0.9040375747082394,
+      "learning_rate": 4.586257018966837e-06,
+      "loss": 0.7382,
+      "step": 326
+    },
+    {
+      "epoch": 2.7948717948717947,
+      "grad_norm": 0.9062791192886377,
+      "learning_rate": 4.583787824219894e-06,
+      "loss": 0.5906,
+      "step": 327
+    },
+    {
+      "epoch": 2.8034188034188032,
+      "grad_norm": 1.0120723259107143,
+      "learning_rate": 4.5813119522037765e-06,
+      "loss": 0.6833,
+      "step": 328
+    },
+    {
+      "epoch": 2.8119658119658117,
+      "grad_norm": 0.865348091293651,
+      "learning_rate": 4.578829410852145e-06,
+      "loss": 0.7787,
+      "step": 329
+    },
+    {
+      "epoch": 2.8205128205128203,
+      "grad_norm": 0.8348180991577413,
+      "learning_rate": 4.5763402081200295e-06,
+      "loss": 0.4408,
+      "step": 330
+    },
+    {
+      "epoch": 2.8290598290598292,
+      "grad_norm": 0.8011745133788588,
+      "learning_rate": 4.573844351983807e-06,
+      "loss": 0.4633,
+      "step": 331
+    },
+    {
+      "epoch": 2.8376068376068377,
+      "grad_norm": 0.8293170512614328,
+      "learning_rate": 4.571341850441175e-06,
+      "loss": 0.4687,
+      "step": 332
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.8232430705506593,
+      "learning_rate": 4.568832711511125e-06,
+      "loss": 0.6406,
+      "step": 333
+    },
+    {
+      "epoch": 2.8547008547008548,
+      "grad_norm": 0.6611830954064016,
+      "learning_rate": 4.566316943233916e-06,
+      "loss": 0.5333,
+      "step": 334
+    },
+    {
+      "epoch": 2.8632478632478633,
+      "grad_norm": 0.7427595171913142,
+      "learning_rate": 4.56379455367105e-06,
+      "loss": 0.4849,
+      "step": 335
+    },
+    {
+      "epoch": 2.871794871794872,
+      "grad_norm": 0.776740740513331,
+      "learning_rate": 4.561265550905251e-06,
+      "loss": 0.3279,
+      "step": 336
+    },
+    {
+      "epoch": 2.8803418803418803,
+      "grad_norm": 0.8914208083154344,
+      "learning_rate": 4.558729943040427e-06,
+      "loss": 0.5588,
+      "step": 337
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.9028547313318527,
+      "learning_rate": 4.556187738201656e-06,
+      "loss": 0.5921,
+      "step": 338
+    },
+    {
+      "epoch": 2.8974358974358974,
+      "grad_norm": 0.9286852889816591,
+      "learning_rate": 4.553638944535155e-06,
+      "loss": 0.5297,
+      "step": 339
+    },
+    {
+      "epoch": 2.905982905982906,
+      "grad_norm": 0.8351144281666316,
+      "learning_rate": 4.551083570208251e-06,
+      "loss": 0.5956,
+      "step": 340
+    },
+    {
+      "epoch": 2.9145299145299144,
+      "grad_norm": 0.7309337413729242,
+      "learning_rate": 4.548521623409364e-06,
+      "loss": 0.4151,
+      "step": 341
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.9346639117201893,
+      "learning_rate": 4.545953112347967e-06,
+      "loss": 0.5964,
+      "step": 342
+    },
+    {
+      "epoch": 2.931623931623932,
+      "grad_norm": 0.81834023562854,
+      "learning_rate": 4.543378045254575e-06,
+      "loss": 0.5455,
+      "step": 343
+    },
+    {
+      "epoch": 2.9401709401709404,
+      "grad_norm": 0.9994568354494042,
+      "learning_rate": 4.540796430380706e-06,
+      "loss": 0.4972,
+      "step": 344
+    },
+    {
+      "epoch": 2.948717948717949,
+      "grad_norm": 0.9623762851964687,
+      "learning_rate": 4.538208275998861e-06,
+      "loss": 0.6229,
+      "step": 345
+    },
+    {
+      "epoch": 2.9572649572649574,
+      "grad_norm": 0.8503131665697095,
+      "learning_rate": 4.535613590402497e-06,
+      "loss": 0.5439,
+      "step": 346
+    },
+    {
+      "epoch": 2.965811965811966,
+      "grad_norm": 0.8914338237199254,
+      "learning_rate": 4.533012381905999e-06,
+      "loss": 0.4731,
+      "step": 347
+    },
+    {
+      "epoch": 2.9743589743589745,
+      "grad_norm": 0.8050974375681713,
+      "learning_rate": 4.530404658844654e-06,
+      "loss": 0.3666,
+      "step": 348
+    },
+    {
+      "epoch": 2.982905982905983,
+      "grad_norm": 0.9766521405817877,
+      "learning_rate": 4.527790429574623e-06,
+      "loss": 0.6397,
+      "step": 349
+    },
+    {
+      "epoch": 2.9914529914529915,
+      "grad_norm": 1.0716768145207356,
+      "learning_rate": 4.525169702472917e-06,
+      "loss": 0.3234,
+      "step": 350
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7414645916740304,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.4125,
+      "step": 351
+    },
+    {
+      "epoch": 3.0085470085470085,
+      "grad_norm": 0.8304413817295728,
+      "learning_rate": 4.519908788386605e-06,
+      "loss": 0.4659,
+      "step": 352
+    },
+    {
+      "epoch": 3.017094017094017,
+      "grad_norm": 0.7735458981648151,
+      "learning_rate": 4.51726861826002e-06,
+      "loss": 0.5033,
+      "step": 353
+    },
+    {
+      "epoch": 3.0256410256410255,
+      "grad_norm": 0.7595322185814292,
+      "learning_rate": 4.514621984017748e-06,
+      "loss": 0.303,
+      "step": 354
+    },
+    {
+      "epoch": 3.034188034188034,
+      "grad_norm": 0.7875034261369316,
+      "learning_rate": 4.511968894140639e-06,
+      "loss": 0.648,
+      "step": 355
+    },
+    {
+      "epoch": 3.0427350427350426,
+      "grad_norm": 0.8203545208978309,
+      "learning_rate": 4.509309357130227e-06,
+      "loss": 0.4485,
+      "step": 356
+    },
+    {
+      "epoch": 3.051282051282051,
+      "grad_norm": 0.7946456857985079,
+      "learning_rate": 4.5066433815087076e-06,
+      "loss": 0.4074,
+      "step": 357
+    },
+    {
+      "epoch": 3.0598290598290596,
+      "grad_norm": 1.2003143687959523,
+      "learning_rate": 4.503970975818905e-06,
+      "loss": 0.4335,
+      "step": 358
+    },
+    {
+      "epoch": 3.0683760683760686,
+      "grad_norm": 0.9408760380102702,
+      "learning_rate": 4.501292148624251e-06,
+      "loss": 0.4196,
+      "step": 359
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.9144025827497951,
+      "learning_rate": 4.498606908508754e-06,
+      "loss": 0.4479,
+      "step": 360
+    },
+    {
+      "epoch": 3.0854700854700856,
+      "grad_norm": 0.7687844481709335,
+      "learning_rate": 4.495915264076967e-06,
+      "loss": 0.4121,
+      "step": 361
+    },
+    {
+      "epoch": 3.094017094017094,
+      "grad_norm": 0.7537385468723475,
+      "learning_rate": 4.493217223953974e-06,
+      "loss": 0.4461,
+      "step": 362
+    },
+    {
+      "epoch": 3.1025641025641026,
+      "grad_norm": 0.7718050015382629,
+      "learning_rate": 4.490512796785344e-06,
+      "loss": 0.3579,
+      "step": 363
+    },
+    {
+      "epoch": 3.111111111111111,
+      "grad_norm": 0.8146700557707305,
+      "learning_rate": 4.48780199123712e-06,
+      "loss": 0.3657,
+      "step": 364
+    },
+    {
+      "epoch": 3.1196581196581197,
+      "grad_norm": 0.7766206702534757,
+      "learning_rate": 4.485084815995778e-06,
+      "loss": 0.3605,
+      "step": 365
+    },
+    {
+      "epoch": 3.128205128205128,
+      "grad_norm": 0.9320602711654982,
+      "learning_rate": 4.482361279768209e-06,
+      "loss": 0.4754,
+      "step": 366
+    },
+    {
+      "epoch": 3.1367521367521367,
+      "grad_norm": 1.002563816761898,
+      "learning_rate": 4.479631391281685e-06,
+      "loss": 0.725,
+      "step": 367
+    },
+    {
+      "epoch": 3.1452991452991452,
+      "grad_norm": 0.8764474287712758,
+      "learning_rate": 4.476895159283835e-06,
+      "loss": 0.6423,
+      "step": 368
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.8441568311922192,
+      "learning_rate": 4.474152592542613e-06,
+      "loss": 0.5277,
+      "step": 369
+    },
+    {
+      "epoch": 3.1623931623931623,
+      "grad_norm": 0.8743031497554583,
+      "learning_rate": 4.4714036998462715e-06,
+      "loss": 0.5851,
+      "step": 370
+    },
+    {
+      "epoch": 3.1709401709401708,
+      "grad_norm": 0.933207995427482,
+      "learning_rate": 4.4686484900033375e-06,
+      "loss": 0.433,
+      "step": 371
+    },
+    {
+      "epoch": 3.1794871794871793,
+      "grad_norm": 0.9190018249217231,
+      "learning_rate": 4.465886971842578e-06,
+      "loss": 0.4419,
+      "step": 372
+    },
+    {
+      "epoch": 3.1880341880341883,
+      "grad_norm": 0.8266660789991668,
+      "learning_rate": 4.463119154212972e-06,
+      "loss": 0.4605,
+      "step": 373
+    },
+    {
+      "epoch": 3.1965811965811968,
+      "grad_norm": 0.76565165126929,
+      "learning_rate": 4.46034504598369e-06,
+      "loss": 0.5145,
+      "step": 374
+    },
+    {
+      "epoch": 3.2051282051282053,
+      "grad_norm": 0.8511966627383084,
+      "learning_rate": 4.457564656044056e-06,
+      "loss": 0.3751,
+      "step": 375
+    },
+    {
+      "epoch": 3.213675213675214,
+      "grad_norm": 0.8298685003297344,
+      "learning_rate": 4.454777993303524e-06,
+      "loss": 0.7063,
+      "step": 376
+    },
+    {
+      "epoch": 3.2222222222222223,
+      "grad_norm": 0.8842731250398438,
+      "learning_rate": 4.451985066691649e-06,
+      "loss": 0.3819,
+      "step": 377
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.7150831598781322,
+      "learning_rate": 4.449185885158056e-06,
+      "loss": 0.3984,
+      "step": 378
+    },
+    {
+      "epoch": 3.2393162393162394,
+      "grad_norm": 0.8069520223481081,
+      "learning_rate": 4.446380457672417e-06,
+      "loss": 0.3641,
+      "step": 379
+    },
+    {
+      "epoch": 3.247863247863248,
+      "grad_norm": 0.8676544763614075,
+      "learning_rate": 4.443568793224415e-06,
+      "loss": 0.5014,
+      "step": 380
+    },
+    {
+      "epoch": 3.2564102564102564,
+      "grad_norm": 0.8556378002114667,
+      "learning_rate": 4.44075090082372e-06,
+      "loss": 0.4846,
+      "step": 381
+    },
+    {
+      "epoch": 3.264957264957265,
+      "grad_norm": 0.889152669192053,
+      "learning_rate": 4.437926789499959e-06,
+      "loss": 0.5168,
+      "step": 382
+    },
+    {
+      "epoch": 3.2735042735042734,
+      "grad_norm": 0.6725425463954209,
+      "learning_rate": 4.435096468302687e-06,
+      "loss": 0.308,
+      "step": 383
+    },
+    {
+      "epoch": 3.282051282051282,
+      "grad_norm": 0.8877402712011035,
+      "learning_rate": 4.432259946301355e-06,
+      "loss": 0.4128,
+      "step": 384
+    },
+    {
+      "epoch": 3.2905982905982905,
+      "grad_norm": 0.8445915729478649,
+      "learning_rate": 4.429417232585288e-06,
+      "loss": 0.3867,
+      "step": 385
+    },
+    {
+      "epoch": 3.299145299145299,
+      "grad_norm": 0.7327654841257518,
+      "learning_rate": 4.42656833626365e-06,
+      "loss": 0.2801,
+      "step": 386
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.9029002294542466,
+      "learning_rate": 4.423713266465415e-06,
+      "loss": 0.4407,
+      "step": 387
+    },
+    {
+      "epoch": 3.316239316239316,
+      "grad_norm": 0.8184482649220772,
+      "learning_rate": 4.4208520323393425e-06,
+      "loss": 0.3226,
+      "step": 388
+    },
+    {
+      "epoch": 3.324786324786325,
+      "grad_norm": 0.9024192814094689,
+      "learning_rate": 4.417984643053941e-06,
+      "loss": 0.6365,
+      "step": 389
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.7111467138011386,
+      "learning_rate": 4.415111107797445e-06,
+      "loss": 0.4378,
+      "step": 390
+    },
+    {
+      "epoch": 3.341880341880342,
+      "grad_norm": 0.9441614638935245,
+      "learning_rate": 4.412231435777784e-06,
+      "loss": 0.4225,
+      "step": 391
+    },
+    {
+      "epoch": 3.3504273504273505,
+      "grad_norm": 2.221227624135174,
+      "learning_rate": 4.409345636222549e-06,
+      "loss": 0.5193,
+      "step": 392
+    },
+    {
+      "epoch": 3.358974358974359,
+      "grad_norm": 0.8877919988116371,
+      "learning_rate": 4.406453718378968e-06,
+      "loss": 0.7117,
+      "step": 393
+    },
+    {
+      "epoch": 3.3675213675213675,
+      "grad_norm": 0.8771932944932836,
+      "learning_rate": 4.4035556915138745e-06,
+      "loss": 0.5035,
+      "step": 394
+    },
+    {
+      "epoch": 3.376068376068376,
+      "grad_norm": 0.9387814704326412,
+      "learning_rate": 4.400651564913676e-06,
+      "loss": 0.5797,
+      "step": 395
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.8244733294329837,
+      "learning_rate": 4.397741347884329e-06,
+      "loss": 0.5068,
+      "step": 396
+    },
+    {
+      "epoch": 3.393162393162393,
+      "grad_norm": 0.9494021493478192,
+      "learning_rate": 4.394825049751303e-06,
+      "loss": 0.5993,
+      "step": 397
+    },
+    {
+      "epoch": 3.4017094017094016,
+      "grad_norm": 0.7221780905312348,
+      "learning_rate": 4.391902679859557e-06,
+      "loss": 0.251,
+      "step": 398
+    },
+    {
+      "epoch": 3.41025641025641,
+      "grad_norm": 0.8268617275414766,
+      "learning_rate": 4.388974247573501e-06,
+      "loss": 0.4459,
+      "step": 399
+    },
+    {
+      "epoch": 3.4188034188034186,
+      "grad_norm": 1.1369326581432304,
+      "learning_rate": 4.386039762276976e-06,
+      "loss": 0.4272,
+      "step": 400
+    },
+    {
+      "epoch": 3.427350427350427,
+      "grad_norm": 0.8576090900488215,
+      "learning_rate": 4.3830992333732185e-06,
+      "loss": 0.4957,
+      "step": 401
+    },
+    {
+      "epoch": 3.435897435897436,
+      "grad_norm": 0.8819239615707368,
+      "learning_rate": 4.3801526702848306e-06,
+      "loss": 0.5069,
+      "step": 402
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 0.7180964553933423,
+      "learning_rate": 4.377200082453748e-06,
+      "loss": 0.4055,
+      "step": 403
+    },
+    {
+      "epoch": 3.452991452991453,
+      "grad_norm": 0.8424182432690462,
+      "learning_rate": 4.374241479341216e-06,
+      "loss": 0.3744,
+      "step": 404
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.8448473931955931,
+      "learning_rate": 4.3712768704277535e-06,
+      "loss": 0.4433,
+      "step": 405
+    },
+    {
+      "epoch": 3.47008547008547,
+      "grad_norm": 0.8768307747327547,
+      "learning_rate": 4.368306265213122e-06,
+      "loss": 0.4239,
+      "step": 406
+    },
+    {
+      "epoch": 3.4786324786324787,
+      "grad_norm": 0.7559664092738253,
+      "learning_rate": 4.365329673216301e-06,
+      "loss": 0.3114,
+      "step": 407
+    },
+    {
+      "epoch": 3.4871794871794872,
+      "grad_norm": 0.7811272037089495,
+      "learning_rate": 4.3623471039754525e-06,
+      "loss": 0.6167,
+      "step": 408
+    },
+    {
+      "epoch": 3.4957264957264957,
+      "grad_norm": 0.8050998843783346,
+      "learning_rate": 4.359358567047892e-06,
+      "loss": 0.4296,
+      "step": 409
+    },
+    {
+      "epoch": 3.5042735042735043,
+      "grad_norm": 0.764913405706021,
+      "learning_rate": 4.356364072010059e-06,
+      "loss": 0.5632,
+      "step": 410
+    },
+    {
+      "epoch": 3.5128205128205128,
+      "grad_norm": 1.0200860018866134,
+      "learning_rate": 4.35336362845748e-06,
+      "loss": 0.7716,
+      "step": 411
+    },
+    {
+      "epoch": 3.5213675213675213,
+      "grad_norm": 0.8071339648683941,
+      "learning_rate": 4.35035724600475e-06,
+      "loss": 0.6755,
+      "step": 412
+    },
+    {
+      "epoch": 3.52991452991453,
+      "grad_norm": 0.8934834178549457,
+      "learning_rate": 4.347344934285492e-06,
+      "loss": 0.5716,
+      "step": 413
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.9013447212285678,
+      "learning_rate": 4.3443267029523265e-06,
+      "loss": 0.4951,
+      "step": 414
+    },
+    {
+      "epoch": 3.547008547008547,
+      "grad_norm": 0.7467615404534375,
+      "learning_rate": 4.3413025616768426e-06,
+      "loss": 0.5141,
+      "step": 415
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 0.8109187006489191,
+      "learning_rate": 4.338272520149572e-06,
+      "loss": 0.4856,
+      "step": 416
+    },
+    {
+      "epoch": 3.564102564102564,
+      "grad_norm": 0.7978268359836372,
+      "learning_rate": 4.335236588079949e-06,
+      "loss": 0.4645,
+      "step": 417
+    },
+    {
+      "epoch": 3.5726495726495724,
+      "grad_norm": 0.7976611760330177,
+      "learning_rate": 4.332194775196282e-06,
+      "loss": 0.5088,
+      "step": 418
+    },
+    {
+      "epoch": 3.5811965811965814,
+      "grad_norm": 0.7273065066610752,
+      "learning_rate": 4.329147091245729e-06,
+      "loss": 0.3724,
+      "step": 419
+    },
+    {
+      "epoch": 3.58974358974359,
+      "grad_norm": 0.7627333803959988,
+      "learning_rate": 4.326093545994258e-06,
+      "loss": 0.3721,
+      "step": 420
+    },
+    {
+      "epoch": 3.5982905982905984,
+      "grad_norm": 0.8883554964464617,
+      "learning_rate": 4.3230341492266195e-06,
+      "loss": 0.3726,
+      "step": 421
+    },
+    {
+      "epoch": 3.606837606837607,
+      "grad_norm": 0.766873829622007,
+      "learning_rate": 4.3199689107463125e-06,
+      "loss": 0.3011,
+      "step": 422
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 1.050080591838294,
+      "learning_rate": 4.316897840375558e-06,
+      "loss": 0.4624,
+      "step": 423
+    },
+    {
+      "epoch": 3.623931623931624,
+      "grad_norm": 0.7431746820429679,
+      "learning_rate": 4.313820947955265e-06,
+      "loss": 0.4298,
+      "step": 424
+    },
+    {
+      "epoch": 3.6324786324786325,
+      "grad_norm": 0.7784354492853087,
+      "learning_rate": 4.310738243344996e-06,
+      "loss": 0.3922,
+      "step": 425
+    },
+    {
+      "epoch": 3.641025641025641,
+      "grad_norm": 0.7444834058751557,
+      "learning_rate": 4.307649736422939e-06,
+      "loss": 0.2623,
+      "step": 426
+    },
+    {
+      "epoch": 3.6495726495726495,
+      "grad_norm": 0.7977973139058897,
+      "learning_rate": 4.304555437085876e-06,
+      "loss": 0.373,
+      "step": 427
+    },
+    {
+      "epoch": 3.658119658119658,
+      "grad_norm": 0.8485658441757654,
+      "learning_rate": 4.301455355249148e-06,
+      "loss": 0.5303,
+      "step": 428
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 0.8224556079644156,
+      "learning_rate": 4.2983495008466285e-06,
+      "loss": 0.3876,
+      "step": 429
+    },
+    {
+      "epoch": 3.6752136752136755,
+      "grad_norm": 0.7754703882264418,
+      "learning_rate": 4.2952378838306855e-06,
+      "loss": 0.4484,
+      "step": 430
+    },
+    {
+      "epoch": 3.683760683760684,
+      "grad_norm": 0.8570419199997498,
+      "learning_rate": 4.292120514172154e-06,
+      "loss": 0.4782,
+      "step": 431
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.7756027229434488,
+      "learning_rate": 4.288997401860303e-06,
+      "loss": 0.3856,
+      "step": 432
+    },
+    {
+      "epoch": 3.700854700854701,
+      "grad_norm": 0.8073968441637798,
+      "learning_rate": 4.285868556902803e-06,
+      "loss": 0.4776,
+      "step": 433
+    },
+    {
+      "epoch": 3.7094017094017095,
+      "grad_norm": 0.8790251187999025,
+      "learning_rate": 4.2827339893256935e-06,
+      "loss": 0.4384,
+      "step": 434
+    },
+    {
+      "epoch": 3.717948717948718,
+      "grad_norm": 0.8528022888286432,
+      "learning_rate": 4.279593709173352e-06,
+      "loss": 0.4662,
+      "step": 435
+    },
+    {
+      "epoch": 3.7264957264957266,
+      "grad_norm": 0.8226819943495567,
+      "learning_rate": 4.276447726508461e-06,
+      "loss": 0.5625,
+      "step": 436
+    },
+    {
+      "epoch": 3.735042735042735,
+      "grad_norm": 0.85329699161591,
+      "learning_rate": 4.273296051411978e-06,
+      "loss": 0.548,
+      "step": 437
+    },
+    {
+      "epoch": 3.7435897435897436,
+      "grad_norm": 0.8815606194105096,
+      "learning_rate": 4.2701386939830966e-06,
+      "loss": 0.6318,
+      "step": 438
+    },
+    {
+      "epoch": 3.752136752136752,
+      "grad_norm": 0.9148527978934294,
+      "learning_rate": 4.2669756643392255e-06,
+      "loss": 0.6247,
+      "step": 439
+    },
+    {
+      "epoch": 3.7606837606837606,
+      "grad_norm": 0.7579395274439618,
+      "learning_rate": 4.263806972615943e-06,
+      "loss": 0.3085,
+      "step": 440
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.8419223552392247,
+      "learning_rate": 4.260632628966974e-06,
+      "loss": 0.3967,
+      "step": 441
+    },
+    {
+      "epoch": 3.7777777777777777,
+      "grad_norm": 0.7554240130558576,
+      "learning_rate": 4.257452643564155e-06,
+      "loss": 0.398,
+      "step": 442
+    },
+    {
+      "epoch": 3.786324786324786,
+      "grad_norm": 0.8507117836828787,
+      "learning_rate": 4.254267026597399e-06,
+      "loss": 0.5881,
+      "step": 443
+    },
+    {
+      "epoch": 3.7948717948717947,
+      "grad_norm": 0.7353099260117935,
+      "learning_rate": 4.251075788274667e-06,
+      "loss": 0.2005,
+      "step": 444
+    },
+    {
+      "epoch": 3.8034188034188032,
+      "grad_norm": 0.8712901741487685,
+      "learning_rate": 4.247878938821929e-06,
+      "loss": 0.4499,
+      "step": 445
+    },
+    {
+      "epoch": 3.8119658119658117,
+      "grad_norm": 0.8974249580810392,
+      "learning_rate": 4.2446764884831404e-06,
+      "loss": 0.5255,
+      "step": 446
+    },
+    {
+      "epoch": 3.8205128205128203,
+      "grad_norm": 0.8829948352253479,
+      "learning_rate": 4.2414684475202014e-06,
+      "loss": 0.4319,
+      "step": 447
+    },
+    {
+      "epoch": 3.8290598290598292,
+      "grad_norm": 0.8605823079252658,
+      "learning_rate": 4.238254826212925e-06,
+      "loss": 0.4846,
+      "step": 448
+    },
+    {
+      "epoch": 3.8376068376068377,
+      "grad_norm": 0.8041945944691401,
+      "learning_rate": 4.2350356348590096e-06,
+      "loss": 0.4267,
+      "step": 449
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.870824196997254,
+      "learning_rate": 4.231810883773999e-06,
+      "loss": 0.5941,
+      "step": 450
+    },
+    {
+      "epoch": 3.8547008547008548,
+      "grad_norm": 0.9593235942305645,
+      "learning_rate": 4.228580583291254e-06,
+      "loss": 0.4393,
+      "step": 451
+    },
+    {
+      "epoch": 3.8632478632478633,
+      "grad_norm": 1.0063955556890787,
+      "learning_rate": 4.225344743761918e-06,
+      "loss": 0.4246,
+      "step": 452
+    },
+    {
+      "epoch": 3.871794871794872,
+      "grad_norm": 0.8202784309796189,
+      "learning_rate": 4.2221033755548835e-06,
+      "loss": 0.3854,
+      "step": 453
+    },
+    {
+      "epoch": 3.8803418803418803,
+      "grad_norm": 1.0562117712056716,
+      "learning_rate": 4.218856489056758e-06,
+      "loss": 0.2155,
+      "step": 454
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 0.8464596519960794,
+      "learning_rate": 4.215604094671835e-06,
+      "loss": 0.5801,
+      "step": 455
+    },
+    {
+      "epoch": 3.8974358974358974,
+      "grad_norm": 0.8504550548583248,
+      "learning_rate": 4.2123462028220505e-06,
+      "loss": 0.5785,
+      "step": 456
+    },
+    {
+      "epoch": 3.905982905982906,
+      "grad_norm": 0.9630832827635597,
+      "learning_rate": 4.209082823946965e-06,
+      "loss": 0.7302,
+      "step": 457
+    },
+    {
+      "epoch": 3.9145299145299144,
+      "grad_norm": 0.8968524846983676,
+      "learning_rate": 4.205813968503717e-06,
+      "loss": 0.4906,
+      "step": 458
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.8547398378994598,
+      "learning_rate": 4.202539646966993e-06,
+      "loss": 0.414,
+      "step": 459
+    },
+    {
+      "epoch": 3.931623931623932,
+      "grad_norm": 0.6844619258431609,
+      "learning_rate": 4.1992598698289985e-06,
+      "loss": 0.3025,
+      "step": 460
+    },
+    {
+      "epoch": 3.9401709401709404,
+      "grad_norm": 0.8949935454247467,
+      "learning_rate": 4.1959746475994175e-06,
+      "loss": 0.5164,
+      "step": 461
+    },
+    {
+      "epoch": 3.948717948717949,
+      "grad_norm": 0.8379831635152304,
+      "learning_rate": 4.1926839908053855e-06,
+      "loss": 0.4645,
+      "step": 462
+    },
+    {
+      "epoch": 3.9572649572649574,
+      "grad_norm": 0.762265401669729,
+      "learning_rate": 4.189387909991448e-06,
+      "loss": 0.4256,
+      "step": 463
+    },
+    {
+      "epoch": 3.965811965811966,
+      "grad_norm": 0.9100475030814085,
+      "learning_rate": 4.186086415719537e-06,
+      "loss": 0.4741,
+      "step": 464
+    },
+    {
+      "epoch": 3.9743589743589745,
+      "grad_norm": 0.8453866689986883,
+      "learning_rate": 4.182779518568925e-06,
+      "loss": 0.504,
+      "step": 465
+    },
+    {
+      "epoch": 3.982905982905983,
+      "grad_norm": 0.8103875920447905,
+      "learning_rate": 4.179467229136205e-06,
+      "loss": 0.2742,
+      "step": 466
+    },
+    {
+      "epoch": 3.9914529914529915,
+      "grad_norm": 1.1325517449404694,
+      "learning_rate": 4.176149558035241e-06,
+      "loss": 0.5041,
+      "step": 467
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.8145184159975724,
+      "learning_rate": 4.172826515897146e-06,
+      "loss": 0.3232,
+      "step": 468
+    },
+    {
+      "epoch": 4.0085470085470085,
+      "grad_norm": 0.8456967300947733,
+      "learning_rate": 4.169498113370245e-06,
+      "loss": 0.5478,
+      "step": 469
+    },
+    {
+      "epoch": 4.017094017094017,
+      "grad_norm": 0.7549586169194719,
+      "learning_rate": 4.166164361120036e-06,
+      "loss": 0.2411,
+      "step": 470
+    },
+    {
+      "epoch": 4.0256410256410255,
+      "grad_norm": 0.8089576849557826,
+      "learning_rate": 4.162825269829165e-06,
+      "loss": 0.4338,
+      "step": 471
+    },
+    {
+      "epoch": 4.034188034188034,
+      "grad_norm": 0.6762049989116309,
+      "learning_rate": 4.15948085019738e-06,
+      "loss": 0.2678,
+      "step": 472
+    },
+    {
+      "epoch": 4.042735042735043,
+      "grad_norm": 0.8918685147268757,
+      "learning_rate": 4.156131112941509e-06,
+      "loss": 0.4549,
+      "step": 473
+    },
+    {
+      "epoch": 4.051282051282051,
+      "grad_norm": 1.028450458974256,
+      "learning_rate": 4.152776068795416e-06,
+      "loss": 0.5193,
+      "step": 474
+    },
+    {
+      "epoch": 4.05982905982906,
+      "grad_norm": 0.8593042738505275,
+      "learning_rate": 4.149415728509971e-06,
+      "loss": 0.4555,
+      "step": 475
+    },
+    {
+      "epoch": 4.068376068376068,
+      "grad_norm": 0.8941008995651688,
+      "learning_rate": 4.146050102853015e-06,
+      "loss": 0.3665,
+      "step": 476
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.9327966593805307,
+      "learning_rate": 4.1426792026093274e-06,
+      "loss": 0.6724,
+      "step": 477
+    },
+    {
+      "epoch": 4.085470085470085,
+      "grad_norm": 0.8100547548689432,
+      "learning_rate": 4.139303038580586e-06,
+      "loss": 0.3283,
+      "step": 478
+    },
+    {
+      "epoch": 4.094017094017094,
+      "grad_norm": 0.719264457249393,
+      "learning_rate": 4.135921621585338e-06,
+      "loss": 0.2572,
+      "step": 479
+    },
+    {
+      "epoch": 4.102564102564102,
+      "grad_norm": 0.8641442922634487,
+      "learning_rate": 4.1325349624589625e-06,
+      "loss": 0.4605,
+      "step": 480
+    },
+    {
+      "epoch": 4.111111111111111,
+      "grad_norm": 0.8430072813650188,
+      "learning_rate": 4.129143072053639e-06,
+      "loss": 0.3969,
+      "step": 481
+    },
+    {
+      "epoch": 4.119658119658119,
+      "grad_norm": 0.7899792800908239,
+      "learning_rate": 4.125745961238305e-06,
+      "loss": 0.2622,
+      "step": 482
+    },
+    {
+      "epoch": 4.128205128205128,
+      "grad_norm": 0.8939354624686937,
+      "learning_rate": 4.122343640898628e-06,
+      "loss": 0.4555,
+      "step": 483
+    },
+    {
+      "epoch": 4.136752136752137,
+      "grad_norm": 0.8543201855817723,
+      "learning_rate": 4.118936121936973e-06,
+      "loss": 0.4876,
+      "step": 484
+    },
+    {
+      "epoch": 4.145299145299146,
+      "grad_norm": 0.6680196413010284,
+      "learning_rate": 4.115523415272358e-06,
+      "loss": 0.3388,
+      "step": 485
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.9087424068651639,
+      "learning_rate": 4.112105531840427e-06,
+      "loss": 0.6114,
+      "step": 486
+    },
+    {
+      "epoch": 4.162393162393163,
+      "grad_norm": 0.8714388993623177,
+      "learning_rate": 4.1086824825934126e-06,
+      "loss": 0.2663,
+      "step": 487
+    },
+    {
+      "epoch": 4.170940170940171,
+      "grad_norm": 0.9931837553159588,
+      "learning_rate": 4.1052542785001e-06,
+      "loss": 0.264,
+      "step": 488
+    },
+    {
+      "epoch": 4.17948717948718,
+      "grad_norm": 0.9239576219551864,
+      "learning_rate": 4.101820930545792e-06,
+      "loss": 0.4188,
+      "step": 489
+    },
+    {
+      "epoch": 4.188034188034188,
+      "grad_norm": 0.8835490304209346,
+      "learning_rate": 4.098382449732276e-06,
+      "loss": 0.4669,
+      "step": 490
+    },
+    {
+      "epoch": 4.196581196581197,
+      "grad_norm": 1.0141618276410034,
+      "learning_rate": 4.094938847077784e-06,
+      "loss": 0.5037,
+      "step": 491
+    },
+    {
+      "epoch": 4.205128205128205,
+      "grad_norm": 0.8527934027685333,
+      "learning_rate": 4.091490133616965e-06,
+      "loss": 0.22,
+      "step": 492
+    },
+    {
+      "epoch": 4.213675213675214,
+      "grad_norm": 0.8461432923951197,
+      "learning_rate": 4.08803632040084e-06,
+      "loss": 0.2893,
+      "step": 493
+    },
+    {
+      "epoch": 4.222222222222222,
+      "grad_norm": 0.741023345781096,
+      "learning_rate": 4.084577418496775e-06,
+      "loss": 0.3295,
+      "step": 494
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.7688995720250164,
+      "learning_rate": 4.081113438988443e-06,
+      "loss": 0.2804,
+      "step": 495
+    },
+    {
+      "epoch": 4.239316239316239,
+      "grad_norm": 0.9418546280795675,
+      "learning_rate": 4.077644392975785e-06,
+      "loss": 0.2362,
+      "step": 496
+    },
+    {
+      "epoch": 4.247863247863248,
+      "grad_norm": 0.8808691764925719,
+      "learning_rate": 4.074170291574975e-06,
+      "loss": 0.4459,
+      "step": 497
+    },
+    {
+      "epoch": 4.256410256410256,
+      "grad_norm": 0.8611529751464202,
+      "learning_rate": 4.0706911459183915e-06,
+      "loss": 0.4481,
+      "step": 498
+    },
+    {
+      "epoch": 4.264957264957265,
+      "grad_norm": 0.689017158073508,
+      "learning_rate": 4.067206967154575e-06,
+      "loss": 0.2915,
+      "step": 499
+    },
+    {
+      "epoch": 4.273504273504273,
+      "grad_norm": 0.8342711752591144,
+      "learning_rate": 4.063717766448194e-06,
+      "loss": 0.3025,
+      "step": 500
+    },
+    {
+      "epoch": 4.282051282051282,
+      "grad_norm": 0.7552302159412333,
+      "learning_rate": 4.060223554980007e-06,
+      "loss": 0.4213,
+      "step": 501
+    },
+    {
+      "epoch": 4.2905982905982905,
+      "grad_norm": 0.7055560735263742,
+      "learning_rate": 4.056724343946832e-06,
+      "loss": 0.168,
+      "step": 502
+    },
+    {
+      "epoch": 4.299145299145299,
+      "grad_norm": 0.8345661950385564,
+      "learning_rate": 4.053220144561506e-06,
+      "loss": 0.2601,
+      "step": 503
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.7767602930561285,
+      "learning_rate": 4.049710968052851e-06,
+      "loss": 0.4115,
+      "step": 504
+    },
+    {
+      "epoch": 4.316239316239316,
+      "grad_norm": 0.7597842473349811,
+      "learning_rate": 4.046196825665638e-06,
+      "loss": 0.3013,
+      "step": 505
+    },
+    {
+      "epoch": 4.3247863247863245,
+      "grad_norm": 0.8218417665919933,
+      "learning_rate": 4.042677728660549e-06,
+      "loss": 0.2268,
+      "step": 506
+    },
+    {
+      "epoch": 4.333333333333333,
+      "grad_norm": 0.8387501902514646,
+      "learning_rate": 4.039153688314146e-06,
+      "loss": 0.4442,
+      "step": 507
+    },
+    {
+      "epoch": 4.3418803418803416,
+      "grad_norm": 0.9426516698668568,
+      "learning_rate": 4.035624715918827e-06,
+      "loss": 0.4359,
+      "step": 508
+    },
+    {
+      "epoch": 4.35042735042735,
+      "grad_norm": 0.8098999588083045,
+      "learning_rate": 4.032090822782798e-06,
+      "loss": 0.3859,
+      "step": 509
+    },
+    {
+      "epoch": 4.358974358974359,
+      "grad_norm": 0.8245264905162911,
+      "learning_rate": 4.028552020230031e-06,
+      "loss": 0.29,
+      "step": 510
+    },
+    {
+      "epoch": 4.367521367521368,
+      "grad_norm": 0.8375221036962769,
+      "learning_rate": 4.0250083196002285e-06,
+      "loss": 0.3568,
+      "step": 511
+    },
+    {
+      "epoch": 4.3760683760683765,
+      "grad_norm": 0.7413180645147043,
+      "learning_rate": 4.021459732248792e-06,
+      "loss": 0.2319,
+      "step": 512
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 6.263164745538754,
+      "learning_rate": 4.017906269546778e-06,
+      "loss": 0.2853,
+      "step": 513
+    },
+    {
+      "epoch": 4.3931623931623935,
+      "grad_norm": 0.9453800593068873,
+      "learning_rate": 4.014347942880869e-06,
+      "loss": 0.2482,
+      "step": 514
+    },
+    {
+      "epoch": 4.401709401709402,
+      "grad_norm": 0.803412607202504,
+      "learning_rate": 4.0107847636533314e-06,
+      "loss": 0.274,
+      "step": 515
+    },
+    {
+      "epoch": 4.410256410256411,
+      "grad_norm": 0.6071379342770246,
+      "learning_rate": 4.0072167432819804e-06,
+      "loss": 0.2204,
+      "step": 516
+    },
+    {
+      "epoch": 4.418803418803419,
+      "grad_norm": 0.8526809501053975,
+      "learning_rate": 4.003643893200148e-06,
+      "loss": 0.3746,
+      "step": 517
+    },
+    {
+      "epoch": 4.427350427350428,
+      "grad_norm": 1.1170624575831807,
+      "learning_rate": 4.000066224856636e-06,
+      "loss": 0.4389,
+      "step": 518
+    },
+    {
+      "epoch": 4.435897435897436,
+      "grad_norm": 0.8714209677695344,
+      "learning_rate": 3.996483749715694e-06,
+      "loss": 0.3663,
+      "step": 519
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.8681016473178148,
+      "learning_rate": 3.992896479256966e-06,
+      "loss": 0.3868,
+      "step": 520
+    },
+    {
+      "epoch": 4.452991452991453,
+      "grad_norm": 0.8431474276532333,
+      "learning_rate": 3.989304424975468e-06,
+      "loss": 0.2684,
+      "step": 521
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.8768461034809396,
+      "learning_rate": 3.985707598381544e-06,
+      "loss": 0.439,
+      "step": 522
+    },
+    {
+      "epoch": 4.47008547008547,
+      "grad_norm": 0.8417846294982461,
+      "learning_rate": 3.9821060110008295e-06,
+      "loss": 0.422,
+      "step": 523
+    },
+    {
+      "epoch": 4.478632478632479,
+      "grad_norm": 0.8997660894045065,
+      "learning_rate": 3.978499674374214e-06,
+      "loss": 0.4011,
+      "step": 524
+    },
+    {
+      "epoch": 4.487179487179487,
+      "grad_norm": 0.8281409585698364,
+      "learning_rate": 3.974888600057808e-06,
+      "loss": 0.2545,
+      "step": 525
+    },
+    {
+      "epoch": 4.495726495726496,
+      "grad_norm": 0.9489099979655634,
+      "learning_rate": 3.971272799622903e-06,
+      "loss": 0.4931,
+      "step": 526
+    },
+    {
+      "epoch": 4.504273504273504,
+      "grad_norm": 0.8608170335327134,
+      "learning_rate": 3.967652284655933e-06,
+      "loss": 0.2562,
+      "step": 527
+    },
+    {
+      "epoch": 4.512820512820513,
+      "grad_norm": 0.8290547376675292,
+      "learning_rate": 3.964027066758442e-06,
+      "loss": 0.2171,
+      "step": 528
+    },
+    {
+      "epoch": 4.521367521367521,
+      "grad_norm": 0.8608815039412893,
+      "learning_rate": 3.960397157547043e-06,
+      "loss": 0.3892,
+      "step": 529
+    },
+    {
+      "epoch": 4.52991452991453,
+      "grad_norm": 0.965861951508712,
+      "learning_rate": 3.956762568653378e-06,
+      "loss": 0.4057,
+      "step": 530
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.8733549209558379,
+      "learning_rate": 3.953123311724092e-06,
+      "loss": 0.4154,
+      "step": 531
+    },
+    {
+      "epoch": 4.547008547008547,
+      "grad_norm": 0.6653677165284325,
+      "learning_rate": 3.9494793984207815e-06,
+      "loss": 0.2331,
+      "step": 532
+    },
+    {
+      "epoch": 4.555555555555555,
+      "grad_norm": 1.004281844464304,
+      "learning_rate": 3.945830840419966e-06,
+      "loss": 0.4135,
+      "step": 533
+    },
+    {
+      "epoch": 4.564102564102564,
+      "grad_norm": 0.7585768676680708,
+      "learning_rate": 3.942177649413051e-06,
+      "loss": 0.4307,
+      "step": 534
+    },
+    {
+      "epoch": 4.572649572649572,
+      "grad_norm": 0.8861421749427683,
+      "learning_rate": 3.938519837106284e-06,
+      "loss": 0.2649,
+      "step": 535
+    },
+    {
+      "epoch": 4.581196581196581,
+      "grad_norm": 0.7497323044095112,
+      "learning_rate": 3.9348574152207245e-06,
+      "loss": 0.3153,
+      "step": 536
+    },
+    {
+      "epoch": 4.589743589743589,
+      "grad_norm": 0.7485592898441834,
+      "learning_rate": 3.931190395492198e-06,
+      "loss": 0.2989,
+      "step": 537
+    },
+    {
+      "epoch": 4.598290598290598,
+      "grad_norm": 0.8109828165333799,
+      "learning_rate": 3.92751878967127e-06,
+      "loss": 0.4264,
+      "step": 538
+    },
+    {
+      "epoch": 4.6068376068376065,
+      "grad_norm": 0.7435044944321713,
+      "learning_rate": 3.923842609523195e-06,
+      "loss": 0.3741,
+      "step": 539
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.6527824018088557,
+      "learning_rate": 3.92016186682789e-06,
+      "loss": 0.2872,
+      "step": 540
+    },
+    {
+      "epoch": 4.6239316239316235,
+      "grad_norm": 0.8364632365210702,
+      "learning_rate": 3.91647657337989e-06,
+      "loss": 0.4051,
+      "step": 541
+    },
+    {
+      "epoch": 4.632478632478632,
+      "grad_norm": 0.7308772252045037,
+      "learning_rate": 3.9127867409883145e-06,
+      "loss": 0.4665,
+      "step": 542
+    },
+    {
+      "epoch": 4.641025641025641,
+      "grad_norm": 0.714687471906677,
+      "learning_rate": 3.909092381476824e-06,
+      "loss": 0.3013,
+      "step": 543
+    },
+    {
+      "epoch": 4.64957264957265,
+      "grad_norm": 0.9203048234486699,
+      "learning_rate": 3.905393506683589e-06,
+      "loss": 0.5556,
+      "step": 544
+    },
+    {
+      "epoch": 4.6581196581196584,
+      "grad_norm": 0.8296100085678216,
+      "learning_rate": 3.901690128461248e-06,
+      "loss": 0.3578,
+      "step": 545
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 1.057400490981615,
+      "learning_rate": 3.897982258676867e-06,
+      "loss": 0.4101,
+      "step": 546
+    },
+    {
+      "epoch": 4.6752136752136755,
+      "grad_norm": 1.0095630702142258,
+      "learning_rate": 3.894269909211911e-06,
+      "loss": 0.4108,
+      "step": 547
+    },
+    {
+      "epoch": 4.683760683760684,
+      "grad_norm": 0.8490424531903906,
+      "learning_rate": 3.890553091962193e-06,
+      "loss": 0.3049,
+      "step": 548
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.8836815839622476,
+      "learning_rate": 3.8868318188378475e-06,
+      "loss": 0.2335,
+      "step": 549
+    },
+    {
+      "epoch": 4.700854700854701,
+      "grad_norm": 0.7553422146759685,
+      "learning_rate": 3.883106101763285e-06,
+      "loss": 0.3286,
+      "step": 550
+    },
+    {
+      "epoch": 4.7094017094017095,
+      "grad_norm": 0.9266604794055423,
+      "learning_rate": 3.879375952677156e-06,
+      "loss": 0.3113,
+      "step": 551
+    },
+    {
+      "epoch": 4.717948717948718,
+      "grad_norm": 0.7998462296426596,
+      "learning_rate": 3.875641383532313e-06,
+      "loss": 0.3974,
+      "step": 552
+    },
+    {
+      "epoch": 4.726495726495727,
+      "grad_norm": 0.7721339739950999,
+      "learning_rate": 3.871902406295775e-06,
+      "loss": 0.2705,
+      "step": 553
+    },
+    {
+      "epoch": 4.735042735042735,
+      "grad_norm": 0.7208568188475144,
+      "learning_rate": 3.868159032948681e-06,
+      "loss": 0.3908,
+      "step": 554
+    },
+    {
+      "epoch": 4.743589743589744,
+      "grad_norm": 0.886277830497899,
+      "learning_rate": 3.8644112754862614e-06,
+      "loss": 0.4494,
+      "step": 555
+    },
+    {
+      "epoch": 4.752136752136752,
+      "grad_norm": 0.9008645068145198,
+      "learning_rate": 3.860659145917794e-06,
+      "loss": 0.3228,
+      "step": 556
+    },
+    {
+      "epoch": 4.760683760683761,
+      "grad_norm": 0.7673395207334885,
+      "learning_rate": 3.856902656266563e-06,
+      "loss": 0.3048,
+      "step": 557
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.8698675752196239,
+      "learning_rate": 3.853141818569829e-06,
+      "loss": 0.3509,
+      "step": 558
+    },
+    {
+      "epoch": 4.777777777777778,
+      "grad_norm": 1.050646507319603,
+      "learning_rate": 3.849376644878783e-06,
+      "loss": 0.2981,
+      "step": 559
+    },
+    {
+      "epoch": 4.786324786324786,
+      "grad_norm": 0.7929756502011445,
+      "learning_rate": 3.84560714725851e-06,
+      "loss": 0.3696,
+      "step": 560
+    },
+    {
+      "epoch": 4.794871794871795,
+      "grad_norm": 0.7303218759520699,
+      "learning_rate": 3.841833337787951e-06,
+      "loss": 0.4221,
+      "step": 561
+    },
+    {
+      "epoch": 4.803418803418803,
+      "grad_norm": 0.9455279566847924,
+      "learning_rate": 3.838055228559864e-06,
+      "loss": 0.3173,
+      "step": 562
+    },
+    {
+      "epoch": 4.811965811965812,
+      "grad_norm": 0.8120892388463951,
+      "learning_rate": 3.834272831680785e-06,
+      "loss": 0.4401,
+      "step": 563
+    },
+    {
+      "epoch": 4.82051282051282,
+      "grad_norm": 0.9040278606499692,
+      "learning_rate": 3.830486159270991e-06,
+      "loss": 0.4469,
+      "step": 564
+    },
+    {
+      "epoch": 4.829059829059829,
+      "grad_norm": 0.7713925407833625,
+      "learning_rate": 3.826695223464455e-06,
+      "loss": 0.4895,
+      "step": 565
+    },
+    {
+      "epoch": 4.837606837606837,
+      "grad_norm": 0.7906937990955305,
+      "learning_rate": 3.822900036408815e-06,
+      "loss": 0.4848,
+      "step": 566
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.7618954650882707,
+      "learning_rate": 3.819100610265332e-06,
+      "loss": 0.3965,
+      "step": 567
+    },
+    {
+      "epoch": 4.854700854700854,
+      "grad_norm": 0.8218662694671668,
+      "learning_rate": 3.815296957208849e-06,
+      "loss": 0.3004,
+      "step": 568
+    },
+    {
+      "epoch": 4.863247863247864,
+      "grad_norm": 0.9075356752746051,
+      "learning_rate": 3.811489089427756e-06,
+      "loss": 0.3602,
+      "step": 569
+    },
+    {
+      "epoch": 4.871794871794872,
+      "grad_norm": 1.0079500894091658,
+      "learning_rate": 3.8076770191239444e-06,
+      "loss": 0.6662,
+      "step": 570
+    },
+    {
+      "epoch": 4.880341880341881,
+      "grad_norm": 0.9358756861041553,
+      "learning_rate": 3.8038607585127762e-06,
+      "loss": 0.6168,
+      "step": 571
+    },
+    {
+      "epoch": 4.888888888888889,
+      "grad_norm": 0.7803465983190693,
+      "learning_rate": 3.8000403198230385e-06,
+      "loss": 0.2579,
+      "step": 572
+    },
+    {
+      "epoch": 4.897435897435898,
+      "grad_norm": 0.7743839576202268,
+      "learning_rate": 3.7962157152969093e-06,
+      "loss": 0.2667,
+      "step": 573
+    },
+    {
+      "epoch": 4.905982905982906,
+      "grad_norm": 0.9137844362982184,
+      "learning_rate": 3.7923869571899115e-06,
+      "loss": 0.2822,
+      "step": 574
+    },
+    {
+      "epoch": 4.914529914529915,
+      "grad_norm": 0.9527902955445767,
+      "learning_rate": 3.7885540577708806e-06,
+      "loss": 0.4376,
+      "step": 575
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.9681874312358377,
+      "learning_rate": 3.7847170293219223e-06,
+      "loss": 0.3069,
+      "step": 576
+    },
+    {
+      "epoch": 4.931623931623932,
+      "grad_norm": 0.8238895402593656,
+      "learning_rate": 3.780875884138372e-06,
+      "loss": 0.1889,
+      "step": 577
+    },
+    {
+      "epoch": 4.94017094017094,
+      "grad_norm": 0.918981736118876,
+      "learning_rate": 3.7770306345287577e-06,
+      "loss": 0.4307,
+      "step": 578
+    },
+    {
+      "epoch": 4.948717948717949,
+      "grad_norm": 0.8454182927023547,
+      "learning_rate": 3.7731812928147593e-06,
+      "loss": 0.3387,
+      "step": 579
+    },
+    {
+      "epoch": 4.957264957264957,
+      "grad_norm": 0.7835233380602107,
+      "learning_rate": 3.76932787133117e-06,
+      "loss": 0.2332,
+      "step": 580
+    },
+    {
+      "epoch": 4.965811965811966,
+      "grad_norm": 0.8919386512222267,
+      "learning_rate": 3.7654703824258544e-06,
+      "loss": 0.4062,
+      "step": 581
+    },
+    {
+      "epoch": 4.9743589743589745,
+      "grad_norm": 1.0307902409547174,
+      "learning_rate": 3.7616088384597138e-06,
+      "loss": 0.6783,
+      "step": 582
+    },
+    {
+      "epoch": 4.982905982905983,
+      "grad_norm": 0.8082634397054688,
+      "learning_rate": 3.757743251806639e-06,
+      "loss": 0.3732,
+      "step": 583
+    },
+    {
+      "epoch": 4.9914529914529915,
+      "grad_norm": 0.7740530355915068,
+      "learning_rate": 3.753873634853481e-06,
+      "loss": 0.376,
+      "step": 584
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.8908301366802671,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.3181,
+      "step": 585
+    },
+    {
+      "epoch": 5.0085470085470085,
+      "grad_norm": 0.8069432751813113,
+      "learning_rate": 3.746122359658834e-06,
+      "loss": 0.3528,
+      "step": 586
+    },
+    {
+      "epoch": 5.017094017094017,
+      "grad_norm": 0.8379587111320108,
+      "learning_rate": 3.7422407262554567e-06,
+      "loss": 0.5588,
+      "step": 587
+    },
+    {
+      "epoch": 5.0256410256410255,
+      "grad_norm": 0.7122032449476201,
+      "learning_rate": 3.738355112228134e-06,
+      "loss": 0.2672,
+      "step": 588
+    },
+    {
+      "epoch": 5.034188034188034,
+      "grad_norm": 0.9524584213957907,
+      "learning_rate": 3.7344655300278887e-06,
+      "loss": 0.4974,
+      "step": 589
+    },
+    {
+      "epoch": 5.042735042735043,
+      "grad_norm": 0.8103317648549345,
+      "learning_rate": 3.7305719921184626e-06,
+      "loss": 0.2324,
+      "step": 590
+    },
+    {
+      "epoch": 5.051282051282051,
+      "grad_norm": 0.7079055681673191,
+      "learning_rate": 3.7266745109762668e-06,
+      "loss": 0.2059,
+      "step": 591
+    },
+    {
+      "epoch": 5.05982905982906,
+      "grad_norm": 0.9947949971700039,
+      "learning_rate": 3.7227730990903556e-06,
+      "loss": 0.4619,
+      "step": 592
+    },
+    {
+      "epoch": 5.068376068376068,
+      "grad_norm": 1.233347361263019,
+      "learning_rate": 3.718867768962371e-06,
+      "loss": 0.2852,
+      "step": 593
+    },
+    {
+      "epoch": 5.076923076923077,
+      "grad_norm": 0.9632110639765008,
+      "learning_rate": 3.714958533106515e-06,
+      "loss": 0.4309,
+      "step": 594
+    },
+    {
+      "epoch": 5.085470085470085,
+      "grad_norm": 0.8286550755670258,
+      "learning_rate": 3.711045404049507e-06,
+      "loss": 0.3396,
+      "step": 595
+    },
+    {
+      "epoch": 5.094017094017094,
+      "grad_norm": 0.9713598976350867,
+      "learning_rate": 3.7071283943305367e-06,
+      "loss": 0.3712,
+      "step": 596
+    },
+    {
+      "epoch": 5.102564102564102,
+      "grad_norm": 0.8550775572776252,
+      "learning_rate": 3.7032075165012323e-06,
+      "loss": 0.261,
+      "step": 597
+    },
+    {
+      "epoch": 5.111111111111111,
+      "grad_norm": 1.181204605664654,
+      "learning_rate": 3.699282783125616e-06,
+      "loss": 0.4192,
+      "step": 598
+    },
+    {
+      "epoch": 5.119658119658119,
+      "grad_norm": 0.8230527538892907,
+      "learning_rate": 3.6953542067800647e-06,
+      "loss": 0.4015,
+      "step": 599
+    },
+    {
+      "epoch": 5.128205128205128,
+      "grad_norm": 0.662879386627582,
+      "learning_rate": 3.6914218000532697e-06,
+      "loss": 0.2604,
+      "step": 600
+    },
+    {
+      "epoch": 5.136752136752137,
+      "grad_norm": 0.8571575604386643,
+      "learning_rate": 3.6874855755461975e-06,
+      "loss": 0.2278,
+      "step": 601
+    },
+    {
+      "epoch": 5.145299145299146,
+      "grad_norm": 0.7657497421323586,
+      "learning_rate": 3.683545545872045e-06,
+      "loss": 0.236,
+      "step": 602
+    },
+    {
+      "epoch": 5.153846153846154,
+      "grad_norm": 0.9438673580943562,
+      "learning_rate": 3.679601723656205e-06,
+      "loss": 0.4039,
+      "step": 603
+    },
+    {
+      "epoch": 5.162393162393163,
+      "grad_norm": 0.7320407216705076,
+      "learning_rate": 3.675654121536225e-06,
+      "loss": 0.1617,
+      "step": 604
+    },
+    {
+      "epoch": 5.170940170940171,
+      "grad_norm": 0.9168657405228899,
+      "learning_rate": 3.6717027521617593e-06,
+      "loss": 0.3958,
+      "step": 605
+    },
+    {
+      "epoch": 5.17948717948718,
+      "grad_norm": 0.8856897705058108,
+      "learning_rate": 3.667747628194539e-06,
+      "loss": 0.3021,
+      "step": 606
+    },
+    {
+      "epoch": 5.188034188034188,
+      "grad_norm": 0.8848451433593771,
+      "learning_rate": 3.6637887623083235e-06,
+      "loss": 0.284,
+      "step": 607
+    },
+    {
+      "epoch": 5.196581196581197,
+      "grad_norm": 0.7931777936733362,
+      "learning_rate": 3.6598261671888623e-06,
+      "loss": 0.2598,
+      "step": 608
+    },
+    {
+      "epoch": 5.205128205128205,
+      "grad_norm": 0.9647752279410728,
+      "learning_rate": 3.655859855533859e-06,
+      "loss": 0.2347,
+      "step": 609
+    },
+    {
+      "epoch": 5.213675213675214,
+      "grad_norm": 0.8027939690593247,
+      "learning_rate": 3.651889840052922e-06,
+      "loss": 0.2356,
+      "step": 610
+    },
+    {
+      "epoch": 5.222222222222222,
+      "grad_norm": 0.8976635229956506,
+      "learning_rate": 3.6479161334675294e-06,
+      "loss": 0.3416,
+      "step": 611
+    },
+    {
+      "epoch": 5.230769230769231,
+      "grad_norm": 0.7907973289926297,
+      "learning_rate": 3.643938748510989e-06,
+      "loss": 0.2042,
+      "step": 612
+    },
+    {
+      "epoch": 5.239316239316239,
+      "grad_norm": 0.9851284719548342,
+      "learning_rate": 3.6399576979283914e-06,
+      "loss": 0.4832,
+      "step": 613
+    },
+    {
+      "epoch": 5.247863247863248,
+      "grad_norm": 0.757028631944513,
+      "learning_rate": 3.6359729944765785e-06,
+      "loss": 0.3568,
+      "step": 614
+    },
+    {
+      "epoch": 5.256410256410256,
+      "grad_norm": 1.0804566273764746,
+      "learning_rate": 3.631984650924094e-06,
+      "loss": 0.3044,
+      "step": 615
+    },
+    {
+      "epoch": 5.264957264957265,
+      "grad_norm": 1.0035092854743617,
+      "learning_rate": 3.6279926800511455e-06,
+      "loss": 0.5795,
+      "step": 616
+    },
+    {
+      "epoch": 5.273504273504273,
+      "grad_norm": 0.833070471688987,
+      "learning_rate": 3.623997094649566e-06,
+      "loss": 0.2358,
+      "step": 617
+    },
+    {
+      "epoch": 5.282051282051282,
+      "grad_norm": 10.073832467148229,
+      "learning_rate": 3.6199979075227707e-06,
+      "loss": 0.2122,
+      "step": 618
+    },
+    {
+      "epoch": 5.2905982905982905,
+      "grad_norm": 0.9327671566598944,
+      "learning_rate": 3.6159951314857145e-06,
+      "loss": 0.1887,
+      "step": 619
+    },
+    {
+      "epoch": 5.299145299145299,
+      "grad_norm": 0.8129982683347555,
+      "learning_rate": 3.6119887793648535e-06,
+      "loss": 0.3573,
+      "step": 620
+    },
+    {
+      "epoch": 5.3076923076923075,
+      "grad_norm": 0.9313154955914471,
+      "learning_rate": 3.607978863998104e-06,
+      "loss": 0.2051,
+      "step": 621
+    },
+    {
+      "epoch": 5.316239316239316,
+      "grad_norm": 0.7081388827230642,
+      "learning_rate": 3.6039653982347977e-06,
+      "loss": 0.136,
+      "step": 622
+    },
+    {
+      "epoch": 5.3247863247863245,
+      "grad_norm": 0.9219903379895183,
+      "learning_rate": 3.5999483949356458e-06,
+      "loss": 0.1471,
+      "step": 623
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 0.7251881597091862,
+      "learning_rate": 3.595927866972694e-06,
+      "loss": 0.2349,
+      "step": 624
+    },
+    {
+      "epoch": 5.3418803418803416,
+      "grad_norm": 0.9020851683327609,
+      "learning_rate": 3.5919038272292824e-06,
+      "loss": 0.3872,
+      "step": 625
+    },
+    {
+      "epoch": 5.35042735042735,
+      "grad_norm": 0.6985204767840774,
+      "learning_rate": 3.587876288600004e-06,
+      "loss": 0.1627,
+      "step": 626
+    },
+    {
+      "epoch": 5.358974358974359,
+      "grad_norm": 0.7857839377226197,
+      "learning_rate": 3.583845263990664e-06,
+      "loss": 0.1391,
+      "step": 627
+    },
+    {
+      "epoch": 5.367521367521368,
+      "grad_norm": 0.7809786139005731,
+      "learning_rate": 3.5798107663182386e-06,
+      "loss": 0.2627,
+      "step": 628
+    },
+    {
+      "epoch": 5.3760683760683765,
+      "grad_norm": 0.626075541867184,
+      "learning_rate": 3.5757728085108318e-06,
+      "loss": 0.1824,
+      "step": 629
+    },
+    {
+      "epoch": 5.384615384615385,
+      "grad_norm": 0.9968721295278316,
+      "learning_rate": 3.5717314035076355e-06,
+      "loss": 0.2707,
+      "step": 630
+    },
+    {
+      "epoch": 5.3931623931623935,
+      "grad_norm": 0.643224962362624,
+      "learning_rate": 3.5676865642588894e-06,
+      "loss": 0.206,
+      "step": 631
+    },
+    {
+      "epoch": 5.401709401709402,
+      "grad_norm": 0.8976091064580924,
+      "learning_rate": 3.563638303725835e-06,
+      "loss": 0.3809,
+      "step": 632
+    },
+    {
+      "epoch": 5.410256410256411,
+      "grad_norm": 0.8108000905897182,
+      "learning_rate": 3.559586634880679e-06,
+      "loss": 0.3308,
+      "step": 633
+    },
+    {
+      "epoch": 5.418803418803419,
+      "grad_norm": 0.9784876560845137,
+      "learning_rate": 3.5555315707065496e-06,
+      "loss": 0.2254,
+      "step": 634
+    },
+    {
+      "epoch": 5.427350427350428,
+      "grad_norm": 0.9203036555176567,
+      "learning_rate": 3.551473124197454e-06,
+      "loss": 0.223,
+      "step": 635
+    },
+    {
+      "epoch": 5.435897435897436,
+      "grad_norm": 1.0744310232489362,
+      "learning_rate": 3.5474113083582382e-06,
+      "loss": 0.4916,
+      "step": 636
+    },
+    {
+      "epoch": 5.444444444444445,
+      "grad_norm": 0.8487492865506904,
+      "learning_rate": 3.543346136204545e-06,
+      "loss": 0.4219,
+      "step": 637
+    },
+    {
+      "epoch": 5.452991452991453,
+      "grad_norm": 1.067112173154662,
+      "learning_rate": 3.539277620762772e-06,
+      "loss": 0.4644,
+      "step": 638
+    },
+    {
+      "epoch": 5.461538461538462,
+      "grad_norm": 0.7961984213609168,
+      "learning_rate": 3.53520577507003e-06,
+      "loss": 0.3426,
+      "step": 639
+    },
+    {
+      "epoch": 5.47008547008547,
+      "grad_norm": 0.9902480110932708,
+      "learning_rate": 3.5311306121741017e-06,
+      "loss": 0.2114,
+      "step": 640
+    },
+    {
+      "epoch": 5.478632478632479,
+      "grad_norm": 0.7549750491704087,
+      "learning_rate": 3.5270521451333984e-06,
+      "loss": 0.2697,
+      "step": 641
+    },
+    {
+      "epoch": 5.487179487179487,
+      "grad_norm": 0.8077593813783904,
+      "learning_rate": 3.522970387016919e-06,
+      "loss": 0.235,
+      "step": 642
+    },
+    {
+      "epoch": 5.495726495726496,
+      "grad_norm": 1.1397711670646056,
+      "learning_rate": 3.5188853509042105e-06,
+      "loss": 0.3049,
+      "step": 643
+    },
+    {
+      "epoch": 5.504273504273504,
+      "grad_norm": 0.9371126916927187,
+      "learning_rate": 3.5147970498853214e-06,
+      "loss": 0.2272,
+      "step": 644
+    },
+    {
+      "epoch": 5.512820512820513,
+      "grad_norm": 0.85247512093159,
+      "learning_rate": 3.5107054970607624e-06,
+      "loss": 0.1913,
+      "step": 645
+    },
+    {
+      "epoch": 5.521367521367521,
+      "grad_norm": 0.6709218528404395,
+      "learning_rate": 3.5066107055414677e-06,
+      "loss": 0.1069,
+      "step": 646
+    },
+    {
+      "epoch": 5.52991452991453,
+      "grad_norm": 0.9707847068514217,
+      "learning_rate": 3.5025126884487447e-06,
+      "loss": 0.3735,
+      "step": 647
+    },
+    {
+      "epoch": 5.538461538461538,
+      "grad_norm": 0.8774435456307486,
+      "learning_rate": 3.4984114589142388e-06,
+      "loss": 0.1399,
+      "step": 648
+    },
+    {
+      "epoch": 5.547008547008547,
+      "grad_norm": 0.703068990656212,
+      "learning_rate": 3.4943070300798913e-06,
+      "loss": 0.1523,
+      "step": 649
+    },
+    {
+      "epoch": 5.555555555555555,
+      "grad_norm": 0.9322284365828467,
+      "learning_rate": 3.4901994150978926e-06,
+      "loss": 0.542,
+      "step": 650
+    },
+    {
+      "epoch": 5.564102564102564,
+      "grad_norm": 0.8731981374849843,
+      "learning_rate": 3.4860886271306433e-06,
+      "loss": 0.2596,
+      "step": 651
+    },
+    {
+      "epoch": 5.572649572649572,
+      "grad_norm": 0.7056414707665756,
+      "learning_rate": 3.481974679350712e-06,
+      "loss": 0.2854,
+      "step": 652
+    },
+    {
+      "epoch": 5.581196581196581,
+      "grad_norm": 0.9683966340382449,
+      "learning_rate": 3.4778575849407924e-06,
+      "loss": 0.1741,
+      "step": 653
+    },
+    {
+      "epoch": 5.589743589743589,
+      "grad_norm": 0.8837532066130346,
+      "learning_rate": 3.473737357093662e-06,
+      "loss": 0.1644,
+      "step": 654
+    },
+    {
+      "epoch": 5.598290598290598,
+      "grad_norm": 0.885697744276433,
+      "learning_rate": 3.4696140090121377e-06,
+      "loss": 0.1639,
+      "step": 655
+    },
+    {
+      "epoch": 5.6068376068376065,
+      "grad_norm": 0.6822269980868856,
+      "learning_rate": 3.465487553909035e-06,
+      "loss": 0.2013,
+      "step": 656
+    },
+    {
+      "epoch": 5.615384615384615,
+      "grad_norm": 0.8666554158855085,
+      "learning_rate": 3.461358005007128e-06,
+      "loss": 0.3809,
+      "step": 657
+    },
+    {
+      "epoch": 5.6239316239316235,
+      "grad_norm": 0.8424133430662399,
+      "learning_rate": 3.4572253755390996e-06,
+      "loss": 0.4638,
+      "step": 658
+    },
+    {
+      "epoch": 5.632478632478632,
+      "grad_norm": 0.7084488043778324,
+      "learning_rate": 3.4530896787475083e-06,
+      "loss": 0.2495,
+      "step": 659
+    },
+    {
+      "epoch": 5.641025641025641,
+      "grad_norm": 0.7454913434451906,
+      "learning_rate": 3.4489509278847415e-06,
+      "loss": 0.1966,
+      "step": 660
+    },
+    {
+      "epoch": 5.64957264957265,
+      "grad_norm": 0.7489666539547467,
+      "learning_rate": 3.44480913621297e-06,
+      "loss": 0.3321,
+      "step": 661
+    },
+    {
+      "epoch": 5.6581196581196584,
+      "grad_norm": 0.8058163715427511,
+      "learning_rate": 3.44066431700411e-06,
+      "loss": 0.1957,
+      "step": 662
+    },
+    {
+      "epoch": 5.666666666666667,
+      "grad_norm": 0.7440212094445409,
+      "learning_rate": 3.436516483539781e-06,
+      "loss": 0.2801,
+      "step": 663
+    },
+    {
+      "epoch": 5.6752136752136755,
+      "grad_norm": 0.8916063043156528,
+      "learning_rate": 3.432365649111257e-06,
+      "loss": 0.3145,
+      "step": 664
+    },
+    {
+      "epoch": 5.683760683760684,
+      "grad_norm": 0.7370090850118274,
+      "learning_rate": 3.428211827019434e-06,
+      "loss": 0.2048,
+      "step": 665
+    },
+    {
+      "epoch": 5.6923076923076925,
+      "grad_norm": 0.5871642655097474,
+      "learning_rate": 3.4240550305747776e-06,
+      "loss": 0.1435,
+      "step": 666
+    },
+    {
+      "epoch": 5.700854700854701,
+      "grad_norm": 0.891319020868617,
+      "learning_rate": 3.4198952730972845e-06,
+      "loss": 0.1379,
+      "step": 667
+    },
+    {
+      "epoch": 5.7094017094017095,
+      "grad_norm": 0.9071253346493615,
+      "learning_rate": 3.4157325679164416e-06,
+      "loss": 0.2694,
+      "step": 668
+    },
+    {
+      "epoch": 5.717948717948718,
+      "grad_norm": 0.7319345015547237,
+      "learning_rate": 3.4115669283711795e-06,
+      "loss": 0.1798,
+      "step": 669
+    },
+    {
+      "epoch": 5.726495726495727,
+      "grad_norm": 0.8775265149736555,
+      "learning_rate": 3.407398367809832e-06,
+      "loss": 0.3048,
+      "step": 670
+    },
+    {
+      "epoch": 5.735042735042735,
+      "grad_norm": 0.9093807763167042,
+      "learning_rate": 3.403226899590096e-06,
+      "loss": 0.2582,
+      "step": 671
+    },
+    {
+      "epoch": 5.743589743589744,
+      "grad_norm": 0.8072037853041546,
+      "learning_rate": 3.3990525370789793e-06,
+      "loss": 0.3216,
+      "step": 672
+    },
+    {
+      "epoch": 5.752136752136752,
+      "grad_norm": 0.7283202704225692,
+      "learning_rate": 3.3948752936527722e-06,
+      "loss": 0.1974,
+      "step": 673
+    },
+    {
+      "epoch": 5.760683760683761,
+      "grad_norm": 1.052656186926491,
+      "learning_rate": 3.3906951826969905e-06,
+      "loss": 0.2879,
+      "step": 674
+    },
+    {
+      "epoch": 5.769230769230769,
+      "grad_norm": 0.9222727700013065,
+      "learning_rate": 3.386512217606339e-06,
+      "loss": 0.3203,
+      "step": 675
+    },
+    {
+      "epoch": 5.777777777777778,
+      "grad_norm": 0.6737226893721318,
+      "learning_rate": 3.3823264117846722e-06,
+      "loss": 0.2091,
+      "step": 676
+    },
+    {
+      "epoch": 5.786324786324786,
+      "grad_norm": 0.8344812169388631,
+      "learning_rate": 3.378137778644945e-06,
+      "loss": 0.1914,
+      "step": 677
+    },
+    {
+      "epoch": 5.794871794871795,
+      "grad_norm": 0.7313452201207752,
+      "learning_rate": 3.3739463316091696e-06,
+      "loss": 0.2082,
+      "step": 678
+    },
+    {
+      "epoch": 5.803418803418803,
+      "grad_norm": 1.0196584768311239,
+      "learning_rate": 3.369752084108381e-06,
+      "loss": 0.3903,
+      "step": 679
+    },
+    {
+      "epoch": 5.811965811965812,
+      "grad_norm": 0.8240003204407577,
+      "learning_rate": 3.3655550495825824e-06,
+      "loss": 0.2762,
+      "step": 680
+    },
+    {
+      "epoch": 5.82051282051282,
+      "grad_norm": 0.7512931928983033,
+      "learning_rate": 3.3613552414807093e-06,
+      "loss": 0.2175,
+      "step": 681
+    },
+    {
+      "epoch": 5.829059829059829,
+      "grad_norm": 0.6871042497938531,
+      "learning_rate": 3.3571526732605875e-06,
+      "loss": 0.2537,
+      "step": 682
+    },
+    {
+      "epoch": 5.837606837606837,
+      "grad_norm": 0.7647048425944158,
+      "learning_rate": 3.352947358388884e-06,
+      "loss": 0.316,
+      "step": 683
+    },
+    {
+      "epoch": 5.846153846153846,
+      "grad_norm": 1.0523474677444042,
+      "learning_rate": 3.3487393103410683e-06,
+      "loss": 0.3565,
+      "step": 684
+    },
+    {
+      "epoch": 5.854700854700854,
+      "grad_norm": 1.0053910019090018,
+      "learning_rate": 3.3445285426013683e-06,
+      "loss": 0.2339,
+      "step": 685
+    },
+    {
+      "epoch": 5.863247863247864,
+      "grad_norm": 1.1075820032173571,
+      "learning_rate": 3.3403150686627267e-06,
+      "loss": 0.1635,
+      "step": 686
+    },
+    {
+      "epoch": 5.871794871794872,
+      "grad_norm": 0.850285980968383,
+      "learning_rate": 3.336098902026758e-06,
+      "loss": 0.2759,
+      "step": 687
+    },
+    {
+      "epoch": 5.880341880341881,
+      "grad_norm": 0.9845363196465164,
+      "learning_rate": 3.331880056203706e-06,
+      "loss": 0.2653,
+      "step": 688
+    },
+    {
+      "epoch": 5.888888888888889,
+      "grad_norm": 0.7994207294988291,
+      "learning_rate": 3.3276585447123957e-06,
+      "loss": 0.3057,
+      "step": 689
+    },
+    {
+      "epoch": 5.897435897435898,
+      "grad_norm": 0.9930164416674668,
+      "learning_rate": 3.3234343810801995e-06,
+      "loss": 0.3275,
+      "step": 690
+    },
+    {
+      "epoch": 5.905982905982906,
+      "grad_norm": 0.8107243099975281,
+      "learning_rate": 3.319207578842985e-06,
+      "loss": 0.2233,
+      "step": 691
+    },
+    {
+      "epoch": 5.914529914529915,
+      "grad_norm": 0.8979682402306377,
+      "learning_rate": 3.314978151545076e-06,
+      "loss": 0.4968,
+      "step": 692
+    },
+    {
+      "epoch": 5.923076923076923,
+      "grad_norm": 1.037164824170856,
+      "learning_rate": 3.3107461127392072e-06,
+      "loss": 0.2247,
+      "step": 693
+    },
+    {
+      "epoch": 5.931623931623932,
+      "grad_norm": 0.7226994441140007,
+      "learning_rate": 3.306511475986482e-06,
+      "loss": 0.2398,
+      "step": 694
+    },
+    {
+      "epoch": 5.94017094017094,
+      "grad_norm": 0.9559487252695029,
+      "learning_rate": 3.3022742548563293e-06,
+      "loss": 0.4466,
+      "step": 695
+    },
+    {
+      "epoch": 5.948717948717949,
+      "grad_norm": 0.9680850683049722,
+      "learning_rate": 3.2980344629264583e-06,
+      "loss": 0.2387,
+      "step": 696
+    },
+    {
+      "epoch": 5.957264957264957,
+      "grad_norm": 0.7532667312418817,
+      "learning_rate": 3.293792113782816e-06,
+      "loss": 0.1821,
+      "step": 697
+    },
+    {
+      "epoch": 5.965811965811966,
+      "grad_norm": 1.057782960709235,
+      "learning_rate": 3.289547221019546e-06,
+      "loss": 0.202,
+      "step": 698
+    },
+    {
+      "epoch": 5.9743589743589745,
+      "grad_norm": 0.7047064734557095,
+      "learning_rate": 3.285299798238938e-06,
+      "loss": 0.4263,
+      "step": 699
+    },
+    {
+      "epoch": 5.982905982905983,
+      "grad_norm": 0.6085610798312747,
+      "learning_rate": 3.281049859051394e-06,
+      "loss": 0.1122,
+      "step": 700
+    },
+    {
+      "epoch": 5.9914529914529915,
+      "grad_norm": 0.8539344831378322,
+      "learning_rate": 3.276797417075377e-06,
+      "loss": 0.2858,
+      "step": 701
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.0174301806268455,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.2751,
+      "step": 702
+    },
+    {
+      "epoch": 6.0085470085470085,
+      "grad_norm": 0.8594263350344946,
+      "learning_rate": 3.26828507927183e-06,
+      "loss": 0.1316,
+      "step": 703
+    },
+    {
+      "epoch": 6.017094017094017,
+      "grad_norm": 0.7086701921056165,
+      "learning_rate": 3.264025210721153e-06,
+      "loss": 0.1628,
+      "step": 704
+    },
+    {
+      "epoch": 6.0256410256410255,
+      "grad_norm": 0.8721167889196723,
+      "learning_rate": 3.2597628939356174e-06,
+      "loss": 0.2331,
+      "step": 705
+    },
+    {
+      "epoch": 6.034188034188034,
+      "grad_norm": 0.9728775357235206,
+      "learning_rate": 3.25549814257335e-06,
+      "loss": 0.2376,
+      "step": 706
+    },
+    {
+      "epoch": 6.042735042735043,
+      "grad_norm": 0.928726012701592,
+      "learning_rate": 3.2512309703002776e-06,
+      "loss": 0.2871,
+      "step": 707
+    },
+    {
+      "epoch": 6.051282051282051,
+      "grad_norm": 1.111083204994824,
+      "learning_rate": 3.2469613907900847e-06,
+      "loss": 0.166,
+      "step": 708
+    },
+    {
+      "epoch": 6.05982905982906,
+      "grad_norm": 1.1257660148157198,
+      "learning_rate": 3.2426894177241707e-06,
+      "loss": 0.3438,
+      "step": 709
+    },
+    {
+      "epoch": 6.068376068376068,
+      "grad_norm": 1.068916052026842,
+      "learning_rate": 3.2384150647916033e-06,
+      "loss": 0.1003,
+      "step": 710
+    },
+    {
+      "epoch": 6.076923076923077,
+      "grad_norm": 1.3536671412811536,
+      "learning_rate": 3.2341383456890776e-06,
+      "loss": 0.2147,
+      "step": 711
+    },
+    {
+      "epoch": 6.085470085470085,
+      "grad_norm": 0.8748429383440107,
+      "learning_rate": 3.229859274120869e-06,
+      "loss": 0.0892,
+      "step": 712
+    },
+    {
+      "epoch": 6.094017094017094,
+      "grad_norm": 0.8616158755713197,
+      "learning_rate": 3.2255778637987935e-06,
+      "loss": 0.1496,
+      "step": 713
+    },
+    {
+      "epoch": 6.102564102564102,
+      "grad_norm": 0.7495676746876246,
+      "learning_rate": 3.2212941284421595e-06,
+      "loss": 0.1524,
+      "step": 714
+    },
+    {
+      "epoch": 6.111111111111111,
+      "grad_norm": 0.8671271794505848,
+      "learning_rate": 3.217008081777726e-06,
+      "loss": 0.0848,
+      "step": 715
+    },
+    {
+      "epoch": 6.119658119658119,
+      "grad_norm": 1.0037109617812803,
+      "learning_rate": 3.2127197375396596e-06,
+      "loss": 0.5167,
+      "step": 716
+    },
+    {
+      "epoch": 6.128205128205128,
+      "grad_norm": 0.9008170358083694,
+      "learning_rate": 3.208429109469488e-06,
+      "loss": 0.2012,
+      "step": 717
+    },
+    {
+      "epoch": 6.136752136752137,
+      "grad_norm": 0.8869908267151387,
+      "learning_rate": 3.204136211316057e-06,
+      "loss": 0.293,
+      "step": 718
+    },
+    {
+      "epoch": 6.145299145299146,
+      "grad_norm": 0.6786676791676538,
+      "learning_rate": 3.199841056835489e-06,
+      "loss": 0.1845,
+      "step": 719
+    },
+    {
+      "epoch": 6.153846153846154,
+      "grad_norm": 0.9254122112443579,
+      "learning_rate": 3.195543659791132e-06,
+      "loss": 0.2013,
+      "step": 720
+    },
+    {
+      "epoch": 6.162393162393163,
+      "grad_norm": 0.8742089134571687,
+      "learning_rate": 3.191244033953524e-06,
+      "loss": 0.251,
+      "step": 721
+    },
+    {
+      "epoch": 6.170940170940171,
+      "grad_norm": 1.1184341098009263,
+      "learning_rate": 3.1869421931003446e-06,
+      "loss": 0.0712,
+      "step": 722
+    },
+    {
+      "epoch": 6.17948717948718,
+      "grad_norm": 0.8781294540758994,
+      "learning_rate": 3.182638151016369e-06,
+      "loss": 0.2918,
+      "step": 723
+    },
+    {
+      "epoch": 6.188034188034188,
+      "grad_norm": 0.7726159764192209,
+      "learning_rate": 3.1783319214934274e-06,
+      "loss": 0.3152,
+      "step": 724
+    },
+    {
+      "epoch": 6.196581196581197,
+      "grad_norm": 0.7818897804716108,
+      "learning_rate": 3.17402351833036e-06,
+      "loss": 0.2823,
+      "step": 725
+    },
+    {
+      "epoch": 6.205128205128205,
+      "grad_norm": 1.0357434492953472,
+      "learning_rate": 3.1697129553329708e-06,
+      "loss": 0.4138,
+      "step": 726
+    },
+    {
+      "epoch": 6.213675213675214,
+      "grad_norm": 0.7576741761488609,
+      "learning_rate": 3.1654002463139854e-06,
+      "loss": 0.2245,
+      "step": 727
+    },
+    {
+      "epoch": 6.222222222222222,
+      "grad_norm": 1.0668651055363016,
+      "learning_rate": 3.1610854050930063e-06,
+      "loss": 0.1697,
+      "step": 728
+    },
+    {
+      "epoch": 6.230769230769231,
+      "grad_norm": 0.8263885838889453,
+      "learning_rate": 3.1567684454964674e-06,
+      "loss": 0.2849,
+      "step": 729
+    },
+    {
+      "epoch": 6.239316239316239,
+      "grad_norm": 0.897280538956934,
+      "learning_rate": 3.1524493813575936e-06,
+      "loss": 0.3428,
+      "step": 730
+    },
+    {
+      "epoch": 6.247863247863248,
+      "grad_norm": 1.1418192612891902,
+      "learning_rate": 3.1481282265163493e-06,
+      "loss": 0.3214,
+      "step": 731
+    },
+    {
+      "epoch": 6.256410256410256,
+      "grad_norm": 0.8665101374271217,
+      "learning_rate": 3.1438049948194006e-06,
+      "loss": 0.111,
+      "step": 732
+    },
+    {
+      "epoch": 6.264957264957265,
+      "grad_norm": 0.8646101090642209,
+      "learning_rate": 3.1394797001200705e-06,
+      "loss": 0.1649,
+      "step": 733
+    },
+    {
+      "epoch": 6.273504273504273,
+      "grad_norm": 0.7281704329321338,
+      "learning_rate": 3.1351523562782893e-06,
+      "loss": 0.0735,
+      "step": 734
+    },
+    {
+      "epoch": 6.282051282051282,
+      "grad_norm": 0.8724737804025101,
+      "learning_rate": 3.1308229771605546e-06,
+      "loss": 0.0932,
+      "step": 735
+    },
+    {
+      "epoch": 6.2905982905982905,
+      "grad_norm": 0.7411862867946514,
+      "learning_rate": 3.1264915766398872e-06,
+      "loss": 0.1875,
+      "step": 736
+    },
+    {
+      "epoch": 6.299145299145299,
+      "grad_norm": 0.6934460194520883,
+      "learning_rate": 3.1221581685957837e-06,
+      "loss": 0.1958,
+      "step": 737
+    },
+    {
+      "epoch": 6.3076923076923075,
+      "grad_norm": 0.9759139901992064,
+      "learning_rate": 3.117822766914174e-06,
+      "loss": 0.361,
+      "step": 738
+    },
+    {
+      "epoch": 6.316239316239316,
+      "grad_norm": 0.8089748492620781,
+      "learning_rate": 3.1134853854873774e-06,
+      "loss": 0.2047,
+      "step": 739
+    },
+    {
+      "epoch": 6.3247863247863245,
+      "grad_norm": 0.7887119389629432,
+      "learning_rate": 3.109146038214055e-06,
+      "loss": 0.2278,
+      "step": 740
+    },
+    {
+      "epoch": 6.333333333333333,
+      "grad_norm": 0.9218221429987968,
+      "learning_rate": 3.1048047389991693e-06,
+      "loss": 0.2887,
+      "step": 741
+    },
+    {
+      "epoch": 6.3418803418803416,
+      "grad_norm": 0.7478781085448379,
+      "learning_rate": 3.1004615017539375e-06,
+      "loss": 0.1427,
+      "step": 742
+    },
+    {
+      "epoch": 6.35042735042735,
+      "grad_norm": 0.8557863575124964,
+      "learning_rate": 3.096116340395783e-06,
+      "loss": 0.2219,
+      "step": 743
+    },
+    {
+      "epoch": 6.358974358974359,
+      "grad_norm": 1.0726756668060886,
+      "learning_rate": 3.0917692688483023e-06,
+      "loss": 0.1425,
+      "step": 744
+    },
+    {
+      "epoch": 6.367521367521368,
+      "grad_norm": 0.7797929578816128,
+      "learning_rate": 3.0874203010412057e-06,
+      "loss": 0.1801,
+      "step": 745
+    },
+    {
+      "epoch": 6.3760683760683765,
+      "grad_norm": 0.9547590433959044,
+      "learning_rate": 3.0830694509102835e-06,
+      "loss": 0.1125,
+      "step": 746
+    },
+    {
+      "epoch": 6.384615384615385,
+      "grad_norm": 1.2416444033785454,
+      "learning_rate": 3.0787167323973584e-06,
+      "loss": 0.2306,
+      "step": 747
+    },
+    {
+      "epoch": 6.3931623931623935,
+      "grad_norm": 0.6885959295233361,
+      "learning_rate": 3.074362159450236e-06,
+      "loss": 0.1142,
+      "step": 748
+    },
+    {
+      "epoch": 6.401709401709402,
+      "grad_norm": 0.5879587487315899,
+      "learning_rate": 3.070005746022669e-06,
+      "loss": 0.1052,
+      "step": 749
+    },
+    {
+      "epoch": 6.410256410256411,
+      "grad_norm": 1.1695009064228226,
+      "learning_rate": 3.0656475060743065e-06,
+      "loss": 0.3528,
+      "step": 750
+    },
+    {
+      "epoch": 6.418803418803419,
+      "grad_norm": 0.8469776850696916,
+      "learning_rate": 3.061287453570646e-06,
+      "loss": 0.3548,
+      "step": 751
+    },
+    {
+      "epoch": 6.427350427350428,
+      "grad_norm": 0.7680536865979709,
+      "learning_rate": 3.056925602483e-06,
+      "loss": 0.1638,
+      "step": 752
+    },
+    {
+      "epoch": 6.435897435897436,
+      "grad_norm": 0.74533754769051,
+      "learning_rate": 3.052561966788441e-06,
+      "loss": 0.1684,
+      "step": 753
+    },
+    {
+      "epoch": 6.444444444444445,
+      "grad_norm": 0.8649414778571741,
+      "learning_rate": 3.0481965604697582e-06,
+      "loss": 0.1592,
+      "step": 754
+    },
+    {
+      "epoch": 6.452991452991453,
+      "grad_norm": 0.6651871547467371,
+      "learning_rate": 3.043829397515419e-06,
+      "loss": 0.2045,
+      "step": 755
+    },
+    {
+      "epoch": 6.461538461538462,
+      "grad_norm": 0.7896759182994997,
+      "learning_rate": 3.039460491919516e-06,
+      "loss": 0.3469,
+      "step": 756
+    },
+    {
+      "epoch": 6.47008547008547,
+      "grad_norm": 0.7284989154139125,
+      "learning_rate": 3.0350898576817268e-06,
+      "loss": 0.1932,
+      "step": 757
+    },
+    {
+      "epoch": 6.478632478632479,
+      "grad_norm": 0.8736911419143972,
+      "learning_rate": 3.03071750880727e-06,
+      "loss": 0.2464,
+      "step": 758
+    },
+    {
+      "epoch": 6.487179487179487,
+      "grad_norm": 0.7594717864593361,
+      "learning_rate": 3.0263434593068562e-06,
+      "loss": 0.2081,
+      "step": 759
+    },
+    {
+      "epoch": 6.495726495726496,
+      "grad_norm": 0.9355239477373309,
+      "learning_rate": 3.021967723196647e-06,
+      "loss": 0.3858,
+      "step": 760
+    },
+    {
+      "epoch": 6.504273504273504,
+      "grad_norm": 0.9974022702585318,
+      "learning_rate": 3.017590314498208e-06,
+      "loss": 0.1876,
+      "step": 761
+    },
+    {
+      "epoch": 6.512820512820513,
+      "grad_norm": 0.8613360050143386,
+      "learning_rate": 3.0132112472384652e-06,
+      "loss": 0.0735,
+      "step": 762
+    },
+    {
+      "epoch": 6.521367521367521,
+      "grad_norm": 0.8783727189757656,
+      "learning_rate": 3.0088305354496574e-06,
+      "loss": 0.3261,
+      "step": 763
+    },
+    {
+      "epoch": 6.52991452991453,
+      "grad_norm": 0.9569640404194898,
+      "learning_rate": 3.004448193169294e-06,
+      "loss": 0.1975,
+      "step": 764
+    },
+    {
+      "epoch": 6.538461538461538,
+      "grad_norm": 0.8657428651377301,
+      "learning_rate": 3.0000642344401115e-06,
+      "loss": 0.1689,
+      "step": 765
+    },
+    {
+      "epoch": 6.547008547008547,
+      "grad_norm": 0.8116797215365377,
+      "learning_rate": 2.9956786733100225e-06,
+      "loss": 0.2281,
+      "step": 766
+    },
+    {
+      "epoch": 6.555555555555555,
+      "grad_norm": 0.7459806019934861,
+      "learning_rate": 2.9912915238320755e-06,
+      "loss": 0.1812,
+      "step": 767
+    },
+    {
+      "epoch": 6.564102564102564,
+      "grad_norm": 0.9075555801622364,
+      "learning_rate": 2.9869028000644102e-06,
+      "loss": 0.301,
+      "step": 768
+    },
+    {
+      "epoch": 6.572649572649572,
+      "grad_norm": 1.0365280365594374,
+      "learning_rate": 2.9825125160702096e-06,
+      "loss": 0.3363,
+      "step": 769
+    },
+    {
+      "epoch": 6.581196581196581,
+      "grad_norm": 1.0681940817467501,
+      "learning_rate": 2.978120685917656e-06,
+      "loss": 0.1981,
+      "step": 770
+    },
+    {
+      "epoch": 6.589743589743589,
+      "grad_norm": 0.6063787298533598,
+      "learning_rate": 2.9737273236798868e-06,
+      "loss": 0.0629,
+      "step": 771
+    },
+    {
+      "epoch": 6.598290598290598,
+      "grad_norm": 0.7740559201346899,
+      "learning_rate": 2.9693324434349486e-06,
+      "loss": 0.1796,
+      "step": 772
+    },
+    {
+      "epoch": 6.6068376068376065,
+      "grad_norm": 0.874415643424101,
+      "learning_rate": 2.9649360592657526e-06,
+      "loss": 0.153,
+      "step": 773
+    },
+    {
+      "epoch": 6.615384615384615,
+      "grad_norm": 1.0167511568733048,
+      "learning_rate": 2.960538185260029e-06,
+      "loss": 0.2641,
+      "step": 774
+    },
+    {
+      "epoch": 6.6239316239316235,
+      "grad_norm": 0.7831542303453528,
+      "learning_rate": 2.956138835510282e-06,
+      "loss": 0.2015,
+      "step": 775
+    },
+    {
+      "epoch": 6.632478632478632,
+      "grad_norm": 0.9313522049165068,
+      "learning_rate": 2.9517380241137437e-06,
+      "loss": 0.1415,
+      "step": 776
+    },
+    {
+      "epoch": 6.641025641025641,
+      "grad_norm": 0.8601509036277558,
+      "learning_rate": 2.9473357651723324e-06,
+      "loss": 0.1626,
+      "step": 777
+    },
+    {
+      "epoch": 6.64957264957265,
+      "grad_norm": 0.9035371429077345,
+      "learning_rate": 2.942932072792602e-06,
+      "loss": 0.3335,
+      "step": 778
+    },
+    {
+      "epoch": 6.6581196581196584,
+      "grad_norm": 0.8289789818585666,
+      "learning_rate": 2.938526961085701e-06,
+      "loss": 0.096,
+      "step": 779
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 1.04127393541744,
+      "learning_rate": 2.9341204441673267e-06,
+      "loss": 0.1766,
+      "step": 780
+    },
+    {
+      "epoch": 6.6752136752136755,
+      "grad_norm": 0.8898603983416691,
+      "learning_rate": 2.929712536157677e-06,
+      "loss": 0.1996,
+      "step": 781
+    },
+    {
+      "epoch": 6.683760683760684,
+      "grad_norm": 0.8654179207006909,
+      "learning_rate": 2.925303251181411e-06,
+      "loss": 0.3013,
+      "step": 782
+    },
+    {
+      "epoch": 6.6923076923076925,
+      "grad_norm": 0.613657904979391,
+      "learning_rate": 2.920892603367596e-06,
+      "loss": 0.1672,
+      "step": 783
+    },
+    {
+      "epoch": 6.700854700854701,
+      "grad_norm": 0.7781355440372467,
+      "learning_rate": 2.916480606849671e-06,
+      "loss": 0.3259,
+      "step": 784
+    },
+    {
+      "epoch": 6.7094017094017095,
+      "grad_norm": 0.8947088549656963,
+      "learning_rate": 2.9120672757653917e-06,
+      "loss": 0.2424,
+      "step": 785
+    },
+    {
+      "epoch": 6.717948717948718,
+      "grad_norm": 0.960212554161266,
+      "learning_rate": 2.907652624256794e-06,
+      "loss": 0.2845,
+      "step": 786
+    },
+    {
+      "epoch": 6.726495726495727,
+      "grad_norm": 0.8486856515975575,
+      "learning_rate": 2.903236666470143e-06,
+      "loss": 0.2724,
+      "step": 787
+    },
+    {
+      "epoch": 6.735042735042735,
+      "grad_norm": 1.0322254546822134,
+      "learning_rate": 2.89881941655589e-06,
+      "loss": 0.1461,
+      "step": 788
+    },
+    {
+      "epoch": 6.743589743589744,
+      "grad_norm": 0.7674403559804235,
+      "learning_rate": 2.8944008886686288e-06,
+      "loss": 0.1028,
+      "step": 789
+    },
+    {
+      "epoch": 6.752136752136752,
+      "grad_norm": 1.1043201498622655,
+      "learning_rate": 2.889981096967045e-06,
+      "loss": 0.3388,
+      "step": 790
+    },
+    {
+      "epoch": 6.760683760683761,
+      "grad_norm": 1.6187698046626917,
+      "learning_rate": 2.8855600556138757e-06,
+      "loss": 0.1716,
+      "step": 791
+    },
+    {
+      "epoch": 6.769230769230769,
+      "grad_norm": 0.8037761816880599,
+      "learning_rate": 2.881137778775864e-06,
+      "loss": 0.0999,
+      "step": 792
+    },
+    {
+      "epoch": 6.777777777777778,
+      "grad_norm": 0.9542605497236784,
+      "learning_rate": 2.876714280623708e-06,
+      "loss": 0.1289,
+      "step": 793
+    },
+    {
+      "epoch": 6.786324786324786,
+      "grad_norm": 0.6727881003007951,
+      "learning_rate": 2.872289575332023e-06,
+      "loss": 0.1749,
+      "step": 794
+    },
+    {
+      "epoch": 6.794871794871795,
+      "grad_norm": 0.8180347661987107,
+      "learning_rate": 2.8678636770792907e-06,
+      "loss": 0.1947,
+      "step": 795
+    },
+    {
+      "epoch": 6.803418803418803,
+      "grad_norm": 0.670627358457358,
+      "learning_rate": 2.863436600047815e-06,
+      "loss": 0.1994,
+      "step": 796
+    },
+    {
+      "epoch": 6.811965811965812,
+      "grad_norm": 0.9167632135320165,
+      "learning_rate": 2.8590083584236792e-06,
+      "loss": 0.2602,
+      "step": 797
+    },
+    {
+      "epoch": 6.82051282051282,
+      "grad_norm": 0.9336105527397447,
+      "learning_rate": 2.854578966396697e-06,
+      "loss": 0.2185,
+      "step": 798
+    },
+    {
+      "epoch": 6.829059829059829,
+      "grad_norm": 0.5738993328361307,
+      "learning_rate": 2.8501484381603685e-06,
+      "loss": 0.1137,
+      "step": 799
+    },
+    {
+      "epoch": 6.837606837606837,
+      "grad_norm": 1.1698886548057332,
+      "learning_rate": 2.8457167879118332e-06,
+      "loss": 0.3665,
+      "step": 800
+    },
+    {
+      "epoch": 6.846153846153846,
+      "grad_norm": 0.8358950075375864,
+      "learning_rate": 2.8412840298518295e-06,
+      "loss": 0.2201,
+      "step": 801
+    },
+    {
+      "epoch": 6.854700854700854,
+      "grad_norm": 1.0664271372394594,
+      "learning_rate": 2.836850178184642e-06,
+      "loss": 0.13,
+      "step": 802
+    },
+    {
+      "epoch": 6.863247863247864,
+      "grad_norm": 0.7714166530100259,
+      "learning_rate": 2.8324152471180634e-06,
+      "loss": 0.1483,
+      "step": 803
+    },
+    {
+      "epoch": 6.871794871794872,
+      "grad_norm": 0.7116665658763507,
+      "learning_rate": 2.8279792508633415e-06,
+      "loss": 0.0972,
+      "step": 804
+    },
+    {
+      "epoch": 6.880341880341881,
+      "grad_norm": 1.0305206757103451,
+      "learning_rate": 2.8235422036351384e-06,
+      "loss": 0.2386,
+      "step": 805
+    },
+    {
+      "epoch": 6.888888888888889,
+      "grad_norm": 0.6737644564750103,
+      "learning_rate": 2.8191041196514874e-06,
+      "loss": 0.0862,
+      "step": 806
+    },
+    {
+      "epoch": 6.897435897435898,
+      "grad_norm": 1.0213339444859935,
+      "learning_rate": 2.8146650131337376e-06,
+      "loss": 0.2497,
+      "step": 807
+    },
+    {
+      "epoch": 6.905982905982906,
+      "grad_norm": 1.2015414358677925,
+      "learning_rate": 2.81022489830652e-06,
+      "loss": 0.1404,
+      "step": 808
+    },
+    {
+      "epoch": 6.914529914529915,
+      "grad_norm": 0.7713933139465124,
+      "learning_rate": 2.8057837893976958e-06,
+      "loss": 0.2377,
+      "step": 809
+    },
+    {
+      "epoch": 6.923076923076923,
+      "grad_norm": 1.040720942073932,
+      "learning_rate": 2.8013417006383078e-06,
+      "loss": 0.2294,
+      "step": 810
+    },
+    {
+      "epoch": 6.931623931623932,
+      "grad_norm": 0.8131320254780423,
+      "learning_rate": 2.7968986462625436e-06,
+      "loss": 0.2105,
+      "step": 811
+    },
+    {
+      "epoch": 6.94017094017094,
+      "grad_norm": 0.8827385433058145,
+      "learning_rate": 2.7924546405076837e-06,
+      "loss": 0.2545,
+      "step": 812
+    },
+    {
+      "epoch": 6.948717948717949,
+      "grad_norm": 1.0866373931985307,
+      "learning_rate": 2.788009697614053e-06,
+      "loss": 0.1797,
+      "step": 813
+    },
+    {
+      "epoch": 6.957264957264957,
+      "grad_norm": 0.8286348008973615,
+      "learning_rate": 2.7835638318249856e-06,
+      "loss": 0.3427,
+      "step": 814
+    },
+    {
+      "epoch": 6.965811965811966,
+      "grad_norm": 0.7069844316283795,
+      "learning_rate": 2.7791170573867698e-06,
+      "loss": 0.1923,
+      "step": 815
+    },
+    {
+      "epoch": 6.9743589743589745,
+      "grad_norm": 0.8259615583243639,
+      "learning_rate": 2.7746693885486044e-06,
+      "loss": 0.2327,
+      "step": 816
+    },
+    {
+      "epoch": 6.982905982905983,
+      "grad_norm": 0.8224327747519812,
+      "learning_rate": 2.770220839562556e-06,
+      "loss": 0.2679,
+      "step": 817
+    },
+    {
+      "epoch": 6.9914529914529915,
+      "grad_norm": 0.9872759388663668,
+      "learning_rate": 2.765771424683513e-06,
+      "loss": 0.0969,
+      "step": 818
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.7830387364303183,
+      "learning_rate": 2.761321158169134e-06,
+      "loss": 0.2424,
+      "step": 819
+    },
+    {
+      "epoch": 7.0085470085470085,
+      "grad_norm": 0.811449830083304,
+      "learning_rate": 2.7568700542798112e-06,
+      "loss": 0.1186,
+      "step": 820
+    },
+    {
+      "epoch": 7.017094017094017,
+      "grad_norm": 0.8884968910045412,
+      "learning_rate": 2.7524181272786153e-06,
+      "loss": 0.36,
+      "step": 821
+    },
+    {
+      "epoch": 7.0256410256410255,
+      "grad_norm": 0.8762185444900101,
+      "learning_rate": 2.747965391431261e-06,
+      "loss": 0.1558,
+      "step": 822
+    },
+    {
+      "epoch": 7.034188034188034,
+      "grad_norm": 0.6490106067899655,
+      "learning_rate": 2.743511861006049e-06,
+      "loss": 0.1165,
+      "step": 823
+    },
+    {
+      "epoch": 7.042735042735043,
+      "grad_norm": 1.2117907753992274,
+      "learning_rate": 2.739057550273828e-06,
+      "loss": 0.1867,
+      "step": 824
+    },
+    {
+      "epoch": 7.051282051282051,
+      "grad_norm": 0.5267319029801504,
+      "learning_rate": 2.7346024735079483e-06,
+      "loss": 0.0796,
+      "step": 825
+    },
+    {
+      "epoch": 7.05982905982906,
+      "grad_norm": 1.017628519036157,
+      "learning_rate": 2.7301466449842147e-06,
+      "loss": 0.1076,
+      "step": 826
+    },
+    {
+      "epoch": 7.068376068376068,
+      "grad_norm": 1.128531455235049,
+      "learning_rate": 2.725690078980838e-06,
+      "loss": 0.3083,
+      "step": 827
+    },
+    {
+      "epoch": 7.076923076923077,
+      "grad_norm": 0.8562518470839942,
+      "learning_rate": 2.7212327897783963e-06,
+      "loss": 0.2254,
+      "step": 828
+    },
+    {
+      "epoch": 7.085470085470085,
+      "grad_norm": 0.6525538186830495,
+      "learning_rate": 2.7167747916597825e-06,
+      "loss": 0.0773,
+      "step": 829
+    },
+    {
+      "epoch": 7.094017094017094,
+      "grad_norm": 0.7411130620062173,
+      "learning_rate": 2.7123160989101623e-06,
+      "loss": 0.2117,
+      "step": 830
+    },
+    {
+      "epoch": 7.102564102564102,
+      "grad_norm": 0.928232225419419,
+      "learning_rate": 2.7078567258169264e-06,
+      "loss": 0.1877,
+      "step": 831
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 0.566317948447876,
+      "learning_rate": 2.703396686669646e-06,
+      "loss": 0.0725,
+      "step": 832
+    },
+    {
+      "epoch": 7.119658119658119,
+      "grad_norm": 0.5667066947319995,
+      "learning_rate": 2.698935995760027e-06,
+      "loss": 0.1054,
+      "step": 833
+    },
+    {
+      "epoch": 7.128205128205128,
+      "grad_norm": 0.7285759378868003,
+      "learning_rate": 2.6944746673818623e-06,
+      "loss": 0.1234,
+      "step": 834
+    },
+    {
+      "epoch": 7.136752136752137,
+      "grad_norm": 0.9036310236315552,
+      "learning_rate": 2.6900127158309903e-06,
+      "loss": 0.1491,
+      "step": 835
+    },
+    {
+      "epoch": 7.145299145299146,
+      "grad_norm": 1.1011326912978958,
+      "learning_rate": 2.6855501554052433e-06,
+      "loss": 0.2026,
+      "step": 836
+    },
+    {
+      "epoch": 7.153846153846154,
+      "grad_norm": 0.7635833117980484,
+      "learning_rate": 2.6810870004044065e-06,
+      "loss": 0.0551,
+      "step": 837
+    },
+    {
+      "epoch": 7.162393162393163,
+      "grad_norm": 0.6872152173732621,
+      "learning_rate": 2.6766232651301694e-06,
+      "loss": 0.0774,
+      "step": 838
+    },
+    {
+      "epoch": 7.170940170940171,
+      "grad_norm": 0.9105534731343037,
+      "learning_rate": 2.672158963886082e-06,
+      "loss": 0.2154,
+      "step": 839
+    },
+    {
+      "epoch": 7.17948717948718,
+      "grad_norm": 0.8941926255951779,
+      "learning_rate": 2.667694110977506e-06,
+      "loss": 0.1053,
+      "step": 840
+    },
+    {
+      "epoch": 7.188034188034188,
+      "grad_norm": 0.8494727638062671,
+      "learning_rate": 2.6632287207115735e-06,
+      "loss": 0.1311,
+      "step": 841
+    },
+    {
+      "epoch": 7.196581196581197,
+      "grad_norm": 1.052548792291065,
+      "learning_rate": 2.6587628073971366e-06,
+      "loss": 0.3121,
+      "step": 842
+    },
+    {
+      "epoch": 7.205128205128205,
+      "grad_norm": 0.6867370020751574,
+      "learning_rate": 2.654296385344724e-06,
+      "loss": 0.1466,
+      "step": 843
+    },
+    {
+      "epoch": 7.213675213675214,
+      "grad_norm": 1.0121071427225556,
+      "learning_rate": 2.6498294688664937e-06,
+      "loss": 0.2121,
+      "step": 844
+    },
+    {
+      "epoch": 7.222222222222222,
+      "grad_norm": 1.1519304302625446,
+      "learning_rate": 2.6453620722761897e-06,
+      "loss": 0.1331,
+      "step": 845
+    },
+    {
+      "epoch": 7.230769230769231,
+      "grad_norm": 0.8501355517389577,
+      "learning_rate": 2.6408942098890937e-06,
+      "loss": 0.221,
+      "step": 846
+    },
+    {
+      "epoch": 7.239316239316239,
+      "grad_norm": 0.8315914453456078,
+      "learning_rate": 2.6364258960219794e-06,
+      "loss": 0.2355,
+      "step": 847
+    },
+    {
+      "epoch": 7.247863247863248,
+      "grad_norm": 0.6522813347545385,
+      "learning_rate": 2.631957144993068e-06,
+      "loss": 0.095,
+      "step": 848
+    },
+    {
+      "epoch": 7.256410256410256,
+      "grad_norm": 0.8036110735240778,
+      "learning_rate": 2.6274879711219816e-06,
+      "loss": 0.2501,
+      "step": 849
+    },
+    {
+      "epoch": 7.264957264957265,
+      "grad_norm": 1.0254631408066757,
+      "learning_rate": 2.6230183887296955e-06,
+      "loss": 0.3126,
+      "step": 850
+    },
+    {
+      "epoch": 7.273504273504273,
+      "grad_norm": 1.0272875948331592,
+      "learning_rate": 2.6185484121384974e-06,
+      "loss": 0.2403,
+      "step": 851
+    },
+    {
+      "epoch": 7.282051282051282,
+      "grad_norm": 0.7868876719936091,
+      "learning_rate": 2.6140780556719354e-06,
+      "loss": 0.0742,
+      "step": 852
+    },
+    {
+      "epoch": 7.2905982905982905,
+      "grad_norm": 0.6363777029874365,
+      "learning_rate": 2.6096073336547757e-06,
+      "loss": 0.1132,
+      "step": 853
+    },
+    {
+      "epoch": 7.299145299145299,
+      "grad_norm": 0.7336756499861784,
+      "learning_rate": 2.6051362604129553e-06,
+      "loss": 0.1026,
+      "step": 854
+    },
+    {
+      "epoch": 7.3076923076923075,
+      "grad_norm": 0.775541539538005,
+      "learning_rate": 2.6006648502735384e-06,
+      "loss": 0.1338,
+      "step": 855
+    },
+    {
+      "epoch": 7.316239316239316,
+      "grad_norm": 0.7702004755767091,
+      "learning_rate": 2.5961931175646658e-06,
+      "loss": 0.0985,
+      "step": 856
+    },
+    {
+      "epoch": 7.3247863247863245,
+      "grad_norm": 0.6316838953297962,
+      "learning_rate": 2.591721076615517e-06,
+      "loss": 0.1023,
+      "step": 857
+    },
+    {
+      "epoch": 7.333333333333333,
+      "grad_norm": 0.7607344265011406,
+      "learning_rate": 2.587248741756253e-06,
+      "loss": 0.1742,
+      "step": 858
+    },
+    {
+      "epoch": 7.3418803418803416,
+      "grad_norm": 0.6901095678494874,
+      "learning_rate": 2.5827761273179795e-06,
+      "loss": 0.2782,
+      "step": 859
+    },
+    {
+      "epoch": 7.35042735042735,
+      "grad_norm": 1.0303443002912978,
+      "learning_rate": 2.578303247632701e-06,
+      "loss": 0.1444,
+      "step": 860
+    },
+    {
+      "epoch": 7.358974358974359,
+      "grad_norm": 0.879389257012702,
+      "learning_rate": 2.5738301170332665e-06,
+      "loss": 0.2016,
+      "step": 861
+    },
+    {
+      "epoch": 7.367521367521368,
+      "grad_norm": 0.758006315638393,
+      "learning_rate": 2.5693567498533315e-06,
+      "loss": 0.107,
+      "step": 862
+    },
+    {
+      "epoch": 7.3760683760683765,
+      "grad_norm": 0.741991263996712,
+      "learning_rate": 2.5648831604273117e-06,
+      "loss": 0.1058,
+      "step": 863
+    },
+    {
+      "epoch": 7.384615384615385,
+      "grad_norm": 0.5885181500724646,
+      "learning_rate": 2.560409363090331e-06,
+      "loss": 0.0387,
+      "step": 864
+    },
+    {
+      "epoch": 7.3931623931623935,
+      "grad_norm": 0.6251334241704289,
+      "learning_rate": 2.555935372178183e-06,
+      "loss": 0.0529,
+      "step": 865
+    },
+    {
+      "epoch": 7.401709401709402,
+      "grad_norm": 0.7108430489857732,
+      "learning_rate": 2.5514612020272792e-06,
+      "loss": 0.0515,
+      "step": 866
+    },
+    {
+      "epoch": 7.410256410256411,
+      "grad_norm": 0.5671854664656416,
+      "learning_rate": 2.546986866974606e-06,
+      "loss": 0.0936,
+      "step": 867
+    },
+    {
+      "epoch": 7.418803418803419,
+      "grad_norm": 0.7055757070638874,
+      "learning_rate": 2.54251238135768e-06,
+      "loss": 0.1075,
+      "step": 868
+    },
+    {
+      "epoch": 7.427350427350428,
+      "grad_norm": 1.003935743284772,
+      "learning_rate": 2.5380377595144984e-06,
+      "loss": 0.2216,
+      "step": 869
+    },
+    {
+      "epoch": 7.435897435897436,
+      "grad_norm": 0.9449024720361391,
+      "learning_rate": 2.533563015783494e-06,
+      "loss": 0.2309,
+      "step": 870
+    },
+    {
+      "epoch": 7.444444444444445,
+      "grad_norm": 0.7129554344380816,
+      "learning_rate": 2.5290881645034932e-06,
+      "loss": 0.1252,
+      "step": 871
+    },
+    {
+      "epoch": 7.452991452991453,
+      "grad_norm": 1.0412648511502645,
+      "learning_rate": 2.524613220013664e-06,
+      "loss": 0.0476,
+      "step": 872
+    },
+    {
+      "epoch": 7.461538461538462,
+      "grad_norm": 1.4336752325274813,
+      "learning_rate": 2.5201381966534748e-06,
+      "loss": 0.433,
+      "step": 873
+    },
+    {
+      "epoch": 7.47008547008547,
+      "grad_norm": 1.2607405383601524,
+      "learning_rate": 2.515663108762648e-06,
+      "loss": 0.1975,
+      "step": 874
+    },
+    {
+      "epoch": 7.478632478632479,
+      "grad_norm": 0.9558946359084702,
+      "learning_rate": 2.511187970681109e-06,
+      "loss": 0.139,
+      "step": 875
+    },
+    {
+      "epoch": 7.487179487179487,
+      "grad_norm": 0.8425542820829937,
+      "learning_rate": 2.5067127967489464e-06,
+      "loss": 0.1183,
+      "step": 876
+    },
+    {
+      "epoch": 7.495726495726496,
+      "grad_norm": 0.628921282602329,
+      "learning_rate": 2.5022376013063653e-06,
+      "loss": 0.0976,
+      "step": 877
+    },
+    {
+      "epoch": 7.504273504273504,
+      "grad_norm": 0.7245192855197514,
+      "learning_rate": 2.497762398693636e-06,
+      "loss": 0.1356,
+      "step": 878
+    },
+    {
+      "epoch": 7.512820512820513,
+      "grad_norm": 0.8889602741073305,
+      "learning_rate": 2.493287203251054e-06,
+      "loss": 0.2406,
+      "step": 879
+    },
+    {
+      "epoch": 7.521367521367521,
+      "grad_norm": 1.2959721074337187,
+      "learning_rate": 2.4888120293188915e-06,
+      "loss": 0.2891,
+      "step": 880
+    },
+    {
+      "epoch": 7.52991452991453,
+      "grad_norm": 0.927145734964671,
+      "learning_rate": 2.484336891237353e-06,
+      "loss": 0.186,
+      "step": 881
+    },
+    {
+      "epoch": 7.538461538461538,
+      "grad_norm": 0.7231331416771678,
+      "learning_rate": 2.4798618033465256e-06,
+      "loss": 0.1549,
+      "step": 882
+    },
+    {
+      "epoch": 7.547008547008547,
+      "grad_norm": 0.7356989966291199,
+      "learning_rate": 2.4753867799863365e-06,
+      "loss": 0.0729,
+      "step": 883
+    },
+    {
+      "epoch": 7.555555555555555,
+      "grad_norm": 1.1297589765273541,
+      "learning_rate": 2.470911835496508e-06,
+      "loss": 0.1264,
+      "step": 884
+    },
+    {
+      "epoch": 7.564102564102564,
+      "grad_norm": 0.7268249250400882,
+      "learning_rate": 2.466436984216507e-06,
+      "loss": 0.1948,
+      "step": 885
+    },
+    {
+      "epoch": 7.572649572649572,
+      "grad_norm": 1.0746091944481344,
+      "learning_rate": 2.4619622404855025e-06,
+      "loss": 0.0925,
+      "step": 886
+    },
+    {
+      "epoch": 7.581196581196581,
+      "grad_norm": 0.9343825196353718,
+      "learning_rate": 2.4574876186423203e-06,
+      "loss": 0.1994,
+      "step": 887
+    },
+    {
+      "epoch": 7.589743589743589,
+      "grad_norm": 0.8818256523384325,
+      "learning_rate": 2.4530131330253946e-06,
+      "loss": 0.3004,
+      "step": 888
+    },
+    {
+      "epoch": 7.598290598290598,
+      "grad_norm": 0.6933798747538589,
+      "learning_rate": 2.4485387979727216e-06,
+      "loss": 0.0969,
+      "step": 889
+    },
+    {
+      "epoch": 7.6068376068376065,
+      "grad_norm": 1.05137189120055,
+      "learning_rate": 2.4440646278218178e-06,
+      "loss": 0.1033,
+      "step": 890
+    },
+    {
+      "epoch": 7.615384615384615,
+      "grad_norm": 0.8078492077071381,
+      "learning_rate": 2.43959063690967e-06,
+      "loss": 0.1876,
+      "step": 891
+    },
+    {
+      "epoch": 7.6239316239316235,
+      "grad_norm": 0.8109878148999148,
+      "learning_rate": 2.435116839572689e-06,
+      "loss": 0.2169,
+      "step": 892
+    },
+    {
+      "epoch": 7.632478632478632,
+      "grad_norm": 1.056400836892595,
+      "learning_rate": 2.430643250146669e-06,
+      "loss": 0.2165,
+      "step": 893
+    },
+    {
+      "epoch": 7.641025641025641,
+      "grad_norm": 0.8285219700841594,
+      "learning_rate": 2.426169882966735e-06,
+      "loss": 0.2727,
+      "step": 894
+    },
+    {
+      "epoch": 7.64957264957265,
+      "grad_norm": 0.9095468547064846,
+      "learning_rate": 2.4216967523673e-06,
+      "loss": 0.1749,
+      "step": 895
+    },
+    {
+      "epoch": 7.6581196581196584,
+      "grad_norm": 0.5146017507927608,
+      "learning_rate": 2.4172238726820205e-06,
+      "loss": 0.0588,
+      "step": 896
+    },
+    {
+      "epoch": 7.666666666666667,
+      "grad_norm": 0.7662282433470485,
+      "learning_rate": 2.4127512582437486e-06,
+      "loss": 0.1758,
+      "step": 897
+    },
+    {
+      "epoch": 7.6752136752136755,
+      "grad_norm": 0.8890016478017211,
+      "learning_rate": 2.4082789233844837e-06,
+      "loss": 0.1344,
+      "step": 898
+    },
+    {
+      "epoch": 7.683760683760684,
+      "grad_norm": 0.8045854211885298,
+      "learning_rate": 2.403806882435334e-06,
+      "loss": 0.1668,
+      "step": 899
+    },
+    {
+      "epoch": 7.6923076923076925,
+      "grad_norm": 0.9376743164546943,
+      "learning_rate": 2.399335149726463e-06,
+      "loss": 0.1746,
+      "step": 900
+    },
+    {
+      "epoch": 7.700854700854701,
+      "grad_norm": 0.8484344804273012,
+      "learning_rate": 2.394863739587045e-06,
+      "loss": 0.0338,
+      "step": 901
+    },
+    {
+      "epoch": 7.7094017094017095,
+      "grad_norm": 0.8856756769399486,
+      "learning_rate": 2.3903926663452255e-06,
+      "loss": 0.1161,
+      "step": 902
+    },
+    {
+      "epoch": 7.717948717948718,
+      "grad_norm": 0.6232667832934703,
+      "learning_rate": 2.385921944328066e-06,
+      "loss": 0.0549,
+      "step": 903
+    },
+    {
+      "epoch": 7.726495726495727,
+      "grad_norm": 1.0427742949381442,
+      "learning_rate": 2.3814515878615035e-06,
+      "loss": 0.1645,
+      "step": 904
+    },
+    {
+      "epoch": 7.735042735042735,
+      "grad_norm": 0.8121694579839741,
+      "learning_rate": 2.376981611270305e-06,
+      "loss": 0.1664,
+      "step": 905
+    },
+    {
+      "epoch": 7.743589743589744,
+      "grad_norm": 0.7659730280591774,
+      "learning_rate": 2.372512028878019e-06,
+      "loss": 0.1028,
+      "step": 906
+    },
+    {
+      "epoch": 7.752136752136752,
+      "grad_norm": 0.5835938881519352,
+      "learning_rate": 2.3680428550069327e-06,
+      "loss": 0.0853,
+      "step": 907
+    },
+    {
+      "epoch": 7.760683760683761,
+      "grad_norm": 0.6898088371986518,
+      "learning_rate": 2.3635741039780214e-06,
+      "loss": 0.1352,
+      "step": 908
+    },
+    {
+      "epoch": 7.769230769230769,
+      "grad_norm": 1.0941235835301037,
+      "learning_rate": 2.3591057901109063e-06,
+      "loss": 0.2384,
+      "step": 909
+    },
+    {
+      "epoch": 7.777777777777778,
+      "grad_norm": 1.0985158914441384,
+      "learning_rate": 2.3546379277238107e-06,
+      "loss": 0.1356,
+      "step": 910
+    },
+    {
+      "epoch": 7.786324786324786,
+      "grad_norm": 0.8285970493804228,
+      "learning_rate": 2.3501705311335067e-06,
+      "loss": 0.0798,
+      "step": 911
+    },
+    {
+      "epoch": 7.794871794871795,
+      "grad_norm": 0.7048258885001851,
+      "learning_rate": 2.3457036146552766e-06,
+      "loss": 0.0889,
+      "step": 912
+    },
+    {
+      "epoch": 7.803418803418803,
+      "grad_norm": 1.3004201397168884,
+      "learning_rate": 2.341237192602864e-06,
+      "loss": 0.2357,
+      "step": 913
+    },
+    {
+      "epoch": 7.811965811965812,
+      "grad_norm": 0.8101870477166999,
+      "learning_rate": 2.336771279288427e-06,
+      "loss": 0.0772,
+      "step": 914
+    },
+    {
+      "epoch": 7.82051282051282,
+      "grad_norm": 0.93339167381936,
+      "learning_rate": 2.332305889022494e-06,
+      "loss": 0.0886,
+      "step": 915
+    },
+    {
+      "epoch": 7.829059829059829,
+      "grad_norm": 0.7680114645074881,
+      "learning_rate": 2.3278410361139198e-06,
+      "loss": 0.2533,
+      "step": 916
+    },
+    {
+      "epoch": 7.837606837606837,
+      "grad_norm": 1.0280566949846137,
+      "learning_rate": 2.3233767348698314e-06,
+      "loss": 0.1258,
+      "step": 917
+    },
+    {
+      "epoch": 7.846153846153846,
+      "grad_norm": 0.7516417773033046,
+      "learning_rate": 2.3189129995955944e-06,
+      "loss": 0.2266,
+      "step": 918
+    },
+    {
+      "epoch": 7.854700854700854,
+      "grad_norm": 0.9280988237858805,
+      "learning_rate": 2.314449844594758e-06,
+      "loss": 0.2605,
+      "step": 919
+    },
+    {
+      "epoch": 7.863247863247864,
+      "grad_norm": 0.9031395365665326,
+      "learning_rate": 2.3099872841690105e-06,
+      "loss": 0.2555,
+      "step": 920
+    },
+    {
+      "epoch": 7.871794871794872,
+      "grad_norm": 0.8324029751037804,
+      "learning_rate": 2.305525332618138e-06,
+      "loss": 0.1983,
+      "step": 921
+    },
+    {
+      "epoch": 7.880341880341881,
+      "grad_norm": 1.0373932446642553,
+      "learning_rate": 2.3010640042399748e-06,
+      "loss": 0.2108,
+      "step": 922
+    },
+    {
+      "epoch": 7.888888888888889,
+      "grad_norm": 0.7324143576403153,
+      "learning_rate": 2.296603313330355e-06,
+      "loss": 0.0424,
+      "step": 923
+    },
+    {
+      "epoch": 7.897435897435898,
+      "grad_norm": 0.8313171728758167,
+      "learning_rate": 2.2921432741830744e-06,
+      "loss": 0.2036,
+      "step": 924
+    },
+    {
+      "epoch": 7.905982905982906,
+      "grad_norm": 0.5819353015291171,
+      "learning_rate": 2.2876839010898377e-06,
+      "loss": 0.0989,
+      "step": 925
+    },
+    {
+      "epoch": 7.914529914529915,
+      "grad_norm": 0.9657210621158413,
+      "learning_rate": 2.283225208340218e-06,
+      "loss": 0.1124,
+      "step": 926
+    },
+    {
+      "epoch": 7.923076923076923,
+      "grad_norm": 0.8348772843589627,
+      "learning_rate": 2.2787672102216045e-06,
+      "loss": 0.1371,
+      "step": 927
+    },
+    {
+      "epoch": 7.931623931623932,
+      "grad_norm": 1.0071415066110063,
+      "learning_rate": 2.2743099210191623e-06,
+      "loss": 0.2917,
+      "step": 928
+    },
+    {
+      "epoch": 7.94017094017094,
+      "grad_norm": 0.7480676950520273,
+      "learning_rate": 2.2698533550157865e-06,
+      "loss": 0.1117,
+      "step": 929
+    },
+    {
+      "epoch": 7.948717948717949,
+      "grad_norm": 0.6984275302001026,
+      "learning_rate": 2.265397526492052e-06,
+      "loss": 0.1257,
+      "step": 930
+    },
+    {
+      "epoch": 7.957264957264957,
+      "grad_norm": 1.0136360419533816,
+      "learning_rate": 2.2609424497261723e-06,
+      "loss": 0.1716,
+      "step": 931
+    },
+    {
+      "epoch": 7.965811965811966,
+      "grad_norm": 0.789818664864703,
+      "learning_rate": 2.2564881389939524e-06,
+      "loss": 0.1089,
+      "step": 932
+    },
+    {
+      "epoch": 7.9743589743589745,
+      "grad_norm": 0.801543110176476,
+      "learning_rate": 2.25203460856874e-06,
+      "loss": 0.1356,
+      "step": 933
+    },
+    {
+      "epoch": 7.982905982905983,
+      "grad_norm": 1.079998000143528,
+      "learning_rate": 2.2475818727213843e-06,
+      "loss": 0.1016,
+      "step": 934
+    },
+    {
+      "epoch": 7.9914529914529915,
+      "grad_norm": 1.2224492900647317,
+      "learning_rate": 2.24312994572019e-06,
+      "loss": 0.0684,
+      "step": 935
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.7840113283449409,
+      "learning_rate": 2.238678841830867e-06,
+      "loss": 0.26,
+      "step": 936
+    },
+    {
+      "epoch": 8.008547008547009,
+      "grad_norm": 0.6801213213487027,
+      "learning_rate": 2.2342285753164876e-06,
+      "loss": 0.0874,
+      "step": 937
+    },
+    {
+      "epoch": 8.017094017094017,
+      "grad_norm": 0.8205242974812631,
+      "learning_rate": 2.2297791604374443e-06,
+      "loss": 0.0564,
+      "step": 938
+    },
+    {
+      "epoch": 8.025641025641026,
+      "grad_norm": 0.5970829689339425,
+      "learning_rate": 2.2253306114513964e-06,
+      "loss": 0.0707,
+      "step": 939
+    },
+    {
+      "epoch": 8.034188034188034,
+      "grad_norm": 0.7521912608976296,
+      "learning_rate": 2.220882942613231e-06,
+      "loss": 0.1745,
+      "step": 940
+    },
+    {
+      "epoch": 8.042735042735043,
+      "grad_norm": 0.7838835824189015,
+      "learning_rate": 2.2164361681750148e-06,
+      "loss": 0.0816,
+      "step": 941
+    },
+    {
+      "epoch": 8.051282051282051,
+      "grad_norm": 0.8426679690138106,
+      "learning_rate": 2.2119903023859475e-06,
+      "loss": 0.0916,
+      "step": 942
+    },
+    {
+      "epoch": 8.05982905982906,
+      "grad_norm": 1.1073502756448632,
+      "learning_rate": 2.2075453594923175e-06,
+      "loss": 0.1688,
+      "step": 943
+    },
+    {
+      "epoch": 8.068376068376068,
+      "grad_norm": 0.9922388944200498,
+      "learning_rate": 2.2031013537374564e-06,
+      "loss": 0.1957,
+      "step": 944
+    },
+    {
+      "epoch": 8.076923076923077,
+      "grad_norm": 0.6640668878424266,
+      "learning_rate": 2.1986582993616926e-06,
+      "loss": 0.0987,
+      "step": 945
+    },
+    {
+      "epoch": 8.085470085470085,
+      "grad_norm": 1.0112842704537792,
+      "learning_rate": 2.194216210602305e-06,
+      "loss": 0.1037,
+      "step": 946
+    },
+    {
+      "epoch": 8.094017094017094,
+      "grad_norm": 0.8175471256641587,
+      "learning_rate": 2.1897751016934802e-06,
+      "loss": 0.1412,
+      "step": 947
+    },
+    {
+      "epoch": 8.102564102564102,
+      "grad_norm": 0.5118679059079263,
+      "learning_rate": 2.1853349868662637e-06,
+      "loss": 0.0745,
+      "step": 948
+    },
+    {
+      "epoch": 8.11111111111111,
+      "grad_norm": 0.9461630409984287,
+      "learning_rate": 2.1808958803485134e-06,
+      "loss": 0.1013,
+      "step": 949
+    },
+    {
+      "epoch": 8.11965811965812,
+      "grad_norm": 0.9858064756039681,
+      "learning_rate": 2.1764577963648616e-06,
+      "loss": 0.143,
+      "step": 950
+    },
+    {
+      "epoch": 8.128205128205128,
+      "grad_norm": 0.8827829292121262,
+      "learning_rate": 2.1720207491366598e-06,
+      "loss": 0.0624,
+      "step": 951
+    },
+    {
+      "epoch": 8.136752136752136,
+      "grad_norm": 0.6546653319429925,
+      "learning_rate": 2.167584752881937e-06,
+      "loss": 0.0823,
+      "step": 952
+    },
+    {
+      "epoch": 8.145299145299145,
+      "grad_norm": 1.0092688749232046,
+      "learning_rate": 2.163149821815358e-06,
+      "loss": 0.235,
+      "step": 953
+    },
+    {
+      "epoch": 8.153846153846153,
+      "grad_norm": 0.8771490405627712,
+      "learning_rate": 2.1587159701481718e-06,
+      "loss": 0.2528,
+      "step": 954
+    },
+    {
+      "epoch": 8.162393162393162,
+      "grad_norm": 0.946709599885579,
+      "learning_rate": 2.154283212088168e-06,
+      "loss": 0.1608,
+      "step": 955
+    },
+    {
+      "epoch": 8.17094017094017,
+      "grad_norm": 0.5952914827869664,
+      "learning_rate": 2.1498515618396327e-06,
+      "loss": 0.1172,
+      "step": 956
+    },
+    {
+      "epoch": 8.179487179487179,
+      "grad_norm": 0.6912864743108678,
+      "learning_rate": 2.145421033603304e-06,
+      "loss": 0.0112,
+      "step": 957
+    },
+    {
+      "epoch": 8.188034188034187,
+      "grad_norm": 0.6662988373552396,
+      "learning_rate": 2.1409916415763216e-06,
+      "loss": 0.0755,
+      "step": 958
+    },
+    {
+      "epoch": 8.196581196581196,
+      "grad_norm": 0.7960679547494807,
+      "learning_rate": 2.1365633999521852e-06,
+      "loss": 0.0607,
+      "step": 959
+    },
+    {
+      "epoch": 8.205128205128204,
+      "grad_norm": 0.8508884811265452,
+      "learning_rate": 2.1321363229207097e-06,
+      "loss": 0.1314,
+      "step": 960
+    },
+    {
+      "epoch": 8.213675213675213,
+      "grad_norm": 0.8298713309207288,
+      "learning_rate": 2.127710424667978e-06,
+      "loss": 0.1361,
+      "step": 961
+    },
+    {
+      "epoch": 8.222222222222221,
+      "grad_norm": 0.5117722405915415,
+      "learning_rate": 2.1232857193762923e-06,
+      "loss": 0.0464,
+      "step": 962
+    },
+    {
+      "epoch": 8.23076923076923,
+      "grad_norm": 0.5227615536686315,
+      "learning_rate": 2.1188622212241366e-06,
+      "loss": 0.0492,
+      "step": 963
+    },
+    {
+      "epoch": 8.239316239316238,
+      "grad_norm": 1.0146157619895304,
+      "learning_rate": 2.114439944386125e-06,
+      "loss": 0.1391,
+      "step": 964
+    },
+    {
+      "epoch": 8.247863247863247,
+      "grad_norm": 1.002266709631584,
+      "learning_rate": 2.1100189030329557e-06,
+      "loss": 0.1537,
+      "step": 965
+    },
+    {
+      "epoch": 8.256410256410255,
+      "grad_norm": 0.6498632887785525,
+      "learning_rate": 2.105599111331372e-06,
+      "loss": 0.0452,
+      "step": 966
+    },
+    {
+      "epoch": 8.264957264957266,
+      "grad_norm": 0.9333635521000488,
+      "learning_rate": 2.101180583444111e-06,
+      "loss": 0.1001,
+      "step": 967
+    },
+    {
+      "epoch": 8.273504273504274,
+      "grad_norm": 0.8310263706256398,
+      "learning_rate": 2.0967633335298583e-06,
+      "loss": 0.1492,
+      "step": 968
+    },
+    {
+      "epoch": 8.282051282051283,
+      "grad_norm": 0.5899470844056727,
+      "learning_rate": 2.0923473757432073e-06,
+      "loss": 0.0545,
+      "step": 969
+    },
+    {
+      "epoch": 8.290598290598291,
+      "grad_norm": 1.0233804469655359,
+      "learning_rate": 2.0879327242346096e-06,
+      "loss": 0.0926,
+      "step": 970
+    },
+    {
+      "epoch": 8.2991452991453,
+      "grad_norm": 1.0819059667991204,
+      "learning_rate": 2.0835193931503297e-06,
+      "loss": 0.1311,
+      "step": 971
+    },
+    {
+      "epoch": 8.307692307692308,
+      "grad_norm": 0.46285369597354903,
+      "learning_rate": 2.079107396632404e-06,
+      "loss": 0.1046,
+      "step": 972
+    },
+    {
+      "epoch": 8.316239316239317,
+      "grad_norm": 1.0596263414741827,
+      "learning_rate": 2.0746967488185903e-06,
+      "loss": 0.1922,
+      "step": 973
+    },
+    {
+      "epoch": 8.324786324786325,
+      "grad_norm": 0.652779105438659,
+      "learning_rate": 2.0702874638423233e-06,
+      "loss": 0.1108,
+      "step": 974
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 0.8934098392963533,
+      "learning_rate": 2.0658795558326745e-06,
+      "loss": 0.1306,
+      "step": 975
+    },
+    {
+      "epoch": 8.341880341880342,
+      "grad_norm": 0.6486169868086549,
+      "learning_rate": 2.0614730389143004e-06,
+      "loss": 0.0647,
+      "step": 976
+    },
+    {
+      "epoch": 8.350427350427351,
+      "grad_norm": 0.6465097746250824,
+      "learning_rate": 2.057067927207399e-06,
+      "loss": 0.0592,
+      "step": 977
+    },
+    {
+      "epoch": 8.35897435897436,
+      "grad_norm": 0.9112154978141302,
+      "learning_rate": 2.052664234827668e-06,
+      "loss": 0.1674,
+      "step": 978
+    },
+    {
+      "epoch": 8.367521367521368,
+      "grad_norm": 1.049107550845732,
+      "learning_rate": 2.048261975886256e-06,
+      "loss": 0.2277,
+      "step": 979
+    },
+    {
+      "epoch": 8.376068376068377,
+      "grad_norm": 0.6487345522229879,
+      "learning_rate": 2.0438611644897186e-06,
+      "loss": 0.0773,
+      "step": 980
+    },
+    {
+      "epoch": 8.384615384615385,
+      "grad_norm": 0.5377657888294962,
+      "learning_rate": 2.0394618147399713e-06,
+      "loss": 0.1007,
+      "step": 981
+    },
+    {
+      "epoch": 8.393162393162394,
+      "grad_norm": 0.6621738996727564,
+      "learning_rate": 2.0350639407342474e-06,
+      "loss": 0.0513,
+      "step": 982
+    },
+    {
+      "epoch": 8.401709401709402,
+      "grad_norm": 0.4872035195253045,
+      "learning_rate": 2.030667556565052e-06,
+      "loss": 0.0265,
+      "step": 983
+    },
+    {
+      "epoch": 8.41025641025641,
+      "grad_norm": 0.9739209495675087,
+      "learning_rate": 2.026272676320114e-06,
+      "loss": 0.1961,
+      "step": 984
+    },
+    {
+      "epoch": 8.418803418803419,
+      "grad_norm": 0.5372233867525106,
+      "learning_rate": 2.021879314082344e-06,
+      "loss": 0.1294,
+      "step": 985
+    },
+    {
+      "epoch": 8.427350427350428,
+      "grad_norm": 0.8921860227920625,
+      "learning_rate": 2.0174874839297912e-06,
+      "loss": 0.1127,
+      "step": 986
+    },
+    {
+      "epoch": 8.435897435897436,
+      "grad_norm": 0.9625191449275724,
+      "learning_rate": 2.01309719993559e-06,
+      "loss": 0.1532,
+      "step": 987
+    },
+    {
+      "epoch": 8.444444444444445,
+      "grad_norm": 1.028032603670202,
+      "learning_rate": 2.0087084761679245e-06,
+      "loss": 0.1056,
+      "step": 988
+    },
+    {
+      "epoch": 8.452991452991453,
+      "grad_norm": 0.6625782427295153,
+      "learning_rate": 2.0043213266899787e-06,
+      "loss": 0.1335,
+      "step": 989
+    },
+    {
+      "epoch": 8.461538461538462,
+      "grad_norm": 1.0598695524412451,
+      "learning_rate": 1.9999357655598894e-06,
+      "loss": 0.2141,
+      "step": 990
+    },
+    {
+      "epoch": 8.47008547008547,
+      "grad_norm": 0.6690236702990353,
+      "learning_rate": 1.995551806830706e-06,
+      "loss": 0.1276,
+      "step": 991
+    },
+    {
+      "epoch": 8.478632478632479,
+      "grad_norm": 0.60166856759265,
+      "learning_rate": 1.9911694645503443e-06,
+      "loss": 0.0579,
+      "step": 992
+    },
+    {
+      "epoch": 8.487179487179487,
+      "grad_norm": 0.9847593137774459,
+      "learning_rate": 1.986788752761536e-06,
+      "loss": 0.111,
+      "step": 993
+    },
+    {
+      "epoch": 8.495726495726496,
+      "grad_norm": 0.982024009501397,
+      "learning_rate": 1.9824096855017922e-06,
+      "loss": 0.2799,
+      "step": 994
+    },
+    {
+      "epoch": 8.504273504273504,
+      "grad_norm": 0.7761724159948599,
+      "learning_rate": 1.978032276803354e-06,
+      "loss": 0.1645,
+      "step": 995
+    },
+    {
+      "epoch": 8.512820512820513,
+      "grad_norm": 0.8471655309381408,
+      "learning_rate": 1.9736565406931446e-06,
+      "loss": 0.1332,
+      "step": 996
+    },
+    {
+      "epoch": 8.521367521367521,
+      "grad_norm": 0.951981738853562,
+      "learning_rate": 1.969282491192731e-06,
+      "loss": 0.1401,
+      "step": 997
+    },
+    {
+      "epoch": 8.52991452991453,
+      "grad_norm": 0.6923660244868259,
+      "learning_rate": 1.9649101423182732e-06,
+      "loss": 0.0843,
+      "step": 998
+    },
+    {
+      "epoch": 8.538461538461538,
+      "grad_norm": 0.8000814315215867,
+      "learning_rate": 1.960539508080485e-06,
+      "loss": 0.2153,
+      "step": 999
+    },
+    {
+      "epoch": 8.547008547008547,
+      "grad_norm": 0.6617963864096438,
+      "learning_rate": 1.956170602484582e-06,
+      "loss": 0.0804,
+      "step": 1000
+    },
+    {
+      "epoch": 8.555555555555555,
+      "grad_norm": 0.8352900774008456,
+      "learning_rate": 1.9518034395302413e-06,
+      "loss": 0.2491,
+      "step": 1001
+    },
+    {
+      "epoch": 8.564102564102564,
+      "grad_norm": 0.5497872023851514,
+      "learning_rate": 1.94743803321156e-06,
+      "loss": 0.0348,
+      "step": 1002
+    },
+    {
+      "epoch": 8.572649572649572,
+      "grad_norm": 0.5285674845565532,
+      "learning_rate": 1.9430743975170004e-06,
+      "loss": 0.0767,
+      "step": 1003
+    },
+    {
+      "epoch": 8.581196581196581,
+      "grad_norm": 0.9005822868435749,
+      "learning_rate": 1.938712546429354e-06,
+      "loss": 0.1976,
+      "step": 1004
+    },
+    {
+      "epoch": 8.58974358974359,
+      "grad_norm": 0.8375381738831286,
+      "learning_rate": 1.934352493925695e-06,
+      "loss": 0.2831,
+      "step": 1005
+    },
+    {
+      "epoch": 8.598290598290598,
+      "grad_norm": 0.7340948938782793,
+      "learning_rate": 1.9299942539773316e-06,
+      "loss": 0.0736,
+      "step": 1006
+    },
+    {
+      "epoch": 8.606837606837606,
+      "grad_norm": 0.39349115495287446,
+      "learning_rate": 1.925637840549764e-06,
+      "loss": 0.05,
+      "step": 1007
+    },
+    {
+      "epoch": 8.615384615384615,
+      "grad_norm": 0.8352497021594969,
+      "learning_rate": 1.921283267602643e-06,
+      "loss": 0.133,
+      "step": 1008
+    },
+    {
+      "epoch": 8.623931623931623,
+      "grad_norm": 0.8392365988719928,
+      "learning_rate": 1.9169305490897173e-06,
+      "loss": 0.1195,
+      "step": 1009
+    },
+    {
+      "epoch": 8.632478632478632,
+      "grad_norm": 0.7904108060873761,
+      "learning_rate": 1.9125796989587947e-06,
+      "loss": 0.1938,
+      "step": 1010
+    },
+    {
+      "epoch": 8.64102564102564,
+      "grad_norm": 0.8645717752205156,
+      "learning_rate": 1.9082307311516985e-06,
+      "loss": 0.1548,
+      "step": 1011
+    },
+    {
+      "epoch": 8.649572649572649,
+      "grad_norm": 0.5395543064003644,
+      "learning_rate": 1.9038836596042174e-06,
+      "loss": 0.0941,
+      "step": 1012
+    },
+    {
+      "epoch": 8.658119658119658,
+      "grad_norm": 0.9345772701049747,
+      "learning_rate": 1.8995384982460636e-06,
+      "loss": 0.1512,
+      "step": 1013
+    },
+    {
+      "epoch": 8.666666666666666,
+      "grad_norm": 0.7617202728674364,
+      "learning_rate": 1.895195261000831e-06,
+      "loss": 0.0691,
+      "step": 1014
+    },
+    {
+      "epoch": 8.675213675213675,
+      "grad_norm": 0.4683005730689552,
+      "learning_rate": 1.8908539617859455e-06,
+      "loss": 0.0309,
+      "step": 1015
+    },
+    {
+      "epoch": 8.683760683760683,
+      "grad_norm": 1.0820714752748837,
+      "learning_rate": 1.8865146145126228e-06,
+      "loss": 0.095,
+      "step": 1016
+    },
+    {
+      "epoch": 8.692307692307692,
+      "grad_norm": 0.5167947937067587,
+      "learning_rate": 1.8821772330858259e-06,
+      "loss": 0.0576,
+      "step": 1017
+    },
+    {
+      "epoch": 8.7008547008547,
+      "grad_norm": 0.5618671513966412,
+      "learning_rate": 1.877841831404217e-06,
+      "loss": 0.0563,
+      "step": 1018
+    },
+    {
+      "epoch": 8.709401709401709,
+      "grad_norm": 0.7395751100091337,
+      "learning_rate": 1.873508423360113e-06,
+      "loss": 0.0616,
+      "step": 1019
+    },
+    {
+      "epoch": 8.717948717948717,
+      "grad_norm": 0.9613190639149091,
+      "learning_rate": 1.8691770228394458e-06,
+      "loss": 0.3077,
+      "step": 1020
+    },
+    {
+      "epoch": 8.726495726495726,
+      "grad_norm": 0.7355368158189639,
+      "learning_rate": 1.8648476437217117e-06,
+      "loss": 0.1128,
+      "step": 1021
+    },
+    {
+      "epoch": 8.735042735042736,
+      "grad_norm": 0.6420518723233007,
+      "learning_rate": 1.8605202998799299e-06,
+      "loss": 0.138,
+      "step": 1022
+    },
+    {
+      "epoch": 8.743589743589745,
+      "grad_norm": 0.8047403589848431,
+      "learning_rate": 1.8561950051805994e-06,
+      "loss": 0.1726,
+      "step": 1023
+    },
+    {
+      "epoch": 8.752136752136753,
+      "grad_norm": 0.7614318567549201,
+      "learning_rate": 1.8518717734836522e-06,
+      "loss": 0.1479,
+      "step": 1024
+    },
+    {
+      "epoch": 8.760683760683762,
+      "grad_norm": 0.9470748204080588,
+      "learning_rate": 1.8475506186424075e-06,
+      "loss": 0.1853,
+      "step": 1025
+    },
+    {
+      "epoch": 8.76923076923077,
+      "grad_norm": 1.6630512840142453,
+      "learning_rate": 1.8432315545035328e-06,
+      "loss": 0.0757,
+      "step": 1026
+    },
+    {
+      "epoch": 8.777777777777779,
+      "grad_norm": 1.2434105142184977,
+      "learning_rate": 1.8389145949069953e-06,
+      "loss": 0.0686,
+      "step": 1027
+    },
+    {
+      "epoch": 8.786324786324787,
+      "grad_norm": 0.8550828646975722,
+      "learning_rate": 1.8345997536860154e-06,
+      "loss": 0.1574,
+      "step": 1028
+    },
+    {
+      "epoch": 8.794871794871796,
+      "grad_norm": 0.6583015306693577,
+      "learning_rate": 1.83028704466703e-06,
+      "loss": 0.0519,
+      "step": 1029
+    },
+    {
+      "epoch": 8.803418803418804,
+      "grad_norm": 0.8504516863547416,
+      "learning_rate": 1.8259764816696413e-06,
+      "loss": 0.0567,
+      "step": 1030
+    },
+    {
+      "epoch": 8.811965811965813,
+      "grad_norm": 1.2133627934038262,
+      "learning_rate": 1.8216680785065734e-06,
+      "loss": 0.1414,
+      "step": 1031
+    },
+    {
+      "epoch": 8.820512820512821,
+      "grad_norm": 0.7975309833677979,
+      "learning_rate": 1.8173618489836315e-06,
+      "loss": 0.0269,
+      "step": 1032
+    },
+    {
+      "epoch": 8.82905982905983,
+      "grad_norm": 0.8326288867470483,
+      "learning_rate": 1.813057806899656e-06,
+      "loss": 0.1349,
+      "step": 1033
+    },
+    {
+      "epoch": 8.837606837606838,
+      "grad_norm": 0.8892203026048484,
+      "learning_rate": 1.8087559660464766e-06,
+      "loss": 0.1477,
+      "step": 1034
+    },
+    {
+      "epoch": 8.846153846153847,
+      "grad_norm": 1.045254181038715,
+      "learning_rate": 1.8044563402088686e-06,
+      "loss": 0.1232,
+      "step": 1035
+    },
+    {
+      "epoch": 8.854700854700855,
+      "grad_norm": 1.0373914190120757,
+      "learning_rate": 1.800158943164512e-06,
+      "loss": 0.0521,
+      "step": 1036
+    },
+    {
+      "epoch": 8.863247863247864,
+      "grad_norm": 0.4542191780940718,
+      "learning_rate": 1.7958637886839437e-06,
+      "loss": 0.0367,
+      "step": 1037
+    },
+    {
+      "epoch": 8.871794871794872,
+      "grad_norm": 0.8711063804247088,
+      "learning_rate": 1.7915708905305124e-06,
+      "loss": 0.097,
+      "step": 1038
+    },
+    {
+      "epoch": 8.88034188034188,
+      "grad_norm": 0.7418297441426122,
+      "learning_rate": 1.7872802624603408e-06,
+      "loss": 0.1599,
+      "step": 1039
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 0.6832073792079075,
+      "learning_rate": 1.7829919182222752e-06,
+      "loss": 0.103,
+      "step": 1040
+    },
+    {
+      "epoch": 8.897435897435898,
+      "grad_norm": 1.139330698907758,
+      "learning_rate": 1.7787058715578415e-06,
+      "loss": 0.1249,
+      "step": 1041
+    },
+    {
+      "epoch": 8.905982905982906,
+      "grad_norm": 0.7542113764593271,
+      "learning_rate": 1.7744221362012075e-06,
+      "loss": 0.0849,
+      "step": 1042
+    },
+    {
+      "epoch": 8.914529914529915,
+      "grad_norm": 0.8049453679611847,
+      "learning_rate": 1.7701407258791323e-06,
+      "loss": 0.1231,
+      "step": 1043
+    },
+    {
+      "epoch": 8.923076923076923,
+      "grad_norm": 0.7350710259171148,
+      "learning_rate": 1.7658616543109237e-06,
+      "loss": 0.1335,
+      "step": 1044
+    },
+    {
+      "epoch": 8.931623931623932,
+      "grad_norm": 0.8851916034876457,
+      "learning_rate": 1.7615849352083975e-06,
+      "loss": 0.1177,
+      "step": 1045
+    },
+    {
+      "epoch": 8.94017094017094,
+      "grad_norm": 0.48977542938325314,
+      "learning_rate": 1.7573105822758307e-06,
+      "loss": 0.0143,
+      "step": 1046
+    },
+    {
+      "epoch": 8.948717948717949,
+      "grad_norm": 0.8288418037510138,
+      "learning_rate": 1.753038609209916e-06,
+      "loss": 0.0476,
+      "step": 1047
+    },
+    {
+      "epoch": 8.957264957264957,
+      "grad_norm": 0.6358081416813472,
+      "learning_rate": 1.7487690296997234e-06,
+      "loss": 0.0985,
+      "step": 1048
+    },
+    {
+      "epoch": 8.965811965811966,
+      "grad_norm": 0.6362613648171749,
+      "learning_rate": 1.7445018574266514e-06,
+      "loss": 0.0773,
+      "step": 1049
+    },
+    {
+      "epoch": 8.974358974358974,
+      "grad_norm": 0.7461142400839225,
+      "learning_rate": 1.740237106064383e-06,
+      "loss": 0.1196,
+      "step": 1050
+    },
+    {
+      "epoch": 8.982905982905983,
+      "grad_norm": 0.8936483774841266,
+      "learning_rate": 1.7359747892788476e-06,
+      "loss": 0.196,
+      "step": 1051
+    },
+    {
+      "epoch": 8.991452991452991,
+      "grad_norm": 0.4840749262955058,
+      "learning_rate": 1.7317149207281697e-06,
+      "loss": 0.0993,
+      "step": 1052
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.0037508288233707,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.2391,
+      "step": 1053
+    },
+    {
+      "epoch": 9.008547008547009,
+      "grad_norm": 0.5469203775844427,
+      "learning_rate": 1.723202582924624e-06,
+      "loss": 0.1241,
+      "step": 1054
+    },
+    {
+      "epoch": 9.017094017094017,
+      "grad_norm": 0.6115702018377079,
+      "learning_rate": 1.7189501409486061e-06,
+      "loss": 0.0685,
+      "step": 1055
+    },
+    {
+      "epoch": 9.025641025641026,
+      "grad_norm": 0.5787625117983263,
+      "learning_rate": 1.7147002017610626e-06,
+      "loss": 0.1274,
+      "step": 1056
+    },
+    {
+      "epoch": 9.034188034188034,
+      "grad_norm": 0.6270696251364306,
+      "learning_rate": 1.7104527789804554e-06,
+      "loss": 0.0868,
+      "step": 1057
+    },
+    {
+      "epoch": 9.042735042735043,
+      "grad_norm": 0.543963227824456,
+      "learning_rate": 1.7062078862171838e-06,
+      "loss": 0.1412,
+      "step": 1058
+    },
+    {
+      "epoch": 9.051282051282051,
+      "grad_norm": 0.7749740607472595,
+      "learning_rate": 1.7019655370735428e-06,
+      "loss": 0.1233,
+      "step": 1059
+    },
+    {
+      "epoch": 9.05982905982906,
+      "grad_norm": 0.9217914154541708,
+      "learning_rate": 1.6977257451436714e-06,
+      "loss": 0.1321,
+      "step": 1060
+    },
+    {
+      "epoch": 9.068376068376068,
+      "grad_norm": 0.6165308553771175,
+      "learning_rate": 1.6934885240135179e-06,
+      "loss": 0.0547,
+      "step": 1061
+    },
+    {
+      "epoch": 9.076923076923077,
+      "grad_norm": 0.5919180728326213,
+      "learning_rate": 1.6892538872607936e-06,
+      "loss": 0.0612,
+      "step": 1062
+    },
+    {
+      "epoch": 9.085470085470085,
+      "grad_norm": 1.0790808763695527,
+      "learning_rate": 1.6850218484549247e-06,
+      "loss": 0.0709,
+      "step": 1063
+    },
+    {
+      "epoch": 9.094017094017094,
+      "grad_norm": 1.0528333030837587,
+      "learning_rate": 1.6807924211570151e-06,
+      "loss": 0.1405,
+      "step": 1064
+    },
+    {
+      "epoch": 9.102564102564102,
+      "grad_norm": 0.7128081472757782,
+      "learning_rate": 1.6765656189198013e-06,
+      "loss": 0.0618,
+      "step": 1065
+    },
+    {
+      "epoch": 9.11111111111111,
+      "grad_norm": 0.3276318721882367,
+      "learning_rate": 1.6723414552876052e-06,
+      "loss": 0.0148,
+      "step": 1066
+    },
+    {
+      "epoch": 9.11965811965812,
+      "grad_norm": 0.5501475064232987,
+      "learning_rate": 1.6681199437962952e-06,
+      "loss": 0.023,
+      "step": 1067
+    },
+    {
+      "epoch": 9.128205128205128,
+      "grad_norm": 1.0861109374111326,
+      "learning_rate": 1.663901097973243e-06,
+      "loss": 0.2336,
+      "step": 1068
+    },
+    {
+      "epoch": 9.136752136752136,
+      "grad_norm": 1.0153934263214417,
+      "learning_rate": 1.6596849313372737e-06,
+      "loss": 0.0865,
+      "step": 1069
+    },
+    {
+      "epoch": 9.145299145299145,
+      "grad_norm": 0.818922631869137,
+      "learning_rate": 1.6554714573986325e-06,
+      "loss": 0.172,
+      "step": 1070
+    },
+    {
+      "epoch": 9.153846153846153,
+      "grad_norm": 1.0151873298589587,
+      "learning_rate": 1.6512606896589323e-06,
+      "loss": 0.1137,
+      "step": 1071
+    },
+    {
+      "epoch": 9.162393162393162,
+      "grad_norm": 0.8559221976647785,
+      "learning_rate": 1.647052641611117e-06,
+      "loss": 0.3052,
+      "step": 1072
+    },
+    {
+      "epoch": 9.17094017094017,
+      "grad_norm": 0.6823398560394819,
+      "learning_rate": 1.6428473267394135e-06,
+      "loss": 0.0456,
+      "step": 1073
+    },
+    {
+      "epoch": 9.179487179487179,
+      "grad_norm": 0.7232126386293811,
+      "learning_rate": 1.6386447585192911e-06,
+      "loss": 0.0662,
+      "step": 1074
+    },
+    {
+      "epoch": 9.188034188034187,
+      "grad_norm": 0.8760431001553861,
+      "learning_rate": 1.6344449504174193e-06,
+      "loss": 0.0734,
+      "step": 1075
+    },
+    {
+      "epoch": 9.196581196581196,
+      "grad_norm": 0.7741351713881812,
+      "learning_rate": 1.63024791589162e-06,
+      "loss": 0.096,
+      "step": 1076
+    },
+    {
+      "epoch": 9.205128205128204,
+      "grad_norm": 0.5196258263774313,
+      "learning_rate": 1.6260536683908302e-06,
+      "loss": 0.0829,
+      "step": 1077
+    },
+    {
+      "epoch": 9.213675213675213,
+      "grad_norm": 0.8037020755326458,
+      "learning_rate": 1.621862221355056e-06,
+      "loss": 0.1224,
+      "step": 1078
+    },
+    {
+      "epoch": 9.222222222222221,
+      "grad_norm": 0.7282477385753159,
+      "learning_rate": 1.6176735882153284e-06,
+      "loss": 0.1329,
+      "step": 1079
+    },
+    {
+      "epoch": 9.23076923076923,
+      "grad_norm": 0.7135098215001101,
+      "learning_rate": 1.613487782393661e-06,
+      "loss": 0.1125,
+      "step": 1080
+    },
+    {
+      "epoch": 9.239316239316238,
+      "grad_norm": 0.49568130995063514,
+      "learning_rate": 1.6093048173030108e-06,
+      "loss": 0.0488,
+      "step": 1081
+    },
+    {
+      "epoch": 9.247863247863247,
+      "grad_norm": 0.6156142870138479,
+      "learning_rate": 1.6051247063472286e-06,
+      "loss": 0.1364,
+      "step": 1082
+    },
+    {
+      "epoch": 9.256410256410255,
+      "grad_norm": 0.6682930787471676,
+      "learning_rate": 1.6009474629210202e-06,
+      "loss": 0.0712,
+      "step": 1083
+    },
+    {
+      "epoch": 9.264957264957266,
+      "grad_norm": 0.6640845493688758,
+      "learning_rate": 1.5967731004099057e-06,
+      "loss": 0.0641,
+      "step": 1084
+    },
+    {
+      "epoch": 9.273504273504274,
+      "grad_norm": 0.7636323793775699,
+      "learning_rate": 1.5926016321901688e-06,
+      "loss": 0.0429,
+      "step": 1085
+    },
+    {
+      "epoch": 9.282051282051283,
+      "grad_norm": 0.8877230380963692,
+      "learning_rate": 1.5884330716288215e-06,
+      "loss": 0.0643,
+      "step": 1086
+    },
+    {
+      "epoch": 9.290598290598291,
+      "grad_norm": 0.33476740323934123,
+      "learning_rate": 1.5842674320835598e-06,
+      "loss": 0.0205,
+      "step": 1087
+    },
+    {
+      "epoch": 9.2991452991453,
+      "grad_norm": 0.6572203225317508,
+      "learning_rate": 1.5801047269027164e-06,
+      "loss": 0.0686,
+      "step": 1088
+    },
+    {
+      "epoch": 9.307692307692308,
+      "grad_norm": 0.9194169349218544,
+      "learning_rate": 1.5759449694252226e-06,
+      "loss": 0.0975,
+      "step": 1089
+    },
+    {
+      "epoch": 9.316239316239317,
+      "grad_norm": 0.7860733352724719,
+      "learning_rate": 1.571788172980566e-06,
+      "loss": 0.1256,
+      "step": 1090
+    },
+    {
+      "epoch": 9.324786324786325,
+      "grad_norm": 0.35704020366142286,
+      "learning_rate": 1.567634350888743e-06,
+      "loss": 0.0108,
+      "step": 1091
+    },
+    {
+      "epoch": 9.333333333333334,
+      "grad_norm": 0.5165940959681908,
+      "learning_rate": 1.56348351646022e-06,
+      "loss": 0.0392,
+      "step": 1092
+    },
+    {
+      "epoch": 9.341880341880342,
+      "grad_norm": 1.0167691098950664,
+      "learning_rate": 1.5593356829958906e-06,
+      "loss": 0.1348,
+      "step": 1093
+    },
+    {
+      "epoch": 9.350427350427351,
+      "grad_norm": 0.7949727993522546,
+      "learning_rate": 1.5551908637870316e-06,
+      "loss": 0.0468,
+      "step": 1094
+    },
+    {
+      "epoch": 9.35897435897436,
+      "grad_norm": 0.6917120775734487,
+      "learning_rate": 1.5510490721152594e-06,
+      "loss": 0.1252,
+      "step": 1095
+    },
+    {
+      "epoch": 9.367521367521368,
+      "grad_norm": 0.563045044386508,
+      "learning_rate": 1.5469103212524917e-06,
+      "loss": 0.0396,
+      "step": 1096
+    },
+    {
+      "epoch": 9.376068376068377,
+      "grad_norm": 1.189868604604147,
+      "learning_rate": 1.5427746244609015e-06,
+      "loss": 0.1178,
+      "step": 1097
+    },
+    {
+      "epoch": 9.384615384615385,
+      "grad_norm": 0.8655690423744861,
+      "learning_rate": 1.5386419949928732e-06,
+      "loss": 0.1011,
+      "step": 1098
+    },
+    {
+      "epoch": 9.393162393162394,
+      "grad_norm": 0.27480320961328897,
+      "learning_rate": 1.534512446090965e-06,
+      "loss": 0.0076,
+      "step": 1099
+    },
+    {
+      "epoch": 9.401709401709402,
+      "grad_norm": 0.898643735539036,
+      "learning_rate": 1.5303859909878632e-06,
+      "loss": 0.0732,
+      "step": 1100
+    },
+    {
+      "epoch": 9.41025641025641,
+      "grad_norm": 0.6781316543310858,
+      "learning_rate": 1.5262626429063385e-06,
+      "loss": 0.0258,
+      "step": 1101
+    },
+    {
+      "epoch": 9.418803418803419,
+      "grad_norm": 0.5356639554354824,
+      "learning_rate": 1.5221424150592078e-06,
+      "loss": 0.0224,
+      "step": 1102
+    },
+    {
+      "epoch": 9.427350427350428,
+      "grad_norm": 0.7036579428853127,
+      "learning_rate": 1.518025320649289e-06,
+      "loss": 0.1014,
+      "step": 1103
+    },
+    {
+      "epoch": 9.435897435897436,
+      "grad_norm": 0.8127569945947543,
+      "learning_rate": 1.5139113728693575e-06,
+      "loss": 0.0729,
+      "step": 1104
+    },
+    {
+      "epoch": 9.444444444444445,
+      "grad_norm": 0.4071231607971864,
+      "learning_rate": 1.509800584902108e-06,
+      "loss": 0.0244,
+      "step": 1105
+    },
+    {
+      "epoch": 9.452991452991453,
+      "grad_norm": 0.8364670311624951,
+      "learning_rate": 1.5056929699201095e-06,
+      "loss": 0.0945,
+      "step": 1106
+    },
+    {
+      "epoch": 9.461538461538462,
+      "grad_norm": 0.4954553788829718,
+      "learning_rate": 1.5015885410857617e-06,
+      "loss": 0.0606,
+      "step": 1107
+    },
+    {
+      "epoch": 9.47008547008547,
+      "grad_norm": 0.5658789103410351,
+      "learning_rate": 1.4974873115512561e-06,
+      "loss": 0.0747,
+      "step": 1108
+    },
+    {
+      "epoch": 9.478632478632479,
+      "grad_norm": 0.7649082423627214,
+      "learning_rate": 1.4933892944585331e-06,
+      "loss": 0.0679,
+      "step": 1109
+    },
+    {
+      "epoch": 9.487179487179487,
+      "grad_norm": 0.47749330261738837,
+      "learning_rate": 1.489294502939238e-06,
+      "loss": 0.0497,
+      "step": 1110
+    },
+    {
+      "epoch": 9.495726495726496,
+      "grad_norm": 0.5052962598593574,
+      "learning_rate": 1.4852029501146797e-06,
+      "loss": 0.077,
+      "step": 1111
+    },
+    {
+      "epoch": 9.504273504273504,
+      "grad_norm": 0.8130091469943291,
+      "learning_rate": 1.4811146490957903e-06,
+      "loss": 0.1415,
+      "step": 1112
+    },
+    {
+      "epoch": 9.512820512820513,
+      "grad_norm": 0.8200553661974441,
+      "learning_rate": 1.477029612983082e-06,
+      "loss": 0.0908,
+      "step": 1113
+    },
+    {
+      "epoch": 9.521367521367521,
+      "grad_norm": 0.4874352736293081,
+      "learning_rate": 1.4729478548666027e-06,
+      "loss": 0.0833,
+      "step": 1114
+    },
+    {
+      "epoch": 9.52991452991453,
+      "grad_norm": 1.0044705325481182,
+      "learning_rate": 1.468869387825899e-06,
+      "loss": 0.1289,
+      "step": 1115
+    },
+    {
+      "epoch": 9.538461538461538,
+      "grad_norm": 0.6176728966289926,
+      "learning_rate": 1.4647942249299708e-06,
+      "loss": 0.0646,
+      "step": 1116
+    },
+    {
+      "epoch": 9.547008547008547,
+      "grad_norm": 0.2590378983739352,
+      "learning_rate": 1.4607223792372283e-06,
+      "loss": 0.0066,
+      "step": 1117
+    },
+    {
+      "epoch": 9.555555555555555,
+      "grad_norm": 0.9995222239324348,
+      "learning_rate": 1.4566538637954556e-06,
+      "loss": 0.1536,
+      "step": 1118
+    },
+    {
+      "epoch": 9.564102564102564,
+      "grad_norm": 1.0559167819192246,
+      "learning_rate": 1.452588691641763e-06,
+      "loss": 0.112,
+      "step": 1119
+    },
+    {
+      "epoch": 9.572649572649572,
+      "grad_norm": 1.040684684533516,
+      "learning_rate": 1.4485268758025467e-06,
+      "loss": 0.2305,
+      "step": 1120
+    },
+    {
+      "epoch": 9.581196581196581,
+      "grad_norm": 0.813868331480055,
+      "learning_rate": 1.4444684292934508e-06,
+      "loss": 0.0973,
+      "step": 1121
+    },
+    {
+      "epoch": 9.58974358974359,
+      "grad_norm": 0.9200586391191897,
+      "learning_rate": 1.4404133651193214e-06,
+      "loss": 0.0995,
+      "step": 1122
+    },
+    {
+      "epoch": 9.598290598290598,
+      "grad_norm": 0.7501002335879925,
+      "learning_rate": 1.436361696274166e-06,
+      "loss": 0.1735,
+      "step": 1123
+    },
+    {
+      "epoch": 9.606837606837606,
+      "grad_norm": 0.6965668620221329,
+      "learning_rate": 1.4323134357411114e-06,
+      "loss": 0.0925,
+      "step": 1124
+    },
+    {
+      "epoch": 9.615384615384615,
+      "grad_norm": 0.5897020132686239,
+      "learning_rate": 1.4282685964923643e-06,
+      "loss": 0.0364,
+      "step": 1125
+    },
+    {
+      "epoch": 9.623931623931623,
+      "grad_norm": 0.4921094546193598,
+      "learning_rate": 1.4242271914891688e-06,
+      "loss": 0.0612,
+      "step": 1126
+    },
+    {
+      "epoch": 9.632478632478632,
+      "grad_norm": 0.674541085706552,
+      "learning_rate": 1.4201892336817616e-06,
+      "loss": 0.1126,
+      "step": 1127
+    },
+    {
+      "epoch": 9.64102564102564,
+      "grad_norm": 0.9111233476933359,
+      "learning_rate": 1.4161547360093364e-06,
+      "loss": 0.0828,
+      "step": 1128
+    },
+    {
+      "epoch": 9.649572649572649,
+      "grad_norm": 0.7087024813552596,
+      "learning_rate": 1.4121237113999975e-06,
+      "loss": 0.1226,
+      "step": 1129
+    },
+    {
+      "epoch": 9.658119658119658,
+      "grad_norm": 0.6092477040379468,
+      "learning_rate": 1.4080961727707185e-06,
+      "loss": 0.0601,
+      "step": 1130
+    },
+    {
+      "epoch": 9.666666666666666,
+      "grad_norm": 0.6395452684949057,
+      "learning_rate": 1.4040721330273063e-06,
+      "loss": 0.0502,
+      "step": 1131
+    },
+    {
+      "epoch": 9.675213675213675,
+      "grad_norm": 0.5101956373827701,
+      "learning_rate": 1.4000516050643549e-06,
+      "loss": 0.032,
+      "step": 1132
+    },
+    {
+      "epoch": 9.683760683760683,
+      "grad_norm": 0.5625589364023198,
+      "learning_rate": 1.3960346017652027e-06,
+      "loss": 0.0248,
+      "step": 1133
+    },
+    {
+      "epoch": 9.692307692307692,
+      "grad_norm": 0.6634624344092164,
+      "learning_rate": 1.3920211360018971e-06,
+      "loss": 0.0512,
+      "step": 1134
+    },
+    {
+      "epoch": 9.7008547008547,
+      "grad_norm": 0.7224811521885878,
+      "learning_rate": 1.3880112206351476e-06,
+      "loss": 0.0983,
+      "step": 1135
+    },
+    {
+      "epoch": 9.709401709401709,
+      "grad_norm": 0.5523514956679569,
+      "learning_rate": 1.3840048685142863e-06,
+      "loss": 0.1096,
+      "step": 1136
+    },
+    {
+      "epoch": 9.717948717948717,
+      "grad_norm": 0.6629443177230425,
+      "learning_rate": 1.3800020924772295e-06,
+      "loss": 0.0514,
+      "step": 1137
+    },
+    {
+      "epoch": 9.726495726495726,
+      "grad_norm": 0.500052262142771,
+      "learning_rate": 1.3760029053504346e-06,
+      "loss": 0.0756,
+      "step": 1138
+    },
+    {
+      "epoch": 9.735042735042736,
+      "grad_norm": 0.7645112724299366,
+      "learning_rate": 1.372007319948855e-06,
+      "loss": 0.1317,
+      "step": 1139
+    },
+    {
+      "epoch": 9.743589743589745,
+      "grad_norm": 0.4826084146079522,
+      "learning_rate": 1.3680153490759074e-06,
+      "loss": 0.027,
+      "step": 1140
+    },
+    {
+      "epoch": 9.752136752136753,
+      "grad_norm": 0.6166594006265274,
+      "learning_rate": 1.3640270055234227e-06,
+      "loss": 0.1439,
+      "step": 1141
+    },
+    {
+      "epoch": 9.760683760683762,
+      "grad_norm": 0.912758359405943,
+      "learning_rate": 1.360042302071609e-06,
+      "loss": 0.0884,
+      "step": 1142
+    },
+    {
+      "epoch": 9.76923076923077,
+      "grad_norm": 0.8898677341687786,
+      "learning_rate": 1.356061251489012e-06,
+      "loss": 0.1925,
+      "step": 1143
+    },
+    {
+      "epoch": 9.777777777777779,
+      "grad_norm": 0.794175190514908,
+      "learning_rate": 1.3520838665324704e-06,
+      "loss": 0.067,
+      "step": 1144
+    },
+    {
+      "epoch": 9.786324786324787,
+      "grad_norm": 0.5056181154976642,
+      "learning_rate": 1.3481101599470794e-06,
+      "loss": 0.0584,
+      "step": 1145
+    },
+    {
+      "epoch": 9.794871794871796,
+      "grad_norm": 0.7530316981449607,
+      "learning_rate": 1.3441401444661416e-06,
+      "loss": 0.1078,
+      "step": 1146
+    },
+    {
+      "epoch": 9.803418803418804,
+      "grad_norm": 0.46970294469799134,
+      "learning_rate": 1.3401738328111374e-06,
+      "loss": 0.0502,
+      "step": 1147
+    },
+    {
+      "epoch": 9.811965811965813,
+      "grad_norm": 0.4670561098785269,
+      "learning_rate": 1.336211237691678e-06,
+      "loss": 0.08,
+      "step": 1148
+    },
+    {
+      "epoch": 9.820512820512821,
+      "grad_norm": 0.9544088217636255,
+      "learning_rate": 1.3322523718054615e-06,
+      "loss": 0.0836,
+      "step": 1149
+    },
+    {
+      "epoch": 9.82905982905983,
+      "grad_norm": 0.9109700464158846,
+      "learning_rate": 1.328297247838241e-06,
+      "loss": 0.0952,
+      "step": 1150
+    },
+    {
+      "epoch": 9.837606837606838,
+      "grad_norm": 0.7733074193148459,
+      "learning_rate": 1.3243458784637763e-06,
+      "loss": 0.144,
+      "step": 1151
+    },
+    {
+      "epoch": 9.846153846153847,
+      "grad_norm": 0.7022197374436578,
+      "learning_rate": 1.320398276343795e-06,
+      "loss": 0.134,
+      "step": 1152
+    },
+    {
+      "epoch": 9.854700854700855,
+      "grad_norm": 0.9098047058181238,
+      "learning_rate": 1.3164544541279555e-06,
+      "loss": 0.1418,
+      "step": 1153
+    },
+    {
+      "epoch": 9.863247863247864,
+      "grad_norm": 0.532927177425754,
+      "learning_rate": 1.3125144244538038e-06,
+      "loss": 0.0912,
+      "step": 1154
+    },
+    {
+      "epoch": 9.871794871794872,
+      "grad_norm": 0.5578707839247825,
+      "learning_rate": 1.3085781999467303e-06,
+      "loss": 0.0474,
+      "step": 1155
+    },
+    {
+      "epoch": 9.88034188034188,
+      "grad_norm": 1.0787188197476811,
+      "learning_rate": 1.304645793219936e-06,
+      "loss": 0.1303,
+      "step": 1156
+    },
+    {
+      "epoch": 9.88888888888889,
+      "grad_norm": 0.5720067454545272,
+      "learning_rate": 1.3007172168743854e-06,
+      "loss": 0.0295,
+      "step": 1157
+    },
+    {
+      "epoch": 9.897435897435898,
+      "grad_norm": 0.6813484673780763,
+      "learning_rate": 1.2967924834987687e-06,
+      "loss": 0.134,
+      "step": 1158
+    },
+    {
+      "epoch": 9.905982905982906,
+      "grad_norm": 0.4114842099127276,
+      "learning_rate": 1.2928716056694637e-06,
+      "loss": 0.0094,
+      "step": 1159
+    },
+    {
+      "epoch": 9.914529914529915,
+      "grad_norm": 0.8382263936677936,
+      "learning_rate": 1.288954595950494e-06,
+      "loss": 0.1134,
+      "step": 1160
+    },
+    {
+      "epoch": 9.923076923076923,
+      "grad_norm": 1.1972460246350214,
+      "learning_rate": 1.285041466893485e-06,
+      "loss": 0.0304,
+      "step": 1161
+    },
+    {
+      "epoch": 9.931623931623932,
+      "grad_norm": 0.9141699196867842,
+      "learning_rate": 1.2811322310376303e-06,
+      "loss": 0.1376,
+      "step": 1162
+    },
+    {
+      "epoch": 9.94017094017094,
+      "grad_norm": 0.9960126860177222,
+      "learning_rate": 1.2772269009096456e-06,
+      "loss": 0.1626,
+      "step": 1163
+    },
+    {
+      "epoch": 9.948717948717949,
+      "grad_norm": 0.5481120652670447,
+      "learning_rate": 1.2733254890237334e-06,
+      "loss": 0.0397,
+      "step": 1164
+    },
+    {
+      "epoch": 9.957264957264957,
+      "grad_norm": 0.7753907538829564,
+      "learning_rate": 1.2694280078815382e-06,
+      "loss": 0.0258,
+      "step": 1165
+    },
+    {
+      "epoch": 9.965811965811966,
+      "grad_norm": 0.3447580989241403,
+      "learning_rate": 1.2655344699721111e-06,
+      "loss": 0.03,
+      "step": 1166
+    },
+    {
+      "epoch": 9.974358974358974,
+      "grad_norm": 0.5310755009164717,
+      "learning_rate": 1.2616448877718672e-06,
+      "loss": 0.0356,
+      "step": 1167
+    },
+    {
+      "epoch": 9.982905982905983,
+      "grad_norm": 1.1489485415815224,
+      "learning_rate": 1.257759273744545e-06,
+      "loss": 0.1356,
+      "step": 1168
+    },
+    {
+      "epoch": 9.991452991452991,
+      "grad_norm": 0.8228009876608517,
+      "learning_rate": 1.253877640341166e-06,
+      "loss": 0.143,
+      "step": 1169
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.4149657424465853,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.0409,
+      "step": 1170
+    },
+    {
+      "epoch": 10.008547008547009,
+      "grad_norm": 0.7528373835054892,
+      "learning_rate": 1.2461263651465194e-06,
+      "loss": 0.072,
+      "step": 1171
+    },
+    {
+      "epoch": 10.017094017094017,
+      "grad_norm": 0.5693118765989197,
+      "learning_rate": 1.2422567481933604e-06,
+      "loss": 0.0482,
+      "step": 1172
+    },
+    {
+      "epoch": 10.025641025641026,
+      "grad_norm": 0.8967770899898463,
+      "learning_rate": 1.2383911615402873e-06,
+      "loss": 0.0923,
+      "step": 1173
+    },
+    {
+      "epoch": 10.034188034188034,
+      "grad_norm": 0.8133267721334558,
+      "learning_rate": 1.2345296175741466e-06,
+      "loss": 0.119,
+      "step": 1174
+    },
+    {
+      "epoch": 10.042735042735043,
+      "grad_norm": 0.27346171070876685,
+      "learning_rate": 1.2306721286688312e-06,
+      "loss": 0.0074,
+      "step": 1175
+    },
+    {
+      "epoch": 10.051282051282051,
+      "grad_norm": 0.47296973060426656,
+      "learning_rate": 1.226818707185242e-06,
+      "loss": 0.0555,
+      "step": 1176
+    },
+    {
+      "epoch": 10.05982905982906,
+      "grad_norm": 0.7330167223259335,
+      "learning_rate": 1.2229693654712433e-06,
+      "loss": 0.129,
+      "step": 1177
+    },
+    {
+      "epoch": 10.068376068376068,
+      "grad_norm": 0.5383813730777611,
+      "learning_rate": 1.2191241158616284e-06,
+      "loss": 0.0502,
+      "step": 1178
+    },
+    {
+      "epoch": 10.076923076923077,
+      "grad_norm": 0.451721049722562,
+      "learning_rate": 1.2152829706780786e-06,
+      "loss": 0.0178,
+      "step": 1179
+    },
+    {
+      "epoch": 10.085470085470085,
+      "grad_norm": 0.8743661988137066,
+      "learning_rate": 1.2114459422291205e-06,
+      "loss": 0.0818,
+      "step": 1180
+    },
+    {
+      "epoch": 10.094017094017094,
+      "grad_norm": 0.4089134604414421,
+      "learning_rate": 1.2076130428100894e-06,
+      "loss": 0.0176,
+      "step": 1181
+    },
+    {
+      "epoch": 10.102564102564102,
+      "grad_norm": 0.6019224728360073,
+      "learning_rate": 1.203784284703091e-06,
+      "loss": 0.1036,
+      "step": 1182
+    },
+    {
+      "epoch": 10.11111111111111,
+      "grad_norm": 0.6988787974863384,
+      "learning_rate": 1.1999596801769617e-06,
+      "loss": 0.0825,
+      "step": 1183
+    },
+    {
+      "epoch": 10.11965811965812,
+      "grad_norm": 0.3663857880245144,
+      "learning_rate": 1.196139241487225e-06,
+      "loss": 0.0363,
+      "step": 1184
+    },
+    {
+      "epoch": 10.128205128205128,
+      "grad_norm": 0.6280681792226723,
+      "learning_rate": 1.1923229808760565e-06,
+      "loss": 0.0776,
+      "step": 1185
+    },
+    {
+      "epoch": 10.136752136752136,
+      "grad_norm": 0.39634122729164556,
+      "learning_rate": 1.1885109105722454e-06,
+      "loss": 0.0116,
+      "step": 1186
+    },
+    {
+      "epoch": 10.145299145299145,
+      "grad_norm": 0.5081856806271968,
+      "learning_rate": 1.184703042791151e-06,
+      "loss": 0.0541,
+      "step": 1187
+    },
+    {
+      "epoch": 10.153846153846153,
+      "grad_norm": 0.5514164069512796,
+      "learning_rate": 1.1808993897346679e-06,
+      "loss": 0.0355,
+      "step": 1188
+    },
+    {
+      "epoch": 10.162393162393162,
+      "grad_norm": 0.3944092444525037,
+      "learning_rate": 1.1770999635911857e-06,
+      "loss": 0.0263,
+      "step": 1189
+    },
+    {
+      "epoch": 10.17094017094017,
+      "grad_norm": 0.4668680631140163,
+      "learning_rate": 1.1733047765355466e-06,
+      "loss": 0.0283,
+      "step": 1190
+    },
+    {
+      "epoch": 10.179487179487179,
+      "grad_norm": 0.48820262814502324,
+      "learning_rate": 1.1695138407290101e-06,
+      "loss": 0.0519,
+      "step": 1191
+    },
+    {
+      "epoch": 10.188034188034187,
+      "grad_norm": 0.8782440276516138,
+      "learning_rate": 1.1657271683192156e-06,
+      "loss": 0.1342,
+      "step": 1192
+    },
+    {
+      "epoch": 10.196581196581196,
+      "grad_norm": 0.6483304893287919,
+      "learning_rate": 1.1619447714401367e-06,
+      "loss": 0.0909,
+      "step": 1193
+    },
+    {
+      "epoch": 10.205128205128204,
+      "grad_norm": 0.9154554603400102,
+      "learning_rate": 1.1581666622120494e-06,
+      "loss": 0.1041,
+      "step": 1194
+    },
+    {
+      "epoch": 10.213675213675213,
+      "grad_norm": 0.8140879592063411,
+      "learning_rate": 1.154392852741491e-06,
+      "loss": 0.1302,
+      "step": 1195
+    },
+    {
+      "epoch": 10.222222222222221,
+      "grad_norm": 0.632195949813156,
+      "learning_rate": 1.1506233551212186e-06,
+      "loss": 0.0345,
+      "step": 1196
+    },
+    {
+      "epoch": 10.23076923076923,
+      "grad_norm": 0.6331347366030665,
+      "learning_rate": 1.1468581814301718e-06,
+      "loss": 0.0605,
+      "step": 1197
+    },
+    {
+      "epoch": 10.239316239316238,
+      "grad_norm": 0.5852831167921697,
+      "learning_rate": 1.1430973437334375e-06,
+      "loss": 0.0595,
+      "step": 1198
+    },
+    {
+      "epoch": 10.247863247863247,
+      "grad_norm": 0.5502452330425632,
+      "learning_rate": 1.1393408540822073e-06,
+      "loss": 0.0947,
+      "step": 1199
+    },
+    {
+      "epoch": 10.256410256410255,
+      "grad_norm": 0.6661243529811738,
+      "learning_rate": 1.1355887245137383e-06,
+      "loss": 0.1082,
+      "step": 1200
+    },
+    {
+      "epoch": 10.264957264957266,
+      "grad_norm": 0.692366618170434,
+      "learning_rate": 1.1318409670513194e-06,
+      "loss": 0.0552,
+      "step": 1201
+    },
+    {
+      "epoch": 10.273504273504274,
+      "grad_norm": 0.41183248684256724,
+      "learning_rate": 1.1280975937042263e-06,
+      "loss": 0.0138,
+      "step": 1202
+    },
+    {
+      "epoch": 10.282051282051283,
+      "grad_norm": 0.3922345475410547,
+      "learning_rate": 1.1243586164676873e-06,
+      "loss": 0.0332,
+      "step": 1203
+    },
+    {
+      "epoch": 10.290598290598291,
+      "grad_norm": 0.4986802355158306,
+      "learning_rate": 1.1206240473228447e-06,
+      "loss": 0.0464,
+      "step": 1204
+    },
+    {
+      "epoch": 10.2991452991453,
+      "grad_norm": 0.5761580848948239,
+      "learning_rate": 1.1168938982367162e-06,
+      "loss": 0.1282,
+      "step": 1205
+    },
+    {
+      "epoch": 10.307692307692308,
+      "grad_norm": 0.48963370949226304,
+      "learning_rate": 1.1131681811621529e-06,
+      "loss": 0.0494,
+      "step": 1206
+    },
+    {
+      "epoch": 10.316239316239317,
+      "grad_norm": 0.30709075253427076,
+      "learning_rate": 1.1094469080378076e-06,
+      "loss": 0.0372,
+      "step": 1207
+    },
+    {
+      "epoch": 10.324786324786325,
+      "grad_norm": 0.363856094256232,
+      "learning_rate": 1.1057300907880904e-06,
+      "loss": 0.0152,
+      "step": 1208
+    },
+    {
+      "epoch": 10.333333333333334,
+      "grad_norm": 0.3184443824187863,
+      "learning_rate": 1.1020177413231334e-06,
+      "loss": 0.0366,
+      "step": 1209
+    },
+    {
+      "epoch": 10.341880341880342,
+      "grad_norm": 0.2350464111005662,
+      "learning_rate": 1.0983098715387528e-06,
+      "loss": 0.0048,
+      "step": 1210
+    },
+    {
+      "epoch": 10.350427350427351,
+      "grad_norm": 1.2205776925052512,
+      "learning_rate": 1.0946064933164117e-06,
+      "loss": 0.0724,
+      "step": 1211
+    },
+    {
+      "epoch": 10.35897435897436,
+      "grad_norm": 0.7312356567575174,
+      "learning_rate": 1.0909076185231762e-06,
+      "loss": 0.1381,
+      "step": 1212
+    },
+    {
+      "epoch": 10.367521367521368,
+      "grad_norm": 0.5863298364245818,
+      "learning_rate": 1.0872132590116866e-06,
+      "loss": 0.052,
+      "step": 1213
+    },
+    {
+      "epoch": 10.376068376068377,
+      "grad_norm": 0.3717821624930494,
+      "learning_rate": 1.0835234266201109e-06,
+      "loss": 0.0242,
+      "step": 1214
+    },
+    {
+      "epoch": 10.384615384615385,
+      "grad_norm": 0.1928478705969336,
+      "learning_rate": 1.079838133172111e-06,
+      "loss": 0.0049,
+      "step": 1215
+    },
+    {
+      "epoch": 10.393162393162394,
+      "grad_norm": 0.32877528239235754,
+      "learning_rate": 1.0761573904768054e-06,
+      "loss": 0.0143,
+      "step": 1216
+    },
+    {
+      "epoch": 10.401709401709402,
+      "grad_norm": 0.702253447542617,
+      "learning_rate": 1.0724812103287304e-06,
+      "loss": 0.0888,
+      "step": 1217
+    },
+    {
+      "epoch": 10.41025641025641,
+      "grad_norm": 0.4895957665769553,
+      "learning_rate": 1.0688096045078023e-06,
+      "loss": 0.027,
+      "step": 1218
+    },
+    {
+      "epoch": 10.418803418803419,
+      "grad_norm": 0.5389561115166098,
+      "learning_rate": 1.0651425847792767e-06,
+      "loss": 0.042,
+      "step": 1219
+    },
+    {
+      "epoch": 10.427350427350428,
+      "grad_norm": 0.32986659917452915,
+      "learning_rate": 1.061480162893716e-06,
+      "loss": 0.0118,
+      "step": 1220
+    },
+    {
+      "epoch": 10.435897435897436,
+      "grad_norm": 0.5670031788128372,
+      "learning_rate": 1.0578223505869494e-06,
+      "loss": 0.0561,
+      "step": 1221
+    },
+    {
+      "epoch": 10.444444444444445,
+      "grad_norm": 0.9213400434155533,
+      "learning_rate": 1.0541691595800338e-06,
+      "loss": 0.1897,
+      "step": 1222
+    },
+    {
+      "epoch": 10.452991452991453,
+      "grad_norm": 0.7140637315596772,
+      "learning_rate": 1.0505206015792194e-06,
+      "loss": 0.1617,
+      "step": 1223
+    },
+    {
+      "epoch": 10.461538461538462,
+      "grad_norm": 0.6165789747562992,
+      "learning_rate": 1.0468766882759094e-06,
+      "loss": 0.1076,
+      "step": 1224
+    },
+    {
+      "epoch": 10.47008547008547,
+      "grad_norm": 0.47363819608468016,
+      "learning_rate": 1.043237431346622e-06,
+      "loss": 0.0485,
+      "step": 1225
+    },
+    {
+      "epoch": 10.478632478632479,
+      "grad_norm": 0.34733303744291694,
+      "learning_rate": 1.0396028424529578e-06,
+      "loss": 0.014,
+      "step": 1226
+    },
+    {
+      "epoch": 10.487179487179487,
+      "grad_norm": 0.5114382942206728,
+      "learning_rate": 1.0359729332415582e-06,
+      "loss": 0.0347,
+      "step": 1227
+    },
+    {
+      "epoch": 10.495726495726496,
+      "grad_norm": 1.1524497729712029,
+      "learning_rate": 1.032347715344067e-06,
+      "loss": 0.1512,
+      "step": 1228
+    },
+    {
+      "epoch": 10.504273504273504,
+      "grad_norm": 0.39107233323596885,
+      "learning_rate": 1.0287272003770982e-06,
+      "loss": 0.0251,
+      "step": 1229
+    },
+    {
+      "epoch": 10.512820512820513,
+      "grad_norm": 0.5521904639634848,
+      "learning_rate": 1.0251113999421936e-06,
+      "loss": 0.0487,
+      "step": 1230
+    },
+    {
+      "epoch": 10.521367521367521,
+      "grad_norm": 0.865100144503052,
+      "learning_rate": 1.0215003256257874e-06,
+      "loss": 0.031,
+      "step": 1231
+    },
+    {
+      "epoch": 10.52991452991453,
+      "grad_norm": 0.7486109733499217,
+      "learning_rate": 1.0178939889991717e-06,
+      "loss": 0.0586,
+      "step": 1232
+    },
+    {
+      "epoch": 10.538461538461538,
+      "grad_norm": 0.5210171135789035,
+      "learning_rate": 1.014292401618457e-06,
+      "loss": 0.0453,
+      "step": 1233
+    },
+    {
+      "epoch": 10.547008547008547,
+      "grad_norm": 0.6957074639868348,
+      "learning_rate": 1.0106955750245323e-06,
+      "loss": 0.0592,
+      "step": 1234
+    },
+    {
+      "epoch": 10.555555555555555,
+      "grad_norm": 0.46970468661795917,
+      "learning_rate": 1.0071035207430352e-06,
+      "loss": 0.0371,
+      "step": 1235
+    },
+    {
+      "epoch": 10.564102564102564,
+      "grad_norm": 0.9700613365583296,
+      "learning_rate": 1.0035162502843073e-06,
+      "loss": 0.1562,
+      "step": 1236
+    },
+    {
+      "epoch": 10.572649572649572,
+      "grad_norm": 0.5923127833018439,
+      "learning_rate": 9.999337751433643e-07,
+      "loss": 0.0958,
+      "step": 1237
+    },
+    {
+      "epoch": 10.581196581196581,
+      "grad_norm": 0.7836272236314462,
+      "learning_rate": 9.963561067998531e-07,
+      "loss": 0.1466,
+      "step": 1238
+    },
+    {
+      "epoch": 10.58974358974359,
+      "grad_norm": 0.5903667674523515,
+      "learning_rate": 9.927832567180193e-07,
+      "loss": 0.0754,
+      "step": 1239
+    },
+    {
+      "epoch": 10.598290598290598,
+      "grad_norm": 0.5989987601274536,
+      "learning_rate": 9.892152363466692e-07,
+      "loss": 0.083,
+      "step": 1240
+    },
+    {
+      "epoch": 10.606837606837606,
+      "grad_norm": 0.7580993772394548,
+      "learning_rate": 9.856520571191316e-07,
+      "loss": 0.0482,
+      "step": 1241
+    },
+    {
+      "epoch": 10.615384615384615,
+      "grad_norm": 0.6157661057336996,
+      "learning_rate": 9.820937304532221e-07,
+      "loss": 0.0659,
+      "step": 1242
+    },
+    {
+      "epoch": 10.623931623931623,
+      "grad_norm": 0.3535321781443096,
+      "learning_rate": 9.78540267751209e-07,
+      "loss": 0.0128,
+      "step": 1243
+    },
+    {
+      "epoch": 10.632478632478632,
+      "grad_norm": 0.5739629799831032,
+      "learning_rate": 9.749916803997717e-07,
+      "loss": 0.1112,
+      "step": 1244
+    },
+    {
+      "epoch": 10.64102564102564,
+      "grad_norm": 0.5198724203896348,
+      "learning_rate": 9.714479797699695e-07,
+      "loss": 0.0846,
+      "step": 1245
+    },
+    {
+      "epoch": 10.649572649572649,
+      "grad_norm": 1.0956938439727597,
+      "learning_rate": 9.679091772172021e-07,
+      "loss": 0.0334,
+      "step": 1246
+    },
+    {
+      "epoch": 10.658119658119658,
+      "grad_norm": 0.6896219245087443,
+      "learning_rate": 9.643752840811734e-07,
+      "loss": 0.1022,
+      "step": 1247
+    },
+    {
+      "epoch": 10.666666666666666,
+      "grad_norm": 0.7358048567920182,
+      "learning_rate": 9.608463116858544e-07,
+      "loss": 0.1266,
+      "step": 1248
+    },
+    {
+      "epoch": 10.675213675213675,
+      "grad_norm": 0.6241086544923451,
+      "learning_rate": 9.573222713394513e-07,
+      "loss": 0.0542,
+      "step": 1249
+    },
+    {
+      "epoch": 10.683760683760683,
+      "grad_norm": 0.7913785676217334,
+      "learning_rate": 9.538031743343628e-07,
+      "loss": 0.2033,
+      "step": 1250
+    },
+    {
+      "epoch": 10.692307692307692,
+      "grad_norm": 0.6768475684484199,
+      "learning_rate": 9.502890319471491e-07,
+      "loss": 0.1312,
+      "step": 1251
+    },
+    {
+      "epoch": 10.7008547008547,
+      "grad_norm": 0.369422414370594,
+      "learning_rate": 9.467798554384946e-07,
+      "loss": 0.0389,
+      "step": 1252
+    },
+    {
+      "epoch": 10.709401709401709,
+      "grad_norm": 0.5810099654433328,
+      "learning_rate": 9.432756560531691e-07,
+      "loss": 0.1071,
+      "step": 1253
+    },
+    {
+      "epoch": 10.717948717948717,
+      "grad_norm": 0.41263470769639693,
+      "learning_rate": 9.397764450199937e-07,
+      "loss": 0.0129,
+      "step": 1254
+    },
+    {
+      "epoch": 10.726495726495726,
+      "grad_norm": 0.7689917693636449,
+      "learning_rate": 9.362822335518062e-07,
+      "loss": 0.1342,
+      "step": 1255
+    },
+    {
+      "epoch": 10.735042735042736,
+      "grad_norm": 0.3029264104977626,
+      "learning_rate": 9.327930328454249e-07,
+      "loss": 0.0098,
+      "step": 1256
+    },
+    {
+      "epoch": 10.743589743589745,
+      "grad_norm": 0.5769986168100306,
+      "learning_rate": 9.293088540816081e-07,
+      "loss": 0.0681,
+      "step": 1257
+    },
+    {
+      "epoch": 10.752136752136753,
+      "grad_norm": 0.6153157026662167,
+      "learning_rate": 9.258297084250256e-07,
+      "loss": 0.1117,
+      "step": 1258
+    },
+    {
+      "epoch": 10.760683760683762,
+      "grad_norm": 0.5001175270228171,
+      "learning_rate": 9.22355607024217e-07,
+      "loss": 0.0647,
+      "step": 1259
+    },
+    {
+      "epoch": 10.76923076923077,
+      "grad_norm": 0.41031875179006044,
+      "learning_rate": 9.188865610115572e-07,
+      "loss": 0.049,
+      "step": 1260
+    },
+    {
+      "epoch": 10.777777777777779,
+      "grad_norm": 0.6200771849234717,
+      "learning_rate": 9.154225815032242e-07,
+      "loss": 0.0605,
+      "step": 1261
+    },
+    {
+      "epoch": 10.786324786324787,
+      "grad_norm": 0.47359609380322704,
+      "learning_rate": 9.119636795991605e-07,
+      "loss": 0.0341,
+      "step": 1262
+    },
+    {
+      "epoch": 10.794871794871796,
+      "grad_norm": 0.5788256097687582,
+      "learning_rate": 9.085098663830366e-07,
+      "loss": 0.0482,
+      "step": 1263
+    },
+    {
+      "epoch": 10.803418803418804,
+      "grad_norm": 0.5159569439778875,
+      "learning_rate": 9.050611529222167e-07,
+      "loss": 0.0552,
+      "step": 1264
+    },
+    {
+      "epoch": 10.811965811965813,
+      "grad_norm": 0.7146916985301787,
+      "learning_rate": 9.01617550267726e-07,
+      "loss": 0.108,
+      "step": 1265
+    },
+    {
+      "epoch": 10.820512820512821,
+      "grad_norm": 0.4995849685364073,
+      "learning_rate": 8.98179069454209e-07,
+      "loss": 0.0261,
+      "step": 1266
+    },
+    {
+      "epoch": 10.82905982905983,
+      "grad_norm": 0.8280784490328196,
+      "learning_rate": 8.947457214999006e-07,
+      "loss": 0.096,
+      "step": 1267
+    },
+    {
+      "epoch": 10.837606837606838,
+      "grad_norm": 0.7122046826322141,
+      "learning_rate": 8.91317517406588e-07,
+      "loss": 0.0925,
+      "step": 1268
+    },
+    {
+      "epoch": 10.846153846153847,
+      "grad_norm": 0.4857030492009978,
+      "learning_rate": 8.878944681595742e-07,
+      "loss": 0.0481,
+      "step": 1269
+    },
+    {
+      "epoch": 10.854700854700855,
+      "grad_norm": 0.561507137918128,
+      "learning_rate": 8.844765847276432e-07,
+      "loss": 0.0408,
+      "step": 1270
+    },
+    {
+      "epoch": 10.863247863247864,
+      "grad_norm": 0.7758353154061892,
+      "learning_rate": 8.810638780630279e-07,
+      "loss": 0.1096,
+      "step": 1271
+    },
+    {
+      "epoch": 10.871794871794872,
+      "grad_norm": 0.515619620193862,
+      "learning_rate": 8.776563591013729e-07,
+      "loss": 0.005,
+      "step": 1272
+    },
+    {
+      "epoch": 10.88034188034188,
+      "grad_norm": 0.9442077883117976,
+      "learning_rate": 8.742540387616966e-07,
+      "loss": 0.1571,
+      "step": 1273
+    },
+    {
+      "epoch": 10.88888888888889,
+      "grad_norm": 1.0257043051233612,
+      "learning_rate": 8.708569279463622e-07,
+      "loss": 0.1633,
+      "step": 1274
+    },
+    {
+      "epoch": 10.897435897435898,
+      "grad_norm": 0.459285913736046,
+      "learning_rate": 8.674650375410379e-07,
+      "loss": 0.0502,
+      "step": 1275
+    },
+    {
+      "epoch": 10.905982905982906,
+      "grad_norm": 0.9103469998042978,
+      "learning_rate": 8.640783784146625e-07,
+      "loss": 0.0768,
+      "step": 1276
+    },
+    {
+      "epoch": 10.914529914529915,
+      "grad_norm": 0.6507237935820749,
+      "learning_rate": 8.606969614194144e-07,
+      "loss": 0.1222,
+      "step": 1277
+    },
+    {
+      "epoch": 10.923076923076923,
+      "grad_norm": 0.5369364732448973,
+      "learning_rate": 8.573207973906736e-07,
+      "loss": 0.0935,
+      "step": 1278
+    },
+    {
+      "epoch": 10.931623931623932,
+      "grad_norm": 0.4727652663160645,
+      "learning_rate": 8.539498971469848e-07,
+      "loss": 0.0117,
+      "step": 1279
+    },
+    {
+      "epoch": 10.94017094017094,
+      "grad_norm": 0.6326787172262908,
+      "learning_rate": 8.505842714900298e-07,
+      "loss": 0.127,
+      "step": 1280
+    },
+    {
+      "epoch": 10.948717948717949,
+      "grad_norm": 0.5802629230404729,
+      "learning_rate": 8.472239312045851e-07,
+      "loss": 0.0445,
+      "step": 1281
+    },
+    {
+      "epoch": 10.957264957264957,
+      "grad_norm": 0.518789507611408,
+      "learning_rate": 8.438688870584913e-07,
+      "loss": 0.0894,
+      "step": 1282
+    },
+    {
+      "epoch": 10.965811965811966,
+      "grad_norm": 0.736324468031325,
+      "learning_rate": 8.405191498026197e-07,
+      "loss": 0.0655,
+      "step": 1283
+    },
+    {
+      "epoch": 10.974358974358974,
+      "grad_norm": 0.7115113974384445,
+      "learning_rate": 8.371747301708358e-07,
+      "loss": 0.0763,
+      "step": 1284
+    },
+    {
+      "epoch": 10.982905982905983,
+      "grad_norm": 0.5593515387241527,
+      "learning_rate": 8.338356388799637e-07,
+      "loss": 0.0163,
+      "step": 1285
+    },
+    {
+      "epoch": 10.991452991452991,
+      "grad_norm": 0.8641104682123674,
+      "learning_rate": 8.305018866297562e-07,
+      "loss": 0.0364,
+      "step": 1286
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.5402027419043104,
+      "learning_rate": 8.271734841028553e-07,
+      "loss": 0.0477,
+      "step": 1287
+    },
+    {
+      "epoch": 11.008547008547009,
+      "grad_norm": 0.5911273896802919,
+      "learning_rate": 8.238504419647602e-07,
+      "loss": 0.0877,
+      "step": 1288
+    },
+    {
+      "epoch": 11.017094017094017,
+      "grad_norm": 0.33997822462745325,
+      "learning_rate": 8.205327708637958e-07,
+      "loss": 0.0254,
+      "step": 1289
+    },
+    {
+      "epoch": 11.025641025641026,
+      "grad_norm": 0.49353807343817413,
+      "learning_rate": 8.172204814310741e-07,
+      "loss": 0.0769,
+      "step": 1290
+    },
+    {
+      "epoch": 11.034188034188034,
+      "grad_norm": 0.3881327917898042,
+      "learning_rate": 8.139135842804638e-07,
+      "loss": 0.0336,
+      "step": 1291
+    },
+    {
+      "epoch": 11.042735042735043,
+      "grad_norm": 0.691569011124054,
+      "learning_rate": 8.106120900085526e-07,
+      "loss": 0.0267,
+      "step": 1292
+    },
+    {
+      "epoch": 11.051282051282051,
+      "grad_norm": 0.29706237464447505,
+      "learning_rate": 8.073160091946156e-07,
+      "loss": 0.0225,
+      "step": 1293
+    },
+    {
+      "epoch": 11.05982905982906,
+      "grad_norm": 0.5077009563429163,
+      "learning_rate": 8.040253524005834e-07,
+      "loss": 0.0514,
+      "step": 1294
+    },
+    {
+      "epoch": 11.068376068376068,
+      "grad_norm": 0.9891213054149579,
+      "learning_rate": 8.007401301710022e-07,
+      "loss": 0.094,
+      "step": 1295
+    },
+    {
+      "epoch": 11.076923076923077,
+      "grad_norm": 0.41721832260093394,
+      "learning_rate": 7.974603530330069e-07,
+      "loss": 0.0428,
+      "step": 1296
+    },
+    {
+      "epoch": 11.085470085470085,
+      "grad_norm": 0.09346754112243624,
+      "learning_rate": 7.941860314962843e-07,
+      "loss": 0.0027,
+      "step": 1297
+    },
+    {
+      "epoch": 11.094017094017094,
+      "grad_norm": 0.5782118619571033,
+      "learning_rate": 7.909171760530351e-07,
+      "loss": 0.0908,
+      "step": 1298
+    },
+    {
+      "epoch": 11.102564102564102,
+      "grad_norm": 0.37375362031134285,
+      "learning_rate": 7.876537971779493e-07,
+      "loss": 0.0469,
+      "step": 1299
+    },
+    {
+      "epoch": 11.11111111111111,
+      "grad_norm": 0.71782364636737,
+      "learning_rate": 7.843959053281663e-07,
+      "loss": 0.1492,
+      "step": 1300
+    },
+    {
+      "epoch": 11.11965811965812,
+      "grad_norm": 0.6855105794017308,
+      "learning_rate": 7.811435109432417e-07,
+      "loss": 0.0959,
+      "step": 1301
+    },
+    {
+      "epoch": 11.128205128205128,
+      "grad_norm": 0.3250302852365095,
+      "learning_rate": 7.778966244451169e-07,
+      "loss": 0.0226,
+      "step": 1302
+    },
+    {
+      "epoch": 11.136752136752136,
+      "grad_norm": 0.5110437520863196,
+      "learning_rate": 7.746552562380829e-07,
+      "loss": 0.0731,
+      "step": 1303
+    },
+    {
+      "epoch": 11.145299145299145,
+      "grad_norm": 0.4993479038849234,
+      "learning_rate": 7.714194167087466e-07,
+      "loss": 0.0406,
+      "step": 1304
+    },
+    {
+      "epoch": 11.153846153846153,
+      "grad_norm": 0.34456639896835917,
+      "learning_rate": 7.681891162260016e-07,
+      "loss": 0.0337,
+      "step": 1305
+    },
+    {
+      "epoch": 11.162393162393162,
+      "grad_norm": 0.647114798756596,
+      "learning_rate": 7.649643651409916e-07,
+      "loss": 0.1163,
+      "step": 1306
+    },
+    {
+      "epoch": 11.17094017094017,
+      "grad_norm": 0.6597078559115684,
+      "learning_rate": 7.617451737870754e-07,
+      "loss": 0.0987,
+      "step": 1307
+    },
+    {
+      "epoch": 11.179487179487179,
+      "grad_norm": 0.5012062943764185,
+      "learning_rate": 7.585315524797998e-07,
+      "loss": 0.046,
+      "step": 1308
+    },
+    {
+      "epoch": 11.188034188034187,
+      "grad_norm": 0.3789882400899952,
+      "learning_rate": 7.553235115168598e-07,
+      "loss": 0.028,
+      "step": 1309
+    },
+    {
+      "epoch": 11.196581196581196,
+      "grad_norm": 0.18142287558793085,
+      "learning_rate": 7.521210611780715e-07,
+      "loss": 0.0058,
+      "step": 1310
+    },
+    {
+      "epoch": 11.205128205128204,
+      "grad_norm": 0.378803728058699,
+      "learning_rate": 7.489242117253342e-07,
+      "loss": 0.0313,
+      "step": 1311
+    },
+    {
+      "epoch": 11.213675213675213,
+      "grad_norm": 0.4789925234922484,
+      "learning_rate": 7.457329734026012e-07,
+      "loss": 0.072,
+      "step": 1312
+    },
+    {
+      "epoch": 11.222222222222221,
+      "grad_norm": 0.3146472429753484,
+      "learning_rate": 7.425473564358457e-07,
+      "loss": 0.0151,
+      "step": 1313
+    },
+    {
+      "epoch": 11.23076923076923,
+      "grad_norm": 0.34429324177174386,
+      "learning_rate": 7.393673710330271e-07,
+      "loss": 0.0272,
+      "step": 1314
+    },
+    {
+      "epoch": 11.239316239316238,
+      "grad_norm": 0.9134534949643333,
+      "learning_rate": 7.361930273840581e-07,
+      "loss": 0.1517,
+      "step": 1315
+    },
+    {
+      "epoch": 11.247863247863247,
+      "grad_norm": 0.5776250431801766,
+      "learning_rate": 7.330243356607758e-07,
+      "loss": 0.0614,
+      "step": 1316
+    },
+    {
+      "epoch": 11.256410256410255,
+      "grad_norm": 0.6828276002890209,
+      "learning_rate": 7.298613060169035e-07,
+      "loss": 0.1427,
+      "step": 1317
+    },
+    {
+      "epoch": 11.264957264957266,
+      "grad_norm": 0.5870274651894514,
+      "learning_rate": 7.267039485880225e-07,
+      "loss": 0.0464,
+      "step": 1318
+    },
+    {
+      "epoch": 11.273504273504274,
+      "grad_norm": 0.31497786244699366,
+      "learning_rate": 7.235522734915393e-07,
+      "loss": 0.0318,
+      "step": 1319
+    },
+    {
+      "epoch": 11.282051282051283,
+      "grad_norm": 0.38817987373377083,
+      "learning_rate": 7.204062908266491e-07,
+      "loss": 0.0418,
+      "step": 1320
+    },
+    {
+      "epoch": 11.290598290598291,
+      "grad_norm": 0.7395579672928589,
+      "learning_rate": 7.172660106743073e-07,
+      "loss": 0.0726,
+      "step": 1321
+    },
+    {
+      "epoch": 11.2991452991453,
+      "grad_norm": 0.6617468807072606,
+      "learning_rate": 7.14131443097198e-07,
+      "loss": 0.1204,
+      "step": 1322
+    },
+    {
+      "epoch": 11.307692307692308,
+      "grad_norm": 0.39065128812433864,
+      "learning_rate": 7.110025981396976e-07,
+      "loss": 0.0464,
+      "step": 1323
+    },
+    {
+      "epoch": 11.316239316239317,
+      "grad_norm": 0.3353924587905785,
+      "learning_rate": 7.078794858278462e-07,
+      "loss": 0.0195,
+      "step": 1324
+    },
+    {
+      "epoch": 11.324786324786325,
+      "grad_norm": 0.7270883932112006,
+      "learning_rate": 7.047621161693152e-07,
+      "loss": 0.1163,
+      "step": 1325
+    },
+    {
+      "epoch": 11.333333333333334,
+      "grad_norm": 0.2013736973419809,
+      "learning_rate": 7.016504991533727e-07,
+      "loss": 0.0045,
+      "step": 1326
+    },
+    {
+      "epoch": 11.341880341880342,
+      "grad_norm": 0.5377819821360028,
+      "learning_rate": 6.985446447508526e-07,
+      "loss": 0.0629,
+      "step": 1327
+    },
+    {
+      "epoch": 11.350427350427351,
+      "grad_norm": 0.7904997992935608,
+      "learning_rate": 6.954445629141246e-07,
+      "loss": 0.127,
+      "step": 1328
+    },
+    {
+      "epoch": 11.35897435897436,
+      "grad_norm": 0.5448615589144118,
+      "learning_rate": 6.923502635770618e-07,
+      "loss": 0.0704,
+      "step": 1329
+    },
+    {
+      "epoch": 11.367521367521368,
+      "grad_norm": 0.7324833370630733,
+      "learning_rate": 6.892617566550044e-07,
+      "loss": 0.0725,
+      "step": 1330
+    },
+    {
+      "epoch": 11.376068376068377,
+      "grad_norm": 0.4079972569261083,
+      "learning_rate": 6.861790520447356e-07,
+      "loss": 0.0249,
+      "step": 1331
+    },
+    {
+      "epoch": 11.384615384615385,
+      "grad_norm": 0.3360927561523562,
+      "learning_rate": 6.831021596244425e-07,
+      "loss": 0.026,
+      "step": 1332
+    },
+    {
+      "epoch": 11.393162393162394,
+      "grad_norm": 0.4744642260817377,
+      "learning_rate": 6.800310892536884e-07,
+      "loss": 0.0397,
+      "step": 1333
+    },
+    {
+      "epoch": 11.401709401709402,
+      "grad_norm": 0.25327380853084464,
+      "learning_rate": 6.769658507733815e-07,
+      "loss": 0.0114,
+      "step": 1334
+    },
+    {
+      "epoch": 11.41025641025641,
+      "grad_norm": 0.4251209808676598,
+      "learning_rate": 6.739064540057425e-07,
+      "loss": 0.0394,
+      "step": 1335
+    },
+    {
+      "epoch": 11.418803418803419,
+      "grad_norm": 0.5531908728480924,
+      "learning_rate": 6.708529087542717e-07,
+      "loss": 0.0769,
+      "step": 1336
+    },
+    {
+      "epoch": 11.427350427350428,
+      "grad_norm": 0.48252102964953725,
+      "learning_rate": 6.678052248037184e-07,
+      "loss": 0.0842,
+      "step": 1337
+    },
+    {
+      "epoch": 11.435897435897436,
+      "grad_norm": 0.5455617564825324,
+      "learning_rate": 6.64763411920053e-07,
+      "loss": 0.041,
+      "step": 1338
+    },
+    {
+      "epoch": 11.444444444444445,
+      "grad_norm": 0.5918199647503406,
+      "learning_rate": 6.617274798504286e-07,
+      "loss": 0.0656,
+      "step": 1339
+    },
+    {
+      "epoch": 11.452991452991453,
+      "grad_norm": 0.36042607672574983,
+      "learning_rate": 6.586974383231573e-07,
+      "loss": 0.0161,
+      "step": 1340
+    },
+    {
+      "epoch": 11.461538461538462,
+      "grad_norm": 0.5747374588172919,
+      "learning_rate": 6.556732970476748e-07,
+      "loss": 0.0967,
+      "step": 1341
+    },
+    {
+      "epoch": 11.47008547008547,
+      "grad_norm": 0.3626825420456705,
+      "learning_rate": 6.526550657145089e-07,
+      "loss": 0.0325,
+      "step": 1342
+    },
+    {
+      "epoch": 11.478632478632479,
+      "grad_norm": 0.7375881253726265,
+      "learning_rate": 6.496427539952499e-07,
+      "loss": 0.0559,
+      "step": 1343
+    },
+    {
+      "epoch": 11.487179487179487,
+      "grad_norm": 0.4282538497196272,
+      "learning_rate": 6.4663637154252e-07,
+      "loss": 0.0635,
+      "step": 1344
+    },
+    {
+      "epoch": 11.495726495726496,
+      "grad_norm": 0.5249222332292881,
+      "learning_rate": 6.436359279899426e-07,
+      "loss": 0.0575,
+      "step": 1345
+    },
+    {
+      "epoch": 11.504273504273504,
+      "grad_norm": 0.5911125039984814,
+      "learning_rate": 6.406414329521079e-07,
+      "loss": 0.0362,
+      "step": 1346
+    },
+    {
+      "epoch": 11.512820512820513,
+      "grad_norm": 0.6291188878291657,
+      "learning_rate": 6.376528960245476e-07,
+      "loss": 0.0416,
+      "step": 1347
+    },
+    {
+      "epoch": 11.521367521367521,
+      "grad_norm": 1.0438630139472773,
+      "learning_rate": 6.346703267836998e-07,
+      "loss": 0.085,
+      "step": 1348
+    },
+    {
+      "epoch": 11.52991452991453,
+      "grad_norm": 0.3318124571152316,
+      "learning_rate": 6.316937347868787e-07,
+      "loss": 0.0337,
+      "step": 1349
+    },
+    {
+      "epoch": 11.538461538461538,
+      "grad_norm": 0.4881877370841047,
+      "learning_rate": 6.28723129572247e-07,
+      "loss": 0.0265,
+      "step": 1350
+    },
+    {
+      "epoch": 11.547008547008547,
+      "grad_norm": 0.42206065565383477,
+      "learning_rate": 6.257585206587843e-07,
+      "loss": 0.0383,
+      "step": 1351
+    },
+    {
+      "epoch": 11.555555555555555,
+      "grad_norm": 0.7835268070693865,
+      "learning_rate": 6.227999175462521e-07,
+      "loss": 0.0897,
+      "step": 1352
+    },
+    {
+      "epoch": 11.564102564102564,
+      "grad_norm": 0.43435236055674364,
+      "learning_rate": 6.198473297151705e-07,
+      "loss": 0.0424,
+      "step": 1353
+    },
+    {
+      "epoch": 11.572649572649572,
+      "grad_norm": 0.5417439990210136,
+      "learning_rate": 6.169007666267824e-07,
+      "loss": 0.082,
+      "step": 1354
+    },
+    {
+      "epoch": 11.581196581196581,
+      "grad_norm": 0.38045323551448207,
+      "learning_rate": 6.139602377230247e-07,
+      "loss": 0.0378,
+      "step": 1355
+    },
+    {
+      "epoch": 11.58974358974359,
+      "grad_norm": 0.20022917291801898,
+      "learning_rate": 6.110257524264998e-07,
+      "loss": 0.0045,
+      "step": 1356
+    },
+    {
+      "epoch": 11.598290598290598,
+      "grad_norm": 0.8737992978550385,
+      "learning_rate": 6.080973201404444e-07,
+      "loss": 0.1648,
+      "step": 1357
+    },
+    {
+      "epoch": 11.606837606837606,
+      "grad_norm": 0.43375747282423105,
+      "learning_rate": 6.051749502486967e-07,
+      "loss": 0.0603,
+      "step": 1358
+    },
+    {
+      "epoch": 11.615384615384615,
+      "grad_norm": 0.40339600429776185,
+      "learning_rate": 6.022586521156714e-07,
+      "loss": 0.0427,
+      "step": 1359
+    },
+    {
+      "epoch": 11.623931623931623,
+      "grad_norm": 0.32496203424257325,
+      "learning_rate": 5.993484350863246e-07,
+      "loss": 0.0246,
+      "step": 1360
+    },
+    {
+      "epoch": 11.632478632478632,
+      "grad_norm": 0.08515217413894376,
+      "learning_rate": 5.964443084861265e-07,
+      "loss": 0.0018,
+      "step": 1361
+    },
+    {
+      "epoch": 11.64102564102564,
+      "grad_norm": 0.45210961448699233,
+      "learning_rate": 5.935462816210325e-07,
+      "loss": 0.0732,
+      "step": 1362
+    },
+    {
+      "epoch": 11.649572649572649,
+      "grad_norm": 0.6607902661129196,
+      "learning_rate": 5.906543637774512e-07,
+      "loss": 0.1566,
+      "step": 1363
+    },
+    {
+      "epoch": 11.658119658119658,
+      "grad_norm": 0.579241782197535,
+      "learning_rate": 5.877685642222163e-07,
+      "loss": 0.0624,
+      "step": 1364
+    },
+    {
+      "epoch": 11.666666666666666,
+      "grad_norm": 0.6270729600768505,
+      "learning_rate": 5.848888922025553e-07,
+      "loss": 0.0542,
+      "step": 1365
+    },
+    {
+      "epoch": 11.675213675213675,
+      "grad_norm": 0.24964227547767157,
+      "learning_rate": 5.820153569460596e-07,
+      "loss": 0.0082,
+      "step": 1366
+    },
+    {
+      "epoch": 11.683760683760683,
+      "grad_norm": 0.49031923515529713,
+      "learning_rate": 5.791479676606587e-07,
+      "loss": 0.0393,
+      "step": 1367
+    },
+    {
+      "epoch": 11.692307692307692,
+      "grad_norm": 0.6758857364179156,
+      "learning_rate": 5.762867335345851e-07,
+      "loss": 0.0886,
+      "step": 1368
+    },
+    {
+      "epoch": 11.7008547008547,
+      "grad_norm": 0.4810336398940464,
+      "learning_rate": 5.734316637363505e-07,
+      "loss": 0.0568,
+      "step": 1369
+    },
+    {
+      "epoch": 11.709401709401709,
+      "grad_norm": 0.3710878858390517,
+      "learning_rate": 5.705827674147125e-07,
+      "loss": 0.0188,
+      "step": 1370
+    },
+    {
+      "epoch": 11.717948717948717,
+      "grad_norm": 0.5475396711113039,
+      "learning_rate": 5.67740053698646e-07,
+      "loss": 0.0387,
+      "step": 1371
+    },
+    {
+      "epoch": 11.726495726495726,
+      "grad_norm": 0.4567651524798183,
+      "learning_rate": 5.649035316973142e-07,
+      "loss": 0.0337,
+      "step": 1372
+    },
+    {
+      "epoch": 11.735042735042736,
+      "grad_norm": 0.15447178250306332,
+      "learning_rate": 5.620732105000415e-07,
+      "loss": 0.0051,
+      "step": 1373
+    },
+    {
+      "epoch": 11.743589743589745,
+      "grad_norm": 0.5477727950392361,
+      "learning_rate": 5.5924909917628e-07,
+      "loss": 0.0725,
+      "step": 1374
+    },
+    {
+      "epoch": 11.752136752136753,
+      "grad_norm": 0.3721327831035405,
+      "learning_rate": 5.564312067755856e-07,
+      "loss": 0.0257,
+      "step": 1375
+    },
+    {
+      "epoch": 11.760683760683762,
+      "grad_norm": 0.5341727624288521,
+      "learning_rate": 5.536195423275839e-07,
+      "loss": 0.1048,
+      "step": 1376
+    },
+    {
+      "epoch": 11.76923076923077,
+      "grad_norm": 0.31087455820578935,
+      "learning_rate": 5.508141148419443e-07,
+      "loss": 0.0318,
+      "step": 1377
+    },
+    {
+      "epoch": 11.777777777777779,
+      "grad_norm": 0.6628966101866233,
+      "learning_rate": 5.48014933308352e-07,
+      "loss": 0.0887,
+      "step": 1378
+    },
+    {
+      "epoch": 11.786324786324787,
+      "grad_norm": 0.49027928460694736,
+      "learning_rate": 5.45222006696477e-07,
+      "loss": 0.0377,
+      "step": 1379
+    },
+    {
+      "epoch": 11.794871794871796,
+      "grad_norm": 0.5691374119125762,
+      "learning_rate": 5.424353439559446e-07,
+      "loss": 0.0483,
+      "step": 1380
+    },
+    {
+      "epoch": 11.803418803418804,
+      "grad_norm": 0.5610287799733852,
+      "learning_rate": 5.396549540163106e-07,
+      "loss": 0.0316,
+      "step": 1381
+    },
+    {
+      "epoch": 11.811965811965813,
+      "grad_norm": 0.4984471059555396,
+      "learning_rate": 5.36880845787028e-07,
+      "loss": 0.0357,
+      "step": 1382
+    },
+    {
+      "epoch": 11.820512820512821,
+      "grad_norm": 0.6272781584817222,
+      "learning_rate": 5.341130281574233e-07,
+      "loss": 0.0827,
+      "step": 1383
+    },
+    {
+      "epoch": 11.82905982905983,
+      "grad_norm": 0.6115936866920039,
+      "learning_rate": 5.313515099966627e-07,
+      "loss": 0.056,
+      "step": 1384
+    },
+    {
+      "epoch": 11.837606837606838,
+      "grad_norm": 0.4513746929956304,
+      "learning_rate": 5.28596300153728e-07,
+      "loss": 0.0504,
+      "step": 1385
+    },
+    {
+      "epoch": 11.846153846153847,
+      "grad_norm": 0.6712793842418363,
+      "learning_rate": 5.258474074573878e-07,
+      "loss": 0.1243,
+      "step": 1386
+    },
+    {
+      "epoch": 11.854700854700855,
+      "grad_norm": 0.38720868482490745,
+      "learning_rate": 5.231048407161657e-07,
+      "loss": 0.031,
+      "step": 1387
+    },
+    {
+      "epoch": 11.863247863247864,
+      "grad_norm": 0.5695039021760533,
+      "learning_rate": 5.203686087183149e-07,
+      "loss": 0.0785,
+      "step": 1388
+    },
+    {
+      "epoch": 11.871794871794872,
+      "grad_norm": 0.8750851731120838,
+      "learning_rate": 5.176387202317915e-07,
+      "loss": 0.1761,
+      "step": 1389
+    },
+    {
+      "epoch": 11.88034188034188,
+      "grad_norm": 0.6257864639644699,
+      "learning_rate": 5.149151840042224e-07,
+      "loss": 0.125,
+      "step": 1390
+    },
+    {
+      "epoch": 11.88888888888889,
+      "grad_norm": 0.41421763085001945,
+      "learning_rate": 5.121980087628802e-07,
+      "loss": 0.0376,
+      "step": 1391
+    },
+    {
+      "epoch": 11.897435897435898,
+      "grad_norm": 0.5926832943736985,
+      "learning_rate": 5.094872032146562e-07,
+      "loss": 0.0296,
+      "step": 1392
+    },
+    {
+      "epoch": 11.905982905982906,
+      "grad_norm": 0.9497155581880431,
+      "learning_rate": 5.06782776046027e-07,
+      "loss": 0.0237,
+      "step": 1393
+    },
+    {
+      "epoch": 11.914529914529915,
+      "grad_norm": 0.4394719509685079,
+      "learning_rate": 5.040847359230327e-07,
+      "loss": 0.0431,
+      "step": 1394
+    },
+    {
+      "epoch": 11.923076923076923,
+      "grad_norm": 0.5155848596346819,
+      "learning_rate": 5.013930914912477e-07,
+      "loss": 0.0372,
+      "step": 1395
+    },
+    {
+      "epoch": 11.931623931623932,
+      "grad_norm": 0.4328919090524444,
+      "learning_rate": 4.98707851375749e-07,
+      "loss": 0.0421,
+      "step": 1396
+    },
+    {
+      "epoch": 11.94017094017094,
+      "grad_norm": 0.4715230395024146,
+      "learning_rate": 4.96029024181095e-07,
+      "loss": 0.0448,
+      "step": 1397
+    },
+    {
+      "epoch": 11.948717948717949,
+      "grad_norm": 0.35700048562694225,
+      "learning_rate": 4.933566184912931e-07,
+      "loss": 0.023,
+      "step": 1398
+    },
+    {
+      "epoch": 11.957264957264957,
+      "grad_norm": 0.5586094473199622,
+      "learning_rate": 4.906906428697736e-07,
+      "loss": 0.0127,
+      "step": 1399
+    },
+    {
+      "epoch": 11.965811965811966,
+      "grad_norm": 0.7247335039292072,
+      "learning_rate": 4.880311058593617e-07,
+      "loss": 0.0598,
+      "step": 1400
+    },
+    {
+      "epoch": 11.974358974358974,
+      "grad_norm": 0.5470831532818907,
+      "learning_rate": 4.853780159822521e-07,
+      "loss": 0.0436,
+      "step": 1401
+    },
+    {
+      "epoch": 11.982905982905983,
+      "grad_norm": 0.4622223877451198,
+      "learning_rate": 4.827313817399809e-07,
+      "loss": 0.0676,
+      "step": 1402
+    },
+    {
+      "epoch": 11.991452991452991,
+      "grad_norm": 0.4091103436084329,
+      "learning_rate": 4.800912116133955e-07,
+      "loss": 0.0255,
+      "step": 1403
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.7340458945691303,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.0752,
+      "step": 1404
+    },
+    {
+      "epoch": 12.008547008547009,
+      "grad_norm": 0.5210990574780048,
+      "learning_rate": 4.748302975270838e-07,
+      "loss": 0.0693,
+      "step": 1405
+    },
+    {
+      "epoch": 12.017094017094017,
+      "grad_norm": 0.4788263979171196,
+      "learning_rate": 4.7220957042537793e-07,
+      "loss": 0.062,
+      "step": 1406
+    },
+    {
+      "epoch": 12.025641025641026,
+      "grad_norm": 0.36329152290724714,
+      "learning_rate": 4.6959534115534665e-07,
+      "loss": 0.0427,
+      "step": 1407
+    },
+    {
+      "epoch": 12.034188034188034,
+      "grad_norm": 0.34183052198104663,
+      "learning_rate": 4.669876180940014e-07,
+      "loss": 0.0343,
+      "step": 1408
+    },
+    {
+      "epoch": 12.042735042735043,
+      "grad_norm": 0.4294908397852798,
+      "learning_rate": 4.6438640959750285e-07,
+      "loss": 0.03,
+      "step": 1409
+    },
+    {
+      "epoch": 12.051282051282051,
+      "grad_norm": 0.3310588008300683,
+      "learning_rate": 4.617917240011394e-07,
+      "loss": 0.0185,
+      "step": 1410
+    },
+    {
+      "epoch": 12.05982905982906,
+      "grad_norm": 0.26661733443446145,
+      "learning_rate": 4.592035696192948e-07,
+      "loss": 0.0165,
+      "step": 1411
+    },
+    {
+      "epoch": 12.068376068376068,
+      "grad_norm": 0.45245941537813045,
+      "learning_rate": 4.566219547454251e-07,
+      "loss": 0.0546,
+      "step": 1412
+    },
+    {
+      "epoch": 12.076923076923077,
+      "grad_norm": 0.44678544747622206,
+      "learning_rate": 4.5404688765203236e-07,
+      "loss": 0.0422,
+      "step": 1413
+    },
+    {
+      "epoch": 12.085470085470085,
+      "grad_norm": 0.42280662183770285,
+      "learning_rate": 4.514783765906369e-07,
+      "loss": 0.0456,
+      "step": 1414
+    },
+    {
+      "epoch": 12.094017094017094,
+      "grad_norm": 0.630588141911986,
+      "learning_rate": 4.489164297917492e-07,
+      "loss": 0.0725,
+      "step": 1415
+    },
+    {
+      "epoch": 12.102564102564102,
+      "grad_norm": 0.6957962761795572,
+      "learning_rate": 4.463610554648459e-07,
+      "loss": 0.0719,
+      "step": 1416
+    },
+    {
+      "epoch": 12.11111111111111,
+      "grad_norm": 0.2344619909817438,
+      "learning_rate": 4.438122617983442e-07,
+      "loss": 0.0085,
+      "step": 1417
+    },
+    {
+      "epoch": 12.11965811965812,
+      "grad_norm": 0.5297503159148729,
+      "learning_rate": 4.4127005695957374e-07,
+      "loss": 0.0426,
+      "step": 1418
+    },
+    {
+      "epoch": 12.128205128205128,
+      "grad_norm": 0.4180850364931802,
+      "learning_rate": 4.3873444909474985e-07,
+      "loss": 0.0465,
+      "step": 1419
+    },
+    {
+      "epoch": 12.136752136752136,
+      "grad_norm": 0.494857947438109,
+      "learning_rate": 4.3620544632895e-07,
+      "loss": 0.0558,
+      "step": 1420
+    },
+    {
+      "epoch": 12.145299145299145,
+      "grad_norm": 0.5494346854035879,
+      "learning_rate": 4.336830567660855e-07,
+      "loss": 0.072,
+      "step": 1421
+    },
+    {
+      "epoch": 12.153846153846153,
+      "grad_norm": 0.505720729616524,
+      "learning_rate": 4.311672884888757e-07,
+      "loss": 0.0531,
+      "step": 1422
+    },
+    {
+      "epoch": 12.162393162393162,
+      "grad_norm": 0.3882375970304378,
+      "learning_rate": 4.286581495588249e-07,
+      "loss": 0.0254,
+      "step": 1423
+    },
+    {
+      "epoch": 12.17094017094017,
+      "grad_norm": 0.5868112593989185,
+      "learning_rate": 4.2615564801619325e-07,
+      "loss": 0.0882,
+      "step": 1424
+    },
+    {
+      "epoch": 12.179487179487179,
+      "grad_norm": 0.3634578454157606,
+      "learning_rate": 4.2365979187997094e-07,
+      "loss": 0.0161,
+      "step": 1425
+    },
+    {
+      "epoch": 12.188034188034187,
+      "grad_norm": 0.3645718143750022,
+      "learning_rate": 4.21170589147856e-07,
+      "loss": 0.0361,
+      "step": 1426
+    },
+    {
+      "epoch": 12.196581196581196,
+      "grad_norm": 0.21409222209999557,
+      "learning_rate": 4.1868804779622437e-07,
+      "loss": 0.009,
+      "step": 1427
+    },
+    {
+      "epoch": 12.205128205128204,
+      "grad_norm": 0.46176542250006986,
+      "learning_rate": 4.1621217578010686e-07,
+      "loss": 0.0558,
+      "step": 1428
+    },
+    {
+      "epoch": 12.213675213675213,
+      "grad_norm": 0.4530777351923695,
+      "learning_rate": 4.137429810331639e-07,
+      "loss": 0.0356,
+      "step": 1429
+    },
+    {
+      "epoch": 12.222222222222221,
+      "grad_norm": 0.4564673214893363,
+      "learning_rate": 4.1128047146765936e-07,
+      "loss": 0.0738,
+      "step": 1430
+    },
+    {
+      "epoch": 12.23076923076923,
+      "grad_norm": 0.5444892972993393,
+      "learning_rate": 4.088246549744332e-07,
+      "loss": 0.0457,
+      "step": 1431
+    },
+    {
+      "epoch": 12.239316239316238,
+      "grad_norm": 0.46279984642252764,
+      "learning_rate": 4.063755394228811e-07,
+      "loss": 0.0411,
+      "step": 1432
+    },
+    {
+      "epoch": 12.247863247863247,
+      "grad_norm": 0.27693768948677727,
+      "learning_rate": 4.039331326609239e-07,
+      "loss": 0.0232,
+      "step": 1433
+    },
+    {
+      "epoch": 12.256410256410255,
+      "grad_norm": 0.43398332814610746,
+      "learning_rate": 4.0149744251498537e-07,
+      "loss": 0.0338,
+      "step": 1434
+    },
+    {
+      "epoch": 12.264957264957266,
+      "grad_norm": 0.7156656613045878,
+      "learning_rate": 3.990684767899677e-07,
+      "loss": 0.0649,
+      "step": 1435
+    },
+    {
+      "epoch": 12.273504273504274,
+      "grad_norm": 1.6223644181519623,
+      "learning_rate": 3.9664624326922447e-07,
+      "loss": 0.0726,
+      "step": 1436
+    },
+    {
+      "epoch": 12.282051282051283,
+      "grad_norm": 0.5691517529033704,
+      "learning_rate": 3.9423074971453785e-07,
+      "loss": 0.1054,
+      "step": 1437
+    },
+    {
+      "epoch": 12.290598290598291,
+      "grad_norm": 0.25125116996613145,
+      "learning_rate": 3.918220038660908e-07,
+      "loss": 0.0081,
+      "step": 1438
+    },
+    {
+      "epoch": 12.2991452991453,
+      "grad_norm": 0.5363768285869193,
+      "learning_rate": 3.8942001344244416e-07,
+      "loss": 0.1076,
+      "step": 1439
+    },
+    {
+      "epoch": 12.307692307692308,
+      "grad_norm": 0.3908348952351424,
+      "learning_rate": 3.8702478614051353e-07,
+      "loss": 0.0472,
+      "step": 1440
+    },
+    {
+      "epoch": 12.316239316239317,
+      "grad_norm": 0.4980122155106956,
+      "learning_rate": 3.846363296355404e-07,
+      "loss": 0.0684,
+      "step": 1441
+    },
+    {
+      "epoch": 12.324786324786325,
+      "grad_norm": 0.7839866964738976,
+      "learning_rate": 3.822546515810724e-07,
+      "loss": 0.136,
+      "step": 1442
+    },
+    {
+      "epoch": 12.333333333333334,
+      "grad_norm": 0.2920246859904782,
+      "learning_rate": 3.798797596089351e-07,
+      "loss": 0.0154,
+      "step": 1443
+    },
+    {
+      "epoch": 12.341880341880342,
+      "grad_norm": 0.34194465726755663,
+      "learning_rate": 3.7751166132920877e-07,
+      "loss": 0.0424,
+      "step": 1444
+    },
+    {
+      "epoch": 12.350427350427351,
+      "grad_norm": 0.27774192769005757,
+      "learning_rate": 3.751503643302035e-07,
+      "loss": 0.0172,
+      "step": 1445
+    },
+    {
+      "epoch": 12.35897435897436,
+      "grad_norm": 0.4910903215917028,
+      "learning_rate": 3.727958761784375e-07,
+      "loss": 0.0594,
+      "step": 1446
+    },
+    {
+      "epoch": 12.367521367521368,
+      "grad_norm": 0.680826297771005,
+      "learning_rate": 3.7044820441860806e-07,
+      "loss": 0.1152,
+      "step": 1447
+    },
+    {
+      "epoch": 12.376068376068377,
+      "grad_norm": 0.6415763390791678,
+      "learning_rate": 3.681073565735718e-07,
+      "loss": 0.0776,
+      "step": 1448
+    },
+    {
+      "epoch": 12.384615384615385,
+      "grad_norm": 0.5833589699943083,
+      "learning_rate": 3.6577334014432003e-07,
+      "loss": 0.1001,
+      "step": 1449
+    },
+    {
+      "epoch": 12.393162393162394,
+      "grad_norm": 0.5599742675848512,
+      "learning_rate": 3.634461626099495e-07,
+      "loss": 0.0769,
+      "step": 1450
+    },
+    {
+      "epoch": 12.401709401709402,
+      "grad_norm": 0.23408880880648503,
+      "learning_rate": 3.611258314276461e-07,
+      "loss": 0.0095,
+      "step": 1451
+    },
+    {
+      "epoch": 12.41025641025641,
+      "grad_norm": 0.5420648942933317,
+      "learning_rate": 3.5881235403265713e-07,
+      "loss": 0.0709,
+      "step": 1452
+    },
+    {
+      "epoch": 12.418803418803419,
+      "grad_norm": 0.9151307781476735,
+      "learning_rate": 3.56505737838265e-07,
+      "loss": 0.1828,
+      "step": 1453
+    },
+    {
+      "epoch": 12.427350427350428,
+      "grad_norm": 0.5051641079170742,
+      "learning_rate": 3.5420599023576946e-07,
+      "loss": 0.0474,
+      "step": 1454
+    },
+    {
+      "epoch": 12.435897435897436,
+      "grad_norm": 0.43289408007834856,
+      "learning_rate": 3.51913118594458e-07,
+      "loss": 0.0373,
+      "step": 1455
+    },
+    {
+      "epoch": 12.444444444444445,
+      "grad_norm": 0.45036332783573607,
+      "learning_rate": 3.4962713026158697e-07,
+      "loss": 0.0531,
+      "step": 1456
+    },
+    {
+      "epoch": 12.452991452991453,
+      "grad_norm": 0.3970269144308973,
+      "learning_rate": 3.473480325623535e-07,
+      "loss": 0.0402,
+      "step": 1457
+    },
+    {
+      "epoch": 12.461538461538462,
+      "grad_norm": 0.4664586229855834,
+      "learning_rate": 3.450758327998768e-07,
+      "loss": 0.0378,
+      "step": 1458
+    },
+    {
+      "epoch": 12.47008547008547,
+      "grad_norm": 0.5833404030851905,
+      "learning_rate": 3.428105382551716e-07,
+      "loss": 0.1227,
+      "step": 1459
+    },
+    {
+      "epoch": 12.478632478632479,
+      "grad_norm": 0.18900716916813645,
+      "learning_rate": 3.405521561871247e-07,
+      "loss": 0.0071,
+      "step": 1460
+    },
+    {
+      "epoch": 12.487179487179487,
+      "grad_norm": 0.6329825613088673,
+      "learning_rate": 3.3830069383247343e-07,
+      "loss": 0.1368,
+      "step": 1461
+    },
+    {
+      "epoch": 12.495726495726496,
+      "grad_norm": 0.3111158537112211,
+      "learning_rate": 3.3605615840578224e-07,
+      "loss": 0.0153,
+      "step": 1462
+    },
+    {
+      "epoch": 12.504273504273504,
+      "grad_norm": 0.15672363439125053,
+      "learning_rate": 3.3381855709941733e-07,
+      "loss": 0.0025,
+      "step": 1463
+    },
+    {
+      "epoch": 12.512820512820513,
+      "grad_norm": 0.2541815963874005,
+      "learning_rate": 3.315878970835268e-07,
+      "loss": 0.0116,
+      "step": 1464
+    },
+    {
+      "epoch": 12.521367521367521,
+      "grad_norm": 0.48599298068031166,
+      "learning_rate": 3.293641855060162e-07,
+      "loss": 0.0386,
+      "step": 1465
+    },
+    {
+      "epoch": 12.52991452991453,
+      "grad_norm": 0.42569652919107487,
+      "learning_rate": 3.2714742949252447e-07,
+      "loss": 0.0412,
+      "step": 1466
+    },
+    {
+      "epoch": 12.538461538461538,
+      "grad_norm": 0.4651470442081948,
+      "learning_rate": 3.249376361464021e-07,
+      "loss": 0.0621,
+      "step": 1467
+    },
+    {
+      "epoch": 12.547008547008547,
+      "grad_norm": 0.35693088161836933,
+      "learning_rate": 3.227348125486904e-07,
+      "loss": 0.0283,
+      "step": 1468
+    },
+    {
+      "epoch": 12.555555555555555,
+      "grad_norm": 0.5401582449926521,
+      "learning_rate": 3.2053896575809426e-07,
+      "loss": 0.0574,
+      "step": 1469
+    },
+    {
+      "epoch": 12.564102564102564,
+      "grad_norm": 0.5180556722548045,
+      "learning_rate": 3.1835010281096426e-07,
+      "loss": 0.0307,
+      "step": 1470
+    },
+    {
+      "epoch": 12.572649572649572,
+      "grad_norm": 0.5064784455421214,
+      "learning_rate": 3.1616823072127157e-07,
+      "loss": 0.0638,
+      "step": 1471
+    },
+    {
+      "epoch": 12.581196581196581,
+      "grad_norm": 0.2832584057082466,
+      "learning_rate": 3.1399335648058555e-07,
+      "loss": 0.0172,
+      "step": 1472
+    },
+    {
+      "epoch": 12.58974358974359,
+      "grad_norm": 0.07907910657271316,
+      "learning_rate": 3.118254870580506e-07,
+      "loss": 0.0019,
+      "step": 1473
+    },
+    {
+      "epoch": 12.598290598290598,
+      "grad_norm": 0.4856065855582885,
+      "learning_rate": 3.096646294003675e-07,
+      "loss": 0.0786,
+      "step": 1474
+    },
+    {
+      "epoch": 12.606837606837606,
+      "grad_norm": 0.5611585905476629,
+      "learning_rate": 3.075107904317667e-07,
+      "loss": 0.0493,
+      "step": 1475
+    },
+    {
+      "epoch": 12.615384615384615,
+      "grad_norm": 0.48631084188031415,
+      "learning_rate": 3.0536397705398845e-07,
+      "loss": 0.1089,
+      "step": 1476
+    },
+    {
+      "epoch": 12.623931623931623,
+      "grad_norm": 0.5762370364964107,
+      "learning_rate": 3.032241961462612e-07,
+      "loss": 0.0783,
+      "step": 1477
+    },
+    {
+      "epoch": 12.632478632478632,
+      "grad_norm": 0.25360897774454116,
+      "learning_rate": 3.010914545652771e-07,
+      "loss": 0.0135,
+      "step": 1478
+    },
+    {
+      "epoch": 12.64102564102564,
+      "grad_norm": 0.363938511193013,
+      "learning_rate": 2.9896575914517166e-07,
+      "loss": 0.0306,
+      "step": 1479
+    },
+    {
+      "epoch": 12.649572649572649,
+      "grad_norm": 0.1858863424391372,
+      "learning_rate": 2.9684711669750313e-07,
+      "loss": 0.0048,
+      "step": 1480
+    },
+    {
+      "epoch": 12.658119658119658,
+      "grad_norm": 0.24369363096646676,
+      "learning_rate": 2.9473553401122875e-07,
+      "loss": 0.0103,
+      "step": 1481
+    },
+    {
+      "epoch": 12.666666666666666,
+      "grad_norm": 0.41698180965055937,
+      "learning_rate": 2.9263101785268253e-07,
+      "loss": 0.0356,
+      "step": 1482
+    },
+    {
+      "epoch": 12.675213675213675,
+      "grad_norm": 0.2670049194018334,
+      "learning_rate": 2.9053357496555635e-07,
+      "loss": 0.0104,
+      "step": 1483
+    },
+    {
+      "epoch": 12.683760683760683,
+      "grad_norm": 0.17605755616268962,
+      "learning_rate": 2.8844321207087465e-07,
+      "loss": 0.0051,
+      "step": 1484
+    },
+    {
+      "epoch": 12.692307692307692,
+      "grad_norm": 0.4103155743327306,
+      "learning_rate": 2.8635993586697555e-07,
+      "loss": 0.0345,
+      "step": 1485
+    },
+    {
+      "epoch": 12.7008547008547,
+      "grad_norm": 0.4605812519261627,
+      "learning_rate": 2.84283753029489e-07,
+      "loss": 0.0229,
+      "step": 1486
+    },
+    {
+      "epoch": 12.709401709401709,
+      "grad_norm": 0.40217176084041906,
+      "learning_rate": 2.822146702113157e-07,
+      "loss": 0.0486,
+      "step": 1487
+    },
+    {
+      "epoch": 12.717948717948717,
+      "grad_norm": 0.30183863999283517,
+      "learning_rate": 2.8015269404260333e-07,
+      "loss": 0.0234,
+      "step": 1488
+    },
+    {
+      "epoch": 12.726495726495726,
+      "grad_norm": 0.43189722616688536,
+      "learning_rate": 2.780978311307278e-07,
+      "loss": 0.028,
+      "step": 1489
+    },
+    {
+      "epoch": 12.735042735042736,
+      "grad_norm": 0.18951438153753558,
+      "learning_rate": 2.7605008806027205e-07,
+      "loss": 0.0056,
+      "step": 1490
+    },
+    {
+      "epoch": 12.743589743589745,
+      "grad_norm": 0.5437323913691973,
+      "learning_rate": 2.7400947139300443e-07,
+      "loss": 0.0827,
+      "step": 1491
+    },
+    {
+      "epoch": 12.752136752136753,
+      "grad_norm": 0.40461808187778764,
+      "learning_rate": 2.7197598766785544e-07,
+      "loss": 0.0484,
+      "step": 1492
+    },
+    {
+      "epoch": 12.760683760683762,
+      "grad_norm": 0.42001625712471485,
+      "learning_rate": 2.6994964340090163e-07,
+      "loss": 0.0497,
+      "step": 1493
+    },
+    {
+      "epoch": 12.76923076923077,
+      "grad_norm": 0.2970457000400673,
+      "learning_rate": 2.679304450853401e-07,
+      "loss": 0.0166,
+      "step": 1494
+    },
+    {
+      "epoch": 12.777777777777779,
+      "grad_norm": 0.7433458397781355,
+      "learning_rate": 2.6591839919146963e-07,
+      "loss": 0.093,
+      "step": 1495
+    },
+    {
+      "epoch": 12.786324786324787,
+      "grad_norm": 0.5356885326565816,
+      "learning_rate": 2.6391351216667053e-07,
+      "loss": 0.0625,
+      "step": 1496
+    },
+    {
+      "epoch": 12.794871794871796,
+      "grad_norm": 0.30520132963784746,
+      "learning_rate": 2.6191579043538333e-07,
+      "loss": 0.0287,
+      "step": 1497
+    },
+    {
+      "epoch": 12.803418803418804,
+      "grad_norm": 0.4444882375877483,
+      "learning_rate": 2.599252403990873e-07,
+      "loss": 0.0426,
+      "step": 1498
+    },
+    {
+      "epoch": 12.811965811965813,
+      "grad_norm": 0.41946477409959565,
+      "learning_rate": 2.5794186843628247e-07,
+      "loss": 0.0485,
+      "step": 1499
+    },
+    {
+      "epoch": 12.820512820512821,
+      "grad_norm": 0.47615560566109044,
+      "learning_rate": 2.5596568090246546e-07,
+      "loss": 0.0585,
+      "step": 1500
+    },
+    {
+      "epoch": 12.82905982905983,
+      "grad_norm": 0.25278361952778927,
+      "learning_rate": 2.539966841301117e-07,
+      "loss": 0.0188,
+      "step": 1501
+    },
+    {
+      "epoch": 12.837606837606838,
+      "grad_norm": 0.7435079955100731,
+      "learning_rate": 2.5203488442865574e-07,
+      "loss": 0.1664,
+      "step": 1502
+    },
+    {
+      "epoch": 12.846153846153847,
+      "grad_norm": 0.4905397007029981,
+      "learning_rate": 2.5008028808446995e-07,
+      "loss": 0.0644,
+      "step": 1503
+    },
+    {
+      "epoch": 12.854700854700855,
+      "grad_norm": 0.3579408204459078,
+      "learning_rate": 2.481329013608419e-07,
+      "loss": 0.0161,
+      "step": 1504
+    },
+    {
+      "epoch": 12.863247863247864,
+      "grad_norm": 0.3690746454860683,
+      "learning_rate": 2.4619273049796e-07,
+      "loss": 0.0147,
+      "step": 1505
+    },
+    {
+      "epoch": 12.871794871794872,
+      "grad_norm": 0.5539640918674135,
+      "learning_rate": 2.4425978171288807e-07,
+      "loss": 0.0746,
+      "step": 1506
+    },
+    {
+      "epoch": 12.88034188034188,
+      "grad_norm": 0.398801152487831,
+      "learning_rate": 2.4233406119954756e-07,
+      "loss": 0.057,
+      "step": 1507
+    },
+    {
+      "epoch": 12.88888888888889,
+      "grad_norm": 0.28829865983970265,
+      "learning_rate": 2.404155751286988e-07,
+      "loss": 0.0131,
+      "step": 1508
+    },
+    {
+      "epoch": 12.897435897435898,
+      "grad_norm": 0.10754494414168028,
+      "learning_rate": 2.385043296479195e-07,
+      "loss": 0.0024,
+      "step": 1509
+    },
+    {
+      "epoch": 12.905982905982906,
+      "grad_norm": 0.4987503634158719,
+      "learning_rate": 2.3660033088158646e-07,
+      "loss": 0.0357,
+      "step": 1510
+    },
+    {
+      "epoch": 12.914529914529915,
+      "grad_norm": 0.7081484429156921,
+      "learning_rate": 2.3470358493085433e-07,
+      "loss": 0.1405,
+      "step": 1511
+    },
+    {
+      "epoch": 12.923076923076923,
+      "grad_norm": 0.3988280902988456,
+      "learning_rate": 2.3281409787363652e-07,
+      "loss": 0.0436,
+      "step": 1512
+    },
+    {
+      "epoch": 12.931623931623932,
+      "grad_norm": 0.26292834769791484,
+      "learning_rate": 2.309318757645873e-07,
+      "loss": 0.0118,
+      "step": 1513
+    },
+    {
+      "epoch": 12.94017094017094,
+      "grad_norm": 0.2947911611296075,
+      "learning_rate": 2.2905692463508045e-07,
+      "loss": 0.0135,
+      "step": 1514
+    },
+    {
+      "epoch": 12.948717948717949,
+      "grad_norm": 0.36055620968224433,
+      "learning_rate": 2.271892504931905e-07,
+      "loss": 0.0364,
+      "step": 1515
+    },
+    {
+      "epoch": 12.957264957264957,
+      "grad_norm": 0.6805417037367615,
+      "learning_rate": 2.253288593236755e-07,
+      "loss": 0.0844,
+      "step": 1516
+    },
+    {
+      "epoch": 12.965811965811966,
+      "grad_norm": 0.3855234002350498,
+      "learning_rate": 2.234757570879534e-07,
+      "loss": 0.0244,
+      "step": 1517
+    },
+    {
+      "epoch": 12.974358974358974,
+      "grad_norm": 0.7122804793044082,
+      "learning_rate": 2.2162994972408647e-07,
+      "loss": 0.1131,
+      "step": 1518
+    },
+    {
+      "epoch": 12.982905982905983,
+      "grad_norm": 0.2188849119906226,
+      "learning_rate": 2.1979144314676239e-07,
+      "loss": 0.0073,
+      "step": 1519
+    },
+    {
+      "epoch": 12.991452991452991,
+      "grad_norm": 0.47128998500439206,
+      "learning_rate": 2.17960243247273e-07,
+      "loss": 0.0437,
+      "step": 1520
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.5310311333782931,
+      "learning_rate": 2.1613635589349756e-07,
+      "loss": 0.065,
+      "step": 1521
+    },
+    {
+      "epoch": 13.008547008547009,
+      "grad_norm": 0.3479275975126997,
+      "learning_rate": 2.1431978692988298e-07,
+      "loss": 0.0612,
+      "step": 1522
+    },
+    {
+      "epoch": 13.017094017094017,
+      "grad_norm": 0.3523745771232036,
+      "learning_rate": 2.1251054217742484e-07,
+      "loss": 0.0348,
+      "step": 1523
+    },
+    {
+      "epoch": 13.025641025641026,
+      "grad_norm": 0.48983824212131266,
+      "learning_rate": 2.107086274336484e-07,
+      "loss": 0.0506,
+      "step": 1524
+    },
+    {
+      "epoch": 13.034188034188034,
+      "grad_norm": 0.5469787413822318,
+      "learning_rate": 2.0891404847259267e-07,
+      "loss": 0.0375,
+      "step": 1525
+    },
+    {
+      "epoch": 13.042735042735043,
+      "grad_norm": 0.3376707731497175,
+      "learning_rate": 2.0712681104478742e-07,
+      "loss": 0.0378,
+      "step": 1526
+    },
+    {
+      "epoch": 13.051282051282051,
+      "grad_norm": 0.5034242344115563,
+      "learning_rate": 2.0534692087724017e-07,
+      "loss": 0.0555,
+      "step": 1527
+    },
+    {
+      "epoch": 13.05982905982906,
+      "grad_norm": 0.25502809856153064,
+      "learning_rate": 2.0357438367341248e-07,
+      "loss": 0.0124,
+      "step": 1528
+    },
+    {
+      "epoch": 13.068376068376068,
+      "grad_norm": 0.0696046335947878,
+      "learning_rate": 2.0180920511320562e-07,
+      "loss": 0.0017,
+      "step": 1529
+    },
+    {
+      "epoch": 13.076923076923077,
+      "grad_norm": 0.3575175521969691,
+      "learning_rate": 2.0005139085293945e-07,
+      "loss": 0.0353,
+      "step": 1530
+    },
+    {
+      "epoch": 13.085470085470085,
+      "grad_norm": 0.4869086930900781,
+      "learning_rate": 1.983009465253377e-07,
+      "loss": 0.0705,
+      "step": 1531
+    },
+    {
+      "epoch": 13.094017094017094,
+      "grad_norm": 0.5167937928937567,
+      "learning_rate": 1.9655787773950691e-07,
+      "loss": 0.085,
+      "step": 1532
+    },
+    {
+      "epoch": 13.102564102564102,
+      "grad_norm": 0.29527783197774443,
+      "learning_rate": 1.9482219008091885e-07,
+      "loss": 0.0157,
+      "step": 1533
+    },
+    {
+      "epoch": 13.11111111111111,
+      "grad_norm": 0.3252991705388003,
+      "learning_rate": 1.9309388911139427e-07,
+      "loss": 0.0223,
+      "step": 1534
+    },
+    {
+      "epoch": 13.11965811965812,
+      "grad_norm": 0.33335729772770656,
+      "learning_rate": 1.9137298036908392e-07,
+      "loss": 0.025,
+      "step": 1535
+    },
+    {
+      "epoch": 13.128205128205128,
+      "grad_norm": 0.39648258193280794,
+      "learning_rate": 1.896594693684503e-07,
+      "loss": 0.0334,
+      "step": 1536
+    },
+    {
+      "epoch": 13.136752136752136,
+      "grad_norm": 0.35775145766381444,
+      "learning_rate": 1.8795336160025157e-07,
+      "loss": 0.0395,
+      "step": 1537
+    },
+    {
+      "epoch": 13.145299145299145,
+      "grad_norm": 0.5482622815094932,
+      "learning_rate": 1.8625466253152314e-07,
+      "loss": 0.0654,
+      "step": 1538
+    },
+    {
+      "epoch": 13.153846153846153,
+      "grad_norm": 0.3848712493794455,
+      "learning_rate": 1.8456337760555915e-07,
+      "loss": 0.0728,
+      "step": 1539
+    },
+    {
+      "epoch": 13.162393162393162,
+      "grad_norm": 0.47502030198058615,
+      "learning_rate": 1.8287951224189555e-07,
+      "loss": 0.0839,
+      "step": 1540
+    },
+    {
+      "epoch": 13.17094017094017,
+      "grad_norm": 0.3083495744880248,
+      "learning_rate": 1.8120307183629533e-07,
+      "loss": 0.0185,
+      "step": 1541
+    },
+    {
+      "epoch": 13.179487179487179,
+      "grad_norm": 0.12747381613610742,
+      "learning_rate": 1.7953406176072636e-07,
+      "loss": 0.0038,
+      "step": 1542
+    },
+    {
+      "epoch": 13.188034188034187,
+      "grad_norm": 0.4041400206367801,
+      "learning_rate": 1.778724873633486e-07,
+      "loss": 0.0573,
+      "step": 1543
+    },
+    {
+      "epoch": 13.196581196581196,
+      "grad_norm": 0.4612137803899462,
+      "learning_rate": 1.7621835396849528e-07,
+      "loss": 0.0485,
+      "step": 1544
+    },
+    {
+      "epoch": 13.205128205128204,
+      "grad_norm": 0.25259858595605755,
+      "learning_rate": 1.745716668766545e-07,
+      "loss": 0.0119,
+      "step": 1545
+    },
+    {
+      "epoch": 13.213675213675213,
+      "grad_norm": 0.21737171693683865,
+      "learning_rate": 1.7293243136445398e-07,
+      "loss": 0.0084,
+      "step": 1546
+    },
+    {
+      "epoch": 13.222222222222221,
+      "grad_norm": 0.5257948866349633,
+      "learning_rate": 1.713006526846439e-07,
+      "loss": 0.1058,
+      "step": 1547
+    },
+    {
+      "epoch": 13.23076923076923,
+      "grad_norm": 0.3647445265942455,
+      "learning_rate": 1.6967633606608082e-07,
+      "loss": 0.0306,
+      "step": 1548
+    },
+    {
+      "epoch": 13.239316239316238,
+      "grad_norm": 0.6273652237714832,
+      "learning_rate": 1.6805948671370726e-07,
+      "loss": 0.0965,
+      "step": 1549
+    },
+    {
+      "epoch": 13.247863247863247,
+      "grad_norm": 0.4250500011918908,
+      "learning_rate": 1.6645010980854082e-07,
+      "loss": 0.0336,
+      "step": 1550
+    },
+    {
+      "epoch": 13.256410256410255,
+      "grad_norm": 0.2541269183698496,
+      "learning_rate": 1.6484821050765209e-07,
+      "loss": 0.0207,
+      "step": 1551
+    },
+    {
+      "epoch": 13.264957264957266,
+      "grad_norm": 0.6916501583191542,
+      "learning_rate": 1.6325379394415168e-07,
+      "loss": 0.0892,
+      "step": 1552
+    },
+    {
+      "epoch": 13.273504273504274,
+      "grad_norm": 0.5419859168716888,
+      "learning_rate": 1.6166686522717217e-07,
+      "loss": 0.0845,
+      "step": 1553
+    },
+    {
+      "epoch": 13.282051282051283,
+      "grad_norm": 0.31897878123049517,
+      "learning_rate": 1.600874294418528e-07,
+      "loss": 0.0281,
+      "step": 1554
+    },
+    {
+      "epoch": 13.290598290598291,
+      "grad_norm": 0.37425385936599204,
+      "learning_rate": 1.5851549164932118e-07,
+      "loss": 0.0336,
+      "step": 1555
+    },
+    {
+      "epoch": 13.2991452991453,
+      "grad_norm": 0.2631276266077297,
+      "learning_rate": 1.569510568866803e-07,
+      "loss": 0.0182,
+      "step": 1556
+    },
+    {
+      "epoch": 13.307692307692308,
+      "grad_norm": 0.53551321371183,
+      "learning_rate": 1.5539413016698923e-07,
+      "loss": 0.0862,
+      "step": 1557
+    },
+    {
+      "epoch": 13.316239316239317,
+      "grad_norm": 0.573266375875409,
+      "learning_rate": 1.5384471647924781e-07,
+      "loss": 0.0848,
+      "step": 1558
+    },
+    {
+      "epoch": 13.324786324786325,
+      "grad_norm": 0.34799927862110314,
+      "learning_rate": 1.5230282078838255e-07,
+      "loss": 0.0323,
+      "step": 1559
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 0.3502320548848171,
+      "learning_rate": 1.507684480352292e-07,
+      "loss": 0.027,
+      "step": 1560
+    },
+    {
+      "epoch": 13.341880341880342,
+      "grad_norm": 0.47342922032100826,
+      "learning_rate": 1.4924160313651598e-07,
+      "loss": 0.0559,
+      "step": 1561
+    },
+    {
+      "epoch": 13.350427350427351,
+      "grad_norm": 0.3824196536808161,
+      "learning_rate": 1.4772229098485053e-07,
+      "loss": 0.0346,
+      "step": 1562
+    },
+    {
+      "epoch": 13.35897435897436,
+      "grad_norm": 0.6077069983234155,
+      "learning_rate": 1.46210516448701e-07,
+      "loss": 0.0883,
+      "step": 1563
+    },
+    {
+      "epoch": 13.367521367521368,
+      "grad_norm": 0.20174686833329616,
+      "learning_rate": 1.447062843723826e-07,
+      "loss": 0.0076,
+      "step": 1564
+    },
+    {
+      "epoch": 13.376068376068377,
+      "grad_norm": 0.3472339294769054,
+      "learning_rate": 1.432095995760424e-07,
+      "loss": 0.0281,
+      "step": 1565
+    },
+    {
+      "epoch": 13.384615384615385,
+      "grad_norm": 0.5395406632267746,
+      "learning_rate": 1.417204668556421e-07,
+      "loss": 0.0608,
+      "step": 1566
+    },
+    {
+      "epoch": 13.393162393162394,
+      "grad_norm": 0.26952215154290254,
+      "learning_rate": 1.402388909829447e-07,
+      "loss": 0.0118,
+      "step": 1567
+    },
+    {
+      "epoch": 13.401709401709402,
+      "grad_norm": 0.3298420949413113,
+      "learning_rate": 1.387648767054961e-07,
+      "loss": 0.031,
+      "step": 1568
+    },
+    {
+      "epoch": 13.41025641025641,
+      "grad_norm": 0.563406459481279,
+      "learning_rate": 1.3729842874661365e-07,
+      "loss": 0.113,
+      "step": 1569
+    },
+    {
+      "epoch": 13.418803418803419,
+      "grad_norm": 0.574681376341801,
+      "learning_rate": 1.35839551805369e-07,
+      "loss": 0.0804,
+      "step": 1570
+    },
+    {
+      "epoch": 13.427350427350428,
+      "grad_norm": 0.3623717101138977,
+      "learning_rate": 1.3438825055657246e-07,
+      "loss": 0.0234,
+      "step": 1571
+    },
+    {
+      "epoch": 13.435897435897436,
+      "grad_norm": 0.4861599303938593,
+      "learning_rate": 1.3294452965076031e-07,
+      "loss": 0.0372,
+      "step": 1572
+    },
+    {
+      "epoch": 13.444444444444445,
+      "grad_norm": 0.34085334824763863,
+      "learning_rate": 1.31508393714177e-07,
+      "loss": 0.0206,
+      "step": 1573
+    },
+    {
+      "epoch": 13.452991452991453,
+      "grad_norm": 0.3258647492947037,
+      "learning_rate": 1.3007984734876217e-07,
+      "loss": 0.0258,
+      "step": 1574
+    },
+    {
+      "epoch": 13.461538461538462,
+      "grad_norm": 0.6071085443076429,
+      "learning_rate": 1.286588951321363e-07,
+      "loss": 0.0776,
+      "step": 1575
+    },
+    {
+      "epoch": 13.47008547008547,
+      "grad_norm": 0.5031416201199156,
+      "learning_rate": 1.272455416175844e-07,
+      "loss": 0.0207,
+      "step": 1576
+    },
+    {
+      "epoch": 13.478632478632479,
+      "grad_norm": 0.3268535611369224,
+      "learning_rate": 1.258397913340423e-07,
+      "loss": 0.0355,
+      "step": 1577
+    },
+    {
+      "epoch": 13.487179487179487,
+      "grad_norm": 0.2851643628328826,
+      "learning_rate": 1.2444164878608307e-07,
+      "loss": 0.0221,
+      "step": 1578
+    },
+    {
+      "epoch": 13.495726495726496,
+      "grad_norm": 0.2217773892646123,
+      "learning_rate": 1.2305111845390043e-07,
+      "loss": 0.0087,
+      "step": 1579
+    },
+    {
+      "epoch": 13.504273504273504,
+      "grad_norm": 0.5523812907991107,
+      "learning_rate": 1.2166820479329572e-07,
+      "loss": 0.077,
+      "step": 1580
+    },
+    {
+      "epoch": 13.512820512820513,
+      "grad_norm": 0.221037095991797,
+      "learning_rate": 1.2029291223566413e-07,
+      "loss": 0.0121,
+      "step": 1581
+    },
+    {
+      "epoch": 13.521367521367521,
+      "grad_norm": 0.39931183898142225,
+      "learning_rate": 1.1892524518797993e-07,
+      "loss": 0.0256,
+      "step": 1582
+    },
+    {
+      "epoch": 13.52991452991453,
+      "grad_norm": 0.4816009623881913,
+      "learning_rate": 1.1756520803278188e-07,
+      "loss": 0.0561,
+      "step": 1583
+    },
+    {
+      "epoch": 13.538461538461538,
+      "grad_norm": 0.12466048888011573,
+      "learning_rate": 1.1621280512815941e-07,
+      "loss": 0.0039,
+      "step": 1584
+    },
+    {
+      "epoch": 13.547008547008547,
+      "grad_norm": 0.5208821923816866,
+      "learning_rate": 1.1486804080773878e-07,
+      "loss": 0.0711,
+      "step": 1585
+    },
+    {
+      "epoch": 13.555555555555555,
+      "grad_norm": 0.3562879598580212,
+      "learning_rate": 1.1353091938067024e-07,
+      "loss": 0.038,
+      "step": 1586
+    },
+    {
+      "epoch": 13.564102564102564,
+      "grad_norm": 0.4062408749849314,
+      "learning_rate": 1.1220144513161197e-07,
+      "loss": 0.044,
+      "step": 1587
+    },
+    {
+      "epoch": 13.572649572649572,
+      "grad_norm": 0.5458930519363671,
+      "learning_rate": 1.1087962232071814e-07,
+      "loss": 0.0765,
+      "step": 1588
+    },
+    {
+      "epoch": 13.581196581196581,
+      "grad_norm": 0.29040690525508095,
+      "learning_rate": 1.0956545518362532e-07,
+      "loss": 0.0242,
+      "step": 1589
+    },
+    {
+      "epoch": 13.58974358974359,
+      "grad_norm": 0.2792827354829269,
+      "learning_rate": 1.0825894793143721e-07,
+      "loss": 0.0264,
+      "step": 1590
+    },
+    {
+      "epoch": 13.598290598290598,
+      "grad_norm": 0.24223454782550927,
+      "learning_rate": 1.0696010475071267e-07,
+      "loss": 0.014,
+      "step": 1591
+    },
+    {
+      "epoch": 13.606837606837606,
+      "grad_norm": 0.4732093758685518,
+      "learning_rate": 1.0566892980345245e-07,
+      "loss": 0.0445,
+      "step": 1592
+    },
+    {
+      "epoch": 13.615384615384615,
+      "grad_norm": 0.5689145979311558,
+      "learning_rate": 1.0438542722708444e-07,
+      "loss": 0.0748,
+      "step": 1593
+    },
+    {
+      "epoch": 13.623931623931623,
+      "grad_norm": 0.5134034697634273,
+      "learning_rate": 1.0310960113445179e-07,
+      "loss": 0.0577,
+      "step": 1594
+    },
+    {
+      "epoch": 13.632478632478632,
+      "grad_norm": 0.12181138755781606,
+      "learning_rate": 1.0184145561379949e-07,
+      "loss": 0.0024,
+      "step": 1595
+    },
+    {
+      "epoch": 13.64102564102564,
+      "grad_norm": 0.42930052739485963,
+      "learning_rate": 1.0058099472876004e-07,
+      "loss": 0.0726,
+      "step": 1596
+    },
+    {
+      "epoch": 13.649572649572649,
+      "grad_norm": 0.5417065797654803,
+      "learning_rate": 9.932822251834173e-08,
+      "loss": 0.1092,
+      "step": 1597
+    },
+    {
+      "epoch": 13.658119658119658,
+      "grad_norm": 0.44463681603211913,
+      "learning_rate": 9.808314299691591e-08,
+      "loss": 0.0518,
+      "step": 1598
+    },
+    {
+      "epoch": 13.666666666666666,
+      "grad_norm": 0.4935889402856598,
+      "learning_rate": 9.684576015420277e-08,
+      "loss": 0.0572,
+      "step": 1599
+    },
+    {
+      "epoch": 13.675213675213675,
+      "grad_norm": 0.3010999309662844,
+      "learning_rate": 9.561607795526007e-08,
+      "loss": 0.0236,
+      "step": 1600
+    },
+    {
+      "epoch": 13.683760683760683,
+      "grad_norm": 0.4115555617813383,
+      "learning_rate": 9.439410034046942e-08,
+      "loss": 0.0355,
+      "step": 1601
+    },
+    {
+      "epoch": 13.692307692307692,
+      "grad_norm": 0.23478638733192264,
+      "learning_rate": 9.317983122552332e-08,
+      "loss": 0.0054,
+      "step": 1602
+    },
+    {
+      "epoch": 13.7008547008547,
+      "grad_norm": 0.400836154779937,
+      "learning_rate": 9.197327450141402e-08,
+      "loss": 0.0422,
+      "step": 1603
+    },
+    {
+      "epoch": 13.709401709401709,
+      "grad_norm": 0.3816228650221198,
+      "learning_rate": 9.077443403441994e-08,
+      "loss": 0.0352,
+      "step": 1604
+    },
+    {
+      "epoch": 13.717948717948717,
+      "grad_norm": 0.2824770084219213,
+      "learning_rate": 8.958331366609424e-08,
+      "loss": 0.0181,
+      "step": 1605
+    },
+    {
+      "epoch": 13.726495726495726,
+      "grad_norm": 0.40392339013272,
+      "learning_rate": 8.839991721325047e-08,
+      "loss": 0.0338,
+      "step": 1606
+    },
+    {
+      "epoch": 13.735042735042736,
+      "grad_norm": 0.3746265965673062,
+      "learning_rate": 8.72242484679528e-08,
+      "loss": 0.0333,
+      "step": 1607
+    },
+    {
+      "epoch": 13.743589743589745,
+      "grad_norm": 0.4591987932732178,
+      "learning_rate": 8.605631119750297e-08,
+      "loss": 0.0426,
+      "step": 1608
+    },
+    {
+      "epoch": 13.752136752136753,
+      "grad_norm": 0.5298373214853962,
+      "learning_rate": 8.489610914442697e-08,
+      "loss": 0.0919,
+      "step": 1609
+    },
+    {
+      "epoch": 13.760683760683762,
+      "grad_norm": 0.5254343214165913,
+      "learning_rate": 8.374364602646512e-08,
+      "loss": 0.0598,
+      "step": 1610
+    },
+    {
+      "epoch": 13.76923076923077,
+      "grad_norm": 0.2749095870578622,
+      "learning_rate": 8.259892553655946e-08,
+      "loss": 0.0097,
+      "step": 1611
+    },
+    {
+      "epoch": 13.777777777777779,
+      "grad_norm": 0.46888063486471226,
+      "learning_rate": 8.146195134284052e-08,
+      "loss": 0.0509,
+      "step": 1612
+    },
+    {
+      "epoch": 13.786324786324787,
+      "grad_norm": 0.5940355968456708,
+      "learning_rate": 8.033272708861673e-08,
+      "loss": 0.0554,
+      "step": 1613
+    },
+    {
+      "epoch": 13.794871794871796,
+      "grad_norm": 0.3992974882422677,
+      "learning_rate": 7.921125639236416e-08,
+      "loss": 0.0434,
+      "step": 1614
+    },
+    {
+      "epoch": 13.803418803418804,
+      "grad_norm": 0.39756035665555717,
+      "learning_rate": 7.809754284771181e-08,
+      "loss": 0.0545,
+      "step": 1615
+    },
+    {
+      "epoch": 13.811965811965813,
+      "grad_norm": 0.41475580652501604,
+      "learning_rate": 7.699159002343248e-08,
+      "loss": 0.0425,
+      "step": 1616
+    },
+    {
+      "epoch": 13.820512820512821,
+      "grad_norm": 0.2936759248756884,
+      "learning_rate": 7.589340146343077e-08,
+      "loss": 0.0287,
+      "step": 1617
+    },
+    {
+      "epoch": 13.82905982905983,
+      "grad_norm": 0.3087411584867849,
+      "learning_rate": 7.48029806867312e-08,
+      "loss": 0.026,
+      "step": 1618
+    },
+    {
+      "epoch": 13.837606837606838,
+      "grad_norm": 0.5275078873095708,
+      "learning_rate": 7.372033118746708e-08,
+      "loss": 0.0581,
+      "step": 1619
+    },
+    {
+      "epoch": 13.846153846153847,
+      "grad_norm": 0.17224019225984827,
+      "learning_rate": 7.264545643486997e-08,
+      "loss": 0.0059,
+      "step": 1620
+    },
+    {
+      "epoch": 13.854700854700855,
+      "grad_norm": 0.6334224216022728,
+      "learning_rate": 7.157835987325807e-08,
+      "loss": 0.1366,
+      "step": 1621
+    },
+    {
+      "epoch": 13.863247863247864,
+      "grad_norm": 0.46343142790855263,
+      "learning_rate": 7.051904492202472e-08,
+      "loss": 0.0127,
+      "step": 1622
+    },
+    {
+      "epoch": 13.871794871794872,
+      "grad_norm": 0.37134376223079346,
+      "learning_rate": 6.946751497562909e-08,
+      "loss": 0.0255,
+      "step": 1623
+    },
+    {
+      "epoch": 13.88034188034188,
+      "grad_norm": 0.5367950100162845,
+      "learning_rate": 6.842377340358252e-08,
+      "loss": 0.0488,
+      "step": 1624
+    },
+    {
+      "epoch": 13.88888888888889,
+      "grad_norm": 0.4398420912857589,
+      "learning_rate": 6.738782355044048e-08,
+      "loss": 0.0478,
+      "step": 1625
+    },
+    {
+      "epoch": 13.897435897435898,
+      "grad_norm": 0.48278027579001104,
+      "learning_rate": 6.635966873579063e-08,
+      "loss": 0.0691,
+      "step": 1626
+    },
+    {
+      "epoch": 13.905982905982906,
+      "grad_norm": 0.3897204443938729,
+      "learning_rate": 6.5339312254242e-08,
+      "loss": 0.0468,
+      "step": 1627
+    },
+    {
+      "epoch": 13.914529914529915,
+      "grad_norm": 0.450160795007421,
+      "learning_rate": 6.432675737541499e-08,
+      "loss": 0.0468,
+      "step": 1628
+    },
+    {
+      "epoch": 13.923076923076923,
+      "grad_norm": 0.4085390644340357,
+      "learning_rate": 6.332200734393057e-08,
+      "loss": 0.0402,
+      "step": 1629
+    },
+    {
+      "epoch": 13.931623931623932,
+      "grad_norm": 0.34213546747552515,
+      "learning_rate": 6.232506537939942e-08,
+      "loss": 0.0271,
+      "step": 1630
+    },
+    {
+      "epoch": 13.94017094017094,
+      "grad_norm": 0.6332223020088068,
+      "learning_rate": 6.13359346764128e-08,
+      "loss": 0.1059,
+      "step": 1631
+    },
+    {
+      "epoch": 13.948717948717949,
+      "grad_norm": 0.6022072045276918,
+      "learning_rate": 6.035461840453116e-08,
+      "loss": 0.1051,
+      "step": 1632
+    },
+    {
+      "epoch": 13.957264957264957,
+      "grad_norm": 0.514020698868604,
+      "learning_rate": 5.938111970827526e-08,
+      "loss": 0.0802,
+      "step": 1633
+    },
+    {
+      "epoch": 13.965811965811966,
+      "grad_norm": 0.07354635381931279,
+      "learning_rate": 5.841544170711422e-08,
+      "loss": 0.0016,
+      "step": 1634
+    },
+    {
+      "epoch": 13.974358974358974,
+      "grad_norm": 0.27811119952922675,
+      "learning_rate": 5.745758749545749e-08,
+      "loss": 0.0154,
+      "step": 1635
+    },
+    {
+      "epoch": 13.982905982905983,
+      "grad_norm": 0.5656664303415825,
+      "learning_rate": 5.650756014264347e-08,
+      "loss": 0.0775,
+      "step": 1636
+    },
+    {
+      "epoch": 13.991452991452991,
+      "grad_norm": 0.4503223978301062,
+      "learning_rate": 5.556536269293006e-08,
+      "loss": 0.046,
+      "step": 1637
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.4582080471389822,
+      "learning_rate": 5.463099816548578e-08,
+      "loss": 0.0648,
+      "step": 1638
+    },
+    {
+      "epoch": 14.008547008547009,
+      "grad_norm": 0.2483000115491485,
+      "learning_rate": 5.3704469554379527e-08,
+      "loss": 0.0108,
+      "step": 1639
+    },
+    {
+      "epoch": 14.017094017094017,
+      "grad_norm": 0.5165846852271239,
+      "learning_rate": 5.278577982857025e-08,
+      "loss": 0.0615,
+      "step": 1640
+    },
+    {
+      "epoch": 14.025641025641026,
+      "grad_norm": 0.5347615828908445,
+      "learning_rate": 5.1874931931897854e-08,
+      "loss": 0.0691,
+      "step": 1641
+    },
+    {
+      "epoch": 14.034188034188034,
+      "grad_norm": 0.4100464066416389,
+      "learning_rate": 5.097192878307455e-08,
+      "loss": 0.0388,
+      "step": 1642
+    },
+    {
+      "epoch": 14.042735042735043,
+      "grad_norm": 0.6027519857440805,
+      "learning_rate": 5.0076773275675174e-08,
+      "loss": 0.1063,
+      "step": 1643
+    },
+    {
+      "epoch": 14.051282051282051,
+      "grad_norm": 0.2553611985836167,
+      "learning_rate": 4.91894682781266e-08,
+      "loss": 0.0135,
+      "step": 1644
+    },
+    {
+      "epoch": 14.05982905982906,
+      "grad_norm": 0.5578825939428207,
+      "learning_rate": 4.831001663370083e-08,
+      "loss": 0.108,
+      "step": 1645
+    },
+    {
+      "epoch": 14.068376068376068,
+      "grad_norm": 0.4895323748851477,
+      "learning_rate": 4.743842116050334e-08,
+      "loss": 0.043,
+      "step": 1646
+    },
+    {
+      "epoch": 14.076923076923077,
+      "grad_norm": 0.4128574776178798,
+      "learning_rate": 4.657468465146642e-08,
+      "loss": 0.0324,
+      "step": 1647
+    },
+    {
+      "epoch": 14.085470085470085,
+      "grad_norm": 0.46410297469667755,
+      "learning_rate": 4.571880987433886e-08,
+      "loss": 0.0566,
+      "step": 1648
+    },
+    {
+      "epoch": 14.094017094017094,
+      "grad_norm": 0.3848244141143089,
+      "learning_rate": 4.487079957167767e-08,
+      "loss": 0.0359,
+      "step": 1649
+    },
+    {
+      "epoch": 14.102564102564102,
+      "grad_norm": 0.3511233417593998,
+      "learning_rate": 4.40306564608381e-08,
+      "loss": 0.0368,
+      "step": 1650
+    },
+    {
+      "epoch": 14.11111111111111,
+      "grad_norm": 0.1622634968755906,
+      "learning_rate": 4.319838323396691e-08,
+      "loss": 0.0058,
+      "step": 1651
+    },
+    {
+      "epoch": 14.11965811965812,
+      "grad_norm": 0.21186337433136335,
+      "learning_rate": 4.237398255799191e-08,
+      "loss": 0.0081,
+      "step": 1652
+    },
+    {
+      "epoch": 14.128205128205128,
+      "grad_norm": 0.3672513072749568,
+      "learning_rate": 4.155745707461467e-08,
+      "loss": 0.0222,
+      "step": 1653
+    },
+    {
+      "epoch": 14.136752136752136,
+      "grad_norm": 0.4195684480150308,
+      "learning_rate": 4.0748809400301403e-08,
+      "loss": 0.0536,
+      "step": 1654
+    },
+    {
+      "epoch": 14.145299145299145,
+      "grad_norm": 0.42261292114989146,
+      "learning_rate": 3.994804212627462e-08,
+      "loss": 0.0441,
+      "step": 1655
+    },
+    {
+      "epoch": 14.153846153846153,
+      "grad_norm": 0.25385726081335536,
+      "learning_rate": 3.9155157818505654e-08,
+      "loss": 0.0112,
+      "step": 1656
+    },
+    {
+      "epoch": 14.162393162393162,
+      "grad_norm": 0.5415475073424915,
+      "learning_rate": 3.8370159017704636e-08,
+      "loss": 0.0725,
+      "step": 1657
+    },
+    {
+      "epoch": 14.17094017094017,
+      "grad_norm": 0.5554466464885877,
+      "learning_rate": 3.759304823931359e-08,
+      "loss": 0.0651,
+      "step": 1658
+    },
+    {
+      "epoch": 14.179487179487179,
+      "grad_norm": 0.47897150903692604,
+      "learning_rate": 3.682382797349976e-08,
+      "loss": 0.041,
+      "step": 1659
+    },
+    {
+      "epoch": 14.188034188034187,
+      "grad_norm": 0.3348930277986746,
+      "learning_rate": 3.6062500685143943e-08,
+      "loss": 0.0316,
+      "step": 1660
+    },
+    {
+      "epoch": 14.196581196581196,
+      "grad_norm": 0.5813358682096855,
+      "learning_rate": 3.5309068813836056e-08,
+      "loss": 0.0748,
+      "step": 1661
+    },
+    {
+      "epoch": 14.205128205128204,
+      "grad_norm": 0.5989693941469737,
+      "learning_rate": 3.4563534773866256e-08,
+      "loss": 0.1531,
+      "step": 1662
+    },
+    {
+      "epoch": 14.213675213675213,
+      "grad_norm": 0.3821820832883375,
+      "learning_rate": 3.382590095421606e-08,
+      "loss": 0.0352,
+      "step": 1663
+    },
+    {
+      "epoch": 14.222222222222221,
+      "grad_norm": 0.30134449071243946,
+      "learning_rate": 3.309616971855195e-08,
+      "loss": 0.0297,
+      "step": 1664
+    },
+    {
+      "epoch": 14.23076923076923,
+      "grad_norm": 0.4161419876130332,
+      "learning_rate": 3.237434340521789e-08,
+      "loss": 0.0468,
+      "step": 1665
+    },
+    {
+      "epoch": 14.239316239316238,
+      "grad_norm": 0.5095699132041175,
+      "learning_rate": 3.166042432722671e-08,
+      "loss": 0.0964,
+      "step": 1666
+    },
+    {
+      "epoch": 14.247863247863247,
+      "grad_norm": 0.286414200855273,
+      "learning_rate": 3.095441477225347e-08,
+      "loss": 0.0246,
+      "step": 1667
+    },
+    {
+      "epoch": 14.256410256410255,
+      "grad_norm": 0.41134110401011575,
+      "learning_rate": 3.025631700262877e-08,
+      "loss": 0.0423,
+      "step": 1668
+    },
+    {
+      "epoch": 14.264957264957266,
+      "grad_norm": 0.4134548650175047,
+      "learning_rate": 2.9566133255329864e-08,
+      "loss": 0.0493,
+      "step": 1669
+    },
+    {
+      "epoch": 14.273504273504274,
+      "grad_norm": 0.38804088243316787,
+      "learning_rate": 2.888386574197488e-08,
+      "loss": 0.0432,
+      "step": 1670
+    },
+    {
+      "epoch": 14.282051282051283,
+      "grad_norm": 0.11441142627700734,
+      "learning_rate": 2.8209516648814996e-08,
+      "loss": 0.0023,
+      "step": 1671
+    },
+    {
+      "epoch": 14.290598290598291,
+      "grad_norm": 0.7652757240500461,
+      "learning_rate": 2.7543088136727792e-08,
+      "loss": 0.0667,
+      "step": 1672
+    },
+    {
+      "epoch": 14.2991452991453,
+      "grad_norm": 0.2044348005003793,
+      "learning_rate": 2.688458234121033e-08,
+      "loss": 0.0069,
+      "step": 1673
+    },
+    {
+      "epoch": 14.307692307692308,
+      "grad_norm": 0.5310235005102,
+      "learning_rate": 2.6234001372372196e-08,
+      "loss": 0.0796,
+      "step": 1674
+    },
+    {
+      "epoch": 14.316239316239317,
+      "grad_norm": 0.34314194974327356,
+      "learning_rate": 2.5591347314928572e-08,
+      "loss": 0.0364,
+      "step": 1675
+    },
+    {
+      "epoch": 14.324786324786325,
+      "grad_norm": 0.5649921942906795,
+      "learning_rate": 2.495662222819356e-08,
+      "loss": 0.0804,
+      "step": 1676
+    },
+    {
+      "epoch": 14.333333333333334,
+      "grad_norm": 0.3315299484835121,
+      "learning_rate": 2.4329828146074096e-08,
+      "loss": 0.0186,
+      "step": 1677
+    },
+    {
+      "epoch": 14.341880341880342,
+      "grad_norm": 0.38110887495178786,
+      "learning_rate": 2.3710967077063275e-08,
+      "loss": 0.0458,
+      "step": 1678
+    },
+    {
+      "epoch": 14.350427350427351,
+      "grad_norm": 0.3064729882502098,
+      "learning_rate": 2.310004100423313e-08,
+      "loss": 0.0137,
+      "step": 1679
+    },
+    {
+      "epoch": 14.35897435897436,
+      "grad_norm": 0.4986358072772935,
+      "learning_rate": 2.2497051885228825e-08,
+      "loss": 0.0603,
+      "step": 1680
+    },
+    {
+      "epoch": 14.367521367521368,
+      "grad_norm": 0.22672327025207323,
+      "learning_rate": 2.190200165226336e-08,
+      "loss": 0.0097,
+      "step": 1681
+    },
+    {
+      "epoch": 14.376068376068377,
+      "grad_norm": 0.4968406611382955,
+      "learning_rate": 2.131489221210953e-08,
+      "loss": 0.0867,
+      "step": 1682
+    },
+    {
+      "epoch": 14.384615384615385,
+      "grad_norm": 0.3838968595965074,
+      "learning_rate": 2.0735725446094924e-08,
+      "loss": 0.0387,
+      "step": 1683
+    },
+    {
+      "epoch": 14.393162393162394,
+      "grad_norm": 0.3404780059832446,
+      "learning_rate": 2.016450321009611e-08,
+      "loss": 0.028,
+      "step": 1684
+    },
+    {
+      "epoch": 14.401709401709402,
+      "grad_norm": 0.3873785523798569,
+      "learning_rate": 1.9601227334531958e-08,
+      "loss": 0.0353,
+      "step": 1685
+    },
+    {
+      "epoch": 14.41025641025641,
+      "grad_norm": 0.48467481839954696,
+      "learning_rate": 1.904589962435782e-08,
+      "loss": 0.0553,
+      "step": 1686
+    },
+    {
+      "epoch": 14.418803418803419,
+      "grad_norm": 0.4249363956567172,
+      "learning_rate": 1.8498521859060814e-08,
+      "loss": 0.0614,
+      "step": 1687
+    },
+    {
+      "epoch": 14.427350427350428,
+      "grad_norm": 0.5422143620874084,
+      "learning_rate": 1.795909579265259e-08,
+      "loss": 0.0819,
+      "step": 1688
+    },
+    {
+      "epoch": 14.435897435897436,
+      "grad_norm": 0.31554573773900985,
+      "learning_rate": 1.7427623153664364e-08,
+      "loss": 0.0199,
+      "step": 1689
+    },
+    {
+      "epoch": 14.444444444444445,
+      "grad_norm": 0.5292406231146032,
+      "learning_rate": 1.6904105645142443e-08,
+      "loss": 0.0642,
+      "step": 1690
+    },
+    {
+      "epoch": 14.452991452991453,
+      "grad_norm": 0.33194448347383515,
+      "learning_rate": 1.638854494464104e-08,
+      "loss": 0.0441,
+      "step": 1691
+    },
+    {
+      "epoch": 14.461538461538462,
+      "grad_norm": 0.484854828866342,
+      "learning_rate": 1.5880942704217528e-08,
+      "loss": 0.0496,
+      "step": 1692
+    },
+    {
+      "epoch": 14.47008547008547,
+      "grad_norm": 0.43023899300400237,
+      "learning_rate": 1.5381300550427748e-08,
+      "loss": 0.0499,
+      "step": 1693
+    },
+    {
+      "epoch": 14.478632478632479,
+      "grad_norm": 0.5504551057502816,
+      "learning_rate": 1.4889620084319878e-08,
+      "loss": 0.0711,
+      "step": 1694
+    },
+    {
+      "epoch": 14.487179487179487,
+      "grad_norm": 0.17208942331996188,
+      "learning_rate": 1.4405902881430289e-08,
+      "loss": 0.0062,
+      "step": 1695
+    },
+    {
+      "epoch": 14.495726495726496,
+      "grad_norm": 0.4286125913530637,
+      "learning_rate": 1.393015049177715e-08,
+      "loss": 0.0502,
+      "step": 1696
+    },
+    {
+      "epoch": 14.504273504273504,
+      "grad_norm": 0.4889904434508282,
+      "learning_rate": 1.3462364439857379e-08,
+      "loss": 0.0571,
+      "step": 1697
+    },
+    {
+      "epoch": 14.512820512820513,
+      "grad_norm": 0.4541729938104265,
+      "learning_rate": 1.3002546224639146e-08,
+      "loss": 0.0438,
+      "step": 1698
+    },
+    {
+      "epoch": 14.521367521367521,
+      "grad_norm": 0.638295519083425,
+      "learning_rate": 1.2550697319560211e-08,
+      "loss": 0.1397,
+      "step": 1699
+    },
+    {
+      "epoch": 14.52991452991453,
+      "grad_norm": 0.4341569849655031,
+      "learning_rate": 1.2106819172520434e-08,
+      "loss": 0.0517,
+      "step": 1700
+    },
+    {
+      "epoch": 14.538461538461538,
+      "grad_norm": 0.4068406338564505,
+      "learning_rate": 1.1670913205878431e-08,
+      "loss": 0.0718,
+      "step": 1701
+    },
+    {
+      "epoch": 14.547008547008547,
+      "grad_norm": 0.32365525965014147,
+      "learning_rate": 1.1242980816447147e-08,
+      "loss": 0.0196,
+      "step": 1702
+    },
+    {
+      "epoch": 14.555555555555555,
+      "grad_norm": 0.19470073348110065,
+      "learning_rate": 1.0823023375489128e-08,
+      "loss": 0.0076,
+      "step": 1703
+    },
+    {
+      "epoch": 14.564102564102564,
+      "grad_norm": 0.11448802274939457,
+      "learning_rate": 1.0411042228711254e-08,
+      "loss": 0.0028,
+      "step": 1704
+    },
+    {
+      "epoch": 14.572649572649572,
+      "grad_norm": 0.6691841832251686,
+      "learning_rate": 1.0007038696262517e-08,
+      "loss": 0.0916,
+      "step": 1705
+    },
+    {
+      "epoch": 14.581196581196581,
+      "grad_norm": 0.28586370985153986,
+      "learning_rate": 9.611014072727354e-09,
+      "loss": 0.0144,
+      "step": 1706
+    },
+    {
+      "epoch": 14.58974358974359,
+      "grad_norm": 0.5898421301273966,
+      "learning_rate": 9.222969627123435e-09,
+      "loss": 0.0968,
+      "step": 1707
+    },
+    {
+      "epoch": 14.598290598290598,
+      "grad_norm": 0.16072954161372915,
+      "learning_rate": 8.842906602896661e-09,
+      "loss": 0.005,
+      "step": 1708
+    },
+    {
+      "epoch": 14.606837606837606,
+      "grad_norm": 0.2759661423413884,
+      "learning_rate": 8.470826217917006e-09,
+      "loss": 0.0215,
+      "step": 1709
+    },
+    {
+      "epoch": 14.615384615384615,
+      "grad_norm": 0.6260797976325166,
+      "learning_rate": 8.106729664475178e-09,
+      "loss": 0.1173,
+      "step": 1710
+    },
+    {
+      "epoch": 14.623931623931623,
+      "grad_norm": 0.31248553543290747,
+      "learning_rate": 7.750618109278464e-09,
+      "loss": 0.0175,
+      "step": 1711
+    },
+    {
+      "epoch": 14.632478632478632,
+      "grad_norm": 0.27412422037840384,
+      "learning_rate": 7.402492693447671e-09,
+      "loss": 0.0158,
+      "step": 1712
+    },
+    {
+      "epoch": 14.64102564102564,
+      "grad_norm": 0.435526459964807,
+      "learning_rate": 7.062354532512416e-09,
+      "loss": 0.0313,
+      "step": 1713
+    },
+    {
+      "epoch": 14.649572649572649,
+      "grad_norm": 0.5786644177120002,
+      "learning_rate": 6.730204716407507e-09,
+      "loss": 0.0898,
+      "step": 1714
+    },
+    {
+      "epoch": 14.658119658119658,
+      "grad_norm": 0.19256453915909177,
+      "learning_rate": 6.406044309471004e-09,
+      "loss": 0.0083,
+      "step": 1715
+    },
+    {
+      "epoch": 14.666666666666666,
+      "grad_norm": 0.375666393092363,
+      "learning_rate": 6.089874350439507e-09,
+      "loss": 0.0299,
+      "step": 1716
+    },
+    {
+      "epoch": 14.675213675213675,
+      "grad_norm": 0.5319710614208967,
+      "learning_rate": 5.781695852444258e-09,
+      "loss": 0.0983,
+      "step": 1717
+    },
+    {
+      "epoch": 14.683760683760683,
+      "grad_norm": 0.35482923106684494,
+      "learning_rate": 5.481509803009766e-09,
+      "loss": 0.0305,
+      "step": 1718
+    },
+    {
+      "epoch": 14.692307692307692,
+      "grad_norm": 0.19746564564826924,
+      "learning_rate": 5.189317164049634e-09,
+      "loss": 0.0103,
+      "step": 1719
+    },
+    {
+      "epoch": 14.7008547008547,
+      "grad_norm": 0.4282057813149622,
+      "learning_rate": 4.905118871862402e-09,
+      "loss": 0.0368,
+      "step": 1720
+    },
+    {
+      "epoch": 14.709401709401709,
+      "grad_norm": 0.07684527835039032,
+      "learning_rate": 4.62891583713071e-09,
+      "loss": 0.002,
+      "step": 1721
+    },
+    {
+      "epoch": 14.717948717948717,
+      "grad_norm": 0.5770419703333539,
+      "learning_rate": 4.3607089449165806e-09,
+      "loss": 0.1029,
+      "step": 1722
+    },
+    {
+      "epoch": 14.726495726495726,
+      "grad_norm": 0.3026257207483837,
+      "learning_rate": 4.100499054659757e-09,
+      "loss": 0.0175,
+      "step": 1723
+    },
+    {
+      "epoch": 14.735042735042736,
+      "grad_norm": 0.43807229538708947,
+      "learning_rate": 3.848287000174089e-09,
+      "loss": 0.0409,
+      "step": 1724
+    },
+    {
+      "epoch": 14.743589743589745,
+      "grad_norm": 0.32564952319806223,
+      "learning_rate": 3.6040735896455957e-09,
+      "loss": 0.0189,
+      "step": 1725
+    },
+    {
+      "epoch": 14.752136752136753,
+      "grad_norm": 0.42691653228759036,
+      "learning_rate": 3.367859605628854e-09,
+      "loss": 0.0333,
+      "step": 1726
+    },
+    {
+      "epoch": 14.760683760683762,
+      "grad_norm": 0.32786135774227537,
+      "learning_rate": 3.139645805046165e-09,
+      "loss": 0.0214,
+      "step": 1727
+    },
+    {
+      "epoch": 14.76923076923077,
+      "grad_norm": 0.2744239467920148,
+      "learning_rate": 2.919432919183396e-09,
+      "loss": 0.0131,
+      "step": 1728
+    },
+    {
+      "epoch": 14.777777777777779,
+      "grad_norm": 0.30419227440175195,
+      "learning_rate": 2.7072216536885855e-09,
+      "loss": 0.0161,
+      "step": 1729
+    },
+    {
+      "epoch": 14.786324786324787,
+      "grad_norm": 0.6233695528799661,
+      "learning_rate": 2.5030126885694505e-09,
+      "loss": 0.0987,
+      "step": 1730
+    },
+    {
+      "epoch": 14.794871794871796,
+      "grad_norm": 0.29663516180844607,
+      "learning_rate": 2.3068066781908873e-09,
+      "loss": 0.023,
+      "step": 1731
+    },
+    {
+      "epoch": 14.803418803418804,
+      "grad_norm": 0.32598395103645494,
+      "learning_rate": 2.118604251273859e-09,
+      "loss": 0.0308,
+      "step": 1732
+    },
+    {
+      "epoch": 14.811965811965813,
+      "grad_norm": 0.6182730441581095,
+      "learning_rate": 1.9384060108923463e-09,
+      "loss": 0.0955,
+      "step": 1733
+    },
+    {
+      "epoch": 14.820512820512821,
+      "grad_norm": 0.1185843301282404,
+      "learning_rate": 1.766212534471401e-09,
+      "loss": 0.0035,
+      "step": 1734
+    },
+    {
+      "epoch": 14.82905982905983,
+      "grad_norm": 0.5723507056543405,
+      "learning_rate": 1.6020243737865926e-09,
+      "loss": 0.0876,
+      "step": 1735
+    },
+    {
+      "epoch": 14.837606837606838,
+      "grad_norm": 0.2061314944219676,
+      "learning_rate": 1.4458420549606777e-09,
+      "loss": 0.0154,
+      "step": 1736
+    },
+    {
+      "epoch": 14.846153846153847,
+      "grad_norm": 0.6249720933757367,
+      "learning_rate": 1.297666078462767e-09,
+      "loss": 0.1111,
+      "step": 1737
+    },
+    {
+      "epoch": 14.854700854700855,
+      "grad_norm": 0.2860798677708746,
+      "learning_rate": 1.1574969191061047e-09,
+      "loss": 0.0237,
+      "step": 1738
+    },
+    {
+      "epoch": 14.863247863247864,
+      "grad_norm": 0.280044513165246,
+      "learning_rate": 1.0253350260480688e-09,
+      "loss": 0.0228,
+      "step": 1739
+    },
+    {
+      "epoch": 14.871794871794872,
+      "grad_norm": 0.4494241840830116,
+      "learning_rate": 9.011808227865626e-10,
+      "loss": 0.0508,
+      "step": 1740
+    },
+    {
+      "epoch": 14.88034188034188,
+      "grad_norm": 0.35579500426583943,
+      "learning_rate": 7.850347071597376e-10,
+      "loss": 0.0236,
+      "step": 1741
+    },
+    {
+      "epoch": 14.88888888888889,
+      "grad_norm": 0.13002283698556336,
+      "learning_rate": 6.768970513457151e-10,
+      "loss": 0.004,
+      "step": 1742
+    },
+    {
+      "epoch": 14.897435897435898,
+      "grad_norm": 0.3928491595882552,
+      "learning_rate": 5.767682018595344e-10,
+      "loss": 0.0294,
+      "step": 1743
+    },
+    {
+      "epoch": 14.905982905982906,
+      "grad_norm": 0.31808212261880264,
+      "learning_rate": 4.846484795528739e-10,
+      "loss": 0.0263,
+      "step": 1744
+    },
+    {
+      "epoch": 14.914529914529915,
+      "grad_norm": 0.27391217697985504,
+      "learning_rate": 4.0053817961321903e-10,
+      "loss": 0.0188,
+      "step": 1745
+    },
+    {
+      "epoch": 14.923076923076923,
+      "grad_norm": 0.43076628005985973,
+      "learning_rate": 3.2443757156330746e-10,
+      "loss": 0.0529,
+      "step": 1746
+    },
+    {
+      "epoch": 14.931623931623932,
+      "grad_norm": 0.5306801307778587,
+      "learning_rate": 2.563468992586304e-10,
+      "loss": 0.0675,
+      "step": 1747
+    },
+    {
+      "epoch": 14.94017094017094,
+      "grad_norm": 0.3592276721288456,
+      "learning_rate": 1.9626638088854344e-10,
+      "loss": 0.0273,
+      "step": 1748
+    },
+    {
+      "epoch": 14.948717948717949,
+      "grad_norm": 0.11763294528593544,
+      "learning_rate": 1.4419620897432318e-10,
+      "loss": 0.0034,
+      "step": 1749
+    },
+    {
+      "epoch": 14.957264957264957,
+      "grad_norm": 0.44419236728112543,
+      "learning_rate": 1.0013655036916758e-10,
+      "loss": 0.0876,
+      "step": 1750
+    },
+    {
+      "epoch": 14.965811965811966,
+      "grad_norm": 0.06513844509503423,
+      "learning_rate": 6.408754625736313e-11,
+      "loss": 0.0013,
+      "step": 1751
+    },
+    {
+      "epoch": 14.974358974358974,
+      "grad_norm": 0.32609380727131143,
+      "learning_rate": 3.604931215400731e-11,
+      "loss": 0.0196,
+      "step": 1752
+    },
+    {
+      "epoch": 14.982905982905983,
+      "grad_norm": 0.5441264595409655,
+      "learning_rate": 1.6021937904731054e-11,
+      "loss": 0.0642,
+      "step": 1753
+    },
+    {
+      "epoch": 14.991452991452991,
+      "grad_norm": 0.21315543474359916,
+      "learning_rate": 4.005487684866083e-12,
+      "loss": 0.0091,
+      "step": 1754
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.38663943234097126,
+      "learning_rate": 0.0,
+      "loss": 0.034,
+      "step": 1755
+    },
+    {
+      "epoch": 15.0,
+      "step": 1755,
+      "total_flos": 170425732071424.0,
+      "train_loss": 0.2854139506731957,
+      "train_runtime": 13259.2544,
+      "train_samples_per_second": 0.924,
+      "train_steps_per_second": 0.132
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1755,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 170425732071424.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}