diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,56553 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.9968971081047537,
+  "eval_steps": 504,
+  "global_step": 8056,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004964627032394191,
+      "grad_norm": 0.2896804752195731,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.6569,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004964627032394191,
+      "eval_loss": 0.6948701739311218,
+      "eval_runtime": 258.3281,
+      "eval_samples_per_second": 117.498,
+      "eval_steps_per_second": 14.691,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009929254064788382,
+      "grad_norm": 0.3050018791891007,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6721,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014893881097182574,
+      "grad_norm": 0.29882980651965635,
+      "learning_rate": 1.5e-06,
+      "loss": 0.681,
+      "step": 3
+    },
+    {
+      "epoch": 0.0019858508129576764,
+      "grad_norm": 0.2769455424447693,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7166,
+      "step": 4
+    },
+    {
+      "epoch": 0.002482313516197096,
+      "grad_norm": 0.28865369141827507,
+      "learning_rate": 2.5e-06,
+      "loss": 0.7363,
+      "step": 5
+    },
+    {
+      "epoch": 0.002978776219436515,
+      "grad_norm": 0.2722576701771127,
+      "learning_rate": 3e-06,
+      "loss": 0.7034,
+      "step": 6
+    },
+    {
+      "epoch": 0.0034752389226759338,
+      "grad_norm": 0.27760553070368027,
+      "learning_rate": 3.5e-06,
+      "loss": 0.7189,
+      "step": 7
+    },
+    {
+      "epoch": 0.003971701625915353,
+      "grad_norm": 0.2560906764365523,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6724,
+      "step": 8
+    },
+    {
+      "epoch": 0.004468164329154772,
+      "grad_norm": 0.24966702280304454,
+      "learning_rate": 4.5e-06,
+      "loss": 0.7125,
+      "step": 9
+    },
+    {
+      "epoch": 0.004964627032394192,
+      "grad_norm": 0.20972188158050864,
+      "learning_rate": 5e-06,
+      "loss": 0.6863,
+      "step": 10
+    },
+    {
+      "epoch": 0.00546108973563361,
+      "grad_norm": 0.2008074817888515,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.6582,
+      "step": 11
+    },
+    {
+      "epoch": 0.00595755243887303,
+      "grad_norm": 0.22832944172310443,
+      "learning_rate": 6e-06,
+      "loss": 0.642,
+      "step": 12
+    },
+    {
+      "epoch": 0.006454015142112449,
+      "grad_norm": 0.2459156254861726,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.671,
+      "step": 13
+    },
+    {
+      "epoch": 0.0069504778453518675,
+      "grad_norm": 0.31687300308761013,
+      "learning_rate": 7e-06,
+      "loss": 0.6689,
+      "step": 14
+    },
+    {
+      "epoch": 0.007446940548591287,
+      "grad_norm": 0.20119950815160606,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6279,
+      "step": 15
+    },
+    {
+      "epoch": 0.007943403251830706,
+      "grad_norm": 0.21772431498741993,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6732,
+      "step": 16
+    },
+    {
+      "epoch": 0.008439865955070125,
+      "grad_norm": 0.17454628482212584,
+      "learning_rate": 8.5e-06,
+      "loss": 0.614,
+      "step": 17
+    },
+    {
+      "epoch": 0.008936328658309544,
+      "grad_norm": 0.1674880235536819,
+      "learning_rate": 9e-06,
+      "loss": 0.6335,
+      "step": 18
+    },
+    {
+      "epoch": 0.009432791361548964,
+      "grad_norm": 0.18488918178360303,
+      "learning_rate": 9.5e-06,
+      "loss": 0.6027,
+      "step": 19
+    },
+    {
+      "epoch": 0.009929254064788383,
+      "grad_norm": 0.24570525549058408,
+      "learning_rate": 1e-05,
+      "loss": 0.621,
+      "step": 20
+    },
+    {
+      "epoch": 0.010425716768027803,
+      "grad_norm": 0.15022107044829835,
+      "learning_rate": 9.999999617915087e-06,
+      "loss": 0.6039,
+      "step": 21
+    },
+    {
+      "epoch": 0.01092217947126722,
+      "grad_norm": 0.16517798078299958,
+      "learning_rate": 9.999998471660397e-06,
+      "loss": 0.6237,
+      "step": 22
+    },
+    {
+      "epoch": 0.01141864217450664,
+      "grad_norm": 0.19491836617556169,
+      "learning_rate": 9.999996561236111e-06,
+      "loss": 0.6763,
+      "step": 23
+    },
+    {
+      "epoch": 0.01191510487774606,
+      "grad_norm": 0.15511366832327092,
+      "learning_rate": 9.99999388664252e-06,
+      "loss": 0.6079,
+      "step": 24
+    },
+    {
+      "epoch": 0.012411567580985479,
+      "grad_norm": 0.14847965517533848,
+      "learning_rate": 9.999990447880033e-06,
+      "loss": 0.6019,
+      "step": 25
+    },
+    {
+      "epoch": 0.012908030284224898,
+      "grad_norm": 0.1517877987340116,
+      "learning_rate": 9.999986244949173e-06,
+      "loss": 0.5898,
+      "step": 26
+    },
+    {
+      "epoch": 0.013404492987464317,
+      "grad_norm": 0.1467811827869313,
+      "learning_rate": 9.999981277850585e-06,
+      "loss": 0.6418,
+      "step": 27
+    },
+    {
+      "epoch": 0.013900955690703735,
+      "grad_norm": 0.14177217518216353,
+      "learning_rate": 9.999975546585027e-06,
+      "loss": 0.653,
+      "step": 28
+    },
+    {
+      "epoch": 0.014397418393943155,
+      "grad_norm": 0.14013715426768192,
+      "learning_rate": 9.999969051153376e-06,
+      "loss": 0.6336,
+      "step": 29
+    },
+    {
+      "epoch": 0.014893881097182574,
+      "grad_norm": 0.13838564452582985,
+      "learning_rate": 9.999961791556623e-06,
+      "loss": 0.6115,
+      "step": 30
+    },
+    {
+      "epoch": 0.015390343800421993,
+      "grad_norm": 0.14196459752762547,
+      "learning_rate": 9.999953767795879e-06,
+      "loss": 0.6651,
+      "step": 31
+    },
+    {
+      "epoch": 0.01588680650366141,
+      "grad_norm": 0.13929013135958662,
+      "learning_rate": 9.99994497987237e-06,
+      "loss": 0.6196,
+      "step": 32
+    },
+    {
+      "epoch": 0.016383269206900832,
+      "grad_norm": 0.1377890217750465,
+      "learning_rate": 9.999935427787437e-06,
+      "loss": 0.6128,
+      "step": 33
+    },
+    {
+      "epoch": 0.01687973191014025,
+      "grad_norm": 0.13403481454155775,
+      "learning_rate": 9.999925111542544e-06,
+      "loss": 0.6287,
+      "step": 34
+    },
+    {
+      "epoch": 0.01737619461337967,
+      "grad_norm": 0.13706676440981166,
+      "learning_rate": 9.999914031139264e-06,
+      "loss": 0.5834,
+      "step": 35
+    },
+    {
+      "epoch": 0.01787265731661909,
+      "grad_norm": 0.13655131236247117,
+      "learning_rate": 9.99990218657929e-06,
+      "loss": 0.5886,
+      "step": 36
+    },
+    {
+      "epoch": 0.018369120019858506,
+      "grad_norm": 0.13692053883385027,
+      "learning_rate": 9.999889577864439e-06,
+      "loss": 0.6128,
+      "step": 37
+    },
+    {
+      "epoch": 0.018865582723097928,
+      "grad_norm": 0.1284073363888045,
+      "learning_rate": 9.99987620499663e-06,
+      "loss": 0.635,
+      "step": 38
+    },
+    {
+      "epoch": 0.019362045426337345,
+      "grad_norm": 0.13517475428298725,
+      "learning_rate": 9.999862067977911e-06,
+      "loss": 0.6698,
+      "step": 39
+    },
+    {
+      "epoch": 0.019858508129576766,
+      "grad_norm": 0.1355982640698863,
+      "learning_rate": 9.999847166810441e-06,
+      "loss": 0.6236,
+      "step": 40
+    },
+    {
+      "epoch": 0.020354970832816184,
+      "grad_norm": 0.12984313043454054,
+      "learning_rate": 9.999831501496497e-06,
+      "loss": 0.5833,
+      "step": 41
+    },
+    {
+      "epoch": 0.020851433536055605,
+      "grad_norm": 0.1383145715551888,
+      "learning_rate": 9.999815072038476e-06,
+      "loss": 0.5906,
+      "step": 42
+    },
+    {
+      "epoch": 0.021347896239295023,
+      "grad_norm": 0.1288047804494723,
+      "learning_rate": 9.999797878438886e-06,
+      "loss": 0.6161,
+      "step": 43
+    },
+    {
+      "epoch": 0.02184435894253444,
+      "grad_norm": 0.13478081351814658,
+      "learning_rate": 9.999779920700358e-06,
+      "loss": 0.5896,
+      "step": 44
+    },
+    {
+      "epoch": 0.022340821645773862,
+      "grad_norm": 0.13613809579000827,
+      "learning_rate": 9.999761198825633e-06,
+      "loss": 0.6314,
+      "step": 45
+    },
+    {
+      "epoch": 0.02283728434901328,
+      "grad_norm": 0.13321658898143937,
+      "learning_rate": 9.999741712817574e-06,
+      "loss": 0.6312,
+      "step": 46
+    },
+    {
+      "epoch": 0.0233337470522527,
+      "grad_norm": 0.13603038062117492,
+      "learning_rate": 9.999721462679158e-06,
+      "loss": 0.5961,
+      "step": 47
+    },
+    {
+      "epoch": 0.02383020975549212,
+      "grad_norm": 0.13675851643279544,
+      "learning_rate": 9.999700448413483e-06,
+      "loss": 0.6154,
+      "step": 48
+    },
+    {
+      "epoch": 0.02432667245873154,
+      "grad_norm": 0.13729938609740347,
+      "learning_rate": 9.999678670023756e-06,
+      "loss": 0.6242,
+      "step": 49
+    },
+    {
+      "epoch": 0.024823135161970957,
+      "grad_norm": 0.13441177177690306,
+      "learning_rate": 9.99965612751331e-06,
+      "loss": 0.6007,
+      "step": 50
+    },
+    {
+      "epoch": 0.025319597865210375,
+      "grad_norm": 0.13547565356560712,
+      "learning_rate": 9.999632820885588e-06,
+      "loss": 0.6411,
+      "step": 51
+    },
+    {
+      "epoch": 0.025816060568449796,
+      "grad_norm": 0.1324008772576428,
+      "learning_rate": 9.999608750144152e-06,
+      "loss": 0.6177,
+      "step": 52
+    },
+    {
+      "epoch": 0.026312523271689214,
+      "grad_norm": 0.13979089573869022,
+      "learning_rate": 9.999583915292681e-06,
+      "loss": 0.5949,
+      "step": 53
+    },
+    {
+      "epoch": 0.026808985974928635,
+      "grad_norm": 0.13859893177899527,
+      "learning_rate": 9.999558316334971e-06,
+      "loss": 0.6541,
+      "step": 54
+    },
+    {
+      "epoch": 0.027305448678168052,
+      "grad_norm": 0.13635221865975883,
+      "learning_rate": 9.999531953274934e-06,
+      "loss": 0.6091,
+      "step": 55
+    },
+    {
+      "epoch": 0.02780191138140747,
+      "grad_norm": 0.1282420474503579,
+      "learning_rate": 9.9995048261166e-06,
+      "loss": 0.5882,
+      "step": 56
+    },
+    {
+      "epoch": 0.02829837408464689,
+      "grad_norm": 0.13401369032020452,
+      "learning_rate": 9.999476934864113e-06,
+      "loss": 0.5938,
+      "step": 57
+    },
+    {
+      "epoch": 0.02879483678788631,
+      "grad_norm": 0.13628126334123536,
+      "learning_rate": 9.999448279521737e-06,
+      "loss": 0.5985,
+      "step": 58
+    },
+    {
+      "epoch": 0.02929129949112573,
+      "grad_norm": 0.1341361184774213,
+      "learning_rate": 9.999418860093852e-06,
+      "loss": 0.5719,
+      "step": 59
+    },
+    {
+      "epoch": 0.029787762194365148,
+      "grad_norm": 0.13786027891551675,
+      "learning_rate": 9.999388676584956e-06,
+      "loss": 0.6298,
+      "step": 60
+    },
+    {
+      "epoch": 0.03028422489760457,
+      "grad_norm": 0.1286588798863049,
+      "learning_rate": 9.999357728999657e-06,
+      "loss": 0.5584,
+      "step": 61
+    },
+    {
+      "epoch": 0.030780687600843987,
+      "grad_norm": 0.13054259557276518,
+      "learning_rate": 9.999326017342688e-06,
+      "loss": 0.6112,
+      "step": 62
+    },
+    {
+      "epoch": 0.03127715030408341,
+      "grad_norm": 0.12726900077110587,
+      "learning_rate": 9.999293541618898e-06,
+      "loss": 0.6069,
+      "step": 63
+    },
+    {
+      "epoch": 0.03177361300732282,
+      "grad_norm": 0.1284338678113,
+      "learning_rate": 9.999260301833245e-06,
+      "loss": 0.5741,
+      "step": 64
+    },
+    {
+      "epoch": 0.03227007571056224,
+      "grad_norm": 0.13254345170506723,
+      "learning_rate": 9.999226297990812e-06,
+      "loss": 0.6091,
+      "step": 65
+    },
+    {
+      "epoch": 0.032766538413801664,
+      "grad_norm": 0.12810965726821946,
+      "learning_rate": 9.999191530096798e-06,
+      "loss": 0.591,
+      "step": 66
+    },
+    {
+      "epoch": 0.033263001117041086,
+      "grad_norm": 0.13053684029880475,
+      "learning_rate": 9.999155998156511e-06,
+      "loss": 0.6185,
+      "step": 67
+    },
+    {
+      "epoch": 0.0337594638202805,
+      "grad_norm": 0.1249000403363421,
+      "learning_rate": 9.999119702175388e-06,
+      "loss": 0.6039,
+      "step": 68
+    },
+    {
+      "epoch": 0.03425592652351992,
+      "grad_norm": 0.13159419524811636,
+      "learning_rate": 9.999082642158972e-06,
+      "loss": 0.6349,
+      "step": 69
+    },
+    {
+      "epoch": 0.03475238922675934,
+      "grad_norm": 0.12994532055439553,
+      "learning_rate": 9.999044818112929e-06,
+      "loss": 0.5775,
+      "step": 70
+    },
+    {
+      "epoch": 0.035248851929998756,
+      "grad_norm": 0.12935935865475864,
+      "learning_rate": 9.999006230043039e-06,
+      "loss": 0.5856,
+      "step": 71
+    },
+    {
+      "epoch": 0.03574531463323818,
+      "grad_norm": 0.12636961689532966,
+      "learning_rate": 9.9989668779552e-06,
+      "loss": 0.5693,
+      "step": 72
+    },
+    {
+      "epoch": 0.0362417773364776,
+      "grad_norm": 0.1291945677437944,
+      "learning_rate": 9.998926761855425e-06,
+      "loss": 0.5856,
+      "step": 73
+    },
+    {
+      "epoch": 0.03673824003971701,
+      "grad_norm": 0.12423558807838328,
+      "learning_rate": 9.998885881749847e-06,
+      "loss": 0.6412,
+      "step": 74
+    },
+    {
+      "epoch": 0.037234702742956434,
+      "grad_norm": 0.11838472293339844,
+      "learning_rate": 9.998844237644714e-06,
+      "loss": 0.5923,
+      "step": 75
+    },
+    {
+      "epoch": 0.037731165446195855,
+      "grad_norm": 0.12070342951020165,
+      "learning_rate": 9.998801829546387e-06,
+      "loss": 0.6009,
+      "step": 76
+    },
+    {
+      "epoch": 0.038227628149435276,
+      "grad_norm": 0.12133413872988458,
+      "learning_rate": 9.998758657461353e-06,
+      "loss": 0.5928,
+      "step": 77
+    },
+    {
+      "epoch": 0.03872409085267469,
+      "grad_norm": 0.11847302132399616,
+      "learning_rate": 9.998714721396206e-06,
+      "loss": 0.5873,
+      "step": 78
+    },
+    {
+      "epoch": 0.03922055355591411,
+      "grad_norm": 0.11586463974187308,
+      "learning_rate": 9.998670021357662e-06,
+      "loss": 0.599,
+      "step": 79
+    },
+    {
+      "epoch": 0.03971701625915353,
+      "grad_norm": 0.11611971202617413,
+      "learning_rate": 9.998624557352552e-06,
+      "loss": 0.5925,
+      "step": 80
+    },
+    {
+      "epoch": 0.04021347896239295,
+      "grad_norm": 0.11063709969348022,
+      "learning_rate": 9.998578329387826e-06,
+      "loss": 0.5651,
+      "step": 81
+    },
+    {
+      "epoch": 0.04070994166563237,
+      "grad_norm": 0.11228145005616408,
+      "learning_rate": 9.998531337470548e-06,
+      "loss": 0.5596,
+      "step": 82
+    },
+    {
+      "epoch": 0.04120640436887179,
+      "grad_norm": 0.10686949876281465,
+      "learning_rate": 9.9984835816079e-06,
+      "loss": 0.5968,
+      "step": 83
+    },
+    {
+      "epoch": 0.04170286707211121,
+      "grad_norm": 0.10654597088862137,
+      "learning_rate": 9.998435061807184e-06,
+      "loss": 0.5967,
+      "step": 84
+    },
+    {
+      "epoch": 0.042199329775350625,
+      "grad_norm": 0.10033719891548469,
+      "learning_rate": 9.99838577807581e-06,
+      "loss": 0.5501,
+      "step": 85
+    },
+    {
+      "epoch": 0.042695792478590046,
+      "grad_norm": 0.10549280001704904,
+      "learning_rate": 9.998335730421313e-06,
+      "loss": 0.5786,
+      "step": 86
+    },
+    {
+      "epoch": 0.04319225518182947,
+      "grad_norm": 0.102951502668565,
+      "learning_rate": 9.998284918851343e-06,
+      "loss": 0.5747,
+      "step": 87
+    },
+    {
+      "epoch": 0.04368871788506888,
+      "grad_norm": 0.09770170164060958,
+      "learning_rate": 9.998233343373664e-06,
+      "loss": 0.5826,
+      "step": 88
+    },
+    {
+      "epoch": 0.0441851805883083,
+      "grad_norm": 0.0958977534181121,
+      "learning_rate": 9.998181003996159e-06,
+      "loss": 0.6037,
+      "step": 89
+    },
+    {
+      "epoch": 0.044681643291547724,
+      "grad_norm": 0.0920383420000729,
+      "learning_rate": 9.998127900726825e-06,
+      "loss": 0.5708,
+      "step": 90
+    },
+    {
+      "epoch": 0.045178105994787145,
+      "grad_norm": 0.093747232922122,
+      "learning_rate": 9.998074033573783e-06,
+      "loss": 0.6067,
+      "step": 91
+    },
+    {
+      "epoch": 0.04567456869802656,
+      "grad_norm": 0.08851388967347672,
+      "learning_rate": 9.998019402545264e-06,
+      "loss": 0.5706,
+      "step": 92
+    },
+    {
+      "epoch": 0.04617103140126598,
+      "grad_norm": 0.0867258581115883,
+      "learning_rate": 9.997964007649614e-06,
+      "loss": 0.5673,
+      "step": 93
+    },
+    {
+      "epoch": 0.0466674941045054,
+      "grad_norm": 0.08589006292805439,
+      "learning_rate": 9.997907848895304e-06,
+      "loss": 0.5839,
+      "step": 94
+    },
+    {
+      "epoch": 0.047163956807744815,
+      "grad_norm": 0.08567652602207897,
+      "learning_rate": 9.997850926290912e-06,
+      "loss": 0.5604,
+      "step": 95
+    },
+    {
+      "epoch": 0.04766041951098424,
+      "grad_norm": 0.09168476487857387,
+      "learning_rate": 9.997793239845141e-06,
+      "loss": 0.6102,
+      "step": 96
+    },
+    {
+      "epoch": 0.04815688221422366,
+      "grad_norm": 0.0862684302577442,
+      "learning_rate": 9.997734789566809e-06,
+      "loss": 0.6065,
+      "step": 97
+    },
+    {
+      "epoch": 0.04865334491746308,
+      "grad_norm": 0.08721353894012582,
+      "learning_rate": 9.997675575464844e-06,
+      "loss": 0.6129,
+      "step": 98
+    },
+    {
+      "epoch": 0.04914980762070249,
+      "grad_norm": 0.08863015163099411,
+      "learning_rate": 9.997615597548302e-06,
+      "loss": 0.5802,
+      "step": 99
+    },
+    {
+      "epoch": 0.049646270323941914,
+      "grad_norm": 0.08225243222916953,
+      "learning_rate": 9.997554855826343e-06,
+      "loss": 0.599,
+      "step": 100
+    },
+    {
+      "epoch": 0.050142733027181335,
+      "grad_norm": 0.08453041407170588,
+      "learning_rate": 9.997493350308258e-06,
+      "loss": 0.6146,
+      "step": 101
+    },
+    {
+      "epoch": 0.05063919573042075,
+      "grad_norm": 0.07970489213260537,
+      "learning_rate": 9.99743108100344e-06,
+      "loss": 0.574,
+      "step": 102
+    },
+    {
+      "epoch": 0.05113565843366017,
+      "grad_norm": 0.07863283495549847,
+      "learning_rate": 9.99736804792141e-06,
+      "loss": 0.5682,
+      "step": 103
+    },
+    {
+      "epoch": 0.05163212113689959,
+      "grad_norm": 0.07752696100053419,
+      "learning_rate": 9.997304251071802e-06,
+      "loss": 0.5569,
+      "step": 104
+    },
+    {
+      "epoch": 0.052128583840139006,
+      "grad_norm": 0.13358841508181127,
+      "learning_rate": 9.997239690464362e-06,
+      "loss": 0.6168,
+      "step": 105
+    },
+    {
+      "epoch": 0.05262504654337843,
+      "grad_norm": 0.07348128316840971,
+      "learning_rate": 9.997174366108962e-06,
+      "loss": 0.5819,
+      "step": 106
+    },
+    {
+      "epoch": 0.05312150924661785,
+      "grad_norm": 0.07799188911676157,
+      "learning_rate": 9.997108278015583e-06,
+      "loss": 0.5903,
+      "step": 107
+    },
+    {
+      "epoch": 0.05361797194985727,
+      "grad_norm": 0.08319318876985128,
+      "learning_rate": 9.997041426194327e-06,
+      "loss": 0.6054,
+      "step": 108
+    },
+    {
+      "epoch": 0.054114434653096684,
+      "grad_norm": 0.07577969048637133,
+      "learning_rate": 9.996973810655409e-06,
+      "loss": 0.588,
+      "step": 109
+    },
+    {
+      "epoch": 0.054610897356336105,
+      "grad_norm": 0.07929330066652941,
+      "learning_rate": 9.996905431409165e-06,
+      "loss": 0.577,
+      "step": 110
+    },
+    {
+      "epoch": 0.055107360059575526,
+      "grad_norm": 0.07669462435166145,
+      "learning_rate": 9.996836288466046e-06,
+      "loss": 0.5774,
+      "step": 111
+    },
+    {
+      "epoch": 0.05560382276281494,
+      "grad_norm": 0.07442700190295497,
+      "learning_rate": 9.996766381836617e-06,
+      "loss": 0.5979,
+      "step": 112
+    },
+    {
+      "epoch": 0.05610028546605436,
+      "grad_norm": 0.07330916527882274,
+      "learning_rate": 9.996695711531565e-06,
+      "loss": 0.5559,
+      "step": 113
+    },
+    {
+      "epoch": 0.05659674816929378,
+      "grad_norm": 0.07665817971888089,
+      "learning_rate": 9.99662427756169e-06,
+      "loss": 0.5979,
+      "step": 114
+    },
+    {
+      "epoch": 0.057093210872533204,
+      "grad_norm": 0.08003254184842483,
+      "learning_rate": 9.996552079937907e-06,
+      "loss": 0.599,
+      "step": 115
+    },
+    {
+      "epoch": 0.05758967357577262,
+      "grad_norm": 0.07510789705619458,
+      "learning_rate": 9.996479118671255e-06,
+      "loss": 0.6122,
+      "step": 116
+    },
+    {
+      "epoch": 0.05808613627901204,
+      "grad_norm": 0.07624743707532947,
+      "learning_rate": 9.99640539377288e-06,
+      "loss": 0.6061,
+      "step": 117
+    },
+    {
+      "epoch": 0.05858259898225146,
+      "grad_norm": 0.08231933957861946,
+      "learning_rate": 9.99633090525405e-06,
+      "loss": 0.6115,
+      "step": 118
+    },
+    {
+      "epoch": 0.059079061685490875,
+      "grad_norm": 0.07413718802908581,
+      "learning_rate": 9.996255653126155e-06,
+      "loss": 0.5688,
+      "step": 119
+    },
+    {
+      "epoch": 0.059575524388730296,
+      "grad_norm": 0.0715246453925429,
+      "learning_rate": 9.996179637400689e-06,
+      "loss": 0.6037,
+      "step": 120
+    },
+    {
+      "epoch": 0.06007198709196972,
+      "grad_norm": 0.07150754449533633,
+      "learning_rate": 9.996102858089276e-06,
+      "loss": 0.5793,
+      "step": 121
+    },
+    {
+      "epoch": 0.06056844979520914,
+      "grad_norm": 0.07498273940894117,
+      "learning_rate": 9.996025315203645e-06,
+      "loss": 0.5924,
+      "step": 122
+    },
+    {
+      "epoch": 0.06106491249844855,
+      "grad_norm": 0.07613201051621966,
+      "learning_rate": 9.995947008755651e-06,
+      "loss": 0.599,
+      "step": 123
+    },
+    {
+      "epoch": 0.06156137520168797,
+      "grad_norm": 0.07403145307648665,
+      "learning_rate": 9.99586793875726e-06,
+      "loss": 0.5827,
+      "step": 124
+    },
+    {
+      "epoch": 0.062057837904927395,
+      "grad_norm": 0.07365891943371078,
+      "learning_rate": 9.99578810522056e-06,
+      "loss": 0.5593,
+      "step": 125
+    },
+    {
+      "epoch": 0.06255430060816682,
+      "grad_norm": 0.07495245460403768,
+      "learning_rate": 9.995707508157746e-06,
+      "loss": 0.5925,
+      "step": 126
+    },
+    {
+      "epoch": 0.06305076331140623,
+      "grad_norm": 0.07153054219977284,
+      "learning_rate": 9.995626147581141e-06,
+      "loss": 0.5627,
+      "step": 127
+    },
+    {
+      "epoch": 0.06354722601464564,
+      "grad_norm": 0.07620684738653265,
+      "learning_rate": 9.995544023503179e-06,
+      "loss": 0.6354,
+      "step": 128
+    },
+    {
+      "epoch": 0.06404368871788507,
+      "grad_norm": 0.0745933597128834,
+      "learning_rate": 9.995461135936409e-06,
+      "loss": 0.565,
+      "step": 129
+    },
+    {
+      "epoch": 0.06454015142112449,
+      "grad_norm": 0.07494493059038407,
+      "learning_rate": 9.9953774848935e-06,
+      "loss": 0.5732,
+      "step": 130
+    },
+    {
+      "epoch": 0.0650366141243639,
+      "grad_norm": 0.07408197272957125,
+      "learning_rate": 9.995293070387237e-06,
+      "loss": 0.6197,
+      "step": 131
+    },
+    {
+      "epoch": 0.06553307682760333,
+      "grad_norm": 0.07097144151203463,
+      "learning_rate": 9.995207892430525e-06,
+      "loss": 0.5706,
+      "step": 132
+    },
+    {
+      "epoch": 0.06602953953084274,
+      "grad_norm": 0.07411471261841912,
+      "learning_rate": 9.995121951036375e-06,
+      "loss": 0.6672,
+      "step": 133
+    },
+    {
+      "epoch": 0.06652600223408217,
+      "grad_norm": 0.08002634782883043,
+      "learning_rate": 9.995035246217928e-06,
+      "loss": 0.5513,
+      "step": 134
+    },
+    {
+      "epoch": 0.06702246493732159,
+      "grad_norm": 0.07632470968229874,
+      "learning_rate": 9.99494777798843e-06,
+      "loss": 0.6046,
+      "step": 135
+    },
+    {
+      "epoch": 0.067518927640561,
+      "grad_norm": 0.07572691173078472,
+      "learning_rate": 9.994859546361255e-06,
+      "loss": 0.5843,
+      "step": 136
+    },
+    {
+      "epoch": 0.06801539034380043,
+      "grad_norm": 0.06904451720592344,
+      "learning_rate": 9.994770551349884e-06,
+      "loss": 0.5461,
+      "step": 137
+    },
+    {
+      "epoch": 0.06851185304703984,
+      "grad_norm": 0.07299490802645708,
+      "learning_rate": 9.99468079296792e-06,
+      "loss": 0.56,
+      "step": 138
+    },
+    {
+      "epoch": 0.06900831575027926,
+      "grad_norm": 0.07518480559637351,
+      "learning_rate": 9.994590271229077e-06,
+      "loss": 0.5714,
+      "step": 139
+    },
+    {
+      "epoch": 0.06950477845351868,
+      "grad_norm": 0.07587964520731476,
+      "learning_rate": 9.994498986147196e-06,
+      "loss": 0.5743,
+      "step": 140
+    },
+    {
+      "epoch": 0.0700012411567581,
+      "grad_norm": 0.07140668444553665,
+      "learning_rate": 9.994406937736225e-06,
+      "loss": 0.5956,
+      "step": 141
+    },
+    {
+      "epoch": 0.07049770385999751,
+      "grad_norm": 0.08342316385982693,
+      "learning_rate": 9.994314126010234e-06,
+      "loss": 0.6292,
+      "step": 142
+    },
+    {
+      "epoch": 0.07099416656323694,
+      "grad_norm": 0.07516288618827946,
+      "learning_rate": 9.994220550983404e-06,
+      "loss": 0.5675,
+      "step": 143
+    },
+    {
+      "epoch": 0.07149062926647635,
+      "grad_norm": 0.07794453135148774,
+      "learning_rate": 9.994126212670042e-06,
+      "loss": 0.6,
+      "step": 144
+    },
+    {
+      "epoch": 0.07198709196971577,
+      "grad_norm": 0.07508762812017271,
+      "learning_rate": 9.99403111108456e-06,
+      "loss": 0.6324,
+      "step": 145
+    },
+    {
+      "epoch": 0.0724835546729552,
+      "grad_norm": 0.07307616676603801,
+      "learning_rate": 9.9939352462415e-06,
+      "loss": 0.5638,
+      "step": 146
+    },
+    {
+      "epoch": 0.07298001737619461,
+      "grad_norm": 0.07276886958258562,
+      "learning_rate": 9.993838618155505e-06,
+      "loss": 0.5773,
+      "step": 147
+    },
+    {
+      "epoch": 0.07347648007943403,
+      "grad_norm": 0.07557752398442871,
+      "learning_rate": 9.99374122684135e-06,
+      "loss": 0.5808,
+      "step": 148
+    },
+    {
+      "epoch": 0.07397294278267345,
+      "grad_norm": 0.0733881295710281,
+      "learning_rate": 9.993643072313916e-06,
+      "loss": 0.5679,
+      "step": 149
+    },
+    {
+      "epoch": 0.07446940548591287,
+      "grad_norm": 0.07381382684727514,
+      "learning_rate": 9.993544154588206e-06,
+      "loss": 0.5883,
+      "step": 150
+    },
+    {
+      "epoch": 0.0749658681891523,
+      "grad_norm": 0.07520498828814545,
+      "learning_rate": 9.993444473679337e-06,
+      "loss": 0.6129,
+      "step": 151
+    },
+    {
+      "epoch": 0.07546233089239171,
+      "grad_norm": 0.07066703681374047,
+      "learning_rate": 9.993344029602543e-06,
+      "loss": 0.5465,
+      "step": 152
+    },
+    {
+      "epoch": 0.07595879359563112,
+      "grad_norm": 0.06991119947069825,
+      "learning_rate": 9.993242822373178e-06,
+      "loss": 0.5595,
+      "step": 153
+    },
+    {
+      "epoch": 0.07645525629887055,
+      "grad_norm": 0.07303529433121117,
+      "learning_rate": 9.993140852006708e-06,
+      "loss": 0.5904,
+      "step": 154
+    },
+    {
+      "epoch": 0.07695171900210997,
+      "grad_norm": 0.08000746313978774,
+      "learning_rate": 9.993038118518716e-06,
+      "loss": 0.6066,
+      "step": 155
+    },
+    {
+      "epoch": 0.07744818170534938,
+      "grad_norm": 0.07301721545465954,
+      "learning_rate": 9.992934621924906e-06,
+      "loss": 0.5661,
+      "step": 156
+    },
+    {
+      "epoch": 0.07794464440858881,
+      "grad_norm": 0.07220658851768877,
+      "learning_rate": 9.992830362241094e-06,
+      "loss": 0.5603,
+      "step": 157
+    },
+    {
+      "epoch": 0.07844110711182822,
+      "grad_norm": 0.07414895354653442,
+      "learning_rate": 9.992725339483218e-06,
+      "loss": 0.5871,
+      "step": 158
+    },
+    {
+      "epoch": 0.07893756981506764,
+      "grad_norm": 0.0741359241647344,
+      "learning_rate": 9.992619553667321e-06,
+      "loss": 0.5853,
+      "step": 159
+    },
+    {
+      "epoch": 0.07943403251830707,
+      "grad_norm": 0.07264805114336846,
+      "learning_rate": 9.99251300480958e-06,
+      "loss": 0.5678,
+      "step": 160
+    },
+    {
+      "epoch": 0.07993049522154648,
+      "grad_norm": 0.07229450444583423,
+      "learning_rate": 9.992405692926273e-06,
+      "loss": 0.5884,
+      "step": 161
+    },
+    {
+      "epoch": 0.0804269579247859,
+      "grad_norm": 0.0740811801883871,
+      "learning_rate": 9.992297618033803e-06,
+      "loss": 0.5824,
+      "step": 162
+    },
+    {
+      "epoch": 0.08092342062802532,
+      "grad_norm": 0.07435862309763852,
+      "learning_rate": 9.99218878014869e-06,
+      "loss": 0.5959,
+      "step": 163
+    },
+    {
+      "epoch": 0.08141988333126474,
+      "grad_norm": 0.0758562192646611,
+      "learning_rate": 9.992079179287564e-06,
+      "loss": 0.6195,
+      "step": 164
+    },
+    {
+      "epoch": 0.08191634603450416,
+      "grad_norm": 0.07278506836733034,
+      "learning_rate": 9.991968815467176e-06,
+      "loss": 0.5523,
+      "step": 165
+    },
+    {
+      "epoch": 0.08241280873774358,
+      "grad_norm": 0.07266838334484792,
+      "learning_rate": 9.991857688704396e-06,
+      "loss": 0.6246,
+      "step": 166
+    },
+    {
+      "epoch": 0.08290927144098299,
+      "grad_norm": 0.07365765610190685,
+      "learning_rate": 9.991745799016206e-06,
+      "loss": 0.5916,
+      "step": 167
+    },
+    {
+      "epoch": 0.08340573414422242,
+      "grad_norm": 0.07200755697270038,
+      "learning_rate": 9.991633146419707e-06,
+      "loss": 0.5803,
+      "step": 168
+    },
+    {
+      "epoch": 0.08390219684746184,
+      "grad_norm": 0.07210206192758828,
+      "learning_rate": 9.991519730932118e-06,
+      "loss": 0.574,
+      "step": 169
+    },
+    {
+      "epoch": 0.08439865955070125,
+      "grad_norm": 0.07267719726276502,
+      "learning_rate": 9.99140555257077e-06,
+      "loss": 0.5798,
+      "step": 170
+    },
+    {
+      "epoch": 0.08489512225394068,
+      "grad_norm": 0.07155471898330006,
+      "learning_rate": 9.991290611353116e-06,
+      "loss": 0.5677,
+      "step": 171
+    },
+    {
+      "epoch": 0.08539158495718009,
+      "grad_norm": 0.07381712566994895,
+      "learning_rate": 9.991174907296718e-06,
+      "loss": 0.6053,
+      "step": 172
+    },
+    {
+      "epoch": 0.0858880476604195,
+      "grad_norm": 0.07574128504119582,
+      "learning_rate": 9.991058440419265e-06,
+      "loss": 0.562,
+      "step": 173
+    },
+    {
+      "epoch": 0.08638451036365893,
+      "grad_norm": 0.07475073669944175,
+      "learning_rate": 9.990941210738553e-06,
+      "loss": 0.5858,
+      "step": 174
+    },
+    {
+      "epoch": 0.08688097306689835,
+      "grad_norm": 0.07344393479012397,
+      "learning_rate": 9.990823218272503e-06,
+      "loss": 0.5688,
+      "step": 175
+    },
+    {
+      "epoch": 0.08737743577013776,
+      "grad_norm": 0.07812657615336012,
+      "learning_rate": 9.990704463039144e-06,
+      "loss": 0.6131,
+      "step": 176
+    },
+    {
+      "epoch": 0.08787389847337719,
+      "grad_norm": 0.0750686775572329,
+      "learning_rate": 9.99058494505663e-06,
+      "loss": 0.5756,
+      "step": 177
+    },
+    {
+      "epoch": 0.0883703611766166,
+      "grad_norm": 0.07104870179442725,
+      "learning_rate": 9.990464664343223e-06,
+      "loss": 0.5716,
+      "step": 178
+    },
+    {
+      "epoch": 0.08886682387985602,
+      "grad_norm": 0.07287131127888745,
+      "learning_rate": 9.990343620917308e-06,
+      "loss": 0.5997,
+      "step": 179
+    },
+    {
+      "epoch": 0.08936328658309545,
+      "grad_norm": 0.07489206888108628,
+      "learning_rate": 9.990221814797386e-06,
+      "loss": 0.5612,
+      "step": 180
+    },
+    {
+      "epoch": 0.08985974928633486,
+      "grad_norm": 0.07593251655355389,
+      "learning_rate": 9.990099246002071e-06,
+      "loss": 0.5759,
+      "step": 181
+    },
+    {
+      "epoch": 0.09035621198957429,
+      "grad_norm": 0.07450332229553874,
+      "learning_rate": 9.989975914550097e-06,
+      "loss": 0.5707,
+      "step": 182
+    },
+    {
+      "epoch": 0.0908526746928137,
+      "grad_norm": 0.0711423216721135,
+      "learning_rate": 9.989851820460312e-06,
+      "loss": 0.5571,
+      "step": 183
+    },
+    {
+      "epoch": 0.09134913739605312,
+      "grad_norm": 0.07565282940352779,
+      "learning_rate": 9.989726963751683e-06,
+      "loss": 0.5767,
+      "step": 184
+    },
+    {
+      "epoch": 0.09184560009929255,
+      "grad_norm": 0.07365920974609773,
+      "learning_rate": 9.989601344443291e-06,
+      "loss": 0.5579,
+      "step": 185
+    },
+    {
+      "epoch": 0.09234206280253196,
+      "grad_norm": 0.07434713572581113,
+      "learning_rate": 9.989474962554335e-06,
+      "loss": 0.6034,
+      "step": 186
+    },
+    {
+      "epoch": 0.09283852550577137,
+      "grad_norm": 0.07382111167342237,
+      "learning_rate": 9.989347818104134e-06,
+      "loss": 0.5744,
+      "step": 187
+    },
+    {
+      "epoch": 0.0933349882090108,
+      "grad_norm": 0.07136125426616181,
+      "learning_rate": 9.989219911112114e-06,
+      "loss": 0.5631,
+      "step": 188
+    },
+    {
+      "epoch": 0.09383145091225022,
+      "grad_norm": 0.0733337379350001,
+      "learning_rate": 9.989091241597828e-06,
+      "loss": 0.5793,
+      "step": 189
+    },
+    {
+      "epoch": 0.09432791361548963,
+      "grad_norm": 0.07566215084062144,
+      "learning_rate": 9.988961809580939e-06,
+      "loss": 0.5552,
+      "step": 190
+    },
+    {
+      "epoch": 0.09482437631872906,
+      "grad_norm": 0.07440377413674429,
+      "learning_rate": 9.988831615081231e-06,
+      "loss": 0.6103,
+      "step": 191
+    },
+    {
+      "epoch": 0.09532083902196847,
+      "grad_norm": 0.07548618401783257,
+      "learning_rate": 9.9887006581186e-06,
+      "loss": 0.5744,
+      "step": 192
+    },
+    {
+      "epoch": 0.09581730172520789,
+      "grad_norm": 0.0704767952117985,
+      "learning_rate": 9.98856893871306e-06,
+      "loss": 0.569,
+      "step": 193
+    },
+    {
+      "epoch": 0.09631376442844732,
+      "grad_norm": 0.0725464372041384,
+      "learning_rate": 9.988436456884743e-06,
+      "loss": 0.5742,
+      "step": 194
+    },
+    {
+      "epoch": 0.09681022713168673,
+      "grad_norm": 0.0710557842749959,
+      "learning_rate": 9.988303212653898e-06,
+      "loss": 0.5572,
+      "step": 195
+    },
+    {
+      "epoch": 0.09730668983492616,
+      "grad_norm": 0.0735237151063019,
+      "learning_rate": 9.988169206040889e-06,
+      "loss": 0.5654,
+      "step": 196
+    },
+    {
+      "epoch": 0.09780315253816557,
+      "grad_norm": 0.07411964385544856,
+      "learning_rate": 9.988034437066195e-06,
+      "loss": 0.5732,
+      "step": 197
+    },
+    {
+      "epoch": 0.09829961524140499,
+      "grad_norm": 0.07267276911713769,
+      "learning_rate": 9.987898905750416e-06,
+      "loss": 0.5752,
+      "step": 198
+    },
+    {
+      "epoch": 0.09879607794464441,
+      "grad_norm": 0.07442839143860121,
+      "learning_rate": 9.987762612114262e-06,
+      "loss": 0.5654,
+      "step": 199
+    },
+    {
+      "epoch": 0.09929254064788383,
+      "grad_norm": 0.07200329045156166,
+      "learning_rate": 9.987625556178566e-06,
+      "loss": 0.5513,
+      "step": 200
+    },
+    {
+      "epoch": 0.09978900335112324,
+      "grad_norm": 0.07040542641505096,
+      "learning_rate": 9.987487737964274e-06,
+      "loss": 0.5314,
+      "step": 201
+    },
+    {
+      "epoch": 0.10028546605436267,
+      "grad_norm": 0.07287637792348034,
+      "learning_rate": 9.98734915749245e-06,
+      "loss": 0.5962,
+      "step": 202
+    },
+    {
+      "epoch": 0.10078192875760209,
+      "grad_norm": 0.07244379147061296,
+      "learning_rate": 9.987209814784273e-06,
+      "loss": 0.5623,
+      "step": 203
+    },
+    {
+      "epoch": 0.1012783914608415,
+      "grad_norm": 0.07236469504659584,
+      "learning_rate": 9.98706970986104e-06,
+      "loss": 0.5538,
+      "step": 204
+    },
+    {
+      "epoch": 0.10177485416408093,
+      "grad_norm": 0.07466623347714184,
+      "learning_rate": 9.986928842744163e-06,
+      "loss": 0.5645,
+      "step": 205
+    },
+    {
+      "epoch": 0.10227131686732034,
+      "grad_norm": 0.07052940531611344,
+      "learning_rate": 9.986787213455174e-06,
+      "loss": 0.5309,
+      "step": 206
+    },
+    {
+      "epoch": 0.10276777957055976,
+      "grad_norm": 0.07516276980019747,
+      "learning_rate": 9.986644822015715e-06,
+      "loss": 0.5487,
+      "step": 207
+    },
+    {
+      "epoch": 0.10326424227379918,
+      "grad_norm": 0.07671293855982048,
+      "learning_rate": 9.986501668447547e-06,
+      "loss": 0.5879,
+      "step": 208
+    },
+    {
+      "epoch": 0.1037607049770386,
+      "grad_norm": 0.07208784254703426,
+      "learning_rate": 9.986357752772555e-06,
+      "loss": 0.5536,
+      "step": 209
+    },
+    {
+      "epoch": 0.10425716768027801,
+      "grad_norm": 0.07815563722985788,
+      "learning_rate": 9.98621307501273e-06,
+      "loss": 0.5642,
+      "step": 210
+    },
+    {
+      "epoch": 0.10475363038351744,
+      "grad_norm": 0.07531957358708724,
+      "learning_rate": 9.986067635190184e-06,
+      "loss": 0.5764,
+      "step": 211
+    },
+    {
+      "epoch": 0.10525009308675685,
+      "grad_norm": 0.07516730483203957,
+      "learning_rate": 9.985921433327144e-06,
+      "loss": 0.5893,
+      "step": 212
+    },
+    {
+      "epoch": 0.10574655578999628,
+      "grad_norm": 0.07452516834494967,
+      "learning_rate": 9.985774469445957e-06,
+      "loss": 0.5878,
+      "step": 213
+    },
+    {
+      "epoch": 0.1062430184932357,
+      "grad_norm": 0.0767192562209654,
+      "learning_rate": 9.985626743569083e-06,
+      "loss": 0.5586,
+      "step": 214
+    },
+    {
+      "epoch": 0.10673948119647511,
+      "grad_norm": 0.07419626208845845,
+      "learning_rate": 9.9854782557191e-06,
+      "loss": 0.5828,
+      "step": 215
+    },
+    {
+      "epoch": 0.10723594389971454,
+      "grad_norm": 0.07684498206577063,
+      "learning_rate": 9.985329005918702e-06,
+      "loss": 0.5651,
+      "step": 216
+    },
+    {
+      "epoch": 0.10773240660295395,
+      "grad_norm": 0.07714407534776108,
+      "learning_rate": 9.9851789941907e-06,
+      "loss": 0.6076,
+      "step": 217
+    },
+    {
+      "epoch": 0.10822886930619337,
+      "grad_norm": 0.07198389859968202,
+      "learning_rate": 9.98502822055802e-06,
+      "loss": 0.5612,
+      "step": 218
+    },
+    {
+      "epoch": 0.1087253320094328,
+      "grad_norm": 0.07571784717245837,
+      "learning_rate": 9.984876685043703e-06,
+      "loss": 0.6116,
+      "step": 219
+    },
+    {
+      "epoch": 0.10922179471267221,
+      "grad_norm": 0.07953051944371987,
+      "learning_rate": 9.984724387670912e-06,
+      "loss": 0.5886,
+      "step": 220
+    },
+    {
+      "epoch": 0.10971825741591162,
+      "grad_norm": 0.07770772328549033,
+      "learning_rate": 9.984571328462924e-06,
+      "loss": 0.5531,
+      "step": 221
+    },
+    {
+      "epoch": 0.11021472011915105,
+      "grad_norm": 0.07536582074411971,
+      "learning_rate": 9.984417507443128e-06,
+      "loss": 0.595,
+      "step": 222
+    },
+    {
+      "epoch": 0.11071118282239047,
+      "grad_norm": 0.07405994836904185,
+      "learning_rate": 9.984262924635036e-06,
+      "loss": 0.5833,
+      "step": 223
+    },
+    {
+      "epoch": 0.11120764552562988,
+      "grad_norm": 0.08017377083919287,
+      "learning_rate": 9.984107580062273e-06,
+      "loss": 0.6021,
+      "step": 224
+    },
+    {
+      "epoch": 0.11170410822886931,
+      "grad_norm": 0.08072092192849259,
+      "learning_rate": 9.983951473748579e-06,
+      "loss": 0.5736,
+      "step": 225
+    },
+    {
+      "epoch": 0.11220057093210872,
+      "grad_norm": 0.07534405176458785,
+      "learning_rate": 9.983794605717815e-06,
+      "loss": 0.5621,
+      "step": 226
+    },
+    {
+      "epoch": 0.11269703363534815,
+      "grad_norm": 0.07547844096854822,
+      "learning_rate": 9.983636975993953e-06,
+      "loss": 0.6161,
+      "step": 227
+    },
+    {
+      "epoch": 0.11319349633858757,
+      "grad_norm": 0.07934085975212889,
+      "learning_rate": 9.983478584601088e-06,
+      "loss": 0.5867,
+      "step": 228
+    },
+    {
+      "epoch": 0.11368995904182698,
+      "grad_norm": 0.07912470758493029,
+      "learning_rate": 9.983319431563424e-06,
+      "loss": 0.5841,
+      "step": 229
+    },
+    {
+      "epoch": 0.11418642174506641,
+      "grad_norm": 0.07275368539544204,
+      "learning_rate": 9.983159516905287e-06,
+      "loss": 0.5827,
+      "step": 230
+    },
+    {
+      "epoch": 0.11468288444830582,
+      "grad_norm": 0.07569968972355616,
+      "learning_rate": 9.982998840651117e-06,
+      "loss": 0.5851,
+      "step": 231
+    },
+    {
+      "epoch": 0.11517934715154524,
+      "grad_norm": 0.07962842640771996,
+      "learning_rate": 9.98283740282547e-06,
+      "loss": 0.596,
+      "step": 232
+    },
+    {
+      "epoch": 0.11567580985478466,
+      "grad_norm": 0.07804031431161682,
+      "learning_rate": 9.982675203453018e-06,
+      "loss": 0.571,
+      "step": 233
+    },
+    {
+      "epoch": 0.11617227255802408,
+      "grad_norm": 0.07700514481764709,
+      "learning_rate": 9.982512242558555e-06,
+      "loss": 0.565,
+      "step": 234
+    },
+    {
+      "epoch": 0.11666873526126349,
+      "grad_norm": 0.0708208906544335,
+      "learning_rate": 9.982348520166982e-06,
+      "loss": 0.5604,
+      "step": 235
+    },
+    {
+      "epoch": 0.11716519796450292,
+      "grad_norm": 0.0754541608709195,
+      "learning_rate": 9.982184036303326e-06,
+      "loss": 0.5987,
+      "step": 236
+    },
+    {
+      "epoch": 0.11766166066774233,
+      "grad_norm": 0.07554717965041857,
+      "learning_rate": 9.982018790992722e-06,
+      "loss": 0.5691,
+      "step": 237
+    },
+    {
+      "epoch": 0.11815812337098175,
+      "grad_norm": 0.07752289757631088,
+      "learning_rate": 9.981852784260427e-06,
+      "loss": 0.5637,
+      "step": 238
+    },
+    {
+      "epoch": 0.11865458607422118,
+      "grad_norm": 0.0762749443511498,
+      "learning_rate": 9.98168601613181e-06,
+      "loss": 0.6146,
+      "step": 239
+    },
+    {
+      "epoch": 0.11915104877746059,
+      "grad_norm": 0.07530073323828042,
+      "learning_rate": 9.981518486632363e-06,
+      "loss": 0.5591,
+      "step": 240
+    },
+    {
+      "epoch": 0.1196475114807,
+      "grad_norm": 0.07507111474292993,
+      "learning_rate": 9.981350195787685e-06,
+      "loss": 0.551,
+      "step": 241
+    },
+    {
+      "epoch": 0.12014397418393943,
+      "grad_norm": 0.07371412964257293,
+      "learning_rate": 9.981181143623501e-06,
+      "loss": 0.5903,
+      "step": 242
+    },
+    {
+      "epoch": 0.12064043688717885,
+      "grad_norm": 0.0704662212442052,
+      "learning_rate": 9.981011330165648e-06,
+      "loss": 0.5489,
+      "step": 243
+    },
+    {
+      "epoch": 0.12113689959041828,
+      "grad_norm": 0.07236692881961186,
+      "learning_rate": 9.980840755440075e-06,
+      "loss": 0.5612,
+      "step": 244
+    },
+    {
+      "epoch": 0.12163336229365769,
+      "grad_norm": 0.07532610257826663,
+      "learning_rate": 9.980669419472856e-06,
+      "loss": 0.555,
+      "step": 245
+    },
+    {
+      "epoch": 0.1221298249968971,
+      "grad_norm": 0.07443526564494042,
+      "learning_rate": 9.980497322290174e-06,
+      "loss": 0.5804,
+      "step": 246
+    },
+    {
+      "epoch": 0.12262628770013653,
+      "grad_norm": 0.07415250825090586,
+      "learning_rate": 9.980324463918333e-06,
+      "loss": 0.5916,
+      "step": 247
+    },
+    {
+      "epoch": 0.12312275040337595,
+      "grad_norm": 0.07255581056843356,
+      "learning_rate": 9.980150844383753e-06,
+      "loss": 0.5838,
+      "step": 248
+    },
+    {
+      "epoch": 0.12361921310661536,
+      "grad_norm": 0.07386636777765433,
+      "learning_rate": 9.979976463712966e-06,
+      "loss": 0.5536,
+      "step": 249
+    },
+    {
+      "epoch": 0.12411567580985479,
+      "grad_norm": 0.07611005306904636,
+      "learning_rate": 9.979801321932624e-06,
+      "loss": 0.6237,
+      "step": 250
+    },
+    {
+      "epoch": 0.1246121385130942,
+      "grad_norm": 0.07533106950641502,
+      "learning_rate": 9.979625419069495e-06,
+      "loss": 0.5828,
+      "step": 251
+    },
+    {
+      "epoch": 0.12510860121633363,
+      "grad_norm": 0.07288766763589251,
+      "learning_rate": 9.979448755150461e-06,
+      "loss": 0.5625,
+      "step": 252
+    },
+    {
+      "epoch": 0.12560506391957305,
+      "grad_norm": 0.07783186390956214,
+      "learning_rate": 9.979271330202527e-06,
+      "loss": 0.5782,
+      "step": 253
+    },
+    {
+      "epoch": 0.12610152662281246,
+      "grad_norm": 0.0739141983854619,
+      "learning_rate": 9.979093144252804e-06,
+      "loss": 0.6089,
+      "step": 254
+    },
+    {
+      "epoch": 0.12659798932605187,
+      "grad_norm": 0.0776536814927388,
+      "learning_rate": 9.978914197328531e-06,
+      "loss": 0.593,
+      "step": 255
+    },
+    {
+      "epoch": 0.1270944520292913,
+      "grad_norm": 0.08012066470866044,
+      "learning_rate": 9.978734489457051e-06,
+      "loss": 0.5998,
+      "step": 256
+    },
+    {
+      "epoch": 0.12759091473253073,
+      "grad_norm": 0.07669771884662993,
+      "learning_rate": 9.978554020665834e-06,
+      "loss": 0.534,
+      "step": 257
+    },
+    {
+      "epoch": 0.12808737743577014,
+      "grad_norm": 0.07662332363179943,
+      "learning_rate": 9.978372790982457e-06,
+      "loss": 0.5994,
+      "step": 258
+    },
+    {
+      "epoch": 0.12858384013900956,
+      "grad_norm": 0.07043282149907523,
+      "learning_rate": 9.978190800434624e-06,
+      "loss": 0.5717,
+      "step": 259
+    },
+    {
+      "epoch": 0.12908030284224897,
+      "grad_norm": 0.07420387749370701,
+      "learning_rate": 9.978008049050145e-06,
+      "loss": 0.5903,
+      "step": 260
+    },
+    {
+      "epoch": 0.1295767655454884,
+      "grad_norm": 0.07327527104462504,
+      "learning_rate": 9.977824536856953e-06,
+      "loss": 0.562,
+      "step": 261
+    },
+    {
+      "epoch": 0.1300732282487278,
+      "grad_norm": 0.07017862418856051,
+      "learning_rate": 9.977640263883095e-06,
+      "loss": 0.5443,
+      "step": 262
+    },
+    {
+      "epoch": 0.13056969095196724,
+      "grad_norm": 0.07109669443509863,
+      "learning_rate": 9.97745523015673e-06,
+      "loss": 0.5225,
+      "step": 263
+    },
+    {
+      "epoch": 0.13106615365520666,
+      "grad_norm": 0.07932382983195517,
+      "learning_rate": 9.977269435706142e-06,
+      "loss": 0.5852,
+      "step": 264
+    },
+    {
+      "epoch": 0.13156261635844607,
+      "grad_norm": 0.07553526767911642,
+      "learning_rate": 9.977082880559725e-06,
+      "loss": 0.5724,
+      "step": 265
+    },
+    {
+      "epoch": 0.13205907906168549,
+      "grad_norm": 0.07454386060375846,
+      "learning_rate": 9.976895564745993e-06,
+      "loss": 0.5578,
+      "step": 266
+    },
+    {
+      "epoch": 0.1325555417649249,
+      "grad_norm": 0.07445473105285273,
+      "learning_rate": 9.976707488293569e-06,
+      "loss": 0.5321,
+      "step": 267
+    },
+    {
+      "epoch": 0.13305200446816434,
+      "grad_norm": 0.07898175514251754,
+      "learning_rate": 9.976518651231203e-06,
+      "loss": 0.622,
+      "step": 268
+    },
+    {
+      "epoch": 0.13354846717140376,
+      "grad_norm": 0.07381948302565586,
+      "learning_rate": 9.976329053587754e-06,
+      "loss": 0.5391,
+      "step": 269
+    },
+    {
+      "epoch": 0.13404492987464317,
+      "grad_norm": 0.0806159559251752,
+      "learning_rate": 9.976138695392196e-06,
+      "loss": 0.5758,
+      "step": 270
+    },
+    {
+      "epoch": 0.13454139257788258,
+      "grad_norm": 0.08268280266256835,
+      "learning_rate": 9.975947576673628e-06,
+      "loss": 0.5687,
+      "step": 271
+    },
+    {
+      "epoch": 0.135037855281122,
+      "grad_norm": 0.07927875514189735,
+      "learning_rate": 9.975755697461254e-06,
+      "loss": 0.5535,
+      "step": 272
+    },
+    {
+      "epoch": 0.1355343179843614,
+      "grad_norm": 0.07325803338115366,
+      "learning_rate": 9.975563057784402e-06,
+      "loss": 0.5325,
+      "step": 273
+    },
+    {
+      "epoch": 0.13603078068760086,
+      "grad_norm": 0.07448375330798707,
+      "learning_rate": 9.975369657672514e-06,
+      "loss": 0.5782,
+      "step": 274
+    },
+    {
+      "epoch": 0.13652724339084027,
+      "grad_norm": 0.07525898100338235,
+      "learning_rate": 9.975175497155149e-06,
+      "loss": 0.5459,
+      "step": 275
+    },
+    {
+      "epoch": 0.13702370609407968,
+      "grad_norm": 0.0757930579472259,
+      "learning_rate": 9.97498057626198e-06,
+      "loss": 0.5652,
+      "step": 276
+    },
+    {
+      "epoch": 0.1375201687973191,
+      "grad_norm": 0.07625175178868698,
+      "learning_rate": 9.974784895022796e-06,
+      "loss": 0.5101,
+      "step": 277
+    },
+    {
+      "epoch": 0.1380166315005585,
+      "grad_norm": 0.07588767928782164,
+      "learning_rate": 9.974588453467506e-06,
+      "loss": 0.5882,
+      "step": 278
+    },
+    {
+      "epoch": 0.13851309420379793,
+      "grad_norm": 0.0733443443040492,
+      "learning_rate": 9.974391251626132e-06,
+      "loss": 0.5476,
+      "step": 279
+    },
+    {
+      "epoch": 0.13900955690703737,
+      "grad_norm": 0.07328131307535096,
+      "learning_rate": 9.974193289528814e-06,
+      "loss": 0.5836,
+      "step": 280
+    },
+    {
+      "epoch": 0.13950601961027678,
+      "grad_norm": 0.07874953449684817,
+      "learning_rate": 9.973994567205806e-06,
+      "loss": 0.6189,
+      "step": 281
+    },
+    {
+      "epoch": 0.1400024823135162,
+      "grad_norm": 0.07891877733000807,
+      "learning_rate": 9.973795084687481e-06,
+      "loss": 0.5618,
+      "step": 282
+    },
+    {
+      "epoch": 0.1404989450167556,
+      "grad_norm": 0.0785897983413089,
+      "learning_rate": 9.973594842004327e-06,
+      "loss": 0.584,
+      "step": 283
+    },
+    {
+      "epoch": 0.14099540771999503,
+      "grad_norm": 0.07245957200919304,
+      "learning_rate": 9.973393839186946e-06,
+      "loss": 0.549,
+      "step": 284
+    },
+    {
+      "epoch": 0.14149187042323447,
+      "grad_norm": 0.07728054437478935,
+      "learning_rate": 9.973192076266058e-06,
+      "loss": 0.5761,
+      "step": 285
+    },
+    {
+      "epoch": 0.14198833312647388,
+      "grad_norm": 0.09235613660316606,
+      "learning_rate": 9.972989553272501e-06,
+      "loss": 0.6147,
+      "step": 286
+    },
+    {
+      "epoch": 0.1424847958297133,
+      "grad_norm": 0.07380406310732682,
+      "learning_rate": 9.972786270237228e-06,
+      "loss": 0.6105,
+      "step": 287
+    },
+    {
+      "epoch": 0.1429812585329527,
+      "grad_norm": 0.07278111762088124,
+      "learning_rate": 9.972582227191305e-06,
+      "loss": 0.5478,
+      "step": 288
+    },
+    {
+      "epoch": 0.14347772123619212,
+      "grad_norm": 0.07524458713971298,
+      "learning_rate": 9.972377424165918e-06,
+      "loss": 0.5486,
+      "step": 289
+    },
+    {
+      "epoch": 0.14397418393943154,
+      "grad_norm": 0.08081462769097447,
+      "learning_rate": 9.972171861192368e-06,
+      "loss": 0.5772,
+      "step": 290
+    },
+    {
+      "epoch": 0.14447064664267098,
+      "grad_norm": 0.07598914546101065,
+      "learning_rate": 9.97196553830207e-06,
+      "loss": 0.5576,
+      "step": 291
+    },
+    {
+      "epoch": 0.1449671093459104,
+      "grad_norm": 0.07361633688826458,
+      "learning_rate": 9.971758455526562e-06,
+      "loss": 0.6215,
+      "step": 292
+    },
+    {
+      "epoch": 0.1454635720491498,
+      "grad_norm": 0.07371424684661225,
+      "learning_rate": 9.971550612897487e-06,
+      "loss": 0.5612,
+      "step": 293
+    },
+    {
+      "epoch": 0.14596003475238922,
+      "grad_norm": 0.07537311603859617,
+      "learning_rate": 9.971342010446615e-06,
+      "loss": 0.5596,
+      "step": 294
+    },
+    {
+      "epoch": 0.14645649745562864,
+      "grad_norm": 0.07370938816688641,
+      "learning_rate": 9.971132648205826e-06,
+      "loss": 0.5895,
+      "step": 295
+    },
+    {
+      "epoch": 0.14695296015886805,
+      "grad_norm": 0.07938936326373351,
+      "learning_rate": 9.97092252620712e-06,
+      "loss": 0.5979,
+      "step": 296
+    },
+    {
+      "epoch": 0.1474494228621075,
+      "grad_norm": 0.07655933756147211,
+      "learning_rate": 9.970711644482605e-06,
+      "loss": 0.571,
+      "step": 297
+    },
+    {
+      "epoch": 0.1479458855653469,
+      "grad_norm": 0.06978512801359449,
+      "learning_rate": 9.970500003064517e-06,
+      "loss": 0.5677,
+      "step": 298
+    },
+    {
+      "epoch": 0.14844234826858632,
+      "grad_norm": 0.07479095468689204,
+      "learning_rate": 9.970287601985197e-06,
+      "loss": 0.5436,
+      "step": 299
+    },
+    {
+      "epoch": 0.14893881097182574,
+      "grad_norm": 0.07551809721412871,
+      "learning_rate": 9.970074441277111e-06,
+      "loss": 0.5156,
+      "step": 300
+    },
+    {
+      "epoch": 0.14943527367506515,
+      "grad_norm": 0.0726621639022755,
+      "learning_rate": 9.969860520972835e-06,
+      "loss": 0.5155,
+      "step": 301
+    },
+    {
+      "epoch": 0.1499317363783046,
+      "grad_norm": 0.07510700099410206,
+      "learning_rate": 9.969645841105065e-06,
+      "loss": 0.5828,
+      "step": 302
+    },
+    {
+      "epoch": 0.150428199081544,
+      "grad_norm": 0.07620065582710349,
+      "learning_rate": 9.96943040170661e-06,
+      "loss": 0.5825,
+      "step": 303
+    },
+    {
+      "epoch": 0.15092466178478342,
+      "grad_norm": 0.07454115542645492,
+      "learning_rate": 9.969214202810397e-06,
+      "loss": 0.566,
+      "step": 304
+    },
+    {
+      "epoch": 0.15142112448802283,
+      "grad_norm": 0.07571175610952988,
+      "learning_rate": 9.968997244449467e-06,
+      "loss": 0.5883,
+      "step": 305
+    },
+    {
+      "epoch": 0.15191758719126225,
+      "grad_norm": 0.07918443748502917,
+      "learning_rate": 9.968779526656981e-06,
+      "loss": 0.5837,
+      "step": 306
+    },
+    {
+      "epoch": 0.15241404989450166,
+      "grad_norm": 0.08231002500440109,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 0.5906,
+      "step": 307
+    },
+    {
+      "epoch": 0.1529105125977411,
+      "grad_norm": 0.08172637702489811,
+      "learning_rate": 9.968341812910553e-06,
+      "loss": 0.5776,
+      "step": 308
+    },
+    {
+      "epoch": 0.15340697530098052,
+      "grad_norm": 0.07541026588900428,
+      "learning_rate": 9.96812181702351e-06,
+      "loss": 0.5984,
+      "step": 309
+    },
+    {
+      "epoch": 0.15390343800421993,
+      "grad_norm": 0.07833737504960553,
+      "learning_rate": 9.967901061838703e-06,
+      "loss": 0.5902,
+      "step": 310
+    },
+    {
+      "epoch": 0.15439990070745935,
+      "grad_norm": 0.08772300134995734,
+      "learning_rate": 9.967679547389874e-06,
+      "loss": 0.5492,
+      "step": 311
+    },
+    {
+      "epoch": 0.15489636341069876,
+      "grad_norm": 0.0802801470852147,
+      "learning_rate": 9.967457273710877e-06,
+      "loss": 0.6016,
+      "step": 312
+    },
+    {
+      "epoch": 0.1553928261139382,
+      "grad_norm": 0.07234712570516748,
+      "learning_rate": 9.967234240835682e-06,
+      "loss": 0.5536,
+      "step": 313
+    },
+    {
+      "epoch": 0.15588928881717762,
+      "grad_norm": 0.07228246183847936,
+      "learning_rate": 9.967010448798376e-06,
+      "loss": 0.5698,
+      "step": 314
+    },
+    {
+      "epoch": 0.15638575152041703,
+      "grad_norm": 0.0791197093555623,
+      "learning_rate": 9.966785897633164e-06,
+      "loss": 0.5761,
+      "step": 315
+    },
+    {
+      "epoch": 0.15688221422365645,
+      "grad_norm": 0.0809617690187681,
+      "learning_rate": 9.966560587374363e-06,
+      "loss": 0.5784,
+      "step": 316
+    },
+    {
+      "epoch": 0.15737867692689586,
+      "grad_norm": 0.07455628784150602,
+      "learning_rate": 9.96633451805641e-06,
+      "loss": 0.5549,
+      "step": 317
+    },
+    {
+      "epoch": 0.15787513963013527,
+      "grad_norm": 0.07556973516566123,
+      "learning_rate": 9.966107689713855e-06,
+      "loss": 0.5946,
+      "step": 318
+    },
+    {
+      "epoch": 0.15837160233337472,
+      "grad_norm": 0.07815659673820999,
+      "learning_rate": 9.965880102381364e-06,
+      "loss": 0.5446,
+      "step": 319
+    },
+    {
+      "epoch": 0.15886806503661413,
+      "grad_norm": 0.07204342389468431,
+      "learning_rate": 9.965651756093724e-06,
+      "loss": 0.5454,
+      "step": 320
+    },
+    {
+      "epoch": 0.15936452773985355,
+      "grad_norm": 0.07453809229436384,
+      "learning_rate": 9.965422650885829e-06,
+      "loss": 0.5483,
+      "step": 321
+    },
+    {
+      "epoch": 0.15986099044309296,
+      "grad_norm": 0.07787163871393117,
+      "learning_rate": 9.965192786792696e-06,
+      "loss": 0.5361,
+      "step": 322
+    },
+    {
+      "epoch": 0.16035745314633237,
+      "grad_norm": 0.07623670256160507,
+      "learning_rate": 9.964962163849457e-06,
+      "loss": 0.6004,
+      "step": 323
+    },
+    {
+      "epoch": 0.1608539158495718,
+      "grad_norm": 0.073269081360789,
+      "learning_rate": 9.964730782091358e-06,
+      "loss": 0.5865,
+      "step": 324
+    },
+    {
+      "epoch": 0.16135037855281123,
+      "grad_norm": 0.08034710038842101,
+      "learning_rate": 9.964498641553764e-06,
+      "loss": 0.6188,
+      "step": 325
+    },
+    {
+      "epoch": 0.16184684125605064,
+      "grad_norm": 0.07878506103100462,
+      "learning_rate": 9.96426574227215e-06,
+      "loss": 0.5174,
+      "step": 326
+    },
+    {
+      "epoch": 0.16234330395929006,
+      "grad_norm": 0.0811582536605504,
+      "learning_rate": 9.964032084282115e-06,
+      "loss": 0.5869,
+      "step": 327
+    },
+    {
+      "epoch": 0.16283976666252947,
+      "grad_norm": 0.07615396770698715,
+      "learning_rate": 9.963797667619368e-06,
+      "loss": 0.5489,
+      "step": 328
+    },
+    {
+      "epoch": 0.1633362293657689,
+      "grad_norm": 0.07375822341527581,
+      "learning_rate": 9.963562492319733e-06,
+      "loss": 0.5423,
+      "step": 329
+    },
+    {
+      "epoch": 0.16383269206900833,
+      "grad_norm": 0.07342225085820822,
+      "learning_rate": 9.96332655841916e-06,
+      "loss": 0.5443,
+      "step": 330
+    },
+    {
+      "epoch": 0.16432915477224774,
+      "grad_norm": 0.07522646471314104,
+      "learning_rate": 9.963089865953701e-06,
+      "loss": 0.5947,
+      "step": 331
+    },
+    {
+      "epoch": 0.16482561747548716,
+      "grad_norm": 0.07093754634633222,
+      "learning_rate": 9.962852414959534e-06,
+      "loss": 0.5636,
+      "step": 332
+    },
+    {
+      "epoch": 0.16532208017872657,
+      "grad_norm": 0.07554089167522272,
+      "learning_rate": 9.962614205472948e-06,
+      "loss": 0.5728,
+      "step": 333
+    },
+    {
+      "epoch": 0.16581854288196599,
+      "grad_norm": 0.07642283336009119,
+      "learning_rate": 9.96237523753035e-06,
+      "loss": 0.5617,
+      "step": 334
+    },
+    {
+      "epoch": 0.1663150055852054,
+      "grad_norm": 0.0754751987943475,
+      "learning_rate": 9.962135511168263e-06,
+      "loss": 0.5669,
+      "step": 335
+    },
+    {
+      "epoch": 0.16681146828844484,
+      "grad_norm": 0.07530280584445144,
+      "learning_rate": 9.961895026423325e-06,
+      "loss": 0.5727,
+      "step": 336
+    },
+    {
+      "epoch": 0.16730793099168426,
+      "grad_norm": 0.07544117416085358,
+      "learning_rate": 9.96165378333229e-06,
+      "loss": 0.5597,
+      "step": 337
+    },
+    {
+      "epoch": 0.16780439369492367,
+      "grad_norm": 0.07451626557128686,
+      "learning_rate": 9.961411781932029e-06,
+      "loss": 0.5533,
+      "step": 338
+    },
+    {
+      "epoch": 0.16830085639816308,
+      "grad_norm": 0.07645683165516677,
+      "learning_rate": 9.961169022259527e-06,
+      "loss": 0.5832,
+      "step": 339
+    },
+    {
+      "epoch": 0.1687973191014025,
+      "grad_norm": 0.074515428017266,
+      "learning_rate": 9.960925504351885e-06,
+      "loss": 0.5517,
+      "step": 340
+    },
+    {
+      "epoch": 0.1692937818046419,
+      "grad_norm": 0.07453622025099385,
+      "learning_rate": 9.960681228246323e-06,
+      "loss": 0.5404,
+      "step": 341
+    },
+    {
+      "epoch": 0.16979024450788135,
+      "grad_norm": 0.07490735728929213,
+      "learning_rate": 9.960436193980175e-06,
+      "loss": 0.5322,
+      "step": 342
+    },
+    {
+      "epoch": 0.17028670721112077,
+      "grad_norm": 0.07533458746856145,
+      "learning_rate": 9.960190401590886e-06,
+      "loss": 0.556,
+      "step": 343
+    },
+    {
+      "epoch": 0.17078316991436018,
+      "grad_norm": 0.07518188762712508,
+      "learning_rate": 9.959943851116027e-06,
+      "loss": 0.5683,
+      "step": 344
+    },
+    {
+      "epoch": 0.1712796326175996,
+      "grad_norm": 0.07031283405558475,
+      "learning_rate": 9.959696542593278e-06,
+      "loss": 0.5393,
+      "step": 345
+    },
+    {
+      "epoch": 0.171776095320839,
+      "grad_norm": 0.0748003316183382,
+      "learning_rate": 9.959448476060434e-06,
+      "loss": 0.5515,
+      "step": 346
+    },
+    {
+      "epoch": 0.17227255802407845,
+      "grad_norm": 0.07589049005675153,
+      "learning_rate": 9.959199651555409e-06,
+      "loss": 0.587,
+      "step": 347
+    },
+    {
+      "epoch": 0.17276902072731787,
+      "grad_norm": 0.0807346094812496,
+      "learning_rate": 9.95895006911623e-06,
+      "loss": 0.6032,
+      "step": 348
+    },
+    {
+      "epoch": 0.17326548343055728,
+      "grad_norm": 0.07472065836941111,
+      "learning_rate": 9.958699728781046e-06,
+      "loss": 0.5518,
+      "step": 349
+    },
+    {
+      "epoch": 0.1737619461337967,
+      "grad_norm": 0.08115486432870875,
+      "learning_rate": 9.958448630588115e-06,
+      "loss": 0.62,
+      "step": 350
+    },
+    {
+      "epoch": 0.1742584088370361,
+      "grad_norm": 0.07873831391601213,
+      "learning_rate": 9.958196774575814e-06,
+      "loss": 0.5792,
+      "step": 351
+    },
+    {
+      "epoch": 0.17475487154027552,
+      "grad_norm": 0.08112665550357574,
+      "learning_rate": 9.957944160782634e-06,
+      "loss": 0.5438,
+      "step": 352
+    },
+    {
+      "epoch": 0.17525133424351497,
+      "grad_norm": 0.07331800307555766,
+      "learning_rate": 9.957690789247183e-06,
+      "loss": 0.5474,
+      "step": 353
+    },
+    {
+      "epoch": 0.17574779694675438,
+      "grad_norm": 0.07563212018228109,
+      "learning_rate": 9.957436660008187e-06,
+      "loss": 0.5735,
+      "step": 354
+    },
+    {
+      "epoch": 0.1762442596499938,
+      "grad_norm": 0.07469151724699816,
+      "learning_rate": 9.957181773104482e-06,
+      "loss": 0.5557,
+      "step": 355
+    },
+    {
+      "epoch": 0.1767407223532332,
+      "grad_norm": 0.07800059494765482,
+      "learning_rate": 9.956926128575026e-06,
+      "loss": 0.5756,
+      "step": 356
+    },
+    {
+      "epoch": 0.17723718505647262,
+      "grad_norm": 0.07027184828449723,
+      "learning_rate": 9.95666972645889e-06,
+      "loss": 0.5288,
+      "step": 357
+    },
+    {
+      "epoch": 0.17773364775971204,
+      "grad_norm": 0.07683077691940579,
+      "learning_rate": 9.95641256679526e-06,
+      "loss": 0.5958,
+      "step": 358
+    },
+    {
+      "epoch": 0.17823011046295148,
+      "grad_norm": 0.07576235890205875,
+      "learning_rate": 9.95615464962344e-06,
+      "loss": 0.5925,
+      "step": 359
+    },
+    {
+      "epoch": 0.1787265731661909,
+      "grad_norm": 0.07861161538869584,
+      "learning_rate": 9.955895974982848e-06,
+      "loss": 0.5677,
+      "step": 360
+    },
+    {
+      "epoch": 0.1792230358694303,
+      "grad_norm": 0.07894396117012514,
+      "learning_rate": 9.955636542913017e-06,
+      "loss": 0.606,
+      "step": 361
+    },
+    {
+      "epoch": 0.17971949857266972,
+      "grad_norm": 0.08030622968705446,
+      "learning_rate": 9.955376353453599e-06,
+      "loss": 0.5616,
+      "step": 362
+    },
+    {
+      "epoch": 0.18021596127590914,
+      "grad_norm": 0.07372907998809378,
+      "learning_rate": 9.955115406644357e-06,
+      "loss": 0.5405,
+      "step": 363
+    },
+    {
+      "epoch": 0.18071242397914858,
+      "grad_norm": 0.07619467670383653,
+      "learning_rate": 9.954853702525176e-06,
+      "loss": 0.5579,
+      "step": 364
+    },
+    {
+      "epoch": 0.181208886682388,
+      "grad_norm": 0.07535913099715885,
+      "learning_rate": 9.95459124113605e-06,
+      "loss": 0.5446,
+      "step": 365
+    },
+    {
+      "epoch": 0.1817053493856274,
+      "grad_norm": 0.07285682873619495,
+      "learning_rate": 9.954328022517094e-06,
+      "loss": 0.567,
+      "step": 366
+    },
+    {
+      "epoch": 0.18220181208886682,
+      "grad_norm": 0.0794068286097486,
+      "learning_rate": 9.954064046708537e-06,
+      "loss": 0.5818,
+      "step": 367
+    },
+    {
+      "epoch": 0.18269827479210624,
+      "grad_norm": 0.07850596036013317,
+      "learning_rate": 9.953799313750723e-06,
+      "loss": 0.5736,
+      "step": 368
+    },
+    {
+      "epoch": 0.18319473749534565,
+      "grad_norm": 0.07724909085841658,
+      "learning_rate": 9.953533823684112e-06,
+      "loss": 0.5274,
+      "step": 369
+    },
+    {
+      "epoch": 0.1836912001985851,
+      "grad_norm": 0.08096743311474128,
+      "learning_rate": 9.953267576549279e-06,
+      "loss": 0.5494,
+      "step": 370
+    },
+    {
+      "epoch": 0.1841876629018245,
+      "grad_norm": 0.07398016211295078,
+      "learning_rate": 9.953000572386916e-06,
+      "loss": 0.6018,
+      "step": 371
+    },
+    {
+      "epoch": 0.18468412560506392,
+      "grad_norm": 0.07594761131816251,
+      "learning_rate": 9.952732811237833e-06,
+      "loss": 0.5691,
+      "step": 372
+    },
+    {
+      "epoch": 0.18518058830830333,
+      "grad_norm": 0.07473651094317643,
+      "learning_rate": 9.952464293142951e-06,
+      "loss": 0.5498,
+      "step": 373
+    },
+    {
+      "epoch": 0.18567705101154275,
+      "grad_norm": 0.07447128483703518,
+      "learning_rate": 9.952195018143308e-06,
+      "loss": 0.5824,
+      "step": 374
+    },
+    {
+      "epoch": 0.1861735137147822,
+      "grad_norm": 0.07317042373061025,
+      "learning_rate": 9.951924986280057e-06,
+      "loss": 0.5782,
+      "step": 375
+    },
+    {
+      "epoch": 0.1866699764180216,
+      "grad_norm": 0.07459200175934973,
+      "learning_rate": 9.951654197594471e-06,
+      "loss": 0.5294,
+      "step": 376
+    },
+    {
+      "epoch": 0.18716643912126102,
+      "grad_norm": 0.07699746371876615,
+      "learning_rate": 9.951382652127935e-06,
+      "loss": 0.5476,
+      "step": 377
+    },
+    {
+      "epoch": 0.18766290182450043,
+      "grad_norm": 0.07448223810955458,
+      "learning_rate": 9.951110349921951e-06,
+      "loss": 0.543,
+      "step": 378
+    },
+    {
+      "epoch": 0.18815936452773985,
+      "grad_norm": 0.07973769043959483,
+      "learning_rate": 9.950837291018133e-06,
+      "loss": 0.5619,
+      "step": 379
+    },
+    {
+      "epoch": 0.18865582723097926,
+      "grad_norm": 0.07310856359312369,
+      "learning_rate": 9.950563475458218e-06,
+      "loss": 0.5514,
+      "step": 380
+    },
+    {
+      "epoch": 0.1891522899342187,
+      "grad_norm": 0.07375459505878652,
+      "learning_rate": 9.950288903284049e-06,
+      "loss": 0.5505,
+      "step": 381
+    },
+    {
+      "epoch": 0.18964875263745812,
+      "grad_norm": 0.0728782672537732,
+      "learning_rate": 9.950013574537595e-06,
+      "loss": 0.5459,
+      "step": 382
+    },
+    {
+      "epoch": 0.19014521534069753,
+      "grad_norm": 0.07740246677875098,
+      "learning_rate": 9.949737489260933e-06,
+      "loss": 0.5548,
+      "step": 383
+    },
+    {
+      "epoch": 0.19064167804393695,
+      "grad_norm": 0.07391007909762363,
+      "learning_rate": 9.949460647496258e-06,
+      "loss": 0.5493,
+      "step": 384
+    },
+    {
+      "epoch": 0.19113814074717636,
+      "grad_norm": 0.07598698030318797,
+      "learning_rate": 9.949183049285884e-06,
+      "loss": 0.5422,
+      "step": 385
+    },
+    {
+      "epoch": 0.19163460345041577,
+      "grad_norm": 0.07479001021475261,
+      "learning_rate": 9.948904694672232e-06,
+      "loss": 0.5607,
+      "step": 386
+    },
+    {
+      "epoch": 0.19213106615365522,
+      "grad_norm": 0.07615526983796699,
+      "learning_rate": 9.94862558369785e-06,
+      "loss": 0.5476,
+      "step": 387
+    },
+    {
+      "epoch": 0.19262752885689463,
+      "grad_norm": 0.08074869204916085,
+      "learning_rate": 9.94834571640539e-06,
+      "loss": 0.5802,
+      "step": 388
+    },
+    {
+      "epoch": 0.19312399156013405,
+      "grad_norm": 0.07415445663241096,
+      "learning_rate": 9.948065092837631e-06,
+      "loss": 0.5499,
+      "step": 389
+    },
+    {
+      "epoch": 0.19362045426337346,
+      "grad_norm": 0.08328624631064929,
+      "learning_rate": 9.947783713037456e-06,
+      "loss": 0.5712,
+      "step": 390
+    },
+    {
+      "epoch": 0.19411691696661287,
+      "grad_norm": 0.07891866140239774,
+      "learning_rate": 9.947501577047874e-06,
+      "loss": 0.5556,
+      "step": 391
+    },
+    {
+      "epoch": 0.19461337966985232,
+      "grad_norm": 0.07476368567789123,
+      "learning_rate": 9.947218684912001e-06,
+      "loss": 0.5371,
+      "step": 392
+    },
+    {
+      "epoch": 0.19510984237309173,
+      "grad_norm": 0.07675097706660683,
+      "learning_rate": 9.946935036673076e-06,
+      "loss": 0.5569,
+      "step": 393
+    },
+    {
+      "epoch": 0.19560630507633114,
+      "grad_norm": 0.07496823454348485,
+      "learning_rate": 9.94665063237445e-06,
+      "loss": 0.5464,
+      "step": 394
+    },
+    {
+      "epoch": 0.19610276777957056,
+      "grad_norm": 0.07818052652954,
+      "learning_rate": 9.946365472059586e-06,
+      "loss": 0.5942,
+      "step": 395
+    },
+    {
+      "epoch": 0.19659923048280997,
+      "grad_norm": 0.07282933318423163,
+      "learning_rate": 9.94607955577207e-06,
+      "loss": 0.5462,
+      "step": 396
+    },
+    {
+      "epoch": 0.1970956931860494,
+      "grad_norm": 0.0770165134901903,
+      "learning_rate": 9.945792883555597e-06,
+      "loss": 0.556,
+      "step": 397
+    },
+    {
+      "epoch": 0.19759215588928883,
+      "grad_norm": 0.07225926792740466,
+      "learning_rate": 9.945505455453983e-06,
+      "loss": 0.5248,
+      "step": 398
+    },
+    {
+      "epoch": 0.19808861859252824,
+      "grad_norm": 0.07352401208203305,
+      "learning_rate": 9.945217271511154e-06,
+      "loss": 0.5481,
+      "step": 399
+    },
+    {
+      "epoch": 0.19858508129576766,
+      "grad_norm": 0.07321353135194356,
+      "learning_rate": 9.944928331771157e-06,
+      "loss": 0.5812,
+      "step": 400
+    },
+    {
+      "epoch": 0.19908154399900707,
+      "grad_norm": 0.07936442126421304,
+      "learning_rate": 9.944638636278148e-06,
+      "loss": 0.5859,
+      "step": 401
+    },
+    {
+      "epoch": 0.19957800670224649,
+      "grad_norm": 0.08462233434104645,
+      "learning_rate": 9.944348185076406e-06,
+      "loss": 0.561,
+      "step": 402
+    },
+    {
+      "epoch": 0.2000744694054859,
+      "grad_norm": 0.07418308136608491,
+      "learning_rate": 9.94405697821032e-06,
+      "loss": 0.539,
+      "step": 403
+    },
+    {
+      "epoch": 0.20057093210872534,
+      "grad_norm": 0.0734957540044604,
+      "learning_rate": 9.9437650157244e-06,
+      "loss": 0.5333,
+      "step": 404
+    },
+    {
+      "epoch": 0.20106739481196476,
+      "grad_norm": 0.07575478601259587,
+      "learning_rate": 9.943472297663262e-06,
+      "loss": 0.5794,
+      "step": 405
+    },
+    {
+      "epoch": 0.20156385751520417,
+      "grad_norm": 0.07710822943994676,
+      "learning_rate": 9.943178824071646e-06,
+      "loss": 0.5631,
+      "step": 406
+    },
+    {
+      "epoch": 0.20206032021844358,
+      "grad_norm": 0.07942163293112221,
+      "learning_rate": 9.942884594994405e-06,
+      "loss": 0.5554,
+      "step": 407
+    },
+    {
+      "epoch": 0.202556782921683,
+      "grad_norm": 0.07804741682603977,
+      "learning_rate": 9.942589610476505e-06,
+      "loss": 0.5989,
+      "step": 408
+    },
+    {
+      "epoch": 0.20305324562492244,
+      "grad_norm": 0.07451188573309668,
+      "learning_rate": 9.942293870563033e-06,
+      "loss": 0.5556,
+      "step": 409
+    },
+    {
+      "epoch": 0.20354970832816185,
+      "grad_norm": 0.07990449182548652,
+      "learning_rate": 9.941997375299187e-06,
+      "loss": 0.5745,
+      "step": 410
+    },
+    {
+      "epoch": 0.20404617103140127,
+      "grad_norm": 0.07288130572381771,
+      "learning_rate": 9.94170012473028e-06,
+      "loss": 0.5495,
+      "step": 411
+    },
+    {
+      "epoch": 0.20454263373464068,
+      "grad_norm": 0.08238928905167173,
+      "learning_rate": 9.941402118901743e-06,
+      "loss": 0.5754,
+      "step": 412
+    },
+    {
+      "epoch": 0.2050390964378801,
+      "grad_norm": 0.07171358910901696,
+      "learning_rate": 9.941103357859123e-06,
+      "loss": 0.5121,
+      "step": 413
+    },
+    {
+      "epoch": 0.2055355591411195,
+      "grad_norm": 0.07325000035957444,
+      "learning_rate": 9.940803841648077e-06,
+      "loss": 0.5408,
+      "step": 414
+    },
+    {
+      "epoch": 0.20603202184435895,
+      "grad_norm": 0.07568269297866982,
+      "learning_rate": 9.940503570314386e-06,
+      "loss": 0.5639,
+      "step": 415
+    },
+    {
+      "epoch": 0.20652848454759837,
+      "grad_norm": 0.07535697622710859,
+      "learning_rate": 9.940202543903939e-06,
+      "loss": 0.5552,
+      "step": 416
+    },
+    {
+      "epoch": 0.20702494725083778,
+      "grad_norm": 0.07271044147735484,
+      "learning_rate": 9.939900762462741e-06,
+      "loss": 0.5388,
+      "step": 417
+    },
+    {
+      "epoch": 0.2075214099540772,
+      "grad_norm": 0.07604343150784937,
+      "learning_rate": 9.939598226036919e-06,
+      "loss": 0.5462,
+      "step": 418
+    },
+    {
+      "epoch": 0.2080178726573166,
+      "grad_norm": 0.07562199878785764,
+      "learning_rate": 9.939294934672707e-06,
+      "loss": 0.5484,
+      "step": 419
+    },
+    {
+      "epoch": 0.20851433536055602,
+      "grad_norm": 0.07652378289509235,
+      "learning_rate": 9.93899088841646e-06,
+      "loss": 0.5564,
+      "step": 420
+    },
+    {
+      "epoch": 0.20901079806379547,
+      "grad_norm": 0.07287884883749866,
+      "learning_rate": 9.938686087314647e-06,
+      "loss": 0.5504,
+      "step": 421
+    },
+    {
+      "epoch": 0.20950726076703488,
+      "grad_norm": 0.07792523226972259,
+      "learning_rate": 9.938380531413851e-06,
+      "loss": 0.587,
+      "step": 422
+    },
+    {
+      "epoch": 0.2100037234702743,
+      "grad_norm": 0.0762025358754494,
+      "learning_rate": 9.938074220760774e-06,
+      "loss": 0.558,
+      "step": 423
+    },
+    {
+      "epoch": 0.2105001861735137,
+      "grad_norm": 0.07611991158494283,
+      "learning_rate": 9.937767155402224e-06,
+      "loss": 0.5858,
+      "step": 424
+    },
+    {
+      "epoch": 0.21099664887675312,
+      "grad_norm": 0.07515544902399915,
+      "learning_rate": 9.937459335385137e-06,
+      "loss": 0.5216,
+      "step": 425
+    },
+    {
+      "epoch": 0.21149311157999257,
+      "grad_norm": 0.07843087477525403,
+      "learning_rate": 9.93715076075656e-06,
+      "loss": 0.5596,
+      "step": 426
+    },
+    {
+      "epoch": 0.21198957428323198,
+      "grad_norm": 0.07327576359699134,
+      "learning_rate": 9.936841431563646e-06,
+      "loss": 0.5336,
+      "step": 427
+    },
+    {
+      "epoch": 0.2124860369864714,
+      "grad_norm": 0.07622638457443663,
+      "learning_rate": 9.936531347853677e-06,
+      "loss": 0.5381,
+      "step": 428
+    },
+    {
+      "epoch": 0.2129824996897108,
+      "grad_norm": 0.07207576415296685,
+      "learning_rate": 9.936220509674044e-06,
+      "loss": 0.5643,
+      "step": 429
+    },
+    {
+      "epoch": 0.21347896239295022,
+      "grad_norm": 0.0754727197146664,
+      "learning_rate": 9.935908917072253e-06,
+      "loss": 0.5492,
+      "step": 430
+    },
+    {
+      "epoch": 0.21397542509618964,
+      "grad_norm": 0.07524418858314785,
+      "learning_rate": 9.935596570095923e-06,
+      "loss": 0.5614,
+      "step": 431
+    },
+    {
+      "epoch": 0.21447188779942908,
+      "grad_norm": 0.07703493076500291,
+      "learning_rate": 9.935283468792793e-06,
+      "loss": 0.5979,
+      "step": 432
+    },
+    {
+      "epoch": 0.2149683505026685,
+      "grad_norm": 0.07870804145032773,
+      "learning_rate": 9.934969613210718e-06,
+      "loss": 0.5497,
+      "step": 433
+    },
+    {
+      "epoch": 0.2154648132059079,
+      "grad_norm": 0.07806957664461764,
+      "learning_rate": 9.934655003397663e-06,
+      "loss": 0.5389,
+      "step": 434
+    },
+    {
+      "epoch": 0.21596127590914732,
+      "grad_norm": 0.08327261700919904,
+      "learning_rate": 9.934339639401712e-06,
+      "loss": 0.5602,
+      "step": 435
+    },
+    {
+      "epoch": 0.21645773861238674,
+      "grad_norm": 0.07699308770199148,
+      "learning_rate": 9.934023521271063e-06,
+      "loss": 0.5289,
+      "step": 436
+    },
+    {
+      "epoch": 0.21695420131562618,
+      "grad_norm": 0.0835512172071361,
+      "learning_rate": 9.93370664905403e-06,
+      "loss": 0.5343,
+      "step": 437
+    },
+    {
+      "epoch": 0.2174506640188656,
+      "grad_norm": 0.08008747971552731,
+      "learning_rate": 9.933389022799042e-06,
+      "loss": 0.5932,
+      "step": 438
+    },
+    {
+      "epoch": 0.217947126722105,
+      "grad_norm": 0.07758508618190278,
+      "learning_rate": 9.933070642554643e-06,
+      "loss": 0.5512,
+      "step": 439
+    },
+    {
+      "epoch": 0.21844358942534442,
+      "grad_norm": 0.07692264693388792,
+      "learning_rate": 9.932751508369492e-06,
+      "loss": 0.577,
+      "step": 440
+    },
+    {
+      "epoch": 0.21894005212858383,
+      "grad_norm": 0.07702699031893523,
+      "learning_rate": 9.932431620292363e-06,
+      "loss": 0.5902,
+      "step": 441
+    },
+    {
+      "epoch": 0.21943651483182325,
+      "grad_norm": 0.072153289238452,
+      "learning_rate": 9.932110978372145e-06,
+      "loss": 0.5765,
+      "step": 442
+    },
+    {
+      "epoch": 0.2199329775350627,
+      "grad_norm": 0.07279626044785806,
+      "learning_rate": 9.931789582657847e-06,
+      "loss": 0.5358,
+      "step": 443
+    },
+    {
+      "epoch": 0.2204294402383021,
+      "grad_norm": 0.07892917204214381,
+      "learning_rate": 9.931467433198585e-06,
+      "loss": 0.5756,
+      "step": 444
+    },
+    {
+      "epoch": 0.22092590294154152,
+      "grad_norm": 0.07994939514502163,
+      "learning_rate": 9.931144530043597e-06,
+      "loss": 0.5978,
+      "step": 445
+    },
+    {
+      "epoch": 0.22142236564478093,
+      "grad_norm": 0.0718213225002427,
+      "learning_rate": 9.93082087324223e-06,
+      "loss": 0.5351,
+      "step": 446
+    },
+    {
+      "epoch": 0.22191882834802035,
+      "grad_norm": 0.07658771671069661,
+      "learning_rate": 9.930496462843954e-06,
+      "loss": 0.5683,
+      "step": 447
+    },
+    {
+      "epoch": 0.22241529105125976,
+      "grad_norm": 0.07738177647845323,
+      "learning_rate": 9.93017129889835e-06,
+      "loss": 0.5711,
+      "step": 448
+    },
+    {
+      "epoch": 0.2229117537544992,
+      "grad_norm": 0.07708589306273327,
+      "learning_rate": 9.92984538145511e-06,
+      "loss": 0.5684,
+      "step": 449
+    },
+    {
+      "epoch": 0.22340821645773862,
+      "grad_norm": 0.07332682077027314,
+      "learning_rate": 9.929518710564048e-06,
+      "loss": 0.5464,
+      "step": 450
+    },
+    {
+      "epoch": 0.22390467916097803,
+      "grad_norm": 0.0765110806185179,
+      "learning_rate": 9.929191286275088e-06,
+      "loss": 0.5874,
+      "step": 451
+    },
+    {
+      "epoch": 0.22440114186421745,
+      "grad_norm": 0.07381967262364783,
+      "learning_rate": 9.928863108638275e-06,
+      "loss": 0.545,
+      "step": 452
+    },
+    {
+      "epoch": 0.22489760456745686,
+      "grad_norm": 0.11056903774769831,
+      "learning_rate": 9.928534177703766e-06,
+      "loss": 0.5859,
+      "step": 453
+    },
+    {
+      "epoch": 0.2253940672706963,
+      "grad_norm": 0.07627989286727224,
+      "learning_rate": 9.92820449352183e-06,
+      "loss": 0.5719,
+      "step": 454
+    },
+    {
+      "epoch": 0.22589052997393572,
+      "grad_norm": 0.07573328286843277,
+      "learning_rate": 9.927874056142854e-06,
+      "loss": 0.5622,
+      "step": 455
+    },
+    {
+      "epoch": 0.22638699267717513,
+      "grad_norm": 0.07510309770011216,
+      "learning_rate": 9.92754286561734e-06,
+      "loss": 0.5315,
+      "step": 456
+    },
+    {
+      "epoch": 0.22688345538041454,
+      "grad_norm": 0.07782909898671676,
+      "learning_rate": 9.927210921995908e-06,
+      "loss": 0.5428,
+      "step": 457
+    },
+    {
+      "epoch": 0.22737991808365396,
+      "grad_norm": 0.07580841682733405,
+      "learning_rate": 9.926878225329288e-06,
+      "loss": 0.5477,
+      "step": 458
+    },
+    {
+      "epoch": 0.22787638078689337,
+      "grad_norm": 0.0771663444644485,
+      "learning_rate": 9.926544775668327e-06,
+      "loss": 0.5514,
+      "step": 459
+    },
+    {
+      "epoch": 0.22837284349013282,
+      "grad_norm": 0.07642125797083227,
+      "learning_rate": 9.926210573063988e-06,
+      "loss": 0.5357,
+      "step": 460
+    },
+    {
+      "epoch": 0.22886930619337223,
+      "grad_norm": 0.07302968349139202,
+      "learning_rate": 9.92587561756735e-06,
+      "loss": 0.5141,
+      "step": 461
+    },
+    {
+      "epoch": 0.22936576889661164,
+      "grad_norm": 0.07237104897439163,
+      "learning_rate": 9.925539909229604e-06,
+      "loss": 0.566,
+      "step": 462
+    },
+    {
+      "epoch": 0.22986223159985106,
+      "grad_norm": 0.07717542777821745,
+      "learning_rate": 9.925203448102058e-06,
+      "loss": 0.5367,
+      "step": 463
+    },
+    {
+      "epoch": 0.23035869430309047,
+      "grad_norm": 0.07671951294820442,
+      "learning_rate": 9.924866234236134e-06,
+      "loss": 0.5584,
+      "step": 464
+    },
+    {
+      "epoch": 0.2308551570063299,
+      "grad_norm": 0.07746481421960554,
+      "learning_rate": 9.924528267683372e-06,
+      "loss": 0.5457,
+      "step": 465
+    },
+    {
+      "epoch": 0.23135161970956933,
+      "grad_norm": 0.0710642404833701,
+      "learning_rate": 9.92418954849542e-06,
+      "loss": 0.5095,
+      "step": 466
+    },
+    {
+      "epoch": 0.23184808241280874,
+      "grad_norm": 0.07191806610081845,
+      "learning_rate": 9.923850076724051e-06,
+      "loss": 0.5267,
+      "step": 467
+    },
+    {
+      "epoch": 0.23234454511604816,
+      "grad_norm": 0.07714488111055401,
+      "learning_rate": 9.923509852421144e-06,
+      "loss": 0.5202,
+      "step": 468
+    },
+    {
+      "epoch": 0.23284100781928757,
+      "grad_norm": 0.07325260373990966,
+      "learning_rate": 9.923168875638701e-06,
+      "loss": 0.5433,
+      "step": 469
+    },
+    {
+      "epoch": 0.23333747052252699,
+      "grad_norm": 0.07665478368369943,
+      "learning_rate": 9.92282714642883e-06,
+      "loss": 0.5683,
+      "step": 470
+    },
+    {
+      "epoch": 0.23383393322576643,
+      "grad_norm": 0.07516189982030862,
+      "learning_rate": 9.922484664843763e-06,
+      "loss": 0.5862,
+      "step": 471
+    },
+    {
+      "epoch": 0.23433039592900584,
+      "grad_norm": 0.07760388755225243,
+      "learning_rate": 9.922141430935842e-06,
+      "loss": 0.5648,
+      "step": 472
+    },
+    {
+      "epoch": 0.23482685863224526,
+      "grad_norm": 0.07609970967938699,
+      "learning_rate": 9.921797444757521e-06,
+      "loss": 0.5241,
+      "step": 473
+    },
+    {
+      "epoch": 0.23532332133548467,
+      "grad_norm": 0.07592265968380732,
+      "learning_rate": 9.921452706361376e-06,
+      "loss": 0.5935,
+      "step": 474
+    },
+    {
+      "epoch": 0.23581978403872408,
+      "grad_norm": 0.07410650337647612,
+      "learning_rate": 9.921107215800095e-06,
+      "loss": 0.543,
+      "step": 475
+    },
+    {
+      "epoch": 0.2363162467419635,
+      "grad_norm": 0.07506294240734249,
+      "learning_rate": 9.92076097312648e-06,
+      "loss": 0.5379,
+      "step": 476
+    },
+    {
+      "epoch": 0.23681270944520294,
+      "grad_norm": 0.07440602775514321,
+      "learning_rate": 9.920413978393449e-06,
+      "loss": 0.5319,
+      "step": 477
+    },
+    {
+      "epoch": 0.23730917214844235,
+      "grad_norm": 0.07528653646912088,
+      "learning_rate": 9.920066231654035e-06,
+      "loss": 0.5447,
+      "step": 478
+    },
+    {
+      "epoch": 0.23780563485168177,
+      "grad_norm": 0.07376070732880129,
+      "learning_rate": 9.919717732961383e-06,
+      "loss": 0.536,
+      "step": 479
+    },
+    {
+      "epoch": 0.23830209755492118,
+      "grad_norm": 0.0742063997969741,
+      "learning_rate": 9.919368482368758e-06,
+      "loss": 0.5376,
+      "step": 480
+    },
+    {
+      "epoch": 0.2387985602581606,
+      "grad_norm": 0.07499906235302634,
+      "learning_rate": 9.919018479929536e-06,
+      "loss": 0.5609,
+      "step": 481
+    },
+    {
+      "epoch": 0.2392950229614,
+      "grad_norm": 0.07489207589373727,
+      "learning_rate": 9.91866772569721e-06,
+      "loss": 0.5582,
+      "step": 482
+    },
+    {
+      "epoch": 0.23979148566463945,
+      "grad_norm": 0.07892405072079473,
+      "learning_rate": 9.918316219725388e-06,
+      "loss": 0.554,
+      "step": 483
+    },
+    {
+      "epoch": 0.24028794836787887,
+      "grad_norm": 0.07499498235806762,
+      "learning_rate": 9.91796396206779e-06,
+      "loss": 0.5505,
+      "step": 484
+    },
+    {
+      "epoch": 0.24078441107111828,
+      "grad_norm": 0.08189343597090869,
+      "learning_rate": 9.917610952778253e-06,
+      "loss": 0.569,
+      "step": 485
+    },
+    {
+      "epoch": 0.2412808737743577,
+      "grad_norm": 0.07400601949803445,
+      "learning_rate": 9.917257191910732e-06,
+      "loss": 0.5384,
+      "step": 486
+    },
+    {
+      "epoch": 0.2417773364775971,
+      "grad_norm": 0.07901731252692805,
+      "learning_rate": 9.91690267951929e-06,
+      "loss": 0.5531,
+      "step": 487
+    },
+    {
+      "epoch": 0.24227379918083655,
+      "grad_norm": 0.07870573377666493,
+      "learning_rate": 9.916547415658111e-06,
+      "loss": 0.5674,
+      "step": 488
+    },
+    {
+      "epoch": 0.24277026188407597,
+      "grad_norm": 0.07911850022053922,
+      "learning_rate": 9.91619140038149e-06,
+      "loss": 0.5721,
+      "step": 489
+    },
+    {
+      "epoch": 0.24326672458731538,
+      "grad_norm": 0.0761702583766454,
+      "learning_rate": 9.915834633743838e-06,
+      "loss": 0.5678,
+      "step": 490
+    },
+    {
+      "epoch": 0.2437631872905548,
+      "grad_norm": 0.07665990511246726,
+      "learning_rate": 9.915477115799682e-06,
+      "loss": 0.5718,
+      "step": 491
+    },
+    {
+      "epoch": 0.2442596499937942,
+      "grad_norm": 0.07872700425419384,
+      "learning_rate": 9.915118846603661e-06,
+      "loss": 0.5889,
+      "step": 492
+    },
+    {
+      "epoch": 0.24475611269703362,
+      "grad_norm": 0.07395617425590947,
+      "learning_rate": 9.914759826210534e-06,
+      "loss": 0.5295,
+      "step": 493
+    },
+    {
+      "epoch": 0.24525257540027307,
+      "grad_norm": 0.07483689041527114,
+      "learning_rate": 9.91440005467517e-06,
+      "loss": 0.5562,
+      "step": 494
+    },
+    {
+      "epoch": 0.24574903810351248,
+      "grad_norm": 0.07604027911059567,
+      "learning_rate": 9.914039532052553e-06,
+      "loss": 0.5269,
+      "step": 495
+    },
+    {
+      "epoch": 0.2462455008067519,
+      "grad_norm": 0.08297691072826975,
+      "learning_rate": 9.913678258397785e-06,
+      "loss": 0.559,
+      "step": 496
+    },
+    {
+      "epoch": 0.2467419635099913,
+      "grad_norm": 0.08073844389509381,
+      "learning_rate": 9.91331623376608e-06,
+      "loss": 0.5625,
+      "step": 497
+    },
+    {
+      "epoch": 0.24723842621323072,
+      "grad_norm": 0.07564059049053361,
+      "learning_rate": 9.912953458212769e-06,
+      "loss": 0.5461,
+      "step": 498
+    },
+    {
+      "epoch": 0.24773488891647016,
+      "grad_norm": 0.07604428278991741,
+      "learning_rate": 9.912589931793294e-06,
+      "loss": 0.5307,
+      "step": 499
+    },
+    {
+      "epoch": 0.24823135161970958,
+      "grad_norm": 0.08041158897596114,
+      "learning_rate": 9.912225654563214e-06,
+      "loss": 0.5683,
+      "step": 500
+    },
+    {
+      "epoch": 0.248727814322949,
+      "grad_norm": 0.07384877613143606,
+      "learning_rate": 9.911860626578204e-06,
+      "loss": 0.5493,
+      "step": 501
+    },
+    {
+      "epoch": 0.2492242770261884,
+      "grad_norm": 0.07510188927857986,
+      "learning_rate": 9.911494847894055e-06,
+      "loss": 0.5718,
+      "step": 502
+    },
+    {
+      "epoch": 0.24972073972942782,
+      "grad_norm": 0.07700133387805451,
+      "learning_rate": 9.911128318566668e-06,
+      "loss": 0.5205,
+      "step": 503
+    },
+    {
+      "epoch": 0.25021720243266726,
+      "grad_norm": 0.08281754395286774,
+      "learning_rate": 9.91076103865206e-06,
+      "loss": 0.5612,
+      "step": 504
+    },
+    {
+      "epoch": 0.25021720243266726,
+      "eval_loss": 0.5554779171943665,
+      "eval_runtime": 259.1788,
+      "eval_samples_per_second": 117.112,
+      "eval_steps_per_second": 14.642,
+      "step": 504
+    },
+    {
+      "epoch": 0.25071366513590665,
+      "grad_norm": 0.07858966858013353,
+      "learning_rate": 9.910393008206367e-06,
+      "loss": 0.5472,
+      "step": 505
+    },
+    {
+      "epoch": 0.2512101278391461,
+      "grad_norm": 0.07292630081923775,
+      "learning_rate": 9.910024227285832e-06,
+      "loss": 0.5426,
+      "step": 506
+    },
+    {
+      "epoch": 0.2517065905423855,
+      "grad_norm": 0.08028416640572211,
+      "learning_rate": 9.909654695946823e-06,
+      "loss": 0.607,
+      "step": 507
+    },
+    {
+      "epoch": 0.2522030532456249,
+      "grad_norm": 0.0817807763355325,
+      "learning_rate": 9.909284414245815e-06,
+      "loss": 0.598,
+      "step": 508
+    },
+    {
+      "epoch": 0.25269951594886436,
+      "grad_norm": 0.08037659726296809,
+      "learning_rate": 9.908913382239396e-06,
+      "loss": 0.5399,
+      "step": 509
+    },
+    {
+      "epoch": 0.25319597865210375,
+      "grad_norm": 0.07398564863526783,
+      "learning_rate": 9.908541599984276e-06,
+      "loss": 0.5296,
+      "step": 510
+    },
+    {
+      "epoch": 0.2536924413553432,
+      "grad_norm": 0.07735462650188997,
+      "learning_rate": 9.908169067537274e-06,
+      "loss": 0.5356,
+      "step": 511
+    },
+    {
+      "epoch": 0.2541889040585826,
+      "grad_norm": 0.08361063862669793,
+      "learning_rate": 9.907795784955327e-06,
+      "loss": 0.5265,
+      "step": 512
+    },
+    {
+      "epoch": 0.254685366761822,
+      "grad_norm": 0.07798942247321251,
+      "learning_rate": 9.907421752295485e-06,
+      "loss": 0.5287,
+      "step": 513
+    },
+    {
+      "epoch": 0.25518182946506146,
+      "grad_norm": 0.07672209798270986,
+      "learning_rate": 9.907046969614913e-06,
+      "loss": 0.5767,
+      "step": 514
+    },
+    {
+      "epoch": 0.25567829216830085,
+      "grad_norm": 0.07957972986175162,
+      "learning_rate": 9.906671436970891e-06,
+      "loss": 0.5744,
+      "step": 515
+    },
+    {
+      "epoch": 0.2561747548715403,
+      "grad_norm": 0.0846841236085876,
+      "learning_rate": 9.906295154420811e-06,
+      "loss": 0.6052,
+      "step": 516
+    },
+    {
+      "epoch": 0.2566712175747797,
+      "grad_norm": 0.07794538960974101,
+      "learning_rate": 9.905918122022183e-06,
+      "loss": 0.5237,
+      "step": 517
+    },
+    {
+      "epoch": 0.2571676802780191,
+      "grad_norm": 0.07389212609092594,
+      "learning_rate": 9.905540339832632e-06,
+      "loss": 0.5577,
+      "step": 518
+    },
+    {
+      "epoch": 0.25766414298125856,
+      "grad_norm": 0.0747315216036527,
+      "learning_rate": 9.905161807909893e-06,
+      "loss": 0.5305,
+      "step": 519
+    },
+    {
+      "epoch": 0.25816060568449795,
+      "grad_norm": 0.0738348078705221,
+      "learning_rate": 9.90478252631182e-06,
+      "loss": 0.5486,
+      "step": 520
+    },
+    {
+      "epoch": 0.2586570683877374,
+      "grad_norm": 0.07980710964169403,
+      "learning_rate": 9.90440249509638e-06,
+      "loss": 0.5498,
+      "step": 521
+    },
+    {
+      "epoch": 0.2591535310909768,
+      "grad_norm": 0.07444828843776845,
+      "learning_rate": 9.904021714321656e-06,
+      "loss": 0.5437,
+      "step": 522
+    },
+    {
+      "epoch": 0.2596499937942162,
+      "grad_norm": 0.07359542898272906,
+      "learning_rate": 9.903640184045842e-06,
+      "loss": 0.5591,
+      "step": 523
+    },
+    {
+      "epoch": 0.2601464564974556,
+      "grad_norm": 0.07496639130963002,
+      "learning_rate": 9.90325790432725e-06,
+      "loss": 0.5553,
+      "step": 524
+    },
+    {
+      "epoch": 0.26064291920069504,
+      "grad_norm": 0.07983536927042992,
+      "learning_rate": 9.902874875224305e-06,
+      "loss": 0.575,
+      "step": 525
+    },
+    {
+      "epoch": 0.2611393819039345,
+      "grad_norm": 0.08616908954969195,
+      "learning_rate": 9.902491096795546e-06,
+      "loss": 0.5634,
+      "step": 526
+    },
+    {
+      "epoch": 0.2616358446071739,
+      "grad_norm": 0.07576389919139324,
+      "learning_rate": 9.90210656909963e-06,
+      "loss": 0.5424,
+      "step": 527
+    },
+    {
+      "epoch": 0.2621323073104133,
+      "grad_norm": 0.07910649495763554,
+      "learning_rate": 9.901721292195323e-06,
+      "loss": 0.5469,
+      "step": 528
+    },
+    {
+      "epoch": 0.2626287700136527,
+      "grad_norm": 0.08224845731683662,
+      "learning_rate": 9.90133526614151e-06,
+      "loss": 0.5707,
+      "step": 529
+    },
+    {
+      "epoch": 0.26312523271689214,
+      "grad_norm": 0.07698783008012512,
+      "learning_rate": 9.900948490997188e-06,
+      "loss": 0.51,
+      "step": 530
+    },
+    {
+      "epoch": 0.2636216954201316,
+      "grad_norm": 0.07524972064564883,
+      "learning_rate": 9.90056096682147e-06,
+      "loss": 0.5452,
+      "step": 531
+    },
+    {
+      "epoch": 0.26411815812337097,
+      "grad_norm": 0.08116567664979311,
+      "learning_rate": 9.900172693673584e-06,
+      "loss": 0.5794,
+      "step": 532
+    },
+    {
+      "epoch": 0.2646146208266104,
+      "grad_norm": 0.07193670033849783,
+      "learning_rate": 9.899783671612868e-06,
+      "loss": 0.5361,
+      "step": 533
+    },
+    {
+      "epoch": 0.2651110835298498,
+      "grad_norm": 0.08234755603848125,
+      "learning_rate": 9.899393900698781e-06,
+      "loss": 0.5482,
+      "step": 534
+    },
+    {
+      "epoch": 0.26560754623308924,
+      "grad_norm": 0.07336446642087938,
+      "learning_rate": 9.899003380990893e-06,
+      "loss": 0.5652,
+      "step": 535
+    },
+    {
+      "epoch": 0.2661040089363287,
+      "grad_norm": 0.07633905125421954,
+      "learning_rate": 9.898612112548886e-06,
+      "loss": 0.5608,
+      "step": 536
+    },
+    {
+      "epoch": 0.26660047163956807,
+      "grad_norm": 0.07758130640934637,
+      "learning_rate": 9.898220095432562e-06,
+      "loss": 0.567,
+      "step": 537
+    },
+    {
+      "epoch": 0.2670969343428075,
+      "grad_norm": 0.07536999521216046,
+      "learning_rate": 9.897827329701834e-06,
+      "loss": 0.5478,
+      "step": 538
+    },
+    {
+      "epoch": 0.2675933970460469,
+      "grad_norm": 0.07343119222091014,
+      "learning_rate": 9.897433815416729e-06,
+      "loss": 0.5643,
+      "step": 539
+    },
+    {
+      "epoch": 0.26808985974928634,
+      "grad_norm": 0.07673148700528902,
+      "learning_rate": 9.897039552637389e-06,
+      "loss": 0.5488,
+      "step": 540
+    },
+    {
+      "epoch": 0.2685863224525257,
+      "grad_norm": 0.07898713801424365,
+      "learning_rate": 9.896644541424071e-06,
+      "loss": 0.5334,
+      "step": 541
+    },
+    {
+      "epoch": 0.26908278515576517,
+      "grad_norm": 0.07708722025335718,
+      "learning_rate": 9.896248781837148e-06,
+      "loss": 0.6273,
+      "step": 542
+    },
+    {
+      "epoch": 0.2695792478590046,
+      "grad_norm": 0.07444543294418082,
+      "learning_rate": 9.895852273937103e-06,
+      "loss": 0.5379,
+      "step": 543
+    },
+    {
+      "epoch": 0.270075710562244,
+      "grad_norm": 0.07292472383292499,
+      "learning_rate": 9.895455017784536e-06,
+      "loss": 0.5395,
+      "step": 544
+    },
+    {
+      "epoch": 0.27057217326548344,
+      "grad_norm": 0.0765918576393523,
+      "learning_rate": 9.895057013440163e-06,
+      "loss": 0.5501,
+      "step": 545
+    },
+    {
+      "epoch": 0.2710686359687228,
+      "grad_norm": 0.07405716714879264,
+      "learning_rate": 9.894658260964814e-06,
+      "loss": 0.512,
+      "step": 546
+    },
+    {
+      "epoch": 0.27156509867196227,
+      "grad_norm": 0.07477208653408075,
+      "learning_rate": 9.894258760419427e-06,
+      "loss": 0.5439,
+      "step": 547
+    },
+    {
+      "epoch": 0.2720615613752017,
+      "grad_norm": 0.11975148352494779,
+      "learning_rate": 9.893858511865063e-06,
+      "loss": 0.5547,
+      "step": 548
+    },
+    {
+      "epoch": 0.2725580240784411,
+      "grad_norm": 0.07579011485998241,
+      "learning_rate": 9.893457515362892e-06,
+      "loss": 0.568,
+      "step": 549
+    },
+    {
+      "epoch": 0.27305448678168054,
+      "grad_norm": 0.07553870503848775,
+      "learning_rate": 9.893055770974202e-06,
+      "loss": 0.514,
+      "step": 550
+    },
+    {
+      "epoch": 0.2735509494849199,
+      "grad_norm": 0.07576057451860235,
+      "learning_rate": 9.892653278760389e-06,
+      "loss": 0.5584,
+      "step": 551
+    },
+    {
+      "epoch": 0.27404741218815937,
+      "grad_norm": 0.07550325207874099,
+      "learning_rate": 9.892250038782972e-06,
+      "loss": 0.5471,
+      "step": 552
+    },
+    {
+      "epoch": 0.2745438748913988,
+      "grad_norm": 0.07554409900790758,
+      "learning_rate": 9.891846051103578e-06,
+      "loss": 0.547,
+      "step": 553
+    },
+    {
+      "epoch": 0.2750403375946382,
+      "grad_norm": 0.07453073205900222,
+      "learning_rate": 9.89144131578395e-06,
+      "loss": 0.5599,
+      "step": 554
+    },
+    {
+      "epoch": 0.27553680029787764,
+      "grad_norm": 0.07678267422238774,
+      "learning_rate": 9.891035832885942e-06,
+      "loss": 0.5784,
+      "step": 555
+    },
+    {
+      "epoch": 0.276033263001117,
+      "grad_norm": 0.07675969130014183,
+      "learning_rate": 9.890629602471532e-06,
+      "loss": 0.5526,
+      "step": 556
+    },
+    {
+      "epoch": 0.27652972570435647,
+      "grad_norm": 0.07974318262948232,
+      "learning_rate": 9.8902226246028e-06,
+      "loss": 0.576,
+      "step": 557
+    },
+    {
+      "epoch": 0.27702618840759585,
+      "grad_norm": 0.08261341445162391,
+      "learning_rate": 9.889814899341951e-06,
+      "loss": 0.5915,
+      "step": 558
+    },
+    {
+      "epoch": 0.2775226511108353,
+      "grad_norm": 0.08009330000358027,
+      "learning_rate": 9.889406426751296e-06,
+      "loss": 0.5485,
+      "step": 559
+    },
+    {
+      "epoch": 0.27801911381407474,
+      "grad_norm": 0.07270863892033039,
+      "learning_rate": 9.888997206893266e-06,
+      "loss": 0.5429,
+      "step": 560
+    },
+    {
+      "epoch": 0.2785155765173141,
+      "grad_norm": 0.07411980046220777,
+      "learning_rate": 9.8885872398304e-06,
+      "loss": 0.5412,
+      "step": 561
+    },
+    {
+      "epoch": 0.27901203922055356,
+      "grad_norm": 0.07859149685871875,
+      "learning_rate": 9.888176525625358e-06,
+      "loss": 0.5605,
+      "step": 562
+    },
+    {
+      "epoch": 0.27950850192379295,
+      "grad_norm": 0.0750640057546601,
+      "learning_rate": 9.887765064340909e-06,
+      "loss": 0.5223,
+      "step": 563
+    },
+    {
+      "epoch": 0.2800049646270324,
+      "grad_norm": 0.07240010518351624,
+      "learning_rate": 9.88735285603994e-06,
+      "loss": 0.5534,
+      "step": 564
+    },
+    {
+      "epoch": 0.28050142733027184,
+      "grad_norm": 0.0749384510617718,
+      "learning_rate": 9.886939900785448e-06,
+      "loss": 0.5472,
+      "step": 565
+    },
+    {
+      "epoch": 0.2809978900335112,
+      "grad_norm": 0.07800629633952484,
+      "learning_rate": 9.88652619864055e-06,
+      "loss": 0.5542,
+      "step": 566
+    },
+    {
+      "epoch": 0.28149435273675066,
+      "grad_norm": 0.07478324010808358,
+      "learning_rate": 9.886111749668472e-06,
+      "loss": 0.508,
+      "step": 567
+    },
+    {
+      "epoch": 0.28199081543999005,
+      "grad_norm": 0.07731985483902606,
+      "learning_rate": 9.885696553932556e-06,
+      "loss": 0.5436,
+      "step": 568
+    },
+    {
+      "epoch": 0.2824872781432295,
+      "grad_norm": 0.0750464125046347,
+      "learning_rate": 9.885280611496256e-06,
+      "loss": 0.5576,
+      "step": 569
+    },
+    {
+      "epoch": 0.28298374084646893,
+      "grad_norm": 0.07461691953178973,
+      "learning_rate": 9.884863922423147e-06,
+      "loss": 0.5685,
+      "step": 570
+    },
+    {
+      "epoch": 0.2834802035497083,
+      "grad_norm": 0.0794903059853572,
+      "learning_rate": 9.884446486776908e-06,
+      "loss": 0.5607,
+      "step": 571
+    },
+    {
+      "epoch": 0.28397666625294776,
+      "grad_norm": 0.07287052459802834,
+      "learning_rate": 9.884028304621341e-06,
+      "loss": 0.5501,
+      "step": 572
+    },
+    {
+      "epoch": 0.28447312895618715,
+      "grad_norm": 0.0788064114434013,
+      "learning_rate": 9.883609376020356e-06,
+      "loss": 0.5665,
+      "step": 573
+    },
+    {
+      "epoch": 0.2849695916594266,
+      "grad_norm": 0.07663493649905155,
+      "learning_rate": 9.883189701037981e-06,
+      "loss": 0.6092,
+      "step": 574
+    },
+    {
+      "epoch": 0.285466054362666,
+      "grad_norm": 0.07479983199950002,
+      "learning_rate": 9.882769279738355e-06,
+      "loss": 0.5612,
+      "step": 575
+    },
+    {
+      "epoch": 0.2859625170659054,
+      "grad_norm": 0.07959254652565169,
+      "learning_rate": 9.882348112185736e-06,
+      "loss": 0.5183,
+      "step": 576
+    },
+    {
+      "epoch": 0.28645897976914486,
+      "grad_norm": 0.0768553794284454,
+      "learning_rate": 9.881926198444489e-06,
+      "loss": 0.5195,
+      "step": 577
+    },
+    {
+      "epoch": 0.28695544247238425,
+      "grad_norm": 0.07481572126590574,
+      "learning_rate": 9.8815035385791e-06,
+      "loss": 0.5684,
+      "step": 578
+    },
+    {
+      "epoch": 0.2874519051756237,
+      "grad_norm": 0.07493217713328718,
+      "learning_rate": 9.881080132654163e-06,
+      "loss": 0.5947,
+      "step": 579
+    },
+    {
+      "epoch": 0.2879483678788631,
+      "grad_norm": 0.07507225285212296,
+      "learning_rate": 9.880655980734391e-06,
+      "loss": 0.5543,
+      "step": 580
+    },
+    {
+      "epoch": 0.2884448305821025,
+      "grad_norm": 0.07904490711995418,
+      "learning_rate": 9.880231082884605e-06,
+      "loss": 0.5551,
+      "step": 581
+    },
+    {
+      "epoch": 0.28894129328534196,
+      "grad_norm": 0.07295892810312825,
+      "learning_rate": 9.87980543916975e-06,
+      "loss": 0.5464,
+      "step": 582
+    },
+    {
+      "epoch": 0.28943775598858135,
+      "grad_norm": 0.07508250816956526,
+      "learning_rate": 9.879379049654872e-06,
+      "loss": 0.5399,
+      "step": 583
+    },
+    {
+      "epoch": 0.2899342186918208,
+      "grad_norm": 0.07218965100235024,
+      "learning_rate": 9.878951914405144e-06,
+      "loss": 0.5525,
+      "step": 584
+    },
+    {
+      "epoch": 0.2904306813950602,
+      "grad_norm": 0.07712208747682248,
+      "learning_rate": 9.878524033485843e-06,
+      "loss": 0.6231,
+      "step": 585
+    },
+    {
+      "epoch": 0.2909271440982996,
+      "grad_norm": 0.07426488021803046,
+      "learning_rate": 9.878095406962364e-06,
+      "loss": 0.5467,
+      "step": 586
+    },
+    {
+      "epoch": 0.29142360680153906,
+      "grad_norm": 0.07595195548804266,
+      "learning_rate": 9.877666034900216e-06,
+      "loss": 0.5568,
+      "step": 587
+    },
+    {
+      "epoch": 0.29192006950477845,
+      "grad_norm": 0.07309985737679185,
+      "learning_rate": 9.877235917365022e-06,
+      "loss": 0.5023,
+      "step": 588
+    },
+    {
+      "epoch": 0.2924165322080179,
+      "grad_norm": 0.07629964951526776,
+      "learning_rate": 9.87680505442252e-06,
+      "loss": 0.5656,
+      "step": 589
+    },
+    {
+      "epoch": 0.2929129949112573,
+      "grad_norm": 0.07525170076872399,
+      "learning_rate": 9.876373446138559e-06,
+      "loss": 0.5583,
+      "step": 590
+    },
+    {
+      "epoch": 0.2934094576144967,
+      "grad_norm": 0.0719482938902049,
+      "learning_rate": 9.875941092579102e-06,
+      "loss": 0.5843,
+      "step": 591
+    },
+    {
+      "epoch": 0.2939059203177361,
+      "grad_norm": 0.07697109291618406,
+      "learning_rate": 9.87550799381023e-06,
+      "loss": 0.5399,
+      "step": 592
+    },
+    {
+      "epoch": 0.29440238302097554,
+      "grad_norm": 0.0767166776402075,
+      "learning_rate": 9.875074149898133e-06,
+      "loss": 0.5276,
+      "step": 593
+    },
+    {
+      "epoch": 0.294898845724215,
+      "grad_norm": 0.07102348146377147,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.5192,
+      "step": 594
+    },
+    {
+      "epoch": 0.2953953084274544,
+      "grad_norm": 0.07279805219484177,
+      "learning_rate": 9.874204226909607e-06,
+      "loss": 0.5372,
+      "step": 595
+    },
+    {
+      "epoch": 0.2958917711306938,
+      "grad_norm": 0.07433801339092601,
+      "learning_rate": 9.87376814796613e-06,
+      "loss": 0.5315,
+      "step": 596
+    },
+    {
+      "epoch": 0.2963882338339332,
+      "grad_norm": 0.07657789334389173,
+      "learning_rate": 9.873331324145337e-06,
+      "loss": 0.5445,
+      "step": 597
+    },
+    {
+      "epoch": 0.29688469653717264,
+      "grad_norm": 0.07705144270666044,
+      "learning_rate": 9.872893755513987e-06,
+      "loss": 0.5607,
+      "step": 598
+    },
+    {
+      "epoch": 0.2973811592404121,
+      "grad_norm": 0.07917616577221262,
+      "learning_rate": 9.872455442138962e-06,
+      "loss": 0.6034,
+      "step": 599
+    },
+    {
+      "epoch": 0.29787762194365147,
+      "grad_norm": 0.07628640882027855,
+      "learning_rate": 9.872016384087243e-06,
+      "loss": 0.5514,
+      "step": 600
+    },
+    {
+      "epoch": 0.2983740846468909,
+      "grad_norm": 0.07818376880862994,
+      "learning_rate": 9.871576581425937e-06,
+      "loss": 0.6002,
+      "step": 601
+    },
+    {
+      "epoch": 0.2988705473501303,
+      "grad_norm": 0.07665596539606143,
+      "learning_rate": 9.871136034222262e-06,
+      "loss": 0.5429,
+      "step": 602
+    },
+    {
+      "epoch": 0.29936701005336974,
+      "grad_norm": 0.07325381782592988,
+      "learning_rate": 9.870694742543544e-06,
+      "loss": 0.5471,
+      "step": 603
+    },
+    {
+      "epoch": 0.2998634727566092,
+      "grad_norm": 0.07639459489392834,
+      "learning_rate": 9.870252706457233e-06,
+      "loss": 0.5711,
+      "step": 604
+    },
+    {
+      "epoch": 0.30035993545984857,
+      "grad_norm": 0.07623059529194105,
+      "learning_rate": 9.869809926030883e-06,
+      "loss": 0.5278,
+      "step": 605
+    },
+    {
+      "epoch": 0.300856398163088,
+      "grad_norm": 0.08082827370502862,
+      "learning_rate": 9.869366401332169e-06,
+      "loss": 0.5308,
+      "step": 606
+    },
+    {
+      "epoch": 0.3013528608663274,
+      "grad_norm": 0.07480986187389814,
+      "learning_rate": 9.868922132428871e-06,
+      "loss": 0.5428,
+      "step": 607
+    },
+    {
+      "epoch": 0.30184932356956684,
+      "grad_norm": 0.07518561564283756,
+      "learning_rate": 9.868477119388897e-06,
+      "loss": 0.5774,
+      "step": 608
+    },
+    {
+      "epoch": 0.3023457862728063,
+      "grad_norm": 0.07628450056698415,
+      "learning_rate": 9.868031362280253e-06,
+      "loss": 0.5611,
+      "step": 609
+    },
+    {
+      "epoch": 0.30284224897604567,
+      "grad_norm": 0.0766015170643728,
+      "learning_rate": 9.867584861171067e-06,
+      "loss": 0.5309,
+      "step": 610
+    },
+    {
+      "epoch": 0.3033387116792851,
+      "grad_norm": 0.07643369967531052,
+      "learning_rate": 9.867137616129583e-06,
+      "loss": 0.5623,
+      "step": 611
+    },
+    {
+      "epoch": 0.3038351743825245,
+      "grad_norm": 0.0728482813967606,
+      "learning_rate": 9.866689627224152e-06,
+      "loss": 0.5233,
+      "step": 612
+    },
+    {
+      "epoch": 0.30433163708576394,
+      "grad_norm": 0.08393283031423471,
+      "learning_rate": 9.866240894523242e-06,
+      "loss": 0.5962,
+      "step": 613
+    },
+    {
+      "epoch": 0.3048280997890033,
+      "grad_norm": 0.0771796216168662,
+      "learning_rate": 9.865791418095437e-06,
+      "loss": 0.5253,
+      "step": 614
+    },
+    {
+      "epoch": 0.30532456249224277,
+      "grad_norm": 0.08018751861730207,
+      "learning_rate": 9.86534119800943e-06,
+      "loss": 0.517,
+      "step": 615
+    },
+    {
+      "epoch": 0.3058210251954822,
+      "grad_norm": 0.07787379590523566,
+      "learning_rate": 9.864890234334032e-06,
+      "loss": 0.5463,
+      "step": 616
+    },
+    {
+      "epoch": 0.3063174878987216,
+      "grad_norm": 0.07763330346534424,
+      "learning_rate": 9.864438527138163e-06,
+      "loss": 0.5793,
+      "step": 617
+    },
+    {
+      "epoch": 0.30681395060196104,
+      "grad_norm": 0.07826557061221215,
+      "learning_rate": 9.86398607649086e-06,
+      "loss": 0.5621,
+      "step": 618
+    },
+    {
+      "epoch": 0.3073104133052004,
+      "grad_norm": 0.07926385658236025,
+      "learning_rate": 9.863532882461275e-06,
+      "loss": 0.5326,
+      "step": 619
+    },
+    {
+      "epoch": 0.30780687600843987,
+      "grad_norm": 0.07685251032990977,
+      "learning_rate": 9.86307894511867e-06,
+      "loss": 0.5696,
+      "step": 620
+    },
+    {
+      "epoch": 0.3083033387116793,
+      "grad_norm": 0.08400275382323562,
+      "learning_rate": 9.86262426453242e-06,
+      "loss": 0.5638,
+      "step": 621
+    },
+    {
+      "epoch": 0.3087998014149187,
+      "grad_norm": 0.07637247838560719,
+      "learning_rate": 9.862168840772018e-06,
+      "loss": 0.5353,
+      "step": 622
+    },
+    {
+      "epoch": 0.30929626411815814,
+      "grad_norm": 0.07567155856874061,
+      "learning_rate": 9.861712673907067e-06,
+      "loss": 0.5624,
+      "step": 623
+    },
+    {
+      "epoch": 0.3097927268213975,
+      "grad_norm": 0.07599653080274857,
+      "learning_rate": 9.861255764007288e-06,
+      "loss": 0.5851,
+      "step": 624
+    },
+    {
+      "epoch": 0.31028918952463697,
+      "grad_norm": 0.07657107986209026,
+      "learning_rate": 9.860798111142507e-06,
+      "loss": 0.557,
+      "step": 625
+    },
+    {
+      "epoch": 0.3107856522278764,
+      "grad_norm": 0.08159401440687102,
+      "learning_rate": 9.860339715382671e-06,
+      "loss": 0.5407,
+      "step": 626
+    },
+    {
+      "epoch": 0.3112821149311158,
+      "grad_norm": 0.08214172807970598,
+      "learning_rate": 9.859880576797842e-06,
+      "loss": 0.5532,
+      "step": 627
+    },
+    {
+      "epoch": 0.31177857763435524,
+      "grad_norm": 0.07544880084915004,
+      "learning_rate": 9.859420695458187e-06,
+      "loss": 0.5316,
+      "step": 628
+    },
+    {
+      "epoch": 0.3122750403375946,
+      "grad_norm": 0.07386885211105122,
+      "learning_rate": 9.858960071433994e-06,
+      "loss": 0.5423,
+      "step": 629
+    },
+    {
+      "epoch": 0.31277150304083406,
+      "grad_norm": 0.07501408504334331,
+      "learning_rate": 9.858498704795663e-06,
+      "loss": 0.5534,
+      "step": 630
+    },
+    {
+      "epoch": 0.31326796574407345,
+      "grad_norm": 0.0759136680167307,
+      "learning_rate": 9.858036595613704e-06,
+      "loss": 0.5307,
+      "step": 631
+    },
+    {
+      "epoch": 0.3137644284473129,
+      "grad_norm": 0.08163074955687628,
+      "learning_rate": 9.857573743958744e-06,
+      "loss": 0.5957,
+      "step": 632
+    },
+    {
+      "epoch": 0.31426089115055234,
+      "grad_norm": 0.07935519929681213,
+      "learning_rate": 9.857110149901521e-06,
+      "loss": 0.567,
+      "step": 633
+    },
+    {
+      "epoch": 0.3147573538537917,
+      "grad_norm": 0.07839620252500945,
+      "learning_rate": 9.856645813512892e-06,
+      "loss": 0.5368,
+      "step": 634
+    },
+    {
+      "epoch": 0.31525381655703116,
+      "grad_norm": 0.07681988893148366,
+      "learning_rate": 9.85618073486382e-06,
+      "loss": 0.504,
+      "step": 635
+    },
+    {
+      "epoch": 0.31575027926027055,
+      "grad_norm": 0.0830345969868476,
+      "learning_rate": 9.855714914025386e-06,
+      "loss": 0.5278,
+      "step": 636
+    },
+    {
+      "epoch": 0.31624674196351,
+      "grad_norm": 0.08129993849570155,
+      "learning_rate": 9.855248351068781e-06,
+      "loss": 0.5402,
+      "step": 637
+    },
+    {
+      "epoch": 0.31674320466674943,
+      "grad_norm": 0.07952213949351684,
+      "learning_rate": 9.854781046065317e-06,
+      "loss": 0.543,
+      "step": 638
+    },
+    {
+      "epoch": 0.3172396673699888,
+      "grad_norm": 0.0751788341311726,
+      "learning_rate": 9.854312999086406e-06,
+      "loss": 0.5411,
+      "step": 639
+    },
+    {
+      "epoch": 0.31773613007322826,
+      "grad_norm": 0.08444156643795851,
+      "learning_rate": 9.85384421020359e-06,
+      "loss": 0.5884,
+      "step": 640
+    },
+    {
+      "epoch": 0.31823259277646765,
+      "grad_norm": 0.07583571988436319,
+      "learning_rate": 9.85337467948851e-06,
+      "loss": 0.5645,
+      "step": 641
+    },
+    {
+      "epoch": 0.3187290554797071,
+      "grad_norm": 0.07686002390051062,
+      "learning_rate": 9.852904407012929e-06,
+      "loss": 0.5691,
+      "step": 642
+    },
+    {
+      "epoch": 0.31922551818294653,
+      "grad_norm": 0.07795666555244363,
+      "learning_rate": 9.852433392848718e-06,
+      "loss": 0.5192,
+      "step": 643
+    },
+    {
+      "epoch": 0.3197219808861859,
+      "grad_norm": 0.08034340447050264,
+      "learning_rate": 9.851961637067869e-06,
+      "loss": 0.6113,
+      "step": 644
+    },
+    {
+      "epoch": 0.32021844358942536,
+      "grad_norm": 0.0748943142336664,
+      "learning_rate": 9.851489139742476e-06,
+      "loss": 0.5362,
+      "step": 645
+    },
+    {
+      "epoch": 0.32071490629266475,
+      "grad_norm": 0.07507270048060632,
+      "learning_rate": 9.851015900944757e-06,
+      "loss": 0.5509,
+      "step": 646
+    },
+    {
+      "epoch": 0.3212113689959042,
+      "grad_norm": 0.07633316575711642,
+      "learning_rate": 9.850541920747037e-06,
+      "loss": 0.6085,
+      "step": 647
+    },
+    {
+      "epoch": 0.3217078316991436,
+      "grad_norm": 0.07528076533630637,
+      "learning_rate": 9.850067199221758e-06,
+      "loss": 0.526,
+      "step": 648
+    },
+    {
+      "epoch": 0.322204294402383,
+      "grad_norm": 0.07707243425328635,
+      "learning_rate": 9.849591736441473e-06,
+      "loss": 0.5695,
+      "step": 649
+    },
+    {
+      "epoch": 0.32270075710562246,
+      "grad_norm": 0.07654913980669235,
+      "learning_rate": 9.849115532478848e-06,
+      "loss": 0.5705,
+      "step": 650
+    },
+    {
+      "epoch": 0.32319721980886185,
+      "grad_norm": 0.07645374658029656,
+      "learning_rate": 9.848638587406661e-06,
+      "loss": 0.5474,
+      "step": 651
+    },
+    {
+      "epoch": 0.3236936825121013,
+      "grad_norm": 0.07807950547456723,
+      "learning_rate": 9.84816090129781e-06,
+      "loss": 0.5887,
+      "step": 652
+    },
+    {
+      "epoch": 0.3241901452153407,
+      "grad_norm": 0.07646768926967591,
+      "learning_rate": 9.8476824742253e-06,
+      "loss": 0.5449,
+      "step": 653
+    },
+    {
+      "epoch": 0.3246866079185801,
+      "grad_norm": 0.07654133567981597,
+      "learning_rate": 9.84720330626225e-06,
+      "loss": 0.562,
+      "step": 654
+    },
+    {
+      "epoch": 0.32518307062181956,
+      "grad_norm": 0.07443762509568126,
+      "learning_rate": 9.846723397481892e-06,
+      "loss": 0.553,
+      "step": 655
+    },
+    {
+      "epoch": 0.32567953332505895,
+      "grad_norm": 0.07370022698268587,
+      "learning_rate": 9.846242747957578e-06,
+      "loss": 0.5053,
+      "step": 656
+    },
+    {
+      "epoch": 0.3261759960282984,
+      "grad_norm": 0.075692372137261,
+      "learning_rate": 9.84576135776276e-06,
+      "loss": 0.5376,
+      "step": 657
+    },
+    {
+      "epoch": 0.3266724587315378,
+      "grad_norm": 0.07705713356659613,
+      "learning_rate": 9.845279226971016e-06,
+      "loss": 0.5515,
+      "step": 658
+    },
+    {
+      "epoch": 0.3271689214347772,
+      "grad_norm": 0.07295323071068782,
+      "learning_rate": 9.84479635565603e-06,
+      "loss": 0.5299,
+      "step": 659
+    },
+    {
+      "epoch": 0.32766538413801666,
+      "grad_norm": 0.0758470961017997,
+      "learning_rate": 9.8443127438916e-06,
+      "loss": 0.5667,
+      "step": 660
+    },
+    {
+      "epoch": 0.32816184684125604,
+      "grad_norm": 0.07599385356522345,
+      "learning_rate": 9.843828391751642e-06,
+      "loss": 0.5288,
+      "step": 661
+    },
+    {
+      "epoch": 0.3286583095444955,
+      "grad_norm": 0.0746482852812454,
+      "learning_rate": 9.843343299310177e-06,
+      "loss": 0.517,
+      "step": 662
+    },
+    {
+      "epoch": 0.3291547722477349,
+      "grad_norm": 0.07538540101260628,
+      "learning_rate": 9.842857466641348e-06,
+      "loss": 0.5433,
+      "step": 663
+    },
+    {
+      "epoch": 0.3296512349509743,
+      "grad_norm": 0.07526342937730707,
+      "learning_rate": 9.842370893819404e-06,
+      "loss": 0.5547,
+      "step": 664
+    },
+    {
+      "epoch": 0.3301476976542137,
+      "grad_norm": 0.0789429151978629,
+      "learning_rate": 9.84188358091871e-06,
+      "loss": 0.5368,
+      "step": 665
+    },
+    {
+      "epoch": 0.33064416035745314,
+      "grad_norm": 0.07763526839910767,
+      "learning_rate": 9.841395528013744e-06,
+      "loss": 0.5795,
+      "step": 666
+    },
+    {
+      "epoch": 0.3311406230606926,
+      "grad_norm": 0.07494340511997437,
+      "learning_rate": 9.840906735179096e-06,
+      "loss": 0.6035,
+      "step": 667
+    },
+    {
+      "epoch": 0.33163708576393197,
+      "grad_norm": 0.07910952886635866,
+      "learning_rate": 9.840417202489473e-06,
+      "loss": 0.5854,
+      "step": 668
+    },
+    {
+      "epoch": 0.3321335484671714,
+      "grad_norm": 0.07892986798515082,
+      "learning_rate": 9.839926930019692e-06,
+      "loss": 0.5396,
+      "step": 669
+    },
+    {
+      "epoch": 0.3326300111704108,
+      "grad_norm": 0.0729563764448872,
+      "learning_rate": 9.839435917844682e-06,
+      "loss": 0.524,
+      "step": 670
+    },
+    {
+      "epoch": 0.33312647387365024,
+      "grad_norm": 0.07802658438460508,
+      "learning_rate": 9.838944166039486e-06,
+      "loss": 0.553,
+      "step": 671
+    },
+    {
+      "epoch": 0.3336229365768897,
+      "grad_norm": 0.07792180398015403,
+      "learning_rate": 9.83845167467926e-06,
+      "loss": 0.5449,
+      "step": 672
+    },
+    {
+      "epoch": 0.33411939928012907,
+      "grad_norm": 0.07574586523670872,
+      "learning_rate": 9.837958443839274e-06,
+      "loss": 0.595,
+      "step": 673
+    },
+    {
+      "epoch": 0.3346158619833685,
+      "grad_norm": 0.07591400320048404,
+      "learning_rate": 9.837464473594911e-06,
+      "loss": 0.5292,
+      "step": 674
+    },
+    {
+      "epoch": 0.3351123246866079,
+      "grad_norm": 0.07644973731060183,
+      "learning_rate": 9.836969764021666e-06,
+      "loss": 0.5455,
+      "step": 675
+    },
+    {
+      "epoch": 0.33560878738984734,
+      "grad_norm": 0.08101825987060038,
+      "learning_rate": 9.836474315195148e-06,
+      "loss": 0.5767,
+      "step": 676
+    },
+    {
+      "epoch": 0.3361052500930868,
+      "grad_norm": 0.07730742434766155,
+      "learning_rate": 9.835978127191077e-06,
+      "loss": 0.6241,
+      "step": 677
+    },
+    {
+      "epoch": 0.33660171279632617,
+      "grad_norm": 0.07874246801059555,
+      "learning_rate": 9.83548120008529e-06,
+      "loss": 0.5278,
+      "step": 678
+    },
+    {
+      "epoch": 0.3370981754995656,
+      "grad_norm": 0.07397850549930313,
+      "learning_rate": 9.83498353395373e-06,
+      "loss": 0.4994,
+      "step": 679
+    },
+    {
+      "epoch": 0.337594638202805,
+      "grad_norm": 0.07591277824088376,
+      "learning_rate": 9.834485128872462e-06,
+      "loss": 0.5196,
+      "step": 680
+    },
+    {
+      "epoch": 0.33809110090604444,
+      "grad_norm": 0.07229658778373127,
+      "learning_rate": 9.833985984917656e-06,
+      "loss": 0.5368,
+      "step": 681
+    },
+    {
+      "epoch": 0.3385875636092838,
+      "grad_norm": 0.08099111946438035,
+      "learning_rate": 9.8334861021656e-06,
+      "loss": 0.5697,
+      "step": 682
+    },
+    {
+      "epoch": 0.33908402631252327,
+      "grad_norm": 0.07252478515888824,
+      "learning_rate": 9.832985480692691e-06,
+      "loss": 0.5604,
+      "step": 683
+    },
+    {
+      "epoch": 0.3395804890157627,
+      "grad_norm": 0.07239879222289877,
+      "learning_rate": 9.832484120575446e-06,
+      "loss": 0.487,
+      "step": 684
+    },
+    {
+      "epoch": 0.3400769517190021,
+      "grad_norm": 0.07153328866561025,
+      "learning_rate": 9.831982021890483e-06,
+      "loss": 0.5233,
+      "step": 685
+    },
+    {
+      "epoch": 0.34057341442224154,
+      "grad_norm": 0.07439020024051021,
+      "learning_rate": 9.831479184714543e-06,
+      "loss": 0.5375,
+      "step": 686
+    },
+    {
+      "epoch": 0.3410698771254809,
+      "grad_norm": 0.07404184993442114,
+      "learning_rate": 9.830975609124477e-06,
+      "loss": 0.5332,
+      "step": 687
+    },
+    {
+      "epoch": 0.34156633982872037,
+      "grad_norm": 0.07438693972814182,
+      "learning_rate": 9.830471295197248e-06,
+      "loss": 0.5536,
+      "step": 688
+    },
+    {
+      "epoch": 0.3420628025319598,
+      "grad_norm": 0.075698102229209,
+      "learning_rate": 9.829966243009932e-06,
+      "loss": 0.5361,
+      "step": 689
+    },
+    {
+      "epoch": 0.3425592652351992,
+      "grad_norm": 0.07580611925918088,
+      "learning_rate": 9.829460452639718e-06,
+      "loss": 0.5569,
+      "step": 690
+    },
+    {
+      "epoch": 0.34305572793843864,
+      "grad_norm": 0.07367114112343696,
+      "learning_rate": 9.828953924163908e-06,
+      "loss": 0.557,
+      "step": 691
+    },
+    {
+      "epoch": 0.343552190641678,
+      "grad_norm": 0.0733573832874063,
+      "learning_rate": 9.828446657659919e-06,
+      "loss": 0.5857,
+      "step": 692
+    },
+    {
+      "epoch": 0.34404865334491747,
+      "grad_norm": 0.08186579423918036,
+      "learning_rate": 9.827938653205275e-06,
+      "loss": 0.5366,
+      "step": 693
+    },
+    {
+      "epoch": 0.3445451160481569,
+      "grad_norm": 0.07582312092529461,
+      "learning_rate": 9.82742991087762e-06,
+      "loss": 0.5524,
+      "step": 694
+    },
+    {
+      "epoch": 0.3450415787513963,
+      "grad_norm": 0.08324114328209058,
+      "learning_rate": 9.826920430754703e-06,
+      "loss": 0.5471,
+      "step": 695
+    },
+    {
+      "epoch": 0.34553804145463574,
+      "grad_norm": 0.07489973171052326,
+      "learning_rate": 9.826410212914393e-06,
+      "loss": 0.5547,
+      "step": 696
+    },
+    {
+      "epoch": 0.3460345041578751,
+      "grad_norm": 0.07337731136398072,
+      "learning_rate": 9.825899257434667e-06,
+      "loss": 0.5299,
+      "step": 697
+    },
+    {
+      "epoch": 0.34653096686111456,
+      "grad_norm": 0.0813945270751598,
+      "learning_rate": 9.825387564393616e-06,
+      "loss": 0.5469,
+      "step": 698
+    },
+    {
+      "epoch": 0.34702742956435395,
+      "grad_norm": 0.07553165147863128,
+      "learning_rate": 9.824875133869447e-06,
+      "loss": 0.5235,
+      "step": 699
+    },
+    {
+      "epoch": 0.3475238922675934,
+      "grad_norm": 0.07798665622206621,
+      "learning_rate": 9.824361965940475e-06,
+      "loss": 0.5449,
+      "step": 700
+    },
+    {
+      "epoch": 0.34802035497083283,
+      "grad_norm": 0.07720920154984248,
+      "learning_rate": 9.823848060685125e-06,
+      "loss": 0.5516,
+      "step": 701
+    },
+    {
+      "epoch": 0.3485168176740722,
+      "grad_norm": 0.07394946809380538,
+      "learning_rate": 9.823333418181948e-06,
+      "loss": 0.5434,
+      "step": 702
+    },
+    {
+      "epoch": 0.34901328037731166,
+      "grad_norm": 0.07443280799785286,
+      "learning_rate": 9.822818038509593e-06,
+      "loss": 0.5206,
+      "step": 703
+    },
+    {
+      "epoch": 0.34950974308055105,
+      "grad_norm": 0.07109902087405791,
+      "learning_rate": 9.822301921746829e-06,
+      "loss": 0.5129,
+      "step": 704
+    },
+    {
+      "epoch": 0.3500062057837905,
+      "grad_norm": 0.07494737825986637,
+      "learning_rate": 9.821785067972536e-06,
+      "loss": 0.5297,
+      "step": 705
+    },
+    {
+      "epoch": 0.35050266848702993,
+      "grad_norm": 0.07233626869785674,
+      "learning_rate": 9.821267477265705e-06,
+      "loss": 0.5418,
+      "step": 706
+    },
+    {
+      "epoch": 0.3509991311902693,
+      "grad_norm": 0.07147306841083971,
+      "learning_rate": 9.820749149705445e-06,
+      "loss": 0.5428,
+      "step": 707
+    },
+    {
+      "epoch": 0.35149559389350876,
+      "grad_norm": 0.07276839503056264,
+      "learning_rate": 9.820230085370972e-06,
+      "loss": 0.5634,
+      "step": 708
+    },
+    {
+      "epoch": 0.35199205659674815,
+      "grad_norm": 0.07575865554460536,
+      "learning_rate": 9.819710284341618e-06,
+      "loss": 0.5377,
+      "step": 709
+    },
+    {
+      "epoch": 0.3524885192999876,
+      "grad_norm": 0.0739559499625771,
+      "learning_rate": 9.819189746696823e-06,
+      "loss": 0.5323,
+      "step": 710
+    },
+    {
+      "epoch": 0.35298498200322703,
+      "grad_norm": 0.07331544965217528,
+      "learning_rate": 9.818668472516146e-06,
+      "loss": 0.5642,
+      "step": 711
+    },
+    {
+      "epoch": 0.3534814447064664,
+      "grad_norm": 0.07851875713635985,
+      "learning_rate": 9.818146461879256e-06,
+      "loss": 0.598,
+      "step": 712
+    },
+    {
+      "epoch": 0.35397790740970586,
+      "grad_norm": 0.07382231426622198,
+      "learning_rate": 9.817623714865931e-06,
+      "loss": 0.5683,
+      "step": 713
+    },
+    {
+      "epoch": 0.35447437011294525,
+      "grad_norm": 0.07647833075770147,
+      "learning_rate": 9.817100231556065e-06,
+      "loss": 0.5647,
+      "step": 714
+    },
+    {
+      "epoch": 0.3549708328161847,
+      "grad_norm": 0.07453844364837098,
+      "learning_rate": 9.816576012029666e-06,
+      "loss": 0.5479,
+      "step": 715
+    },
+    {
+      "epoch": 0.3554672955194241,
+      "grad_norm": 0.06994784433794665,
+      "learning_rate": 9.816051056366851e-06,
+      "loss": 0.5545,
+      "step": 716
+    },
+    {
+      "epoch": 0.3559637582226635,
+      "grad_norm": 0.0708558727408941,
+      "learning_rate": 9.815525364647853e-06,
+      "loss": 0.5235,
+      "step": 717
+    },
+    {
+      "epoch": 0.35646022092590296,
+      "grad_norm": 0.07211829677081709,
+      "learning_rate": 9.814998936953012e-06,
+      "loss": 0.5375,
+      "step": 718
+    },
+    {
+      "epoch": 0.35695668362914235,
+      "grad_norm": 0.0778707977537702,
+      "learning_rate": 9.814471773362788e-06,
+      "loss": 0.5003,
+      "step": 719
+    },
+    {
+      "epoch": 0.3574531463323818,
+      "grad_norm": 0.07346792577565847,
+      "learning_rate": 9.813943873957748e-06,
+      "loss": 0.5252,
+      "step": 720
+    },
+    {
+      "epoch": 0.3579496090356212,
+      "grad_norm": 0.0780037540912097,
+      "learning_rate": 9.81341523881857e-06,
+      "loss": 0.5609,
+      "step": 721
+    },
+    {
+      "epoch": 0.3584460717388606,
+      "grad_norm": 0.12649906090457236,
+      "learning_rate": 9.812885868026052e-06,
+      "loss": 0.5611,
+      "step": 722
+    },
+    {
+      "epoch": 0.35894253444210006,
+      "grad_norm": 0.07460362088706496,
+      "learning_rate": 9.812355761661096e-06,
+      "loss": 0.538,
+      "step": 723
+    },
+    {
+      "epoch": 0.35943899714533945,
+      "grad_norm": 0.0739528024367322,
+      "learning_rate": 9.811824919804725e-06,
+      "loss": 0.5721,
+      "step": 724
+    },
+    {
+      "epoch": 0.3599354598485789,
+      "grad_norm": 0.07753894839284191,
+      "learning_rate": 9.811293342538063e-06,
+      "loss": 0.5858,
+      "step": 725
+    },
+    {
+      "epoch": 0.3604319225518183,
+      "grad_norm": 0.07368755167074739,
+      "learning_rate": 9.81076102994236e-06,
+      "loss": 0.5774,
+      "step": 726
+    },
+    {
+      "epoch": 0.3609283852550577,
+      "grad_norm": 0.07661834913233602,
+      "learning_rate": 9.810227982098968e-06,
+      "loss": 0.5737,
+      "step": 727
+    },
+    {
+      "epoch": 0.36142484795829716,
+      "grad_norm": 0.07973682947552374,
+      "learning_rate": 9.809694199089352e-06,
+      "loss": 0.5407,
+      "step": 728
+    },
+    {
+      "epoch": 0.36192131066153654,
+      "grad_norm": 0.07276884200881067,
+      "learning_rate": 9.809159680995098e-06,
+      "loss": 0.5372,
+      "step": 729
+    },
+    {
+      "epoch": 0.362417773364776,
+      "grad_norm": 0.07705887562756104,
+      "learning_rate": 9.808624427897896e-06,
+      "loss": 0.5148,
+      "step": 730
+    },
+    {
+      "epoch": 0.3629142360680154,
+      "grad_norm": 0.07781544338816433,
+      "learning_rate": 9.80808843987955e-06,
+      "loss": 0.5545,
+      "step": 731
+    },
+    {
+      "epoch": 0.3634106987712548,
+      "grad_norm": 0.0806221612612232,
+      "learning_rate": 9.807551717021977e-06,
+      "loss": 0.5428,
+      "step": 732
+    },
+    {
+      "epoch": 0.36390716147449426,
+      "grad_norm": 0.07365922527604281,
+      "learning_rate": 9.807014259407209e-06,
+      "loss": 0.5246,
+      "step": 733
+    },
+    {
+      "epoch": 0.36440362417773364,
+      "grad_norm": 0.07405742185015944,
+      "learning_rate": 9.806476067117384e-06,
+      "loss": 0.5392,
+      "step": 734
+    },
+    {
+      "epoch": 0.3649000868809731,
+      "grad_norm": 0.07652890186972217,
+      "learning_rate": 9.80593714023476e-06,
+      "loss": 0.5149,
+      "step": 735
+    },
+    {
+      "epoch": 0.36539654958421247,
+      "grad_norm": 0.07593774024556392,
+      "learning_rate": 9.8053974788417e-06,
+      "loss": 0.5356,
+      "step": 736
+    },
+    {
+      "epoch": 0.3658930122874519,
+      "grad_norm": 0.07415413956682762,
+      "learning_rate": 9.804857083020685e-06,
+      "loss": 0.5415,
+      "step": 737
+    },
+    {
+      "epoch": 0.3663894749906913,
+      "grad_norm": 0.07595510777957903,
+      "learning_rate": 9.804315952854304e-06,
+      "loss": 0.5259,
+      "step": 738
+    },
+    {
+      "epoch": 0.36688593769393074,
+      "grad_norm": 0.07424855325354711,
+      "learning_rate": 9.803774088425262e-06,
+      "loss": 0.5243,
+      "step": 739
+    },
+    {
+      "epoch": 0.3673824003971702,
+      "grad_norm": 0.07812406038583773,
+      "learning_rate": 9.803231489816371e-06,
+      "loss": 0.5584,
+      "step": 740
+    },
+    {
+      "epoch": 0.36787886310040957,
+      "grad_norm": 0.07240708263932845,
+      "learning_rate": 9.802688157110564e-06,
+      "loss": 0.5248,
+      "step": 741
+    },
+    {
+      "epoch": 0.368375325803649,
+      "grad_norm": 0.07605118794849668,
+      "learning_rate": 9.802144090390875e-06,
+      "loss": 0.5226,
+      "step": 742
+    },
+    {
+      "epoch": 0.3688717885068884,
+      "grad_norm": 0.15783281913989158,
+      "learning_rate": 9.80159928974046e-06,
+      "loss": 0.5453,
+      "step": 743
+    },
+    {
+      "epoch": 0.36936825121012784,
+      "grad_norm": 0.07410517260137793,
+      "learning_rate": 9.80105375524258e-06,
+      "loss": 0.5069,
+      "step": 744
+    },
+    {
+      "epoch": 0.3698647139133673,
+      "grad_norm": 0.07734176892139462,
+      "learning_rate": 9.800507486980613e-06,
+      "loss": 0.5359,
+      "step": 745
+    },
+    {
+      "epoch": 0.37036117661660667,
+      "grad_norm": 0.07149204470583237,
+      "learning_rate": 9.799960485038047e-06,
+      "loss": 0.4931,
+      "step": 746
+    },
+    {
+      "epoch": 0.3708576393198461,
+      "grad_norm": 0.0721698434133166,
+      "learning_rate": 9.799412749498483e-06,
+      "loss": 0.5284,
+      "step": 747
+    },
+    {
+      "epoch": 0.3713541020230855,
+      "grad_norm": 0.07167968336400159,
+      "learning_rate": 9.798864280445633e-06,
+      "loss": 0.5619,
+      "step": 748
+    },
+    {
+      "epoch": 0.37185056472632494,
+      "grad_norm": 0.07408407334726448,
+      "learning_rate": 9.798315077963321e-06,
+      "loss": 0.5305,
+      "step": 749
+    },
+    {
+      "epoch": 0.3723470274295644,
+      "grad_norm": 0.07298928263403225,
+      "learning_rate": 9.797765142135483e-06,
+      "loss": 0.5543,
+      "step": 750
+    },
+    {
+      "epoch": 0.37284349013280377,
+      "grad_norm": 0.075371200502978,
+      "learning_rate": 9.797214473046171e-06,
+      "loss": 0.5802,
+      "step": 751
+    },
+    {
+      "epoch": 0.3733399528360432,
+      "grad_norm": 0.0738341822930936,
+      "learning_rate": 9.796663070779545e-06,
+      "loss": 0.5377,
+      "step": 752
+    },
+    {
+      "epoch": 0.3738364155392826,
+      "grad_norm": 0.07475281086820233,
+      "learning_rate": 9.796110935419876e-06,
+      "loss": 0.5433,
+      "step": 753
+    },
+    {
+      "epoch": 0.37433287824252204,
+      "grad_norm": 0.07636423039177996,
+      "learning_rate": 9.79555806705155e-06,
+      "loss": 0.5191,
+      "step": 754
+    },
+    {
+      "epoch": 0.3748293409457614,
+      "grad_norm": 0.07553329273177121,
+      "learning_rate": 9.795004465759067e-06,
+      "loss": 0.5685,
+      "step": 755
+    },
+    {
+      "epoch": 0.37532580364900087,
+      "grad_norm": 0.07669955016383546,
+      "learning_rate": 9.79445013162703e-06,
+      "loss": 0.5391,
+      "step": 756
+    },
+    {
+      "epoch": 0.3758222663522403,
+      "grad_norm": 0.07544294491461324,
+      "learning_rate": 9.793895064740166e-06,
+      "loss": 0.5612,
+      "step": 757
+    },
+    {
+      "epoch": 0.3763187290554797,
+      "grad_norm": 0.0754930880002628,
+      "learning_rate": 9.793339265183303e-06,
+      "loss": 0.5458,
+      "step": 758
+    },
+    {
+      "epoch": 0.37681519175871914,
+      "grad_norm": 0.07749670149783751,
+      "learning_rate": 9.79278273304139e-06,
+      "loss": 0.5937,
+      "step": 759
+    },
+    {
+      "epoch": 0.3773116544619585,
+      "grad_norm": 0.07268005398789518,
+      "learning_rate": 9.792225468399485e-06,
+      "loss": 0.556,
+      "step": 760
+    },
+    {
+      "epoch": 0.37780811716519797,
+      "grad_norm": 0.07239528164494755,
+      "learning_rate": 9.791667471342752e-06,
+      "loss": 0.5385,
+      "step": 761
+    },
+    {
+      "epoch": 0.3783045798684374,
+      "grad_norm": 0.07334739224571676,
+      "learning_rate": 9.791108741956476e-06,
+      "loss": 0.5158,
+      "step": 762
+    },
+    {
+      "epoch": 0.3788010425716768,
+      "grad_norm": 0.07033808439598177,
+      "learning_rate": 9.790549280326046e-06,
+      "loss": 0.5481,
+      "step": 763
+    },
+    {
+      "epoch": 0.37929750527491624,
+      "grad_norm": 0.07552295599412208,
+      "learning_rate": 9.789989086536971e-06,
+      "loss": 0.5242,
+      "step": 764
+    },
+    {
+      "epoch": 0.3797939679781556,
+      "grad_norm": 0.07401484746819591,
+      "learning_rate": 9.789428160674865e-06,
+      "loss": 0.5562,
+      "step": 765
+    },
+    {
+      "epoch": 0.38029043068139506,
+      "grad_norm": 0.11414006306157573,
+      "learning_rate": 9.788866502825458e-06,
+      "loss": 0.5698,
+      "step": 766
+    },
+    {
+      "epoch": 0.3807868933846345,
+      "grad_norm": 0.07121253600333775,
+      "learning_rate": 9.78830411307459e-06,
+      "loss": 0.5284,
+      "step": 767
+    },
+    {
+      "epoch": 0.3812833560878739,
+      "grad_norm": 0.07265804118235417,
+      "learning_rate": 9.787740991508212e-06,
+      "loss": 0.5158,
+      "step": 768
+    },
+    {
+      "epoch": 0.38177981879111333,
+      "grad_norm": 0.07254318353819536,
+      "learning_rate": 9.787177138212391e-06,
+      "loss": 0.5091,
+      "step": 769
+    },
+    {
+      "epoch": 0.3822762814943527,
+      "grad_norm": 0.07281349514641368,
+      "learning_rate": 9.786612553273298e-06,
+      "loss": 0.5231,
+      "step": 770
+    },
+    {
+      "epoch": 0.38277274419759216,
+      "grad_norm": 0.07050000723172054,
+      "learning_rate": 9.786047236777225e-06,
+      "loss": 0.5579,
+      "step": 771
+    },
+    {
+      "epoch": 0.38326920690083155,
+      "grad_norm": 0.07541531424572744,
+      "learning_rate": 9.78548118881057e-06,
+      "loss": 0.5373,
+      "step": 772
+    },
+    {
+      "epoch": 0.383765669604071,
+      "grad_norm": 0.07453492248282302,
+      "learning_rate": 9.784914409459847e-06,
+      "loss": 0.5556,
+      "step": 773
+    },
+    {
+      "epoch": 0.38426213230731043,
+      "grad_norm": 0.0757761604807883,
+      "learning_rate": 9.784346898811675e-06,
+      "loss": 0.5199,
+      "step": 774
+    },
+    {
+      "epoch": 0.3847585950105498,
+      "grad_norm": 0.07568484772982484,
+      "learning_rate": 9.78377865695279e-06,
+      "loss": 0.5324,
+      "step": 775
+    },
+    {
+      "epoch": 0.38525505771378926,
+      "grad_norm": 0.07364200493411856,
+      "learning_rate": 9.78320968397004e-06,
+      "loss": 0.5339,
+      "step": 776
+    },
+    {
+      "epoch": 0.38575152041702865,
+      "grad_norm": 0.07497774875981036,
+      "learning_rate": 9.782639979950382e-06,
+      "loss": 0.5578,
+      "step": 777
+    },
+    {
+      "epoch": 0.3862479831202681,
+      "grad_norm": 0.08179279051563958,
+      "learning_rate": 9.782069544980887e-06,
+      "loss": 0.5467,
+      "step": 778
+    },
+    {
+      "epoch": 0.38674444582350753,
+      "grad_norm": 0.07280472908410666,
+      "learning_rate": 9.781498379148738e-06,
+      "loss": 0.501,
+      "step": 779
+    },
+    {
+      "epoch": 0.3872409085267469,
+      "grad_norm": 0.07490404669769593,
+      "learning_rate": 9.780926482541227e-06,
+      "loss": 0.5342,
+      "step": 780
+    },
+    {
+      "epoch": 0.38773737122998636,
+      "grad_norm": 0.07570507230711694,
+      "learning_rate": 9.780353855245759e-06,
+      "loss": 0.5816,
+      "step": 781
+    },
+    {
+      "epoch": 0.38823383393322575,
+      "grad_norm": 0.07622051872508213,
+      "learning_rate": 9.779780497349852e-06,
+      "loss": 0.5522,
+      "step": 782
+    },
+    {
+      "epoch": 0.3887302966364652,
+      "grad_norm": 0.0765505749512821,
+      "learning_rate": 9.779206408941131e-06,
+      "loss": 0.5141,
+      "step": 783
+    },
+    {
+      "epoch": 0.38922675933970463,
+      "grad_norm": 0.07629599426112826,
+      "learning_rate": 9.778631590107342e-06,
+      "loss": 0.5388,
+      "step": 784
+    },
+    {
+      "epoch": 0.389723222042944,
+      "grad_norm": 0.07362072760938124,
+      "learning_rate": 9.778056040936333e-06,
+      "loss": 0.5531,
+      "step": 785
+    },
+    {
+      "epoch": 0.39021968474618346,
+      "grad_norm": 0.07712485250124279,
+      "learning_rate": 9.777479761516069e-06,
+      "loss": 0.5088,
+      "step": 786
+    },
+    {
+      "epoch": 0.39071614744942285,
+      "grad_norm": 0.07248722869142624,
+      "learning_rate": 9.776902751934625e-06,
+      "loss": 0.5495,
+      "step": 787
+    },
+    {
+      "epoch": 0.3912126101526623,
+      "grad_norm": 0.07969653988684422,
+      "learning_rate": 9.776325012280185e-06,
+      "loss": 0.5586,
+      "step": 788
+    },
+    {
+      "epoch": 0.3917090728559017,
+      "grad_norm": 0.07887747490682782,
+      "learning_rate": 9.775746542641052e-06,
+      "loss": 0.5469,
+      "step": 789
+    },
+    {
+      "epoch": 0.3922055355591411,
+      "grad_norm": 0.07511324662951371,
+      "learning_rate": 9.77516734310563e-06,
+      "loss": 0.5043,
+      "step": 790
+    },
+    {
+      "epoch": 0.39270199826238056,
+      "grad_norm": 0.074682710636718,
+      "learning_rate": 9.774587413762448e-06,
+      "loss": 0.573,
+      "step": 791
+    },
+    {
+      "epoch": 0.39319846096561994,
+      "grad_norm": 0.07823987539462922,
+      "learning_rate": 9.774006754700132e-06,
+      "loss": 0.5012,
+      "step": 792
+    },
+    {
+      "epoch": 0.3936949236688594,
+      "grad_norm": 0.07927560532508765,
+      "learning_rate": 9.77342536600743e-06,
+      "loss": 0.5332,
+      "step": 793
+    },
+    {
+      "epoch": 0.3941913863720988,
+      "grad_norm": 0.081075947027259,
+      "learning_rate": 9.772843247773197e-06,
+      "loss": 0.571,
+      "step": 794
+    },
+    {
+      "epoch": 0.3946878490753382,
+      "grad_norm": 0.07529856682734566,
+      "learning_rate": 9.772260400086397e-06,
+      "loss": 0.5392,
+      "step": 795
+    },
+    {
+      "epoch": 0.39518431177857766,
+      "grad_norm": 0.07593026315167804,
+      "learning_rate": 9.771676823036115e-06,
+      "loss": 0.513,
+      "step": 796
+    },
+    {
+      "epoch": 0.39568077448181704,
+      "grad_norm": 0.07488551320818262,
+      "learning_rate": 9.771092516711538e-06,
+      "loss": 0.5658,
+      "step": 797
+    },
+    {
+      "epoch": 0.3961772371850565,
+      "grad_norm": 0.07448583103314095,
+      "learning_rate": 9.77050748120197e-06,
+      "loss": 0.5064,
+      "step": 798
+    },
+    {
+      "epoch": 0.39667369988829587,
+      "grad_norm": 0.07407263475614544,
+      "learning_rate": 9.76992171659682e-06,
+      "loss": 0.5379,
+      "step": 799
+    },
+    {
+      "epoch": 0.3971701625915353,
+      "grad_norm": 0.0748585105071261,
+      "learning_rate": 9.769335222985617e-06,
+      "loss": 0.5195,
+      "step": 800
+    },
+    {
+      "epoch": 0.39766662529477476,
+      "grad_norm": 0.07391534395442713,
+      "learning_rate": 9.768748000457996e-06,
+      "loss": 0.5131,
+      "step": 801
+    },
+    {
+      "epoch": 0.39816308799801414,
+      "grad_norm": 0.07465639657356582,
+      "learning_rate": 9.768160049103702e-06,
+      "loss": 0.5405,
+      "step": 802
+    },
+    {
+      "epoch": 0.3986595507012536,
+      "grad_norm": 0.07476172520797038,
+      "learning_rate": 9.767571369012599e-06,
+      "loss": 0.5231,
+      "step": 803
+    },
+    {
+      "epoch": 0.39915601340449297,
+      "grad_norm": 0.08064284315067823,
+      "learning_rate": 9.766981960274653e-06,
+      "loss": 0.5443,
+      "step": 804
+    },
+    {
+      "epoch": 0.3996524761077324,
+      "grad_norm": 0.07430645599205549,
+      "learning_rate": 9.766391822979948e-06,
+      "loss": 0.5142,
+      "step": 805
+    },
+    {
+      "epoch": 0.4001489388109718,
+      "grad_norm": 0.07410980228388762,
+      "learning_rate": 9.765800957218677e-06,
+      "loss": 0.5517,
+      "step": 806
+    },
+    {
+      "epoch": 0.40064540151421124,
+      "grad_norm": 0.06867453077929601,
+      "learning_rate": 9.765209363081141e-06,
+      "loss": 0.5649,
+      "step": 807
+    },
+    {
+      "epoch": 0.4011418642174507,
+      "grad_norm": 0.07353979734131227,
+      "learning_rate": 9.764617040657759e-06,
+      "loss": 0.5204,
+      "step": 808
+    },
+    {
+      "epoch": 0.40163832692069007,
+      "grad_norm": 0.07525469655765583,
+      "learning_rate": 9.764023990039058e-06,
+      "loss": 0.5532,
+      "step": 809
+    },
+    {
+      "epoch": 0.4021347896239295,
+      "grad_norm": 0.07078684341894576,
+      "learning_rate": 9.763430211315675e-06,
+      "loss": 0.5532,
+      "step": 810
+    },
+    {
+      "epoch": 0.4026312523271689,
+      "grad_norm": 0.07734947995505723,
+      "learning_rate": 9.76283570457836e-06,
+      "loss": 0.5731,
+      "step": 811
+    },
+    {
+      "epoch": 0.40312771503040834,
+      "grad_norm": 0.0751893291465059,
+      "learning_rate": 9.762240469917972e-06,
+      "loss": 0.5656,
+      "step": 812
+    },
+    {
+      "epoch": 0.4036241777336478,
+      "grad_norm": 0.07498423865346433,
+      "learning_rate": 9.761644507425487e-06,
+      "loss": 0.5528,
+      "step": 813
+    },
+    {
+      "epoch": 0.40412064043688717,
+      "grad_norm": 0.07905003613185434,
+      "learning_rate": 9.761047817191987e-06,
+      "loss": 0.5376,
+      "step": 814
+    },
+    {
+      "epoch": 0.4046171031401266,
+      "grad_norm": 0.0739115656470773,
+      "learning_rate": 9.760450399308662e-06,
+      "loss": 0.5471,
+      "step": 815
+    },
+    {
+      "epoch": 0.405113565843366,
+      "grad_norm": 0.07314428919759865,
+      "learning_rate": 9.759852253866825e-06,
+      "loss": 0.5458,
+      "step": 816
+    },
+    {
+      "epoch": 0.40561002854660544,
+      "grad_norm": 0.0754779528703246,
+      "learning_rate": 9.759253380957889e-06,
+      "loss": 0.5271,
+      "step": 817
+    },
+    {
+      "epoch": 0.4061064912498449,
+      "grad_norm": 0.0767964488723569,
+      "learning_rate": 9.758653780673381e-06,
+      "loss": 0.5481,
+      "step": 818
+    },
+    {
+      "epoch": 0.40660295395308427,
+      "grad_norm": 0.08174582451126397,
+      "learning_rate": 9.758053453104943e-06,
+      "loss": 0.5661,
+      "step": 819
+    },
+    {
+      "epoch": 0.4070994166563237,
+      "grad_norm": 0.07215132459628594,
+      "learning_rate": 9.757452398344324e-06,
+      "loss": 0.5176,
+      "step": 820
+    },
+    {
+      "epoch": 0.4075958793595631,
+      "grad_norm": 0.0716641776433738,
+      "learning_rate": 9.756850616483386e-06,
+      "loss": 0.5315,
+      "step": 821
+    },
+    {
+      "epoch": 0.40809234206280254,
+      "grad_norm": 0.07282126077177561,
+      "learning_rate": 9.7562481076141e-06,
+      "loss": 0.5332,
+      "step": 822
+    },
+    {
+      "epoch": 0.4085888047660419,
+      "grad_norm": 0.07751227831062067,
+      "learning_rate": 9.755644871828555e-06,
+      "loss": 0.5222,
+      "step": 823
+    },
+    {
+      "epoch": 0.40908526746928137,
+      "grad_norm": 0.07485329801902614,
+      "learning_rate": 9.75504090921894e-06,
+      "loss": 0.5188,
+      "step": 824
+    },
+    {
+      "epoch": 0.4095817301725208,
+      "grad_norm": 0.07291242005129558,
+      "learning_rate": 9.754436219877564e-06,
+      "loss": 0.5199,
+      "step": 825
+    },
+    {
+      "epoch": 0.4100781928757602,
+      "grad_norm": 0.07463161826705483,
+      "learning_rate": 9.753830803896842e-06,
+      "loss": 0.5158,
+      "step": 826
+    },
+    {
+      "epoch": 0.41057465557899964,
+      "grad_norm": 0.07832879467891833,
+      "learning_rate": 9.753224661369304e-06,
+      "loss": 0.5607,
+      "step": 827
+    },
+    {
+      "epoch": 0.411071118282239,
+      "grad_norm": 0.07111309195620046,
+      "learning_rate": 9.75261779238759e-06,
+      "loss": 0.5544,
+      "step": 828
+    },
+    {
+      "epoch": 0.41156758098547847,
+      "grad_norm": 0.0819846829516738,
+      "learning_rate": 9.752010197044448e-06,
+      "loss": 0.5557,
+      "step": 829
+    },
+    {
+      "epoch": 0.4120640436887179,
+      "grad_norm": 0.07898817088337341,
+      "learning_rate": 9.75140187543274e-06,
+      "loss": 0.546,
+      "step": 830
+    },
+    {
+      "epoch": 0.4125605063919573,
+      "grad_norm": 0.07647522418031873,
+      "learning_rate": 9.750792827645438e-06,
+      "loss": 0.5149,
+      "step": 831
+    },
+    {
+      "epoch": 0.41305696909519674,
+      "grad_norm": 0.07515552620271704,
+      "learning_rate": 9.750183053775625e-06,
+      "loss": 0.5493,
+      "step": 832
+    },
+    {
+      "epoch": 0.4135534317984361,
+      "grad_norm": 0.08341787971048387,
+      "learning_rate": 9.749572553916497e-06,
+      "loss": 0.5505,
+      "step": 833
+    },
+    {
+      "epoch": 0.41404989450167556,
+      "grad_norm": 0.07297396297855867,
+      "learning_rate": 9.748961328161358e-06,
+      "loss": 0.5288,
+      "step": 834
+    },
+    {
+      "epoch": 0.414546357204915,
+      "grad_norm": 0.0742750111482202,
+      "learning_rate": 9.748349376603622e-06,
+      "loss": 0.541,
+      "step": 835
+    },
+    {
+      "epoch": 0.4150428199081544,
+      "grad_norm": 0.0742391348271075,
+      "learning_rate": 9.747736699336819e-06,
+      "loss": 0.5399,
+      "step": 836
+    },
+    {
+      "epoch": 0.41553928261139383,
+      "grad_norm": 0.07769022255581656,
+      "learning_rate": 9.747123296454584e-06,
+      "loss": 0.5403,
+      "step": 837
+    },
+    {
+      "epoch": 0.4160357453146332,
+      "grad_norm": 0.07478718017828734,
+      "learning_rate": 9.74650916805067e-06,
+      "loss": 0.5456,
+      "step": 838
+    },
+    {
+      "epoch": 0.41653220801787266,
+      "grad_norm": 0.07963698330755821,
+      "learning_rate": 9.745894314218933e-06,
+      "loss": 0.5898,
+      "step": 839
+    },
+    {
+      "epoch": 0.41702867072111205,
+      "grad_norm": 0.07829076168531228,
+      "learning_rate": 9.745278735053345e-06,
+      "loss": 0.515,
+      "step": 840
+    },
+    {
+      "epoch": 0.4175251334243515,
+      "grad_norm": 0.07440206958109287,
+      "learning_rate": 9.744662430647986e-06,
+      "loss": 0.5477,
+      "step": 841
+    },
+    {
+      "epoch": 0.41802159612759093,
+      "grad_norm": 0.07089389200900213,
+      "learning_rate": 9.74404540109705e-06,
+      "loss": 0.5084,
+      "step": 842
+    },
+    {
+      "epoch": 0.4185180588308303,
+      "grad_norm": 0.07302037721807964,
+      "learning_rate": 9.74342764649484e-06,
+      "loss": 0.5318,
+      "step": 843
+    },
+    {
+      "epoch": 0.41901452153406976,
+      "grad_norm": 0.0783625897828163,
+      "learning_rate": 9.742809166935768e-06,
+      "loss": 0.5938,
+      "step": 844
+    },
+    {
+      "epoch": 0.41951098423730915,
+      "grad_norm": 0.07369207278532003,
+      "learning_rate": 9.742189962514361e-06,
+      "loss": 0.5423,
+      "step": 845
+    },
+    {
+      "epoch": 0.4200074469405486,
+      "grad_norm": 0.07785657874090317,
+      "learning_rate": 9.741570033325254e-06,
+      "loss": 0.5319,
+      "step": 846
+    },
+    {
+      "epoch": 0.42050390964378803,
+      "grad_norm": 0.08593813392209855,
+      "learning_rate": 9.740949379463192e-06,
+      "loss": 0.5383,
+      "step": 847
+    },
+    {
+      "epoch": 0.4210003723470274,
+      "grad_norm": 0.07225369547298743,
+      "learning_rate": 9.740328001023032e-06,
+      "loss": 0.506,
+      "step": 848
+    },
+    {
+      "epoch": 0.42149683505026686,
+      "grad_norm": 0.07409992281897824,
+      "learning_rate": 9.739705898099743e-06,
+      "loss": 0.5119,
+      "step": 849
+    },
+    {
+      "epoch": 0.42199329775350625,
+      "grad_norm": 0.07016946900319038,
+      "learning_rate": 9.739083070788405e-06,
+      "loss": 0.5482,
+      "step": 850
+    },
+    {
+      "epoch": 0.4224897604567457,
+      "grad_norm": 0.07800094484472588,
+      "learning_rate": 9.738459519184203e-06,
+      "loss": 0.5533,
+      "step": 851
+    },
+    {
+      "epoch": 0.42298622315998513,
+      "grad_norm": 0.07530706526398431,
+      "learning_rate": 9.737835243382438e-06,
+      "loss": 0.6289,
+      "step": 852
+    },
+    {
+      "epoch": 0.4234826858632245,
+      "grad_norm": 0.07309750256723833,
+      "learning_rate": 9.737210243478522e-06,
+      "loss": 0.5388,
+      "step": 853
+    },
+    {
+      "epoch": 0.42397914856646396,
+      "grad_norm": 0.07402449977170224,
+      "learning_rate": 9.736584519567976e-06,
+      "loss": 0.5155,
+      "step": 854
+    },
+    {
+      "epoch": 0.42447561126970335,
+      "grad_norm": 0.07780518098869948,
+      "learning_rate": 9.735958071746431e-06,
+      "loss": 0.5241,
+      "step": 855
+    },
+    {
+      "epoch": 0.4249720739729428,
+      "grad_norm": 0.07374222764934292,
+      "learning_rate": 9.735330900109631e-06,
+      "loss": 0.5132,
+      "step": 856
+    },
+    {
+      "epoch": 0.4254685366761822,
+      "grad_norm": 0.0745070085758739,
+      "learning_rate": 9.734703004753429e-06,
+      "loss": 0.5271,
+      "step": 857
+    },
+    {
+      "epoch": 0.4259649993794216,
+      "grad_norm": 0.08051238332124257,
+      "learning_rate": 9.734074385773786e-06,
+      "loss": 0.4877,
+      "step": 858
+    },
+    {
+      "epoch": 0.42646146208266106,
+      "grad_norm": 0.07259415930172602,
+      "learning_rate": 9.733445043266779e-06,
+      "loss": 0.5827,
+      "step": 859
+    },
+    {
+      "epoch": 0.42695792478590044,
+      "grad_norm": 0.07557236145733463,
+      "learning_rate": 9.732814977328593e-06,
+      "loss": 0.5468,
+      "step": 860
+    },
+    {
+      "epoch": 0.4274543874891399,
+      "grad_norm": 0.07259157033280643,
+      "learning_rate": 9.732184188055522e-06,
+      "loss": 0.5329,
+      "step": 861
+    },
+    {
+      "epoch": 0.4279508501923793,
+      "grad_norm": 0.07549072522153745,
+      "learning_rate": 9.731552675543972e-06,
+      "loss": 0.5211,
+      "step": 862
+    },
+    {
+      "epoch": 0.4284473128956187,
+      "grad_norm": 0.07288970060708126,
+      "learning_rate": 9.73092043989046e-06,
+      "loss": 0.5334,
+      "step": 863
+    },
+    {
+      "epoch": 0.42894377559885816,
+      "grad_norm": 0.07075634318269011,
+      "learning_rate": 9.730287481191615e-06,
+      "loss": 0.5053,
+      "step": 864
+    },
+    {
+      "epoch": 0.42944023830209754,
+      "grad_norm": 0.07614606694018755,
+      "learning_rate": 9.729653799544171e-06,
+      "loss": 0.5587,
+      "step": 865
+    },
+    {
+      "epoch": 0.429936701005337,
+      "grad_norm": 0.07316335947524487,
+      "learning_rate": 9.729019395044979e-06,
+      "loss": 0.5342,
+      "step": 866
+    },
+    {
+      "epoch": 0.43043316370857637,
+      "grad_norm": 0.07053572019737249,
+      "learning_rate": 9.728384267790997e-06,
+      "loss": 0.5213,
+      "step": 867
+    },
+    {
+      "epoch": 0.4309296264118158,
+      "grad_norm": 0.07164245763524565,
+      "learning_rate": 9.727748417879293e-06,
+      "loss": 0.5445,
+      "step": 868
+    },
+    {
+      "epoch": 0.43142608911505526,
+      "grad_norm": 0.07308996659537338,
+      "learning_rate": 9.727111845407046e-06,
+      "loss": 0.5487,
+      "step": 869
+    },
+    {
+      "epoch": 0.43192255181829464,
+      "grad_norm": 0.0806618601368084,
+      "learning_rate": 9.726474550471549e-06,
+      "loss": 0.5628,
+      "step": 870
+    },
+    {
+      "epoch": 0.4324190145215341,
+      "grad_norm": 0.11999394247964658,
+      "learning_rate": 9.725836533170199e-06,
+      "loss": 0.5501,
+      "step": 871
+    },
+    {
+      "epoch": 0.43291547722477347,
+      "grad_norm": 0.07338156895848527,
+      "learning_rate": 9.725197793600508e-06,
+      "loss": 0.5529,
+      "step": 872
+    },
+    {
+      "epoch": 0.4334119399280129,
+      "grad_norm": 0.07414444706146318,
+      "learning_rate": 9.724558331860097e-06,
+      "loss": 0.522,
+      "step": 873
+    },
+    {
+      "epoch": 0.43390840263125235,
+      "grad_norm": 0.07241558922926562,
+      "learning_rate": 9.723918148046696e-06,
+      "loss": 0.4884,
+      "step": 874
+    },
+    {
+      "epoch": 0.43440486533449174,
+      "grad_norm": 0.07214285111060169,
+      "learning_rate": 9.723277242258151e-06,
+      "loss": 0.5147,
+      "step": 875
+    },
+    {
+      "epoch": 0.4349013280377312,
+      "grad_norm": 0.0754089726353843,
+      "learning_rate": 9.72263561459241e-06,
+      "loss": 0.5296,
+      "step": 876
+    },
+    {
+      "epoch": 0.43539779074097057,
+      "grad_norm": 0.07237060558190542,
+      "learning_rate": 9.721993265147539e-06,
+      "loss": 0.5269,
+      "step": 877
+    },
+    {
+      "epoch": 0.43589425344421,
+      "grad_norm": 0.07486793573002197,
+      "learning_rate": 9.721350194021705e-06,
+      "loss": 0.5243,
+      "step": 878
+    },
+    {
+      "epoch": 0.4363907161474494,
+      "grad_norm": 0.08538442120689331,
+      "learning_rate": 9.720706401313199e-06,
+      "loss": 0.5686,
+      "step": 879
+    },
+    {
+      "epoch": 0.43688717885068884,
+      "grad_norm": 0.06871510871572026,
+      "learning_rate": 9.720061887120408e-06,
+      "loss": 0.5311,
+      "step": 880
+    },
+    {
+      "epoch": 0.4373836415539283,
+      "grad_norm": 0.07273195512905156,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 0.527,
+      "step": 881
+    },
+    {
+      "epoch": 0.43788010425716767,
+      "grad_norm": 0.07259430458734004,
+      "learning_rate": 9.718770694676103e-06,
+      "loss": 0.5396,
+      "step": 882
+    },
+    {
+      "epoch": 0.4383765669604071,
+      "grad_norm": 0.0744120171521945,
+      "learning_rate": 9.718124016621929e-06,
+      "loss": 0.5203,
+      "step": 883
+    },
+    {
+      "epoch": 0.4388730296636465,
+      "grad_norm": 0.0766040114467303,
+      "learning_rate": 9.717476617478146e-06,
+      "loss": 0.5254,
+      "step": 884
+    },
+    {
+      "epoch": 0.43936949236688594,
+      "grad_norm": 0.07952016478925952,
+      "learning_rate": 9.716828497343702e-06,
+      "loss": 0.6041,
+      "step": 885
+    },
+    {
+      "epoch": 0.4398659550701254,
+      "grad_norm": 0.07515893930429524,
+      "learning_rate": 9.716179656317651e-06,
+      "loss": 0.5671,
+      "step": 886
+    },
+    {
+      "epoch": 0.44036241777336477,
+      "grad_norm": 0.07275833676676684,
+      "learning_rate": 9.715530094499157e-06,
+      "loss": 0.5301,
+      "step": 887
+    },
+    {
+      "epoch": 0.4408588804766042,
+      "grad_norm": 0.07382926331753971,
+      "learning_rate": 9.714879811987496e-06,
+      "loss": 0.5506,
+      "step": 888
+    },
+    {
+      "epoch": 0.4413553431798436,
+      "grad_norm": 0.076328236140326,
+      "learning_rate": 9.714228808882054e-06,
+      "loss": 0.5138,
+      "step": 889
+    },
+    {
+      "epoch": 0.44185180588308304,
+      "grad_norm": 0.07171275885569235,
+      "learning_rate": 9.713577085282325e-06,
+      "loss": 0.5414,
+      "step": 890
+    },
+    {
+      "epoch": 0.4423482685863225,
+      "grad_norm": 0.07231082291098188,
+      "learning_rate": 9.712924641287915e-06,
+      "loss": 0.5006,
+      "step": 891
+    },
+    {
+      "epoch": 0.44284473128956187,
+      "grad_norm": 0.0746099129815945,
+      "learning_rate": 9.712271476998538e-06,
+      "loss": 0.5469,
+      "step": 892
+    },
+    {
+      "epoch": 0.4433411939928013,
+      "grad_norm": 0.07493537326196857,
+      "learning_rate": 9.711617592514024e-06,
+      "loss": 0.5532,
+      "step": 893
+    },
+    {
+      "epoch": 0.4438376566960407,
+      "grad_norm": 0.07910135513976987,
+      "learning_rate": 9.710962987934305e-06,
+      "loss": 0.6094,
+      "step": 894
+    },
+    {
+      "epoch": 0.44433411939928014,
+      "grad_norm": 0.07074797883606643,
+      "learning_rate": 9.710307663359426e-06,
+      "loss": 0.5247,
+      "step": 895
+    },
+    {
+      "epoch": 0.4448305821025195,
+      "grad_norm": 0.07081481487079631,
+      "learning_rate": 9.709651618889546e-06,
+      "loss": 0.5358,
+      "step": 896
+    },
+    {
+      "epoch": 0.44532704480575896,
+      "grad_norm": 0.07530093918611445,
+      "learning_rate": 9.70899485462493e-06,
+      "loss": 0.5725,
+      "step": 897
+    },
+    {
+      "epoch": 0.4458235075089984,
+      "grad_norm": 0.07414071773290506,
+      "learning_rate": 9.708337370665954e-06,
+      "loss": 0.5377,
+      "step": 898
+    },
+    {
+      "epoch": 0.4463199702122378,
+      "grad_norm": 0.07189120190783159,
+      "learning_rate": 9.707679167113102e-06,
+      "loss": 0.5394,
+      "step": 899
+    },
+    {
+      "epoch": 0.44681643291547724,
+      "grad_norm": 0.07351739293690249,
+      "learning_rate": 9.707020244066972e-06,
+      "loss": 0.5597,
+      "step": 900
+    },
+    {
+      "epoch": 0.4473128956187166,
+      "grad_norm": 0.07610739851431275,
+      "learning_rate": 9.70636060162827e-06,
+      "loss": 0.5451,
+      "step": 901
+    },
+    {
+      "epoch": 0.44780935832195606,
+      "grad_norm": 0.07981647785507498,
+      "learning_rate": 9.705700239897809e-06,
+      "loss": 0.5411,
+      "step": 902
+    },
+    {
+      "epoch": 0.4483058210251955,
+      "grad_norm": 0.06992496152308564,
+      "learning_rate": 9.705039158976517e-06,
+      "loss": 0.5144,
+      "step": 903
+    },
+    {
+      "epoch": 0.4488022837284349,
+      "grad_norm": 0.0700880272974536,
+      "learning_rate": 9.70437735896543e-06,
+      "loss": 0.5535,
+      "step": 904
+    },
+    {
+      "epoch": 0.44929874643167433,
+      "grad_norm": 0.07061113641641646,
+      "learning_rate": 9.70371483996569e-06,
+      "loss": 0.5413,
+      "step": 905
+    },
+    {
+      "epoch": 0.4497952091349137,
+      "grad_norm": 0.0736462909733856,
+      "learning_rate": 9.703051602078557e-06,
+      "loss": 0.5498,
+      "step": 906
+    },
+    {
+      "epoch": 0.45029167183815316,
+      "grad_norm": 0.07228895838116739,
+      "learning_rate": 9.702387645405396e-06,
+      "loss": 0.5474,
+      "step": 907
+    },
+    {
+      "epoch": 0.4507881345413926,
+      "grad_norm": 0.07671904216956826,
+      "learning_rate": 9.701722970047679e-06,
+      "loss": 0.5754,
+      "step": 908
+    },
+    {
+      "epoch": 0.451284597244632,
+      "grad_norm": 0.07480903224238651,
+      "learning_rate": 9.701057576106991e-06,
+      "loss": 0.5272,
+      "step": 909
+    },
+    {
+      "epoch": 0.45178105994787143,
+      "grad_norm": 0.0748176099276939,
+      "learning_rate": 9.700391463685029e-06,
+      "loss": 0.5494,
+      "step": 910
+    },
+    {
+      "epoch": 0.4522775226511108,
+      "grad_norm": 0.09152256842941854,
+      "learning_rate": 9.699724632883598e-06,
+      "loss": 0.5582,
+      "step": 911
+    },
+    {
+      "epoch": 0.45277398535435026,
+      "grad_norm": 0.07611999755029966,
+      "learning_rate": 9.699057083804609e-06,
+      "loss": 0.5313,
+      "step": 912
+    },
+    {
+      "epoch": 0.45327044805758965,
+      "grad_norm": 0.07116184510313514,
+      "learning_rate": 9.69838881655009e-06,
+      "loss": 0.56,
+      "step": 913
+    },
+    {
+      "epoch": 0.4537669107608291,
+      "grad_norm": 0.07594954740931627,
+      "learning_rate": 9.697719831222173e-06,
+      "loss": 0.5031,
+      "step": 914
+    },
+    {
+      "epoch": 0.45426337346406853,
+      "grad_norm": 0.0746375214764949,
+      "learning_rate": 9.697050127923102e-06,
+      "loss": 0.556,
+      "step": 915
+    },
+    {
+      "epoch": 0.4547598361673079,
+      "grad_norm": 0.07771038247489137,
+      "learning_rate": 9.69637970675523e-06,
+      "loss": 0.5592,
+      "step": 916
+    },
+    {
+      "epoch": 0.45525629887054736,
+      "grad_norm": 0.07327276935059207,
+      "learning_rate": 9.695708567821021e-06,
+      "loss": 0.5774,
+      "step": 917
+    },
+    {
+      "epoch": 0.45575276157378675,
+      "grad_norm": 0.07594839409251314,
+      "learning_rate": 9.695036711223049e-06,
+      "loss": 0.6068,
+      "step": 918
+    },
+    {
+      "epoch": 0.4562492242770262,
+      "grad_norm": 0.07720893984030859,
+      "learning_rate": 9.694364137063993e-06,
+      "loss": 0.5783,
+      "step": 919
+    },
+    {
+      "epoch": 0.45674568698026563,
+      "grad_norm": 0.07205569253228451,
+      "learning_rate": 9.693690845446647e-06,
+      "loss": 0.5348,
+      "step": 920
+    },
+    {
+      "epoch": 0.457242149683505,
+      "grad_norm": 0.07485535224592804,
+      "learning_rate": 9.693016836473913e-06,
+      "loss": 0.506,
+      "step": 921
+    },
+    {
+      "epoch": 0.45773861238674446,
+      "grad_norm": 0.07966542633741693,
+      "learning_rate": 9.692342110248802e-06,
+      "loss": 0.5605,
+      "step": 922
+    },
+    {
+      "epoch": 0.45823507508998385,
+      "grad_norm": 0.07722893541906527,
+      "learning_rate": 9.691666666874438e-06,
+      "loss": 0.5184,
+      "step": 923
+    },
+    {
+      "epoch": 0.4587315377932233,
+      "grad_norm": 0.0737626275345602,
+      "learning_rate": 9.690990506454045e-06,
+      "loss": 0.5655,
+      "step": 924
+    },
+    {
+      "epoch": 0.45922800049646273,
+      "grad_norm": 0.07332977419296617,
+      "learning_rate": 9.69031362909097e-06,
+      "loss": 0.5281,
+      "step": 925
+    },
+    {
+      "epoch": 0.4597244631997021,
+      "grad_norm": 0.07823065260085653,
+      "learning_rate": 9.689636034888662e-06,
+      "loss": 0.5293,
+      "step": 926
+    },
+    {
+      "epoch": 0.46022092590294156,
+      "grad_norm": 0.07332590189924752,
+      "learning_rate": 9.688957723950675e-06,
+      "loss": 0.5392,
+      "step": 927
+    },
+    {
+      "epoch": 0.46071738860618094,
+      "grad_norm": 0.07084330672852263,
+      "learning_rate": 9.688278696380684e-06,
+      "loss": 0.5319,
+      "step": 928
+    },
+    {
+      "epoch": 0.4612138513094204,
+      "grad_norm": 0.0737821820663325,
+      "learning_rate": 9.687598952282462e-06,
+      "loss": 0.5817,
+      "step": 929
+    },
+    {
+      "epoch": 0.4617103140126598,
+      "grad_norm": 0.08094122475419381,
+      "learning_rate": 9.686918491759904e-06,
+      "loss": 0.5447,
+      "step": 930
+    },
+    {
+      "epoch": 0.4622067767158992,
+      "grad_norm": 0.07590994311402643,
+      "learning_rate": 9.686237314917e-06,
+      "loss": 0.5669,
+      "step": 931
+    },
+    {
+      "epoch": 0.46270323941913866,
+      "grad_norm": 0.07528128755147503,
+      "learning_rate": 9.685555421857864e-06,
+      "loss": 0.5497,
+      "step": 932
+    },
+    {
+      "epoch": 0.46319970212237804,
+      "grad_norm": 0.07434670733221073,
+      "learning_rate": 9.684872812686706e-06,
+      "loss": 0.5615,
+      "step": 933
+    },
+    {
+      "epoch": 0.4636961648256175,
+      "grad_norm": 0.0690160514658518,
+      "learning_rate": 9.684189487507857e-06,
+      "loss": 0.5144,
+      "step": 934
+    },
+    {
+      "epoch": 0.46419262752885687,
+      "grad_norm": 0.07220825561231929,
+      "learning_rate": 9.683505446425749e-06,
+      "loss": 0.5099,
+      "step": 935
+    },
+    {
+      "epoch": 0.4646890902320963,
+      "grad_norm": 0.07408178533769717,
+      "learning_rate": 9.682820689544927e-06,
+      "loss": 0.5517,
+      "step": 936
+    },
+    {
+      "epoch": 0.46518555293533576,
+      "grad_norm": 0.0760943537779404,
+      "learning_rate": 9.682135216970048e-06,
+      "loss": 0.5589,
+      "step": 937
+    },
+    {
+      "epoch": 0.46568201563857514,
+      "grad_norm": 0.07109341085075326,
+      "learning_rate": 9.681449028805872e-06,
+      "loss": 0.5042,
+      "step": 938
+    },
+    {
+      "epoch": 0.4661784783418146,
+      "grad_norm": 0.07783378438863178,
+      "learning_rate": 9.680762125157273e-06,
+      "loss": 0.5291,
+      "step": 939
+    },
+    {
+      "epoch": 0.46667494104505397,
+      "grad_norm": 0.07299282855701139,
+      "learning_rate": 9.680074506129235e-06,
+      "loss": 0.5279,
+      "step": 940
+    },
+    {
+      "epoch": 0.4671714037482934,
+      "grad_norm": 0.13566447811720078,
+      "learning_rate": 9.679386171826846e-06,
+      "loss": 0.5316,
+      "step": 941
+    },
+    {
+      "epoch": 0.46766786645153285,
+      "grad_norm": 0.07187672347320467,
+      "learning_rate": 9.678697122355311e-06,
+      "loss": 0.5478,
+      "step": 942
+    },
+    {
+      "epoch": 0.46816432915477224,
+      "grad_norm": 0.07340560611836003,
+      "learning_rate": 9.678007357819936e-06,
+      "loss": 0.5629,
+      "step": 943
+    },
+    {
+      "epoch": 0.4686607918580117,
+      "grad_norm": 0.08015408546831505,
+      "learning_rate": 9.677316878326144e-06,
+      "loss": 0.5513,
+      "step": 944
+    },
+    {
+      "epoch": 0.46915725456125107,
+      "grad_norm": 0.07401280117648018,
+      "learning_rate": 9.676625683979462e-06,
+      "loss": 0.5633,
+      "step": 945
+    },
+    {
+      "epoch": 0.4696537172644905,
+      "grad_norm": 0.07324544085941354,
+      "learning_rate": 9.675933774885529e-06,
+      "loss": 0.5684,
+      "step": 946
+    },
+    {
+      "epoch": 0.4701501799677299,
+      "grad_norm": 0.07175023640004917,
+      "learning_rate": 9.67524115115009e-06,
+      "loss": 0.5248,
+      "step": 947
+    },
+    {
+      "epoch": 0.47064664267096934,
+      "grad_norm": 0.07095640867631502,
+      "learning_rate": 9.674547812879002e-06,
+      "loss": 0.519,
+      "step": 948
+    },
+    {
+      "epoch": 0.4711431053742088,
+      "grad_norm": 0.07508488266902727,
+      "learning_rate": 9.673853760178233e-06,
+      "loss": 0.5225,
+      "step": 949
+    },
+    {
+      "epoch": 0.47163956807744817,
+      "grad_norm": 0.07858038399219275,
+      "learning_rate": 9.673158993153857e-06,
+      "loss": 0.5636,
+      "step": 950
+    },
+    {
+      "epoch": 0.4721360307806876,
+      "grad_norm": 0.07043208671603315,
+      "learning_rate": 9.672463511912056e-06,
+      "loss": 0.5295,
+      "step": 951
+    },
+    {
+      "epoch": 0.472632493483927,
+      "grad_norm": 0.07784227872499093,
+      "learning_rate": 9.671767316559124e-06,
+      "loss": 0.5456,
+      "step": 952
+    },
+    {
+      "epoch": 0.47312895618716644,
+      "grad_norm": 0.07574228547831556,
+      "learning_rate": 9.671070407201465e-06,
+      "loss": 0.5656,
+      "step": 953
+    },
+    {
+      "epoch": 0.4736254188904059,
+      "grad_norm": 0.06957012762899452,
+      "learning_rate": 9.67037278394559e-06,
+      "loss": 0.5382,
+      "step": 954
+    },
+    {
+      "epoch": 0.47412188159364527,
+      "grad_norm": 0.09679534685898586,
+      "learning_rate": 9.669674446898116e-06,
+      "loss": 0.5197,
+      "step": 955
+    },
+    {
+      "epoch": 0.4746183442968847,
+      "grad_norm": 0.06803644728978617,
+      "learning_rate": 9.668975396165776e-06,
+      "loss": 0.4857,
+      "step": 956
+    },
+    {
+      "epoch": 0.4751148070001241,
+      "grad_norm": 0.07328282551492919,
+      "learning_rate": 9.66827563185541e-06,
+      "loss": 0.5177,
+      "step": 957
+    },
+    {
+      "epoch": 0.47561126970336354,
+      "grad_norm": 0.06901815690540343,
+      "learning_rate": 9.667575154073962e-06,
+      "loss": 0.5271,
+      "step": 958
+    },
+    {
+      "epoch": 0.476107732406603,
+      "grad_norm": 0.071811828342126,
+      "learning_rate": 9.666873962928491e-06,
+      "loss": 0.5402,
+      "step": 959
+    },
+    {
+      "epoch": 0.47660419510984237,
+      "grad_norm": 0.0745044806945251,
+      "learning_rate": 9.666172058526162e-06,
+      "loss": 0.5485,
+      "step": 960
+    },
+    {
+      "epoch": 0.4771006578130818,
+      "grad_norm": 0.06727332445949166,
+      "learning_rate": 9.66546944097425e-06,
+      "loss": 0.5302,
+      "step": 961
+    },
+    {
+      "epoch": 0.4775971205163212,
+      "grad_norm": 0.0723126514873621,
+      "learning_rate": 9.664766110380141e-06,
+      "loss": 0.5599,
+      "step": 962
+    },
+    {
+      "epoch": 0.47809358321956064,
+      "grad_norm": 0.07395350429498122,
+      "learning_rate": 9.664062066851325e-06,
+      "loss": 0.5192,
+      "step": 963
+    },
+    {
+      "epoch": 0.4785900459228,
+      "grad_norm": 0.07378023026691215,
+      "learning_rate": 9.663357310495404e-06,
+      "loss": 0.5469,
+      "step": 964
+    },
+    {
+      "epoch": 0.47908650862603946,
+      "grad_norm": 0.07156838728300292,
+      "learning_rate": 9.66265184142009e-06,
+      "loss": 0.5467,
+      "step": 965
+    },
+    {
+      "epoch": 0.4795829713292789,
+      "grad_norm": 0.07649205641151963,
+      "learning_rate": 9.661945659733201e-06,
+      "loss": 0.5472,
+      "step": 966
+    },
+    {
+      "epoch": 0.4800794340325183,
+      "grad_norm": 0.0729425278563696,
+      "learning_rate": 9.661238765542668e-06,
+      "loss": 0.5397,
+      "step": 967
+    },
+    {
+      "epoch": 0.48057589673575773,
+      "grad_norm": 0.07096620529347483,
+      "learning_rate": 9.660531158956525e-06,
+      "loss": 0.5676,
+      "step": 968
+    },
+    {
+      "epoch": 0.4810723594389971,
+      "grad_norm": 0.07513824604507036,
+      "learning_rate": 9.659822840082922e-06,
+      "loss": 0.5125,
+      "step": 969
+    },
+    {
+      "epoch": 0.48156882214223656,
+      "grad_norm": 0.07212660742880082,
+      "learning_rate": 9.659113809030112e-06,
+      "loss": 0.5174,
+      "step": 970
+    },
+    {
+      "epoch": 0.482065284845476,
+      "grad_norm": 0.07547232447246666,
+      "learning_rate": 9.65840406590646e-06,
+      "loss": 0.5458,
+      "step": 971
+    },
+    {
+      "epoch": 0.4825617475487154,
+      "grad_norm": 0.07145748264825304,
+      "learning_rate": 9.657693610820437e-06,
+      "loss": 0.5544,
+      "step": 972
+    },
+    {
+      "epoch": 0.48305821025195483,
+      "grad_norm": 0.07035145889253784,
+      "learning_rate": 9.656982443880626e-06,
+      "loss": 0.5336,
+      "step": 973
+    },
+    {
+      "epoch": 0.4835546729551942,
+      "grad_norm": 0.07483673669785795,
+      "learning_rate": 9.65627056519572e-06,
+      "loss": 0.5237,
+      "step": 974
+    },
+    {
+      "epoch": 0.48405113565843366,
+      "grad_norm": 0.07455887409410537,
+      "learning_rate": 9.655557974874512e-06,
+      "loss": 0.5963,
+      "step": 975
+    },
+    {
+      "epoch": 0.4845475983616731,
+      "grad_norm": 0.09159875498697193,
+      "learning_rate": 9.654844673025917e-06,
+      "loss": 0.6166,
+      "step": 976
+    },
+    {
+      "epoch": 0.4850440610649125,
+      "grad_norm": 0.07885063412079849,
+      "learning_rate": 9.654130659758947e-06,
+      "loss": 0.5867,
+      "step": 977
+    },
+    {
+      "epoch": 0.48554052376815193,
+      "grad_norm": 0.07055769737141665,
+      "learning_rate": 9.653415935182728e-06,
+      "loss": 0.5366,
+      "step": 978
+    },
+    {
+      "epoch": 0.4860369864713913,
+      "grad_norm": 0.07245998065157473,
+      "learning_rate": 9.652700499406497e-06,
+      "loss": 0.5324,
+      "step": 979
+    },
+    {
+      "epoch": 0.48653344917463076,
+      "grad_norm": 0.07611782049187152,
+      "learning_rate": 9.651984352539595e-06,
+      "loss": 0.5655,
+      "step": 980
+    },
+    {
+      "epoch": 0.48702991187787015,
+      "grad_norm": 0.07353257519946095,
+      "learning_rate": 9.651267494691471e-06,
+      "loss": 0.5203,
+      "step": 981
+    },
+    {
+      "epoch": 0.4875263745811096,
+      "grad_norm": 0.07449068355216579,
+      "learning_rate": 9.65054992597169e-06,
+      "loss": 0.5287,
+      "step": 982
+    },
+    {
+      "epoch": 0.48802283728434903,
+      "grad_norm": 0.07332839129254907,
+      "learning_rate": 9.64983164648992e-06,
+      "loss": 0.5633,
+      "step": 983
+    },
+    {
+      "epoch": 0.4885192999875884,
+      "grad_norm": 0.07099306409257887,
+      "learning_rate": 9.649112656355936e-06,
+      "loss": 0.5225,
+      "step": 984
+    },
+    {
+      "epoch": 0.48901576269082786,
+      "grad_norm": 0.07467585354112222,
+      "learning_rate": 9.648392955679624e-06,
+      "loss": 0.5589,
+      "step": 985
+    },
+    {
+      "epoch": 0.48951222539406725,
+      "grad_norm": 0.07407066249938493,
+      "learning_rate": 9.647672544570981e-06,
+      "loss": 0.5499,
+      "step": 986
+    },
+    {
+      "epoch": 0.4900086880973067,
+      "grad_norm": 0.06921009250090104,
+      "learning_rate": 9.64695142314011e-06,
+      "loss": 0.5514,
+      "step": 987
+    },
+    {
+      "epoch": 0.49050515080054613,
+      "grad_norm": 0.07365692083114214,
+      "learning_rate": 9.646229591497222e-06,
+      "loss": 0.5355,
+      "step": 988
+    },
+    {
+      "epoch": 0.4910016135037855,
+      "grad_norm": 0.0718859256778461,
+      "learning_rate": 9.645507049752637e-06,
+      "loss": 0.4963,
+      "step": 989
+    },
+    {
+      "epoch": 0.49149807620702496,
+      "grad_norm": 0.07324913526100718,
+      "learning_rate": 9.644783798016785e-06,
+      "loss": 0.5726,
+      "step": 990
+    },
+    {
+      "epoch": 0.49199453891026435,
+      "grad_norm": 0.08062267007727522,
+      "learning_rate": 9.644059836400203e-06,
+      "loss": 0.5604,
+      "step": 991
+    },
+    {
+      "epoch": 0.4924910016135038,
+      "grad_norm": 0.0721895312156475,
+      "learning_rate": 9.643335165013536e-06,
+      "loss": 0.5275,
+      "step": 992
+    },
+    {
+      "epoch": 0.49298746431674323,
+      "grad_norm": 0.07242784574028176,
+      "learning_rate": 9.642609783967539e-06,
+      "loss": 0.5313,
+      "step": 993
+    },
+    {
+      "epoch": 0.4934839270199826,
+      "grad_norm": 0.07572575082005072,
+      "learning_rate": 9.641883693373077e-06,
+      "loss": 0.5442,
+      "step": 994
+    },
+    {
+      "epoch": 0.49398038972322206,
+      "grad_norm": 0.07228182174445916,
+      "learning_rate": 9.641156893341117e-06,
+      "loss": 0.5612,
+      "step": 995
+    },
+    {
+      "epoch": 0.49447685242646144,
+      "grad_norm": 0.07245474432940825,
+      "learning_rate": 9.640429383982743e-06,
+      "loss": 0.522,
+      "step": 996
+    },
+    {
+      "epoch": 0.4949733151297009,
+      "grad_norm": 0.07547168525249354,
+      "learning_rate": 9.63970116540914e-06,
+      "loss": 0.5637,
+      "step": 997
+    },
+    {
+      "epoch": 0.49546977783294033,
+      "grad_norm": 0.07571162844471516,
+      "learning_rate": 9.638972237731608e-06,
+      "loss": 0.5458,
+      "step": 998
+    },
+    {
+      "epoch": 0.4959662405361797,
+      "grad_norm": 0.06792756036208669,
+      "learning_rate": 9.638242601061547e-06,
+      "loss": 0.5037,
+      "step": 999
+    },
+    {
+      "epoch": 0.49646270323941916,
+      "grad_norm": 0.07543208718463203,
+      "learning_rate": 9.637512255510475e-06,
+      "loss": 0.5627,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49695916594265854,
+      "grad_norm": 0.0744355804303214,
+      "learning_rate": 9.63678120119001e-06,
+      "loss": 0.5292,
+      "step": 1001
+    },
+    {
+      "epoch": 0.497455628645898,
+      "grad_norm": 0.07343036450567166,
+      "learning_rate": 9.636049438211883e-06,
+      "loss": 0.5386,
+      "step": 1002
+    },
+    {
+      "epoch": 0.49795209134913737,
+      "grad_norm": 0.0705676280818943,
+      "learning_rate": 9.635316966687935e-06,
+      "loss": 0.5206,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4984485540523768,
+      "grad_norm": 0.07682429996586587,
+      "learning_rate": 9.63458378673011e-06,
+      "loss": 0.5468,
+      "step": 1004
+    },
+    {
+      "epoch": 0.49894501675561626,
+      "grad_norm": 0.07274415834069523,
+      "learning_rate": 9.633849898450463e-06,
+      "loss": 0.5605,
+      "step": 1005
+    },
+    {
+      "epoch": 0.49944147945885564,
+      "grad_norm": 0.06851151007874164,
+      "learning_rate": 9.633115301961156e-06,
+      "loss": 0.4902,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4999379421620951,
+      "grad_norm": 0.07046304690001261,
+      "learning_rate": 9.632379997374462e-06,
+      "loss": 0.496,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5004344048653345,
+      "grad_norm": 0.07583845542191688,
+      "learning_rate": 9.63164398480276e-06,
+      "loss": 0.5543,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5004344048653345,
+      "eval_loss": 0.5403582453727722,
+      "eval_runtime": 259.5561,
+      "eval_samples_per_second": 116.942,
+      "eval_steps_per_second": 14.621,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5009308675685739,
+      "grad_norm": 0.07170711724531288,
+      "learning_rate": 9.630907264358538e-06,
+      "loss": 0.5503,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5014273302718133,
+      "grad_norm": 0.07524070502561764,
+      "learning_rate": 9.630169836154391e-06,
+      "loss": 0.5661,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5019237929750527,
+      "grad_norm": 0.07608869983603515,
+      "learning_rate": 9.629431700303025e-06,
+      "loss": 0.5455,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5024202556782922,
+      "grad_norm": 0.07104754400679542,
+      "learning_rate": 9.628692856917249e-06,
+      "loss": 0.5146,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5029167183815316,
+      "grad_norm": 0.0782620771521833,
+      "learning_rate": 9.627953306109985e-06,
+      "loss": 0.5954,
+      "step": 1013
+    },
+    {
+      "epoch": 0.503413181084771,
+      "grad_norm": 0.07064620528254402,
+      "learning_rate": 9.627213047994265e-06,
+      "loss": 0.5271,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5039096437880104,
+      "grad_norm": 0.0728348121611852,
+      "learning_rate": 9.62647208268322e-06,
+      "loss": 0.5188,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5044061064912498,
+      "grad_norm": 0.076035071497151,
+      "learning_rate": 9.625730410290097e-06,
+      "loss": 0.5236,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5049025691944893,
+      "grad_norm": 0.07007830649545077,
+      "learning_rate": 9.624988030928248e-06,
+      "loss": 0.539,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5053990318977287,
+      "grad_norm": 0.07422124598706549,
+      "learning_rate": 9.624244944711137e-06,
+      "loss": 0.5177,
+      "step": 1018
+    },
+    {
+      "epoch": 0.505895494600968,
+      "grad_norm": 0.07464012356026203,
+      "learning_rate": 9.623501151752329e-06,
+      "loss": 0.5557,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5063919573042075,
+      "grad_norm": 0.07069286385577062,
+      "learning_rate": 9.622756652165501e-06,
+      "loss": 0.5034,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5068884200074469,
+      "grad_norm": 0.0731620752819763,
+      "learning_rate": 9.622011446064439e-06,
+      "loss": 0.5405,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5073848827106864,
+      "grad_norm": 0.0747099004906738,
+      "learning_rate": 9.621265533563038e-06,
+      "loss": 0.5481,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5078813454139258,
+      "grad_norm": 0.0724421928759601,
+      "learning_rate": 9.620518914775295e-06,
+      "loss": 0.53,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5083778081171652,
+      "grad_norm": 0.07314562621509882,
+      "learning_rate": 9.61977158981532e-06,
+      "loss": 0.5443,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5088742708204046,
+      "grad_norm": 0.07300052887850178,
+      "learning_rate": 9.61902355879733e-06,
+      "loss": 0.5279,
+      "step": 1025
+    },
+    {
+      "epoch": 0.509370733523644,
+      "grad_norm": 0.06762167908780276,
+      "learning_rate": 9.61827482183565e-06,
+      "loss": 0.4825,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5098671962268835,
+      "grad_norm": 0.07544837802714413,
+      "learning_rate": 9.617525379044712e-06,
+      "loss": 0.5472,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5103636589301229,
+      "grad_norm": 0.07456712402407543,
+      "learning_rate": 9.616775230539057e-06,
+      "loss": 0.5526,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5108601216333623,
+      "grad_norm": 0.0686351546215355,
+      "learning_rate": 9.61602437643333e-06,
+      "loss": 0.5186,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5113565843366017,
+      "grad_norm": 0.07819276782161572,
+      "learning_rate": 9.615272816842292e-06,
+      "loss": 0.5653,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5118530470398411,
+      "grad_norm": 0.07382695755499852,
+      "learning_rate": 9.614520551880802e-06,
+      "loss": 0.5593,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5123495097430806,
+      "grad_norm": 0.07567047708008438,
+      "learning_rate": 9.613767581663836e-06,
+      "loss": 0.5334,
+      "step": 1032
+    },
+    {
+      "epoch": 0.51284597244632,
+      "grad_norm": 0.07617923757435242,
+      "learning_rate": 9.61301390630647e-06,
+      "loss": 0.5528,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5133424351495594,
+      "grad_norm": 0.07073358897699591,
+      "learning_rate": 9.612259525923893e-06,
+      "loss": 0.5084,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5138388978527988,
+      "grad_norm": 0.08154535797533109,
+      "learning_rate": 9.611504440631398e-06,
+      "loss": 0.5354,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5143353605560382,
+      "grad_norm": 0.07259202484825346,
+      "learning_rate": 9.610748650544391e-06,
+      "loss": 0.531,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5148318232592777,
+      "grad_norm": 0.07867424820741396,
+      "learning_rate": 9.609992155778377e-06,
+      "loss": 0.5624,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5153282859625171,
+      "grad_norm": 0.07292637512788251,
+      "learning_rate": 9.609234956448983e-06,
+      "loss": 0.5455,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5158247486657564,
+      "grad_norm": 0.07415662463746647,
+      "learning_rate": 9.608477052671926e-06,
+      "loss": 0.5407,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5163212113689959,
+      "grad_norm": 0.07079322361068474,
+      "learning_rate": 9.607718444563044e-06,
+      "loss": 0.5477,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5168176740722353,
+      "grad_norm": 0.07276915746515902,
+      "learning_rate": 9.606959132238276e-06,
+      "loss": 0.5467,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5173141367754748,
+      "grad_norm": 0.07502606002689027,
+      "learning_rate": 9.606199115813672e-06,
+      "loss": 0.5336,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5178105994787141,
+      "grad_norm": 0.07326432550036611,
+      "learning_rate": 9.605438395405388e-06,
+      "loss": 0.557,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5183070621819535,
+      "grad_norm": 0.07453174120545882,
+      "learning_rate": 9.604676971129687e-06,
+      "loss": 0.5426,
+      "step": 1044
+    },
+    {
+      "epoch": 0.518803524885193,
+      "grad_norm": 0.07528215530131163,
+      "learning_rate": 9.603914843102941e-06,
+      "loss": 0.5746,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5192999875884324,
+      "grad_norm": 0.07100176094569843,
+      "learning_rate": 9.603152011441631e-06,
+      "loss": 0.5121,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5197964502916719,
+      "grad_norm": 0.07327171771917512,
+      "learning_rate": 9.602388476262342e-06,
+      "loss": 0.507,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5202929129949112,
+      "grad_norm": 0.0748795466452681,
+      "learning_rate": 9.601624237681769e-06,
+      "loss": 0.5467,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5207893756981506,
+      "grad_norm": 0.07123730635813282,
+      "learning_rate": 9.600859295816708e-06,
+      "loss": 0.5062,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5212858384013901,
+      "grad_norm": 0.07552662977747024,
+      "learning_rate": 9.600093650784077e-06,
+      "loss": 0.541,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5217823011046295,
+      "grad_norm": 0.069532099508212,
+      "learning_rate": 9.599327302700888e-06,
+      "loss": 0.5098,
+      "step": 1051
+    },
+    {
+      "epoch": 0.522278763807869,
+      "grad_norm": 0.07438165535163875,
+      "learning_rate": 9.598560251684265e-06,
+      "loss": 0.5462,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5227752265111083,
+      "grad_norm": 0.07334192360035025,
+      "learning_rate": 9.59779249785144e-06,
+      "loss": 0.5125,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5232716892143477,
+      "grad_norm": 0.0712734477574473,
+      "learning_rate": 9.597024041319752e-06,
+      "loss": 0.5463,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5237681519175872,
+      "grad_norm": 0.0737924898215157,
+      "learning_rate": 9.596254882206645e-06,
+      "loss": 0.5285,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5242646146208266,
+      "grad_norm": 0.07618318262965273,
+      "learning_rate": 9.595485020629676e-06,
+      "loss": 0.5492,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5247610773240661,
+      "grad_norm": 0.07006285871473625,
+      "learning_rate": 9.594714456706504e-06,
+      "loss": 0.504,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5252575400273054,
+      "grad_norm": 0.07399743684753544,
+      "learning_rate": 9.593943190554899e-06,
+      "loss": 0.5292,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5257540027305448,
+      "grad_norm": 0.07619518300934514,
+      "learning_rate": 9.593171222292734e-06,
+      "loss": 0.5795,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5262504654337843,
+      "grad_norm": 0.071100769018996,
+      "learning_rate": 9.592398552037995e-06,
+      "loss": 0.5385,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5267469281370237,
+      "grad_norm": 0.06999405161525964,
+      "learning_rate": 9.59162517990877e-06,
+      "loss": 0.5259,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5272433908402632,
+      "grad_norm": 0.0741871177310929,
+      "learning_rate": 9.590851106023257e-06,
+      "loss": 0.5522,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5277398535435025,
+      "grad_norm": 0.06944998603394864,
+      "learning_rate": 9.590076330499763e-06,
+      "loss": 0.5268,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5282363162467419,
+      "grad_norm": 0.07174352781617274,
+      "learning_rate": 9.589300853456698e-06,
+      "loss": 0.5529,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5287327789499814,
+      "grad_norm": 0.07175214299724526,
+      "learning_rate": 9.58852467501258e-06,
+      "loss": 0.5444,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5292292416532208,
+      "grad_norm": 0.07400820280039133,
+      "learning_rate": 9.587747795286037e-06,
+      "loss": 0.5185,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5297257043564603,
+      "grad_norm": 0.07475007237984427,
+      "learning_rate": 9.586970214395804e-06,
+      "loss": 0.5226,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5302221670596996,
+      "grad_norm": 0.07264770723943843,
+      "learning_rate": 9.586191932460718e-06,
+      "loss": 0.529,
+      "step": 1068
+    },
+    {
+      "epoch": 0.530718629762939,
+      "grad_norm": 0.07050995046572377,
+      "learning_rate": 9.58541294959973e-06,
+      "loss": 0.4913,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5312150924661785,
+      "grad_norm": 0.07586926085270411,
+      "learning_rate": 9.584633265931894e-06,
+      "loss": 0.5405,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5317115551694179,
+      "grad_norm": 0.06959496775952001,
+      "learning_rate": 9.583852881576372e-06,
+      "loss": 0.5308,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5322080178726574,
+      "grad_norm": 0.078697865178032,
+      "learning_rate": 9.583071796652434e-06,
+      "loss": 0.5671,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5327044805758967,
+      "grad_norm": 0.0706600680730896,
+      "learning_rate": 9.582290011279457e-06,
+      "loss": 0.5489,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5332009432791361,
+      "grad_norm": 0.07534168289206372,
+      "learning_rate": 9.581507525576922e-06,
+      "loss": 0.502,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5336974059823756,
+      "grad_norm": 0.07268063895409936,
+      "learning_rate": 9.58072433966442e-06,
+      "loss": 0.5371,
+      "step": 1075
+    },
+    {
+      "epoch": 0.534193868685615,
+      "grad_norm": 0.07665390496410063,
+      "learning_rate": 9.57994045366165e-06,
+      "loss": 0.5468,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5346903313888545,
+      "grad_norm": 0.06923978802164854,
+      "learning_rate": 9.579155867688415e-06,
+      "loss": 0.522,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5351867940920938,
+      "grad_norm": 0.07948336708054222,
+      "learning_rate": 9.578370581864627e-06,
+      "loss": 0.5454,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5356832567953332,
+      "grad_norm": 0.07172413843820762,
+      "learning_rate": 9.577584596310305e-06,
+      "loss": 0.5099,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5361797194985727,
+      "grad_norm": 0.07080159293682067,
+      "learning_rate": 9.576797911145572e-06,
+      "loss": 0.5212,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5366761822018121,
+      "grad_norm": 0.07070283593873587,
+      "learning_rate": 9.576010526490662e-06,
+      "loss": 0.5213,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5371726449050515,
+      "grad_norm": 0.07139818749133874,
+      "learning_rate": 9.575222442465915e-06,
+      "loss": 0.5458,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5376691076082909,
+      "grad_norm": 0.07288959818990044,
+      "learning_rate": 9.574433659191775e-06,
+      "loss": 0.5262,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5381655703115303,
+      "grad_norm": 0.0727858790407329,
+      "learning_rate": 9.573644176788795e-06,
+      "loss": 0.5592,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5386620330147698,
+      "grad_norm": 0.07284011673299268,
+      "learning_rate": 9.572853995377635e-06,
+      "loss": 0.5534,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5391584957180092,
+      "grad_norm": 0.07010630635054382,
+      "learning_rate": 9.572063115079063e-06,
+      "loss": 0.5031,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5396549584212486,
+      "grad_norm": 0.07812136413943721,
+      "learning_rate": 9.57127153601395e-06,
+      "loss": 0.5893,
+      "step": 1087
+    },
+    {
+      "epoch": 0.540151421124488,
+      "grad_norm": 0.07326998894383631,
+      "learning_rate": 9.57047925830328e-06,
+      "loss": 0.5561,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5406478838277274,
+      "grad_norm": 0.06966971851917615,
+      "learning_rate": 9.569686282068135e-06,
+      "loss": 0.5104,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5411443465309669,
+      "grad_norm": 0.07507880630445724,
+      "learning_rate": 9.568892607429712e-06,
+      "loss": 0.5341,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5416408092342063,
+      "grad_norm": 0.07202610071319031,
+      "learning_rate": 9.568098234509312e-06,
+      "loss": 0.555,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5421372719374457,
+      "grad_norm": 0.07189145724012203,
+      "learning_rate": 9.567303163428338e-06,
+      "loss": 0.539,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5426337346406851,
+      "grad_norm": 0.07689989643038242,
+      "learning_rate": 9.566507394308309e-06,
+      "loss": 0.5525,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5431301973439245,
+      "grad_norm": 0.07067817954470944,
+      "learning_rate": 9.565710927270843e-06,
+      "loss": 0.5403,
+      "step": 1094
+    },
+    {
+      "epoch": 0.543626660047164,
+      "grad_norm": 0.07050407504479753,
+      "learning_rate": 9.564913762437667e-06,
+      "loss": 0.5197,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5441231227504034,
+      "grad_norm": 0.07475858179014994,
+      "learning_rate": 9.564115899930614e-06,
+      "loss": 0.54,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5446195854536428,
+      "grad_norm": 0.07052195765532154,
+      "learning_rate": 9.563317339871626e-06,
+      "loss": 0.5485,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5451160481568822,
+      "grad_norm": 0.07254645987577656,
+      "learning_rate": 9.562518082382751e-06,
+      "loss": 0.5237,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5456125108601216,
+      "grad_norm": 0.07384503839511744,
+      "learning_rate": 9.561718127586141e-06,
+      "loss": 0.5251,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5461089735633611,
+      "grad_norm": 0.07695343336131556,
+      "learning_rate": 9.560917475604057e-06,
+      "loss": 0.5794,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5466054362666005,
+      "grad_norm": 0.07275898006626658,
+      "learning_rate": 9.560116126558864e-06,
+      "loss": 0.5573,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5471018989698399,
+      "grad_norm": 0.07389739244860348,
+      "learning_rate": 9.559314080573038e-06,
+      "loss": 0.5215,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5475983616730793,
+      "grad_norm": 0.07810033073554441,
+      "learning_rate": 9.558511337769158e-06,
+      "loss": 0.539,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5480948243763187,
+      "grad_norm": 0.0731889932716798,
+      "learning_rate": 9.557707898269912e-06,
+      "loss": 0.5211,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5485912870795582,
+      "grad_norm": 0.07607334826510194,
+      "learning_rate": 9.55690376219809e-06,
+      "loss": 0.5702,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5490877497827976,
+      "grad_norm": 0.07426832365368147,
+      "learning_rate": 9.556098929676591e-06,
+      "loss": 0.5429,
+      "step": 1106
+    },
+    {
+      "epoch": 0.549584212486037,
+      "grad_norm": 0.07795795275763351,
+      "learning_rate": 9.555293400828422e-06,
+      "loss": 0.5335,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5500806751892764,
+      "grad_norm": 0.07490574809345175,
+      "learning_rate": 9.554487175776697e-06,
+      "loss": 0.5489,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5505771378925158,
+      "grad_norm": 0.07261446271483263,
+      "learning_rate": 9.553680254644631e-06,
+      "loss": 0.5622,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5510736005957553,
+      "grad_norm": 0.07465238791494326,
+      "learning_rate": 9.552872637555553e-06,
+      "loss": 0.5249,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5515700632989947,
+      "grad_norm": 0.07195004690331216,
+      "learning_rate": 9.55206432463289e-06,
+      "loss": 0.5016,
+      "step": 1111
+    },
+    {
+      "epoch": 0.552066526002234,
+      "grad_norm": 0.07826854935318284,
+      "learning_rate": 9.551255316000183e-06,
+      "loss": 0.5317,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5525629887054735,
+      "grad_norm": 0.07676533756395308,
+      "learning_rate": 9.550445611781073e-06,
+      "loss": 0.569,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5530594514087129,
+      "grad_norm": 0.07932871190286644,
+      "learning_rate": 9.549635212099315e-06,
+      "loss": 0.5872,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5535559141119524,
+      "grad_norm": 0.0730298940147842,
+      "learning_rate": 9.54882411707876e-06,
+      "loss": 0.594,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5540523768151917,
+      "grad_norm": 0.07036776038592685,
+      "learning_rate": 9.548012326843374e-06,
+      "loss": 0.5188,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5545488395184311,
+      "grad_norm": 0.0714932448289326,
+      "learning_rate": 9.547199841517228e-06,
+      "loss": 0.5192,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5550453022216706,
+      "grad_norm": 0.0750650728139119,
+      "learning_rate": 9.546386661224492e-06,
+      "loss": 0.5213,
+      "step": 1118
+    },
+    {
+      "epoch": 0.55554176492491,
+      "grad_norm": 0.0718806510750882,
+      "learning_rate": 9.545572786089452e-06,
+      "loss": 0.5466,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5560382276281495,
+      "grad_norm": 0.07230351199367406,
+      "learning_rate": 9.544758216236494e-06,
+      "loss": 0.5218,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5565346903313888,
+      "grad_norm": 0.06979791491219231,
+      "learning_rate": 9.543942951790113e-06,
+      "loss": 0.5168,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5570311530346282,
+      "grad_norm": 0.07957089363539423,
+      "learning_rate": 9.543126992874909e-06,
+      "loss": 0.5539,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5575276157378677,
+      "grad_norm": 0.07492299237904469,
+      "learning_rate": 9.542310339615586e-06,
+      "loss": 0.5481,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5580240784411071,
+      "grad_norm": 0.07431977064563716,
+      "learning_rate": 9.541492992136958e-06,
+      "loss": 0.5316,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5585205411443466,
+      "grad_norm": 0.07458363210156903,
+      "learning_rate": 9.540674950563943e-06,
+      "loss": 0.4946,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5590170038475859,
+      "grad_norm": 0.07055614119573127,
+      "learning_rate": 9.539856215021568e-06,
+      "loss": 0.5266,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5595134665508253,
+      "grad_norm": 0.07321974300708112,
+      "learning_rate": 9.539036785634961e-06,
+      "loss": 0.5345,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5600099292540648,
+      "grad_norm": 0.06997988509814153,
+      "learning_rate": 9.53821666252936e-06,
+      "loss": 0.543,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5605063919573042,
+      "grad_norm": 0.06901480956636961,
+      "learning_rate": 9.537395845830105e-06,
+      "loss": 0.5296,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5610028546605437,
+      "grad_norm": 0.06729539521159718,
+      "learning_rate": 9.536574335662647e-06,
+      "loss": 0.508,
+      "step": 1130
+    },
+    {
+      "epoch": 0.561499317363783,
+      "grad_norm": 0.07692036381800664,
+      "learning_rate": 9.535752132152542e-06,
+      "loss": 0.5671,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5619957800670224,
+      "grad_norm": 0.07201167584301307,
+      "learning_rate": 9.534929235425447e-06,
+      "loss": 0.5745,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5624922427702619,
+      "grad_norm": 0.07778363775383548,
+      "learning_rate": 9.53410564560713e-06,
+      "loss": 0.5466,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5629887054735013,
+      "grad_norm": 0.07348871677637435,
+      "learning_rate": 9.533281362823465e-06,
+      "loss": 0.5215,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5634851681767408,
+      "grad_norm": 0.07291607632635455,
+      "learning_rate": 9.532456387200431e-06,
+      "loss": 0.5355,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5639816308799801,
+      "grad_norm": 0.07240540144940869,
+      "learning_rate": 9.531630718864108e-06,
+      "loss": 0.5696,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5644780935832195,
+      "grad_norm": 0.08262533186527253,
+      "learning_rate": 9.53080435794069e-06,
+      "loss": 0.5232,
+      "step": 1137
+    },
+    {
+      "epoch": 0.564974556286459,
+      "grad_norm": 0.07723729011533378,
+      "learning_rate": 9.52997730455647e-06,
+      "loss": 0.5464,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5654710189896984,
+      "grad_norm": 0.07315058133761201,
+      "learning_rate": 9.529149558837853e-06,
+      "loss": 0.5389,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5659674816929379,
+      "grad_norm": 0.06782611091937678,
+      "learning_rate": 9.528321120911345e-06,
+      "loss": 0.4988,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5664639443961772,
+      "grad_norm": 0.07674985492643183,
+      "learning_rate": 9.527491990903562e-06,
+      "loss": 0.5416,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5669604070994166,
+      "grad_norm": 0.07480549990375804,
+      "learning_rate": 9.526662168941219e-06,
+      "loss": 0.5337,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5674568698026561,
+      "grad_norm": 0.07146162705187283,
+      "learning_rate": 9.525831655151143e-06,
+      "loss": 0.5153,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5679533325058955,
+      "grad_norm": 0.07313978051760994,
+      "learning_rate": 9.525000449660264e-06,
+      "loss": 0.5112,
+      "step": 1144
+    },
+    {
+      "epoch": 0.568449795209135,
+      "grad_norm": 0.07549591293564042,
+      "learning_rate": 9.524168552595621e-06,
+      "loss": 0.5301,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5689462579123743,
+      "grad_norm": 0.07296790622958364,
+      "learning_rate": 9.523335964084352e-06,
+      "loss": 0.5089,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5694427206156137,
+      "grad_norm": 0.074992297201754,
+      "learning_rate": 9.522502684253709e-06,
+      "loss": 0.543,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5699391833188532,
+      "grad_norm": 0.07555928711395048,
+      "learning_rate": 9.521668713231042e-06,
+      "loss": 0.5485,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5704356460220926,
+      "grad_norm": 0.07238072731934045,
+      "learning_rate": 9.520834051143814e-06,
+      "loss": 0.5176,
+      "step": 1149
+    },
+    {
+      "epoch": 0.570932108725332,
+      "grad_norm": 0.07105926905059287,
+      "learning_rate": 9.519998698119586e-06,
+      "loss": 0.5119,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.07693800848614878,
+      "learning_rate": 9.51916265428603e-06,
+      "loss": 0.5527,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5719250341318108,
+      "grad_norm": 0.07346670069398563,
+      "learning_rate": 9.518325919770919e-06,
+      "loss": 0.574,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5724214968350503,
+      "grad_norm": 0.06977185406549447,
+      "learning_rate": 9.517488494702141e-06,
+      "loss": 0.5303,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5729179595382897,
+      "grad_norm": 0.07082572887503037,
+      "learning_rate": 9.516650379207677e-06,
+      "loss": 0.4947,
+      "step": 1154
+    },
+    {
+      "epoch": 0.573414422241529,
+      "grad_norm": 0.07376675912920903,
+      "learning_rate": 9.515811573415621e-06,
+      "loss": 0.5449,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5739108849447685,
+      "grad_norm": 0.07164889694586393,
+      "learning_rate": 9.514972077454171e-06,
+      "loss": 0.5522,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5744073476480079,
+      "grad_norm": 0.07192965800935237,
+      "learning_rate": 9.514131891451632e-06,
+      "loss": 0.5016,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5749038103512474,
+      "grad_norm": 0.07115596024143077,
+      "learning_rate": 9.513291015536413e-06,
+      "loss": 0.516,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5754002730544868,
+      "grad_norm": 0.07566922011447912,
+      "learning_rate": 9.512449449837026e-06,
+      "loss": 0.5476,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5758967357577262,
+      "grad_norm": 0.07194571071415225,
+      "learning_rate": 9.511607194482093e-06,
+      "loss": 0.5302,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5763931984609656,
+      "grad_norm": 0.07100717600020252,
+      "learning_rate": 9.510764249600339e-06,
+      "loss": 0.5227,
+      "step": 1161
+    },
+    {
+      "epoch": 0.576889661164205,
+      "grad_norm": 0.07209205885217734,
+      "learning_rate": 9.509920615320593e-06,
+      "loss": 0.5497,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5773861238674445,
+      "grad_norm": 0.07457087613777853,
+      "learning_rate": 9.509076291771793e-06,
+      "loss": 0.5196,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5778825865706839,
+      "grad_norm": 0.07480024464625332,
+      "learning_rate": 9.508231279082978e-06,
+      "loss": 0.5484,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5783790492739233,
+      "grad_norm": 0.07444790838817751,
+      "learning_rate": 9.507385577383297e-06,
+      "loss": 0.5677,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5788755119771627,
+      "grad_norm": 0.07413996735633248,
+      "learning_rate": 9.506539186802e-06,
+      "loss": 0.5577,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5793719746804021,
+      "grad_norm": 0.0750851696775246,
+      "learning_rate": 9.505692107468446e-06,
+      "loss": 0.5659,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5798684373836416,
+      "grad_norm": 0.07516933716182636,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.5069,
+      "step": 1168
+    },
+    {
+      "epoch": 0.580364900086881,
+      "grad_norm": 0.07516454409413244,
+      "learning_rate": 9.503995883062519e-06,
+      "loss": 0.5497,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5808613627901204,
+      "grad_norm": 0.074536429781554,
+      "learning_rate": 9.503146738249386e-06,
+      "loss": 0.5364,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5813578254933598,
+      "grad_norm": 0.07812964313209408,
+      "learning_rate": 9.502296905202479e-06,
+      "loss": 0.5796,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5818542881965992,
+      "grad_norm": 0.07219971092833606,
+      "learning_rate": 9.501446384051678e-06,
+      "loss": 0.5287,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5823507508998387,
+      "grad_norm": 0.07519688233518797,
+      "learning_rate": 9.500595174926972e-06,
+      "loss": 0.5642,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5828472136030781,
+      "grad_norm": 0.07012348080736677,
+      "learning_rate": 9.499743277958453e-06,
+      "loss": 0.4809,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5833436763063174,
+      "grad_norm": 0.06959138377666908,
+      "learning_rate": 9.498890693276326e-06,
+      "loss": 0.5087,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5838401390095569,
+      "grad_norm": 0.07524711459895272,
+      "learning_rate": 9.498037421010888e-06,
+      "loss": 0.539,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5843366017127963,
+      "grad_norm": 0.07190994525714839,
+      "learning_rate": 9.497183461292552e-06,
+      "loss": 0.5616,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5848330644160358,
+      "grad_norm": 0.07368602441399731,
+      "learning_rate": 9.49632881425183e-06,
+      "loss": 0.5239,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5853295271192752,
+      "grad_norm": 0.07519071489636418,
+      "learning_rate": 9.495473480019341e-06,
+      "loss": 0.5277,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5858259898225145,
+      "grad_norm": 0.07510671666338055,
+      "learning_rate": 9.494617458725812e-06,
+      "loss": 0.5218,
+      "step": 1180
+    },
+    {
+      "epoch": 0.586322452525754,
+      "grad_norm": 0.07479722093999988,
+      "learning_rate": 9.493760750502068e-06,
+      "loss": 0.5061,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5868189152289934,
+      "grad_norm": 0.07429647836890872,
+      "learning_rate": 9.492903355479047e-06,
+      "loss": 0.5201,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5873153779322329,
+      "grad_norm": 0.07243145250868639,
+      "learning_rate": 9.492045273787787e-06,
+      "loss": 0.5262,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5878118406354722,
+      "grad_norm": 0.07714760997242102,
+      "learning_rate": 9.49118650555943e-06,
+      "loss": 0.5683,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5883083033387116,
+      "grad_norm": 0.07303272362298142,
+      "learning_rate": 9.490327050925225e-06,
+      "loss": 0.5389,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5888047660419511,
+      "grad_norm": 0.07126612983172714,
+      "learning_rate": 9.48946691001653e-06,
+      "loss": 0.5098,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5893012287451905,
+      "grad_norm": 0.07285329202992465,
+      "learning_rate": 9.4886060829648e-06,
+      "loss": 0.5297,
+      "step": 1187
+    },
+    {
+      "epoch": 0.58979769144843,
+      "grad_norm": 0.07647879877688098,
+      "learning_rate": 9.487744569901598e-06,
+      "loss": 0.5249,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5902941541516693,
+      "grad_norm": 0.07188442696621349,
+      "learning_rate": 9.486882370958596e-06,
+      "loss": 0.5199,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5907906168549087,
+      "grad_norm": 0.07496456315600855,
+      "learning_rate": 9.486019486267563e-06,
+      "loss": 0.5416,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5912870795581482,
+      "grad_norm": 0.07576728726087202,
+      "learning_rate": 9.485155915960383e-06,
+      "loss": 0.5398,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5917835422613876,
+      "grad_norm": 0.07305444440941704,
+      "learning_rate": 9.484291660169031e-06,
+      "loss": 0.5462,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5922800049646271,
+      "grad_norm": 0.07182476967325477,
+      "learning_rate": 9.4834267190256e-06,
+      "loss": 0.5501,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5927764676678664,
+      "grad_norm": 0.07024451668694567,
+      "learning_rate": 9.482561092662284e-06,
+      "loss": 0.532,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5932729303711058,
+      "grad_norm": 0.07155857699067651,
+      "learning_rate": 9.481694781211375e-06,
+      "loss": 0.5501,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5937693930743453,
+      "grad_norm": 0.07197262136863748,
+      "learning_rate": 9.480827784805278e-06,
+      "loss": 0.5288,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5942658557775847,
+      "grad_norm": 0.07585902794239587,
+      "learning_rate": 9.4799601035765e-06,
+      "loss": 0.5626,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5947623184808242,
+      "grad_norm": 0.07401948158206198,
+      "learning_rate": 9.479091737657649e-06,
+      "loss": 0.5111,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5952587811840635,
+      "grad_norm": 0.07407351265240464,
+      "learning_rate": 9.478222687181444e-06,
+      "loss": 0.5287,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5957552438873029,
+      "grad_norm": 0.07262590404108443,
+      "learning_rate": 9.477352952280703e-06,
+      "loss": 0.5257,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5962517065905424,
+      "grad_norm": 0.07358832154964039,
+      "learning_rate": 9.476482533088351e-06,
+      "loss": 0.5292,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5967481692937818,
+      "grad_norm": 0.07178211125756767,
+      "learning_rate": 9.475611429737422e-06,
+      "loss": 0.5368,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5972446319970213,
+      "grad_norm": 0.07339341623133162,
+      "learning_rate": 9.474739642361043e-06,
+      "loss": 0.5549,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5977410947002606,
+      "grad_norm": 0.07412432663341953,
+      "learning_rate": 9.473867171092458e-06,
+      "loss": 0.5407,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5982375574035,
+      "grad_norm": 0.07436474908834359,
+      "learning_rate": 9.47299401606501e-06,
+      "loss": 0.5611,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5987340201067395,
+      "grad_norm": 0.07736756388873084,
+      "learning_rate": 9.472120177412147e-06,
+      "loss": 0.5572,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5992304828099789,
+      "grad_norm": 0.07576101680111423,
+      "learning_rate": 9.471245655267419e-06,
+      "loss": 0.5409,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5997269455132184,
+      "grad_norm": 0.07436447613810461,
+      "learning_rate": 9.470370449764481e-06,
+      "loss": 0.5538,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6002234082164577,
+      "grad_norm": 0.07528145343081451,
+      "learning_rate": 9.469494561037097e-06,
+      "loss": 0.5238,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6007198709196971,
+      "grad_norm": 0.06902241269601483,
+      "learning_rate": 9.468617989219136e-06,
+      "loss": 0.5147,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6012163336229366,
+      "grad_norm": 0.07674309299936645,
+      "learning_rate": 9.46774073444456e-06,
+      "loss": 0.5588,
+      "step": 1211
+    },
+    {
+      "epoch": 0.601712796326176,
+      "grad_norm": 0.07272454622505915,
+      "learning_rate": 9.46686279684745e-06,
+      "loss": 0.5379,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6022092590294155,
+      "grad_norm": 0.07281798474656986,
+      "learning_rate": 9.465984176561982e-06,
+      "loss": 0.5513,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6027057217326548,
+      "grad_norm": 0.07450970928418141,
+      "learning_rate": 9.46510487372244e-06,
+      "loss": 0.544,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6032021844358942,
+      "grad_norm": 0.08725930469560868,
+      "learning_rate": 9.464224888463208e-06,
+      "loss": 0.5453,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6036986471391337,
+      "grad_norm": 0.07862582216165505,
+      "learning_rate": 9.463344220918781e-06,
+      "loss": 0.5306,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6041951098423731,
+      "grad_norm": 0.07528725574920143,
+      "learning_rate": 9.462462871223755e-06,
+      "loss": 0.5449,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6046915725456126,
+      "grad_norm": 0.07316670669943198,
+      "learning_rate": 9.461580839512829e-06,
+      "loss": 0.5591,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6051880352488519,
+      "grad_norm": 0.07230039876916608,
+      "learning_rate": 9.46069812592081e-06,
+      "loss": 0.4814,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6056844979520913,
+      "grad_norm": 0.07364861679682433,
+      "learning_rate": 9.459814730582599e-06,
+      "loss": 0.5321,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6061809606553308,
+      "grad_norm": 0.07241383458550736,
+      "learning_rate": 9.458930653633218e-06,
+      "loss": 0.5114,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6066774233585702,
+      "grad_norm": 0.0731938940819185,
+      "learning_rate": 9.45804589520778e-06,
+      "loss": 0.577,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6071738860618096,
+      "grad_norm": 0.0723275474160098,
+      "learning_rate": 9.457160455441505e-06,
+      "loss": 0.5483,
+      "step": 1223
+    },
+    {
+      "epoch": 0.607670348765049,
+      "grad_norm": 0.07637854588101618,
+      "learning_rate": 9.45627433446972e-06,
+      "loss": 0.5549,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6081668114682884,
+      "grad_norm": 0.07879115938829902,
+      "learning_rate": 9.455387532427854e-06,
+      "loss": 0.6051,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6086632741715279,
+      "grad_norm": 0.07309959660536655,
+      "learning_rate": 9.45450004945144e-06,
+      "loss": 0.5086,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6091597368747673,
+      "grad_norm": 0.07694131558099677,
+      "learning_rate": 9.453611885676115e-06,
+      "loss": 0.5543,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6096561995780067,
+      "grad_norm": 0.0755960681922189,
+      "learning_rate": 9.452723041237624e-06,
+      "loss": 0.5145,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6101526622812461,
+      "grad_norm": 0.07295682109883318,
+      "learning_rate": 9.45183351627181e-06,
+      "loss": 0.5224,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6106491249844855,
+      "grad_norm": 0.07128517723475501,
+      "learning_rate": 9.45094331091462e-06,
+      "loss": 0.5116,
+      "step": 1230
+    },
+    {
+      "epoch": 0.611145587687725,
+      "grad_norm": 0.08115875461185187,
+      "learning_rate": 9.450052425302112e-06,
+      "loss": 0.513,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6116420503909644,
+      "grad_norm": 0.07377422236820412,
+      "learning_rate": 9.44916085957044e-06,
+      "loss": 0.5435,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6121385130942038,
+      "grad_norm": 0.07775316082126749,
+      "learning_rate": 9.448268613855871e-06,
+      "loss": 0.5434,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6126349757974432,
+      "grad_norm": 0.0731516601538171,
+      "learning_rate": 9.447375688294765e-06,
+      "loss": 0.5603,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6131314385006826,
+      "grad_norm": 0.07441731588595143,
+      "learning_rate": 9.446482083023594e-06,
+      "loss": 0.5506,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6136279012039221,
+      "grad_norm": 0.07198200814304813,
+      "learning_rate": 9.44558779817893e-06,
+      "loss": 0.5357,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6141243639071615,
+      "grad_norm": 0.07315295770611031,
+      "learning_rate": 9.444692833897451e-06,
+      "loss": 0.5532,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6146208266104008,
+      "grad_norm": 0.07174351007581338,
+      "learning_rate": 9.443797190315938e-06,
+      "loss": 0.5331,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6151172893136403,
+      "grad_norm": 0.06979723079086682,
+      "learning_rate": 9.442900867571274e-06,
+      "loss": 0.5135,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6156137520168797,
+      "grad_norm": 0.07178731116164377,
+      "learning_rate": 9.442003865800448e-06,
+      "loss": 0.4953,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6161102147201192,
+      "grad_norm": 0.07054593892362157,
+      "learning_rate": 9.441106185140557e-06,
+      "loss": 0.5395,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6166066774233586,
+      "grad_norm": 0.0736531086968794,
+      "learning_rate": 9.44020782572879e-06,
+      "loss": 0.5721,
+      "step": 1242
+    },
+    {
+      "epoch": 0.617103140126598,
+      "grad_norm": 0.07586757881170157,
+      "learning_rate": 9.43930878770245e-06,
+      "loss": 0.5557,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6175996028298374,
+      "grad_norm": 0.07191430014087756,
+      "learning_rate": 9.438409071198944e-06,
+      "loss": 0.5094,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6180960655330768,
+      "grad_norm": 0.07452596818342855,
+      "learning_rate": 9.437508676355774e-06,
+      "loss": 0.5543,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6185925282363163,
+      "grad_norm": 0.08010039702376043,
+      "learning_rate": 9.436607603310553e-06,
+      "loss": 0.5622,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6190889909395557,
+      "grad_norm": 0.07695327823931182,
+      "learning_rate": 9.435705852200994e-06,
+      "loss": 0.5231,
+      "step": 1247
+    },
+    {
+      "epoch": 0.619585453642795,
+      "grad_norm": 0.07276200871027631,
+      "learning_rate": 9.434803423164917e-06,
+      "loss": 0.518,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6200819163460345,
+      "grad_norm": 0.07104138436967752,
+      "learning_rate": 9.433900316340246e-06,
+      "loss": 0.4994,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6205783790492739,
+      "grad_norm": 0.07404188758170115,
+      "learning_rate": 9.432996531865001e-06,
+      "loss": 0.5293,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6210748417525134,
+      "grad_norm": 0.0716565229661667,
+      "learning_rate": 9.432092069877315e-06,
+      "loss": 0.5112,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6215713044557528,
+      "grad_norm": 0.08010756662952559,
+      "learning_rate": 9.431186930515419e-06,
+      "loss": 0.5259,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6220677671589921,
+      "grad_norm": 0.07413909114414897,
+      "learning_rate": 9.430281113917649e-06,
+      "loss": 0.5255,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6225642298622316,
+      "grad_norm": 0.07327540089614044,
+      "learning_rate": 9.429374620222448e-06,
+      "loss": 0.53,
+      "step": 1254
+    },
+    {
+      "epoch": 0.623060692565471,
+      "grad_norm": 0.07091853000727939,
+      "learning_rate": 9.428467449568352e-06,
+      "loss": 0.5135,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6235571552687105,
+      "grad_norm": 0.07410900989671004,
+      "learning_rate": 9.427559602094011e-06,
+      "loss": 0.5276,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6240536179719498,
+      "grad_norm": 0.07515041114499699,
+      "learning_rate": 9.426651077938178e-06,
+      "loss": 0.528,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6245500806751892,
+      "grad_norm": 0.07444010187119487,
+      "learning_rate": 9.425741877239703e-06,
+      "loss": 0.5479,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6250465433784287,
+      "grad_norm": 0.07370774747030723,
+      "learning_rate": 9.424832000137542e-06,
+      "loss": 0.5541,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6255430060816681,
+      "grad_norm": 0.0714035032190127,
+      "learning_rate": 9.423921446770759e-06,
+      "loss": 0.4864,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6260394687849076,
+      "grad_norm": 0.07730674579399738,
+      "learning_rate": 9.423010217278515e-06,
+      "loss": 0.5534,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6265359314881469,
+      "grad_norm": 0.07138774544095353,
+      "learning_rate": 9.422098311800074e-06,
+      "loss": 0.5399,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6270323941913863,
+      "grad_norm": 0.06852941610757017,
+      "learning_rate": 9.421185730474811e-06,
+      "loss": 0.519,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6275288568946258,
+      "grad_norm": 0.07367955240222782,
+      "learning_rate": 9.420272473442198e-06,
+      "loss": 0.5456,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6280253195978652,
+      "grad_norm": 0.07541228844186217,
+      "learning_rate": 9.41935854084181e-06,
+      "loss": 0.5314,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6285217823011047,
+      "grad_norm": 0.07435558350951817,
+      "learning_rate": 9.418443932813328e-06,
+      "loss": 0.5269,
+      "step": 1266
+    },
+    {
+      "epoch": 0.629018245004344,
+      "grad_norm": 0.0728163151724603,
+      "learning_rate": 9.417528649496535e-06,
+      "loss": 0.547,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6295147077075834,
+      "grad_norm": 0.07338689354836506,
+      "learning_rate": 9.41661269103132e-06,
+      "loss": 0.5555,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6300111704108229,
+      "grad_norm": 0.07076974627201735,
+      "learning_rate": 9.415696057557667e-06,
+      "loss": 0.5584,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6305076331140623,
+      "grad_norm": 0.07256759974301101,
+      "learning_rate": 9.414778749215673e-06,
+      "loss": 0.5451,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6310040958173018,
+      "grad_norm": 0.07319635272423798,
+      "learning_rate": 9.413860766145533e-06,
+      "loss": 0.5177,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6315005585205411,
+      "grad_norm": 0.07689312407289102,
+      "learning_rate": 9.412942108487545e-06,
+      "loss": 0.5209,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6319970212237805,
+      "grad_norm": 0.0734130755794059,
+      "learning_rate": 9.412022776382113e-06,
+      "loss": 0.5675,
+      "step": 1273
+    },
+    {
+      "epoch": 0.63249348392702,
+      "grad_norm": 0.07228991038198292,
+      "learning_rate": 9.411102769969742e-06,
+      "loss": 0.5065,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6329899466302594,
+      "grad_norm": 0.07467310042619545,
+      "learning_rate": 9.410182089391039e-06,
+      "loss": 0.5473,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6334864093334989,
+      "grad_norm": 0.0730424746593062,
+      "learning_rate": 9.409260734786713e-06,
+      "loss": 0.5342,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6339828720367382,
+      "grad_norm": 0.07040514946923192,
+      "learning_rate": 9.408338706297581e-06,
+      "loss": 0.5214,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6344793347399776,
+      "grad_norm": 0.07277326050517614,
+      "learning_rate": 9.407416004064562e-06,
+      "loss": 0.5741,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6349757974432171,
+      "grad_norm": 0.07219826740008305,
+      "learning_rate": 9.406492628228674e-06,
+      "loss": 0.5273,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6354722601464565,
+      "grad_norm": 0.06850912173296844,
+      "learning_rate": 9.405568578931042e-06,
+      "loss": 0.5055,
+      "step": 1280
+    },
+    {
+      "epoch": 0.635968722849696,
+      "grad_norm": 0.07229403813315836,
+      "learning_rate": 9.404643856312887e-06,
+      "loss": 0.5369,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6364651855529353,
+      "grad_norm": 0.06994299050471268,
+      "learning_rate": 9.403718460515544e-06,
+      "loss": 0.5254,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6369616482561747,
+      "grad_norm": 0.06823196465634472,
+      "learning_rate": 9.402792391680443e-06,
+      "loss": 0.4799,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6374581109594142,
+      "grad_norm": 0.07588187341041024,
+      "learning_rate": 9.401865649949116e-06,
+      "loss": 0.5718,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6379545736626536,
+      "grad_norm": 0.07594938287429129,
+      "learning_rate": 9.400938235463203e-06,
+      "loss": 0.526,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6384510363658931,
+      "grad_norm": 0.07332308146260103,
+      "learning_rate": 9.400010148364447e-06,
+      "loss": 0.5429,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6389474990691324,
+      "grad_norm": 0.07072214782920225,
+      "learning_rate": 9.399081388794688e-06,
+      "loss": 0.5202,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6394439617723718,
+      "grad_norm": 0.0746382662046732,
+      "learning_rate": 9.398151956895872e-06,
+      "loss": 0.5072,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6399404244756113,
+      "grad_norm": 0.0735533069677649,
+      "learning_rate": 9.397221852810049e-06,
+      "loss": 0.5329,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6404368871788507,
+      "grad_norm": 0.07615515782140855,
+      "learning_rate": 9.396291076679369e-06,
+      "loss": 0.5768,
+      "step": 1290
+    },
+    {
+      "epoch": 0.64093334988209,
+      "grad_norm": 0.07001774866514658,
+      "learning_rate": 9.395359628646087e-06,
+      "loss": 0.5257,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6414298125853295,
+      "grad_norm": 0.07375326519499116,
+      "learning_rate": 9.39442750885256e-06,
+      "loss": 0.5092,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6419262752885689,
+      "grad_norm": 0.07259307932350538,
+      "learning_rate": 9.393494717441248e-06,
+      "loss": 0.5599,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6424227379918084,
+      "grad_norm": 0.07088258695341772,
+      "learning_rate": 9.392561254554712e-06,
+      "loss": 0.4888,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6429192006950478,
+      "grad_norm": 0.07264741884460572,
+      "learning_rate": 9.391627120335618e-06,
+      "loss": 0.5168,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6434156633982872,
+      "grad_norm": 0.07619595372888417,
+      "learning_rate": 9.390692314926734e-06,
+      "loss": 0.5409,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6439121261015266,
+      "grad_norm": 0.07051164086514715,
+      "learning_rate": 9.389756838470929e-06,
+      "loss": 0.5437,
+      "step": 1297
+    },
+    {
+      "epoch": 0.644408588804766,
+      "grad_norm": 0.07311117048289703,
+      "learning_rate": 9.388820691111175e-06,
+      "loss": 0.4895,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6449050515080055,
+      "grad_norm": 0.07300862751347871,
+      "learning_rate": 9.387883872990547e-06,
+      "loss": 0.5208,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6454015142112449,
+      "grad_norm": 0.07600173346135768,
+      "learning_rate": 9.386946384252225e-06,
+      "loss": 0.5377,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6458979769144843,
+      "grad_norm": 0.06913090974935486,
+      "learning_rate": 9.386008225039486e-06,
+      "loss": 0.549,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6463944396177237,
+      "grad_norm": 0.07293082318748247,
+      "learning_rate": 9.385069395495715e-06,
+      "loss": 0.5494,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6468909023209631,
+      "grad_norm": 0.07192021828481457,
+      "learning_rate": 9.384129895764396e-06,
+      "loss": 0.5539,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6473873650242026,
+      "grad_norm": 0.0677715996313107,
+      "learning_rate": 9.383189725989117e-06,
+      "loss": 0.4885,
+      "step": 1304
+    },
+    {
+      "epoch": 0.647883827727442,
+      "grad_norm": 0.06993925668460728,
+      "learning_rate": 9.382248886313568e-06,
+      "loss": 0.5666,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6483802904306813,
+      "grad_norm": 0.0732675510502527,
+      "learning_rate": 9.381307376881538e-06,
+      "loss": 0.5244,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6488767531339208,
+      "grad_norm": 0.07833454832413211,
+      "learning_rate": 9.380365197836927e-06,
+      "loss": 0.5291,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6493732158371602,
+      "grad_norm": 0.07246897473704703,
+      "learning_rate": 9.379422349323728e-06,
+      "loss": 0.5336,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6498696785403997,
+      "grad_norm": 0.08306837527169122,
+      "learning_rate": 9.378478831486042e-06,
+      "loss": 0.5433,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6503661412436391,
+      "grad_norm": 0.07296698699268978,
+      "learning_rate": 9.37753464446807e-06,
+      "loss": 0.5181,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6508626039468784,
+      "grad_norm": 0.07287062275676794,
+      "learning_rate": 9.376589788414116e-06,
+      "loss": 0.5266,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6513590666501179,
+      "grad_norm": 0.07284381245821252,
+      "learning_rate": 9.375644263468586e-06,
+      "loss": 0.547,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6518555293533573,
+      "grad_norm": 0.07656099448383531,
+      "learning_rate": 9.374698069775989e-06,
+      "loss": 0.5341,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6523519920565968,
+      "grad_norm": 0.0708265851795634,
+      "learning_rate": 9.373751207480935e-06,
+      "loss": 0.5366,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6528484547598362,
+      "grad_norm": 0.07593480603910471,
+      "learning_rate": 9.372803676728138e-06,
+      "loss": 0.5393,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6533449174630755,
+      "grad_norm": 0.07672621233019099,
+      "learning_rate": 9.371855477662409e-06,
+      "loss": 0.5437,
+      "step": 1316
+    },
+    {
+      "epoch": 0.653841380166315,
+      "grad_norm": 0.07057674890321015,
+      "learning_rate": 9.37090661042867e-06,
+      "loss": 0.5431,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6543378428695544,
+      "grad_norm": 0.07035606365375213,
+      "learning_rate": 9.369957075171935e-06,
+      "loss": 0.5158,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6548343055727939,
+      "grad_norm": 0.07424226230822006,
+      "learning_rate": 9.369006872037329e-06,
+      "loss": 0.533,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6553307682760333,
+      "grad_norm": 0.07381144539992311,
+      "learning_rate": 9.368056001170077e-06,
+      "loss": 0.5094,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6558272309792726,
+      "grad_norm": 0.06949672074730547,
+      "learning_rate": 9.367104462715498e-06,
+      "loss": 0.5161,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6563236936825121,
+      "grad_norm": 0.07454285891198276,
+      "learning_rate": 9.366152256819025e-06,
+      "loss": 0.5661,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6568201563857515,
+      "grad_norm": 0.0710626384867322,
+      "learning_rate": 9.365199383626184e-06,
+      "loss": 0.5173,
+      "step": 1323
+    },
+    {
+      "epoch": 0.657316619088991,
+      "grad_norm": 0.07589984877939893,
+      "learning_rate": 9.36424584328261e-06,
+      "loss": 0.5494,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6578130817922304,
+      "grad_norm": 0.07614774842121315,
+      "learning_rate": 9.363291635934033e-06,
+      "loss": 0.5115,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6583095444954697,
+      "grad_norm": 0.07569403053633093,
+      "learning_rate": 9.36233676172629e-06,
+      "loss": 0.5351,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6588060071987092,
+      "grad_norm": 0.07563821512274932,
+      "learning_rate": 9.361381220805317e-06,
+      "loss": 0.5736,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6593024699019486,
+      "grad_norm": 0.07013748950442422,
+      "learning_rate": 9.360425013317153e-06,
+      "loss": 0.4882,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6597989326051881,
+      "grad_norm": 0.07779315687414581,
+      "learning_rate": 9.359468139407942e-06,
+      "loss": 0.5296,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6602953953084274,
+      "grad_norm": 0.07400049233265299,
+      "learning_rate": 9.358510599223922e-06,
+      "loss": 0.5161,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6607918580116668,
+      "grad_norm": 0.06767323424859467,
+      "learning_rate": 9.357552392911444e-06,
+      "loss": 0.5141,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6612883207149063,
+      "grad_norm": 0.07102033958900007,
+      "learning_rate": 9.356593520616948e-06,
+      "loss": 0.5587,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6617847834181457,
+      "grad_norm": 0.07588371408730442,
+      "learning_rate": 9.355633982486986e-06,
+      "loss": 0.5885,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6622812461213852,
+      "grad_norm": 0.07715392657364961,
+      "learning_rate": 9.354673778668206e-06,
+      "loss": 0.5403,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6627777088246245,
+      "grad_norm": 0.07455958529451663,
+      "learning_rate": 9.353712909307361e-06,
+      "loss": 0.5134,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6632741715278639,
+      "grad_norm": 0.07191533309938422,
+      "learning_rate": 9.352751374551305e-06,
+      "loss": 0.5087,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6637706342311034,
+      "grad_norm": 0.07186294289718925,
+      "learning_rate": 9.351789174546993e-06,
+      "loss": 0.539,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6642670969343428,
+      "grad_norm": 0.0754358841299627,
+      "learning_rate": 9.350826309441481e-06,
+      "loss": 0.4992,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6647635596375823,
+      "grad_norm": 0.07627703380934699,
+      "learning_rate": 9.349862779381926e-06,
+      "loss": 0.5061,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6652600223408216,
+      "grad_norm": 0.0734366299613896,
+      "learning_rate": 9.348898584515593e-06,
+      "loss": 0.5316,
+      "step": 1340
+    },
+    {
+      "epoch": 0.665756485044061,
+      "grad_norm": 0.07674901990204858,
+      "learning_rate": 9.347933724989839e-06,
+      "loss": 0.5538,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6662529477473005,
+      "grad_norm": 0.07020692260151944,
+      "learning_rate": 9.346968200952129e-06,
+      "loss": 0.5035,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6667494104505399,
+      "grad_norm": 0.07093936922660374,
+      "learning_rate": 9.346002012550027e-06,
+      "loss": 0.5251,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6672458731537794,
+      "grad_norm": 0.07673979882696808,
+      "learning_rate": 9.345035159931201e-06,
+      "loss": 0.5519,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6677423358570187,
+      "grad_norm": 0.06821342541699567,
+      "learning_rate": 9.344067643243419e-06,
+      "loss": 0.5204,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6682387985602581,
+      "grad_norm": 0.07188197485984522,
+      "learning_rate": 9.343099462634548e-06,
+      "loss": 0.5607,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6687352612634976,
+      "grad_norm": 0.07745552205514158,
+      "learning_rate": 9.34213061825256e-06,
+      "loss": 0.5732,
+      "step": 1347
+    },
+    {
+      "epoch": 0.669231723966737,
+      "grad_norm": 0.07068348754676128,
+      "learning_rate": 9.34116111024553e-06,
+      "loss": 0.4946,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6697281866699765,
+      "grad_norm": 0.07112019014542299,
+      "learning_rate": 9.340190938761628e-06,
+      "loss": 0.5266,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6702246493732158,
+      "grad_norm": 0.07273078228929754,
+      "learning_rate": 9.339220103949132e-06,
+      "loss": 0.5487,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6707211120764552,
+      "grad_norm": 0.07688997752683249,
+      "learning_rate": 9.338248605956416e-06,
+      "loss": 0.5285,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6712175747796947,
+      "grad_norm": 0.06971510713006104,
+      "learning_rate": 9.337276444931959e-06,
+      "loss": 0.4981,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6717140374829341,
+      "grad_norm": 0.0722029002376623,
+      "learning_rate": 9.33630362102434e-06,
+      "loss": 0.5059,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6722105001861736,
+      "grad_norm": 0.08013613409130184,
+      "learning_rate": 9.335330134382242e-06,
+      "loss": 0.5564,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6727069628894129,
+      "grad_norm": 0.07975096625901282,
+      "learning_rate": 9.334355985154444e-06,
+      "loss": 0.5727,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6732034255926523,
+      "grad_norm": 0.07276394010589855,
+      "learning_rate": 9.333381173489828e-06,
+      "loss": 0.5428,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6736998882958918,
+      "grad_norm": 0.07450651270520357,
+      "learning_rate": 9.332405699537382e-06,
+      "loss": 0.5589,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6741963509991312,
+      "grad_norm": 0.07503406849766957,
+      "learning_rate": 9.331429563446189e-06,
+      "loss": 0.5258,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6746928137023707,
+      "grad_norm": 0.07217189294540513,
+      "learning_rate": 9.330452765365436e-06,
+      "loss": 0.5145,
+      "step": 1359
+    },
+    {
+      "epoch": 0.67518927640561,
+      "grad_norm": 0.07214669475300672,
+      "learning_rate": 9.32947530544441e-06,
+      "loss": 0.5306,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6756857391088494,
+      "grad_norm": 0.07055049962573819,
+      "learning_rate": 9.328497183832505e-06,
+      "loss": 0.5267,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6761822018120889,
+      "grad_norm": 0.07106910357432814,
+      "learning_rate": 9.327518400679206e-06,
+      "loss": 0.52,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6766786645153283,
+      "grad_norm": 0.07301694174327626,
+      "learning_rate": 9.326538956134106e-06,
+      "loss": 0.5302,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6771751272185677,
+      "grad_norm": 0.07032215298745946,
+      "learning_rate": 9.325558850346897e-06,
+      "loss": 0.5197,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6776715899218071,
+      "grad_norm": 0.06935838081944155,
+      "learning_rate": 9.324578083467372e-06,
+      "loss": 0.5363,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6781680526250465,
+      "grad_norm": 0.07423776754352923,
+      "learning_rate": 9.323596655645427e-06,
+      "loss": 0.5485,
+      "step": 1366
+    },
+    {
+      "epoch": 0.678664515328286,
+      "grad_norm": 0.07315712610205513,
+      "learning_rate": 9.322614567031056e-06,
+      "loss": 0.5104,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6791609780315254,
+      "grad_norm": 0.07445358870399109,
+      "learning_rate": 9.321631817774358e-06,
+      "loss": 0.5472,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6796574407347648,
+      "grad_norm": 0.06898749856725994,
+      "learning_rate": 9.320648408025528e-06,
+      "loss": 0.4936,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6801539034380042,
+      "grad_norm": 0.07381816406555464,
+      "learning_rate": 9.319664337934865e-06,
+      "loss": 0.5203,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6806503661412436,
+      "grad_norm": 0.07735751452122537,
+      "learning_rate": 9.318679607652768e-06,
+      "loss": 0.5364,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6811468288444831,
+      "grad_norm": 0.07326646920478877,
+      "learning_rate": 9.317694217329737e-06,
+      "loss": 0.5142,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6816432915477225,
+      "grad_norm": 0.073764703258718,
+      "learning_rate": 9.316708167116377e-06,
+      "loss": 0.5554,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6821397542509618,
+      "grad_norm": 0.07390414723701343,
+      "learning_rate": 9.315721457163384e-06,
+      "loss": 0.567,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6826362169542013,
+      "grad_norm": 0.07393942919613289,
+      "learning_rate": 9.314734087621566e-06,
+      "loss": 0.5123,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6831326796574407,
+      "grad_norm": 0.07620045589780511,
+      "learning_rate": 9.313746058641822e-06,
+      "loss": 0.5372,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6836291423606802,
+      "grad_norm": 0.07583882363513543,
+      "learning_rate": 9.312757370375159e-06,
+      "loss": 0.5732,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6841256050639196,
+      "grad_norm": 0.07726479654671939,
+      "learning_rate": 9.311768022972682e-06,
+      "loss": 0.5422,
+      "step": 1378
+    },
+    {
+      "epoch": 0.684622067767159,
+      "grad_norm": 0.0760180487581988,
+      "learning_rate": 9.310778016585597e-06,
+      "loss": 0.558,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6851185304703984,
+      "grad_norm": 0.0691920787710429,
+      "learning_rate": 9.30978735136521e-06,
+      "loss": 0.4954,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6856149931736378,
+      "grad_norm": 0.07452724187979319,
+      "learning_rate": 9.308796027462928e-06,
+      "loss": 0.5316,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6861114558768773,
+      "grad_norm": 0.0756477838695681,
+      "learning_rate": 9.30780404503026e-06,
+      "loss": 0.5423,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6866079185801167,
+      "grad_norm": 0.07516990810148905,
+      "learning_rate": 9.306811404218814e-06,
+      "loss": 0.5624,
+      "step": 1383
+    },
+    {
+      "epoch": 0.687104381283356,
+      "grad_norm": 0.07346320934704692,
+      "learning_rate": 9.3058181051803e-06,
+      "loss": 0.5434,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6876008439865955,
+      "grad_norm": 0.07279939873545586,
+      "learning_rate": 9.304824148066526e-06,
+      "loss": 0.5349,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6880973066898349,
+      "grad_norm": 0.08384339788396475,
+      "learning_rate": 9.303829533029406e-06,
+      "loss": 0.5348,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6885937693930744,
+      "grad_norm": 0.07095408782774806,
+      "learning_rate": 9.302834260220945e-06,
+      "loss": 0.4933,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6890902320963138,
+      "grad_norm": 0.07549768019197384,
+      "learning_rate": 9.30183832979326e-06,
+      "loss": 0.5506,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6895866947995531,
+      "grad_norm": 0.0766368547269983,
+      "learning_rate": 9.30084174189856e-06,
+      "loss": 0.5791,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6900831575027926,
+      "grad_norm": 0.07073419123255918,
+      "learning_rate": 9.29984449668916e-06,
+      "loss": 0.4825,
+      "step": 1390
+    },
+    {
+      "epoch": 0.690579620206032,
+      "grad_norm": 0.07395858346898046,
+      "learning_rate": 9.298846594317471e-06,
+      "loss": 0.5832,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6910760829092715,
+      "grad_norm": 0.06889946396344103,
+      "learning_rate": 9.297848034936007e-06,
+      "loss": 0.5146,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6915725456125109,
+      "grad_norm": 0.07150224954000071,
+      "learning_rate": 9.296848818697381e-06,
+      "loss": 0.5077,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6920690083157502,
+      "grad_norm": 0.07339810430184905,
+      "learning_rate": 9.295848945754308e-06,
+      "loss": 0.5329,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6925654710189897,
+      "grad_norm": 0.06917499027819218,
+      "learning_rate": 9.294848416259603e-06,
+      "loss": 0.515,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6930619337222291,
+      "grad_norm": 0.07238605456545845,
+      "learning_rate": 9.293847230366178e-06,
+      "loss": 0.553,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6935583964254686,
+      "grad_norm": 0.07049884661873568,
+      "learning_rate": 9.292845388227052e-06,
+      "loss": 0.5143,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6940548591287079,
+      "grad_norm": 0.07546666489246001,
+      "learning_rate": 9.291842889995339e-06,
+      "loss": 0.5639,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6945513218319473,
+      "grad_norm": 0.0751436423514526,
+      "learning_rate": 9.290839735824254e-06,
+      "loss": 0.5102,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6950477845351868,
+      "grad_norm": 0.07389138982609979,
+      "learning_rate": 9.289835925867116e-06,
+      "loss": 0.528,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6955442472384262,
+      "grad_norm": 0.06625971976683562,
+      "learning_rate": 9.288831460277337e-06,
+      "loss": 0.5063,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6960407099416657,
+      "grad_norm": 0.07734680372898531,
+      "learning_rate": 9.287826339208436e-06,
+      "loss": 0.5726,
+      "step": 1402
+    },
+    {
+      "epoch": 0.696537172644905,
+      "grad_norm": 0.06951228908969029,
+      "learning_rate": 9.286820562814029e-06,
+      "loss": 0.5191,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6970336353481444,
+      "grad_norm": 0.07287713737550759,
+      "learning_rate": 9.285814131247831e-06,
+      "loss": 0.5476,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6975300980513839,
+      "grad_norm": 0.06956108091891039,
+      "learning_rate": 9.284807044663663e-06,
+      "loss": 0.5176,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6980265607546233,
+      "grad_norm": 0.08380941217622373,
+      "learning_rate": 9.283799303215442e-06,
+      "loss": 0.5468,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6985230234578628,
+      "grad_norm": 0.07138672000727564,
+      "learning_rate": 9.28279090705718e-06,
+      "loss": 0.4963,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6990194861611021,
+      "grad_norm": 0.07292069605598205,
+      "learning_rate": 9.281781856342998e-06,
+      "loss": 0.542,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6995159488643415,
+      "grad_norm": 0.07332813612145221,
+      "learning_rate": 9.280772151227112e-06,
+      "loss": 0.5548,
+      "step": 1409
+    },
+    {
+      "epoch": 0.700012411567581,
+      "grad_norm": 0.07246818987746459,
+      "learning_rate": 9.279761791863839e-06,
+      "loss": 0.5503,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7005088742708204,
+      "grad_norm": 0.07005478459710343,
+      "learning_rate": 9.2787507784076e-06,
+      "loss": 0.5004,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7010053369740599,
+      "grad_norm": 0.06934183608207299,
+      "learning_rate": 9.277739111012905e-06,
+      "loss": 0.5528,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7015017996772992,
+      "grad_norm": 0.07052000189461478,
+      "learning_rate": 9.276726789834378e-06,
+      "loss": 0.5087,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7019982623805386,
+      "grad_norm": 0.07301837264825195,
+      "learning_rate": 9.275713815026732e-06,
+      "loss": 0.53,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7024947250837781,
+      "grad_norm": 0.0722789778678723,
+      "learning_rate": 9.274700186744786e-06,
+      "loss": 0.5401,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7029911877870175,
+      "grad_norm": 0.07527888887630624,
+      "learning_rate": 9.273685905143454e-06,
+      "loss": 0.5287,
+      "step": 1416
+    },
+    {
+      "epoch": 0.703487650490257,
+      "grad_norm": 0.07285186180656061,
+      "learning_rate": 9.272670970377758e-06,
+      "loss": 0.5305,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7039841131934963,
+      "grad_norm": 0.07059684009978115,
+      "learning_rate": 9.271655382602809e-06,
+      "loss": 0.5263,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7044805758967357,
+      "grad_norm": 0.07370364807475144,
+      "learning_rate": 9.270639141973826e-06,
+      "loss": 0.536,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7049770385999752,
+      "grad_norm": 0.07584063600861424,
+      "learning_rate": 9.269622248646124e-06,
+      "loss": 0.5108,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7054735013032146,
+      "grad_norm": 0.11190360389061123,
+      "learning_rate": 9.26860470277512e-06,
+      "loss": 0.5217,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7059699640064541,
+      "grad_norm": 0.07173583317483305,
+      "learning_rate": 9.267586504516331e-06,
+      "loss": 0.5281,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7064664267096934,
+      "grad_norm": 0.07299583837400489,
+      "learning_rate": 9.266567654025369e-06,
+      "loss": 0.5074,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7069628894129328,
+      "grad_norm": 0.07453659993742369,
+      "learning_rate": 9.265548151457949e-06,
+      "loss": 0.5803,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7074593521161723,
+      "grad_norm": 0.07321720883021786,
+      "learning_rate": 9.264527996969888e-06,
+      "loss": 0.5109,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7079558148194117,
+      "grad_norm": 0.07222177910537633,
+      "learning_rate": 9.2635071907171e-06,
+      "loss": 0.5059,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7084522775226512,
+      "grad_norm": 0.07247620140102892,
+      "learning_rate": 9.262485732855597e-06,
+      "loss": 0.5467,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7089487402258905,
+      "grad_norm": 0.07077818847784131,
+      "learning_rate": 9.261463623541493e-06,
+      "loss": 0.5223,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7094452029291299,
+      "grad_norm": 0.07067550789715772,
+      "learning_rate": 9.260440862931002e-06,
+      "loss": 0.4934,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7099416656323694,
+      "grad_norm": 0.07365823742825753,
+      "learning_rate": 9.259417451180437e-06,
+      "loss": 0.5329,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7104381283356088,
+      "grad_norm": 0.07127326575777367,
+      "learning_rate": 9.258393388446208e-06,
+      "loss": 0.5535,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7109345910388482,
+      "grad_norm": 0.07488318035883733,
+      "learning_rate": 9.257368674884829e-06,
+      "loss": 0.5271,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7114310537420876,
+      "grad_norm": 0.07724788510343007,
+      "learning_rate": 9.256343310652907e-06,
+      "loss": 0.5321,
+      "step": 1433
+    },
+    {
+      "epoch": 0.711927516445327,
+      "grad_norm": 0.07923242956914898,
+      "learning_rate": 9.255317295907158e-06,
+      "loss": 0.547,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7124239791485665,
+      "grad_norm": 0.07674530249324395,
+      "learning_rate": 9.254290630804387e-06,
+      "loss": 0.5312,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7129204418518059,
+      "grad_norm": 0.07180989622822076,
+      "learning_rate": 9.253263315501508e-06,
+      "loss": 0.5144,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7134169045550453,
+      "grad_norm": 0.09719368557984771,
+      "learning_rate": 9.252235350155524e-06,
+      "loss": 0.5375,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7139133672582847,
+      "grad_norm": 0.07200959908689893,
+      "learning_rate": 9.25120673492355e-06,
+      "loss": 0.5303,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7144098299615241,
+      "grad_norm": 0.07490489574435455,
+      "learning_rate": 9.250177469962787e-06,
+      "loss": 0.5515,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7149062926647636,
+      "grad_norm": 0.07704389871771329,
+      "learning_rate": 9.249147555430545e-06,
+      "loss": 0.5139,
+      "step": 1440
+    },
+    {
+      "epoch": 0.715402755368003,
+      "grad_norm": 0.07348214641843492,
+      "learning_rate": 9.24811699148423e-06,
+      "loss": 0.5126,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7158992180712423,
+      "grad_norm": 0.07586501625655766,
+      "learning_rate": 9.247085778281342e-06,
+      "loss": 0.55,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7163956807744818,
+      "grad_norm": 0.07483427178766174,
+      "learning_rate": 9.246053915979492e-06,
+      "loss": 0.5241,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7168921434777212,
+      "grad_norm": 0.07846284947721777,
+      "learning_rate": 9.245021404736382e-06,
+      "loss": 0.5398,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7173886061809607,
+      "grad_norm": 0.06935259803429858,
+      "learning_rate": 9.243988244709815e-06,
+      "loss": 0.5141,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7178850688842001,
+      "grad_norm": 0.07133423082003033,
+      "learning_rate": 9.24295443605769e-06,
+      "loss": 0.5239,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7183815315874394,
+      "grad_norm": 0.0789951285815156,
+      "learning_rate": 9.24191997893801e-06,
+      "loss": 0.5489,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7188779942906789,
+      "grad_norm": 0.07789738837985204,
+      "learning_rate": 9.240884873508876e-06,
+      "loss": 0.5433,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7193744569939183,
+      "grad_norm": 0.07407387509806816,
+      "learning_rate": 9.239849119928486e-06,
+      "loss": 0.515,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7198709196971578,
+      "grad_norm": 0.07376724827013004,
+      "learning_rate": 9.23881271835514e-06,
+      "loss": 0.5465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7203673824003972,
+      "grad_norm": 0.07614036890621281,
+      "learning_rate": 9.237775668947233e-06,
+      "loss": 0.5728,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7208638451036365,
+      "grad_norm": 0.07088363970569353,
+      "learning_rate": 9.236737971863263e-06,
+      "loss": 0.506,
+      "step": 1452
+    },
+    {
+      "epoch": 0.721360307806876,
+      "grad_norm": 0.0731833784182166,
+      "learning_rate": 9.235699627261825e-06,
+      "loss": 0.485,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7218567705101154,
+      "grad_norm": 0.07542331219080749,
+      "learning_rate": 9.234660635301613e-06,
+      "loss": 0.5164,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7223532332133549,
+      "grad_norm": 0.06950421429090972,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 0.509,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7228496959165943,
+      "grad_norm": 0.06914956109999934,
+      "learning_rate": 9.23258070994014e-06,
+      "loss": 0.5121,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7233461586198336,
+      "grad_norm": 0.07448552133247559,
+      "learning_rate": 9.231539776856764e-06,
+      "loss": 0.5373,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7238426213230731,
+      "grad_norm": 0.07433314213228254,
+      "learning_rate": 9.230498197050377e-06,
+      "loss": 0.5063,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7243390840263125,
+      "grad_norm": 0.07175828404165463,
+      "learning_rate": 9.229455970680175e-06,
+      "loss": 0.4841,
+      "step": 1459
+    },
+    {
+      "epoch": 0.724835546729552,
+      "grad_norm": 0.07405100866158164,
+      "learning_rate": 9.22841309790544e-06,
+      "loss": 0.543,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7253320094327914,
+      "grad_norm": 0.07068543317117736,
+      "learning_rate": 9.227369578885561e-06,
+      "loss": 0.5044,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7258284721360307,
+      "grad_norm": 0.07369786122206998,
+      "learning_rate": 9.226325413780021e-06,
+      "loss": 0.5316,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7263249348392702,
+      "grad_norm": 0.07139828290659295,
+      "learning_rate": 9.225280602748408e-06,
+      "loss": 0.5262,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7268213975425096,
+      "grad_norm": 0.06840457873936871,
+      "learning_rate": 9.2242351459504e-06,
+      "loss": 0.521,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7273178602457491,
+      "grad_norm": 0.0730202453066913,
+      "learning_rate": 9.223189043545783e-06,
+      "loss": 0.4982,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7278143229489885,
+      "grad_norm": 0.07738082853275433,
+      "learning_rate": 9.222142295694432e-06,
+      "loss": 0.5325,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7283107856522278,
+      "grad_norm": 0.10268618786682156,
+      "learning_rate": 9.221094902556329e-06,
+      "loss": 0.5434,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7288072483554673,
+      "grad_norm": 0.07328876949375858,
+      "learning_rate": 9.220046864291549e-06,
+      "loss": 0.5175,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7293037110587067,
+      "grad_norm": 0.07254546474047666,
+      "learning_rate": 9.218998181060271e-06,
+      "loss": 0.5406,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7298001737619462,
+      "grad_norm": 0.07561106514048177,
+      "learning_rate": 9.217948853022766e-06,
+      "loss": 0.5229,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7302966364651855,
+      "grad_norm": 0.08213765210369488,
+      "learning_rate": 9.21689888033941e-06,
+      "loss": 0.5214,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7307930991684249,
+      "grad_norm": 0.07417262119469888,
+      "learning_rate": 9.215848263170672e-06,
+      "loss": 0.5309,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7312895618716644,
+      "grad_norm": 0.06898938755371146,
+      "learning_rate": 9.214797001677122e-06,
+      "loss": 0.4872,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7317860245749038,
+      "grad_norm": 0.1042117632122199,
+      "learning_rate": 9.213745096019432e-06,
+      "loss": 0.5298,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7322824872781433,
+      "grad_norm": 0.07050395956024427,
+      "learning_rate": 9.212692546358364e-06,
+      "loss": 0.491,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7327789499813826,
+      "grad_norm": 0.07451854816196715,
+      "learning_rate": 9.211639352854786e-06,
+      "loss": 0.5502,
+      "step": 1476
+    },
+    {
+      "epoch": 0.733275412684622,
+      "grad_norm": 0.06968755487318913,
+      "learning_rate": 9.210585515669664e-06,
+      "loss": 0.4877,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7337718753878615,
+      "grad_norm": 0.07622242603043373,
+      "learning_rate": 9.209531034964055e-06,
+      "loss": 0.555,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7342683380911009,
+      "grad_norm": 0.06999707168192029,
+      "learning_rate": 9.208475910899121e-06,
+      "loss": 0.5393,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7347648007943404,
+      "grad_norm": 0.07107633271501804,
+      "learning_rate": 9.207420143636124e-06,
+      "loss": 0.484,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7352612634975797,
+      "grad_norm": 0.07559982394848742,
+      "learning_rate": 9.206363733336419e-06,
+      "loss": 0.5636,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7357577262008191,
+      "grad_norm": 0.2065710551000181,
+      "learning_rate": 9.20530668016146e-06,
+      "loss": 0.5293,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7362541889040586,
+      "grad_norm": 0.07006483401468061,
+      "learning_rate": 9.204248984272802e-06,
+      "loss": 0.5436,
+      "step": 1483
+    },
+    {
+      "epoch": 0.736750651607298,
+      "grad_norm": 0.08039091210606134,
+      "learning_rate": 9.203190645832098e-06,
+      "loss": 0.5436,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7372471143105375,
+      "grad_norm": 0.07361816607337918,
+      "learning_rate": 9.202131665001096e-06,
+      "loss": 0.5386,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7377435770137768,
+      "grad_norm": 0.07428889038193494,
+      "learning_rate": 9.201072041941644e-06,
+      "loss": 0.5428,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7382400397170162,
+      "grad_norm": 0.0741239104187118,
+      "learning_rate": 9.200011776815691e-06,
+      "loss": 0.564,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7387365024202557,
+      "grad_norm": 0.0761608368835551,
+      "learning_rate": 9.19895086978528e-06,
+      "loss": 0.511,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7392329651234951,
+      "grad_norm": 0.07634971872131308,
+      "learning_rate": 9.197889321012552e-06,
+      "loss": 0.5189,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7397294278267346,
+      "grad_norm": 0.07451573011243572,
+      "learning_rate": 9.196827130659752e-06,
+      "loss": 0.5581,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7402258905299739,
+      "grad_norm": 0.0752122816148327,
+      "learning_rate": 9.195764298889213e-06,
+      "loss": 0.5336,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7407223532332133,
+      "grad_norm": 0.07846376180300381,
+      "learning_rate": 9.194700825863377e-06,
+      "loss": 0.5395,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7412188159364528,
+      "grad_norm": 0.0724115759426848,
+      "learning_rate": 9.193636711744775e-06,
+      "loss": 0.5405,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7417152786396922,
+      "grad_norm": 0.07415138326255291,
+      "learning_rate": 9.192571956696044e-06,
+      "loss": 0.5617,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7422117413429317,
+      "grad_norm": 0.07195572234281511,
+      "learning_rate": 9.19150656087991e-06,
+      "loss": 0.5277,
+      "step": 1495
+    },
+    {
+      "epoch": 0.742708204046171,
+      "grad_norm": 0.07114858026308328,
+      "learning_rate": 9.190440524459203e-06,
+      "loss": 0.5088,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7432046667494104,
+      "grad_norm": 0.07622345987455985,
+      "learning_rate": 9.189373847596853e-06,
+      "loss": 0.534,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7437011294526499,
+      "grad_norm": 0.07656296285911206,
+      "learning_rate": 9.188306530455882e-06,
+      "loss": 0.4895,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7441975921558893,
+      "grad_norm": 0.07356439663118937,
+      "learning_rate": 9.187238573199411e-06,
+      "loss": 0.5347,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7446940548591288,
+      "grad_norm": 0.07131573683028186,
+      "learning_rate": 9.18616997599066e-06,
+      "loss": 0.5155,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7451905175623681,
+      "grad_norm": 0.07289576946445583,
+      "learning_rate": 9.18510073899295e-06,
+      "loss": 0.5572,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7456869802656075,
+      "grad_norm": 0.26584442872430764,
+      "learning_rate": 9.184030862369694e-06,
+      "loss": 0.5308,
+      "step": 1502
+    },
+    {
+      "epoch": 0.746183442968847,
+      "grad_norm": 0.07300736520636775,
+      "learning_rate": 9.182960346284408e-06,
+      "loss": 0.5297,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7466799056720864,
+      "grad_norm": 0.07341036866622214,
+      "learning_rate": 9.181889190900702e-06,
+      "loss": 0.5043,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7471763683753258,
+      "grad_norm": 0.07191242330312968,
+      "learning_rate": 9.180817396382283e-06,
+      "loss": 0.5342,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7476728310785652,
+      "grad_norm": 0.0695672717047009,
+      "learning_rate": 9.17974496289296e-06,
+      "loss": 0.5275,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7481692937818046,
+      "grad_norm": 0.07559861194849239,
+      "learning_rate": 9.178671890596636e-06,
+      "loss": 0.553,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7486657564850441,
+      "grad_norm": 0.08380571084657057,
+      "learning_rate": 9.177598179657314e-06,
+      "loss": 0.5684,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7491622191882835,
+      "grad_norm": 0.07901289401940544,
+      "learning_rate": 9.176523830239093e-06,
+      "loss": 0.5268,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7496586818915228,
+      "grad_norm": 0.0688523499951963,
+      "learning_rate": 9.17544884250617e-06,
+      "loss": 0.5068,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7501551445947623,
+      "grad_norm": 0.06947079013396563,
+      "learning_rate": 9.174373216622841e-06,
+      "loss": 0.4972,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7506516072980017,
+      "grad_norm": 0.07146732680663952,
+      "learning_rate": 9.173296952753494e-06,
+      "loss": 0.5429,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7506516072980017,
+      "eval_loss": 0.5325629115104675,
+      "eval_runtime": 259.2283,
+      "eval_samples_per_second": 117.09,
+      "eval_steps_per_second": 14.64,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7511480700012412,
+      "grad_norm": 0.07362115281936982,
+      "learning_rate": 9.172220051062624e-06,
+      "loss": 0.4972,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7516445327044806,
+      "grad_norm": 0.07398697191188404,
+      "learning_rate": 9.171142511714815e-06,
+      "loss": 0.5469,
+      "step": 1514
+    },
+    {
+      "epoch": 0.75214099540772,
+      "grad_norm": 0.0726269568320206,
+      "learning_rate": 9.17006433487475e-06,
+      "loss": 0.524,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7526374581109594,
+      "grad_norm": 0.07055758850826166,
+      "learning_rate": 9.168985520707215e-06,
+      "loss": 0.5181,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7531339208141988,
+      "grad_norm": 0.07245448024401645,
+      "learning_rate": 9.167906069377088e-06,
+      "loss": 0.5171,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7536303835174383,
+      "grad_norm": 0.07100972584744263,
+      "learning_rate": 9.166825981049345e-06,
+      "loss": 0.5148,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7541268462206777,
+      "grad_norm": 0.0724354876187257,
+      "learning_rate": 9.165745255889062e-06,
+      "loss": 0.5152,
+      "step": 1519
+    },
+    {
+      "epoch": 0.754623308923917,
+      "grad_norm": 0.07341777423839119,
+      "learning_rate": 9.164663894061408e-06,
+      "loss": 0.5398,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7551197716271565,
+      "grad_norm": 0.08055372063895748,
+      "learning_rate": 9.163581895731654e-06,
+      "loss": 0.5235,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7556162343303959,
+      "grad_norm": 0.07199240388013335,
+      "learning_rate": 9.162499261065164e-06,
+      "loss": 0.5076,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7561126970336354,
+      "grad_norm": 0.06963714899756415,
+      "learning_rate": 9.161415990227405e-06,
+      "loss": 0.5237,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7566091597368748,
+      "grad_norm": 0.0770440952254681,
+      "learning_rate": 9.160332083383933e-06,
+      "loss": 0.5329,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7571056224401141,
+      "grad_norm": 0.07151756468144972,
+      "learning_rate": 9.15924754070041e-06,
+      "loss": 0.5189,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7576020851433536,
+      "grad_norm": 0.0729147292162753,
+      "learning_rate": 9.158162362342584e-06,
+      "loss": 0.5355,
+      "step": 1526
+    },
+    {
+      "epoch": 0.758098547846593,
+      "grad_norm": 0.07368817324840382,
+      "learning_rate": 9.157076548476317e-06,
+      "loss": 0.5393,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7585950105498325,
+      "grad_norm": 0.0968328493278401,
+      "learning_rate": 9.155990099267551e-06,
+      "loss": 0.4916,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7590914732530719,
+      "grad_norm": 0.06967081513272996,
+      "learning_rate": 9.154903014882334e-06,
+      "loss": 0.512,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7595879359563112,
+      "grad_norm": 0.06960396037018117,
+      "learning_rate": 9.153815295486811e-06,
+      "loss": 0.5085,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7600843986595507,
+      "grad_norm": 0.06801654295562962,
+      "learning_rate": 9.152726941247223e-06,
+      "loss": 0.5055,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7605808613627901,
+      "grad_norm": 0.0752300824946068,
+      "learning_rate": 9.151637952329903e-06,
+      "loss": 0.5427,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7610773240660296,
+      "grad_norm": 0.07452158768511066,
+      "learning_rate": 9.15054832890129e-06,
+      "loss": 0.5597,
+      "step": 1533
+    },
+    {
+      "epoch": 0.761573786769269,
+      "grad_norm": 0.07028632531598884,
+      "learning_rate": 9.149458071127914e-06,
+      "loss": 0.5202,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7620702494725083,
+      "grad_norm": 0.07239151178732862,
+      "learning_rate": 9.148367179176405e-06,
+      "loss": 0.5144,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7625667121757478,
+      "grad_norm": 0.07357075200741311,
+      "learning_rate": 9.147275653213484e-06,
+      "loss": 0.5582,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7630631748789872,
+      "grad_norm": 0.07390560783887343,
+      "learning_rate": 9.146183493405976e-06,
+      "loss": 0.5662,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7635596375822267,
+      "grad_norm": 0.07055852319994438,
+      "learning_rate": 9.145090699920801e-06,
+      "loss": 0.5061,
+      "step": 1538
+    },
+    {
+      "epoch": 0.764056100285466,
+      "grad_norm": 0.0736301172340416,
+      "learning_rate": 9.143997272924974e-06,
+      "loss": 0.5179,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7645525629887054,
+      "grad_norm": 0.06884159473420354,
+      "learning_rate": 9.142903212585607e-06,
+      "loss": 0.5153,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7650490256919449,
+      "grad_norm": 0.07591997410813622,
+      "learning_rate": 9.14180851906991e-06,
+      "loss": 0.5573,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7655454883951843,
+      "grad_norm": 0.06765520349623405,
+      "learning_rate": 9.140713192545193e-06,
+      "loss": 0.5003,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7660419510984238,
+      "grad_norm": 0.07139622939405862,
+      "learning_rate": 9.139617233178853e-06,
+      "loss": 0.54,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7665384138016631,
+      "grad_norm": 0.0734046035089621,
+      "learning_rate": 9.138520641138391e-06,
+      "loss": 0.5298,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7670348765049025,
+      "grad_norm": 0.07129862203399848,
+      "learning_rate": 9.137423416591408e-06,
+      "loss": 0.537,
+      "step": 1545
+    },
+    {
+      "epoch": 0.767531339208142,
+      "grad_norm": 0.07123101461490239,
+      "learning_rate": 9.136325559705593e-06,
+      "loss": 0.5424,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7680278019113814,
+      "grad_norm": 0.0680050575809704,
+      "learning_rate": 9.135227070648737e-06,
+      "loss": 0.507,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7685242646146209,
+      "grad_norm": 0.07185264556036576,
+      "learning_rate": 9.134127949588727e-06,
+      "loss": 0.5327,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7690207273178602,
+      "grad_norm": 0.07539063274123536,
+      "learning_rate": 9.133028196693548e-06,
+      "loss": 0.5232,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7695171900210996,
+      "grad_norm": 0.07386130365503506,
+      "learning_rate": 9.131927812131273e-06,
+      "loss": 0.5364,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7700136527243391,
+      "grad_norm": 0.06818654256924661,
+      "learning_rate": 9.130826796070085e-06,
+      "loss": 0.5088,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7705101154275785,
+      "grad_norm": 0.06906607950611202,
+      "learning_rate": 9.129725148678252e-06,
+      "loss": 0.497,
+      "step": 1552
+    },
+    {
+      "epoch": 0.771006578130818,
+      "grad_norm": 7.740988013167634,
+      "learning_rate": 9.128622870124147e-06,
+      "loss": 0.5968,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7715030408340573,
+      "grad_norm": 0.07199899183090658,
+      "learning_rate": 9.127519960576234e-06,
+      "loss": 0.5153,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7719995035372967,
+      "grad_norm": 0.07229222543077109,
+      "learning_rate": 9.126416420203072e-06,
+      "loss": 0.5207,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7724959662405362,
+      "grad_norm": 0.07716579011178949,
+      "learning_rate": 9.125312249173325e-06,
+      "loss": 0.5099,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7729924289437756,
+      "grad_norm": 0.07253532602344766,
+      "learning_rate": 9.124207447655744e-06,
+      "loss": 0.5323,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7734888916470151,
+      "grad_norm": 0.07497907003215884,
+      "learning_rate": 9.123102015819184e-06,
+      "loss": 0.5201,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7739853543502544,
+      "grad_norm": 0.07354183539985416,
+      "learning_rate": 9.121995953832585e-06,
+      "loss": 0.5053,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7744818170534938,
+      "grad_norm": 0.07035972316181158,
+      "learning_rate": 9.120889261864999e-06,
+      "loss": 0.5156,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7749782797567333,
+      "grad_norm": 0.08187819899227436,
+      "learning_rate": 9.119781940085561e-06,
+      "loss": 0.5326,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7754747424599727,
+      "grad_norm": 0.0736392754058841,
+      "learning_rate": 9.11867398866351e-06,
+      "loss": 0.5146,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7759712051632122,
+      "grad_norm": 0.07456619741450733,
+      "learning_rate": 9.117565407768178e-06,
+      "loss": 0.5476,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7764676678664515,
+      "grad_norm": 0.07319757416197845,
+      "learning_rate": 9.116456197568993e-06,
+      "loss": 0.5591,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7769641305696909,
+      "grad_norm": 0.06997134853606297,
+      "learning_rate": 9.11534635823548e-06,
+      "loss": 0.4941,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7774605932729304,
+      "grad_norm": 0.07205444260596973,
+      "learning_rate": 9.114235889937262e-06,
+      "loss": 0.5427,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7779570559761698,
+      "grad_norm": 0.07788512971157865,
+      "learning_rate": 9.113124792844053e-06,
+      "loss": 0.5207,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7784535186794093,
+      "grad_norm": 0.08002426048799419,
+      "learning_rate": 9.112013067125671e-06,
+      "loss": 0.5307,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7789499813826486,
+      "grad_norm": 0.07468268399263038,
+      "learning_rate": 9.11090071295202e-06,
+      "loss": 0.5262,
+      "step": 1569
+    },
+    {
+      "epoch": 0.779446444085888,
+      "grad_norm": 0.07298474341759807,
+      "learning_rate": 9.109787730493111e-06,
+      "loss": 0.5407,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7799429067891275,
+      "grad_norm": 0.07728077776334637,
+      "learning_rate": 9.10867411991904e-06,
+      "loss": 0.5579,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7804393694923669,
+      "grad_norm": 0.07344685556951021,
+      "learning_rate": 9.10755988140001e-06,
+      "loss": 0.5205,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7809358321956062,
+      "grad_norm": 0.07014883880403816,
+      "learning_rate": 9.10644501510631e-06,
+      "loss": 0.4727,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7814322948988457,
+      "grad_norm": 0.07213557704529419,
+      "learning_rate": 9.105329521208334e-06,
+      "loss": 0.5279,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7819287576020851,
+      "grad_norm": 0.0757860432054162,
+      "learning_rate": 9.104213399876562e-06,
+      "loss": 0.5387,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7824252203053246,
+      "grad_norm": 0.07057256214391969,
+      "learning_rate": 9.103096651281578e-06,
+      "loss": 0.5345,
+      "step": 1576
+    },
+    {
+      "epoch": 0.782921683008564,
+      "grad_norm": 0.07132201540873075,
+      "learning_rate": 9.101979275594061e-06,
+      "loss": 0.5098,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7834181457118033,
+      "grad_norm": 0.07152535103531563,
+      "learning_rate": 9.10086127298478e-06,
+      "loss": 0.5062,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7839146084150428,
+      "grad_norm": 0.07093279415393061,
+      "learning_rate": 9.099742643624607e-06,
+      "loss": 0.5225,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7844110711182822,
+      "grad_norm": 0.07636988653396815,
+      "learning_rate": 9.098623387684504e-06,
+      "loss": 0.5901,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7849075338215217,
+      "grad_norm": 0.0726902485186396,
+      "learning_rate": 9.097503505335534e-06,
+      "loss": 0.5992,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7854039965247611,
+      "grad_norm": 0.07148367002169637,
+      "learning_rate": 9.09638299674885e-06,
+      "loss": 0.5249,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7859004592280004,
+      "grad_norm": 0.07187191854489607,
+      "learning_rate": 9.095261862095706e-06,
+      "loss": 0.5149,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7863969219312399,
+      "grad_norm": 0.07433331234158458,
+      "learning_rate": 9.09414010154745e-06,
+      "loss": 0.4996,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7868933846344793,
+      "grad_norm": 0.07421628137732915,
+      "learning_rate": 9.093017715275523e-06,
+      "loss": 0.5373,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7873898473377188,
+      "grad_norm": 0.0745071207582901,
+      "learning_rate": 9.091894703451464e-06,
+      "loss": 0.5044,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7878863100409582,
+      "grad_norm": 0.07565857135893618,
+      "learning_rate": 9.090771066246911e-06,
+      "loss": 0.536,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7883827727441975,
+      "grad_norm": 0.07130484120336626,
+      "learning_rate": 9.089646803833589e-06,
+      "loss": 0.4945,
+      "step": 1588
+    },
+    {
+      "epoch": 0.788879235447437,
+      "grad_norm": 0.07351509883364295,
+      "learning_rate": 9.088521916383326e-06,
+      "loss": 0.544,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7893756981506764,
+      "grad_norm": 0.07229735716128917,
+      "learning_rate": 9.087396404068043e-06,
+      "loss": 0.5031,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7898721608539159,
+      "grad_norm": 0.07640419900882879,
+      "learning_rate": 9.086270267059755e-06,
+      "loss": 0.5397,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7903686235571553,
+      "grad_norm": 0.07396099607093055,
+      "learning_rate": 9.085143505530576e-06,
+      "loss": 0.564,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7908650862603946,
+      "grad_norm": 0.06974958002197082,
+      "learning_rate": 9.084016119652711e-06,
+      "loss": 0.5277,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7913615489636341,
+      "grad_norm": 0.07711746671171255,
+      "learning_rate": 9.082888109598465e-06,
+      "loss": 0.52,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7918580116668735,
+      "grad_norm": 0.07562328283238078,
+      "learning_rate": 9.081759475540236e-06,
+      "loss": 0.5187,
+      "step": 1595
+    },
+    {
+      "epoch": 0.792354474370113,
+      "grad_norm": 0.07163646761900765,
+      "learning_rate": 9.080630217650516e-06,
+      "loss": 0.5116,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7928509370733524,
+      "grad_norm": 0.07470949012871686,
+      "learning_rate": 9.079500336101898e-06,
+      "loss": 0.511,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7933473997765917,
+      "grad_norm": 0.07250875886295932,
+      "learning_rate": 9.078369831067062e-06,
+      "loss": 0.5349,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7938438624798312,
+      "grad_norm": 0.0739653329382637,
+      "learning_rate": 9.077238702718786e-06,
+      "loss": 0.5229,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7943403251830706,
+      "grad_norm": 0.07107778628692231,
+      "learning_rate": 9.076106951229952e-06,
+      "loss": 0.516,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7948367878863101,
+      "grad_norm": 0.0725789052221172,
+      "learning_rate": 9.074974576773525e-06,
+      "loss": 0.5106,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7953332505895495,
+      "grad_norm": 0.07305095035904646,
+      "learning_rate": 9.073841579522571e-06,
+      "loss": 0.5172,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7958297132927888,
+      "grad_norm": 0.07264990359957456,
+      "learning_rate": 9.07270795965025e-06,
+      "loss": 0.508,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7963261759960283,
+      "grad_norm": 0.0826982852729199,
+      "learning_rate": 9.071573717329818e-06,
+      "loss": 0.5049,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7968226386992677,
+      "grad_norm": 0.07656668555306413,
+      "learning_rate": 9.070438852734627e-06,
+      "loss": 0.5313,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7973191014025072,
+      "grad_norm": 0.07069294942555418,
+      "learning_rate": 9.069303366038122e-06,
+      "loss": 0.5177,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7978155641057466,
+      "grad_norm": 0.0777471019112174,
+      "learning_rate": 9.068167257413842e-06,
+      "loss": 0.5439,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7983120268089859,
+      "grad_norm": 0.07682960008512643,
+      "learning_rate": 9.067030527035426e-06,
+      "loss": 0.5406,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7988084895122254,
+      "grad_norm": 0.0807223836950902,
+      "learning_rate": 9.065893175076604e-06,
+      "loss": 0.561,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7993049522154648,
+      "grad_norm": 0.07047014648160267,
+      "learning_rate": 9.064755201711202e-06,
+      "loss": 0.5154,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7998014149187043,
+      "grad_norm": 0.07595726981086956,
+      "learning_rate": 9.06361660711314e-06,
+      "loss": 0.5027,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8002978776219436,
+      "grad_norm": 0.08005685606981752,
+      "learning_rate": 9.062477391456436e-06,
+      "loss": 0.5118,
+      "step": 1612
+    },
+    {
+      "epoch": 0.800794340325183,
+      "grad_norm": 0.07718134842573832,
+      "learning_rate": 9.0613375549152e-06,
+      "loss": 0.565,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8012908030284225,
+      "grad_norm": 0.07158710304795181,
+      "learning_rate": 9.060197097663634e-06,
+      "loss": 0.5297,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8017872657316619,
+      "grad_norm": 0.07203230378497506,
+      "learning_rate": 9.059056019876044e-06,
+      "loss": 0.5164,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8022837284349014,
+      "grad_norm": 0.07406422518581404,
+      "learning_rate": 9.057914321726824e-06,
+      "loss": 0.5327,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8027801911381407,
+      "grad_norm": 0.07251157226369823,
+      "learning_rate": 9.056772003390464e-06,
+      "loss": 0.5002,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8032766538413801,
+      "grad_norm": 0.07167096759529387,
+      "learning_rate": 9.055629065041547e-06,
+      "loss": 0.492,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8037731165446196,
+      "grad_norm": 0.07455451424426704,
+      "learning_rate": 9.054485506854756e-06,
+      "loss": 0.5225,
+      "step": 1619
+    },
+    {
+      "epoch": 0.804269579247859,
+      "grad_norm": 0.07080913355805624,
+      "learning_rate": 9.053341329004863e-06,
+      "loss": 0.5133,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8047660419510985,
+      "grad_norm": 0.07413789776565281,
+      "learning_rate": 9.05219653166674e-06,
+      "loss": 0.5035,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8052625046543378,
+      "grad_norm": 0.0703480239098969,
+      "learning_rate": 9.051051115015346e-06,
+      "loss": 0.5276,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8057589673575772,
+      "grad_norm": 0.0765687761483791,
+      "learning_rate": 9.049905079225744e-06,
+      "loss": 0.5311,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8062554300608167,
+      "grad_norm": 0.07143355651802613,
+      "learning_rate": 9.048758424473088e-06,
+      "loss": 0.5458,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8067518927640561,
+      "grad_norm": 0.07465766054033428,
+      "learning_rate": 9.047611150932621e-06,
+      "loss": 0.5764,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8072483554672956,
+      "grad_norm": 0.07113501350075914,
+      "learning_rate": 9.04646325877969e-06,
+      "loss": 0.5159,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8077448181705349,
+      "grad_norm": 0.0710495387744484,
+      "learning_rate": 9.045314748189728e-06,
+      "loss": 0.5323,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8082412808737743,
+      "grad_norm": 0.07185246674290095,
+      "learning_rate": 9.04416561933827e-06,
+      "loss": 0.5299,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8087377435770138,
+      "grad_norm": 0.06825471797050817,
+      "learning_rate": 9.04301587240094e-06,
+      "loss": 0.4911,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8092342062802532,
+      "grad_norm": 0.07128246257879234,
+      "learning_rate": 9.041865507553458e-06,
+      "loss": 0.5186,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8097306689834927,
+      "grad_norm": 0.0731020684241034,
+      "learning_rate": 9.04071452497164e-06,
+      "loss": 0.5209,
+      "step": 1631
+    },
+    {
+      "epoch": 0.810227131686732,
+      "grad_norm": 0.07250812004228605,
+      "learning_rate": 9.039562924831395e-06,
+      "loss": 0.5277,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8107235943899714,
+      "grad_norm": 0.0768123954655183,
+      "learning_rate": 9.038410707308727e-06,
+      "loss": 0.5584,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8112200570932109,
+      "grad_norm": 0.07224607675157957,
+      "learning_rate": 9.037257872579733e-06,
+      "loss": 0.5241,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8117165197964503,
+      "grad_norm": 0.07069975133524385,
+      "learning_rate": 9.036104420820606e-06,
+      "loss": 0.5242,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8122129824996898,
+      "grad_norm": 0.07184499072459798,
+      "learning_rate": 9.034950352207632e-06,
+      "loss": 0.5434,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8127094452029291,
+      "grad_norm": 0.07144391298900397,
+      "learning_rate": 9.033795666917191e-06,
+      "loss": 0.5169,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8132059079061685,
+      "grad_norm": 0.06912538260937971,
+      "learning_rate": 9.032640365125761e-06,
+      "loss": 0.5312,
+      "step": 1638
+    },
+    {
+      "epoch": 0.813702370609408,
+      "grad_norm": 0.07046782884763288,
+      "learning_rate": 9.031484447009908e-06,
+      "loss": 0.5351,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8141988333126474,
+      "grad_norm": 0.06976119450345018,
+      "learning_rate": 9.0303279127463e-06,
+      "loss": 0.4855,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8146952960158869,
+      "grad_norm": 0.06869843149728422,
+      "learning_rate": 9.02917076251169e-06,
+      "loss": 0.5057,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8151917587191262,
+      "grad_norm": 0.07434287432224491,
+      "learning_rate": 9.02801299648293e-06,
+      "loss": 0.5234,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8156882214223656,
+      "grad_norm": 0.06957914337939076,
+      "learning_rate": 9.02685461483697e-06,
+      "loss": 0.541,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8161846841256051,
+      "grad_norm": 0.07218018252814719,
+      "learning_rate": 9.025695617750848e-06,
+      "loss": 0.5486,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8166811468288445,
+      "grad_norm": 0.07168288119557607,
+      "learning_rate": 9.024536005401697e-06,
+      "loss": 0.5013,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8171776095320838,
+      "grad_norm": 0.0750914692398203,
+      "learning_rate": 9.023375777966747e-06,
+      "loss": 0.5278,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8176740722353233,
+      "grad_norm": 0.07583223590925382,
+      "learning_rate": 9.022214935623318e-06,
+      "loss": 0.5198,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8181705349385627,
+      "grad_norm": 0.07173398671171025,
+      "learning_rate": 9.02105347854883e-06,
+      "loss": 0.5073,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8186669976418022,
+      "grad_norm": 0.0745847654854046,
+      "learning_rate": 9.019891406920788e-06,
+      "loss": 0.5383,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8191634603450416,
+      "grad_norm": 0.07086350747239142,
+      "learning_rate": 9.018728720916798e-06,
+      "loss": 0.5287,
+      "step": 1650
+    },
+    {
+      "epoch": 0.819659923048281,
+      "grad_norm": 0.07777382699121162,
+      "learning_rate": 9.01756542071456e-06,
+      "loss": 0.5203,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8201563857515204,
+      "grad_norm": 0.07603501493619808,
+      "learning_rate": 9.016401506491863e-06,
+      "loss": 0.5363,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8206528484547598,
+      "grad_norm": 0.07394262339412508,
+      "learning_rate": 9.015236978426595e-06,
+      "loss": 0.5478,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8211493111579993,
+      "grad_norm": 0.07074353754605808,
+      "learning_rate": 9.014071836696734e-06,
+      "loss": 0.4966,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8216457738612387,
+      "grad_norm": 0.07423057235044962,
+      "learning_rate": 9.012906081480354e-06,
+      "loss": 0.5172,
+      "step": 1655
+    },
+    {
+      "epoch": 0.822142236564478,
+      "grad_norm": 0.07522410434633135,
+      "learning_rate": 9.011739712955621e-06,
+      "loss": 0.5228,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8226386992677175,
+      "grad_norm": 0.07547811533928037,
+      "learning_rate": 9.010572731300796e-06,
+      "loss": 0.5376,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8231351619709569,
+      "grad_norm": 0.07471384072083828,
+      "learning_rate": 9.009405136694234e-06,
+      "loss": 0.5194,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8236316246741964,
+      "grad_norm": 0.07608351907222413,
+      "learning_rate": 9.008236929314383e-06,
+      "loss": 0.5367,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8241280873774358,
+      "grad_norm": 0.07635352589296011,
+      "learning_rate": 9.007068109339783e-06,
+      "loss": 0.5292,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8246245500806751,
+      "grad_norm": 0.07282084669410545,
+      "learning_rate": 9.005898676949073e-06,
+      "loss": 0.5448,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8251210127839146,
+      "grad_norm": 0.0707264456584643,
+      "learning_rate": 9.00472863232098e-06,
+      "loss": 0.492,
+      "step": 1662
+    },
+    {
+      "epoch": 0.825617475487154,
+      "grad_norm": 0.0731668244630585,
+      "learning_rate": 9.003557975634325e-06,
+      "loss": 0.5227,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8261139381903935,
+      "grad_norm": 0.07211067458154444,
+      "learning_rate": 9.002386707068026e-06,
+      "loss": 0.5333,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8266104008936329,
+      "grad_norm": 0.08009101745689773,
+      "learning_rate": 9.001214826801092e-06,
+      "loss": 0.5483,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8271068635968722,
+      "grad_norm": 0.0698658302792023,
+      "learning_rate": 9.000042335012627e-06,
+      "loss": 0.537,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8276033263001117,
+      "grad_norm": 0.0739225167598179,
+      "learning_rate": 8.998869231881827e-06,
+      "loss": 0.5246,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8280997890033511,
+      "grad_norm": 0.07383535775288107,
+      "learning_rate": 8.997695517587981e-06,
+      "loss": 0.5328,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8285962517065906,
+      "grad_norm": 0.07322555009863138,
+      "learning_rate": 8.996521192310474e-06,
+      "loss": 0.5529,
+      "step": 1669
+    },
+    {
+      "epoch": 0.82909271440983,
+      "grad_norm": 0.07112105943789013,
+      "learning_rate": 8.995346256228782e-06,
+      "loss": 0.5422,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8295891771130693,
+      "grad_norm": 0.07542664568841825,
+      "learning_rate": 8.994170709522473e-06,
+      "loss": 0.5439,
+      "step": 1671
+    },
+    {
+      "epoch": 0.8300856398163088,
+      "grad_norm": 0.07050701387729236,
+      "learning_rate": 8.992994552371217e-06,
+      "loss": 0.5269,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8305821025195482,
+      "grad_norm": 0.07092483775981985,
+      "learning_rate": 8.991817784954764e-06,
+      "loss": 0.5481,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8310785652227877,
+      "grad_norm": 0.06851211677419224,
+      "learning_rate": 8.990640407452966e-06,
+      "loss": 0.5005,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8315750279260271,
+      "grad_norm": 0.074848670606186,
+      "learning_rate": 8.989462420045768e-06,
+      "loss": 0.5087,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8320714906292664,
+      "grad_norm": 0.07915295738445463,
+      "learning_rate": 8.988283822913205e-06,
+      "loss": 0.5564,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8325679533325059,
+      "grad_norm": 0.07375116283085385,
+      "learning_rate": 8.987104616235407e-06,
+      "loss": 0.5184,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8330644160357453,
+      "grad_norm": 0.07598742406736596,
+      "learning_rate": 8.985924800192597e-06,
+      "loss": 0.5732,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8335608787389848,
+      "grad_norm": 0.0733126506443,
+      "learning_rate": 8.98474437496509e-06,
+      "loss": 0.596,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8340573414422241,
+      "grad_norm": 0.07315827370163273,
+      "learning_rate": 8.983563340733296e-06,
+      "loss": 0.5723,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8345538041454635,
+      "grad_norm": 0.07562637951263981,
+      "learning_rate": 8.982381697677717e-06,
+      "loss": 0.5499,
+      "step": 1681
+    },
+    {
+      "epoch": 0.835050266848703,
+      "grad_norm": 0.07599468393821707,
+      "learning_rate": 8.981199445978947e-06,
+      "loss": 0.5214,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8355467295519424,
+      "grad_norm": 0.07295503864472273,
+      "learning_rate": 8.980016585817677e-06,
+      "loss": 0.5008,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8360431922551819,
+      "grad_norm": 0.07626131253063163,
+      "learning_rate": 8.978833117374685e-06,
+      "loss": 0.5449,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8365396549584212,
+      "grad_norm": 0.06955667754604454,
+      "learning_rate": 8.97764904083085e-06,
+      "loss": 0.5214,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8370361176616606,
+      "grad_norm": 0.0730561593210296,
+      "learning_rate": 8.976464356367133e-06,
+      "loss": 0.548,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8375325803649001,
+      "grad_norm": 0.06996728191269769,
+      "learning_rate": 8.975279064164597e-06,
+      "loss": 0.5328,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8380290430681395,
+      "grad_norm": 0.07041962149072396,
+      "learning_rate": 8.974093164404396e-06,
+      "loss": 0.5198,
+      "step": 1688
+    },
+    {
+      "epoch": 0.838525505771379,
+      "grad_norm": 0.07163834216651387,
+      "learning_rate": 8.972906657267773e-06,
+      "loss": 0.5137,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8390219684746183,
+      "grad_norm": 0.07387087026768005,
+      "learning_rate": 8.97171954293607e-06,
+      "loss": 0.6089,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8395184311778577,
+      "grad_norm": 0.07356742299332733,
+      "learning_rate": 8.970531821590715e-06,
+      "loss": 0.5234,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8400148938810972,
+      "grad_norm": 0.07259975026238215,
+      "learning_rate": 8.969343493413234e-06,
+      "loss": 0.5185,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8405113565843366,
+      "grad_norm": 0.07374705958699512,
+      "learning_rate": 8.968154558585244e-06,
+      "loss": 0.5354,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8410078192875761,
+      "grad_norm": 0.07018652218556445,
+      "learning_rate": 8.966965017288456e-06,
+      "loss": 0.5333,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8415042819908154,
+      "grad_norm": 0.0723460304446156,
+      "learning_rate": 8.965774869704669e-06,
+      "loss": 0.549,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8420007446940548,
+      "grad_norm": 0.07257751184105775,
+      "learning_rate": 8.964584116015777e-06,
+      "loss": 0.5398,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8424972073972943,
+      "grad_norm": 0.07080762620664613,
+      "learning_rate": 8.963392756403774e-06,
+      "loss": 0.5359,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8429936701005337,
+      "grad_norm": 0.0716877086229164,
+      "learning_rate": 8.962200791050734e-06,
+      "loss": 0.5142,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8434901328037732,
+      "grad_norm": 0.07454251146437897,
+      "learning_rate": 8.961008220138833e-06,
+      "loss": 0.5383,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8439865955070125,
+      "grad_norm": 0.0736072882247968,
+      "learning_rate": 8.959815043850336e-06,
+      "loss": 0.545,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8444830582102519,
+      "grad_norm": 0.07201680235850795,
+      "learning_rate": 8.9586212623676e-06,
+      "loss": 0.5033,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8449795209134914,
+      "grad_norm": 0.07282467805491234,
+      "learning_rate": 8.957426875873075e-06,
+      "loss": 0.5163,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8454759836167308,
+      "grad_norm": 0.07208805465158266,
+      "learning_rate": 8.956231884549304e-06,
+      "loss": 0.502,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8459724463199703,
+      "grad_norm": 0.07232790534446927,
+      "learning_rate": 8.955036288578924e-06,
+      "loss": 0.5406,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8464689090232096,
+      "grad_norm": 0.07481224199301578,
+      "learning_rate": 8.953840088144663e-06,
+      "loss": 0.546,
+      "step": 1705
+    },
+    {
+      "epoch": 0.846965371726449,
+      "grad_norm": 0.06953858909134908,
+      "learning_rate": 8.952643283429337e-06,
+      "loss": 0.4947,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8474618344296885,
+      "grad_norm": 0.07353616438918857,
+      "learning_rate": 8.951445874615862e-06,
+      "loss": 0.5537,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8479582971329279,
+      "grad_norm": 0.06906196166656396,
+      "learning_rate": 8.950247861887242e-06,
+      "loss": 0.4881,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8484547598361674,
+      "grad_norm": 0.07103588869751419,
+      "learning_rate": 8.949049245426573e-06,
+      "loss": 0.4984,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8489512225394067,
+      "grad_norm": 0.07620017467523198,
+      "learning_rate": 8.947850025417044e-06,
+      "loss": 0.5356,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8494476852426461,
+      "grad_norm": 0.07734988462561786,
+      "learning_rate": 8.94665020204194e-06,
+      "loss": 0.5357,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8499441479458856,
+      "grad_norm": 0.07415145536146497,
+      "learning_rate": 8.945449775484631e-06,
+      "loss": 0.5621,
+      "step": 1712
+    },
+    {
+      "epoch": 0.850440610649125,
+      "grad_norm": 0.07228473869129433,
+      "learning_rate": 8.944248745928584e-06,
+      "loss": 0.5114,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8509370733523643,
+      "grad_norm": 0.06940824642647671,
+      "learning_rate": 8.943047113557358e-06,
+      "loss": 0.5199,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8514335360556038,
+      "grad_norm": 0.07607055327588681,
+      "learning_rate": 8.941844878554602e-06,
+      "loss": 0.5556,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8519299987588432,
+      "grad_norm": 0.07198339251155979,
+      "learning_rate": 8.94064204110406e-06,
+      "loss": 0.5257,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8524264614620827,
+      "grad_norm": 0.06934540644110943,
+      "learning_rate": 8.939438601389566e-06,
+      "loss": 0.5085,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8529229241653221,
+      "grad_norm": 0.07389823313277775,
+      "learning_rate": 8.938234559595044e-06,
+      "loss": 0.5853,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8534193868685614,
+      "grad_norm": 0.07106273787523364,
+      "learning_rate": 8.937029915904515e-06,
+      "loss": 0.5172,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8539158495718009,
+      "grad_norm": 0.07626576874496946,
+      "learning_rate": 8.93582467050209e-06,
+      "loss": 0.5902,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8544123122750403,
+      "grad_norm": 0.07104800686298383,
+      "learning_rate": 8.934618823571968e-06,
+      "loss": 0.5669,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8549087749782798,
+      "grad_norm": 0.07098134787197038,
+      "learning_rate": 8.933412375298447e-06,
+      "loss": 0.5071,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8554052376815192,
+      "grad_norm": 0.07178430609385278,
+      "learning_rate": 8.932205325865912e-06,
+      "loss": 0.5663,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8559017003847585,
+      "grad_norm": 0.07321592545745911,
+      "learning_rate": 8.93099767545884e-06,
+      "loss": 0.5684,
+      "step": 1724
+    },
+    {
+      "epoch": 0.856398163087998,
+      "grad_norm": 0.07405468733080081,
+      "learning_rate": 8.929789424261804e-06,
+      "loss": 0.5093,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8568946257912374,
+      "grad_norm": 0.06958960035844272,
+      "learning_rate": 8.928580572459462e-06,
+      "loss": 0.516,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8573910884944769,
+      "grad_norm": 0.07477229658813306,
+      "learning_rate": 8.92737112023657e-06,
+      "loss": 0.5131,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8578875511977163,
+      "grad_norm": 0.07175257894439932,
+      "learning_rate": 8.926161067777973e-06,
+      "loss": 0.5119,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8583840139009556,
+      "grad_norm": 0.07751441304214718,
+      "learning_rate": 8.924950415268609e-06,
+      "loss": 0.563,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8588804766041951,
+      "grad_norm": 0.07420172426697907,
+      "learning_rate": 8.923739162893505e-06,
+      "loss": 0.5324,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8593769393074345,
+      "grad_norm": 0.07407924635811906,
+      "learning_rate": 8.922527310837782e-06,
+      "loss": 0.4887,
+      "step": 1731
+    },
+    {
+      "epoch": 0.859873402010674,
+      "grad_norm": 0.07350646391304734,
+      "learning_rate": 8.921314859286651e-06,
+      "loss": 0.5163,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8603698647139134,
+      "grad_norm": 0.07043961460651518,
+      "learning_rate": 8.920101808425422e-06,
+      "loss": 0.5057,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8608663274171527,
+      "grad_norm": 0.17234638710081665,
+      "learning_rate": 8.91888815843948e-06,
+      "loss": 0.5448,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8613627901203922,
+      "grad_norm": 0.0732175517309694,
+      "learning_rate": 8.917673909514321e-06,
+      "loss": 0.5299,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8618592528236316,
+      "grad_norm": 0.07453686320781494,
+      "learning_rate": 8.916459061835519e-06,
+      "loss": 0.546,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8623557155268711,
+      "grad_norm": 0.07567184139704818,
+      "learning_rate": 8.915243615588745e-06,
+      "loss": 0.5669,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8628521782301105,
+      "grad_norm": 0.07425811097859154,
+      "learning_rate": 8.914027570959762e-06,
+      "loss": 0.5029,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8633486409333498,
+      "grad_norm": 0.07391851841657099,
+      "learning_rate": 8.91281092813442e-06,
+      "loss": 0.532,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8638451036365893,
+      "grad_norm": 0.09916635466093679,
+      "learning_rate": 8.911593687298665e-06,
+      "loss": 0.5352,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8643415663398287,
+      "grad_norm": 0.0723850348145785,
+      "learning_rate": 8.910375848638532e-06,
+      "loss": 0.5222,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8648380290430682,
+      "grad_norm": 0.0707452473794512,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.5296,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8653344917463076,
+      "grad_norm": 0.07299743926849714,
+      "learning_rate": 8.907938378589735e-06,
+      "loss": 0.5235,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8658309544495469,
+      "grad_norm": 0.0714751596943647,
+      "learning_rate": 8.906718747573598e-06,
+      "loss": 0.5276,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8663274171527864,
+      "grad_norm": 0.07178899786010984,
+      "learning_rate": 8.90549851947814e-06,
+      "loss": 0.5128,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8668238798560258,
+      "grad_norm": 0.0731518645963831,
+      "learning_rate": 8.904277694489853e-06,
+      "loss": 0.5194,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8673203425592653,
+      "grad_norm": 0.07149517704501569,
+      "learning_rate": 8.903056272795322e-06,
+      "loss": 0.5135,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8678168052625047,
+      "grad_norm": 0.0706661260761562,
+      "learning_rate": 8.901834254581219e-06,
+      "loss": 0.5,
+      "step": 1748
+    },
+    {
+      "epoch": 0.868313267965744,
+      "grad_norm": 0.0699038656327828,
+      "learning_rate": 8.900611640034313e-06,
+      "loss": 0.5424,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8688097306689835,
+      "grad_norm": 0.07062531372070045,
+      "learning_rate": 8.899388429341459e-06,
+      "loss": 0.5001,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8693061933722229,
+      "grad_norm": 0.0734084689002575,
+      "learning_rate": 8.898164622689604e-06,
+      "loss": 0.4946,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8698026560754624,
+      "grad_norm": 0.07054036867742379,
+      "learning_rate": 8.896940220265789e-06,
+      "loss": 0.4898,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8702991187787017,
+      "grad_norm": 0.07268417521194488,
+      "learning_rate": 8.895715222257144e-06,
+      "loss": 0.539,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8707955814819411,
+      "grad_norm": 0.08161951973156148,
+      "learning_rate": 8.894489628850891e-06,
+      "loss": 0.501,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8712920441851806,
+      "grad_norm": 0.07063657796104031,
+      "learning_rate": 8.893263440234341e-06,
+      "loss": 0.5108,
+      "step": 1755
+    },
+    {
+      "epoch": 0.87178850688842,
+      "grad_norm": 0.07093034108492971,
+      "learning_rate": 8.892036656594898e-06,
+      "loss": 0.5045,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8722849695916595,
+      "grad_norm": 0.07173587549839525,
+      "learning_rate": 8.890809278120056e-06,
+      "loss": 0.5442,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8727814322948988,
+      "grad_norm": 0.07594294618433388,
+      "learning_rate": 8.889581304997401e-06,
+      "loss": 0.5691,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8732778949981382,
+      "grad_norm": 0.07023217187033555,
+      "learning_rate": 8.88835273741461e-06,
+      "loss": 0.494,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8737743577013777,
+      "grad_norm": 0.0747344031630131,
+      "learning_rate": 8.887123575559445e-06,
+      "loss": 0.5287,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8742708204046171,
+      "grad_norm": 0.070787626979846,
+      "learning_rate": 8.885893819619768e-06,
+      "loss": 0.4994,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8747672831078566,
+      "grad_norm": 0.07053403609249528,
+      "learning_rate": 8.884663469783526e-06,
+      "loss": 0.5206,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8752637458110959,
+      "grad_norm": 0.07093547787233846,
+      "learning_rate": 8.883432526238757e-06,
+      "loss": 0.5518,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8757602085143353,
+      "grad_norm": 0.07623805022634478,
+      "learning_rate": 8.882200989173595e-06,
+      "loss": 0.5243,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8762566712175748,
+      "grad_norm": 0.07046322887327075,
+      "learning_rate": 8.880968858776257e-06,
+      "loss": 0.5434,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8767531339208142,
+      "grad_norm": 0.07111009011290098,
+      "learning_rate": 8.879736135235055e-06,
+      "loss": 0.5327,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8772495966240537,
+      "grad_norm": 0.07166896583234532,
+      "learning_rate": 8.878502818738393e-06,
+      "loss": 0.5299,
+      "step": 1767
+    },
+    {
+      "epoch": 0.877746059327293,
+      "grad_norm": 0.07148408961382326,
+      "learning_rate": 8.87726890947476e-06,
+      "loss": 0.5107,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8782425220305324,
+      "grad_norm": 0.07238265024706987,
+      "learning_rate": 8.876034407632743e-06,
+      "loss": 0.5135,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8787389847337719,
+      "grad_norm": 0.07047522806209866,
+      "learning_rate": 8.874799313401014e-06,
+      "loss": 0.5153,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8792354474370113,
+      "grad_norm": 0.07174926494461373,
+      "learning_rate": 8.873563626968337e-06,
+      "loss": 0.5344,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8797319101402508,
+      "grad_norm": 0.07469035136029122,
+      "learning_rate": 8.87232734852357e-06,
+      "loss": 0.5491,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8802283728434901,
+      "grad_norm": 0.07023138060368163,
+      "learning_rate": 8.871090478255654e-06,
+      "loss": 0.5239,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8807248355467295,
+      "grad_norm": 0.07037733604789666,
+      "learning_rate": 8.869853016353627e-06,
+      "loss": 0.5273,
+      "step": 1774
+    },
+    {
+      "epoch": 0.881221298249969,
+      "grad_norm": 0.06824188101306318,
+      "learning_rate": 8.868614963006615e-06,
+      "loss": 0.5118,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8817177609532084,
+      "grad_norm": 0.06912832231818924,
+      "learning_rate": 8.867376318403834e-06,
+      "loss": 0.4811,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8822142236564479,
+      "grad_norm": 0.07774906294910376,
+      "learning_rate": 8.866137082734591e-06,
+      "loss": 0.5466,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8827106863596872,
+      "grad_norm": 0.07045859944368564,
+      "learning_rate": 8.864897256188283e-06,
+      "loss": 0.5163,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8832071490629266,
+      "grad_norm": 0.07003492490692469,
+      "learning_rate": 8.8636568389544e-06,
+      "loss": 0.5414,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8837036117661661,
+      "grad_norm": 0.07307705639634823,
+      "learning_rate": 8.862415831222518e-06,
+      "loss": 0.5571,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8842000744694055,
+      "grad_norm": 0.07483261513383652,
+      "learning_rate": 8.861174233182303e-06,
+      "loss": 0.5431,
+      "step": 1781
+    },
+    {
+      "epoch": 0.884696537172645,
+      "grad_norm": 0.07173435811049399,
+      "learning_rate": 8.859932045023518e-06,
+      "loss": 0.5266,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8851929998758843,
+      "grad_norm": 0.07344114936635032,
+      "learning_rate": 8.85868926693601e-06,
+      "loss": 0.5074,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8856894625791237,
+      "grad_norm": 0.0749870011007574,
+      "learning_rate": 8.857445899109716e-06,
+      "loss": 0.537,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8861859252823632,
+      "grad_norm": 0.07353741724151482,
+      "learning_rate": 8.856201941734664e-06,
+      "loss": 0.5381,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8866823879856026,
+      "grad_norm": 0.06935062733647068,
+      "learning_rate": 8.854957395000977e-06,
+      "loss": 0.5384,
+      "step": 1786
+    },
+    {
+      "epoch": 0.887178850688842,
+      "grad_norm": 0.06984395557189015,
+      "learning_rate": 8.853712259098862e-06,
+      "loss": 0.5432,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8876753133920814,
+      "grad_norm": 0.07009190296729087,
+      "learning_rate": 8.852466534218615e-06,
+      "loss": 0.5282,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8881717760953208,
+      "grad_norm": 0.0713077561850042,
+      "learning_rate": 8.851220220550631e-06,
+      "loss": 0.5399,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8886682387985603,
+      "grad_norm": 0.07335240913305834,
+      "learning_rate": 8.849973318285382e-06,
+      "loss": 0.5419,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8891647015017997,
+      "grad_norm": 0.06929966245122514,
+      "learning_rate": 8.848725827613445e-06,
+      "loss": 0.547,
+      "step": 1791
+    },
+    {
+      "epoch": 0.889661164205039,
+      "grad_norm": 0.06926234064819861,
+      "learning_rate": 8.847477748725473e-06,
+      "loss": 0.5263,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8901576269082785,
+      "grad_norm": 0.07213756111054107,
+      "learning_rate": 8.846229081812217e-06,
+      "loss": 0.5545,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8906540896115179,
+      "grad_norm": 0.07418020782540029,
+      "learning_rate": 8.844979827064517e-06,
+      "loss": 0.4989,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8911505523147574,
+      "grad_norm": 0.07369827208226862,
+      "learning_rate": 8.843729984673296e-06,
+      "loss": 0.511,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8916470150179968,
+      "grad_norm": 0.0731866798675951,
+      "learning_rate": 8.842479554829579e-06,
+      "loss": 0.509,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8921434777212361,
+      "grad_norm": 0.07227353713539185,
+      "learning_rate": 8.841228537724473e-06,
+      "loss": 0.5569,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8926399404244756,
+      "grad_norm": 0.07354933028438398,
+      "learning_rate": 8.839976933549173e-06,
+      "loss": 0.5679,
+      "step": 1798
+    },
+    {
+      "epoch": 0.893136403127715,
+      "grad_norm": 0.07250987685680538,
+      "learning_rate": 8.838724742494966e-06,
+      "loss": 0.4917,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8936328658309545,
+      "grad_norm": 0.07615518221178341,
+      "learning_rate": 8.837471964753234e-06,
+      "loss": 0.5601,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8941293285341939,
+      "grad_norm": 0.07152492656736423,
+      "learning_rate": 8.83621860051544e-06,
+      "loss": 0.5565,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8946257912374332,
+      "grad_norm": 0.07224335571893448,
+      "learning_rate": 8.834964649973144e-06,
+      "loss": 0.5633,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8951222539406727,
+      "grad_norm": 0.07269768082488755,
+      "learning_rate": 8.833710113317988e-06,
+      "loss": 0.5332,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8956187166439121,
+      "grad_norm": 0.07315786298863783,
+      "learning_rate": 8.83245499074171e-06,
+      "loss": 0.4924,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8961151793471516,
+      "grad_norm": 0.06806955055055416,
+      "learning_rate": 8.831199282436136e-06,
+      "loss": 0.5016,
+      "step": 1805
+    },
+    {
+      "epoch": 0.896611642050391,
+      "grad_norm": 0.07498148185530268,
+      "learning_rate": 8.829942988593181e-06,
+      "loss": 0.5304,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8971081047536303,
+      "grad_norm": 0.07445796417367874,
+      "learning_rate": 8.828686109404848e-06,
+      "loss": 0.5823,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8976045674568698,
+      "grad_norm": 0.07278033300821443,
+      "learning_rate": 8.827428645063231e-06,
+      "loss": 0.5328,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8981010301601092,
+      "grad_norm": 0.07392660401254864,
+      "learning_rate": 8.826170595760515e-06,
+      "loss": 0.5118,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8985974928633487,
+      "grad_norm": 0.07188258391302393,
+      "learning_rate": 8.824911961688971e-06,
+      "loss": 0.5152,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8990939555665881,
+      "grad_norm": 0.07434385570738812,
+      "learning_rate": 8.823652743040961e-06,
+      "loss": 0.5152,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8995904182698274,
+      "grad_norm": 0.07354820990692497,
+      "learning_rate": 8.822392940008937e-06,
+      "loss": 0.5516,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9000868809730669,
+      "grad_norm": 0.07453766204422067,
+      "learning_rate": 8.82113255278544e-06,
+      "loss": 0.5199,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9005833436763063,
+      "grad_norm": 0.07462682684863291,
+      "learning_rate": 8.819871581563098e-06,
+      "loss": 0.5774,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9010798063795458,
+      "grad_norm": 0.07218553639506078,
+      "learning_rate": 8.818610026534633e-06,
+      "loss": 0.5245,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9015762690827852,
+      "grad_norm": 0.07027550804968397,
+      "learning_rate": 8.817347887892852e-06,
+      "loss": 0.531,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9020727317860245,
+      "grad_norm": 0.06859338763619163,
+      "learning_rate": 8.816085165830654e-06,
+      "loss": 0.5051,
+      "step": 1817
+    },
+    {
+      "epoch": 0.902569194489264,
+      "grad_norm": 0.07183332754922603,
+      "learning_rate": 8.814821860541024e-06,
+      "loss": 0.5111,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9030656571925034,
+      "grad_norm": 0.07322285377385977,
+      "learning_rate": 8.813557972217038e-06,
+      "loss": 0.5105,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9035621198957429,
+      "grad_norm": 0.07539086388590406,
+      "learning_rate": 8.812293501051862e-06,
+      "loss": 0.5726,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9040585825989822,
+      "grad_norm": 0.07114256778119082,
+      "learning_rate": 8.81102844723875e-06,
+      "loss": 0.5378,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9045550453022216,
+      "grad_norm": 0.07529377230408558,
+      "learning_rate": 8.809762810971044e-06,
+      "loss": 0.5446,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9050515080054611,
+      "grad_norm": 0.0744774028212076,
+      "learning_rate": 8.808496592442178e-06,
+      "loss": 0.5322,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9055479707087005,
+      "grad_norm": 0.07169001776450037,
+      "learning_rate": 8.807229791845673e-06,
+      "loss": 0.5588,
+      "step": 1824
+    },
+    {
+      "epoch": 0.90604443341194,
+      "grad_norm": 0.07010171567219378,
+      "learning_rate": 8.805962409375138e-06,
+      "loss": 0.5316,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9065408961151793,
+      "grad_norm": 0.07048707759482971,
+      "learning_rate": 8.804694445224274e-06,
+      "loss": 0.5199,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9070373588184187,
+      "grad_norm": 0.08194347977828213,
+      "learning_rate": 8.803425899586865e-06,
+      "loss": 0.6072,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9075338215216582,
+      "grad_norm": 0.07117753158187239,
+      "learning_rate": 8.802156772656793e-06,
+      "loss": 0.5572,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9080302842248976,
+      "grad_norm": 0.06968828674618492,
+      "learning_rate": 8.80088706462802e-06,
+      "loss": 0.5047,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9085267469281371,
+      "grad_norm": 0.07433432012948353,
+      "learning_rate": 8.799616775694601e-06,
+      "loss": 0.4883,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9090232096313764,
+      "grad_norm": 0.06853377833470747,
+      "learning_rate": 8.798345906050683e-06,
+      "loss": 0.4933,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9095196723346158,
+      "grad_norm": 0.07198218840214864,
+      "learning_rate": 8.797074455890493e-06,
+      "loss": 0.5172,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9100161350378553,
+      "grad_norm": 0.07144054209959581,
+      "learning_rate": 8.795802425408352e-06,
+      "loss": 0.5285,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9105125977410947,
+      "grad_norm": 0.07241882710808062,
+      "learning_rate": 8.794529814798674e-06,
+      "loss": 0.5499,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9110090604443342,
+      "grad_norm": 0.07128523526894211,
+      "learning_rate": 8.793256624255954e-06,
+      "loss": 0.4895,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9115055231475735,
+      "grad_norm": 0.07431748476081376,
+      "learning_rate": 8.79198285397478e-06,
+      "loss": 0.531,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9120019858508129,
+      "grad_norm": 0.07264275110965394,
+      "learning_rate": 8.790708504149825e-06,
+      "loss": 0.5144,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9124984485540524,
+      "grad_norm": 0.07383415119938352,
+      "learning_rate": 8.789433574975856e-06,
+      "loss": 0.5325,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9129949112572918,
+      "grad_norm": 0.0732949206741093,
+      "learning_rate": 8.788158066647724e-06,
+      "loss": 0.5266,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9134913739605313,
+      "grad_norm": 0.07200921781459274,
+      "learning_rate": 8.786881979360368e-06,
+      "loss": 0.5584,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9139878366637706,
+      "grad_norm": 0.0723153842242348,
+      "learning_rate": 8.78560531330882e-06,
+      "loss": 0.5425,
+      "step": 1841
+    },
+    {
+      "epoch": 0.91448429936701,
+      "grad_norm": 0.07059693624782344,
+      "learning_rate": 8.784328068688199e-06,
+      "loss": 0.5163,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9149807620702495,
+      "grad_norm": 0.07088592353046123,
+      "learning_rate": 8.78305024569371e-06,
+      "loss": 0.5158,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9154772247734889,
+      "grad_norm": 0.07026501902575008,
+      "learning_rate": 8.781771844520646e-06,
+      "loss": 0.4983,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9159736874767284,
+      "grad_norm": 0.07359600617343613,
+      "learning_rate": 8.780492865364392e-06,
+      "loss": 0.5598,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9164701501799677,
+      "grad_norm": 0.0729931146847884,
+      "learning_rate": 8.779213308420418e-06,
+      "loss": 0.5495,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9169666128832071,
+      "grad_norm": 0.07492975377099075,
+      "learning_rate": 8.777933173884288e-06,
+      "loss": 0.5328,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9174630755864466,
+      "grad_norm": 0.07235530445499065,
+      "learning_rate": 8.776652461951644e-06,
+      "loss": 0.4969,
+      "step": 1848
+    },
+    {
+      "epoch": 0.917959538289686,
+      "grad_norm": 0.08357922589860901,
+      "learning_rate": 8.775371172818226e-06,
+      "loss": 0.5883,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9184560009929255,
+      "grad_norm": 0.07047535892914779,
+      "learning_rate": 8.774089306679859e-06,
+      "loss": 0.5016,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9189524636961648,
+      "grad_norm": 0.07455933744307272,
+      "learning_rate": 8.772806863732454e-06,
+      "loss": 0.4909,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9194489263994042,
+      "grad_norm": 0.0718812451067364,
+      "learning_rate": 8.771523844172012e-06,
+      "loss": 0.5203,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9199453891026437,
+      "grad_norm": 0.07798810160053044,
+      "learning_rate": 8.770240248194622e-06,
+      "loss": 0.5235,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9204418518058831,
+      "grad_norm": 0.07655454615769729,
+      "learning_rate": 8.76895607599646e-06,
+      "loss": 0.5527,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9209383145091226,
+      "grad_norm": 0.07714897953371495,
+      "learning_rate": 8.767671327773793e-06,
+      "loss": 0.5689,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9214347772123619,
+      "grad_norm": 0.06788417157684977,
+      "learning_rate": 8.766386003722975e-06,
+      "loss": 0.5284,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9219312399156013,
+      "grad_norm": 0.07465274826817246,
+      "learning_rate": 8.765100104040446e-06,
+      "loss": 0.5368,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9224277026188408,
+      "grad_norm": 0.06873679249158743,
+      "learning_rate": 8.763813628922732e-06,
+      "loss": 0.5331,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9229241653220802,
+      "grad_norm": 0.07176820186024782,
+      "learning_rate": 8.762526578566455e-06,
+      "loss": 0.5211,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9234206280253195,
+      "grad_norm": 0.07147325275088282,
+      "learning_rate": 8.76123895316832e-06,
+      "loss": 0.5231,
+      "step": 1860
+    },
+    {
+      "epoch": 0.923917090728559,
+      "grad_norm": 0.06921441012509941,
+      "learning_rate": 8.759950752925114e-06,
+      "loss": 0.5153,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9244135534317984,
+      "grad_norm": 0.07374828268990942,
+      "learning_rate": 8.758661978033723e-06,
+      "loss": 0.5341,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9249100161350379,
+      "grad_norm": 0.07350976624933628,
+      "learning_rate": 8.757372628691115e-06,
+      "loss": 0.5232,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9254064788382773,
+      "grad_norm": 0.07539026737326146,
+      "learning_rate": 8.756082705094344e-06,
+      "loss": 0.5454,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9259029415415166,
+      "grad_norm": 0.07085472578928324,
+      "learning_rate": 8.754792207440557e-06,
+      "loss": 0.5424,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9263994042447561,
+      "grad_norm": 0.07878932254764395,
+      "learning_rate": 8.753501135926985e-06,
+      "loss": 0.5218,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9268958669479955,
+      "grad_norm": 0.06901455357252793,
+      "learning_rate": 8.752209490750947e-06,
+      "loss": 0.5258,
+      "step": 1867
+    },
+    {
+      "epoch": 0.927392329651235,
+      "grad_norm": 0.0713378413072867,
+      "learning_rate": 8.750917272109849e-06,
+      "loss": 0.5579,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9278887923544744,
+      "grad_norm": 0.07256870817533925,
+      "learning_rate": 8.749624480201188e-06,
+      "loss": 0.5068,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9283852550577137,
+      "grad_norm": 0.07103653504023608,
+      "learning_rate": 8.748331115222546e-06,
+      "loss": 0.515,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9288817177609532,
+      "grad_norm": 0.0738251937073301,
+      "learning_rate": 8.747037177371593e-06,
+      "loss": 0.5185,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9293781804641926,
+      "grad_norm": 0.07390345582909139,
+      "learning_rate": 8.745742666846088e-06,
+      "loss": 0.5159,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9298746431674321,
+      "grad_norm": 0.07285618531105278,
+      "learning_rate": 8.744447583843874e-06,
+      "loss": 0.5519,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9303711058706715,
+      "grad_norm": 0.06928191759599445,
+      "learning_rate": 8.743151928562883e-06,
+      "loss": 0.5162,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9308675685739108,
+      "grad_norm": 0.07672934530948829,
+      "learning_rate": 8.741855701201138e-06,
+      "loss": 0.5844,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9313640312771503,
+      "grad_norm": 0.07565381558885703,
+      "learning_rate": 8.740558901956745e-06,
+      "loss": 0.512,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9318604939803897,
+      "grad_norm": 0.07428988447726668,
+      "learning_rate": 8.739261531027899e-06,
+      "loss": 0.4979,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9323569566836292,
+      "grad_norm": 0.07272692110414847,
+      "learning_rate": 8.737963588612882e-06,
+      "loss": 0.5256,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9328534193868686,
+      "grad_norm": 0.07394818427713572,
+      "learning_rate": 8.736665074910064e-06,
+      "loss": 0.5442,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9333498820901079,
+      "grad_norm": 0.07489595106458445,
+      "learning_rate": 8.735365990117904e-06,
+      "loss": 0.5009,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9338463447933474,
+      "grad_norm": 0.07379647498011625,
+      "learning_rate": 8.734066334434944e-06,
+      "loss": 0.4879,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9343428074965868,
+      "grad_norm": 0.07110250204025945,
+      "learning_rate": 8.732766108059814e-06,
+      "loss": 0.5298,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9348392701998263,
+      "grad_norm": 0.07081073435018918,
+      "learning_rate": 8.731465311191237e-06,
+      "loss": 0.5221,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9353357329030657,
+      "grad_norm": 0.07000873579274325,
+      "learning_rate": 8.730163944028013e-06,
+      "loss": 0.5187,
+      "step": 1884
+    },
+    {
+      "epoch": 0.935832195606305,
+      "grad_norm": 0.07240937120624676,
+      "learning_rate": 8.728862006769043e-06,
+      "loss": 0.5041,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9363286583095445,
+      "grad_norm": 0.0746119118897645,
+      "learning_rate": 8.7275594996133e-06,
+      "loss": 0.5379,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9368251210127839,
+      "grad_norm": 0.07015740462906615,
+      "learning_rate": 8.726256422759857e-06,
+      "loss": 0.5386,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9373215837160234,
+      "grad_norm": 0.06777154576680991,
+      "learning_rate": 8.724952776407864e-06,
+      "loss": 0.4789,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9378180464192628,
+      "grad_norm": 0.07200557741488726,
+      "learning_rate": 8.723648560756565e-06,
+      "loss": 0.5483,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9383145091225021,
+      "grad_norm": 0.07146897275049505,
+      "learning_rate": 8.722343776005288e-06,
+      "loss": 0.4842,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9388109718257416,
+      "grad_norm": 0.07349104098205175,
+      "learning_rate": 8.721038422353447e-06,
+      "loss": 0.5368,
+      "step": 1891
+    },
+    {
+      "epoch": 0.939307434528981,
+      "grad_norm": 0.07452676274413209,
+      "learning_rate": 8.719732500000547e-06,
+      "loss": 0.4931,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9398038972322205,
+      "grad_norm": 0.07592669585842883,
+      "learning_rate": 8.718426009146174e-06,
+      "loss": 0.5249,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9403003599354598,
+      "grad_norm": 0.06953574446858567,
+      "learning_rate": 8.717118949990006e-06,
+      "loss": 0.4989,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9407968226386992,
+      "grad_norm": 0.06939013011968077,
+      "learning_rate": 8.715811322731808e-06,
+      "loss": 0.5243,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9412932853419387,
+      "grad_norm": 0.07460509153996832,
+      "learning_rate": 8.714503127571425e-06,
+      "loss": 0.55,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9417897480451781,
+      "grad_norm": 0.07492121354167032,
+      "learning_rate": 8.713194364708799e-06,
+      "loss": 0.5162,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9422862107484176,
+      "grad_norm": 0.07201694827718819,
+      "learning_rate": 8.71188503434395e-06,
+      "loss": 0.5435,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9427826734516569,
+      "grad_norm": 0.0698106849011963,
+      "learning_rate": 8.710575136676988e-06,
+      "loss": 0.5325,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9432791361548963,
+      "grad_norm": 0.07457961112892264,
+      "learning_rate": 8.709264671908113e-06,
+      "loss": 0.5269,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9437755988581358,
+      "grad_norm": 0.07363246085888084,
+      "learning_rate": 8.707953640237605e-06,
+      "loss": 0.5467,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9442720615613752,
+      "grad_norm": 0.07217700611369388,
+      "learning_rate": 8.706642041865836e-06,
+      "loss": 0.5387,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9447685242646147,
+      "grad_norm": 0.0692734553474916,
+      "learning_rate": 8.705329876993262e-06,
+      "loss": 0.506,
+      "step": 1903
+    },
+    {
+      "epoch": 0.945264986967854,
+      "grad_norm": 0.07196674390301115,
+      "learning_rate": 8.704017145820427e-06,
+      "loss": 0.549,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9457614496710934,
+      "grad_norm": 0.07512068711896185,
+      "learning_rate": 8.70270384854796e-06,
+      "loss": 0.5663,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9462579123743329,
+      "grad_norm": 0.07421738591859378,
+      "learning_rate": 8.701389985376578e-06,
+      "loss": 0.5622,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9467543750775723,
+      "grad_norm": 0.06964931407494165,
+      "learning_rate": 8.700075556507085e-06,
+      "loss": 0.4877,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9472508377808118,
+      "grad_norm": 0.07205777339082249,
+      "learning_rate": 8.698760562140369e-06,
+      "loss": 0.5148,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9477473004840511,
+      "grad_norm": 0.06972774499188475,
+      "learning_rate": 8.697445002477408e-06,
+      "loss": 0.5172,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9482437631872905,
+      "grad_norm": 0.0730143838918345,
+      "learning_rate": 8.696128877719258e-06,
+      "loss": 0.5825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.94874022589053,
+      "grad_norm": 0.06930828954474016,
+      "learning_rate": 8.694812188067077e-06,
+      "loss": 0.5488,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9492366885937694,
+      "grad_norm": 0.06817526079016235,
+      "learning_rate": 8.693494933722091e-06,
+      "loss": 0.5137,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9497331512970089,
+      "grad_norm": 0.07130049301765591,
+      "learning_rate": 8.692177114885626e-06,
+      "loss": 0.5108,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9502296140002482,
+      "grad_norm": 0.06952714928758798,
+      "learning_rate": 8.69085873175909e-06,
+      "loss": 0.5593,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9507260767034876,
+      "grad_norm": 0.07028375429072933,
+      "learning_rate": 8.689539784543975e-06,
+      "loss": 0.5354,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9512225394067271,
+      "grad_norm": 0.06774708549780749,
+      "learning_rate": 8.68822027344186e-06,
+      "loss": 0.4871,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9517190021099665,
+      "grad_norm": 0.07171222831888815,
+      "learning_rate": 8.686900198654413e-06,
+      "loss": 0.5434,
+      "step": 1917
+    },
+    {
+      "epoch": 0.952215464813206,
+      "grad_norm": 0.07452854923353851,
+      "learning_rate": 8.685579560383386e-06,
+      "loss": 0.5635,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9527119275164453,
+      "grad_norm": 0.07754262739318235,
+      "learning_rate": 8.684258358830617e-06,
+      "loss": 0.5291,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9532083902196847,
+      "grad_norm": 0.07210569027672796,
+      "learning_rate": 8.682936594198029e-06,
+      "loss": 0.5101,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9537048529229242,
+      "grad_norm": 0.07585209431668576,
+      "learning_rate": 8.681614266687634e-06,
+      "loss": 0.5186,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9542013156261636,
+      "grad_norm": 0.07508056085158579,
+      "learning_rate": 8.680291376501531e-06,
+      "loss": 0.5519,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9546977783294031,
+      "grad_norm": 0.06911989074015167,
+      "learning_rate": 8.678967923841897e-06,
+      "loss": 0.5203,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9551942410326424,
+      "grad_norm": 0.07000431990955368,
+      "learning_rate": 8.677643908911007e-06,
+      "loss": 0.5072,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9556907037358818,
+      "grad_norm": 0.0706651754552142,
+      "learning_rate": 8.67631933191121e-06,
+      "loss": 0.5236,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9561871664391213,
+      "grad_norm": 0.07233656910668078,
+      "learning_rate": 8.674994193044947e-06,
+      "loss": 0.5408,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9566836291423607,
+      "grad_norm": 0.07165155095334301,
+      "learning_rate": 8.673668492514748e-06,
+      "loss": 0.5361,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9571800918456,
+      "grad_norm": 0.07104402132471122,
+      "learning_rate": 8.672342230523222e-06,
+      "loss": 0.5252,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9576765545488395,
+      "grad_norm": 0.07062543323954087,
+      "learning_rate": 8.671015407273067e-06,
+      "loss": 0.5412,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9581730172520789,
+      "grad_norm": 0.06926853707239683,
+      "learning_rate": 8.669688022967068e-06,
+      "loss": 0.5511,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9586694799553184,
+      "grad_norm": 0.07159547068682602,
+      "learning_rate": 8.668360077808093e-06,
+      "loss": 0.5512,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9591659426585578,
+      "grad_norm": 0.07115586840594067,
+      "learning_rate": 8.667031571999098e-06,
+      "loss": 0.5133,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9596624053617971,
+      "grad_norm": 0.07200342418348525,
+      "learning_rate": 8.665702505743125e-06,
+      "loss": 0.5322,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9601588680650366,
+      "grad_norm": 0.06825765717088793,
+      "learning_rate": 8.664372879243297e-06,
+      "loss": 0.5292,
+      "step": 1934
+    },
+    {
+      "epoch": 0.960655330768276,
+      "grad_norm": 0.07283728005005825,
+      "learning_rate": 8.66304269270283e-06,
+      "loss": 0.5449,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9611517934715155,
+      "grad_norm": 0.07304900897173891,
+      "learning_rate": 8.661711946325018e-06,
+      "loss": 0.5512,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9616482561747549,
+      "grad_norm": 0.06853811178509077,
+      "learning_rate": 8.660380640313247e-06,
+      "loss": 0.4965,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9621447188779942,
+      "grad_norm": 0.0751622154001969,
+      "learning_rate": 8.659048774870986e-06,
+      "loss": 0.5175,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9626411815812337,
+      "grad_norm": 0.07096935969983968,
+      "learning_rate": 8.657716350201786e-06,
+      "loss": 0.5653,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9631376442844731,
+      "grad_norm": 0.07358237864616872,
+      "learning_rate": 8.656383366509292e-06,
+      "loss": 0.5213,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9636341069877126,
+      "grad_norm": 0.06865112950199855,
+      "learning_rate": 8.655049823997222e-06,
+      "loss": 0.5004,
+      "step": 1941
+    },
+    {
+      "epoch": 0.964130569690952,
+      "grad_norm": 0.06913087429662826,
+      "learning_rate": 8.653715722869394e-06,
+      "loss": 0.5091,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9646270323941913,
+      "grad_norm": 0.07334073189778997,
+      "learning_rate": 8.652381063329697e-06,
+      "loss": 0.5273,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9651234950974308,
+      "grad_norm": 0.07154596833896756,
+      "learning_rate": 8.65104584558212e-06,
+      "loss": 0.5221,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9656199578006702,
+      "grad_norm": 0.0669766137567477,
+      "learning_rate": 8.649710069830723e-06,
+      "loss": 0.4856,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9661164205039097,
+      "grad_norm": 0.06695881028521661,
+      "learning_rate": 8.648373736279662e-06,
+      "loss": 0.5006,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9666128832071491,
+      "grad_norm": 0.07056372764666827,
+      "learning_rate": 8.647036845133171e-06,
+      "loss": 0.5038,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9671093459103884,
+      "grad_norm": 0.06721853167954736,
+      "learning_rate": 8.645699396595574e-06,
+      "loss": 0.5192,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9676058086136279,
+      "grad_norm": 0.07297678022854343,
+      "learning_rate": 8.644361390871281e-06,
+      "loss": 0.5508,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9681022713168673,
+      "grad_norm": 0.07278377444176168,
+      "learning_rate": 8.64302282816478e-06,
+      "loss": 0.5509,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9685987340201068,
+      "grad_norm": 0.06846087625519698,
+      "learning_rate": 8.641683708680653e-06,
+      "loss": 0.4928,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9690951967233462,
+      "grad_norm": 0.06533270866026043,
+      "learning_rate": 8.64034403262356e-06,
+      "loss": 0.485,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9695916594265855,
+      "grad_norm": 0.07265110323601376,
+      "learning_rate": 8.63900380019825e-06,
+      "loss": 0.5153,
+      "step": 1953
+    },
+    {
+      "epoch": 0.970088122129825,
+      "grad_norm": 0.07263717621963031,
+      "learning_rate": 8.637663011609556e-06,
+      "loss": 0.5068,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9705845848330644,
+      "grad_norm": 0.06931969752092298,
+      "learning_rate": 8.636321667062398e-06,
+      "loss": 0.4819,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9710810475363039,
+      "grad_norm": 0.07329123137887754,
+      "learning_rate": 8.634979766761775e-06,
+      "loss": 0.5111,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9715775102395433,
+      "grad_norm": 0.0776376074481898,
+      "learning_rate": 8.633637310912777e-06,
+      "loss": 0.5299,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9720739729427826,
+      "grad_norm": 0.07357164179881222,
+      "learning_rate": 8.632294299720578e-06,
+      "loss": 0.5242,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9725704356460221,
+      "grad_norm": 0.06918945191060966,
+      "learning_rate": 8.630950733390434e-06,
+      "loss": 0.5112,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9730668983492615,
+      "grad_norm": 0.07259280346654948,
+      "learning_rate": 8.62960661212769e-06,
+      "loss": 0.5416,
+      "step": 1960
+    },
+    {
+      "epoch": 0.973563361052501,
+      "grad_norm": 0.0710695063103676,
+      "learning_rate": 8.628261936137769e-06,
+      "loss": 0.543,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9740598237557403,
+      "grad_norm": 0.07278661425053735,
+      "learning_rate": 8.626916705626186e-06,
+      "loss": 0.5096,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9745562864589797,
+      "grad_norm": 0.07217270046199414,
+      "learning_rate": 8.62557092079854e-06,
+      "loss": 0.5251,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9750527491622192,
+      "grad_norm": 0.07443189862074188,
+      "learning_rate": 8.62422458186051e-06,
+      "loss": 0.5263,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9755492118654586,
+      "grad_norm": 0.0720078190705775,
+      "learning_rate": 8.62287768901786e-06,
+      "loss": 0.5163,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9760456745686981,
+      "grad_norm": 0.07216624095705255,
+      "learning_rate": 8.621530242476446e-06,
+      "loss": 0.5491,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9765421372719374,
+      "grad_norm": 0.07025665529312518,
+      "learning_rate": 8.620182242442202e-06,
+      "loss": 0.5003,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9770385999751768,
+      "grad_norm": 0.07990291482485633,
+      "learning_rate": 8.618833689121147e-06,
+      "loss": 0.5196,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9775350626784163,
+      "grad_norm": 0.07182987590456455,
+      "learning_rate": 8.617484582719384e-06,
+      "loss": 0.4937,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9780315253816557,
+      "grad_norm": 0.07065785829898263,
+      "learning_rate": 8.616134923443107e-06,
+      "loss": 0.5059,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9785279880848952,
+      "grad_norm": 0.07052696820744597,
+      "learning_rate": 8.614784711498586e-06,
+      "loss": 0.5154,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9790244507881345,
+      "grad_norm": 0.07208828528634863,
+      "learning_rate": 8.61343394709218e-06,
+      "loss": 0.5148,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9795209134913739,
+      "grad_norm": 0.07218283214224182,
+      "learning_rate": 8.612082630430333e-06,
+      "loss": 0.5254,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9800173761946134,
+      "grad_norm": 0.0700016743353793,
+      "learning_rate": 8.610730761719573e-06,
+      "loss": 0.524,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9805138388978528,
+      "grad_norm": 0.06872083929445431,
+      "learning_rate": 8.609378341166508e-06,
+      "loss": 0.5152,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9810103016010923,
+      "grad_norm": 0.07548071827402791,
+      "learning_rate": 8.608025368977834e-06,
+      "loss": 0.5443,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9815067643043316,
+      "grad_norm": 0.07492045099973432,
+      "learning_rate": 8.606671845360334e-06,
+      "loss": 0.5235,
+      "step": 1977
+    },
+    {
+      "epoch": 0.982003227007571,
+      "grad_norm": 0.0725796525707599,
+      "learning_rate": 8.605317770520871e-06,
+      "loss": 0.5093,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9824996897108105,
+      "grad_norm": 0.0716698247485905,
+      "learning_rate": 8.603963144666393e-06,
+      "loss": 0.537,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9829961524140499,
+      "grad_norm": 0.07084273949601609,
+      "learning_rate": 8.602607968003935e-06,
+      "loss": 0.5064,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9834926151172894,
+      "grad_norm": 0.07171289777818614,
+      "learning_rate": 8.601252240740611e-06,
+      "loss": 0.5377,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9839890778205287,
+      "grad_norm": 0.07193685972435261,
+      "learning_rate": 8.599895963083627e-06,
+      "loss": 0.5365,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9844855405237681,
+      "grad_norm": 0.0716912144267546,
+      "learning_rate": 8.598539135240263e-06,
+      "loss": 0.501,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9849820032270076,
+      "grad_norm": 0.07125142000863152,
+      "learning_rate": 8.597181757417889e-06,
+      "loss": 0.5495,
+      "step": 1984
+    },
+    {
+      "epoch": 0.985478465930247,
+      "grad_norm": 0.07060360439051618,
+      "learning_rate": 8.59582382982396e-06,
+      "loss": 0.5344,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9859749286334865,
+      "grad_norm": 0.07445172176295355,
+      "learning_rate": 8.594465352666015e-06,
+      "loss": 0.57,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9864713913367258,
+      "grad_norm": 0.07371532452638814,
+      "learning_rate": 8.593106326151672e-06,
+      "loss": 0.5311,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9869678540399652,
+      "grad_norm": 0.0696119232310848,
+      "learning_rate": 8.591746750488639e-06,
+      "loss": 0.4858,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9874643167432047,
+      "grad_norm": 0.07048082002862621,
+      "learning_rate": 8.590386625884703e-06,
+      "loss": 0.5171,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9879607794464441,
+      "grad_norm": 0.07171447996292699,
+      "learning_rate": 8.58902595254774e-06,
+      "loss": 0.5028,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9884572421496836,
+      "grad_norm": 0.07039308554220271,
+      "learning_rate": 8.587664730685707e-06,
+      "loss": 0.5248,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9889537048529229,
+      "grad_norm": 0.07283522641859247,
+      "learning_rate": 8.586302960506643e-06,
+      "loss": 0.5451,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9894501675561623,
+      "grad_norm": 0.07227866851044816,
+      "learning_rate": 8.584940642218672e-06,
+      "loss": 0.5738,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9899466302594018,
+      "grad_norm": 0.07329916195849069,
+      "learning_rate": 8.583577776030005e-06,
+      "loss": 0.5331,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9904430929626412,
+      "grad_norm": 0.07113529640045456,
+      "learning_rate": 8.582214362148932e-06,
+      "loss": 0.4991,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9909395556658807,
+      "grad_norm": 0.07416831869367692,
+      "learning_rate": 8.580850400783833e-06,
+      "loss": 0.5589,
+      "step": 1996
+    },
+    {
+      "epoch": 0.99143601836912,
+      "grad_norm": 0.0719901887328746,
+      "learning_rate": 8.579485892143163e-06,
+      "loss": 0.535,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9919324810723594,
+      "grad_norm": 0.07290232210396123,
+      "learning_rate": 8.578120836435467e-06,
+      "loss": 0.5309,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9924289437755989,
+      "grad_norm": 0.07268201060562912,
+      "learning_rate": 8.576755233869372e-06,
+      "loss": 0.5039,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9929254064788383,
+      "grad_norm": 0.069235314012733,
+      "learning_rate": 8.57538908465359e-06,
+      "loss": 0.4977,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9934218691820776,
+      "grad_norm": 0.07348961416704078,
+      "learning_rate": 8.574022388996913e-06,
+      "loss": 0.5569,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9939183318853171,
+      "grad_norm": 0.0729022760562524,
+      "learning_rate": 8.572655147108217e-06,
+      "loss": 0.5113,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9944147945885565,
+      "grad_norm": 0.07172086453966273,
+      "learning_rate": 8.571287359196466e-06,
+      "loss": 0.5105,
+      "step": 2003
+    },
+    {
+      "epoch": 0.994911257291796,
+      "grad_norm": 0.06853033454269805,
+      "learning_rate": 8.569919025470704e-06,
+      "loss": 0.4886,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9954077199950354,
+      "grad_norm": 0.06903654454945618,
+      "learning_rate": 8.568550146140056e-06,
+      "loss": 0.5157,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9959041826982747,
+      "grad_norm": 0.07258062579104714,
+      "learning_rate": 8.567180721413736e-06,
+      "loss": 0.5318,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9964006454015142,
+      "grad_norm": 0.07499795243636775,
+      "learning_rate": 8.56581075150104e-06,
+      "loss": 0.5757,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9968971081047536,
+      "grad_norm": 0.07104244022202039,
+      "learning_rate": 8.564440236611344e-06,
+      "loss": 0.4889,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9973935708079931,
+      "grad_norm": 0.07249943092428517,
+      "learning_rate": 8.563069176954108e-06,
+      "loss": 0.552,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9978900335112325,
+      "grad_norm": 0.07144120704844,
+      "learning_rate": 8.561697572738878e-06,
+      "loss": 0.5075,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9983864962144718,
+      "grad_norm": 0.0742860541852747,
+      "learning_rate": 8.560325424175282e-06,
+      "loss": 0.501,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9988829589177113,
+      "grad_norm": 0.06847104594866535,
+      "learning_rate": 8.558952731473031e-06,
+      "loss": 0.5224,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9993794216209507,
+      "grad_norm": 0.07003316710867157,
+      "learning_rate": 8.557579494841918e-06,
+      "loss": 0.5415,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9998758843241902,
+      "grad_norm": 0.0748671769203952,
+      "learning_rate": 8.55620571449182e-06,
+      "loss": 0.6028,
+      "step": 2014
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0748671769203952,
+      "learning_rate": 8.5548313906327e-06,
+      "loss": 0.1345,
+      "step": 2015
+    },
+    {
+      "epoch": 1.0003723470274295,
+      "grad_norm": 0.07559504504545322,
+      "learning_rate": 8.553456523474596e-06,
+      "loss": 0.4075,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0003723470274295,
+      "eval_loss": 0.5260358452796936,
+      "eval_runtime": 258.7864,
+      "eval_samples_per_second": 117.29,
+      "eval_steps_per_second": 14.665,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0004964627032393,
+      "grad_norm": 0.09547152201342973,
+      "learning_rate": 8.55208111322764e-06,
+      "loss": 0.4972,
+      "step": 2017
+    },
+    {
+      "epoch": 1.0009929254064789,
+      "grad_norm": 0.08024846390865537,
+      "learning_rate": 8.550705160102037e-06,
+      "loss": 0.4913,
+      "step": 2018
+    },
+    {
+      "epoch": 1.0014893881097182,
+      "grad_norm": 0.07164772320273062,
+      "learning_rate": 8.549328664308084e-06,
+      "loss": 0.4719,
+      "step": 2019
+    },
+    {
+      "epoch": 1.0019858508129578,
+      "grad_norm": 0.0791799829849966,
+      "learning_rate": 8.547951626056152e-06,
+      "loss": 0.4754,
+      "step": 2020
+    },
+    {
+      "epoch": 1.002482313516197,
+      "grad_norm": 0.08827479564010493,
+      "learning_rate": 8.546574045556702e-06,
+      "loss": 0.5106,
+      "step": 2021
+    },
+    {
+      "epoch": 1.0029787762194364,
+      "grad_norm": 0.08009835838102199,
+      "learning_rate": 8.545195923020273e-06,
+      "loss": 0.4828,
+      "step": 2022
+    },
+    {
+      "epoch": 1.003475238922676,
+      "grad_norm": 0.07696883614047778,
+      "learning_rate": 8.543817258657493e-06,
+      "loss": 0.4869,
+      "step": 2023
+    },
+    {
+      "epoch": 1.0039717016259153,
+      "grad_norm": 0.08329873785631868,
+      "learning_rate": 8.542438052679063e-06,
+      "loss": 0.5088,
+      "step": 2024
+    },
+    {
+      "epoch": 1.0044681643291549,
+      "grad_norm": 0.08107257398321653,
+      "learning_rate": 8.541058305295777e-06,
+      "loss": 0.5295,
+      "step": 2025
+    },
+    {
+      "epoch": 1.0049646270323942,
+      "grad_norm": 0.07242648419214391,
+      "learning_rate": 8.539678016718505e-06,
+      "loss": 0.4733,
+      "step": 2026
+    },
+    {
+      "epoch": 1.0054610897356335,
+      "grad_norm": 0.07294681808712751,
+      "learning_rate": 8.538297187158202e-06,
+      "loss": 0.4937,
+      "step": 2027
+    },
+    {
+      "epoch": 1.005957552438873,
+      "grad_norm": 0.07774560057953825,
+      "learning_rate": 8.536915816825906e-06,
+      "loss": 0.4661,
+      "step": 2028
+    },
+    {
+      "epoch": 1.0064540151421124,
+      "grad_norm": 0.07321460359687063,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.4938,
+      "step": 2029
+    },
+    {
+      "epoch": 1.006950477845352,
+      "grad_norm": 0.07353088039144523,
+      "learning_rate": 8.534151454689901e-06,
+      "loss": 0.4904,
+      "step": 2030
+    },
+    {
+      "epoch": 1.0074469405485913,
+      "grad_norm": 0.07221853677357111,
+      "learning_rate": 8.532768463308679e-06,
+      "loss": 0.4859,
+      "step": 2031
+    },
+    {
+      "epoch": 1.0079434032518306,
+      "grad_norm": 0.07626837115175593,
+      "learning_rate": 8.531384932000442e-06,
+      "loss": 0.5063,
+      "step": 2032
+    },
+    {
+      "epoch": 1.0084398659550702,
+      "grad_norm": 0.07611380745129669,
+      "learning_rate": 8.530000860976639e-06,
+      "loss": 0.5051,
+      "step": 2033
+    },
+    {
+      "epoch": 1.0089363286583095,
+      "grad_norm": 0.07174543909991783,
+      "learning_rate": 8.528616250448805e-06,
+      "loss": 0.4955,
+      "step": 2034
+    },
+    {
+      "epoch": 1.009432791361549,
+      "grad_norm": 0.07442453470054473,
+      "learning_rate": 8.527231100628553e-06,
+      "loss": 0.4755,
+      "step": 2035
+    },
+    {
+      "epoch": 1.0099292540647884,
+      "grad_norm": 0.0724289281191574,
+      "learning_rate": 8.525845411727581e-06,
+      "loss": 0.4908,
+      "step": 2036
+    },
+    {
+      "epoch": 1.0104257167680277,
+      "grad_norm": 0.07872472635051855,
+      "learning_rate": 8.524459183957673e-06,
+      "loss": 0.5112,
+      "step": 2037
+    },
+    {
+      "epoch": 1.0109221794712673,
+      "grad_norm": 0.06971999213309993,
+      "learning_rate": 8.523072417530686e-06,
+      "loss": 0.4562,
+      "step": 2038
+    },
+    {
+      "epoch": 1.0114186421745066,
+      "grad_norm": 0.07312421396196309,
+      "learning_rate": 8.52168511265857e-06,
+      "loss": 0.5046,
+      "step": 2039
+    },
+    {
+      "epoch": 1.0119151048777462,
+      "grad_norm": 0.07995669672272424,
+      "learning_rate": 8.52029726955335e-06,
+      "loss": 0.5058,
+      "step": 2040
+    },
+    {
+      "epoch": 1.0124115675809855,
+      "grad_norm": 0.0752474945014717,
+      "learning_rate": 8.518908888427137e-06,
+      "loss": 0.4896,
+      "step": 2041
+    },
+    {
+      "epoch": 1.0129080302842248,
+      "grad_norm": 0.07367661958562566,
+      "learning_rate": 8.51751996949212e-06,
+      "loss": 0.5159,
+      "step": 2042
+    },
+    {
+      "epoch": 1.0134044929874644,
+      "grad_norm": 0.06979249530045643,
+      "learning_rate": 8.516130512960576e-06,
+      "loss": 0.4974,
+      "step": 2043
+    },
+    {
+      "epoch": 1.0139009556907037,
+      "grad_norm": 0.07119884891727783,
+      "learning_rate": 8.51474051904486e-06,
+      "loss": 0.4919,
+      "step": 2044
+    },
+    {
+      "epoch": 1.0143974183939433,
+      "grad_norm": 0.07609361710536897,
+      "learning_rate": 8.513349987957411e-06,
+      "loss": 0.4723,
+      "step": 2045
+    },
+    {
+      "epoch": 1.0148938810971826,
+      "grad_norm": 0.06703588203654677,
+      "learning_rate": 8.511958919910748e-06,
+      "loss": 0.4399,
+      "step": 2046
+    },
+    {
+      "epoch": 1.015390343800422,
+      "grad_norm": 0.07309725028188202,
+      "learning_rate": 8.510567315117472e-06,
+      "loss": 0.4804,
+      "step": 2047
+    },
+    {
+      "epoch": 1.0158868065036615,
+      "grad_norm": 0.068314468194369,
+      "learning_rate": 8.509175173790271e-06,
+      "loss": 0.4567,
+      "step": 2048
+    },
+    {
+      "epoch": 1.0163832692069008,
+      "grad_norm": 0.07906321230711548,
+      "learning_rate": 8.507782496141911e-06,
+      "loss": 0.5249,
+      "step": 2049
+    },
+    {
+      "epoch": 1.0168797319101404,
+      "grad_norm": 0.07035411917162568,
+      "learning_rate": 8.506389282385242e-06,
+      "loss": 0.4569,
+      "step": 2050
+    },
+    {
+      "epoch": 1.0173761946133797,
+      "grad_norm": 0.07087902433432516,
+      "learning_rate": 8.504995532733187e-06,
+      "loss": 0.4752,
+      "step": 2051
+    },
+    {
+      "epoch": 1.017872657316619,
+      "grad_norm": 0.09428637704107151,
+      "learning_rate": 8.503601247398765e-06,
+      "loss": 0.5069,
+      "step": 2052
+    },
+    {
+      "epoch": 1.0183691200198586,
+      "grad_norm": 0.07313205171308455,
+      "learning_rate": 8.502206426595069e-06,
+      "loss": 0.48,
+      "step": 2053
+    },
+    {
+      "epoch": 1.018865582723098,
+      "grad_norm": 0.07444256500494913,
+      "learning_rate": 8.500811070535271e-06,
+      "loss": 0.505,
+      "step": 2054
+    },
+    {
+      "epoch": 1.0193620454263372,
+      "grad_norm": 0.07187260748173696,
+      "learning_rate": 8.499415179432635e-06,
+      "loss": 0.4878,
+      "step": 2055
+    },
+    {
+      "epoch": 1.0198585081295768,
+      "grad_norm": 0.07521015775447627,
+      "learning_rate": 8.4980187535005e-06,
+      "loss": 0.5347,
+      "step": 2056
+    },
+    {
+      "epoch": 1.0203549708328161,
+      "grad_norm": 0.07519459961664571,
+      "learning_rate": 8.49662179295228e-06,
+      "loss": 0.4939,
+      "step": 2057
+    },
+    {
+      "epoch": 1.0208514335360557,
+      "grad_norm": 0.07478071090183817,
+      "learning_rate": 8.495224298001487e-06,
+      "loss": 0.4952,
+      "step": 2058
+    },
+    {
+      "epoch": 1.021347896239295,
+      "grad_norm": 0.0730378035645826,
+      "learning_rate": 8.4938262688617e-06,
+      "loss": 0.4901,
+      "step": 2059
+    },
+    {
+      "epoch": 1.0218443589425343,
+      "grad_norm": 0.07650825318167469,
+      "learning_rate": 8.492427705746587e-06,
+      "loss": 0.4727,
+      "step": 2060
+    },
+    {
+      "epoch": 1.022340821645774,
+      "grad_norm": 0.07265259240781462,
+      "learning_rate": 8.491028608869895e-06,
+      "loss": 0.4697,
+      "step": 2061
+    },
+    {
+      "epoch": 1.0228372843490132,
+      "grad_norm": 0.07192624745986412,
+      "learning_rate": 8.489628978445456e-06,
+      "loss": 0.475,
+      "step": 2062
+    },
+    {
+      "epoch": 1.0233337470522528,
+      "grad_norm": 0.07911963050617478,
+      "learning_rate": 8.488228814687178e-06,
+      "loss": 0.5122,
+      "step": 2063
+    },
+    {
+      "epoch": 1.023830209755492,
+      "grad_norm": 0.0741852579227699,
+      "learning_rate": 8.486828117809057e-06,
+      "loss": 0.5003,
+      "step": 2064
+    },
+    {
+      "epoch": 1.0243266724587314,
+      "grad_norm": 0.07311097676007818,
+      "learning_rate": 8.485426888025166e-06,
+      "loss": 0.5349,
+      "step": 2065
+    },
+    {
+      "epoch": 1.024823135161971,
+      "grad_norm": 0.07532799313479474,
+      "learning_rate": 8.484025125549658e-06,
+      "loss": 0.4889,
+      "step": 2066
+    },
+    {
+      "epoch": 1.0253195978652103,
+      "grad_norm": 0.08160693762566441,
+      "learning_rate": 8.482622830596772e-06,
+      "loss": 0.5053,
+      "step": 2067
+    },
+    {
+      "epoch": 1.0258160605684499,
+      "grad_norm": 0.07068127338378674,
+      "learning_rate": 8.481220003380826e-06,
+      "loss": 0.5036,
+      "step": 2068
+    },
+    {
+      "epoch": 1.0263125232716892,
+      "grad_norm": 0.07424244914892054,
+      "learning_rate": 8.479816644116218e-06,
+      "loss": 0.4704,
+      "step": 2069
+    },
+    {
+      "epoch": 1.0268089859749285,
+      "grad_norm": 0.07315128718542926,
+      "learning_rate": 8.478412753017433e-06,
+      "loss": 0.4754,
+      "step": 2070
+    },
+    {
+      "epoch": 1.027305448678168,
+      "grad_norm": 0.07772853008795041,
+      "learning_rate": 8.47700833029903e-06,
+      "loss": 0.5077,
+      "step": 2071
+    },
+    {
+      "epoch": 1.0278019113814074,
+      "grad_norm": 0.07141108432687718,
+      "learning_rate": 8.475603376175654e-06,
+      "loss": 0.4918,
+      "step": 2072
+    },
+    {
+      "epoch": 1.028298374084647,
+      "grad_norm": 0.08059188331356318,
+      "learning_rate": 8.474197890862028e-06,
+      "loss": 0.5103,
+      "step": 2073
+    },
+    {
+      "epoch": 1.0287948367878863,
+      "grad_norm": 0.07290453191714018,
+      "learning_rate": 8.472791874572958e-06,
+      "loss": 0.4811,
+      "step": 2074
+    },
+    {
+      "epoch": 1.0292912994911256,
+      "grad_norm": 0.07126996934228423,
+      "learning_rate": 8.471385327523333e-06,
+      "loss": 0.4988,
+      "step": 2075
+    },
+    {
+      "epoch": 1.0297877621943652,
+      "grad_norm": 0.07682736893829598,
+      "learning_rate": 8.469978249928122e-06,
+      "loss": 0.4967,
+      "step": 2076
+    },
+    {
+      "epoch": 1.0302842248976045,
+      "grad_norm": 0.07159491409294527,
+      "learning_rate": 8.46857064200237e-06,
+      "loss": 0.5015,
+      "step": 2077
+    },
+    {
+      "epoch": 1.030780687600844,
+      "grad_norm": 0.07150838517569402,
+      "learning_rate": 8.467162503961209e-06,
+      "loss": 0.5317,
+      "step": 2078
+    },
+    {
+      "epoch": 1.0312771503040834,
+      "grad_norm": 0.07504062599965208,
+      "learning_rate": 8.465753836019853e-06,
+      "loss": 0.4991,
+      "step": 2079
+    },
+    {
+      "epoch": 1.0317736130073227,
+      "grad_norm": 0.06785624085900396,
+      "learning_rate": 8.46434463839359e-06,
+      "loss": 0.4836,
+      "step": 2080
+    },
+    {
+      "epoch": 1.0322700757105623,
+      "grad_norm": 0.07273111308848569,
+      "learning_rate": 8.462934911297797e-06,
+      "loss": 0.5016,
+      "step": 2081
+    },
+    {
+      "epoch": 1.0327665384138016,
+      "grad_norm": 0.07520341003875997,
+      "learning_rate": 8.461524654947927e-06,
+      "loss": 0.4808,
+      "step": 2082
+    },
+    {
+      "epoch": 1.0332630011170412,
+      "grad_norm": 0.07682200973431808,
+      "learning_rate": 8.460113869559517e-06,
+      "loss": 0.4908,
+      "step": 2083
+    },
+    {
+      "epoch": 1.0337594638202805,
+      "grad_norm": 0.07078292265875742,
+      "learning_rate": 8.458702555348176e-06,
+      "loss": 0.4962,
+      "step": 2084
+    },
+    {
+      "epoch": 1.0342559265235198,
+      "grad_norm": 0.0772565123724033,
+      "learning_rate": 8.45729071252961e-06,
+      "loss": 0.5454,
+      "step": 2085
+    },
+    {
+      "epoch": 1.0347523892267594,
+      "grad_norm": 0.076563402523467,
+      "learning_rate": 8.45587834131959e-06,
+      "loss": 0.517,
+      "step": 2086
+    },
+    {
+      "epoch": 1.0352488519299987,
+      "grad_norm": 0.0725286596603905,
+      "learning_rate": 8.454465441933976e-06,
+      "loss": 0.4779,
+      "step": 2087
+    },
+    {
+      "epoch": 1.0357453146332383,
+      "grad_norm": 0.07453353990137378,
+      "learning_rate": 8.453052014588707e-06,
+      "loss": 0.477,
+      "step": 2088
+    },
+    {
+      "epoch": 1.0362417773364776,
+      "grad_norm": 0.07335625469651262,
+      "learning_rate": 8.451638059499803e-06,
+      "loss": 0.5118,
+      "step": 2089
+    },
+    {
+      "epoch": 1.036738240039717,
+      "grad_norm": 0.07587133404150129,
+      "learning_rate": 8.450223576883365e-06,
+      "loss": 0.5302,
+      "step": 2090
+    },
+    {
+      "epoch": 1.0372347027429565,
+      "grad_norm": 0.07153404212574806,
+      "learning_rate": 8.448808566955575e-06,
+      "loss": 0.4998,
+      "step": 2091
+    },
+    {
+      "epoch": 1.0377311654461958,
+      "grad_norm": 0.07745704974069888,
+      "learning_rate": 8.447393029932692e-06,
+      "loss": 0.5115,
+      "step": 2092
+    },
+    {
+      "epoch": 1.0382276281494354,
+      "grad_norm": 0.07431588797598375,
+      "learning_rate": 8.445976966031057e-06,
+      "loss": 0.4692,
+      "step": 2093
+    },
+    {
+      "epoch": 1.0387240908526747,
+      "grad_norm": 0.07436571219611558,
+      "learning_rate": 8.444560375467098e-06,
+      "loss": 0.5276,
+      "step": 2094
+    },
+    {
+      "epoch": 1.039220553555914,
+      "grad_norm": 0.07294391240711395,
+      "learning_rate": 8.443143258457311e-06,
+      "loss": 0.4775,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0397170162591536,
+      "grad_norm": 0.07387829012012691,
+      "learning_rate": 8.441725615218287e-06,
+      "loss": 0.4563,
+      "step": 2096
+    },
+    {
+      "epoch": 1.040213478962393,
+      "grad_norm": 0.07242229489258531,
+      "learning_rate": 8.440307445966684e-06,
+      "loss": 0.4983,
+      "step": 2097
+    },
+    {
+      "epoch": 1.0407099416656325,
+      "grad_norm": 0.07064880963925171,
+      "learning_rate": 8.438888750919252e-06,
+      "loss": 0.5024,
+      "step": 2098
+    },
+    {
+      "epoch": 1.0412064043688718,
+      "grad_norm": 0.07072758643070051,
+      "learning_rate": 8.43746953029281e-06,
+      "loss": 0.4797,
+      "step": 2099
+    },
+    {
+      "epoch": 1.0417028670721111,
+      "grad_norm": 0.07118181371356039,
+      "learning_rate": 8.436049784304268e-06,
+      "loss": 0.4881,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0421993297753507,
+      "grad_norm": 0.07017625392750985,
+      "learning_rate": 8.43462951317061e-06,
+      "loss": 0.4812,
+      "step": 2101
+    },
+    {
+      "epoch": 1.04269579247859,
+      "grad_norm": 0.07377904358026054,
+      "learning_rate": 8.433208717108899e-06,
+      "loss": 0.5008,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0431922551818296,
+      "grad_norm": 0.07445763458420322,
+      "learning_rate": 8.431787396336283e-06,
+      "loss": 0.4829,
+      "step": 2103
+    },
+    {
+      "epoch": 1.043688717885069,
+      "grad_norm": 0.07470552441745613,
+      "learning_rate": 8.430365551069989e-06,
+      "loss": 0.5315,
+      "step": 2104
+    },
+    {
+      "epoch": 1.0441851805883082,
+      "grad_norm": 0.07295084291229291,
+      "learning_rate": 8.42894318152732e-06,
+      "loss": 0.5156,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0446816432915478,
+      "grad_norm": 0.07646343472351082,
+      "learning_rate": 8.427520287925669e-06,
+      "loss": 0.5085,
+      "step": 2106
+    },
+    {
+      "epoch": 1.045178105994787,
+      "grad_norm": 0.07191136555573914,
+      "learning_rate": 8.426096870482495e-06,
+      "loss": 0.4682,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0456745686980267,
+      "grad_norm": 0.07060992139300228,
+      "learning_rate": 8.424672929415347e-06,
+      "loss": 0.4731,
+      "step": 2108
+    },
+    {
+      "epoch": 1.046171031401266,
+      "grad_norm": 0.07325524222008324,
+      "learning_rate": 8.423248464941854e-06,
+      "loss": 0.5259,
+      "step": 2109
+    },
+    {
+      "epoch": 1.0466674941045053,
+      "grad_norm": 0.07085481033742728,
+      "learning_rate": 8.421823477279719e-06,
+      "loss": 0.4817,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0471639568077449,
+      "grad_norm": 0.07275605555140592,
+      "learning_rate": 8.420397966646732e-06,
+      "loss": 0.5319,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0476604195109842,
+      "grad_norm": 0.07397160654166765,
+      "learning_rate": 8.418971933260755e-06,
+      "loss": 0.4864,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0481568822142238,
+      "grad_norm": 0.07219789497661702,
+      "learning_rate": 8.417545377339739e-06,
+      "loss": 0.5307,
+      "step": 2113
+    },
+    {
+      "epoch": 1.048653344917463,
+      "grad_norm": 0.06817862255465229,
+      "learning_rate": 8.41611829910171e-06,
+      "loss": 0.5066,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0491498076207024,
+      "grad_norm": 0.07474161367633145,
+      "learning_rate": 8.41469069876477e-06,
+      "loss": 0.5049,
+      "step": 2115
+    },
+    {
+      "epoch": 1.049646270323942,
+      "grad_norm": 0.07441496396684265,
+      "learning_rate": 8.413262576547108e-06,
+      "loss": 0.4809,
+      "step": 2116
+    },
+    {
+      "epoch": 1.0501427330271813,
+      "grad_norm": 0.07004502242911621,
+      "learning_rate": 8.411833932666989e-06,
+      "loss": 0.4678,
+      "step": 2117
+    },
+    {
+      "epoch": 1.0506391957304206,
+      "grad_norm": 0.0737960103890059,
+      "learning_rate": 8.410404767342757e-06,
+      "loss": 0.496,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0511356584336602,
+      "grad_norm": 0.07332685142857046,
+      "learning_rate": 8.408975080792839e-06,
+      "loss": 0.462,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0516321211368995,
+      "grad_norm": 0.07372161091283125,
+      "learning_rate": 8.407544873235736e-06,
+      "loss": 0.5025,
+      "step": 2120
+    },
+    {
+      "epoch": 1.052128583840139,
+      "grad_norm": 0.07211336962240478,
+      "learning_rate": 8.406114144890038e-06,
+      "loss": 0.4897,
+      "step": 2121
+    },
+    {
+      "epoch": 1.0526250465433784,
+      "grad_norm": 0.07149511493399731,
+      "learning_rate": 8.404682895974404e-06,
+      "loss": 0.4974,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0531215092466177,
+      "grad_norm": 0.07374106989538137,
+      "learning_rate": 8.403251126707581e-06,
+      "loss": 0.5235,
+      "step": 2123
+    },
+    {
+      "epoch": 1.0536179719498573,
+      "grad_norm": 0.07446852193118496,
+      "learning_rate": 8.401818837308388e-06,
+      "loss": 0.4998,
+      "step": 2124
+    },
+    {
+      "epoch": 1.0541144346530966,
+      "grad_norm": 0.07002579988117166,
+      "learning_rate": 8.400386027995732e-06,
+      "loss": 0.4651,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0546108973563362,
+      "grad_norm": 0.07254323023643783,
+      "learning_rate": 8.398952698988592e-06,
+      "loss": 0.4839,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0551073600595755,
+      "grad_norm": 0.07723775990560815,
+      "learning_rate": 8.39751885050603e-06,
+      "loss": 0.4951,
+      "step": 2127
+    },
+    {
+      "epoch": 1.0556038227628148,
+      "grad_norm": 0.07673925620439852,
+      "learning_rate": 8.396084482767186e-06,
+      "loss": 0.5034,
+      "step": 2128
+    },
+    {
+      "epoch": 1.0561002854660544,
+      "grad_norm": 0.07694300866783188,
+      "learning_rate": 8.39464959599128e-06,
+      "loss": 0.4752,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0565967481692937,
+      "grad_norm": 0.07237850482434734,
+      "learning_rate": 8.393214190397615e-06,
+      "loss": 0.5097,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0570932108725333,
+      "grad_norm": 0.07628338215290212,
+      "learning_rate": 8.391778266205565e-06,
+      "loss": 0.5026,
+      "step": 2131
+    },
+    {
+      "epoch": 1.0575896735757726,
+      "grad_norm": 0.07604150356394877,
+      "learning_rate": 8.390341823634591e-06,
+      "loss": 0.5094,
+      "step": 2132
+    },
+    {
+      "epoch": 1.058086136279012,
+      "grad_norm": 0.07318041529403152,
+      "learning_rate": 8.38890486290423e-06,
+      "loss": 0.513,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0585825989822515,
+      "grad_norm": 0.07150538307929709,
+      "learning_rate": 8.387467384234096e-06,
+      "loss": 0.495,
+      "step": 2134
+    },
+    {
+      "epoch": 1.0590790616854908,
+      "grad_norm": 0.0738979649614615,
+      "learning_rate": 8.386029387843888e-06,
+      "loss": 0.4943,
+      "step": 2135
+    },
+    {
+      "epoch": 1.0595755243887304,
+      "grad_norm": 0.07223352302598267,
+      "learning_rate": 8.384590873953376e-06,
+      "loss": 0.5002,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0600719870919697,
+      "grad_norm": 0.07703756058610264,
+      "learning_rate": 8.38315184278242e-06,
+      "loss": 0.4888,
+      "step": 2137
+    },
+    {
+      "epoch": 1.060568449795209,
+      "grad_norm": 0.06992834400071618,
+      "learning_rate": 8.381712294550948e-06,
+      "loss": 0.4591,
+      "step": 2138
+    },
+    {
+      "epoch": 1.0610649124984486,
+      "grad_norm": 0.07346783438489786,
+      "learning_rate": 8.380272229478974e-06,
+      "loss": 0.4877,
+      "step": 2139
+    },
+    {
+      "epoch": 1.061561375201688,
+      "grad_norm": 0.07244028096269002,
+      "learning_rate": 8.378831647786586e-06,
+      "loss": 0.4724,
+      "step": 2140
+    },
+    {
+      "epoch": 1.0620578379049275,
+      "grad_norm": 0.07201885190546764,
+      "learning_rate": 8.377390549693959e-06,
+      "loss": 0.493,
+      "step": 2141
+    },
+    {
+      "epoch": 1.0625543006081668,
+      "grad_norm": 0.07556559260892799,
+      "learning_rate": 8.375948935421337e-06,
+      "loss": 0.491,
+      "step": 2142
+    },
+    {
+      "epoch": 1.0630507633114061,
+      "grad_norm": 0.07291427292666455,
+      "learning_rate": 8.374506805189051e-06,
+      "loss": 0.5273,
+      "step": 2143
+    },
+    {
+      "epoch": 1.0635472260146457,
+      "grad_norm": 0.07549265071798042,
+      "learning_rate": 8.373064159217506e-06,
+      "loss": 0.5096,
+      "step": 2144
+    },
+    {
+      "epoch": 1.064043688717885,
+      "grad_norm": 0.07475791606852278,
+      "learning_rate": 8.371620997727184e-06,
+      "loss": 0.4717,
+      "step": 2145
+    },
+    {
+      "epoch": 1.0645401514211246,
+      "grad_norm": 0.07413928468511259,
+      "learning_rate": 8.370177320938656e-06,
+      "loss": 0.5129,
+      "step": 2146
+    },
+    {
+      "epoch": 1.065036614124364,
+      "grad_norm": 0.07800399161763176,
+      "learning_rate": 8.36873312907256e-06,
+      "loss": 0.5194,
+      "step": 2147
+    },
+    {
+      "epoch": 1.0655330768276032,
+      "grad_norm": 0.07524291986138339,
+      "learning_rate": 8.367288422349617e-06,
+      "loss": 0.5008,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0660295395308428,
+      "grad_norm": 0.07511255026748975,
+      "learning_rate": 8.365843200990632e-06,
+      "loss": 0.4985,
+      "step": 2149
+    },
+    {
+      "epoch": 1.0665260022340821,
+      "grad_norm": 0.07054144916309989,
+      "learning_rate": 8.364397465216479e-06,
+      "loss": 0.4743,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0670224649373217,
+      "grad_norm": 0.07193153540851546,
+      "learning_rate": 8.362951215248118e-06,
+      "loss": 0.5347,
+      "step": 2151
+    },
+    {
+      "epoch": 1.067518927640561,
+      "grad_norm": 0.07176991910719552,
+      "learning_rate": 8.361504451306585e-06,
+      "loss": 0.5031,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0680153903438003,
+      "grad_norm": 0.07170040930551358,
+      "learning_rate": 8.360057173612993e-06,
+      "loss": 0.492,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0685118530470399,
+      "grad_norm": 0.07157465397253797,
+      "learning_rate": 8.358609382388538e-06,
+      "loss": 0.5148,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0690083157502792,
+      "grad_norm": 0.07006770452583463,
+      "learning_rate": 8.35716107785449e-06,
+      "loss": 0.4604,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0695047784535188,
+      "grad_norm": 0.07385970105527723,
+      "learning_rate": 8.355712260232197e-06,
+      "loss": 0.5139,
+      "step": 2156
+    },
+    {
+      "epoch": 1.070001241156758,
+      "grad_norm": 0.07298273056849365,
+      "learning_rate": 8.35426292974309e-06,
+      "loss": 0.4724,
+      "step": 2157
+    },
+    {
+      "epoch": 1.0704977038599974,
+      "grad_norm": 0.07226179617360365,
+      "learning_rate": 8.352813086608678e-06,
+      "loss": 0.4996,
+      "step": 2158
+    },
+    {
+      "epoch": 1.070994166563237,
+      "grad_norm": 0.07083077084368294,
+      "learning_rate": 8.351362731050542e-06,
+      "loss": 0.4659,
+      "step": 2159
+    },
+    {
+      "epoch": 1.0714906292664763,
+      "grad_norm": 0.07357266169838603,
+      "learning_rate": 8.34991186329035e-06,
+      "loss": 0.4889,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0719870919697159,
+      "grad_norm": 0.07734031191337602,
+      "learning_rate": 8.348460483549841e-06,
+      "loss": 0.5021,
+      "step": 2161
+    },
+    {
+      "epoch": 1.0724835546729552,
+      "grad_norm": 0.07378794415957397,
+      "learning_rate": 8.347008592050834e-06,
+      "loss": 0.5122,
+      "step": 2162
+    },
+    {
+      "epoch": 1.0729800173761945,
+      "grad_norm": 0.07201934091861806,
+      "learning_rate": 8.345556189015231e-06,
+      "loss": 0.4443,
+      "step": 2163
+    },
+    {
+      "epoch": 1.073476480079434,
+      "grad_norm": 0.07040976076289147,
+      "learning_rate": 8.344103274665002e-06,
+      "loss": 0.4995,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0739729427826734,
+      "grad_norm": 0.07088121726877565,
+      "learning_rate": 8.34264984922221e-06,
+      "loss": 0.4891,
+      "step": 2165
+    },
+    {
+      "epoch": 1.074469405485913,
+      "grad_norm": 0.07686251463455322,
+      "learning_rate": 8.341195912908984e-06,
+      "loss": 0.5565,
+      "step": 2166
+    },
+    {
+      "epoch": 1.0749658681891523,
+      "grad_norm": 0.07252309150564194,
+      "learning_rate": 8.339741465947533e-06,
+      "loss": 0.4759,
+      "step": 2167
+    },
+    {
+      "epoch": 1.0754623308923916,
+      "grad_norm": 0.07087699232546477,
+      "learning_rate": 8.338286508560148e-06,
+      "loss": 0.4977,
+      "step": 2168
+    },
+    {
+      "epoch": 1.0759587935956312,
+      "grad_norm": 0.07432595022478114,
+      "learning_rate": 8.336831040969196e-06,
+      "loss": 0.5416,
+      "step": 2169
+    },
+    {
+      "epoch": 1.0764552562988705,
+      "grad_norm": 0.0734004314269313,
+      "learning_rate": 8.335375063397123e-06,
+      "loss": 0.4883,
+      "step": 2170
+    },
+    {
+      "epoch": 1.07695171900211,
+      "grad_norm": 0.07405951233615124,
+      "learning_rate": 8.333918576066446e-06,
+      "loss": 0.4907,
+      "step": 2171
+    },
+    {
+      "epoch": 1.0774481817053494,
+      "grad_norm": 0.07060359600736023,
+      "learning_rate": 8.332461579199773e-06,
+      "loss": 0.4841,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0779446444085887,
+      "grad_norm": 0.06899558938544381,
+      "learning_rate": 8.33100407301978e-06,
+      "loss": 0.4777,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0784411071118283,
+      "grad_norm": 0.07481858213782207,
+      "learning_rate": 8.32954605774922e-06,
+      "loss": 0.4673,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0789375698150676,
+      "grad_norm": 0.075847623970544,
+      "learning_rate": 8.328087533610933e-06,
+      "loss": 0.4756,
+      "step": 2175
+    },
+    {
+      "epoch": 1.0794340325183072,
+      "grad_norm": 0.07120553324275082,
+      "learning_rate": 8.326628500827826e-06,
+      "loss": 0.4693,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0799304952215465,
+      "grad_norm": 0.07415639230784418,
+      "learning_rate": 8.325168959622893e-06,
+      "loss": 0.4867,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0804269579247858,
+      "grad_norm": 0.07374489312987371,
+      "learning_rate": 8.323708910219201e-06,
+      "loss": 0.4627,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0809234206280254,
+      "grad_norm": 0.07383508955193396,
+      "learning_rate": 8.322248352839893e-06,
+      "loss": 0.5022,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0814198833312647,
+      "grad_norm": 0.0708623476675988,
+      "learning_rate": 8.32078728770819e-06,
+      "loss": 0.4897,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0819163460345043,
+      "grad_norm": 0.07441658954175509,
+      "learning_rate": 8.319325715047394e-06,
+      "loss": 0.4839,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0824128087377436,
+      "grad_norm": 0.07083548626853245,
+      "learning_rate": 8.317863635080886e-06,
+      "loss": 0.5122,
+      "step": 2182
+    },
+    {
+      "epoch": 1.082909271440983,
+      "grad_norm": 0.07421527166175029,
+      "learning_rate": 8.316401048032121e-06,
+      "loss": 0.4773,
+      "step": 2183
+    },
+    {
+      "epoch": 1.0834057341442225,
+      "grad_norm": 0.07327130466932932,
+      "learning_rate": 8.314937954124629e-06,
+      "loss": 0.5122,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0839021968474618,
+      "grad_norm": 0.07536245786155,
+      "learning_rate": 8.31347435358202e-06,
+      "loss": 0.499,
+      "step": 2185
+    },
+    {
+      "epoch": 1.0843986595507014,
+      "grad_norm": 0.07593717142738794,
+      "learning_rate": 8.312010246627986e-06,
+      "loss": 0.485,
+      "step": 2186
+    },
+    {
+      "epoch": 1.0848951222539407,
+      "grad_norm": 0.07167997632106814,
+      "learning_rate": 8.31054563348629e-06,
+      "loss": 0.459,
+      "step": 2187
+    },
+    {
+      "epoch": 1.08539158495718,
+      "grad_norm": 0.07270113397492473,
+      "learning_rate": 8.309080514380771e-06,
+      "loss": 0.4727,
+      "step": 2188
+    },
+    {
+      "epoch": 1.0858880476604196,
+      "grad_norm": 0.07627450526912384,
+      "learning_rate": 8.307614889535354e-06,
+      "loss": 0.4767,
+      "step": 2189
+    },
+    {
+      "epoch": 1.086384510363659,
+      "grad_norm": 0.08054828675705718,
+      "learning_rate": 8.306148759174036e-06,
+      "loss": 0.5516,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0868809730668985,
+      "grad_norm": 0.07374128808206212,
+      "learning_rate": 8.30468212352089e-06,
+      "loss": 0.4905,
+      "step": 2191
+    },
+    {
+      "epoch": 1.0873774357701378,
+      "grad_norm": 0.07487202792279594,
+      "learning_rate": 8.303214982800067e-06,
+      "loss": 0.5264,
+      "step": 2192
+    },
+    {
+      "epoch": 1.0878738984733771,
+      "grad_norm": 0.07381112821428028,
+      "learning_rate": 8.301747337235798e-06,
+      "loss": 0.4757,
+      "step": 2193
+    },
+    {
+      "epoch": 1.0883703611766167,
+      "grad_norm": 0.07339639597409396,
+      "learning_rate": 8.300279187052386e-06,
+      "loss": 0.48,
+      "step": 2194
+    },
+    {
+      "epoch": 1.088866823879856,
+      "grad_norm": 0.076185034343091,
+      "learning_rate": 8.298810532474218e-06,
+      "loss": 0.4781,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0893632865830956,
+      "grad_norm": 0.068948413488983,
+      "learning_rate": 8.297341373725754e-06,
+      "loss": 0.4594,
+      "step": 2196
+    },
+    {
+      "epoch": 1.089859749286335,
+      "grad_norm": 0.07517076494598852,
+      "learning_rate": 8.295871711031527e-06,
+      "loss": 0.5134,
+      "step": 2197
+    },
+    {
+      "epoch": 1.0903562119895742,
+      "grad_norm": 0.07353154295953639,
+      "learning_rate": 8.294401544616155e-06,
+      "loss": 0.5114,
+      "step": 2198
+    },
+    {
+      "epoch": 1.0908526746928138,
+      "grad_norm": 0.07625897388957875,
+      "learning_rate": 8.292930874704328e-06,
+      "loss": 0.4849,
+      "step": 2199
+    },
+    {
+      "epoch": 1.091349137396053,
+      "grad_norm": 0.07626683441495624,
+      "learning_rate": 8.291459701520816e-06,
+      "loss": 0.4931,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0918456000992927,
+      "grad_norm": 0.07301701910481151,
+      "learning_rate": 8.289988025290463e-06,
+      "loss": 0.4972,
+      "step": 2201
+    },
+    {
+      "epoch": 1.092342062802532,
+      "grad_norm": 0.0725810819115238,
+      "learning_rate": 8.288515846238193e-06,
+      "loss": 0.4686,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0928385255057713,
+      "grad_norm": 0.07479074559006994,
+      "learning_rate": 8.287043164589001e-06,
+      "loss": 0.5003,
+      "step": 2203
+    },
+    {
+      "epoch": 1.0933349882090109,
+      "grad_norm": 0.07312842594353675,
+      "learning_rate": 8.285569980567965e-06,
+      "loss": 0.4772,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0938314509122502,
+      "grad_norm": 0.07054737968378164,
+      "learning_rate": 8.284096294400238e-06,
+      "loss": 0.5029,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0943279136154895,
+      "grad_norm": 0.07641103035051312,
+      "learning_rate": 8.282622106311049e-06,
+      "loss": 0.5053,
+      "step": 2206
+    },
+    {
+      "epoch": 1.094824376318729,
+      "grad_norm": 0.07449614033992102,
+      "learning_rate": 8.281147416525704e-06,
+      "loss": 0.5007,
+      "step": 2207
+    },
+    {
+      "epoch": 1.0953208390219684,
+      "grad_norm": 0.0757024611392332,
+      "learning_rate": 8.279672225269584e-06,
+      "loss": 0.5168,
+      "step": 2208
+    },
+    {
+      "epoch": 1.095817301725208,
+      "grad_norm": 0.07543978367284243,
+      "learning_rate": 8.278196532768152e-06,
+      "loss": 0.4986,
+      "step": 2209
+    },
+    {
+      "epoch": 1.0963137644284473,
+      "grad_norm": 0.07885663632789892,
+      "learning_rate": 8.276720339246942e-06,
+      "loss": 0.5038,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0968102271316866,
+      "grad_norm": 0.07595383345135583,
+      "learning_rate": 8.275243644931565e-06,
+      "loss": 0.4982,
+      "step": 2211
+    },
+    {
+      "epoch": 1.0973066898349262,
+      "grad_norm": 0.0739185354406121,
+      "learning_rate": 8.273766450047713e-06,
+      "loss": 0.493,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0978031525381655,
+      "grad_norm": 0.07391837862098494,
+      "learning_rate": 8.272288754821149e-06,
+      "loss": 0.533,
+      "step": 2213
+    },
+    {
+      "epoch": 1.098299615241405,
+      "grad_norm": 0.08110713367122144,
+      "learning_rate": 8.270810559477716e-06,
+      "loss": 0.4997,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0987960779446444,
+      "grad_norm": 0.0758720474097381,
+      "learning_rate": 8.26933186424333e-06,
+      "loss": 0.5158,
+      "step": 2215
+    },
+    {
+      "epoch": 1.0992925406478837,
+      "grad_norm": 0.0718223545522586,
+      "learning_rate": 8.267852669343991e-06,
+      "loss": 0.4516,
+      "step": 2216
+    },
+    {
+      "epoch": 1.0997890033511233,
+      "grad_norm": 0.07859584280020389,
+      "learning_rate": 8.266372975005768e-06,
+      "loss": 0.5332,
+      "step": 2217
+    },
+    {
+      "epoch": 1.1002854660543626,
+      "grad_norm": 0.07291038437725997,
+      "learning_rate": 8.264892781454807e-06,
+      "loss": 0.4982,
+      "step": 2218
+    },
+    {
+      "epoch": 1.1007819287576022,
+      "grad_norm": 0.07372572346588169,
+      "learning_rate": 8.263412088917333e-06,
+      "loss": 0.4804,
+      "step": 2219
+    },
+    {
+      "epoch": 1.1012783914608415,
+      "grad_norm": 0.06939509607738924,
+      "learning_rate": 8.261930897619647e-06,
+      "loss": 0.4397,
+      "step": 2220
+    },
+    {
+      "epoch": 1.1017748541640808,
+      "grad_norm": 0.0728653700593787,
+      "learning_rate": 8.260449207788124e-06,
+      "loss": 0.4753,
+      "step": 2221
+    },
+    {
+      "epoch": 1.1022713168673204,
+      "grad_norm": 0.07454856223228062,
+      "learning_rate": 8.258967019649216e-06,
+      "loss": 0.5085,
+      "step": 2222
+    },
+    {
+      "epoch": 1.1027677795705597,
+      "grad_norm": 0.074275864871691,
+      "learning_rate": 8.257484333429452e-06,
+      "loss": 0.4788,
+      "step": 2223
+    },
+    {
+      "epoch": 1.1032642422737993,
+      "grad_norm": 0.07351109388785956,
+      "learning_rate": 8.256001149355439e-06,
+      "loss": 0.5037,
+      "step": 2224
+    },
+    {
+      "epoch": 1.1037607049770386,
+      "grad_norm": 0.07596144114073311,
+      "learning_rate": 8.254517467653858e-06,
+      "loss": 0.4851,
+      "step": 2225
+    },
+    {
+      "epoch": 1.104257167680278,
+      "grad_norm": 0.07320020631956403,
+      "learning_rate": 8.253033288551463e-06,
+      "loss": 0.5113,
+      "step": 2226
+    },
+    {
+      "epoch": 1.1047536303835175,
+      "grad_norm": 0.07343420892205225,
+      "learning_rate": 8.251548612275086e-06,
+      "loss": 0.5087,
+      "step": 2227
+    },
+    {
+      "epoch": 1.1052500930867568,
+      "grad_norm": 0.0763838063102248,
+      "learning_rate": 8.25006343905164e-06,
+      "loss": 0.4955,
+      "step": 2228
+    },
+    {
+      "epoch": 1.1057465557899964,
+      "grad_norm": 0.07339654182978661,
+      "learning_rate": 8.248577769108106e-06,
+      "loss": 0.4988,
+      "step": 2229
+    },
+    {
+      "epoch": 1.1062430184932357,
+      "grad_norm": 0.07362892885262858,
+      "learning_rate": 8.247091602671551e-06,
+      "loss": 0.4855,
+      "step": 2230
+    },
+    {
+      "epoch": 1.106739481196475,
+      "grad_norm": 0.06953026409193383,
+      "learning_rate": 8.245604939969104e-06,
+      "loss": 0.4896,
+      "step": 2231
+    },
+    {
+      "epoch": 1.1072359438997146,
+      "grad_norm": 0.07111640843092504,
+      "learning_rate": 8.244117781227982e-06,
+      "loss": 0.5095,
+      "step": 2232
+    },
+    {
+      "epoch": 1.107732406602954,
+      "grad_norm": 0.07481021626932494,
+      "learning_rate": 8.242630126675475e-06,
+      "loss": 0.5125,
+      "step": 2233
+    },
+    {
+      "epoch": 1.1082288693061935,
+      "grad_norm": 0.07395283841841717,
+      "learning_rate": 8.241141976538944e-06,
+      "loss": 0.5416,
+      "step": 2234
+    },
+    {
+      "epoch": 1.1087253320094328,
+      "grad_norm": 0.07130473116026037,
+      "learning_rate": 8.239653331045827e-06,
+      "loss": 0.4824,
+      "step": 2235
+    },
+    {
+      "epoch": 1.1092217947126721,
+      "grad_norm": 0.06947962941312745,
+      "learning_rate": 8.238164190423645e-06,
+      "loss": 0.4668,
+      "step": 2236
+    },
+    {
+      "epoch": 1.1097182574159117,
+      "grad_norm": 0.07191577528119829,
+      "learning_rate": 8.236674554899985e-06,
+      "loss": 0.4853,
+      "step": 2237
+    },
+    {
+      "epoch": 1.110214720119151,
+      "grad_norm": 0.07496165833604053,
+      "learning_rate": 8.235184424702516e-06,
+      "loss": 0.4787,
+      "step": 2238
+    },
+    {
+      "epoch": 1.1107111828223906,
+      "grad_norm": 0.07213792735155145,
+      "learning_rate": 8.23369380005898e-06,
+      "loss": 0.5311,
+      "step": 2239
+    },
+    {
+      "epoch": 1.11120764552563,
+      "grad_norm": 0.07053117748052583,
+      "learning_rate": 8.232202681197194e-06,
+      "loss": 0.4654,
+      "step": 2240
+    },
+    {
+      "epoch": 1.1117041082288692,
+      "grad_norm": 0.07528814724049783,
+      "learning_rate": 8.230711068345055e-06,
+      "loss": 0.5237,
+      "step": 2241
+    },
+    {
+      "epoch": 1.1122005709321088,
+      "grad_norm": 0.06950506610855729,
+      "learning_rate": 8.229218961730527e-06,
+      "loss": 0.4976,
+      "step": 2242
+    },
+    {
+      "epoch": 1.112697033635348,
+      "grad_norm": 0.07195918947291367,
+      "learning_rate": 8.227726361581659e-06,
+      "loss": 0.4846,
+      "step": 2243
+    },
+    {
+      "epoch": 1.1131934963385877,
+      "grad_norm": 0.06959712282406523,
+      "learning_rate": 8.22623326812657e-06,
+      "loss": 0.4849,
+      "step": 2244
+    },
+    {
+      "epoch": 1.113689959041827,
+      "grad_norm": 0.07051471449226025,
+      "learning_rate": 8.224739681593453e-06,
+      "loss": 0.4855,
+      "step": 2245
+    },
+    {
+      "epoch": 1.1141864217450663,
+      "grad_norm": 0.07379793580293327,
+      "learning_rate": 8.22324560221058e-06,
+      "loss": 0.5028,
+      "step": 2246
+    },
+    {
+      "epoch": 1.1146828844483059,
+      "grad_norm": 0.07176973461803826,
+      "learning_rate": 8.221751030206297e-06,
+      "loss": 0.4903,
+      "step": 2247
+    },
+    {
+      "epoch": 1.1151793471515452,
+      "grad_norm": 0.07613295929530475,
+      "learning_rate": 8.220255965809027e-06,
+      "loss": 0.498,
+      "step": 2248
+    },
+    {
+      "epoch": 1.1156758098547848,
+      "grad_norm": 0.07481677436822458,
+      "learning_rate": 8.218760409247267e-06,
+      "loss": 0.4896,
+      "step": 2249
+    },
+    {
+      "epoch": 1.116172272558024,
+      "grad_norm": 0.07606898831692863,
+      "learning_rate": 8.217264360749587e-06,
+      "loss": 0.4571,
+      "step": 2250
+    },
+    {
+      "epoch": 1.1166687352612634,
+      "grad_norm": 0.07151037999784068,
+      "learning_rate": 8.215767820544633e-06,
+      "loss": 0.5038,
+      "step": 2251
+    },
+    {
+      "epoch": 1.117165197964503,
+      "grad_norm": 0.0750769122463959,
+      "learning_rate": 8.21427078886113e-06,
+      "loss": 0.5074,
+      "step": 2252
+    },
+    {
+      "epoch": 1.1176616606677423,
+      "grad_norm": 0.0749081827284495,
+      "learning_rate": 8.21277326592787e-06,
+      "loss": 0.4922,
+      "step": 2253
+    },
+    {
+      "epoch": 1.1181581233709816,
+      "grad_norm": 0.07152106501318276,
+      "learning_rate": 8.211275251973734e-06,
+      "loss": 0.4843,
+      "step": 2254
+    },
+    {
+      "epoch": 1.1186545860742212,
+      "grad_norm": 0.07098442469069642,
+      "learning_rate": 8.209776747227663e-06,
+      "loss": 0.4735,
+      "step": 2255
+    },
+    {
+      "epoch": 1.1191510487774605,
+      "grad_norm": 0.0719612578777751,
+      "learning_rate": 8.20827775191868e-06,
+      "loss": 0.5094,
+      "step": 2256
+    },
+    {
+      "epoch": 1.1196475114807,
+      "grad_norm": 0.07179858593800634,
+      "learning_rate": 8.206778266275885e-06,
+      "loss": 0.488,
+      "step": 2257
+    },
+    {
+      "epoch": 1.1201439741839394,
+      "grad_norm": 0.07118495482126498,
+      "learning_rate": 8.205278290528446e-06,
+      "loss": 0.465,
+      "step": 2258
+    },
+    {
+      "epoch": 1.1206404368871787,
+      "grad_norm": 0.07419162410687083,
+      "learning_rate": 8.203777824905617e-06,
+      "loss": 0.4747,
+      "step": 2259
+    },
+    {
+      "epoch": 1.1211368995904183,
+      "grad_norm": 0.0710781813716537,
+      "learning_rate": 8.202276869636713e-06,
+      "loss": 0.494,
+      "step": 2260
+    },
+    {
+      "epoch": 1.1216333622936576,
+      "grad_norm": 0.07732360536034513,
+      "learning_rate": 8.200775424951137e-06,
+      "loss": 0.504,
+      "step": 2261
+    },
+    {
+      "epoch": 1.1221298249968972,
+      "grad_norm": 0.06777372530014544,
+      "learning_rate": 8.199273491078355e-06,
+      "loss": 0.4637,
+      "step": 2262
+    },
+    {
+      "epoch": 1.1226262877001365,
+      "grad_norm": 0.07084811372777965,
+      "learning_rate": 8.197771068247917e-06,
+      "loss": 0.485,
+      "step": 2263
+    },
+    {
+      "epoch": 1.1231227504033758,
+      "grad_norm": 0.07658689251813063,
+      "learning_rate": 8.196268156689444e-06,
+      "loss": 0.5295,
+      "step": 2264
+    },
+    {
+      "epoch": 1.1236192131066154,
+      "grad_norm": 0.07540282347213605,
+      "learning_rate": 8.194764756632632e-06,
+      "loss": 0.4946,
+      "step": 2265
+    },
+    {
+      "epoch": 1.1241156758098547,
+      "grad_norm": 0.07431439448491374,
+      "learning_rate": 8.193260868307251e-06,
+      "loss": 0.5025,
+      "step": 2266
+    },
+    {
+      "epoch": 1.1246121385130943,
+      "grad_norm": 0.08623105926325221,
+      "learning_rate": 8.191756491943146e-06,
+      "loss": 0.5448,
+      "step": 2267
+    },
+    {
+      "epoch": 1.1251086012163336,
+      "grad_norm": 0.07666257532724999,
+      "learning_rate": 8.190251627770237e-06,
+      "loss": 0.5008,
+      "step": 2268
+    },
+    {
+      "epoch": 1.125605063919573,
+      "grad_norm": 0.0760010099142199,
+      "learning_rate": 8.188746276018518e-06,
+      "loss": 0.4786,
+      "step": 2269
+    },
+    {
+      "epoch": 1.1261015266228125,
+      "grad_norm": 0.07651610079909663,
+      "learning_rate": 8.187240436918057e-06,
+      "loss": 0.4941,
+      "step": 2270
+    },
+    {
+      "epoch": 1.1265979893260518,
+      "grad_norm": 0.07217604222108294,
+      "learning_rate": 8.185734110699002e-06,
+      "loss": 0.4981,
+      "step": 2271
+    },
+    {
+      "epoch": 1.1270944520292914,
+      "grad_norm": 0.07313906323366962,
+      "learning_rate": 8.184227297591568e-06,
+      "loss": 0.4865,
+      "step": 2272
+    },
+    {
+      "epoch": 1.1275909147325307,
+      "grad_norm": 0.07579452675806608,
+      "learning_rate": 8.182719997826043e-06,
+      "loss": 0.4815,
+      "step": 2273
+    },
+    {
+      "epoch": 1.12808737743577,
+      "grad_norm": 0.07607016220528141,
+      "learning_rate": 8.1812122116328e-06,
+      "loss": 0.4973,
+      "step": 2274
+    },
+    {
+      "epoch": 1.1285838401390096,
+      "grad_norm": 0.06925031457091706,
+      "learning_rate": 8.179703939242276e-06,
+      "loss": 0.47,
+      "step": 2275
+    },
+    {
+      "epoch": 1.129080302842249,
+      "grad_norm": 0.07533759199309992,
+      "learning_rate": 8.178195180884989e-06,
+      "loss": 0.4871,
+      "step": 2276
+    },
+    {
+      "epoch": 1.1295767655454885,
+      "grad_norm": 0.0749165977272755,
+      "learning_rate": 8.176685936791526e-06,
+      "loss": 0.4734,
+      "step": 2277
+    },
+    {
+      "epoch": 1.1300732282487278,
+      "grad_norm": 0.07436005370185177,
+      "learning_rate": 8.175176207192552e-06,
+      "loss": 0.5079,
+      "step": 2278
+    },
+    {
+      "epoch": 1.1305696909519671,
+      "grad_norm": 0.07557004980259102,
+      "learning_rate": 8.173665992318805e-06,
+      "loss": 0.5002,
+      "step": 2279
+    },
+    {
+      "epoch": 1.1310661536552067,
+      "grad_norm": 0.07567715209828088,
+      "learning_rate": 8.172155292401096e-06,
+      "loss": 0.5125,
+      "step": 2280
+    },
+    {
+      "epoch": 1.131562616358446,
+      "grad_norm": 0.07764282615456619,
+      "learning_rate": 8.170644107670313e-06,
+      "loss": 0.4966,
+      "step": 2281
+    },
+    {
+      "epoch": 1.1320590790616856,
+      "grad_norm": 0.07372865116940043,
+      "learning_rate": 8.169132438357416e-06,
+      "loss": 0.5107,
+      "step": 2282
+    },
+    {
+      "epoch": 1.132555541764925,
+      "grad_norm": 0.07638680398573897,
+      "learning_rate": 8.16762028469344e-06,
+      "loss": 0.4882,
+      "step": 2283
+    },
+    {
+      "epoch": 1.1330520044681642,
+      "grad_norm": 0.07121039450346492,
+      "learning_rate": 8.166107646909491e-06,
+      "loss": 0.5046,
+      "step": 2284
+    },
+    {
+      "epoch": 1.1335484671714038,
+      "grad_norm": 0.07401683052681642,
+      "learning_rate": 8.164594525236752e-06,
+      "loss": 0.4988,
+      "step": 2285
+    },
+    {
+      "epoch": 1.1340449298746431,
+      "grad_norm": 0.07398370488195749,
+      "learning_rate": 8.163080919906482e-06,
+      "loss": 0.4866,
+      "step": 2286
+    },
+    {
+      "epoch": 1.1345413925778827,
+      "grad_norm": 0.07422405936633944,
+      "learning_rate": 8.161566831150007e-06,
+      "loss": 0.4901,
+      "step": 2287
+    },
+    {
+      "epoch": 1.135037855281122,
+      "grad_norm": 0.06830231848502405,
+      "learning_rate": 8.160052259198737e-06,
+      "loss": 0.4868,
+      "step": 2288
+    },
+    {
+      "epoch": 1.1355343179843613,
+      "grad_norm": 0.07658053146961001,
+      "learning_rate": 8.158537204284145e-06,
+      "loss": 0.4899,
+      "step": 2289
+    },
+    {
+      "epoch": 1.1360307806876009,
+      "grad_norm": 0.07266971889913688,
+      "learning_rate": 8.157021666637783e-06,
+      "loss": 0.4585,
+      "step": 2290
+    },
+    {
+      "epoch": 1.1365272433908402,
+      "grad_norm": 0.07396138422667788,
+      "learning_rate": 8.155505646491282e-06,
+      "loss": 0.4892,
+      "step": 2291
+    },
+    {
+      "epoch": 1.1370237060940798,
+      "grad_norm": 0.07774988658009194,
+      "learning_rate": 8.153989144076335e-06,
+      "loss": 0.5008,
+      "step": 2292
+    },
+    {
+      "epoch": 1.137520168797319,
+      "grad_norm": 0.07188359362083234,
+      "learning_rate": 8.152472159624718e-06,
+      "loss": 0.5156,
+      "step": 2293
+    },
+    {
+      "epoch": 1.1380166315005584,
+      "grad_norm": 0.07587558951701104,
+      "learning_rate": 8.150954693368278e-06,
+      "loss": 0.4923,
+      "step": 2294
+    },
+    {
+      "epoch": 1.138513094203798,
+      "grad_norm": 0.07985374664068505,
+      "learning_rate": 8.149436745538934e-06,
+      "loss": 0.5302,
+      "step": 2295
+    },
+    {
+      "epoch": 1.1390095569070373,
+      "grad_norm": 0.07351485081740546,
+      "learning_rate": 8.14791831636868e-06,
+      "loss": 0.5146,
+      "step": 2296
+    },
+    {
+      "epoch": 1.1395060196102769,
+      "grad_norm": 0.07543744388533724,
+      "learning_rate": 8.146399406089587e-06,
+      "loss": 0.5112,
+      "step": 2297
+    },
+    {
+      "epoch": 1.1400024823135162,
+      "grad_norm": 0.07015283401717387,
+      "learning_rate": 8.144880014933791e-06,
+      "loss": 0.4671,
+      "step": 2298
+    },
+    {
+      "epoch": 1.1404989450167555,
+      "grad_norm": 0.07424863703283004,
+      "learning_rate": 8.143360143133512e-06,
+      "loss": 0.5043,
+      "step": 2299
+    },
+    {
+      "epoch": 1.140995407719995,
+      "grad_norm": 0.07182043679790903,
+      "learning_rate": 8.141839790921033e-06,
+      "loss": 0.4948,
+      "step": 2300
+    },
+    {
+      "epoch": 1.1414918704232344,
+      "grad_norm": 0.07320854472829157,
+      "learning_rate": 8.140318958528717e-06,
+      "loss": 0.4831,
+      "step": 2301
+    },
+    {
+      "epoch": 1.141988333126474,
+      "grad_norm": 0.07046526965939379,
+      "learning_rate": 8.138797646189e-06,
+      "loss": 0.5034,
+      "step": 2302
+    },
+    {
+      "epoch": 1.1424847958297133,
+      "grad_norm": 0.0738887751278947,
+      "learning_rate": 8.137275854134391e-06,
+      "loss": 0.5195,
+      "step": 2303
+    },
+    {
+      "epoch": 1.1429812585329526,
+      "grad_norm": 0.0734698833161908,
+      "learning_rate": 8.135753582597468e-06,
+      "loss": 0.5302,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1434777212361922,
+      "grad_norm": 0.07969162891614834,
+      "learning_rate": 8.13423083181089e-06,
+      "loss": 0.5185,
+      "step": 2305
+    },
+    {
+      "epoch": 1.1439741839394315,
+      "grad_norm": 0.07469688708542015,
+      "learning_rate": 8.132707602007381e-06,
+      "loss": 0.5084,
+      "step": 2306
+    },
+    {
+      "epoch": 1.144470646642671,
+      "grad_norm": 0.07579153342194478,
+      "learning_rate": 8.131183893419746e-06,
+      "loss": 0.505,
+      "step": 2307
+    },
+    {
+      "epoch": 1.1449671093459104,
+      "grad_norm": 0.06906370518919978,
+      "learning_rate": 8.129659706280856e-06,
+      "loss": 0.4814,
+      "step": 2308
+    },
+    {
+      "epoch": 1.1454635720491497,
+      "grad_norm": 0.07068549367332468,
+      "learning_rate": 8.128135040823661e-06,
+      "loss": 0.4601,
+      "step": 2309
+    },
+    {
+      "epoch": 1.1459600347523893,
+      "grad_norm": 0.07520549253264013,
+      "learning_rate": 8.12660989728118e-06,
+      "loss": 0.5018,
+      "step": 2310
+    },
+    {
+      "epoch": 1.1464564974556286,
+      "grad_norm": 0.07694693218495899,
+      "learning_rate": 8.125084275886507e-06,
+      "loss": 0.5548,
+      "step": 2311
+    },
+    {
+      "epoch": 1.1469529601588682,
+      "grad_norm": 0.08540895386379119,
+      "learning_rate": 8.123558176872812e-06,
+      "loss": 0.4933,
+      "step": 2312
+    },
+    {
+      "epoch": 1.1474494228621075,
+      "grad_norm": 0.07170201494666945,
+      "learning_rate": 8.12203160047333e-06,
+      "loss": 0.4806,
+      "step": 2313
+    },
+    {
+      "epoch": 1.1479458855653468,
+      "grad_norm": 0.07479344185796176,
+      "learning_rate": 8.120504546921377e-06,
+      "loss": 0.495,
+      "step": 2314
+    },
+    {
+      "epoch": 1.1484423482685864,
+      "grad_norm": 0.07586036128037843,
+      "learning_rate": 8.118977016450337e-06,
+      "loss": 0.4885,
+      "step": 2315
+    },
+    {
+      "epoch": 1.1489388109718257,
+      "grad_norm": 0.07298836535827842,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.4822,
+      "step": 2316
+    },
+    {
+      "epoch": 1.1494352736750653,
+      "grad_norm": 0.07424610160942582,
+      "learning_rate": 8.115920525684904e-06,
+      "loss": 0.4881,
+      "step": 2317
+    },
+    {
+      "epoch": 1.1499317363783046,
+      "grad_norm": 0.07711573726644844,
+      "learning_rate": 8.114391565857647e-06,
+      "loss": 0.4964,
+      "step": 2318
+    },
+    {
+      "epoch": 1.150428199081544,
+      "grad_norm": 0.07179898256858769,
+      "learning_rate": 8.112862130045574e-06,
+      "loss": 0.496,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1509246617847835,
+      "grad_norm": 0.06936652859448023,
+      "learning_rate": 8.111332218482436e-06,
+      "loss": 0.4672,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1514211244880228,
+      "grad_norm": 0.07455917920637714,
+      "learning_rate": 8.109801831402056e-06,
+      "loss": 0.4856,
+      "step": 2321
+    },
+    {
+      "epoch": 1.1519175871912624,
+      "grad_norm": 0.07669239126681408,
+      "learning_rate": 8.108270969038326e-06,
+      "loss": 0.467,
+      "step": 2322
+    },
+    {
+      "epoch": 1.1524140498945017,
+      "grad_norm": 0.0754927369489174,
+      "learning_rate": 8.106739631625216e-06,
+      "loss": 0.4973,
+      "step": 2323
+    },
+    {
+      "epoch": 1.152910512597741,
+      "grad_norm": 0.07373289963214065,
+      "learning_rate": 8.105207819396767e-06,
+      "loss": 0.4617,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1534069753009806,
+      "grad_norm": 0.07214062657617648,
+      "learning_rate": 8.10367553258709e-06,
+      "loss": 0.517,
+      "step": 2325
+    },
+    {
+      "epoch": 1.15390343800422,
+      "grad_norm": 0.07328961733541389,
+      "learning_rate": 8.102142771430373e-06,
+      "loss": 0.5141,
+      "step": 2326
+    },
+    {
+      "epoch": 1.1543999007074595,
+      "grad_norm": 0.07648118911976606,
+      "learning_rate": 8.100609536160871e-06,
+      "loss": 0.5044,
+      "step": 2327
+    },
+    {
+      "epoch": 1.1548963634106988,
+      "grad_norm": 0.07367333108806713,
+      "learning_rate": 8.099075827012917e-06,
+      "loss": 0.4904,
+      "step": 2328
+    },
+    {
+      "epoch": 1.1553928261139381,
+      "grad_norm": 0.07368647392363427,
+      "learning_rate": 8.097541644220912e-06,
+      "loss": 0.4628,
+      "step": 2329
+    },
+    {
+      "epoch": 1.1558892888171777,
+      "grad_norm": 0.07439036359702104,
+      "learning_rate": 8.096006988019331e-06,
+      "loss": 0.497,
+      "step": 2330
+    },
+    {
+      "epoch": 1.156385751520417,
+      "grad_norm": 0.0698387579277891,
+      "learning_rate": 8.094471858642726e-06,
+      "loss": 0.4735,
+      "step": 2331
+    },
+    {
+      "epoch": 1.1568822142236566,
+      "grad_norm": 0.07457852845134195,
+      "learning_rate": 8.092936256325709e-06,
+      "loss": 0.466,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1573786769268959,
+      "grad_norm": 0.07697323290341726,
+      "learning_rate": 8.09140018130298e-06,
+      "loss": 0.5094,
+      "step": 2333
+    },
+    {
+      "epoch": 1.1578751396301352,
+      "grad_norm": 0.06998446316294323,
+      "learning_rate": 8.089863633809298e-06,
+      "loss": 0.4626,
+      "step": 2334
+    },
+    {
+      "epoch": 1.1583716023333748,
+      "grad_norm": 0.07158518666753631,
+      "learning_rate": 8.088326614079503e-06,
+      "loss": 0.4773,
+      "step": 2335
+    },
+    {
+      "epoch": 1.158868065036614,
+      "grad_norm": 0.07123059869176439,
+      "learning_rate": 8.086789122348504e-06,
+      "loss": 0.4914,
+      "step": 2336
+    },
+    {
+      "epoch": 1.1593645277398537,
+      "grad_norm": 0.07132547049758975,
+      "learning_rate": 8.085251158851278e-06,
+      "loss": 0.4692,
+      "step": 2337
+    },
+    {
+      "epoch": 1.159860990443093,
+      "grad_norm": 0.07362286127579233,
+      "learning_rate": 8.08371272382288e-06,
+      "loss": 0.5461,
+      "step": 2338
+    },
+    {
+      "epoch": 1.1603574531463323,
+      "grad_norm": 0.07510848961258125,
+      "learning_rate": 8.08217381749844e-06,
+      "loss": 0.495,
+      "step": 2339
+    },
+    {
+      "epoch": 1.1608539158495719,
+      "grad_norm": 0.06994507718975675,
+      "learning_rate": 8.080634440113147e-06,
+      "loss": 0.4877,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1613503785528112,
+      "grad_norm": 0.07484978381716109,
+      "learning_rate": 8.079094591902275e-06,
+      "loss": 0.4768,
+      "step": 2341
+    },
+    {
+      "epoch": 1.1618468412560508,
+      "grad_norm": 0.07331691403913798,
+      "learning_rate": 8.077554273101165e-06,
+      "loss": 0.4996,
+      "step": 2342
+    },
+    {
+      "epoch": 1.16234330395929,
+      "grad_norm": 0.07655827308992964,
+      "learning_rate": 8.076013483945228e-06,
+      "loss": 0.5221,
+      "step": 2343
+    },
+    {
+      "epoch": 1.1628397666625294,
+      "grad_norm": 0.07121203009120934,
+      "learning_rate": 8.074472224669952e-06,
+      "loss": 0.4673,
+      "step": 2344
+    },
+    {
+      "epoch": 1.163336229365769,
+      "grad_norm": 0.07461640470824088,
+      "learning_rate": 8.072930495510888e-06,
+      "loss": 0.4544,
+      "step": 2345
+    },
+    {
+      "epoch": 1.1638326920690083,
+      "grad_norm": 0.07204944499795159,
+      "learning_rate": 8.071388296703672e-06,
+      "loss": 0.4868,
+      "step": 2346
+    },
+    {
+      "epoch": 1.1643291547722479,
+      "grad_norm": 0.0716988652101321,
+      "learning_rate": 8.069845628484002e-06,
+      "loss": 0.4686,
+      "step": 2347
+    },
+    {
+      "epoch": 1.1648256174754872,
+      "grad_norm": 0.077316414868866,
+      "learning_rate": 8.068302491087645e-06,
+      "loss": 0.4858,
+      "step": 2348
+    },
+    {
+      "epoch": 1.1653220801787265,
+      "grad_norm": 0.07517214266739652,
+      "learning_rate": 8.06675888475045e-06,
+      "loss": 0.5266,
+      "step": 2349
+    },
+    {
+      "epoch": 1.165818542881966,
+      "grad_norm": 0.07561697197455328,
+      "learning_rate": 8.065214809708332e-06,
+      "loss": 0.4932,
+      "step": 2350
+    },
+    {
+      "epoch": 1.1663150055852054,
+      "grad_norm": 0.07393071703889437,
+      "learning_rate": 8.063670266197278e-06,
+      "loss": 0.4945,
+      "step": 2351
+    },
+    {
+      "epoch": 1.166811468288445,
+      "grad_norm": 0.07018592225608884,
+      "learning_rate": 8.062125254453343e-06,
+      "loss": 0.4697,
+      "step": 2352
+    },
+    {
+      "epoch": 1.1673079309916843,
+      "grad_norm": 0.0704428825737941,
+      "learning_rate": 8.060579774712664e-06,
+      "loss": 0.4706,
+      "step": 2353
+    },
+    {
+      "epoch": 1.1678043936949236,
+      "grad_norm": 0.07140627571096919,
+      "learning_rate": 8.059033827211438e-06,
+      "loss": 0.496,
+      "step": 2354
+    },
+    {
+      "epoch": 1.1683008563981632,
+      "grad_norm": 0.0734276480348377,
+      "learning_rate": 8.057487412185937e-06,
+      "loss": 0.5266,
+      "step": 2355
+    },
+    {
+      "epoch": 1.1687973191014025,
+      "grad_norm": 0.0770116552694326,
+      "learning_rate": 8.055940529872512e-06,
+      "loss": 0.4968,
+      "step": 2356
+    },
+    {
+      "epoch": 1.1692937818046418,
+      "grad_norm": 0.0752006355628804,
+      "learning_rate": 8.054393180507572e-06,
+      "loss": 0.5169,
+      "step": 2357
+    },
+    {
+      "epoch": 1.1697902445078814,
+      "grad_norm": 0.07104098237478916,
+      "learning_rate": 8.052845364327609e-06,
+      "loss": 0.4847,
+      "step": 2358
+    },
+    {
+      "epoch": 1.1702867072111207,
+      "grad_norm": 0.0732561308111403,
+      "learning_rate": 8.05129708156918e-06,
+      "loss": 0.4757,
+      "step": 2359
+    },
+    {
+      "epoch": 1.1707831699143603,
+      "grad_norm": 0.0726216716118377,
+      "learning_rate": 8.049748332468917e-06,
+      "loss": 0.4785,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1712796326175996,
+      "grad_norm": 0.07151328576643422,
+      "learning_rate": 8.04819911726352e-06,
+      "loss": 0.5001,
+      "step": 2361
+    },
+    {
+      "epoch": 1.171776095320839,
+      "grad_norm": 0.0711654992225538,
+      "learning_rate": 8.046649436189763e-06,
+      "loss": 0.4915,
+      "step": 2362
+    },
+    {
+      "epoch": 1.1722725580240785,
+      "grad_norm": 0.07352594087862271,
+      "learning_rate": 8.045099289484488e-06,
+      "loss": 0.4671,
+      "step": 2363
+    },
+    {
+      "epoch": 1.1727690207273178,
+      "grad_norm": 0.07484080237772299,
+      "learning_rate": 8.043548677384611e-06,
+      "loss": 0.4861,
+      "step": 2364
+    },
+    {
+      "epoch": 1.1732654834305574,
+      "grad_norm": 0.07550160233111058,
+      "learning_rate": 8.041997600127118e-06,
+      "loss": 0.5023,
+      "step": 2365
+    },
+    {
+      "epoch": 1.1737619461337967,
+      "grad_norm": 0.07114299993253431,
+      "learning_rate": 8.040446057949067e-06,
+      "loss": 0.5116,
+      "step": 2366
+    },
+    {
+      "epoch": 1.174258408837036,
+      "grad_norm": 0.07022660054047274,
+      "learning_rate": 8.038894051087587e-06,
+      "loss": 0.4845,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1747548715402756,
+      "grad_norm": 0.08090894423687531,
+      "learning_rate": 8.037341579779875e-06,
+      "loss": 0.5432,
+      "step": 2368
+    },
+    {
+      "epoch": 1.175251334243515,
+      "grad_norm": 0.07285277362700689,
+      "learning_rate": 8.035788644263203e-06,
+      "loss": 0.4921,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1757477969467545,
+      "grad_norm": 0.07243641654934144,
+      "learning_rate": 8.034235244774911e-06,
+      "loss": 0.431,
+      "step": 2370
+    },
+    {
+      "epoch": 1.1762442596499938,
+      "grad_norm": 0.07669360482068058,
+      "learning_rate": 8.032681381552415e-06,
+      "loss": 0.5152,
+      "step": 2371
+    },
+    {
+      "epoch": 1.1767407223532331,
+      "grad_norm": 0.07436716811386304,
+      "learning_rate": 8.031127054833192e-06,
+      "loss": 0.5089,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1772371850564727,
+      "grad_norm": 0.0720366346955759,
+      "learning_rate": 8.029572264854799e-06,
+      "loss": 0.4832,
+      "step": 2373
+    },
+    {
+      "epoch": 1.177733647759712,
+      "grad_norm": 0.0702137374413376,
+      "learning_rate": 8.028017011854861e-06,
+      "loss": 0.4817,
+      "step": 2374
+    },
+    {
+      "epoch": 1.1782301104629516,
+      "grad_norm": 0.0720483512228606,
+      "learning_rate": 8.026461296071075e-06,
+      "loss": 0.4676,
+      "step": 2375
+    },
+    {
+      "epoch": 1.178726573166191,
+      "grad_norm": 0.07397180830163641,
+      "learning_rate": 8.024905117741204e-06,
+      "loss": 0.5442,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1792230358694302,
+      "grad_norm": 0.07716130964866487,
+      "learning_rate": 8.023348477103088e-06,
+      "loss": 0.5113,
+      "step": 2377
+    },
+    {
+      "epoch": 1.1797194985726698,
+      "grad_norm": 0.07335130875745567,
+      "learning_rate": 8.021791374394631e-06,
+      "loss": 0.5032,
+      "step": 2378
+    },
+    {
+      "epoch": 1.180215961275909,
+      "grad_norm": 0.07146794267621975,
+      "learning_rate": 8.020233809853815e-06,
+      "loss": 0.4932,
+      "step": 2379
+    },
+    {
+      "epoch": 1.1807124239791487,
+      "grad_norm": 0.07252371872661585,
+      "learning_rate": 8.018675783718686e-06,
+      "loss": 0.5086,
+      "step": 2380
+    },
+    {
+      "epoch": 1.181208886682388,
+      "grad_norm": 0.07261609295457848,
+      "learning_rate": 8.017117296227364e-06,
+      "loss": 0.4856,
+      "step": 2381
+    },
+    {
+      "epoch": 1.1817053493856273,
+      "grad_norm": 0.07734803531243527,
+      "learning_rate": 8.015558347618039e-06,
+      "loss": 0.5243,
+      "step": 2382
+    },
+    {
+      "epoch": 1.1822018120888669,
+      "grad_norm": 0.06979903430013636,
+      "learning_rate": 8.013998938128973e-06,
+      "loss": 0.4699,
+      "step": 2383
+    },
+    {
+      "epoch": 1.1826982747921062,
+      "grad_norm": 0.07386585758579457,
+      "learning_rate": 8.012439067998494e-06,
+      "loss": 0.4932,
+      "step": 2384
+    },
+    {
+      "epoch": 1.1831947374953455,
+      "grad_norm": 0.07541637636041616,
+      "learning_rate": 8.010878737465003e-06,
+      "loss": 0.5202,
+      "step": 2385
+    },
+    {
+      "epoch": 1.183691200198585,
+      "grad_norm": 0.07199214991500814,
+      "learning_rate": 8.009317946766975e-06,
+      "loss": 0.4632,
+      "step": 2386
+    },
+    {
+      "epoch": 1.1841876629018244,
+      "grad_norm": 0.07387274022313904,
+      "learning_rate": 8.007756696142948e-06,
+      "loss": 0.4928,
+      "step": 2387
+    },
+    {
+      "epoch": 1.184684125605064,
+      "grad_norm": 0.0709506582818121,
+      "learning_rate": 8.006194985831537e-06,
+      "loss": 0.4685,
+      "step": 2388
+    },
+    {
+      "epoch": 1.1851805883083033,
+      "grad_norm": 0.07741639300132216,
+      "learning_rate": 8.004632816071422e-06,
+      "loss": 0.4852,
+      "step": 2389
+    },
+    {
+      "epoch": 1.1856770510115426,
+      "grad_norm": 0.07182104672096723,
+      "learning_rate": 8.003070187101356e-06,
+      "loss": 0.4759,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1861735137147822,
+      "grad_norm": 0.0745573604205288,
+      "learning_rate": 8.001507099160164e-06,
+      "loss": 0.4728,
+      "step": 2391
+    },
+    {
+      "epoch": 1.1866699764180215,
+      "grad_norm": 0.07176520107798823,
+      "learning_rate": 7.999943552486737e-06,
+      "loss": 0.4857,
+      "step": 2392
+    },
+    {
+      "epoch": 1.187166439121261,
+      "grad_norm": 0.07181856926868868,
+      "learning_rate": 7.998379547320038e-06,
+      "loss": 0.4712,
+      "step": 2393
+    },
+    {
+      "epoch": 1.1876629018245004,
+      "grad_norm": 0.07050318637166479,
+      "learning_rate": 7.996815083899102e-06,
+      "loss": 0.4546,
+      "step": 2394
+    },
+    {
+      "epoch": 1.1881593645277397,
+      "grad_norm": 0.07293329762823501,
+      "learning_rate": 7.995250162463028e-06,
+      "loss": 0.4993,
+      "step": 2395
+    },
+    {
+      "epoch": 1.1886558272309793,
+      "grad_norm": 0.07689205680729931,
+      "learning_rate": 7.993684783250994e-06,
+      "loss": 0.4835,
+      "step": 2396
+    },
+    {
+      "epoch": 1.1891522899342186,
+      "grad_norm": 0.07560318192495996,
+      "learning_rate": 7.99211894650224e-06,
+      "loss": 0.461,
+      "step": 2397
+    },
+    {
+      "epoch": 1.1896487526374582,
+      "grad_norm": 0.07309599735206149,
+      "learning_rate": 7.99055265245608e-06,
+      "loss": 0.4779,
+      "step": 2398
+    },
+    {
+      "epoch": 1.1901452153406975,
+      "grad_norm": 0.07095152261698759,
+      "learning_rate": 7.988985901351898e-06,
+      "loss": 0.4905,
+      "step": 2399
+    },
+    {
+      "epoch": 1.1906416780439368,
+      "grad_norm": 0.07164890116119788,
+      "learning_rate": 7.987418693429145e-06,
+      "loss": 0.5099,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1911381407471764,
+      "grad_norm": 0.07124693531064728,
+      "learning_rate": 7.985851028927344e-06,
+      "loss": 0.4604,
+      "step": 2401
+    },
+    {
+      "epoch": 1.1916346034504157,
+      "grad_norm": 0.07435149050158306,
+      "learning_rate": 7.98428290808609e-06,
+      "loss": 0.4801,
+      "step": 2402
+    },
+    {
+      "epoch": 1.1921310661536553,
+      "grad_norm": 0.07294313694584678,
+      "learning_rate": 7.98271433114504e-06,
+      "loss": 0.509,
+      "step": 2403
+    },
+    {
+      "epoch": 1.1926275288568946,
+      "grad_norm": 0.06978769812413184,
+      "learning_rate": 7.981145298343929e-06,
+      "loss": 0.4892,
+      "step": 2404
+    },
+    {
+      "epoch": 1.193123991560134,
+      "grad_norm": 0.06912663308966963,
+      "learning_rate": 7.979575809922559e-06,
+      "loss": 0.4898,
+      "step": 2405
+    },
+    {
+      "epoch": 1.1936204542633735,
+      "grad_norm": 0.07301215898745836,
+      "learning_rate": 7.9780058661208e-06,
+      "loss": 0.4756,
+      "step": 2406
+    },
+    {
+      "epoch": 1.1941169169666128,
+      "grad_norm": 0.07229724679847097,
+      "learning_rate": 7.976435467178592e-06,
+      "loss": 0.5056,
+      "step": 2407
+    },
+    {
+      "epoch": 1.1946133796698524,
+      "grad_norm": 0.07301499091902912,
+      "learning_rate": 7.97486461333595e-06,
+      "loss": 0.5187,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1951098423730917,
+      "grad_norm": 0.07395695237003778,
+      "learning_rate": 7.973293304832946e-06,
+      "loss": 0.4859,
+      "step": 2409
+    },
+    {
+      "epoch": 1.195606305076331,
+      "grad_norm": 0.07147072664149673,
+      "learning_rate": 7.971721541909734e-06,
+      "loss": 0.4762,
+      "step": 2410
+    },
+    {
+      "epoch": 1.1961027677795706,
+      "grad_norm": 0.0710124719486651,
+      "learning_rate": 7.970149324806535e-06,
+      "loss": 0.4821,
+      "step": 2411
+    },
+    {
+      "epoch": 1.19659923048281,
+      "grad_norm": 0.06924011820147728,
+      "learning_rate": 7.968576653763633e-06,
+      "loss": 0.4777,
+      "step": 2412
+    },
+    {
+      "epoch": 1.1970956931860495,
+      "grad_norm": 0.07631306801702314,
+      "learning_rate": 7.967003529021386e-06,
+      "loss": 0.4963,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1975921558892888,
+      "grad_norm": 0.0724947892734663,
+      "learning_rate": 7.965429950820222e-06,
+      "loss": 0.4936,
+      "step": 2414
+    },
+    {
+      "epoch": 1.1980886185925281,
+      "grad_norm": 0.07164010492469318,
+      "learning_rate": 7.963855919400639e-06,
+      "loss": 0.4679,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1985850812957677,
+      "grad_norm": 0.06825874139662513,
+      "learning_rate": 7.962281435003199e-06,
+      "loss": 0.4449,
+      "step": 2416
+    },
+    {
+      "epoch": 1.199081543999007,
+      "grad_norm": 0.07409910389087015,
+      "learning_rate": 7.960706497868537e-06,
+      "loss": 0.4958,
+      "step": 2417
+    },
+    {
+      "epoch": 1.1995780067022466,
+      "grad_norm": 0.0739553558009401,
+      "learning_rate": 7.959131108237361e-06,
+      "loss": 0.5019,
+      "step": 2418
+    },
+    {
+      "epoch": 1.200074469405486,
+      "grad_norm": 0.07170215535474178,
+      "learning_rate": 7.95755526635044e-06,
+      "loss": 0.4887,
+      "step": 2419
+    },
+    {
+      "epoch": 1.2005709321087252,
+      "grad_norm": 0.07683710435298069,
+      "learning_rate": 7.955978972448618e-06,
+      "loss": 0.4925,
+      "step": 2420
+    },
+    {
+      "epoch": 1.2010673948119648,
+      "grad_norm": 0.07377097983405335,
+      "learning_rate": 7.954402226772804e-06,
+      "loss": 0.4788,
+      "step": 2421
+    },
+    {
+      "epoch": 1.2015638575152041,
+      "grad_norm": 0.07655586471619548,
+      "learning_rate": 7.95282502956398e-06,
+      "loss": 0.5347,
+      "step": 2422
+    },
+    {
+      "epoch": 1.2020603202184437,
+      "grad_norm": 0.07406558881865472,
+      "learning_rate": 7.951247381063195e-06,
+      "loss": 0.5173,
+      "step": 2423
+    },
+    {
+      "epoch": 1.202556782921683,
+      "grad_norm": 0.07322830499207313,
+      "learning_rate": 7.949669281511569e-06,
+      "loss": 0.4733,
+      "step": 2424
+    },
+    {
+      "epoch": 1.2030532456249223,
+      "grad_norm": 0.07039907851339344,
+      "learning_rate": 7.948090731150287e-06,
+      "loss": 0.477,
+      "step": 2425
+    },
+    {
+      "epoch": 1.2035497083281619,
+      "grad_norm": 0.07037418167901223,
+      "learning_rate": 7.946511730220605e-06,
+      "loss": 0.4692,
+      "step": 2426
+    },
+    {
+      "epoch": 1.2040461710314012,
+      "grad_norm": 0.07368459831288712,
+      "learning_rate": 7.94493227896385e-06,
+      "loss": 0.4857,
+      "step": 2427
+    },
+    {
+      "epoch": 1.2045426337346408,
+      "grad_norm": 0.07426055953297137,
+      "learning_rate": 7.943352377621414e-06,
+      "loss": 0.489,
+      "step": 2428
+    },
+    {
+      "epoch": 1.20503909643788,
+      "grad_norm": 0.06945786010896897,
+      "learning_rate": 7.941772026434759e-06,
+      "loss": 0.4716,
+      "step": 2429
+    },
+    {
+      "epoch": 1.2055355591411194,
+      "grad_norm": 0.07502359795174329,
+      "learning_rate": 7.94019122564542e-06,
+      "loss": 0.4873,
+      "step": 2430
+    },
+    {
+      "epoch": 1.206032021844359,
+      "grad_norm": 0.07253641226610841,
+      "learning_rate": 7.938609975494992e-06,
+      "loss": 0.4767,
+      "step": 2431
+    },
+    {
+      "epoch": 1.2065284845475983,
+      "grad_norm": 0.07172366435546958,
+      "learning_rate": 7.937028276225149e-06,
+      "loss": 0.475,
+      "step": 2432
+    },
+    {
+      "epoch": 1.2070249472508379,
+      "grad_norm": 0.07075938334855661,
+      "learning_rate": 7.935446128077624e-06,
+      "loss": 0.4911,
+      "step": 2433
+    },
+    {
+      "epoch": 1.2075214099540772,
+      "grad_norm": 0.0729907659202211,
+      "learning_rate": 7.933863531294224e-06,
+      "loss": 0.4841,
+      "step": 2434
+    },
+    {
+      "epoch": 1.2080178726573165,
+      "grad_norm": 0.07374115829562204,
+      "learning_rate": 7.932280486116825e-06,
+      "loss": 0.4853,
+      "step": 2435
+    },
+    {
+      "epoch": 1.208514335360556,
+      "grad_norm": 0.07806369217645258,
+      "learning_rate": 7.93069699278737e-06,
+      "loss": 0.5576,
+      "step": 2436
+    },
+    {
+      "epoch": 1.2090107980637954,
+      "grad_norm": 0.0718331092080611,
+      "learning_rate": 7.92911305154787e-06,
+      "loss": 0.4735,
+      "step": 2437
+    },
+    {
+      "epoch": 1.209507260767035,
+      "grad_norm": 0.06731176175485117,
+      "learning_rate": 7.927528662640402e-06,
+      "loss": 0.4893,
+      "step": 2438
+    },
+    {
+      "epoch": 1.2100037234702743,
+      "grad_norm": 0.07186021670426744,
+      "learning_rate": 7.925943826307119e-06,
+      "loss": 0.5123,
+      "step": 2439
+    },
+    {
+      "epoch": 1.2105001861735136,
+      "grad_norm": 0.07473382580979585,
+      "learning_rate": 7.924358542790236e-06,
+      "loss": 0.5024,
+      "step": 2440
+    },
+    {
+      "epoch": 1.2109966488767532,
+      "grad_norm": 0.07386057347004561,
+      "learning_rate": 7.922772812332038e-06,
+      "loss": 0.4854,
+      "step": 2441
+    },
+    {
+      "epoch": 1.2114931115799925,
+      "grad_norm": 0.07206280929710557,
+      "learning_rate": 7.92118663517488e-06,
+      "loss": 0.5094,
+      "step": 2442
+    },
+    {
+      "epoch": 1.211989574283232,
+      "grad_norm": 0.07061427887302714,
+      "learning_rate": 7.919600011561181e-06,
+      "loss": 0.4914,
+      "step": 2443
+    },
+    {
+      "epoch": 1.2124860369864714,
+      "grad_norm": 0.07364474032932036,
+      "learning_rate": 7.918012941733434e-06,
+      "loss": 0.5141,
+      "step": 2444
+    },
+    {
+      "epoch": 1.2129824996897107,
+      "grad_norm": 0.07157728506982447,
+      "learning_rate": 7.916425425934195e-06,
+      "loss": 0.5009,
+      "step": 2445
+    },
+    {
+      "epoch": 1.2134789623929503,
+      "grad_norm": 0.07574472192509478,
+      "learning_rate": 7.91483746440609e-06,
+      "loss": 0.4826,
+      "step": 2446
+    },
+    {
+      "epoch": 1.2139754250961896,
+      "grad_norm": 0.07518967003949961,
+      "learning_rate": 7.913249057391815e-06,
+      "loss": 0.5062,
+      "step": 2447
+    },
+    {
+      "epoch": 1.2144718877994292,
+      "grad_norm": 0.07260291659745174,
+      "learning_rate": 7.911660205134132e-06,
+      "loss": 0.5081,
+      "step": 2448
+    },
+    {
+      "epoch": 1.2149683505026685,
+      "grad_norm": 0.07482200230732527,
+      "learning_rate": 7.910070907875871e-06,
+      "loss": 0.4947,
+      "step": 2449
+    },
+    {
+      "epoch": 1.2154648132059078,
+      "grad_norm": 0.07251676217475807,
+      "learning_rate": 7.90848116585993e-06,
+      "loss": 0.479,
+      "step": 2450
+    },
+    {
+      "epoch": 1.2159612759091474,
+      "grad_norm": 0.0737268000170673,
+      "learning_rate": 7.906890979329282e-06,
+      "loss": 0.5006,
+      "step": 2451
+    },
+    {
+      "epoch": 1.2164577386123867,
+      "grad_norm": 0.07387770278193574,
+      "learning_rate": 7.905300348526951e-06,
+      "loss": 0.4883,
+      "step": 2452
+    },
+    {
+      "epoch": 1.2169542013156263,
+      "grad_norm": 0.07216039271740472,
+      "learning_rate": 7.903709273696047e-06,
+      "loss": 0.469,
+      "step": 2453
+    },
+    {
+      "epoch": 1.2174506640188656,
+      "grad_norm": 0.07259470892183868,
+      "learning_rate": 7.902117755079738e-06,
+      "loss": 0.4811,
+      "step": 2454
+    },
+    {
+      "epoch": 1.217947126722105,
+      "grad_norm": 0.07458876832011727,
+      "learning_rate": 7.90052579292126e-06,
+      "loss": 0.5109,
+      "step": 2455
+    },
+    {
+      "epoch": 1.2184435894253445,
+      "grad_norm": 0.06901029085737141,
+      "learning_rate": 7.898933387463924e-06,
+      "loss": 0.5012,
+      "step": 2456
+    },
+    {
+      "epoch": 1.2189400521285838,
+      "grad_norm": 0.07324817525795049,
+      "learning_rate": 7.897340538951099e-06,
+      "loss": 0.4823,
+      "step": 2457
+    },
+    {
+      "epoch": 1.2194365148318234,
+      "grad_norm": 0.08003178451935818,
+      "learning_rate": 7.895747247626228e-06,
+      "loss": 0.5125,
+      "step": 2458
+    },
+    {
+      "epoch": 1.2199329775350627,
+      "grad_norm": 0.07274596398321198,
+      "learning_rate": 7.89415351373282e-06,
+      "loss": 0.4934,
+      "step": 2459
+    },
+    {
+      "epoch": 1.220429440238302,
+      "grad_norm": 0.06964652576032265,
+      "learning_rate": 7.892559337514451e-06,
+      "loss": 0.4658,
+      "step": 2460
+    },
+    {
+      "epoch": 1.2209259029415416,
+      "grad_norm": 0.07116022602457564,
+      "learning_rate": 7.890964719214767e-06,
+      "loss": 0.4764,
+      "step": 2461
+    },
+    {
+      "epoch": 1.221422365644781,
+      "grad_norm": 0.06952619399608187,
+      "learning_rate": 7.88936965907748e-06,
+      "loss": 0.4558,
+      "step": 2462
+    },
+    {
+      "epoch": 1.2219188283480205,
+      "grad_norm": 0.07518655995107235,
+      "learning_rate": 7.887774157346365e-06,
+      "loss": 0.4865,
+      "step": 2463
+    },
+    {
+      "epoch": 1.2224152910512598,
+      "grad_norm": 0.07376865854194593,
+      "learning_rate": 7.886178214265274e-06,
+      "loss": 0.4992,
+      "step": 2464
+    },
+    {
+      "epoch": 1.2229117537544991,
+      "grad_norm": 0.07342072581864792,
+      "learning_rate": 7.884581830078118e-06,
+      "loss": 0.5262,
+      "step": 2465
+    },
+    {
+      "epoch": 1.2234082164577387,
+      "grad_norm": 0.07167032318175433,
+      "learning_rate": 7.88298500502888e-06,
+      "loss": 0.5072,
+      "step": 2466
+    },
+    {
+      "epoch": 1.223904679160978,
+      "grad_norm": 0.07097269675747088,
+      "learning_rate": 7.88138773936161e-06,
+      "loss": 0.4793,
+      "step": 2467
+    },
+    {
+      "epoch": 1.2244011418642176,
+      "grad_norm": 0.07013672385308536,
+      "learning_rate": 7.879790033320424e-06,
+      "loss": 0.4929,
+      "step": 2468
+    },
+    {
+      "epoch": 1.2248976045674569,
+      "grad_norm": 0.07223102583442119,
+      "learning_rate": 7.878191887149504e-06,
+      "loss": 0.511,
+      "step": 2469
+    },
+    {
+      "epoch": 1.2253940672706962,
+      "grad_norm": 0.07669917513271436,
+      "learning_rate": 7.876593301093104e-06,
+      "loss": 0.5289,
+      "step": 2470
+    },
+    {
+      "epoch": 1.2258905299739358,
+      "grad_norm": 0.07446008605700466,
+      "learning_rate": 7.87499427539554e-06,
+      "loss": 0.5093,
+      "step": 2471
+    },
+    {
+      "epoch": 1.226386992677175,
+      "grad_norm": 0.07618064450545196,
+      "learning_rate": 7.873394810301198e-06,
+      "loss": 0.4884,
+      "step": 2472
+    },
+    {
+      "epoch": 1.2268834553804147,
+      "grad_norm": 0.07167284442434069,
+      "learning_rate": 7.87179490605453e-06,
+      "loss": 0.4948,
+      "step": 2473
+    },
+    {
+      "epoch": 1.227379918083654,
+      "grad_norm": 0.07401798466940147,
+      "learning_rate": 7.870194562900055e-06,
+      "loss": 0.4849,
+      "step": 2474
+    },
+    {
+      "epoch": 1.2278763807868933,
+      "grad_norm": 0.07186829556894118,
+      "learning_rate": 7.868593781082364e-06,
+      "loss": 0.5079,
+      "step": 2475
+    },
+    {
+      "epoch": 1.2283728434901329,
+      "grad_norm": 0.06863177399733751,
+      "learning_rate": 7.866992560846107e-06,
+      "loss": 0.4458,
+      "step": 2476
+    },
+    {
+      "epoch": 1.2288693061933722,
+      "grad_norm": 0.07297315411470365,
+      "learning_rate": 7.865390902436005e-06,
+      "loss": 0.4827,
+      "step": 2477
+    },
+    {
+      "epoch": 1.2293657688966118,
+      "grad_norm": 0.07130156565218151,
+      "learning_rate": 7.863788806096847e-06,
+      "loss": 0.4858,
+      "step": 2478
+    },
+    {
+      "epoch": 1.229862231599851,
+      "grad_norm": 0.07309681317328279,
+      "learning_rate": 7.862186272073489e-06,
+      "loss": 0.4791,
+      "step": 2479
+    },
+    {
+      "epoch": 1.2303586943030904,
+      "grad_norm": 0.07059455181087342,
+      "learning_rate": 7.860583300610849e-06,
+      "loss": 0.4786,
+      "step": 2480
+    },
+    {
+      "epoch": 1.23085515700633,
+      "grad_norm": 0.07278780610477634,
+      "learning_rate": 7.858979891953918e-06,
+      "loss": 0.487,
+      "step": 2481
+    },
+    {
+      "epoch": 1.2313516197095693,
+      "grad_norm": 0.07228312983604328,
+      "learning_rate": 7.85737604634775e-06,
+      "loss": 0.4954,
+      "step": 2482
+    },
+    {
+      "epoch": 1.2318480824128089,
+      "grad_norm": 0.07191368746552866,
+      "learning_rate": 7.85577176403747e-06,
+      "loss": 0.4872,
+      "step": 2483
+    },
+    {
+      "epoch": 1.2323445451160482,
+      "grad_norm": 0.07136655375077475,
+      "learning_rate": 7.854167045268265e-06,
+      "loss": 0.502,
+      "step": 2484
+    },
+    {
+      "epoch": 1.2328410078192875,
+      "grad_norm": 0.07208287450277875,
+      "learning_rate": 7.852561890285385e-06,
+      "loss": 0.4857,
+      "step": 2485
+    },
+    {
+      "epoch": 1.233337470522527,
+      "grad_norm": 0.07296424887863194,
+      "learning_rate": 7.850956299334162e-06,
+      "loss": 0.5175,
+      "step": 2486
+    },
+    {
+      "epoch": 1.2338339332257664,
+      "grad_norm": 0.07366881702863355,
+      "learning_rate": 7.84935027265998e-06,
+      "loss": 0.5011,
+      "step": 2487
+    },
+    {
+      "epoch": 1.234330395929006,
+      "grad_norm": 0.07599137022216883,
+      "learning_rate": 7.847743810508292e-06,
+      "loss": 0.5042,
+      "step": 2488
+    },
+    {
+      "epoch": 1.2348268586322453,
+      "grad_norm": 0.07111617274127456,
+      "learning_rate": 7.846136913124627e-06,
+      "loss": 0.4698,
+      "step": 2489
+    },
+    {
+      "epoch": 1.2353233213354846,
+      "grad_norm": 0.07201232200572243,
+      "learning_rate": 7.844529580754566e-06,
+      "loss": 0.472,
+      "step": 2490
+    },
+    {
+      "epoch": 1.2358197840387242,
+      "grad_norm": 0.07341460404524983,
+      "learning_rate": 7.842921813643767e-06,
+      "loss": 0.4736,
+      "step": 2491
+    },
+    {
+      "epoch": 1.2363162467419635,
+      "grad_norm": 0.07516599673488107,
+      "learning_rate": 7.841313612037953e-06,
+      "loss": 0.536,
+      "step": 2492
+    },
+    {
+      "epoch": 1.236812709445203,
+      "grad_norm": 0.07612536960775762,
+      "learning_rate": 7.83970497618291e-06,
+      "loss": 0.5107,
+      "step": 2493
+    },
+    {
+      "epoch": 1.2373091721484424,
+      "grad_norm": 0.0712196177649546,
+      "learning_rate": 7.838095906324493e-06,
+      "loss": 0.5272,
+      "step": 2494
+    },
+    {
+      "epoch": 1.2378056348516817,
+      "grad_norm": 0.07174382506085918,
+      "learning_rate": 7.83648640270862e-06,
+      "loss": 0.4699,
+      "step": 2495
+    },
+    {
+      "epoch": 1.2383020975549213,
+      "grad_norm": 0.07176500264816929,
+      "learning_rate": 7.834876465581283e-06,
+      "loss": 0.4689,
+      "step": 2496
+    },
+    {
+      "epoch": 1.2387985602581606,
+      "grad_norm": 0.07375994219799964,
+      "learning_rate": 7.83326609518853e-06,
+      "loss": 0.4781,
+      "step": 2497
+    },
+    {
+      "epoch": 1.2392950229614,
+      "grad_norm": 0.0706960013355042,
+      "learning_rate": 7.831655291776484e-06,
+      "loss": 0.493,
+      "step": 2498
+    },
+    {
+      "epoch": 1.2397914856646395,
+      "grad_norm": 0.07198655800007961,
+      "learning_rate": 7.830044055591326e-06,
+      "loss": 0.4644,
+      "step": 2499
+    },
+    {
+      "epoch": 1.2402879483678788,
+      "grad_norm": 0.07948572448179321,
+      "learning_rate": 7.828432386879314e-06,
+      "loss": 0.5537,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2407844110711184,
+      "grad_norm": 0.07649974228423813,
+      "learning_rate": 7.82682028588676e-06,
+      "loss": 0.508,
+      "step": 2501
+    },
+    {
+      "epoch": 1.2412808737743577,
+      "grad_norm": 0.07004967280284491,
+      "learning_rate": 7.82520775286005e-06,
+      "loss": 0.4786,
+      "step": 2502
+    },
+    {
+      "epoch": 1.241777336477597,
+      "grad_norm": 0.07152110383482235,
+      "learning_rate": 7.823594788045633e-06,
+      "loss": 0.5309,
+      "step": 2503
+    },
+    {
+      "epoch": 1.2422737991808366,
+      "grad_norm": 0.07063608077865494,
+      "learning_rate": 7.821981391690026e-06,
+      "loss": 0.4706,
+      "step": 2504
+    },
+    {
+      "epoch": 1.242770261884076,
+      "grad_norm": 0.07701316889800286,
+      "learning_rate": 7.82036756403981e-06,
+      "loss": 0.5144,
+      "step": 2505
+    },
+    {
+      "epoch": 1.2432667245873155,
+      "grad_norm": 0.07226679143037919,
+      "learning_rate": 7.818753305341635e-06,
+      "loss": 0.5024,
+      "step": 2506
+    },
+    {
+      "epoch": 1.2437631872905548,
+      "grad_norm": 0.07865435918533266,
+      "learning_rate": 7.817138615842212e-06,
+      "loss": 0.5544,
+      "step": 2507
+    },
+    {
+      "epoch": 1.2442596499937941,
+      "grad_norm": 0.07980179701888628,
+      "learning_rate": 7.81552349578832e-06,
+      "loss": 0.5101,
+      "step": 2508
+    },
+    {
+      "epoch": 1.2447561126970337,
+      "grad_norm": 0.0720486211523932,
+      "learning_rate": 7.813907945426806e-06,
+      "loss": 0.4818,
+      "step": 2509
+    },
+    {
+      "epoch": 1.245252575400273,
+      "grad_norm": 0.0697060574436996,
+      "learning_rate": 7.81229196500458e-06,
+      "loss": 0.4905,
+      "step": 2510
+    },
+    {
+      "epoch": 1.2457490381035126,
+      "grad_norm": 0.0712762778017296,
+      "learning_rate": 7.810675554768616e-06,
+      "loss": 0.4704,
+      "step": 2511
+    },
+    {
+      "epoch": 1.246245500806752,
+      "grad_norm": 0.07355863446675028,
+      "learning_rate": 7.809058714965962e-06,
+      "loss": 0.479,
+      "step": 2512
+    },
+    {
+      "epoch": 1.2467419635099912,
+      "grad_norm": 0.07255719671250284,
+      "learning_rate": 7.807441445843723e-06,
+      "loss": 0.4959,
+      "step": 2513
+    },
+    {
+      "epoch": 1.2472384262132308,
+      "grad_norm": 0.07805962130769145,
+      "learning_rate": 7.805823747649073e-06,
+      "loss": 0.4964,
+      "step": 2514
+    },
+    {
+      "epoch": 1.24773488891647,
+      "grad_norm": 0.07922959031722938,
+      "learning_rate": 7.80420562062925e-06,
+      "loss": 0.4982,
+      "step": 2515
+    },
+    {
+      "epoch": 1.2482313516197097,
+      "grad_norm": 0.07529660466681241,
+      "learning_rate": 7.802587065031561e-06,
+      "loss": 0.4909,
+      "step": 2516
+    },
+    {
+      "epoch": 1.248727814322949,
+      "grad_norm": 0.07433614085477507,
+      "learning_rate": 7.800968081103375e-06,
+      "loss": 0.5054,
+      "step": 2517
+    },
+    {
+      "epoch": 1.2492242770261883,
+      "grad_norm": 0.07091809704367669,
+      "learning_rate": 7.799348669092128e-06,
+      "loss": 0.4713,
+      "step": 2518
+    },
+    {
+      "epoch": 1.2497207397294279,
+      "grad_norm": 0.07280705234842633,
+      "learning_rate": 7.797728829245321e-06,
+      "loss": 0.4936,
+      "step": 2519
+    },
+    {
+      "epoch": 1.2502172024326672,
+      "grad_norm": 0.07158759022082693,
+      "learning_rate": 7.79610856181052e-06,
+      "loss": 0.4401,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2502172024326672,
+      "eval_loss": 0.5242970585823059,
+      "eval_runtime": 259.1392,
+      "eval_samples_per_second": 117.13,
+      "eval_steps_per_second": 14.645,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2507136651359065,
+      "grad_norm": 0.07377314632312748,
+      "learning_rate": 7.794487867035358e-06,
+      "loss": 0.4947,
+      "step": 2521
+    },
+    {
+      "epoch": 1.251210127839146,
+      "grad_norm": 0.07476997296963254,
+      "learning_rate": 7.792866745167532e-06,
+      "loss": 0.4837,
+      "step": 2522
+    },
+    {
+      "epoch": 1.2517065905423854,
+      "grad_norm": 0.06860856143991592,
+      "learning_rate": 7.791245196454803e-06,
+      "loss": 0.4677,
+      "step": 2523
+    },
+    {
+      "epoch": 1.252203053245625,
+      "grad_norm": 0.07706477950077137,
+      "learning_rate": 7.789623221145002e-06,
+      "loss": 0.5447,
+      "step": 2524
+    },
+    {
+      "epoch": 1.2526995159488643,
+      "grad_norm": 0.07191297047243664,
+      "learning_rate": 7.788000819486019e-06,
+      "loss": 0.4662,
+      "step": 2525
+    },
+    {
+      "epoch": 1.2531959786521036,
+      "grad_norm": 0.07723494482816311,
+      "learning_rate": 7.786377991725813e-06,
+      "loss": 0.5004,
+      "step": 2526
+    },
+    {
+      "epoch": 1.2536924413553432,
+      "grad_norm": 0.07573264886896905,
+      "learning_rate": 7.784754738112406e-06,
+      "loss": 0.4696,
+      "step": 2527
+    },
+    {
+      "epoch": 1.2541889040585825,
+      "grad_norm": 0.07425027782995233,
+      "learning_rate": 7.783131058893889e-06,
+      "loss": 0.5003,
+      "step": 2528
+    },
+    {
+      "epoch": 1.254685366761822,
+      "grad_norm": 0.07825615066186287,
+      "learning_rate": 7.781506954318413e-06,
+      "loss": 0.5098,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2551818294650614,
+      "grad_norm": 0.08341839776678486,
+      "learning_rate": 7.779882424634197e-06,
+      "loss": 0.5447,
+      "step": 2530
+    },
+    {
+      "epoch": 1.2556782921683007,
+      "grad_norm": 0.07146841653369992,
+      "learning_rate": 7.778257470089524e-06,
+      "loss": 0.5097,
+      "step": 2531
+    },
+    {
+      "epoch": 1.2561747548715403,
+      "grad_norm": 0.07257801387976101,
+      "learning_rate": 7.776632090932745e-06,
+      "loss": 0.4839,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2566712175747796,
+      "grad_norm": 0.0735635305923275,
+      "learning_rate": 7.775006287412268e-06,
+      "loss": 0.4542,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2571676802780192,
+      "grad_norm": 0.07223004129649539,
+      "learning_rate": 7.773380059776575e-06,
+      "loss": 0.4778,
+      "step": 2534
+    },
+    {
+      "epoch": 1.2576641429812585,
+      "grad_norm": 0.07769649472723285,
+      "learning_rate": 7.771753408274208e-06,
+      "loss": 0.5006,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2581606056844978,
+      "grad_norm": 0.0733052305747097,
+      "learning_rate": 7.770126333153772e-06,
+      "loss": 0.5167,
+      "step": 2536
+    },
+    {
+      "epoch": 1.2586570683877374,
+      "grad_norm": 0.07351854004910238,
+      "learning_rate": 7.768498834663945e-06,
+      "loss": 0.4859,
+      "step": 2537
+    },
+    {
+      "epoch": 1.2591535310909767,
+      "grad_norm": 0.07657201822534493,
+      "learning_rate": 7.766870913053456e-06,
+      "loss": 0.4979,
+      "step": 2538
+    },
+    {
+      "epoch": 1.2596499937942163,
+      "grad_norm": 0.0742793733950408,
+      "learning_rate": 7.765242568571116e-06,
+      "loss": 0.5173,
+      "step": 2539
+    },
+    {
+      "epoch": 1.2601464564974556,
+      "grad_norm": 0.07628534364392862,
+      "learning_rate": 7.763613801465785e-06,
+      "loss": 0.4761,
+      "step": 2540
+    },
+    {
+      "epoch": 1.260642919200695,
+      "grad_norm": 0.07099537482619052,
+      "learning_rate": 7.761984611986396e-06,
+      "loss": 0.4917,
+      "step": 2541
+    },
+    {
+      "epoch": 1.2611393819039345,
+      "grad_norm": 0.07077673464762375,
+      "learning_rate": 7.760355000381942e-06,
+      "loss": 0.5155,
+      "step": 2542
+    },
+    {
+      "epoch": 1.2616358446071738,
+      "grad_norm": 0.07090392931999898,
+      "learning_rate": 7.758724966901487e-06,
+      "loss": 0.4918,
+      "step": 2543
+    },
+    {
+      "epoch": 1.2621323073104134,
+      "grad_norm": 0.07125484939936946,
+      "learning_rate": 7.757094511794155e-06,
+      "loss": 0.472,
+      "step": 2544
+    },
+    {
+      "epoch": 1.2626287700136527,
+      "grad_norm": 0.0718501582192958,
+      "learning_rate": 7.755463635309131e-06,
+      "loss": 0.5081,
+      "step": 2545
+    },
+    {
+      "epoch": 1.263125232716892,
+      "grad_norm": 0.07819740213883132,
+      "learning_rate": 7.753832337695672e-06,
+      "loss": 0.524,
+      "step": 2546
+    },
+    {
+      "epoch": 1.2636216954201316,
+      "grad_norm": 0.07653866302850325,
+      "learning_rate": 7.752200619203094e-06,
+      "loss": 0.5639,
+      "step": 2547
+    },
+    {
+      "epoch": 1.264118158123371,
+      "grad_norm": 0.07348045181586392,
+      "learning_rate": 7.75056848008078e-06,
+      "loss": 0.4803,
+      "step": 2548
+    },
+    {
+      "epoch": 1.2646146208266105,
+      "grad_norm": 0.07076246994678682,
+      "learning_rate": 7.748935920578176e-06,
+      "loss": 0.4713,
+      "step": 2549
+    },
+    {
+      "epoch": 1.2651110835298498,
+      "grad_norm": 0.07313613013227323,
+      "learning_rate": 7.747302940944791e-06,
+      "loss": 0.5083,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2656075462330891,
+      "grad_norm": 0.07349928306237019,
+      "learning_rate": 7.745669541430204e-06,
+      "loss": 0.5042,
+      "step": 2551
+    },
+    {
+      "epoch": 1.2661040089363287,
+      "grad_norm": 0.07157914735485121,
+      "learning_rate": 7.744035722284049e-06,
+      "loss": 0.5005,
+      "step": 2552
+    },
+    {
+      "epoch": 1.266600471639568,
+      "grad_norm": 0.07520177710019567,
+      "learning_rate": 7.74240148375603e-06,
+      "loss": 0.5022,
+      "step": 2553
+    },
+    {
+      "epoch": 1.2670969343428076,
+      "grad_norm": 0.07163077315954698,
+      "learning_rate": 7.740766826095918e-06,
+      "loss": 0.4607,
+      "step": 2554
+    },
+    {
+      "epoch": 1.267593397046047,
+      "grad_norm": 0.0743815357073748,
+      "learning_rate": 7.73913174955354e-06,
+      "loss": 0.4875,
+      "step": 2555
+    },
+    {
+      "epoch": 1.2680898597492862,
+      "grad_norm": 0.06912879169110557,
+      "learning_rate": 7.737496254378794e-06,
+      "loss": 0.4677,
+      "step": 2556
+    },
+    {
+      "epoch": 1.2685863224525258,
+      "grad_norm": 0.07262641887138097,
+      "learning_rate": 7.735860340821635e-06,
+      "loss": 0.496,
+      "step": 2557
+    },
+    {
+      "epoch": 1.2690827851557651,
+      "grad_norm": 0.07555540297786063,
+      "learning_rate": 7.734224009132091e-06,
+      "loss": 0.4857,
+      "step": 2558
+    },
+    {
+      "epoch": 1.2695792478590047,
+      "grad_norm": 0.07481020040902342,
+      "learning_rate": 7.732587259560247e-06,
+      "loss": 0.4845,
+      "step": 2559
+    },
+    {
+      "epoch": 1.270075710562244,
+      "grad_norm": 0.07347608023900698,
+      "learning_rate": 7.730950092356254e-06,
+      "loss": 0.4766,
+      "step": 2560
+    },
+    {
+      "epoch": 1.2705721732654833,
+      "grad_norm": 0.07310849706261968,
+      "learning_rate": 7.729312507770326e-06,
+      "loss": 0.4632,
+      "step": 2561
+    },
+    {
+      "epoch": 1.2710686359687229,
+      "grad_norm": 0.07375541777210778,
+      "learning_rate": 7.727674506052744e-06,
+      "loss": 0.5228,
+      "step": 2562
+    },
+    {
+      "epoch": 1.2715650986719622,
+      "grad_norm": 0.07333928626803109,
+      "learning_rate": 7.726036087453848e-06,
+      "loss": 0.4628,
+      "step": 2563
+    },
+    {
+      "epoch": 1.2720615613752018,
+      "grad_norm": 0.07561533426069027,
+      "learning_rate": 7.724397252224045e-06,
+      "loss": 0.5148,
+      "step": 2564
+    },
+    {
+      "epoch": 1.272558024078441,
+      "grad_norm": 0.07474417547503953,
+      "learning_rate": 7.722758000613804e-06,
+      "loss": 0.5097,
+      "step": 2565
+    },
+    {
+      "epoch": 1.2730544867816804,
+      "grad_norm": 0.07102407219450897,
+      "learning_rate": 7.721118332873659e-06,
+      "loss": 0.476,
+      "step": 2566
+    },
+    {
+      "epoch": 1.27355094948492,
+      "grad_norm": 0.07287426452910022,
+      "learning_rate": 7.719478249254206e-06,
+      "loss": 0.47,
+      "step": 2567
+    },
+    {
+      "epoch": 1.2740474121881593,
+      "grad_norm": 0.07176497136363207,
+      "learning_rate": 7.717837750006106e-06,
+      "loss": 0.4692,
+      "step": 2568
+    },
+    {
+      "epoch": 1.2745438748913989,
+      "grad_norm": 0.07219630348957971,
+      "learning_rate": 7.716196835380084e-06,
+      "loss": 0.4837,
+      "step": 2569
+    },
+    {
+      "epoch": 1.2750403375946382,
+      "grad_norm": 0.07536620917998516,
+      "learning_rate": 7.714555505626927e-06,
+      "loss": 0.5168,
+      "step": 2570
+    },
+    {
+      "epoch": 1.2755368002978775,
+      "grad_norm": 0.0714646589840732,
+      "learning_rate": 7.712913760997484e-06,
+      "loss": 0.4711,
+      "step": 2571
+    },
+    {
+      "epoch": 1.276033263001117,
+      "grad_norm": 0.07502031913388156,
+      "learning_rate": 7.71127160174267e-06,
+      "loss": 0.5066,
+      "step": 2572
+    },
+    {
+      "epoch": 1.2765297257043564,
+      "grad_norm": 0.07134378711895495,
+      "learning_rate": 7.709629028113468e-06,
+      "loss": 0.5159,
+      "step": 2573
+    },
+    {
+      "epoch": 1.277026188407596,
+      "grad_norm": 0.07332695455794223,
+      "learning_rate": 7.707986040360911e-06,
+      "loss": 0.5021,
+      "step": 2574
+    },
+    {
+      "epoch": 1.2775226511108353,
+      "grad_norm": 0.07145066063460748,
+      "learning_rate": 7.706342638736108e-06,
+      "loss": 0.512,
+      "step": 2575
+    },
+    {
+      "epoch": 1.2780191138140746,
+      "grad_norm": 0.07354359727508271,
+      "learning_rate": 7.704698823490226e-06,
+      "loss": 0.4953,
+      "step": 2576
+    },
+    {
+      "epoch": 1.2785155765173142,
+      "grad_norm": 0.07270371860886368,
+      "learning_rate": 7.703054594874495e-06,
+      "loss": 0.5099,
+      "step": 2577
+    },
+    {
+      "epoch": 1.2790120392205535,
+      "grad_norm": 0.06624808865885844,
+      "learning_rate": 7.701409953140209e-06,
+      "loss": 0.4564,
+      "step": 2578
+    },
+    {
+      "epoch": 1.279508501923793,
+      "grad_norm": 0.07412750614095082,
+      "learning_rate": 7.699764898538726e-06,
+      "loss": 0.5055,
+      "step": 2579
+    },
+    {
+      "epoch": 1.2800049646270324,
+      "grad_norm": 0.07275195995178917,
+      "learning_rate": 7.698119431321464e-06,
+      "loss": 0.4772,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2805014273302717,
+      "grad_norm": 0.07425362001957296,
+      "learning_rate": 7.69647355173991e-06,
+      "loss": 0.5233,
+      "step": 2581
+    },
+    {
+      "epoch": 1.2809978900335113,
+      "grad_norm": 0.07469451138696086,
+      "learning_rate": 7.694827260045608e-06,
+      "loss": 0.4987,
+      "step": 2582
+    },
+    {
+      "epoch": 1.2814943527367506,
+      "grad_norm": 0.0708213062123937,
+      "learning_rate": 7.693180556490167e-06,
+      "loss": 0.5357,
+      "step": 2583
+    },
+    {
+      "epoch": 1.2819908154399902,
+      "grad_norm": 0.07497939814476405,
+      "learning_rate": 7.691533441325261e-06,
+      "loss": 0.5359,
+      "step": 2584
+    },
+    {
+      "epoch": 1.2824872781432295,
+      "grad_norm": 0.07164073703137543,
+      "learning_rate": 7.689885914802622e-06,
+      "loss": 0.4742,
+      "step": 2585
+    },
+    {
+      "epoch": 1.2829837408464688,
+      "grad_norm": 0.07677183840172884,
+      "learning_rate": 7.68823797717405e-06,
+      "loss": 0.4812,
+      "step": 2586
+    },
+    {
+      "epoch": 1.2834802035497084,
+      "grad_norm": 0.07077909405950275,
+      "learning_rate": 7.68658962869141e-06,
+      "loss": 0.4696,
+      "step": 2587
+    },
+    {
+      "epoch": 1.2839766662529477,
+      "grad_norm": 0.07354323740522496,
+      "learning_rate": 7.684940869606617e-06,
+      "loss": 0.4724,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2844731289561873,
+      "grad_norm": 0.06993500834216944,
+      "learning_rate": 7.683291700171663e-06,
+      "loss": 0.4929,
+      "step": 2589
+    },
+    {
+      "epoch": 1.2849695916594266,
+      "grad_norm": 0.07345783185058559,
+      "learning_rate": 7.681642120638596e-06,
+      "loss": 0.507,
+      "step": 2590
+    },
+    {
+      "epoch": 1.285466054362666,
+      "grad_norm": 0.07367018462400143,
+      "learning_rate": 7.679992131259528e-06,
+      "loss": 0.497,
+      "step": 2591
+    },
+    {
+      "epoch": 1.2859625170659055,
+      "grad_norm": 0.07077074820163252,
+      "learning_rate": 7.678341732286633e-06,
+      "loss": 0.4722,
+      "step": 2592
+    },
+    {
+      "epoch": 1.2864589797691448,
+      "grad_norm": 0.07093567714122909,
+      "learning_rate": 7.676690923972148e-06,
+      "loss": 0.4881,
+      "step": 2593
+    },
+    {
+      "epoch": 1.2869554424723844,
+      "grad_norm": 0.0692286985895299,
+      "learning_rate": 7.675039706568373e-06,
+      "loss": 0.4654,
+      "step": 2594
+    },
+    {
+      "epoch": 1.2874519051756237,
+      "grad_norm": 0.07470922721879901,
+      "learning_rate": 7.673388080327669e-06,
+      "loss": 0.464,
+      "step": 2595
+    },
+    {
+      "epoch": 1.287948367878863,
+      "grad_norm": 0.07267808804570815,
+      "learning_rate": 7.671736045502462e-06,
+      "loss": 0.525,
+      "step": 2596
+    },
+    {
+      "epoch": 1.2884448305821026,
+      "grad_norm": 0.07322570248638664,
+      "learning_rate": 7.670083602345239e-06,
+      "loss": 0.4661,
+      "step": 2597
+    },
+    {
+      "epoch": 1.288941293285342,
+      "grad_norm": 0.07374076711966049,
+      "learning_rate": 7.66843075110855e-06,
+      "loss": 0.4996,
+      "step": 2598
+    },
+    {
+      "epoch": 1.2894377559885815,
+      "grad_norm": 0.07305715181136896,
+      "learning_rate": 7.666777492045003e-06,
+      "loss": 0.4805,
+      "step": 2599
+    },
+    {
+      "epoch": 1.2899342186918208,
+      "grad_norm": 0.07041169265149858,
+      "learning_rate": 7.665123825407276e-06,
+      "loss": 0.465,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2904306813950601,
+      "grad_norm": 0.07711313194745137,
+      "learning_rate": 7.663469751448104e-06,
+      "loss": 0.4995,
+      "step": 2601
+    },
+    {
+      "epoch": 1.2909271440982997,
+      "grad_norm": 0.07276220005027466,
+      "learning_rate": 7.661815270420286e-06,
+      "loss": 0.5004,
+      "step": 2602
+    },
+    {
+      "epoch": 1.291423606801539,
+      "grad_norm": 0.06923528529241481,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 0.4784,
+      "step": 2603
+    },
+    {
+      "epoch": 1.2919200695047786,
+      "grad_norm": 0.07204538421004412,
+      "learning_rate": 7.65850508817022e-06,
+      "loss": 0.504,
+      "step": 2604
+    },
+    {
+      "epoch": 1.2924165322080179,
+      "grad_norm": 0.0729160710003885,
+      "learning_rate": 7.656849387453878e-06,
+      "loss": 0.4906,
+      "step": 2605
+    },
+    {
+      "epoch": 1.2929129949112572,
+      "grad_norm": 0.07535310209672472,
+      "learning_rate": 7.655193280680706e-06,
+      "loss": 0.5436,
+      "step": 2606
+    },
+    {
+      "epoch": 1.2934094576144968,
+      "grad_norm": 0.0726456432774452,
+      "learning_rate": 7.653536768103814e-06,
+      "loss": 0.4971,
+      "step": 2607
+    },
+    {
+      "epoch": 1.293905920317736,
+      "grad_norm": 0.07302223221909274,
+      "learning_rate": 7.651879849976374e-06,
+      "loss": 0.4966,
+      "step": 2608
+    },
+    {
+      "epoch": 1.2944023830209757,
+      "grad_norm": 0.07004922600667704,
+      "learning_rate": 7.650222526551618e-06,
+      "loss": 0.4657,
+      "step": 2609
+    },
+    {
+      "epoch": 1.294898845724215,
+      "grad_norm": 0.07015347440478134,
+      "learning_rate": 7.648564798082842e-06,
+      "loss": 0.459,
+      "step": 2610
+    },
+    {
+      "epoch": 1.2953953084274543,
+      "grad_norm": 0.0745936418355579,
+      "learning_rate": 7.646906664823403e-06,
+      "loss": 0.4771,
+      "step": 2611
+    },
+    {
+      "epoch": 1.2958917711306939,
+      "grad_norm": 0.07667816509699976,
+      "learning_rate": 7.645248127026723e-06,
+      "loss": 0.5219,
+      "step": 2612
+    },
+    {
+      "epoch": 1.2963882338339332,
+      "grad_norm": 0.07029635717147024,
+      "learning_rate": 7.643589184946277e-06,
+      "loss": 0.4832,
+      "step": 2613
+    },
+    {
+      "epoch": 1.2968846965371728,
+      "grad_norm": 0.07366949480303261,
+      "learning_rate": 7.641929838835613e-06,
+      "loss": 0.5084,
+      "step": 2614
+    },
+    {
+      "epoch": 1.297381159240412,
+      "grad_norm": 0.07378733016384503,
+      "learning_rate": 7.640270088948332e-06,
+      "loss": 0.502,
+      "step": 2615
+    },
+    {
+      "epoch": 1.2978776219436514,
+      "grad_norm": 0.07359555858747464,
+      "learning_rate": 7.6386099355381e-06,
+      "loss": 0.4991,
+      "step": 2616
+    },
+    {
+      "epoch": 1.298374084646891,
+      "grad_norm": 0.07487529791386079,
+      "learning_rate": 7.636949378858647e-06,
+      "loss": 0.5023,
+      "step": 2617
+    },
+    {
+      "epoch": 1.2988705473501303,
+      "grad_norm": 0.07261861801231229,
+      "learning_rate": 7.635288419163763e-06,
+      "loss": 0.4968,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2993670100533699,
+      "grad_norm": 0.07170644816281213,
+      "learning_rate": 7.633627056707297e-06,
+      "loss": 0.495,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2998634727566092,
+      "grad_norm": 0.07326782739072436,
+      "learning_rate": 7.631965291743163e-06,
+      "loss": 0.4749,
+      "step": 2620
+    },
+    {
+      "epoch": 1.3003599354598485,
+      "grad_norm": 0.07354949520465813,
+      "learning_rate": 7.630303124525333e-06,
+      "loss": 0.4887,
+      "step": 2621
+    },
+    {
+      "epoch": 1.300856398163088,
+      "grad_norm": 0.07677994339072212,
+      "learning_rate": 7.628640555307845e-06,
+      "loss": 0.4941,
+      "step": 2622
+    },
+    {
+      "epoch": 1.3013528608663274,
+      "grad_norm": 0.07331225391921062,
+      "learning_rate": 7.626977584344795e-06,
+      "loss": 0.4858,
+      "step": 2623
+    },
+    {
+      "epoch": 1.301849323569567,
+      "grad_norm": 0.07170856926012528,
+      "learning_rate": 7.625314211890342e-06,
+      "loss": 0.4455,
+      "step": 2624
+    },
+    {
+      "epoch": 1.3023457862728063,
+      "grad_norm": 0.07475766562711376,
+      "learning_rate": 7.623650438198707e-06,
+      "loss": 0.4724,
+      "step": 2625
+    },
+    {
+      "epoch": 1.3028422489760456,
+      "grad_norm": 0.07303322858619267,
+      "learning_rate": 7.621986263524166e-06,
+      "loss": 0.5025,
+      "step": 2626
+    },
+    {
+      "epoch": 1.3033387116792852,
+      "grad_norm": 0.07209015796425527,
+      "learning_rate": 7.620321688121066e-06,
+      "loss": 0.533,
+      "step": 2627
+    },
+    {
+      "epoch": 1.3038351743825245,
+      "grad_norm": 0.07418574085022885,
+      "learning_rate": 7.618656712243813e-06,
+      "loss": 0.479,
+      "step": 2628
+    },
+    {
+      "epoch": 1.304331637085764,
+      "grad_norm": 0.07566746702664336,
+      "learning_rate": 7.616991336146864e-06,
+      "loss": 0.4857,
+      "step": 2629
+    },
+    {
+      "epoch": 1.3048280997890034,
+      "grad_norm": 0.06931794377147632,
+      "learning_rate": 7.615325560084752e-06,
+      "loss": 0.4614,
+      "step": 2630
+    },
+    {
+      "epoch": 1.3053245624922427,
+      "grad_norm": 0.07616031124079053,
+      "learning_rate": 7.613659384312062e-06,
+      "loss": 0.5274,
+      "step": 2631
+    },
+    {
+      "epoch": 1.3058210251954823,
+      "grad_norm": 0.07085893200757314,
+      "learning_rate": 7.611992809083439e-06,
+      "loss": 0.4691,
+      "step": 2632
+    },
+    {
+      "epoch": 1.3063174878987216,
+      "grad_norm": 0.07319646074018087,
+      "learning_rate": 7.610325834653598e-06,
+      "loss": 0.517,
+      "step": 2633
+    },
+    {
+      "epoch": 1.3068139506019611,
+      "grad_norm": 0.07459044022717175,
+      "learning_rate": 7.6086584612773055e-06,
+      "loss": 0.4582,
+      "step": 2634
+    },
+    {
+      "epoch": 1.3073104133052005,
+      "grad_norm": 0.07570613366111596,
+      "learning_rate": 7.606990689209395e-06,
+      "loss": 0.5111,
+      "step": 2635
+    },
+    {
+      "epoch": 1.3078068760084398,
+      "grad_norm": 0.07428287381621476,
+      "learning_rate": 7.605322518704759e-06,
+      "loss": 0.4913,
+      "step": 2636
+    },
+    {
+      "epoch": 1.3083033387116794,
+      "grad_norm": 0.0734226364221571,
+      "learning_rate": 7.603653950018346e-06,
+      "loss": 0.4762,
+      "step": 2637
+    },
+    {
+      "epoch": 1.3087998014149187,
+      "grad_norm": 0.07332280586009639,
+      "learning_rate": 7.601984983405173e-06,
+      "loss": 0.4946,
+      "step": 2638
+    },
+    {
+      "epoch": 1.3092962641181582,
+      "grad_norm": 0.07120489183911462,
+      "learning_rate": 7.600315619120317e-06,
+      "loss": 0.4762,
+      "step": 2639
+    },
+    {
+      "epoch": 1.3097927268213976,
+      "grad_norm": 0.07303767095401219,
+      "learning_rate": 7.59864585741891e-06,
+      "loss": 0.475,
+      "step": 2640
+    },
+    {
+      "epoch": 1.310289189524637,
+      "grad_norm": 0.07187071930757646,
+      "learning_rate": 7.596975698556151e-06,
+      "loss": 0.474,
+      "step": 2641
+    },
+    {
+      "epoch": 1.3107856522278765,
+      "grad_norm": 0.07415230799455923,
+      "learning_rate": 7.595305142787294e-06,
+      "loss": 0.5229,
+      "step": 2642
+    },
+    {
+      "epoch": 1.3112821149311158,
+      "grad_norm": 0.07447483592407889,
+      "learning_rate": 7.59363419036766e-06,
+      "loss": 0.5035,
+      "step": 2643
+    },
+    {
+      "epoch": 1.3117785776343553,
+      "grad_norm": 0.07040880107400416,
+      "learning_rate": 7.591962841552627e-06,
+      "loss": 0.4725,
+      "step": 2644
+    },
+    {
+      "epoch": 1.3122750403375947,
+      "grad_norm": 0.06793138068471151,
+      "learning_rate": 7.590291096597631e-06,
+      "loss": 0.4412,
+      "step": 2645
+    },
+    {
+      "epoch": 1.312771503040834,
+      "grad_norm": 0.07495097651516075,
+      "learning_rate": 7.588618955758173e-06,
+      "loss": 0.5002,
+      "step": 2646
+    },
+    {
+      "epoch": 1.3132679657440733,
+      "grad_norm": 0.07565513056199885,
+      "learning_rate": 7.586946419289813e-06,
+      "loss": 0.495,
+      "step": 2647
+    },
+    {
+      "epoch": 1.313764428447313,
+      "grad_norm": 0.07846426523907774,
+      "learning_rate": 7.58527348744817e-06,
+      "loss": 0.4963,
+      "step": 2648
+    },
+    {
+      "epoch": 1.3142608911505524,
+      "grad_norm": 0.07726618154433754,
+      "learning_rate": 7.583600160488929e-06,
+      "loss": 0.5038,
+      "step": 2649
+    },
+    {
+      "epoch": 1.3147573538537918,
+      "grad_norm": 0.07177971035598558,
+      "learning_rate": 7.581926438667826e-06,
+      "loss": 0.4977,
+      "step": 2650
+    },
+    {
+      "epoch": 1.315253816557031,
+      "grad_norm": 0.07234611189725769,
+      "learning_rate": 7.580252322240666e-06,
+      "loss": 0.4719,
+      "step": 2651
+    },
+    {
+      "epoch": 1.3157502792602704,
+      "grad_norm": 0.07201594842948567,
+      "learning_rate": 7.57857781146331e-06,
+      "loss": 0.5089,
+      "step": 2652
+    },
+    {
+      "epoch": 1.31624674196351,
+      "grad_norm": 0.07498288918290671,
+      "learning_rate": 7.57690290659168e-06,
+      "loss": 0.4814,
+      "step": 2653
+    },
+    {
+      "epoch": 1.3167432046667495,
+      "grad_norm": 0.07618518497314423,
+      "learning_rate": 7.575227607881757e-06,
+      "loss": 0.494,
+      "step": 2654
+    },
+    {
+      "epoch": 1.3172396673699889,
+      "grad_norm": 0.07245035215910761,
+      "learning_rate": 7.573551915589586e-06,
+      "loss": 0.5111,
+      "step": 2655
+    },
+    {
+      "epoch": 1.3177361300732282,
+      "grad_norm": 0.07063014586866156,
+      "learning_rate": 7.571875829971267e-06,
+      "loss": 0.5076,
+      "step": 2656
+    },
+    {
+      "epoch": 1.3182325927764675,
+      "grad_norm": 0.07384313400197565,
+      "learning_rate": 7.5701993512829664e-06,
+      "loss": 0.4998,
+      "step": 2657
+    },
+    {
+      "epoch": 1.318729055479707,
+      "grad_norm": 0.0752424884466584,
+      "learning_rate": 7.568522479780903e-06,
+      "loss": 0.4881,
+      "step": 2658
+    },
+    {
+      "epoch": 1.3192255181829466,
+      "grad_norm": 0.08052341158611934,
+      "learning_rate": 7.566845215721362e-06,
+      "loss": 0.495,
+      "step": 2659
+    },
+    {
+      "epoch": 1.319721980886186,
+      "grad_norm": 0.07208145165414205,
+      "learning_rate": 7.5651675593606876e-06,
+      "loss": 0.4981,
+      "step": 2660
+    },
+    {
+      "epoch": 1.3202184435894253,
+      "grad_norm": 0.07313301936435801,
+      "learning_rate": 7.5634895109552795e-06,
+      "loss": 0.4965,
+      "step": 2661
+    },
+    {
+      "epoch": 1.3207149062926646,
+      "grad_norm": 0.07556523446833772,
+      "learning_rate": 7.561811070761602e-06,
+      "loss": 0.5176,
+      "step": 2662
+    },
+    {
+      "epoch": 1.3212113689959042,
+      "grad_norm": 0.07319725779490793,
+      "learning_rate": 7.56013223903618e-06,
+      "loss": 0.4585,
+      "step": 2663
+    },
+    {
+      "epoch": 1.3217078316991435,
+      "grad_norm": 0.07083563037225474,
+      "learning_rate": 7.558453016035592e-06,
+      "loss": 0.482,
+      "step": 2664
+    },
+    {
+      "epoch": 1.322204294402383,
+      "grad_norm": 0.07220725027684433,
+      "learning_rate": 7.556773402016482e-06,
+      "loss": 0.5062,
+      "step": 2665
+    },
+    {
+      "epoch": 1.3227007571056224,
+      "grad_norm": 0.07152061737141101,
+      "learning_rate": 7.555093397235553e-06,
+      "loss": 0.4993,
+      "step": 2666
+    },
+    {
+      "epoch": 1.3231972198088617,
+      "grad_norm": 0.0731269231629725,
+      "learning_rate": 7.553413001949566e-06,
+      "loss": 0.5071,
+      "step": 2667
+    },
+    {
+      "epoch": 1.3236936825121013,
+      "grad_norm": 0.07211546920533103,
+      "learning_rate": 7.551732216415342e-06,
+      "loss": 0.4899,
+      "step": 2668
+    },
+    {
+      "epoch": 1.3241901452153406,
+      "grad_norm": 0.07427869438620326,
+      "learning_rate": 7.5500510408897634e-06,
+      "loss": 0.518,
+      "step": 2669
+    },
+    {
+      "epoch": 1.3246866079185802,
+      "grad_norm": 0.07411425941651224,
+      "learning_rate": 7.548369475629769e-06,
+      "loss": 0.5099,
+      "step": 2670
+    },
+    {
+      "epoch": 1.3251830706218195,
+      "grad_norm": 0.07102919893674516,
+      "learning_rate": 7.546687520892361e-06,
+      "loss": 0.4577,
+      "step": 2671
+    },
+    {
+      "epoch": 1.3256795333250588,
+      "grad_norm": 0.07124317917931709,
+      "learning_rate": 7.545005176934597e-06,
+      "loss": 0.4857,
+      "step": 2672
+    },
+    {
+      "epoch": 1.3261759960282984,
+      "grad_norm": 0.06769496175031949,
+      "learning_rate": 7.543322444013601e-06,
+      "loss": 0.4641,
+      "step": 2673
+    },
+    {
+      "epoch": 1.3266724587315377,
+      "grad_norm": 0.06899557389748309,
+      "learning_rate": 7.541639322386546e-06,
+      "loss": 0.4664,
+      "step": 2674
+    },
+    {
+      "epoch": 1.3271689214347773,
+      "grad_norm": 0.07481434699614627,
+      "learning_rate": 7.539955812310673e-06,
+      "loss": 0.4744,
+      "step": 2675
+    },
+    {
+      "epoch": 1.3276653841380166,
+      "grad_norm": 0.07304587484098414,
+      "learning_rate": 7.538271914043281e-06,
+      "loss": 0.4559,
+      "step": 2676
+    },
+    {
+      "epoch": 1.328161846841256,
+      "grad_norm": 0.07415770338497507,
+      "learning_rate": 7.536587627841723e-06,
+      "loss": 0.5121,
+      "step": 2677
+    },
+    {
+      "epoch": 1.3286583095444955,
+      "grad_norm": 0.07228770607947253,
+      "learning_rate": 7.534902953963417e-06,
+      "loss": 0.5157,
+      "step": 2678
+    },
+    {
+      "epoch": 1.3291547722477348,
+      "grad_norm": 0.07319401385123882,
+      "learning_rate": 7.533217892665839e-06,
+      "loss": 0.504,
+      "step": 2679
+    },
+    {
+      "epoch": 1.3296512349509744,
+      "grad_norm": 0.07045236666155964,
+      "learning_rate": 7.531532444206524e-06,
+      "loss": 0.5008,
+      "step": 2680
+    },
+    {
+      "epoch": 1.3301476976542137,
+      "grad_norm": 0.07117873637101506,
+      "learning_rate": 7.529846608843063e-06,
+      "loss": 0.4847,
+      "step": 2681
+    },
+    {
+      "epoch": 1.330644160357453,
+      "grad_norm": 0.07067616323712225,
+      "learning_rate": 7.528160386833112e-06,
+      "loss": 0.4804,
+      "step": 2682
+    },
+    {
+      "epoch": 1.3311406230606926,
+      "grad_norm": 0.0759415536662072,
+      "learning_rate": 7.526473778434383e-06,
+      "loss": 0.4969,
+      "step": 2683
+    },
+    {
+      "epoch": 1.331637085763932,
+      "grad_norm": 0.07086435551103258,
+      "learning_rate": 7.524786783904645e-06,
+      "loss": 0.4638,
+      "step": 2684
+    },
+    {
+      "epoch": 1.3321335484671715,
+      "grad_norm": 0.07035449113846659,
+      "learning_rate": 7.52309940350173e-06,
+      "loss": 0.4932,
+      "step": 2685
+    },
+    {
+      "epoch": 1.3326300111704108,
+      "grad_norm": 0.07525591700961363,
+      "learning_rate": 7.521411637483525e-06,
+      "loss": 0.4691,
+      "step": 2686
+    },
+    {
+      "epoch": 1.3331264738736501,
+      "grad_norm": 0.07621426823298727,
+      "learning_rate": 7.519723486107977e-06,
+      "loss": 0.5114,
+      "step": 2687
+    },
+    {
+      "epoch": 1.3336229365768897,
+      "grad_norm": 0.07019664010034184,
+      "learning_rate": 7.518034949633097e-06,
+      "loss": 0.4966,
+      "step": 2688
+    },
+    {
+      "epoch": 1.334119399280129,
+      "grad_norm": 0.07454034140674029,
+      "learning_rate": 7.51634602831695e-06,
+      "loss": 0.4781,
+      "step": 2689
+    },
+    {
+      "epoch": 1.3346158619833686,
+      "grad_norm": 0.07418990783466832,
+      "learning_rate": 7.514656722417656e-06,
+      "loss": 0.5156,
+      "step": 2690
+    },
+    {
+      "epoch": 1.335112324686608,
+      "grad_norm": 0.07262590410613991,
+      "learning_rate": 7.512967032193404e-06,
+      "loss": 0.4759,
+      "step": 2691
+    },
+    {
+      "epoch": 1.3356087873898472,
+      "grad_norm": 0.07326886517366066,
+      "learning_rate": 7.511276957902431e-06,
+      "loss": 0.5035,
+      "step": 2692
+    },
+    {
+      "epoch": 1.3361052500930868,
+      "grad_norm": 0.07557567814539083,
+      "learning_rate": 7.509586499803042e-06,
+      "loss": 0.5098,
+      "step": 2693
+    },
+    {
+      "epoch": 1.3366017127963261,
+      "grad_norm": 0.07269601973403042,
+      "learning_rate": 7.507895658153594e-06,
+      "loss": 0.4628,
+      "step": 2694
+    },
+    {
+      "epoch": 1.3370981754995657,
+      "grad_norm": 0.0717655547977608,
+      "learning_rate": 7.5062044332125076e-06,
+      "loss": 0.5035,
+      "step": 2695
+    },
+    {
+      "epoch": 1.337594638202805,
+      "grad_norm": 0.06961351252939559,
+      "learning_rate": 7.504512825238255e-06,
+      "loss": 0.47,
+      "step": 2696
+    },
+    {
+      "epoch": 1.3380911009060443,
+      "grad_norm": 0.07492553610892763,
+      "learning_rate": 7.502820834489374e-06,
+      "loss": 0.5463,
+      "step": 2697
+    },
+    {
+      "epoch": 1.3385875636092839,
+      "grad_norm": 0.07210512256673479,
+      "learning_rate": 7.5011284612244585e-06,
+      "loss": 0.4816,
+      "step": 2698
+    },
+    {
+      "epoch": 1.3390840263125232,
+      "grad_norm": 0.0779198103694556,
+      "learning_rate": 7.499435705702161e-06,
+      "loss": 0.462,
+      "step": 2699
+    },
+    {
+      "epoch": 1.3395804890157628,
+      "grad_norm": 0.07158341114436462,
+      "learning_rate": 7.497742568181191e-06,
+      "loss": 0.4965,
+      "step": 2700
+    },
+    {
+      "epoch": 1.340076951719002,
+      "grad_norm": 0.07158678321680322,
+      "learning_rate": 7.496049048920317e-06,
+      "loss": 0.4917,
+      "step": 2701
+    },
+    {
+      "epoch": 1.3405734144222414,
+      "grad_norm": 0.07424116914927884,
+      "learning_rate": 7.494355148178368e-06,
+      "loss": 0.5064,
+      "step": 2702
+    },
+    {
+      "epoch": 1.341069877125481,
+      "grad_norm": 0.08002397795499651,
+      "learning_rate": 7.492660866214228e-06,
+      "loss": 0.5164,
+      "step": 2703
+    },
+    {
+      "epoch": 1.3415663398287203,
+      "grad_norm": 0.07652500765458362,
+      "learning_rate": 7.490966203286841e-06,
+      "loss": 0.4886,
+      "step": 2704
+    },
+    {
+      "epoch": 1.3420628025319599,
+      "grad_norm": 0.075845565077813,
+      "learning_rate": 7.489271159655212e-06,
+      "loss": 0.5047,
+      "step": 2705
+    },
+    {
+      "epoch": 1.3425592652351992,
+      "grad_norm": 0.07487393856531731,
+      "learning_rate": 7.4875757355783955e-06,
+      "loss": 0.4791,
+      "step": 2706
+    },
+    {
+      "epoch": 1.3430557279384385,
+      "grad_norm": 0.07701656968563915,
+      "learning_rate": 7.485879931315514e-06,
+      "loss": 0.5021,
+      "step": 2707
+    },
+    {
+      "epoch": 1.343552190641678,
+      "grad_norm": 0.07312866184878818,
+      "learning_rate": 7.484183747125743e-06,
+      "loss": 0.5169,
+      "step": 2708
+    },
+    {
+      "epoch": 1.3440486533449174,
+      "grad_norm": 0.07242385146219911,
+      "learning_rate": 7.482487183268318e-06,
+      "loss": 0.5085,
+      "step": 2709
+    },
+    {
+      "epoch": 1.344545116048157,
+      "grad_norm": 0.07506511763062933,
+      "learning_rate": 7.480790240002529e-06,
+      "loss": 0.4956,
+      "step": 2710
+    },
+    {
+      "epoch": 1.3450415787513963,
+      "grad_norm": 0.07677366310399905,
+      "learning_rate": 7.4790929175877305e-06,
+      "loss": 0.5031,
+      "step": 2711
+    },
+    {
+      "epoch": 1.3455380414546356,
+      "grad_norm": 0.07216912197089925,
+      "learning_rate": 7.477395216283328e-06,
+      "loss": 0.4683,
+      "step": 2712
+    },
+    {
+      "epoch": 1.3460345041578752,
+      "grad_norm": 0.07071635918706311,
+      "learning_rate": 7.475697136348787e-06,
+      "loss": 0.4681,
+      "step": 2713
+    },
+    {
+      "epoch": 1.3465309668611145,
+      "grad_norm": 0.07775974984507308,
+      "learning_rate": 7.4739986780436345e-06,
+      "loss": 0.5062,
+      "step": 2714
+    },
+    {
+      "epoch": 1.347027429564354,
+      "grad_norm": 0.07973251419899774,
+      "learning_rate": 7.472299841627452e-06,
+      "loss": 0.4732,
+      "step": 2715
+    },
+    {
+      "epoch": 1.3475238922675934,
+      "grad_norm": 0.0776256133543414,
+      "learning_rate": 7.470600627359879e-06,
+      "loss": 0.4989,
+      "step": 2716
+    },
+    {
+      "epoch": 1.3480203549708327,
+      "grad_norm": 0.07218639025971221,
+      "learning_rate": 7.468901035500613e-06,
+      "loss": 0.4695,
+      "step": 2717
+    },
+    {
+      "epoch": 1.3485168176740723,
+      "grad_norm": 0.07177156812514981,
+      "learning_rate": 7.46720106630941e-06,
+      "loss": 0.4525,
+      "step": 2718
+    },
+    {
+      "epoch": 1.3490132803773116,
+      "grad_norm": 0.0743645465552789,
+      "learning_rate": 7.465500720046082e-06,
+      "loss": 0.4782,
+      "step": 2719
+    },
+    {
+      "epoch": 1.3495097430805512,
+      "grad_norm": 0.07670649036827379,
+      "learning_rate": 7.4637999969705e-06,
+      "loss": 0.4898,
+      "step": 2720
+    },
+    {
+      "epoch": 1.3500062057837905,
+      "grad_norm": 0.08775144330304403,
+      "learning_rate": 7.462098897342593e-06,
+      "loss": 0.4931,
+      "step": 2721
+    },
+    {
+      "epoch": 1.3505026684870298,
+      "grad_norm": 0.07840866502242834,
+      "learning_rate": 7.460397421422346e-06,
+      "loss": 0.5265,
+      "step": 2722
+    },
+    {
+      "epoch": 1.3509991311902694,
+      "grad_norm": 0.07196359299149803,
+      "learning_rate": 7.458695569469802e-06,
+      "loss": 0.4771,
+      "step": 2723
+    },
+    {
+      "epoch": 1.3514955938935087,
+      "grad_norm": 0.06953071754532851,
+      "learning_rate": 7.456993341745063e-06,
+      "loss": 0.4897,
+      "step": 2724
+    },
+    {
+      "epoch": 1.3519920565967483,
+      "grad_norm": 0.07063937728790774,
+      "learning_rate": 7.455290738508288e-06,
+      "loss": 0.4808,
+      "step": 2725
+    },
+    {
+      "epoch": 1.3524885192999876,
+      "grad_norm": 0.08088110766023791,
+      "learning_rate": 7.453587760019691e-06,
+      "loss": 0.4938,
+      "step": 2726
+    },
+    {
+      "epoch": 1.352984982003227,
+      "grad_norm": 0.0736673081073693,
+      "learning_rate": 7.451884406539545e-06,
+      "loss": 0.4748,
+      "step": 2727
+    },
+    {
+      "epoch": 1.3534814447064665,
+      "grad_norm": 0.06947625515604816,
+      "learning_rate": 7.4501806783281785e-06,
+      "loss": 0.4733,
+      "step": 2728
+    },
+    {
+      "epoch": 1.3539779074097058,
+      "grad_norm": 0.07218112018283263,
+      "learning_rate": 7.448476575645982e-06,
+      "loss": 0.5189,
+      "step": 2729
+    },
+    {
+      "epoch": 1.3544743701129454,
+      "grad_norm": 0.07421666853034066,
+      "learning_rate": 7.446772098753398e-06,
+      "loss": 0.4864,
+      "step": 2730
+    },
+    {
+      "epoch": 1.3549708328161847,
+      "grad_norm": 0.0689127465215305,
+      "learning_rate": 7.445067247910931e-06,
+      "loss": 0.4633,
+      "step": 2731
+    },
+    {
+      "epoch": 1.355467295519424,
+      "grad_norm": 0.07185351280383354,
+      "learning_rate": 7.44336202337914e-06,
+      "loss": 0.4856,
+      "step": 2732
+    },
+    {
+      "epoch": 1.3559637582226636,
+      "grad_norm": 0.07237908595669429,
+      "learning_rate": 7.441656425418639e-06,
+      "loss": 0.4672,
+      "step": 2733
+    },
+    {
+      "epoch": 1.356460220925903,
+      "grad_norm": 0.0717292512341406,
+      "learning_rate": 7.439950454290103e-06,
+      "loss": 0.4777,
+      "step": 2734
+    },
+    {
+      "epoch": 1.3569566836291425,
+      "grad_norm": 0.07388599914969228,
+      "learning_rate": 7.43824411025426e-06,
+      "loss": 0.4854,
+      "step": 2735
+    },
+    {
+      "epoch": 1.3574531463323818,
+      "grad_norm": 0.07748978640369678,
+      "learning_rate": 7.4365373935719e-06,
+      "loss": 0.4898,
+      "step": 2736
+    },
+    {
+      "epoch": 1.3579496090356211,
+      "grad_norm": 0.07054638978491355,
+      "learning_rate": 7.434830304503866e-06,
+      "loss": 0.4763,
+      "step": 2737
+    },
+    {
+      "epoch": 1.3584460717388607,
+      "grad_norm": 0.07292503890545296,
+      "learning_rate": 7.43312284331106e-06,
+      "loss": 0.5078,
+      "step": 2738
+    },
+    {
+      "epoch": 1.3589425344421,
+      "grad_norm": 0.07105075474486573,
+      "learning_rate": 7.431415010254439e-06,
+      "loss": 0.4839,
+      "step": 2739
+    },
+    {
+      "epoch": 1.3594389971453396,
+      "grad_norm": 0.07334885127452315,
+      "learning_rate": 7.42970680559502e-06,
+      "loss": 0.5086,
+      "step": 2740
+    },
+    {
+      "epoch": 1.3599354598485789,
+      "grad_norm": 0.07527855807145163,
+      "learning_rate": 7.427998229593873e-06,
+      "loss": 0.4987,
+      "step": 2741
+    },
+    {
+      "epoch": 1.3604319225518182,
+      "grad_norm": 0.07325157344838848,
+      "learning_rate": 7.426289282512125e-06,
+      "loss": 0.5008,
+      "step": 2742
+    },
+    {
+      "epoch": 1.3609283852550578,
+      "grad_norm": 0.07267084590873454,
+      "learning_rate": 7.424579964610963e-06,
+      "loss": 0.4874,
+      "step": 2743
+    },
+    {
+      "epoch": 1.361424847958297,
+      "grad_norm": 0.07193623146080504,
+      "learning_rate": 7.422870276151629e-06,
+      "loss": 0.471,
+      "step": 2744
+    },
+    {
+      "epoch": 1.3619213106615367,
+      "grad_norm": 0.07179868825468037,
+      "learning_rate": 7.42116021739542e-06,
+      "loss": 0.497,
+      "step": 2745
+    },
+    {
+      "epoch": 1.362417773364776,
+      "grad_norm": 0.0734773074878661,
+      "learning_rate": 7.419449788603693e-06,
+      "loss": 0.4774,
+      "step": 2746
+    },
+    {
+      "epoch": 1.3629142360680153,
+      "grad_norm": 0.07023057493932133,
+      "learning_rate": 7.417738990037859e-06,
+      "loss": 0.4829,
+      "step": 2747
+    },
+    {
+      "epoch": 1.3634106987712549,
+      "grad_norm": 0.07282086501091081,
+      "learning_rate": 7.416027821959387e-06,
+      "loss": 0.4958,
+      "step": 2748
+    },
+    {
+      "epoch": 1.3639071614744942,
+      "grad_norm": 0.07089773487569606,
+      "learning_rate": 7.414316284629799e-06,
+      "loss": 0.4971,
+      "step": 2749
+    },
+    {
+      "epoch": 1.3644036241777338,
+      "grad_norm": 0.0677423809061766,
+      "learning_rate": 7.412604378310677e-06,
+      "loss": 0.496,
+      "step": 2750
+    },
+    {
+      "epoch": 1.364900086880973,
+      "grad_norm": 0.06777700725873004,
+      "learning_rate": 7.4108921032636605e-06,
+      "loss": 0.4729,
+      "step": 2751
+    },
+    {
+      "epoch": 1.3653965495842124,
+      "grad_norm": 0.07065655577723433,
+      "learning_rate": 7.409179459750439e-06,
+      "loss": 0.521,
+      "step": 2752
+    },
+    {
+      "epoch": 1.365893012287452,
+      "grad_norm": 0.07392407970241713,
+      "learning_rate": 7.407466448032768e-06,
+      "loss": 0.4896,
+      "step": 2753
+    },
+    {
+      "epoch": 1.3663894749906913,
+      "grad_norm": 0.06934154352736671,
+      "learning_rate": 7.405753068372451e-06,
+      "loss": 0.4528,
+      "step": 2754
+    },
+    {
+      "epoch": 1.3668859376939309,
+      "grad_norm": 0.06887110619053861,
+      "learning_rate": 7.40403932103135e-06,
+      "loss": 0.4637,
+      "step": 2755
+    },
+    {
+      "epoch": 1.3673824003971702,
+      "grad_norm": 0.07123362248000452,
+      "learning_rate": 7.402325206271385e-06,
+      "loss": 0.5373,
+      "step": 2756
+    },
+    {
+      "epoch": 1.3678788631004095,
+      "grad_norm": 0.07431889053149492,
+      "learning_rate": 7.400610724354531e-06,
+      "loss": 0.4968,
+      "step": 2757
+    },
+    {
+      "epoch": 1.368375325803649,
+      "grad_norm": 0.07044786714598965,
+      "learning_rate": 7.398895875542818e-06,
+      "loss": 0.4992,
+      "step": 2758
+    },
+    {
+      "epoch": 1.3688717885068884,
+      "grad_norm": 0.08173014827815345,
+      "learning_rate": 7.397180660098334e-06,
+      "loss": 0.5075,
+      "step": 2759
+    },
+    {
+      "epoch": 1.369368251210128,
+      "grad_norm": 0.07510902820516983,
+      "learning_rate": 7.395465078283222e-06,
+      "loss": 0.4954,
+      "step": 2760
+    },
+    {
+      "epoch": 1.3698647139133673,
+      "grad_norm": 0.0717772886882652,
+      "learning_rate": 7.393749130359681e-06,
+      "loss": 0.5002,
+      "step": 2761
+    },
+    {
+      "epoch": 1.3703611766166066,
+      "grad_norm": 0.07860853949103024,
+      "learning_rate": 7.392032816589965e-06,
+      "loss": 0.5241,
+      "step": 2762
+    },
+    {
+      "epoch": 1.3708576393198462,
+      "grad_norm": 0.1974487115378719,
+      "learning_rate": 7.390316137236389e-06,
+      "loss": 0.5133,
+      "step": 2763
+    },
+    {
+      "epoch": 1.3713541020230855,
+      "grad_norm": 0.07342381822412332,
+      "learning_rate": 7.388599092561315e-06,
+      "loss": 0.5204,
+      "step": 2764
+    },
+    {
+      "epoch": 1.371850564726325,
+      "grad_norm": 0.07415428627463654,
+      "learning_rate": 7.38688168282717e-06,
+      "loss": 0.4918,
+      "step": 2765
+    },
+    {
+      "epoch": 1.3723470274295644,
+      "grad_norm": 0.07375186245968954,
+      "learning_rate": 7.3851639082964285e-06,
+      "loss": 0.4917,
+      "step": 2766
+    },
+    {
+      "epoch": 1.3728434901328037,
+      "grad_norm": 0.07311946684531592,
+      "learning_rate": 7.383445769231628e-06,
+      "loss": 0.54,
+      "step": 2767
+    },
+    {
+      "epoch": 1.3733399528360433,
+      "grad_norm": 0.07154186959129635,
+      "learning_rate": 7.381727265895356e-06,
+      "loss": 0.4945,
+      "step": 2768
+    },
+    {
+      "epoch": 1.3738364155392826,
+      "grad_norm": 0.07311048103614315,
+      "learning_rate": 7.38000839855026e-06,
+      "loss": 0.5018,
+      "step": 2769
+    },
+    {
+      "epoch": 1.3743328782425221,
+      "grad_norm": 0.07076489048383193,
+      "learning_rate": 7.378289167459043e-06,
+      "loss": 0.5027,
+      "step": 2770
+    },
+    {
+      "epoch": 1.3748293409457615,
+      "grad_norm": 0.07211214606420607,
+      "learning_rate": 7.376569572884457e-06,
+      "loss": 0.4903,
+      "step": 2771
+    },
+    {
+      "epoch": 1.3753258036490008,
+      "grad_norm": 0.07315568940001275,
+      "learning_rate": 7.374849615089318e-06,
+      "loss": 0.497,
+      "step": 2772
+    },
+    {
+      "epoch": 1.3758222663522404,
+      "grad_norm": 0.07074353994217372,
+      "learning_rate": 7.373129294336494e-06,
+      "loss": 0.4647,
+      "step": 2773
+    },
+    {
+      "epoch": 1.3763187290554797,
+      "grad_norm": 0.07573576492885328,
+      "learning_rate": 7.371408610888907e-06,
+      "loss": 0.4607,
+      "step": 2774
+    },
+    {
+      "epoch": 1.3768151917587192,
+      "grad_norm": 0.07374494787897591,
+      "learning_rate": 7.3696875650095355e-06,
+      "loss": 0.474,
+      "step": 2775
+    },
+    {
+      "epoch": 1.3773116544619586,
+      "grad_norm": 0.06844979894385408,
+      "learning_rate": 7.367966156961417e-06,
+      "loss": 0.463,
+      "step": 2776
+    },
+    {
+      "epoch": 1.377808117165198,
+      "grad_norm": 0.07494696046611961,
+      "learning_rate": 7.366244387007637e-06,
+      "loss": 0.4904,
+      "step": 2777
+    },
+    {
+      "epoch": 1.3783045798684375,
+      "grad_norm": 0.07245749426844175,
+      "learning_rate": 7.364522255411342e-06,
+      "loss": 0.4692,
+      "step": 2778
+    },
+    {
+      "epoch": 1.3788010425716768,
+      "grad_norm": 0.07166509196102683,
+      "learning_rate": 7.362799762435733e-06,
+      "loss": 0.4576,
+      "step": 2779
+    },
+    {
+      "epoch": 1.3792975052749163,
+      "grad_norm": 0.0692664814713353,
+      "learning_rate": 7.361076908344066e-06,
+      "loss": 0.481,
+      "step": 2780
+    },
+    {
+      "epoch": 1.3797939679781557,
+      "grad_norm": 0.07418340449606887,
+      "learning_rate": 7.359353693399651e-06,
+      "loss": 0.4919,
+      "step": 2781
+    },
+    {
+      "epoch": 1.380290430681395,
+      "grad_norm": 0.07669373600032768,
+      "learning_rate": 7.357630117865852e-06,
+      "loss": 0.5127,
+      "step": 2782
+    },
+    {
+      "epoch": 1.3807868933846346,
+      "grad_norm": 0.07323914105177598,
+      "learning_rate": 7.355906182006091e-06,
+      "loss": 0.4794,
+      "step": 2783
+    },
+    {
+      "epoch": 1.381283356087874,
+      "grad_norm": 0.07130416980727011,
+      "learning_rate": 7.354181886083843e-06,
+      "loss": 0.5066,
+      "step": 2784
+    },
+    {
+      "epoch": 1.3817798187911134,
+      "grad_norm": 0.07414140959170529,
+      "learning_rate": 7.3524572303626415e-06,
+      "loss": 0.481,
+      "step": 2785
+    },
+    {
+      "epoch": 1.3822762814943528,
+      "grad_norm": 0.07282023842599468,
+      "learning_rate": 7.3507322151060725e-06,
+      "loss": 0.5111,
+      "step": 2786
+    },
+    {
+      "epoch": 1.382772744197592,
+      "grad_norm": 0.07371876509619173,
+      "learning_rate": 7.3490068405777736e-06,
+      "loss": 0.4518,
+      "step": 2787
+    },
+    {
+      "epoch": 1.3832692069008314,
+      "grad_norm": 0.07242717641607478,
+      "learning_rate": 7.347281107041443e-06,
+      "loss": 0.4797,
+      "step": 2788
+    },
+    {
+      "epoch": 1.383765669604071,
+      "grad_norm": 0.0745013084369651,
+      "learning_rate": 7.345555014760832e-06,
+      "loss": 0.5479,
+      "step": 2789
+    },
+    {
+      "epoch": 1.3842621323073105,
+      "grad_norm": 0.07221130133499926,
+      "learning_rate": 7.343828563999744e-06,
+      "loss": 0.4875,
+      "step": 2790
+    },
+    {
+      "epoch": 1.3847585950105499,
+      "grad_norm": 0.07154952038380806,
+      "learning_rate": 7.342101755022041e-06,
+      "loss": 0.4759,
+      "step": 2791
+    },
+    {
+      "epoch": 1.3852550577137892,
+      "grad_norm": 0.07185399247769826,
+      "learning_rate": 7.340374588091638e-06,
+      "loss": 0.4831,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3857515204170285,
+      "grad_norm": 0.07022152259713238,
+      "learning_rate": 7.338647063472503e-06,
+      "loss": 0.4607,
+      "step": 2793
+    },
+    {
+      "epoch": 1.386247983120268,
+      "grad_norm": 0.07530734274608732,
+      "learning_rate": 7.336919181428661e-06,
+      "loss": 0.4538,
+      "step": 2794
+    },
+    {
+      "epoch": 1.3867444458235076,
+      "grad_norm": 0.0800273980847199,
+      "learning_rate": 7.335190942224193e-06,
+      "loss": 0.5545,
+      "step": 2795
+    },
+    {
+      "epoch": 1.387240908526747,
+      "grad_norm": 0.07118403274784829,
+      "learning_rate": 7.333462346123232e-06,
+      "loss": 0.51,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3877373712299863,
+      "grad_norm": 0.07267712142071889,
+      "learning_rate": 7.331733393389965e-06,
+      "loss": 0.4748,
+      "step": 2797
+    },
+    {
+      "epoch": 1.3882338339332256,
+      "grad_norm": 0.07444760909833327,
+      "learning_rate": 7.330004084288636e-06,
+      "loss": 0.5047,
+      "step": 2798
+    },
+    {
+      "epoch": 1.3887302966364652,
+      "grad_norm": 0.07464387036798184,
+      "learning_rate": 7.328274419083541e-06,
+      "loss": 0.5287,
+      "step": 2799
+    },
+    {
+      "epoch": 1.3892267593397047,
+      "grad_norm": 0.07055343310815258,
+      "learning_rate": 7.326544398039032e-06,
+      "loss": 0.4614,
+      "step": 2800
+    },
+    {
+      "epoch": 1.389723222042944,
+      "grad_norm": 0.07357320411155402,
+      "learning_rate": 7.324814021419514e-06,
+      "loss": 0.4733,
+      "step": 2801
+    },
+    {
+      "epoch": 1.3902196847461834,
+      "grad_norm": 0.06994120383038073,
+      "learning_rate": 7.32308328948945e-06,
+      "loss": 0.4627,
+      "step": 2802
+    },
+    {
+      "epoch": 1.3907161474494227,
+      "grad_norm": 0.07092734310261352,
+      "learning_rate": 7.321352202513352e-06,
+      "loss": 0.4506,
+      "step": 2803
+    },
+    {
+      "epoch": 1.3912126101526623,
+      "grad_norm": 0.0725892321942883,
+      "learning_rate": 7.31962076075579e-06,
+      "loss": 0.4783,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3917090728559016,
+      "grad_norm": 0.07270086524200547,
+      "learning_rate": 7.3178889644813875e-06,
+      "loss": 0.5169,
+      "step": 2805
+    },
+    {
+      "epoch": 1.3922055355591412,
+      "grad_norm": 0.0710477813020673,
+      "learning_rate": 7.316156813954821e-06,
+      "loss": 0.4634,
+      "step": 2806
+    },
+    {
+      "epoch": 1.3927019982623805,
+      "grad_norm": 0.0737346523243725,
+      "learning_rate": 7.314424309440822e-06,
+      "loss": 0.5189,
+      "step": 2807
+    },
+    {
+      "epoch": 1.3931984609656198,
+      "grad_norm": 0.07447635620495711,
+      "learning_rate": 7.312691451204178e-06,
+      "loss": 0.496,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3936949236688594,
+      "grad_norm": 0.07035201190670005,
+      "learning_rate": 7.310958239509725e-06,
+      "loss": 0.477,
+      "step": 2809
+    },
+    {
+      "epoch": 1.3941913863720987,
+      "grad_norm": 0.07253712416056099,
+      "learning_rate": 7.309224674622358e-06,
+      "loss": 0.4931,
+      "step": 2810
+    },
+    {
+      "epoch": 1.3946878490753383,
+      "grad_norm": 0.07026343239920593,
+      "learning_rate": 7.3074907568070266e-06,
+      "loss": 0.4995,
+      "step": 2811
+    },
+    {
+      "epoch": 1.3951843117785776,
+      "grad_norm": 0.07375095563668935,
+      "learning_rate": 7.3057564863287304e-06,
+      "loss": 0.4753,
+      "step": 2812
+    },
+    {
+      "epoch": 1.395680774481817,
+      "grad_norm": 0.0723516046829901,
+      "learning_rate": 7.304021863452525e-06,
+      "loss": 0.4861,
+      "step": 2813
+    },
+    {
+      "epoch": 1.3961772371850565,
+      "grad_norm": 0.07226116513343771,
+      "learning_rate": 7.30228688844352e-06,
+      "loss": 0.5004,
+      "step": 2814
+    },
+    {
+      "epoch": 1.3966736998882958,
+      "grad_norm": 0.07093942414377284,
+      "learning_rate": 7.3005515615668785e-06,
+      "loss": 0.5041,
+      "step": 2815
+    },
+    {
+      "epoch": 1.3971701625915354,
+      "grad_norm": 0.07556771300420517,
+      "learning_rate": 7.2988158830878174e-06,
+      "loss": 0.5133,
+      "step": 2816
+    },
+    {
+      "epoch": 1.3976666252947747,
+      "grad_norm": 0.07498104781592332,
+      "learning_rate": 7.297079853271607e-06,
+      "loss": 0.508,
+      "step": 2817
+    },
+    {
+      "epoch": 1.398163087998014,
+      "grad_norm": 0.07281928634469736,
+      "learning_rate": 7.295343472383573e-06,
+      "loss": 0.4827,
+      "step": 2818
+    },
+    {
+      "epoch": 1.3986595507012536,
+      "grad_norm": 0.07150001155960911,
+      "learning_rate": 7.293606740689091e-06,
+      "loss": 0.4767,
+      "step": 2819
+    },
+    {
+      "epoch": 1.399156013404493,
+      "grad_norm": 0.0696624964643031,
+      "learning_rate": 7.291869658453594e-06,
+      "loss": 0.4941,
+      "step": 2820
+    },
+    {
+      "epoch": 1.3996524761077325,
+      "grad_norm": 0.07346208294226515,
+      "learning_rate": 7.2901322259425675e-06,
+      "loss": 0.5148,
+      "step": 2821
+    },
+    {
+      "epoch": 1.4001489388109718,
+      "grad_norm": 0.07387505696804075,
+      "learning_rate": 7.28839444342155e-06,
+      "loss": 0.5054,
+      "step": 2822
+    },
+    {
+      "epoch": 1.4006454015142111,
+      "grad_norm": 0.07202026689241921,
+      "learning_rate": 7.286656311156133e-06,
+      "loss": 0.5098,
+      "step": 2823
+    },
+    {
+      "epoch": 1.4011418642174507,
+      "grad_norm": 0.07254707901648703,
+      "learning_rate": 7.2849178294119635e-06,
+      "loss": 0.4801,
+      "step": 2824
+    },
+    {
+      "epoch": 1.40163832692069,
+      "grad_norm": 0.07426847450752543,
+      "learning_rate": 7.283178998454738e-06,
+      "loss": 0.5112,
+      "step": 2825
+    },
+    {
+      "epoch": 1.4021347896239296,
+      "grad_norm": 0.0742033526393686,
+      "learning_rate": 7.281439818550211e-06,
+      "loss": 0.5388,
+      "step": 2826
+    },
+    {
+      "epoch": 1.402631252327169,
+      "grad_norm": 0.07310257455356693,
+      "learning_rate": 7.279700289964187e-06,
+      "loss": 0.4672,
+      "step": 2827
+    },
+    {
+      "epoch": 1.4031277150304082,
+      "grad_norm": 0.07575871999901848,
+      "learning_rate": 7.277960412962528e-06,
+      "loss": 0.4735,
+      "step": 2828
+    },
+    {
+      "epoch": 1.4036241777336478,
+      "grad_norm": 0.07061926156970885,
+      "learning_rate": 7.276220187811144e-06,
+      "loss": 0.5155,
+      "step": 2829
+    },
+    {
+      "epoch": 1.4041206404368871,
+      "grad_norm": 0.0745984279597669,
+      "learning_rate": 7.274479614776001e-06,
+      "loss": 0.4973,
+      "step": 2830
+    },
+    {
+      "epoch": 1.4046171031401267,
+      "grad_norm": 0.07292084635183517,
+      "learning_rate": 7.272738694123116e-06,
+      "loss": 0.5167,
+      "step": 2831
+    },
+    {
+      "epoch": 1.405113565843366,
+      "grad_norm": 0.07291076477353618,
+      "learning_rate": 7.270997426118563e-06,
+      "loss": 0.4759,
+      "step": 2832
+    },
+    {
+      "epoch": 1.4056100285466053,
+      "grad_norm": 0.06936979630980541,
+      "learning_rate": 7.269255811028464e-06,
+      "loss": 0.4838,
+      "step": 2833
+    },
+    {
+      "epoch": 1.4061064912498449,
+      "grad_norm": 0.06729679362346919,
+      "learning_rate": 7.267513849119001e-06,
+      "loss": 0.466,
+      "step": 2834
+    },
+    {
+      "epoch": 1.4066029539530842,
+      "grad_norm": 0.07515873936232569,
+      "learning_rate": 7.265771540656404e-06,
+      "loss": 0.5145,
+      "step": 2835
+    },
+    {
+      "epoch": 1.4070994166563238,
+      "grad_norm": 0.07335281716005654,
+      "learning_rate": 7.264028885906953e-06,
+      "loss": 0.5136,
+      "step": 2836
+    },
+    {
+      "epoch": 1.407595879359563,
+      "grad_norm": 0.07390288119336746,
+      "learning_rate": 7.26228588513699e-06,
+      "loss": 0.4892,
+      "step": 2837
+    },
+    {
+      "epoch": 1.4080923420628024,
+      "grad_norm": 0.07085400460434918,
+      "learning_rate": 7.260542538612902e-06,
+      "loss": 0.4625,
+      "step": 2838
+    },
+    {
+      "epoch": 1.408588804766042,
+      "grad_norm": 0.07153303266929528,
+      "learning_rate": 7.258798846601132e-06,
+      "loss": 0.4791,
+      "step": 2839
+    },
+    {
+      "epoch": 1.4090852674692813,
+      "grad_norm": 0.07431484763570721,
+      "learning_rate": 7.257054809368176e-06,
+      "loss": 0.4823,
+      "step": 2840
+    },
+    {
+      "epoch": 1.4095817301725209,
+      "grad_norm": 0.07242693206760291,
+      "learning_rate": 7.255310427180579e-06,
+      "loss": 0.4911,
+      "step": 2841
+    },
+    {
+      "epoch": 1.4100781928757602,
+      "grad_norm": 0.06913102542680617,
+      "learning_rate": 7.253565700304946e-06,
+      "loss": 0.4698,
+      "step": 2842
+    },
+    {
+      "epoch": 1.4105746555789995,
+      "grad_norm": 0.07432955588039054,
+      "learning_rate": 7.25182062900793e-06,
+      "loss": 0.4929,
+      "step": 2843
+    },
+    {
+      "epoch": 1.411071118282239,
+      "grad_norm": 0.07030319072588502,
+      "learning_rate": 7.250075213556234e-06,
+      "loss": 0.5069,
+      "step": 2844
+    },
+    {
+      "epoch": 1.4115675809854784,
+      "grad_norm": 0.06783577787923013,
+      "learning_rate": 7.24832945421662e-06,
+      "loss": 0.4538,
+      "step": 2845
+    },
+    {
+      "epoch": 1.412064043688718,
+      "grad_norm": 0.07226793652217142,
+      "learning_rate": 7.246583351255899e-06,
+      "loss": 0.4924,
+      "step": 2846
+    },
+    {
+      "epoch": 1.4125605063919573,
+      "grad_norm": 0.07379520096028509,
+      "learning_rate": 7.244836904940933e-06,
+      "loss": 0.4852,
+      "step": 2847
+    },
+    {
+      "epoch": 1.4130569690951966,
+      "grad_norm": 0.0696803836747072,
+      "learning_rate": 7.243090115538639e-06,
+      "loss": 0.4686,
+      "step": 2848
+    },
+    {
+      "epoch": 1.4135534317984362,
+      "grad_norm": 0.07439574493493696,
+      "learning_rate": 7.241342983315985e-06,
+      "loss": 0.517,
+      "step": 2849
+    },
+    {
+      "epoch": 1.4140498945016755,
+      "grad_norm": 0.07341222404947037,
+      "learning_rate": 7.239595508539995e-06,
+      "loss": 0.4851,
+      "step": 2850
+    },
+    {
+      "epoch": 1.414546357204915,
+      "grad_norm": 0.07253343281665837,
+      "learning_rate": 7.237847691477741e-06,
+      "loss": 0.4916,
+      "step": 2851
+    },
+    {
+      "epoch": 1.4150428199081544,
+      "grad_norm": 0.07444987740405767,
+      "learning_rate": 7.236099532396347e-06,
+      "loss": 0.4789,
+      "step": 2852
+    },
+    {
+      "epoch": 1.4155392826113937,
+      "grad_norm": 0.07095562862050957,
+      "learning_rate": 7.234351031562994e-06,
+      "loss": 0.4731,
+      "step": 2853
+    },
+    {
+      "epoch": 1.4160357453146333,
+      "grad_norm": 0.07390168113853861,
+      "learning_rate": 7.2326021892449105e-06,
+      "loss": 0.4769,
+      "step": 2854
+    },
+    {
+      "epoch": 1.4165322080178726,
+      "grad_norm": 0.07134364630264015,
+      "learning_rate": 7.230853005709378e-06,
+      "loss": 0.4674,
+      "step": 2855
+    },
+    {
+      "epoch": 1.4170286707211122,
+      "grad_norm": 0.07607452559927425,
+      "learning_rate": 7.229103481223735e-06,
+      "loss": 0.4917,
+      "step": 2856
+    },
+    {
+      "epoch": 1.4175251334243515,
+      "grad_norm": 0.07022996173204776,
+      "learning_rate": 7.227353616055364e-06,
+      "loss": 0.4808,
+      "step": 2857
+    },
+    {
+      "epoch": 1.4180215961275908,
+      "grad_norm": 0.06904514897261274,
+      "learning_rate": 7.225603410471707e-06,
+      "loss": 0.5062,
+      "step": 2858
+    },
+    {
+      "epoch": 1.4185180588308304,
+      "grad_norm": 0.07173966517539798,
+      "learning_rate": 7.223852864740251e-06,
+      "loss": 0.4777,
+      "step": 2859
+    },
+    {
+      "epoch": 1.4190145215340697,
+      "grad_norm": 0.07105840102585351,
+      "learning_rate": 7.222101979128544e-06,
+      "loss": 0.4838,
+      "step": 2860
+    },
+    {
+      "epoch": 1.4195109842373093,
+      "grad_norm": 0.0693002905559142,
+      "learning_rate": 7.220350753904177e-06,
+      "loss": 0.4807,
+      "step": 2861
+    },
+    {
+      "epoch": 1.4200074469405486,
+      "grad_norm": 0.07366922326583022,
+      "learning_rate": 7.218599189334799e-06,
+      "loss": 0.4974,
+      "step": 2862
+    },
+    {
+      "epoch": 1.420503909643788,
+      "grad_norm": 0.07105657624613147,
+      "learning_rate": 7.216847285688106e-06,
+      "loss": 0.495,
+      "step": 2863
+    },
+    {
+      "epoch": 1.4210003723470275,
+      "grad_norm": 0.0759918025493373,
+      "learning_rate": 7.215095043231852e-06,
+      "loss": 0.5141,
+      "step": 2864
+    },
+    {
+      "epoch": 1.4214968350502668,
+      "grad_norm": 0.07233086192834333,
+      "learning_rate": 7.213342462233835e-06,
+      "loss": 0.508,
+      "step": 2865
+    },
+    {
+      "epoch": 1.4219932977535064,
+      "grad_norm": 0.07490086058497883,
+      "learning_rate": 7.211589542961911e-06,
+      "loss": 0.5053,
+      "step": 2866
+    },
+    {
+      "epoch": 1.4224897604567457,
+      "grad_norm": 0.0710248696653551,
+      "learning_rate": 7.209836285683987e-06,
+      "loss": 0.4889,
+      "step": 2867
+    },
+    {
+      "epoch": 1.422986223159985,
+      "grad_norm": 0.07134268814850037,
+      "learning_rate": 7.208082690668017e-06,
+      "loss": 0.4672,
+      "step": 2868
+    },
+    {
+      "epoch": 1.4234826858632246,
+      "grad_norm": 0.0724334537858711,
+      "learning_rate": 7.206328758182013e-06,
+      "loss": 0.4637,
+      "step": 2869
+    },
+    {
+      "epoch": 1.423979148566464,
+      "grad_norm": 0.07165662584752473,
+      "learning_rate": 7.204574488494034e-06,
+      "loss": 0.473,
+      "step": 2870
+    },
+    {
+      "epoch": 1.4244756112697035,
+      "grad_norm": 0.07574200923260407,
+      "learning_rate": 7.202819881872191e-06,
+      "loss": 0.5067,
+      "step": 2871
+    },
+    {
+      "epoch": 1.4249720739729428,
+      "grad_norm": 0.07262595793489623,
+      "learning_rate": 7.2010649385846484e-06,
+      "loss": 0.4653,
+      "step": 2872
+    },
+    {
+      "epoch": 1.4254685366761821,
+      "grad_norm": 0.07190810770150632,
+      "learning_rate": 7.199309658899623e-06,
+      "loss": 0.5003,
+      "step": 2873
+    },
+    {
+      "epoch": 1.4259649993794217,
+      "grad_norm": 0.07146006958831883,
+      "learning_rate": 7.197554043085378e-06,
+      "loss": 0.477,
+      "step": 2874
+    },
+    {
+      "epoch": 1.426461462082661,
+      "grad_norm": 0.07287315250247191,
+      "learning_rate": 7.195798091410233e-06,
+      "loss": 0.4694,
+      "step": 2875
+    },
+    {
+      "epoch": 1.4269579247859006,
+      "grad_norm": 0.07145354823728287,
+      "learning_rate": 7.194041804142556e-06,
+      "loss": 0.5067,
+      "step": 2876
+    },
+    {
+      "epoch": 1.4274543874891399,
+      "grad_norm": 0.07374432502560274,
+      "learning_rate": 7.19228518155077e-06,
+      "loss": 0.5045,
+      "step": 2877
+    },
+    {
+      "epoch": 1.4279508501923792,
+      "grad_norm": 0.06905942432712643,
+      "learning_rate": 7.190528223903345e-06,
+      "loss": 0.4945,
+      "step": 2878
+    },
+    {
+      "epoch": 1.4284473128956188,
+      "grad_norm": 0.07235972820329792,
+      "learning_rate": 7.188770931468802e-06,
+      "loss": 0.4949,
+      "step": 2879
+    },
+    {
+      "epoch": 1.428943775598858,
+      "grad_norm": 0.07379607473437916,
+      "learning_rate": 7.187013304515715e-06,
+      "loss": 0.4942,
+      "step": 2880
+    },
+    {
+      "epoch": 1.4294402383020977,
+      "grad_norm": 0.07148160710339348,
+      "learning_rate": 7.185255343312712e-06,
+      "loss": 0.4643,
+      "step": 2881
+    },
+    {
+      "epoch": 1.429936701005337,
+      "grad_norm": 0.07115263600446331,
+      "learning_rate": 7.183497048128467e-06,
+      "loss": 0.4849,
+      "step": 2882
+    },
+    {
+      "epoch": 1.4304331637085763,
+      "grad_norm": 0.07313485288182806,
+      "learning_rate": 7.181738419231708e-06,
+      "loss": 0.4904,
+      "step": 2883
+    },
+    {
+      "epoch": 1.4309296264118159,
+      "grad_norm": 0.07403245647540267,
+      "learning_rate": 7.179979456891214e-06,
+      "loss": 0.5176,
+      "step": 2884
+    },
+    {
+      "epoch": 1.4314260891150552,
+      "grad_norm": 0.07430889143593936,
+      "learning_rate": 7.178220161375814e-06,
+      "loss": 0.4958,
+      "step": 2885
+    },
+    {
+      "epoch": 1.4319225518182948,
+      "grad_norm": 0.07146626429946829,
+      "learning_rate": 7.176460532954386e-06,
+      "loss": 0.4735,
+      "step": 2886
+    },
+    {
+      "epoch": 1.432419014521534,
+      "grad_norm": 0.07133541636516295,
+      "learning_rate": 7.174700571895863e-06,
+      "loss": 0.4905,
+      "step": 2887
+    },
+    {
+      "epoch": 1.4329154772247734,
+      "grad_norm": 0.07064549164282272,
+      "learning_rate": 7.172940278469225e-06,
+      "loss": 0.483,
+      "step": 2888
+    },
+    {
+      "epoch": 1.433411939928013,
+      "grad_norm": 0.0718501679320543,
+      "learning_rate": 7.171179652943507e-06,
+      "loss": 0.4555,
+      "step": 2889
+    },
+    {
+      "epoch": 1.4339084026312523,
+      "grad_norm": 0.07331695386537057,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.4912,
+      "step": 2890
+    },
+    {
+      "epoch": 1.4344048653344919,
+      "grad_norm": 0.07150200496801334,
+      "learning_rate": 7.167657406671212e-06,
+      "loss": 0.4946,
+      "step": 2891
+    },
+    {
+      "epoch": 1.4349013280377312,
+      "grad_norm": 0.06955567648220769,
+      "learning_rate": 7.165895786462953e-06,
+      "loss": 0.4716,
+      "step": 2892
+    },
+    {
+      "epoch": 1.4353977907409705,
+      "grad_norm": 0.07185399643784783,
+      "learning_rate": 7.164133835232252e-06,
+      "loss": 0.491,
+      "step": 2893
+    },
+    {
+      "epoch": 1.43589425344421,
+      "grad_norm": 0.0705794713571087,
+      "learning_rate": 7.162371553248393e-06,
+      "loss": 0.474,
+      "step": 2894
+    },
+    {
+      "epoch": 1.4363907161474494,
+      "grad_norm": 0.07619159211445736,
+      "learning_rate": 7.160608940780713e-06,
+      "loss": 0.4653,
+      "step": 2895
+    },
+    {
+      "epoch": 1.436887178850689,
+      "grad_norm": 0.07272819639781909,
+      "learning_rate": 7.158845998098598e-06,
+      "loss": 0.478,
+      "step": 2896
+    },
+    {
+      "epoch": 1.4373836415539283,
+      "grad_norm": 0.07353706245344895,
+      "learning_rate": 7.157082725471488e-06,
+      "loss": 0.5007,
+      "step": 2897
+    },
+    {
+      "epoch": 1.4378801042571676,
+      "grad_norm": 0.0710932004339634,
+      "learning_rate": 7.155319123168869e-06,
+      "loss": 0.5049,
+      "step": 2898
+    },
+    {
+      "epoch": 1.4383765669604072,
+      "grad_norm": 0.07036816763446126,
+      "learning_rate": 7.1535551914602804e-06,
+      "loss": 0.5002,
+      "step": 2899
+    },
+    {
+      "epoch": 1.4388730296636465,
+      "grad_norm": 0.07323068489466551,
+      "learning_rate": 7.15179093061531e-06,
+      "loss": 0.4988,
+      "step": 2900
+    },
+    {
+      "epoch": 1.439369492366886,
+      "grad_norm": 0.07103226205608117,
+      "learning_rate": 7.150026340903597e-06,
+      "loss": 0.4955,
+      "step": 2901
+    },
+    {
+      "epoch": 1.4398659550701254,
+      "grad_norm": 0.07179736174595919,
+      "learning_rate": 7.148261422594832e-06,
+      "loss": 0.4768,
+      "step": 2902
+    },
+    {
+      "epoch": 1.4403624177733647,
+      "grad_norm": 0.07171588950765849,
+      "learning_rate": 7.146496175958753e-06,
+      "loss": 0.4781,
+      "step": 2903
+    },
+    {
+      "epoch": 1.4408588804766043,
+      "grad_norm": 0.07386979687028543,
+      "learning_rate": 7.144730601265148e-06,
+      "loss": 0.4838,
+      "step": 2904
+    },
+    {
+      "epoch": 1.4413553431798436,
+      "grad_norm": 0.07078290659161332,
+      "learning_rate": 7.142964698783861e-06,
+      "loss": 0.4747,
+      "step": 2905
+    },
+    {
+      "epoch": 1.4418518058830831,
+      "grad_norm": 0.07123642642593622,
+      "learning_rate": 7.141198468784778e-06,
+      "loss": 0.4782,
+      "step": 2906
+    },
+    {
+      "epoch": 1.4423482685863225,
+      "grad_norm": 0.07365790707179352,
+      "learning_rate": 7.139431911537842e-06,
+      "loss": 0.4699,
+      "step": 2907
+    },
+    {
+      "epoch": 1.4428447312895618,
+      "grad_norm": 0.07225394246849842,
+      "learning_rate": 7.13766502731304e-06,
+      "loss": 0.5103,
+      "step": 2908
+    },
+    {
+      "epoch": 1.4433411939928014,
+      "grad_norm": 0.07235066461489845,
+      "learning_rate": 7.135897816380415e-06,
+      "loss": 0.5005,
+      "step": 2909
+    },
+    {
+      "epoch": 1.4438376566960407,
+      "grad_norm": 0.07100236241952648,
+      "learning_rate": 7.1341302790100546e-06,
+      "loss": 0.478,
+      "step": 2910
+    },
+    {
+      "epoch": 1.4443341193992802,
+      "grad_norm": 0.0699430966374579,
+      "learning_rate": 7.132362415472099e-06,
+      "loss": 0.5328,
+      "step": 2911
+    },
+    {
+      "epoch": 1.4448305821025196,
+      "grad_norm": 0.07467905849188111,
+      "learning_rate": 7.130594226036739e-06,
+      "loss": 0.5144,
+      "step": 2912
+    },
+    {
+      "epoch": 1.445327044805759,
+      "grad_norm": 0.06929574174293675,
+      "learning_rate": 7.128825710974212e-06,
+      "loss": 0.4824,
+      "step": 2913
+    },
+    {
+      "epoch": 1.4458235075089985,
+      "grad_norm": 0.07491153198596696,
+      "learning_rate": 7.127056870554807e-06,
+      "loss": 0.5402,
+      "step": 2914
+    },
+    {
+      "epoch": 1.4463199702122378,
+      "grad_norm": 0.07041488508688841,
+      "learning_rate": 7.125287705048867e-06,
+      "loss": 0.4973,
+      "step": 2915
+    },
+    {
+      "epoch": 1.4468164329154773,
+      "grad_norm": 0.07148072216701858,
+      "learning_rate": 7.123518214726775e-06,
+      "loss": 0.5064,
+      "step": 2916
+    },
+    {
+      "epoch": 1.4473128956187167,
+      "grad_norm": 0.07176842480825346,
+      "learning_rate": 7.121748399858974e-06,
+      "loss": 0.504,
+      "step": 2917
+    },
+    {
+      "epoch": 1.447809358321956,
+      "grad_norm": 0.07080899273892569,
+      "learning_rate": 7.1199782607159494e-06,
+      "loss": 0.4491,
+      "step": 2918
+    },
+    {
+      "epoch": 1.4483058210251956,
+      "grad_norm": 0.07295946872020888,
+      "learning_rate": 7.118207797568238e-06,
+      "loss": 0.5054,
+      "step": 2919
+    },
+    {
+      "epoch": 1.448802283728435,
+      "grad_norm": 0.07196455584447783,
+      "learning_rate": 7.116437010686427e-06,
+      "loss": 0.4746,
+      "step": 2920
+    },
+    {
+      "epoch": 1.4492987464316744,
+      "grad_norm": 0.0711831036893069,
+      "learning_rate": 7.1146659003411554e-06,
+      "loss": 0.5072,
+      "step": 2921
+    },
+    {
+      "epoch": 1.4497952091349138,
+      "grad_norm": 0.07047858314978178,
+      "learning_rate": 7.112894466803106e-06,
+      "loss": 0.4919,
+      "step": 2922
+    },
+    {
+      "epoch": 1.450291671838153,
+      "grad_norm": 0.07333316027850086,
+      "learning_rate": 7.1111227103430145e-06,
+      "loss": 0.5077,
+      "step": 2923
+    },
+    {
+      "epoch": 1.4507881345413927,
+      "grad_norm": 0.07565813630229153,
+      "learning_rate": 7.109350631231666e-06,
+      "loss": 0.4875,
+      "step": 2924
+    },
+    {
+      "epoch": 1.451284597244632,
+      "grad_norm": 0.07422857378436408,
+      "learning_rate": 7.107578229739895e-06,
+      "loss": 0.5018,
+      "step": 2925
+    },
+    {
+      "epoch": 1.4517810599478715,
+      "grad_norm": 0.07004713204598993,
+      "learning_rate": 7.105805506138586e-06,
+      "loss": 0.4845,
+      "step": 2926
+    },
+    {
+      "epoch": 1.4522775226511109,
+      "grad_norm": 0.07332153238042952,
+      "learning_rate": 7.104032460698668e-06,
+      "loss": 0.5136,
+      "step": 2927
+    },
+    {
+      "epoch": 1.4527739853543502,
+      "grad_norm": 0.07400446058276221,
+      "learning_rate": 7.102259093691122e-06,
+      "loss": 0.4801,
+      "step": 2928
+    },
+    {
+      "epoch": 1.4532704480575895,
+      "grad_norm": 0.0774114970002744,
+      "learning_rate": 7.100485405386982e-06,
+      "loss": 0.4807,
+      "step": 2929
+    },
+    {
+      "epoch": 1.453766910760829,
+      "grad_norm": 0.0734254664750455,
+      "learning_rate": 7.098711396057326e-06,
+      "loss": 0.4963,
+      "step": 2930
+    },
+    {
+      "epoch": 1.4542633734640686,
+      "grad_norm": 0.0694875881388645,
+      "learning_rate": 7.096937065973285e-06,
+      "loss": 0.4768,
+      "step": 2931
+    },
+    {
+      "epoch": 1.454759836167308,
+      "grad_norm": 0.07529738097359195,
+      "learning_rate": 7.095162415406034e-06,
+      "loss": 0.5126,
+      "step": 2932
+    },
+    {
+      "epoch": 1.4552562988705473,
+      "grad_norm": 0.07667108396414482,
+      "learning_rate": 7.093387444626801e-06,
+      "loss": 0.4962,
+      "step": 2933
+    },
+    {
+      "epoch": 1.4557527615737866,
+      "grad_norm": 0.07234479864658316,
+      "learning_rate": 7.0916121539068635e-06,
+      "loss": 0.4739,
+      "step": 2934
+    },
+    {
+      "epoch": 1.4562492242770262,
+      "grad_norm": 0.07242903221297743,
+      "learning_rate": 7.0898365435175435e-06,
+      "loss": 0.515,
+      "step": 2935
+    },
+    {
+      "epoch": 1.4567456869802657,
+      "grad_norm": 0.07507384079677762,
+      "learning_rate": 7.088060613730215e-06,
+      "loss": 0.5275,
+      "step": 2936
+    },
+    {
+      "epoch": 1.457242149683505,
+      "grad_norm": 0.07458928181042776,
+      "learning_rate": 7.0862843648163024e-06,
+      "loss": 0.5047,
+      "step": 2937
+    },
+    {
+      "epoch": 1.4577386123867444,
+      "grad_norm": 0.07596881392589007,
+      "learning_rate": 7.084507797047276e-06,
+      "loss": 0.5599,
+      "step": 2938
+    },
+    {
+      "epoch": 1.4582350750899837,
+      "grad_norm": 0.07243129312968741,
+      "learning_rate": 7.082730910694655e-06,
+      "loss": 0.4748,
+      "step": 2939
+    },
+    {
+      "epoch": 1.4587315377932233,
+      "grad_norm": 0.07248717359550946,
+      "learning_rate": 7.080953706030007e-06,
+      "loss": 0.5158,
+      "step": 2940
+    },
+    {
+      "epoch": 1.4592280004964628,
+      "grad_norm": 0.0731657949003018,
+      "learning_rate": 7.079176183324952e-06,
+      "loss": 0.4821,
+      "step": 2941
+    },
+    {
+      "epoch": 1.4597244631997022,
+      "grad_norm": 0.07581383406250891,
+      "learning_rate": 7.077398342851155e-06,
+      "loss": 0.5324,
+      "step": 2942
+    },
+    {
+      "epoch": 1.4602209259029415,
+      "grad_norm": 0.06881955653615653,
+      "learning_rate": 7.07562018488033e-06,
+      "loss": 0.4593,
+      "step": 2943
+    },
+    {
+      "epoch": 1.4607173886061808,
+      "grad_norm": 0.07432003151885896,
+      "learning_rate": 7.073841709684238e-06,
+      "loss": 0.5089,
+      "step": 2944
+    },
+    {
+      "epoch": 1.4612138513094204,
+      "grad_norm": 0.06822754389937877,
+      "learning_rate": 7.072062917534693e-06,
+      "loss": 0.5,
+      "step": 2945
+    },
+    {
+      "epoch": 1.4617103140126597,
+      "grad_norm": 0.0747311313862388,
+      "learning_rate": 7.070283808703553e-06,
+      "loss": 0.5058,
+      "step": 2946
+    },
+    {
+      "epoch": 1.4622067767158993,
+      "grad_norm": 0.07102663976968088,
+      "learning_rate": 7.068504383462729e-06,
+      "loss": 0.4682,
+      "step": 2947
+    },
+    {
+      "epoch": 1.4627032394191386,
+      "grad_norm": 0.07725801668567019,
+      "learning_rate": 7.0667246420841754e-06,
+      "loss": 0.5274,
+      "step": 2948
+    },
+    {
+      "epoch": 1.463199702122378,
+      "grad_norm": 0.07117806160550469,
+      "learning_rate": 7.064944584839898e-06,
+      "loss": 0.5052,
+      "step": 2949
+    },
+    {
+      "epoch": 1.4636961648256175,
+      "grad_norm": 0.07271174891546413,
+      "learning_rate": 7.06316421200195e-06,
+      "loss": 0.4849,
+      "step": 2950
+    },
+    {
+      "epoch": 1.4641926275288568,
+      "grad_norm": 0.07361228563816642,
+      "learning_rate": 7.061383523842431e-06,
+      "loss": 0.4672,
+      "step": 2951
+    },
+    {
+      "epoch": 1.4646890902320964,
+      "grad_norm": 0.07345309673612148,
+      "learning_rate": 7.0596025206334925e-06,
+      "loss": 0.4988,
+      "step": 2952
+    },
+    {
+      "epoch": 1.4651855529353357,
+      "grad_norm": 0.0750803661826231,
+      "learning_rate": 7.057821202647332e-06,
+      "loss": 0.4906,
+      "step": 2953
+    },
+    {
+      "epoch": 1.465682015638575,
+      "grad_norm": 0.0725746788290917,
+      "learning_rate": 7.056039570156197e-06,
+      "loss": 0.491,
+      "step": 2954
+    },
+    {
+      "epoch": 1.4661784783418146,
+      "grad_norm": 0.07240507467234208,
+      "learning_rate": 7.054257623432378e-06,
+      "loss": 0.5088,
+      "step": 2955
+    },
+    {
+      "epoch": 1.466674941045054,
+      "grad_norm": 0.07216315731512568,
+      "learning_rate": 7.052475362748219e-06,
+      "loss": 0.4697,
+      "step": 2956
+    },
+    {
+      "epoch": 1.4671714037482935,
+      "grad_norm": 0.07076545303447647,
+      "learning_rate": 7.05069278837611e-06,
+      "loss": 0.4697,
+      "step": 2957
+    },
+    {
+      "epoch": 1.4676678664515328,
+      "grad_norm": 0.07005482997181905,
+      "learning_rate": 7.048909900588488e-06,
+      "loss": 0.4614,
+      "step": 2958
+    },
+    {
+      "epoch": 1.4681643291547721,
+      "grad_norm": 0.07570364017599765,
+      "learning_rate": 7.047126699657842e-06,
+      "loss": 0.5188,
+      "step": 2959
+    },
+    {
+      "epoch": 1.4686607918580117,
+      "grad_norm": 0.07325286216717752,
+      "learning_rate": 7.045343185856701e-06,
+      "loss": 0.4743,
+      "step": 2960
+    },
+    {
+      "epoch": 1.469157254561251,
+      "grad_norm": 0.07542132612527033,
+      "learning_rate": 7.043559359457648e-06,
+      "loss": 0.5081,
+      "step": 2961
+    },
+    {
+      "epoch": 1.4696537172644906,
+      "grad_norm": 0.10995884031213667,
+      "learning_rate": 7.041775220733313e-06,
+      "loss": 0.4853,
+      "step": 2962
+    },
+    {
+      "epoch": 1.47015017996773,
+      "grad_norm": 0.07255210762978338,
+      "learning_rate": 7.039990769956374e-06,
+      "loss": 0.4904,
+      "step": 2963
+    },
+    {
+      "epoch": 1.4706466426709692,
+      "grad_norm": 0.07431973502861508,
+      "learning_rate": 7.038206007399555e-06,
+      "loss": 0.5069,
+      "step": 2964
+    },
+    {
+      "epoch": 1.4711431053742088,
+      "grad_norm": 0.07192623705516626,
+      "learning_rate": 7.036420933335627e-06,
+      "loss": 0.4764,
+      "step": 2965
+    },
+    {
+      "epoch": 1.4716395680774481,
+      "grad_norm": 0.07182207582734508,
+      "learning_rate": 7.034635548037412e-06,
+      "loss": 0.4933,
+      "step": 2966
+    },
+    {
+      "epoch": 1.4721360307806877,
+      "grad_norm": 0.07525017370629115,
+      "learning_rate": 7.032849851777774e-06,
+      "loss": 0.4918,
+      "step": 2967
+    },
+    {
+      "epoch": 1.472632493483927,
+      "grad_norm": 0.07277899449762383,
+      "learning_rate": 7.031063844829632e-06,
+      "loss": 0.4659,
+      "step": 2968
+    },
+    {
+      "epoch": 1.4731289561871663,
+      "grad_norm": 0.0699788593870598,
+      "learning_rate": 7.029277527465948e-06,
+      "loss": 0.4704,
+      "step": 2969
+    },
+    {
+      "epoch": 1.4736254188904059,
+      "grad_norm": 0.07548125236861504,
+      "learning_rate": 7.027490899959729e-06,
+      "loss": 0.5407,
+      "step": 2970
+    },
+    {
+      "epoch": 1.4741218815936452,
+      "grad_norm": 0.07361918602319495,
+      "learning_rate": 7.025703962584035e-06,
+      "loss": 0.5167,
+      "step": 2971
+    },
+    {
+      "epoch": 1.4746183442968848,
+      "grad_norm": 0.07164684601867893,
+      "learning_rate": 7.023916715611969e-06,
+      "loss": 0.476,
+      "step": 2972
+    },
+    {
+      "epoch": 1.475114807000124,
+      "grad_norm": 0.07342930896696674,
+      "learning_rate": 7.022129159316685e-06,
+      "loss": 0.4862,
+      "step": 2973
+    },
+    {
+      "epoch": 1.4756112697033634,
+      "grad_norm": 0.07448008055751498,
+      "learning_rate": 7.020341293971383e-06,
+      "loss": 0.5077,
+      "step": 2974
+    },
+    {
+      "epoch": 1.476107732406603,
+      "grad_norm": 0.07590357611172563,
+      "learning_rate": 7.018553119849306e-06,
+      "loss": 0.4977,
+      "step": 2975
+    },
+    {
+      "epoch": 1.4766041951098423,
+      "grad_norm": 0.07000464301210924,
+      "learning_rate": 7.0167646372237495e-06,
+      "loss": 0.4695,
+      "step": 2976
+    },
+    {
+      "epoch": 1.4771006578130819,
+      "grad_norm": 0.06744354825036339,
+      "learning_rate": 7.014975846368055e-06,
+      "loss": 0.4525,
+      "step": 2977
+    },
+    {
+      "epoch": 1.4775971205163212,
+      "grad_norm": 0.07206037360333768,
+      "learning_rate": 7.013186747555611e-06,
+      "loss": 0.4837,
+      "step": 2978
+    },
+    {
+      "epoch": 1.4780935832195605,
+      "grad_norm": 0.07395975276717683,
+      "learning_rate": 7.01139734105985e-06,
+      "loss": 0.4924,
+      "step": 2979
+    },
+    {
+      "epoch": 1.4785900459228,
+      "grad_norm": 0.07113804829814217,
+      "learning_rate": 7.009607627154257e-06,
+      "loss": 0.4946,
+      "step": 2980
+    },
+    {
+      "epoch": 1.4790865086260394,
+      "grad_norm": 0.07233617024386177,
+      "learning_rate": 7.0078176061123595e-06,
+      "loss": 0.5144,
+      "step": 2981
+    },
+    {
+      "epoch": 1.479582971329279,
+      "grad_norm": 0.07350380568403078,
+      "learning_rate": 7.006027278207734e-06,
+      "loss": 0.4951,
+      "step": 2982
+    },
+    {
+      "epoch": 1.4800794340325183,
+      "grad_norm": 0.07026993393268165,
+      "learning_rate": 7.004236643714002e-06,
+      "loss": 0.4697,
+      "step": 2983
+    },
+    {
+      "epoch": 1.4805758967357576,
+      "grad_norm": 0.07222164558693017,
+      "learning_rate": 7.002445702904835e-06,
+      "loss": 0.4953,
+      "step": 2984
+    },
+    {
+      "epoch": 1.4810723594389972,
+      "grad_norm": 0.07143971798306151,
+      "learning_rate": 7.000654456053949e-06,
+      "loss": 0.4687,
+      "step": 2985
+    },
+    {
+      "epoch": 1.4815688221422365,
+      "grad_norm": 0.07459913209718101,
+      "learning_rate": 6.998862903435109e-06,
+      "loss": 0.4708,
+      "step": 2986
+    },
+    {
+      "epoch": 1.482065284845476,
+      "grad_norm": 0.07352362619317326,
+      "learning_rate": 6.997071045322123e-06,
+      "loss": 0.5225,
+      "step": 2987
+    },
+    {
+      "epoch": 1.4825617475487154,
+      "grad_norm": 0.07326362477169654,
+      "learning_rate": 6.995278881988847e-06,
+      "loss": 0.4651,
+      "step": 2988
+    },
+    {
+      "epoch": 1.4830582102519547,
+      "grad_norm": 0.07617706210167993,
+      "learning_rate": 6.993486413709187e-06,
+      "loss": 0.4713,
+      "step": 2989
+    },
+    {
+      "epoch": 1.4835546729551943,
+      "grad_norm": 0.07272746269808053,
+      "learning_rate": 6.991693640757091e-06,
+      "loss": 0.5036,
+      "step": 2990
+    },
+    {
+      "epoch": 1.4840511356584336,
+      "grad_norm": 0.0735407652235623,
+      "learning_rate": 6.989900563406557e-06,
+      "loss": 0.5171,
+      "step": 2991
+    },
+    {
+      "epoch": 1.4845475983616732,
+      "grad_norm": 0.07481908248468541,
+      "learning_rate": 6.988107181931627e-06,
+      "loss": 0.4865,
+      "step": 2992
+    },
+    {
+      "epoch": 1.4850440610649125,
+      "grad_norm": 0.07569078793936909,
+      "learning_rate": 6.986313496606392e-06,
+      "loss": 0.5161,
+      "step": 2993
+    },
+    {
+      "epoch": 1.4855405237681518,
+      "grad_norm": 0.06977321199854734,
+      "learning_rate": 6.984519507704985e-06,
+      "loss": 0.5026,
+      "step": 2994
+    },
+    {
+      "epoch": 1.4860369864713914,
+      "grad_norm": 0.07160834128804278,
+      "learning_rate": 6.982725215501592e-06,
+      "loss": 0.4607,
+      "step": 2995
+    },
+    {
+      "epoch": 1.4865334491746307,
+      "grad_norm": 0.07161162488316418,
+      "learning_rate": 6.980930620270441e-06,
+      "loss": 0.518,
+      "step": 2996
+    },
+    {
+      "epoch": 1.4870299118778703,
+      "grad_norm": 0.07373692206462576,
+      "learning_rate": 6.9791357222858054e-06,
+      "loss": 0.4838,
+      "step": 2997
+    },
+    {
+      "epoch": 1.4875263745811096,
+      "grad_norm": 0.07201153669770845,
+      "learning_rate": 6.977340521822009e-06,
+      "loss": 0.4788,
+      "step": 2998
+    },
+    {
+      "epoch": 1.488022837284349,
+      "grad_norm": 0.07630818774232069,
+      "learning_rate": 6.975545019153418e-06,
+      "loss": 0.4986,
+      "step": 2999
+    },
+    {
+      "epoch": 1.4885192999875885,
+      "grad_norm": 0.07161586866226113,
+      "learning_rate": 6.973749214554445e-06,
+      "loss": 0.4932,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4890157626908278,
+      "grad_norm": 0.07103890527803151,
+      "learning_rate": 6.9719531082995516e-06,
+      "loss": 0.4894,
+      "step": 3001
+    },
+    {
+      "epoch": 1.4895122253940674,
+      "grad_norm": 0.070312542505285,
+      "learning_rate": 6.970156700663244e-06,
+      "loss": 0.481,
+      "step": 3002
+    },
+    {
+      "epoch": 1.4900086880973067,
+      "grad_norm": 0.07497732442687259,
+      "learning_rate": 6.968359991920073e-06,
+      "loss": 0.5113,
+      "step": 3003
+    },
+    {
+      "epoch": 1.490505150800546,
+      "grad_norm": 0.07381885890576345,
+      "learning_rate": 6.9665629823446375e-06,
+      "loss": 0.4943,
+      "step": 3004
+    },
+    {
+      "epoch": 1.4910016135037856,
+      "grad_norm": 0.07282606616955749,
+      "learning_rate": 6.964765672211582e-06,
+      "loss": 0.4979,
+      "step": 3005
+    },
+    {
+      "epoch": 1.491498076207025,
+      "grad_norm": 0.07314271604221277,
+      "learning_rate": 6.962968061795596e-06,
+      "loss": 0.564,
+      "step": 3006
+    },
+    {
+      "epoch": 1.4919945389102645,
+      "grad_norm": 0.07348280758443781,
+      "learning_rate": 6.9611701513714165e-06,
+      "loss": 0.5347,
+      "step": 3007
+    },
+    {
+      "epoch": 1.4924910016135038,
+      "grad_norm": 0.07602272659194373,
+      "learning_rate": 6.959371941213824e-06,
+      "loss": 0.4952,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4929874643167431,
+      "grad_norm": 0.07447850718939307,
+      "learning_rate": 6.957573431597646e-06,
+      "loss": 0.519,
+      "step": 3009
+    },
+    {
+      "epoch": 1.4934839270199827,
+      "grad_norm": 0.06810881816124986,
+      "learning_rate": 6.955774622797755e-06,
+      "loss": 0.4765,
+      "step": 3010
+    },
+    {
+      "epoch": 1.493980389723222,
+      "grad_norm": 0.07211306225922738,
+      "learning_rate": 6.953975515089073e-06,
+      "loss": 0.4819,
+      "step": 3011
+    },
+    {
+      "epoch": 1.4944768524264616,
+      "grad_norm": 0.07325489823289166,
+      "learning_rate": 6.952176108746563e-06,
+      "loss": 0.4665,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4949733151297009,
+      "grad_norm": 0.07668839966145699,
+      "learning_rate": 6.950376404045235e-06,
+      "loss": 0.5321,
+      "step": 3013
+    },
+    {
+      "epoch": 1.4954697778329402,
+      "grad_norm": 0.06965813102342724,
+      "learning_rate": 6.948576401260147e-06,
+      "loss": 0.456,
+      "step": 3014
+    },
+    {
+      "epoch": 1.4959662405361798,
+      "grad_norm": 0.0714214738156806,
+      "learning_rate": 6.946776100666397e-06,
+      "loss": 0.4824,
+      "step": 3015
+    },
+    {
+      "epoch": 1.496462703239419,
+      "grad_norm": 0.0728874617313902,
+      "learning_rate": 6.9449755025391355e-06,
+      "loss": 0.4896,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4969591659426587,
+      "grad_norm": 0.07702335377688295,
+      "learning_rate": 6.943174607153553e-06,
+      "loss": 0.4881,
+      "step": 3017
+    },
+    {
+      "epoch": 1.497455628645898,
+      "grad_norm": 0.07239817990267242,
+      "learning_rate": 6.941373414784889e-06,
+      "loss": 0.4867,
+      "step": 3018
+    },
+    {
+      "epoch": 1.4979520913491373,
+      "grad_norm": 0.07323119244865894,
+      "learning_rate": 6.939571925708426e-06,
+      "loss": 0.493,
+      "step": 3019
+    },
+    {
+      "epoch": 1.4984485540523769,
+      "grad_norm": 0.07165300199457361,
+      "learning_rate": 6.937770140199491e-06,
+      "loss": 0.4982,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4989450167556162,
+      "grad_norm": 0.06964794952545811,
+      "learning_rate": 6.935968058533462e-06,
+      "loss": 0.4833,
+      "step": 3021
+    },
+    {
+      "epoch": 1.4994414794588558,
+      "grad_norm": 0.07611127268165498,
+      "learning_rate": 6.934165680985756e-06,
+      "loss": 0.5043,
+      "step": 3022
+    },
+    {
+      "epoch": 1.499937942162095,
+      "grad_norm": 0.07052388962092157,
+      "learning_rate": 6.932363007831837e-06,
+      "loss": 0.4756,
+      "step": 3023
+    },
+    {
+      "epoch": 1.5004344048653344,
+      "grad_norm": 0.06976733179848893,
+      "learning_rate": 6.930560039347216e-06,
+      "loss": 0.477,
+      "step": 3024
+    },
+    {
+      "epoch": 1.5004344048653344,
+      "eval_loss": 0.5206817984580994,
+      "eval_runtime": 259.3872,
+      "eval_samples_per_second": 117.018,
+      "eval_steps_per_second": 14.631,
+      "step": 3024
+    },
+    {
+      "epoch": 1.5009308675685737,
+      "grad_norm": 0.07903128042684261,
+      "learning_rate": 6.928756775807447e-06,
+      "loss": 0.4959,
+      "step": 3025
+    },
+    {
+      "epoch": 1.5014273302718133,
+      "grad_norm": 0.08126205695713004,
+      "learning_rate": 6.926953217488129e-06,
+      "loss": 0.5213,
+      "step": 3026
+    },
+    {
+      "epoch": 1.5019237929750529,
+      "grad_norm": 0.07352870292325349,
+      "learning_rate": 6.925149364664909e-06,
+      "loss": 0.4891,
+      "step": 3027
+    },
+    {
+      "epoch": 1.5024202556782922,
+      "grad_norm": 0.07176814755640172,
+      "learning_rate": 6.923345217613477e-06,
+      "loss": 0.5191,
+      "step": 3028
+    },
+    {
+      "epoch": 1.5029167183815315,
+      "grad_norm": 0.07282967027183829,
+      "learning_rate": 6.921540776609564e-06,
+      "loss": 0.4886,
+      "step": 3029
+    },
+    {
+      "epoch": 1.5034131810847708,
+      "grad_norm": 0.07341157454941727,
+      "learning_rate": 6.919736041928956e-06,
+      "loss": 0.4969,
+      "step": 3030
+    },
+    {
+      "epoch": 1.5039096437880104,
+      "grad_norm": 0.07174957878205697,
+      "learning_rate": 6.9179310138474734e-06,
+      "loss": 0.5151,
+      "step": 3031
+    },
+    {
+      "epoch": 1.50440610649125,
+      "grad_norm": 0.07186055210570085,
+      "learning_rate": 6.916125692640987e-06,
+      "loss": 0.4882,
+      "step": 3032
+    },
+    {
+      "epoch": 1.5049025691944893,
+      "grad_norm": 0.06818602296286949,
+      "learning_rate": 6.91432007858541e-06,
+      "loss": 0.4667,
+      "step": 3033
+    },
+    {
+      "epoch": 1.5053990318977286,
+      "grad_norm": 0.07199100368580628,
+      "learning_rate": 6.912514171956704e-06,
+      "loss": 0.5049,
+      "step": 3034
+    },
+    {
+      "epoch": 1.505895494600968,
+      "grad_norm": 0.07260935212384795,
+      "learning_rate": 6.9107079730308724e-06,
+      "loss": 0.4931,
+      "step": 3035
+    },
+    {
+      "epoch": 1.5063919573042075,
+      "grad_norm": 0.0727951239071628,
+      "learning_rate": 6.908901482083961e-06,
+      "loss": 0.5121,
+      "step": 3036
+    },
+    {
+      "epoch": 1.506888420007447,
+      "grad_norm": 0.07189572060344285,
+      "learning_rate": 6.907094699392066e-06,
+      "loss": 0.4765,
+      "step": 3037
+    },
+    {
+      "epoch": 1.5073848827106864,
+      "grad_norm": 0.07077554594688146,
+      "learning_rate": 6.905287625231325e-06,
+      "loss": 0.4861,
+      "step": 3038
+    },
+    {
+      "epoch": 1.5078813454139257,
+      "grad_norm": 0.07032031769025249,
+      "learning_rate": 6.90348025987792e-06,
+      "loss": 0.4684,
+      "step": 3039
+    },
+    {
+      "epoch": 1.508377808117165,
+      "grad_norm": 0.07212837688355783,
+      "learning_rate": 6.901672603608076e-06,
+      "loss": 0.493,
+      "step": 3040
+    },
+    {
+      "epoch": 1.5088742708204046,
+      "grad_norm": 0.07459184873506873,
+      "learning_rate": 6.899864656698066e-06,
+      "loss": 0.4786,
+      "step": 3041
+    },
+    {
+      "epoch": 1.5093707335236441,
+      "grad_norm": 0.0695658996301349,
+      "learning_rate": 6.898056419424204e-06,
+      "loss": 0.4674,
+      "step": 3042
+    },
+    {
+      "epoch": 1.5098671962268835,
+      "grad_norm": 0.07066812767150565,
+      "learning_rate": 6.8962478920628505e-06,
+      "loss": 0.4729,
+      "step": 3043
+    },
+    {
+      "epoch": 1.5103636589301228,
+      "grad_norm": 0.07346041759217889,
+      "learning_rate": 6.894439074890413e-06,
+      "loss": 0.5077,
+      "step": 3044
+    },
+    {
+      "epoch": 1.5108601216333621,
+      "grad_norm": 0.07619677857322961,
+      "learning_rate": 6.892629968183338e-06,
+      "loss": 0.513,
+      "step": 3045
+    },
+    {
+      "epoch": 1.5113565843366017,
+      "grad_norm": 0.07663086607755477,
+      "learning_rate": 6.890820572218118e-06,
+      "loss": 0.4849,
+      "step": 3046
+    },
+    {
+      "epoch": 1.5118530470398412,
+      "grad_norm": 0.07527384356601459,
+      "learning_rate": 6.88901088727129e-06,
+      "loss": 0.488,
+      "step": 3047
+    },
+    {
+      "epoch": 1.5123495097430806,
+      "grad_norm": 0.07613931254967439,
+      "learning_rate": 6.887200913619435e-06,
+      "loss": 0.5246,
+      "step": 3048
+    },
+    {
+      "epoch": 1.51284597244632,
+      "grad_norm": 0.07351039591100084,
+      "learning_rate": 6.885390651539181e-06,
+      "loss": 0.4986,
+      "step": 3049
+    },
+    {
+      "epoch": 1.5133424351495592,
+      "grad_norm": 0.07283507380458602,
+      "learning_rate": 6.883580101307195e-06,
+      "loss": 0.4696,
+      "step": 3050
+    },
+    {
+      "epoch": 1.5138388978527988,
+      "grad_norm": 0.07057028032237944,
+      "learning_rate": 6.881769263200192e-06,
+      "loss": 0.4747,
+      "step": 3051
+    },
+    {
+      "epoch": 1.5143353605560383,
+      "grad_norm": 0.07179846588719133,
+      "learning_rate": 6.8799581374949276e-06,
+      "loss": 0.4866,
+      "step": 3052
+    },
+    {
+      "epoch": 1.5148318232592777,
+      "grad_norm": 0.07496502673400987,
+      "learning_rate": 6.878146724468205e-06,
+      "loss": 0.5174,
+      "step": 3053
+    },
+    {
+      "epoch": 1.515328285962517,
+      "grad_norm": 0.07430870063782538,
+      "learning_rate": 6.876335024396872e-06,
+      "loss": 0.5139,
+      "step": 3054
+    },
+    {
+      "epoch": 1.5158247486657563,
+      "grad_norm": 0.07581456093878954,
+      "learning_rate": 6.874523037557812e-06,
+      "loss": 0.4792,
+      "step": 3055
+    },
+    {
+      "epoch": 1.516321211368996,
+      "grad_norm": 0.07193302206420817,
+      "learning_rate": 6.8727107642279645e-06,
+      "loss": 0.4771,
+      "step": 3056
+    },
+    {
+      "epoch": 1.5168176740722354,
+      "grad_norm": 0.07423534271844814,
+      "learning_rate": 6.8708982046843005e-06,
+      "loss": 0.4922,
+      "step": 3057
+    },
+    {
+      "epoch": 1.5173141367754748,
+      "grad_norm": 0.06944464704893721,
+      "learning_rate": 6.869085359203844e-06,
+      "loss": 0.4767,
+      "step": 3058
+    },
+    {
+      "epoch": 1.517810599478714,
+      "grad_norm": 0.07422967148546548,
+      "learning_rate": 6.8672722280636595e-06,
+      "loss": 0.4891,
+      "step": 3059
+    },
+    {
+      "epoch": 1.5183070621819534,
+      "grad_norm": 0.07764248517174992,
+      "learning_rate": 6.865458811540854e-06,
+      "loss": 0.5052,
+      "step": 3060
+    },
+    {
+      "epoch": 1.518803524885193,
+      "grad_norm": 0.07449831386708612,
+      "learning_rate": 6.863645109912581e-06,
+      "loss": 0.4694,
+      "step": 3061
+    },
+    {
+      "epoch": 1.5192999875884325,
+      "grad_norm": 0.06960098049635088,
+      "learning_rate": 6.861831123456033e-06,
+      "loss": 0.4961,
+      "step": 3062
+    },
+    {
+      "epoch": 1.5197964502916719,
+      "grad_norm": 0.0726928648817254,
+      "learning_rate": 6.86001685244845e-06,
+      "loss": 0.4877,
+      "step": 3063
+    },
+    {
+      "epoch": 1.5202929129949112,
+      "grad_norm": 0.07271711349404424,
+      "learning_rate": 6.858202297167114e-06,
+      "loss": 0.4965,
+      "step": 3064
+    },
+    {
+      "epoch": 1.5207893756981505,
+      "grad_norm": 0.07172467737507537,
+      "learning_rate": 6.8563874578893505e-06,
+      "loss": 0.4776,
+      "step": 3065
+    },
+    {
+      "epoch": 1.52128583840139,
+      "grad_norm": 0.06994923402408372,
+      "learning_rate": 6.854572334892531e-06,
+      "loss": 0.4631,
+      "step": 3066
+    },
+    {
+      "epoch": 1.5217823011046296,
+      "grad_norm": 0.07151187949027787,
+      "learning_rate": 6.852756928454064e-06,
+      "loss": 0.4661,
+      "step": 3067
+    },
+    {
+      "epoch": 1.522278763807869,
+      "grad_norm": 0.07336019831993339,
+      "learning_rate": 6.850941238851408e-06,
+      "loss": 0.4841,
+      "step": 3068
+    },
+    {
+      "epoch": 1.5227752265111083,
+      "grad_norm": 0.07386078183875347,
+      "learning_rate": 6.84912526636206e-06,
+      "loss": 0.523,
+      "step": 3069
+    },
+    {
+      "epoch": 1.5232716892143476,
+      "grad_norm": 0.07181896557591305,
+      "learning_rate": 6.8473090112635656e-06,
+      "loss": 0.4877,
+      "step": 3070
+    },
+    {
+      "epoch": 1.5237681519175872,
+      "grad_norm": 0.07203431208701555,
+      "learning_rate": 6.845492473833506e-06,
+      "loss": 0.4741,
+      "step": 3071
+    },
+    {
+      "epoch": 1.5242646146208267,
+      "grad_norm": 0.07472847969899919,
+      "learning_rate": 6.843675654349513e-06,
+      "loss": 0.4788,
+      "step": 3072
+    },
+    {
+      "epoch": 1.524761077324066,
+      "grad_norm": 0.07256373704505083,
+      "learning_rate": 6.841858553089258e-06,
+      "loss": 0.4665,
+      "step": 3073
+    },
+    {
+      "epoch": 1.5252575400273054,
+      "grad_norm": 0.0722750199855971,
+      "learning_rate": 6.840041170330454e-06,
+      "loss": 0.4601,
+      "step": 3074
+    },
+    {
+      "epoch": 1.5257540027305447,
+      "grad_norm": 0.0721037848913084,
+      "learning_rate": 6.838223506350859e-06,
+      "loss": 0.4798,
+      "step": 3075
+    },
+    {
+      "epoch": 1.5262504654337843,
+      "grad_norm": 0.07215571549842134,
+      "learning_rate": 6.836405561428276e-06,
+      "loss": 0.4738,
+      "step": 3076
+    },
+    {
+      "epoch": 1.5267469281370238,
+      "grad_norm": 0.07342767459463316,
+      "learning_rate": 6.834587335840549e-06,
+      "loss": 0.5264,
+      "step": 3077
+    },
+    {
+      "epoch": 1.5272433908402632,
+      "grad_norm": 0.0707351475232259,
+      "learning_rate": 6.832768829865561e-06,
+      "loss": 0.4981,
+      "step": 3078
+    },
+    {
+      "epoch": 1.5277398535435025,
+      "grad_norm": 0.07421547342728028,
+      "learning_rate": 6.830950043781245e-06,
+      "loss": 0.5003,
+      "step": 3079
+    },
+    {
+      "epoch": 1.5282363162467418,
+      "grad_norm": 0.07731539811606819,
+      "learning_rate": 6.82913097786557e-06,
+      "loss": 0.4683,
+      "step": 3080
+    },
+    {
+      "epoch": 1.5287327789499814,
+      "grad_norm": 0.06928467278023695,
+      "learning_rate": 6.827311632396553e-06,
+      "loss": 0.4626,
+      "step": 3081
+    },
+    {
+      "epoch": 1.529229241653221,
+      "grad_norm": 0.0745738000242575,
+      "learning_rate": 6.825492007652255e-06,
+      "loss": 0.5103,
+      "step": 3082
+    },
+    {
+      "epoch": 1.5297257043564603,
+      "grad_norm": 0.07364417846410318,
+      "learning_rate": 6.82367210391077e-06,
+      "loss": 0.4731,
+      "step": 3083
+    },
+    {
+      "epoch": 1.5302221670596996,
+      "grad_norm": 0.07028000247402512,
+      "learning_rate": 6.821851921450246e-06,
+      "loss": 0.4717,
+      "step": 3084
+    },
+    {
+      "epoch": 1.530718629762939,
+      "grad_norm": 0.07445663844660433,
+      "learning_rate": 6.820031460548865e-06,
+      "loss": 0.4728,
+      "step": 3085
+    },
+    {
+      "epoch": 1.5312150924661785,
+      "grad_norm": 0.07642475577058762,
+      "learning_rate": 6.818210721484859e-06,
+      "loss": 0.4866,
+      "step": 3086
+    },
+    {
+      "epoch": 1.531711555169418,
+      "grad_norm": 0.07272701730211181,
+      "learning_rate": 6.816389704536496e-06,
+      "loss": 0.5144,
+      "step": 3087
+    },
+    {
+      "epoch": 1.5322080178726574,
+      "grad_norm": 0.07226212774546704,
+      "learning_rate": 6.8145684099820915e-06,
+      "loss": 0.4789,
+      "step": 3088
+    },
+    {
+      "epoch": 1.5327044805758967,
+      "grad_norm": 0.06972510732857357,
+      "learning_rate": 6.812746838099998e-06,
+      "loss": 0.4781,
+      "step": 3089
+    },
+    {
+      "epoch": 1.533200943279136,
+      "grad_norm": 0.07047457761812748,
+      "learning_rate": 6.8109249891686165e-06,
+      "loss": 0.4789,
+      "step": 3090
+    },
+    {
+      "epoch": 1.5336974059823756,
+      "grad_norm": 0.07024127145574385,
+      "learning_rate": 6.809102863466386e-06,
+      "loss": 0.471,
+      "step": 3091
+    },
+    {
+      "epoch": 1.5341938686856151,
+      "grad_norm": 0.07162497294503184,
+      "learning_rate": 6.80728046127179e-06,
+      "loss": 0.5013,
+      "step": 3092
+    },
+    {
+      "epoch": 1.5346903313888545,
+      "grad_norm": 0.07229883941244337,
+      "learning_rate": 6.805457782863354e-06,
+      "loss": 0.497,
+      "step": 3093
+    },
+    {
+      "epoch": 1.5351867940920938,
+      "grad_norm": 0.07135143607274454,
+      "learning_rate": 6.803634828519643e-06,
+      "loss": 0.4552,
+      "step": 3094
+    },
+    {
+      "epoch": 1.5356832567953331,
+      "grad_norm": 0.0664211340599298,
+      "learning_rate": 6.801811598519268e-06,
+      "loss": 0.4231,
+      "step": 3095
+    },
+    {
+      "epoch": 1.5361797194985727,
+      "grad_norm": 0.07023830081055071,
+      "learning_rate": 6.799988093140879e-06,
+      "loss": 0.489,
+      "step": 3096
+    },
+    {
+      "epoch": 1.5366761822018122,
+      "grad_norm": 0.07314527625696632,
+      "learning_rate": 6.7981643126631714e-06,
+      "loss": 0.5141,
+      "step": 3097
+    },
+    {
+      "epoch": 1.5371726449050516,
+      "grad_norm": 0.07361069625191462,
+      "learning_rate": 6.796340257364879e-06,
+      "loss": 0.5258,
+      "step": 3098
+    },
+    {
+      "epoch": 1.537669107608291,
+      "grad_norm": 0.07198866226971654,
+      "learning_rate": 6.794515927524783e-06,
+      "loss": 0.4824,
+      "step": 3099
+    },
+    {
+      "epoch": 1.5381655703115302,
+      "grad_norm": 0.0690803990888876,
+      "learning_rate": 6.792691323421698e-06,
+      "loss": 0.4608,
+      "step": 3100
+    },
+    {
+      "epoch": 1.5386620330147698,
+      "grad_norm": 0.07486582879406296,
+      "learning_rate": 6.790866445334489e-06,
+      "loss": 0.5164,
+      "step": 3101
+    },
+    {
+      "epoch": 1.5391584957180093,
+      "grad_norm": 0.07599038459310221,
+      "learning_rate": 6.789041293542058e-06,
+      "loss": 0.516,
+      "step": 3102
+    },
+    {
+      "epoch": 1.5396549584212487,
+      "grad_norm": 0.07342375419107655,
+      "learning_rate": 6.787215868323349e-06,
+      "loss": 0.4985,
+      "step": 3103
+    },
+    {
+      "epoch": 1.540151421124488,
+      "grad_norm": 0.07336496676278476,
+      "learning_rate": 6.785390169957354e-06,
+      "loss": 0.4882,
+      "step": 3104
+    },
+    {
+      "epoch": 1.5406478838277273,
+      "grad_norm": 0.0733944918690652,
+      "learning_rate": 6.783564198723094e-06,
+      "loss": 0.4793,
+      "step": 3105
+    },
+    {
+      "epoch": 1.5411443465309669,
+      "grad_norm": 0.07278949180379689,
+      "learning_rate": 6.781737954899644e-06,
+      "loss": 0.5187,
+      "step": 3106
+    },
+    {
+      "epoch": 1.5416408092342064,
+      "grad_norm": 0.06996250280626114,
+      "learning_rate": 6.779911438766117e-06,
+      "loss": 0.4756,
+      "step": 3107
+    },
+    {
+      "epoch": 1.5421372719374458,
+      "grad_norm": 0.07095598603026688,
+      "learning_rate": 6.778084650601664e-06,
+      "loss": 0.508,
+      "step": 3108
+    },
+    {
+      "epoch": 1.542633734640685,
+      "grad_norm": 0.07131546994283615,
+      "learning_rate": 6.776257590685485e-06,
+      "loss": 0.4602,
+      "step": 3109
+    },
+    {
+      "epoch": 1.5431301973439244,
+      "grad_norm": 0.07280569367230365,
+      "learning_rate": 6.77443025929681e-06,
+      "loss": 0.5203,
+      "step": 3110
+    },
+    {
+      "epoch": 1.543626660047164,
+      "grad_norm": 0.0707173933711939,
+      "learning_rate": 6.772602656714922e-06,
+      "loss": 0.4857,
+      "step": 3111
+    },
+    {
+      "epoch": 1.5441231227504035,
+      "grad_norm": 0.07197803101114433,
+      "learning_rate": 6.770774783219139e-06,
+      "loss": 0.4693,
+      "step": 3112
+    },
+    {
+      "epoch": 1.5446195854536429,
+      "grad_norm": 0.07114587955675691,
+      "learning_rate": 6.7689466390888216e-06,
+      "loss": 0.4832,
+      "step": 3113
+    },
+    {
+      "epoch": 1.5451160481568822,
+      "grad_norm": 0.07396760660848735,
+      "learning_rate": 6.767118224603374e-06,
+      "loss": 0.4828,
+      "step": 3114
+    },
+    {
+      "epoch": 1.5456125108601215,
+      "grad_norm": 0.07517982329949065,
+      "learning_rate": 6.76528954004224e-06,
+      "loss": 0.4974,
+      "step": 3115
+    },
+    {
+      "epoch": 1.546108973563361,
+      "grad_norm": 0.07043798379945213,
+      "learning_rate": 6.763460585684903e-06,
+      "loss": 0.4972,
+      "step": 3116
+    },
+    {
+      "epoch": 1.5466054362666006,
+      "grad_norm": 0.07254963724358023,
+      "learning_rate": 6.761631361810892e-06,
+      "loss": 0.4786,
+      "step": 3117
+    },
+    {
+      "epoch": 1.54710189896984,
+      "grad_norm": 0.07611923909342257,
+      "learning_rate": 6.7598018686997725e-06,
+      "loss": 0.537,
+      "step": 3118
+    },
+    {
+      "epoch": 1.5475983616730793,
+      "grad_norm": 0.07540310011286495,
+      "learning_rate": 6.757972106631153e-06,
+      "loss": 0.4943,
+      "step": 3119
+    },
+    {
+      "epoch": 1.5480948243763186,
+      "grad_norm": 0.07568744832809565,
+      "learning_rate": 6.756142075884685e-06,
+      "loss": 0.4948,
+      "step": 3120
+    },
+    {
+      "epoch": 1.5485912870795582,
+      "grad_norm": 0.07178159273392977,
+      "learning_rate": 6.754311776740057e-06,
+      "loss": 0.493,
+      "step": 3121
+    },
+    {
+      "epoch": 1.5490877497827977,
+      "grad_norm": 0.07210617256147947,
+      "learning_rate": 6.7524812094770024e-06,
+      "loss": 0.4569,
+      "step": 3122
+    },
+    {
+      "epoch": 1.549584212486037,
+      "grad_norm": 0.07661066086921471,
+      "learning_rate": 6.750650374375293e-06,
+      "loss": 0.5068,
+      "step": 3123
+    },
+    {
+      "epoch": 1.5500806751892764,
+      "grad_norm": 0.07394053303684799,
+      "learning_rate": 6.748819271714745e-06,
+      "loss": 0.5012,
+      "step": 3124
+    },
+    {
+      "epoch": 1.5505771378925157,
+      "grad_norm": 0.07531830271554768,
+      "learning_rate": 6.746987901775211e-06,
+      "loss": 0.5163,
+      "step": 3125
+    },
+    {
+      "epoch": 1.5510736005957553,
+      "grad_norm": 0.07362018369668626,
+      "learning_rate": 6.745156264836589e-06,
+      "loss": 0.4705,
+      "step": 3126
+    },
+    {
+      "epoch": 1.5515700632989948,
+      "grad_norm": 0.07498203965559165,
+      "learning_rate": 6.743324361178811e-06,
+      "loss": 0.5239,
+      "step": 3127
+    },
+    {
+      "epoch": 1.5520665260022342,
+      "grad_norm": 0.07425384522239706,
+      "learning_rate": 6.741492191081856e-06,
+      "loss": 0.4918,
+      "step": 3128
+    },
+    {
+      "epoch": 1.5525629887054735,
+      "grad_norm": 0.07268237609705687,
+      "learning_rate": 6.739659754825742e-06,
+      "loss": 0.5026,
+      "step": 3129
+    },
+    {
+      "epoch": 1.5530594514087128,
+      "grad_norm": 0.0743248504811937,
+      "learning_rate": 6.73782705269053e-06,
+      "loss": 0.4999,
+      "step": 3130
+    },
+    {
+      "epoch": 1.5535559141119524,
+      "grad_norm": 0.07084637700551116,
+      "learning_rate": 6.735994084956317e-06,
+      "loss": 0.4823,
+      "step": 3131
+    },
+    {
+      "epoch": 1.5540523768151917,
+      "grad_norm": 0.07331080609658902,
+      "learning_rate": 6.734160851903241e-06,
+      "loss": 0.4629,
+      "step": 3132
+    },
+    {
+      "epoch": 1.5545488395184313,
+      "grad_norm": 0.07280582900730688,
+      "learning_rate": 6.732327353811484e-06,
+      "loss": 0.4908,
+      "step": 3133
+    },
+    {
+      "epoch": 1.5550453022216706,
+      "grad_norm": 0.07356499549409012,
+      "learning_rate": 6.730493590961268e-06,
+      "loss": 0.5275,
+      "step": 3134
+    },
+    {
+      "epoch": 1.55554176492491,
+      "grad_norm": 0.06992613113899178,
+      "learning_rate": 6.728659563632853e-06,
+      "loss": 0.4796,
+      "step": 3135
+    },
+    {
+      "epoch": 1.5560382276281495,
+      "grad_norm": 0.07211726923050378,
+      "learning_rate": 6.726825272106539e-06,
+      "loss": 0.5181,
+      "step": 3136
+    },
+    {
+      "epoch": 1.5565346903313888,
+      "grad_norm": 0.07333092531161171,
+      "learning_rate": 6.724990716662672e-06,
+      "loss": 0.4988,
+      "step": 3137
+    },
+    {
+      "epoch": 1.5570311530346284,
+      "grad_norm": 0.07593080180263549,
+      "learning_rate": 6.7231558975816294e-06,
+      "loss": 0.5345,
+      "step": 3138
+    },
+    {
+      "epoch": 1.5575276157378677,
+      "grad_norm": 0.06944979950702822,
+      "learning_rate": 6.721320815143837e-06,
+      "loss": 0.4673,
+      "step": 3139
+    },
+    {
+      "epoch": 1.558024078441107,
+      "grad_norm": 0.07427402559622737,
+      "learning_rate": 6.719485469629758e-06,
+      "loss": 0.4742,
+      "step": 3140
+    },
+    {
+      "epoch": 1.5585205411443466,
+      "grad_norm": 0.07355680254283553,
+      "learning_rate": 6.717649861319896e-06,
+      "loss": 0.4671,
+      "step": 3141
+    },
+    {
+      "epoch": 1.559017003847586,
+      "grad_norm": 0.07307892379944664,
+      "learning_rate": 6.715813990494793e-06,
+      "loss": 0.4799,
+      "step": 3142
+    },
+    {
+      "epoch": 1.5595134665508255,
+      "grad_norm": 0.06942389465400658,
+      "learning_rate": 6.713977857435031e-06,
+      "loss": 0.4788,
+      "step": 3143
+    },
+    {
+      "epoch": 1.5600099292540648,
+      "grad_norm": 0.07329210200157485,
+      "learning_rate": 6.712141462421236e-06,
+      "loss": 0.4861,
+      "step": 3144
+    },
+    {
+      "epoch": 1.5605063919573041,
+      "grad_norm": 0.07106259480596694,
+      "learning_rate": 6.7103048057340696e-06,
+      "loss": 0.5108,
+      "step": 3145
+    },
+    {
+      "epoch": 1.5610028546605437,
+      "grad_norm": 0.07588428796875119,
+      "learning_rate": 6.708467887654237e-06,
+      "loss": 0.5474,
+      "step": 3146
+    },
+    {
+      "epoch": 1.561499317363783,
+      "grad_norm": 0.07137860313281386,
+      "learning_rate": 6.70663070846248e-06,
+      "loss": 0.4952,
+      "step": 3147
+    },
+    {
+      "epoch": 1.5619957800670226,
+      "grad_norm": 0.06988977725921999,
+      "learning_rate": 6.704793268439584e-06,
+      "loss": 0.4708,
+      "step": 3148
+    },
+    {
+      "epoch": 1.5624922427702619,
+      "grad_norm": 0.0709502308910497,
+      "learning_rate": 6.702955567866372e-06,
+      "loss": 0.5429,
+      "step": 3149
+    },
+    {
+      "epoch": 1.5629887054735012,
+      "grad_norm": 0.07302469676059627,
+      "learning_rate": 6.7011176070237035e-06,
+      "loss": 0.4873,
+      "step": 3150
+    },
+    {
+      "epoch": 1.5634851681767408,
+      "grad_norm": 0.07060128527275819,
+      "learning_rate": 6.699279386192487e-06,
+      "loss": 0.4626,
+      "step": 3151
+    },
+    {
+      "epoch": 1.56398163087998,
+      "grad_norm": 0.07539679176181197,
+      "learning_rate": 6.6974409056536605e-06,
+      "loss": 0.4675,
+      "step": 3152
+    },
+    {
+      "epoch": 1.5644780935832197,
+      "grad_norm": 0.0745636123342702,
+      "learning_rate": 6.69560216568821e-06,
+      "loss": 0.481,
+      "step": 3153
+    },
+    {
+      "epoch": 1.564974556286459,
+      "grad_norm": 0.06906599990319741,
+      "learning_rate": 6.6937631665771545e-06,
+      "loss": 0.4497,
+      "step": 3154
+    },
+    {
+      "epoch": 1.5654710189896983,
+      "grad_norm": 0.07462061344881958,
+      "learning_rate": 6.6919239086015545e-06,
+      "loss": 0.4897,
+      "step": 3155
+    },
+    {
+      "epoch": 1.5659674816929379,
+      "grad_norm": 0.07077635712710982,
+      "learning_rate": 6.690084392042514e-06,
+      "loss": 0.5235,
+      "step": 3156
+    },
+    {
+      "epoch": 1.5664639443961772,
+      "grad_norm": 0.06944489107858984,
+      "learning_rate": 6.6882446171811746e-06,
+      "loss": 0.4841,
+      "step": 3157
+    },
+    {
+      "epoch": 1.5669604070994168,
+      "grad_norm": 0.07012427338846192,
+      "learning_rate": 6.686404584298711e-06,
+      "loss": 0.4871,
+      "step": 3158
+    },
+    {
+      "epoch": 1.567456869802656,
+      "grad_norm": 0.07176612922444559,
+      "learning_rate": 6.68456429367635e-06,
+      "loss": 0.4732,
+      "step": 3159
+    },
+    {
+      "epoch": 1.5679533325058954,
+      "grad_norm": 0.07298056998831044,
+      "learning_rate": 6.6827237455953435e-06,
+      "loss": 0.5041,
+      "step": 3160
+    },
+    {
+      "epoch": 1.568449795209135,
+      "grad_norm": 0.06987812192112296,
+      "learning_rate": 6.680882940336993e-06,
+      "loss": 0.4722,
+      "step": 3161
+    },
+    {
+      "epoch": 1.5689462579123743,
+      "grad_norm": 0.0760172434224406,
+      "learning_rate": 6.679041878182637e-06,
+      "loss": 0.5077,
+      "step": 3162
+    },
+    {
+      "epoch": 1.5694427206156139,
+      "grad_norm": 0.06996288342494414,
+      "learning_rate": 6.677200559413652e-06,
+      "loss": 0.4952,
+      "step": 3163
+    },
+    {
+      "epoch": 1.5699391833188532,
+      "grad_norm": 0.06879675843507983,
+      "learning_rate": 6.675358984311453e-06,
+      "loss": 0.4839,
+      "step": 3164
+    },
+    {
+      "epoch": 1.5704356460220925,
+      "grad_norm": 0.0720529922914632,
+      "learning_rate": 6.673517153157495e-06,
+      "loss": 0.4694,
+      "step": 3165
+    },
+    {
+      "epoch": 1.5709321087253318,
+      "grad_norm": 0.0728027783553152,
+      "learning_rate": 6.671675066233273e-06,
+      "loss": 0.511,
+      "step": 3166
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.07316311296453923,
+      "learning_rate": 6.669832723820321e-06,
+      "loss": 0.4949,
+      "step": 3167
+    },
+    {
+      "epoch": 1.571925034131811,
+      "grad_norm": 0.0780852417204941,
+      "learning_rate": 6.66799012620021e-06,
+      "loss": 0.4751,
+      "step": 3168
+    },
+    {
+      "epoch": 1.5724214968350503,
+      "grad_norm": 0.07366023643229092,
+      "learning_rate": 6.666147273654554e-06,
+      "loss": 0.4773,
+      "step": 3169
+    },
+    {
+      "epoch": 1.5729179595382896,
+      "grad_norm": 0.07005105873310478,
+      "learning_rate": 6.664304166465e-06,
+      "loss": 0.4912,
+      "step": 3170
+    },
+    {
+      "epoch": 1.573414422241529,
+      "grad_norm": 0.07330296651903939,
+      "learning_rate": 6.66246080491324e-06,
+      "loss": 0.4769,
+      "step": 3171
+    },
+    {
+      "epoch": 1.5739108849447685,
+      "grad_norm": 0.07135761963420995,
+      "learning_rate": 6.660617189281001e-06,
+      "loss": 0.506,
+      "step": 3172
+    },
+    {
+      "epoch": 1.574407347648008,
+      "grad_norm": 0.07246079314389194,
+      "learning_rate": 6.65877331985005e-06,
+      "loss": 0.5013,
+      "step": 3173
+    },
+    {
+      "epoch": 1.5749038103512474,
+      "grad_norm": 0.07514102098587554,
+      "learning_rate": 6.656929196902195e-06,
+      "loss": 0.4871,
+      "step": 3174
+    },
+    {
+      "epoch": 1.5754002730544867,
+      "grad_norm": 0.07291536990626205,
+      "learning_rate": 6.65508482071928e-06,
+      "loss": 0.4911,
+      "step": 3175
+    },
+    {
+      "epoch": 1.575896735757726,
+      "grad_norm": 0.07250201842234996,
+      "learning_rate": 6.6532401915831855e-06,
+      "loss": 0.4773,
+      "step": 3176
+    },
+    {
+      "epoch": 1.5763931984609656,
+      "grad_norm": 0.07149374200149058,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 0.4966,
+      "step": 3177
+    },
+    {
+      "epoch": 1.5768896611642051,
+      "grad_norm": 0.07309712040914346,
+      "learning_rate": 6.649550175579191e-06,
+      "loss": 0.4702,
+      "step": 3178
+    },
+    {
+      "epoch": 1.5773861238674445,
+      "grad_norm": 0.07264584874883896,
+      "learning_rate": 6.647704789275251e-06,
+      "loss": 0.4663,
+      "step": 3179
+    },
+    {
+      "epoch": 1.5778825865706838,
+      "grad_norm": 0.07569991453689745,
+      "learning_rate": 6.645859151146052e-06,
+      "loss": 0.4978,
+      "step": 3180
+    },
+    {
+      "epoch": 1.5783790492739231,
+      "grad_norm": 0.0703518000533615,
+      "learning_rate": 6.644013261473672e-06,
+      "loss": 0.4821,
+      "step": 3181
+    },
+    {
+      "epoch": 1.5788755119771627,
+      "grad_norm": 0.07370794045437885,
+      "learning_rate": 6.642167120540224e-06,
+      "loss": 0.5049,
+      "step": 3182
+    },
+    {
+      "epoch": 1.5793719746804022,
+      "grad_norm": 0.07207862109622967,
+      "learning_rate": 6.640320728627863e-06,
+      "loss": 0.5071,
+      "step": 3183
+    },
+    {
+      "epoch": 1.5798684373836416,
+      "grad_norm": 0.07270111894209261,
+      "learning_rate": 6.638474086018778e-06,
+      "loss": 0.4872,
+      "step": 3184
+    },
+    {
+      "epoch": 1.580364900086881,
+      "grad_norm": 0.07115484824977912,
+      "learning_rate": 6.636627192995201e-06,
+      "loss": 0.4982,
+      "step": 3185
+    },
+    {
+      "epoch": 1.5808613627901202,
+      "grad_norm": 0.07401912439597283,
+      "learning_rate": 6.634780049839397e-06,
+      "loss": 0.5041,
+      "step": 3186
+    },
+    {
+      "epoch": 1.5813578254933598,
+      "grad_norm": 0.07265465103238414,
+      "learning_rate": 6.6329326568336764e-06,
+      "loss": 0.4624,
+      "step": 3187
+    },
+    {
+      "epoch": 1.5818542881965993,
+      "grad_norm": 0.07348558258067905,
+      "learning_rate": 6.631085014260379e-06,
+      "loss": 0.4777,
+      "step": 3188
+    },
+    {
+      "epoch": 1.5823507508998387,
+      "grad_norm": 0.06853816265650765,
+      "learning_rate": 6.629237122401891e-06,
+      "loss": 0.4736,
+      "step": 3189
+    },
+    {
+      "epoch": 1.582847213603078,
+      "grad_norm": 0.07117687394840508,
+      "learning_rate": 6.627388981540632e-06,
+      "loss": 0.4864,
+      "step": 3190
+    },
+    {
+      "epoch": 1.5833436763063173,
+      "grad_norm": 0.07322973094031512,
+      "learning_rate": 6.62554059195906e-06,
+      "loss": 0.4937,
+      "step": 3191
+    },
+    {
+      "epoch": 1.583840139009557,
+      "grad_norm": 0.07230162933491331,
+      "learning_rate": 6.623691953939672e-06,
+      "loss": 0.503,
+      "step": 3192
+    },
+    {
+      "epoch": 1.5843366017127964,
+      "grad_norm": 0.08004626806270065,
+      "learning_rate": 6.6218430677650015e-06,
+      "loss": 0.5251,
+      "step": 3193
+    },
+    {
+      "epoch": 1.5848330644160358,
+      "grad_norm": 0.07451252608556116,
+      "learning_rate": 6.619993933717623e-06,
+      "loss": 0.5029,
+      "step": 3194
+    },
+    {
+      "epoch": 1.585329527119275,
+      "grad_norm": 0.07152721415325579,
+      "learning_rate": 6.618144552080148e-06,
+      "loss": 0.5249,
+      "step": 3195
+    },
+    {
+      "epoch": 1.5858259898225144,
+      "grad_norm": 0.06830156338501069,
+      "learning_rate": 6.616294923135221e-06,
+      "loss": 0.4986,
+      "step": 3196
+    },
+    {
+      "epoch": 1.586322452525754,
+      "grad_norm": 0.07340013043109206,
+      "learning_rate": 6.614445047165532e-06,
+      "loss": 0.4847,
+      "step": 3197
+    },
+    {
+      "epoch": 1.5868189152289935,
+      "grad_norm": 0.07484398073094696,
+      "learning_rate": 6.612594924453801e-06,
+      "loss": 0.5109,
+      "step": 3198
+    },
+    {
+      "epoch": 1.5873153779322329,
+      "grad_norm": 0.07501950325032929,
+      "learning_rate": 6.6107445552827955e-06,
+      "loss": 0.4683,
+      "step": 3199
+    },
+    {
+      "epoch": 1.5878118406354722,
+      "grad_norm": 0.07061001694221351,
+      "learning_rate": 6.608893939935308e-06,
+      "loss": 0.4614,
+      "step": 3200
+    },
+    {
+      "epoch": 1.5883083033387115,
+      "grad_norm": 0.07201468221091423,
+      "learning_rate": 6.607043078694179e-06,
+      "loss": 0.4681,
+      "step": 3201
+    },
+    {
+      "epoch": 1.588804766041951,
+      "grad_norm": 0.07102035784420759,
+      "learning_rate": 6.605191971842283e-06,
+      "loss": 0.4838,
+      "step": 3202
+    },
+    {
+      "epoch": 1.5893012287451906,
+      "grad_norm": 0.07356265885096674,
+      "learning_rate": 6.6033406196625306e-06,
+      "loss": 0.5241,
+      "step": 3203
+    },
+    {
+      "epoch": 1.58979769144843,
+      "grad_norm": 0.07013443767166846,
+      "learning_rate": 6.6014890224378724e-06,
+      "loss": 0.4752,
+      "step": 3204
+    },
+    {
+      "epoch": 1.5902941541516693,
+      "grad_norm": 0.07009450309863739,
+      "learning_rate": 6.599637180451295e-06,
+      "loss": 0.4941,
+      "step": 3205
+    },
+    {
+      "epoch": 1.5907906168549086,
+      "grad_norm": 0.07345078430314567,
+      "learning_rate": 6.597785093985824e-06,
+      "loss": 0.5187,
+      "step": 3206
+    },
+    {
+      "epoch": 1.5912870795581482,
+      "grad_norm": 0.07212784017604812,
+      "learning_rate": 6.59593276332452e-06,
+      "loss": 0.4917,
+      "step": 3207
+    },
+    {
+      "epoch": 1.5917835422613877,
+      "grad_norm": 0.07108259073549884,
+      "learning_rate": 6.59408018875048e-06,
+      "loss": 0.4793,
+      "step": 3208
+    },
+    {
+      "epoch": 1.592280004964627,
+      "grad_norm": 0.0715048074956627,
+      "learning_rate": 6.592227370546843e-06,
+      "loss": 0.5008,
+      "step": 3209
+    },
+    {
+      "epoch": 1.5927764676678664,
+      "grad_norm": 0.07546006913679203,
+      "learning_rate": 6.590374308996782e-06,
+      "loss": 0.4917,
+      "step": 3210
+    },
+    {
+      "epoch": 1.5932729303711057,
+      "grad_norm": 0.0716530556098064,
+      "learning_rate": 6.588521004383507e-06,
+      "loss": 0.4968,
+      "step": 3211
+    },
+    {
+      "epoch": 1.5937693930743453,
+      "grad_norm": 0.07083408847887936,
+      "learning_rate": 6.5866674569902676e-06,
+      "loss": 0.5028,
+      "step": 3212
+    },
+    {
+      "epoch": 1.5942658557775848,
+      "grad_norm": 0.07115108608146581,
+      "learning_rate": 6.584813667100347e-06,
+      "loss": 0.5092,
+      "step": 3213
+    },
+    {
+      "epoch": 1.5947623184808242,
+      "grad_norm": 0.07038895988678812,
+      "learning_rate": 6.582959634997068e-06,
+      "loss": 0.4479,
+      "step": 3214
+    },
+    {
+      "epoch": 1.5952587811840635,
+      "grad_norm": 0.06975115394749672,
+      "learning_rate": 6.581105360963791e-06,
+      "loss": 0.4841,
+      "step": 3215
+    },
+    {
+      "epoch": 1.5957552438873028,
+      "grad_norm": 0.07197478829125538,
+      "learning_rate": 6.579250845283909e-06,
+      "loss": 0.4971,
+      "step": 3216
+    },
+    {
+      "epoch": 1.5962517065905424,
+      "grad_norm": 0.0711203164305087,
+      "learning_rate": 6.577396088240857e-06,
+      "loss": 0.4959,
+      "step": 3217
+    },
+    {
+      "epoch": 1.596748169293782,
+      "grad_norm": 0.07016673110860351,
+      "learning_rate": 6.575541090118105e-06,
+      "loss": 0.4606,
+      "step": 3218
+    },
+    {
+      "epoch": 1.5972446319970213,
+      "grad_norm": 0.0717158281669078,
+      "learning_rate": 6.5736858511991585e-06,
+      "loss": 0.5236,
+      "step": 3219
+    },
+    {
+      "epoch": 1.5977410947002606,
+      "grad_norm": 0.07207805937373091,
+      "learning_rate": 6.5718303717675614e-06,
+      "loss": 0.4935,
+      "step": 3220
+    },
+    {
+      "epoch": 1.5982375574035,
+      "grad_norm": 0.07306215422798554,
+      "learning_rate": 6.5699746521068944e-06,
+      "loss": 0.4995,
+      "step": 3221
+    },
+    {
+      "epoch": 1.5987340201067395,
+      "grad_norm": 0.07178842625175158,
+      "learning_rate": 6.5681186925007755e-06,
+      "loss": 0.4961,
+      "step": 3222
+    },
+    {
+      "epoch": 1.599230482809979,
+      "grad_norm": 0.07054850163936612,
+      "learning_rate": 6.566262493232858e-06,
+      "loss": 0.47,
+      "step": 3223
+    },
+    {
+      "epoch": 1.5997269455132184,
+      "grad_norm": 0.07202065099054866,
+      "learning_rate": 6.564406054586831e-06,
+      "loss": 0.512,
+      "step": 3224
+    },
+    {
+      "epoch": 1.6002234082164577,
+      "grad_norm": 0.07115337760806216,
+      "learning_rate": 6.562549376846421e-06,
+      "loss": 0.4661,
+      "step": 3225
+    },
+    {
+      "epoch": 1.600719870919697,
+      "grad_norm": 0.07217097163771145,
+      "learning_rate": 6.5606924602953925e-06,
+      "loss": 0.4872,
+      "step": 3226
+    },
+    {
+      "epoch": 1.6012163336229366,
+      "grad_norm": 0.07184519047556386,
+      "learning_rate": 6.558835305217545e-06,
+      "loss": 0.4918,
+      "step": 3227
+    },
+    {
+      "epoch": 1.6017127963261761,
+      "grad_norm": 0.068517155266461,
+      "learning_rate": 6.556977911896716e-06,
+      "loss": 0.4715,
+      "step": 3228
+    },
+    {
+      "epoch": 1.6022092590294155,
+      "grad_norm": 0.06873824143142059,
+      "learning_rate": 6.555120280616777e-06,
+      "loss": 0.4702,
+      "step": 3229
+    },
+    {
+      "epoch": 1.6027057217326548,
+      "grad_norm": 0.07054371263644203,
+      "learning_rate": 6.553262411661636e-06,
+      "loss": 0.476,
+      "step": 3230
+    },
+    {
+      "epoch": 1.6032021844358941,
+      "grad_norm": 0.07448870421576329,
+      "learning_rate": 6.551404305315243e-06,
+      "loss": 0.4909,
+      "step": 3231
+    },
+    {
+      "epoch": 1.6036986471391337,
+      "grad_norm": 0.07600252195798972,
+      "learning_rate": 6.5495459618615756e-06,
+      "loss": 0.5184,
+      "step": 3232
+    },
+    {
+      "epoch": 1.6041951098423732,
+      "grad_norm": 0.0748000187133278,
+      "learning_rate": 6.547687381584653e-06,
+      "loss": 0.4955,
+      "step": 3233
+    },
+    {
+      "epoch": 1.6046915725456126,
+      "grad_norm": 0.07127300203261505,
+      "learning_rate": 6.545828564768529e-06,
+      "loss": 0.4823,
+      "step": 3234
+    },
+    {
+      "epoch": 1.605188035248852,
+      "grad_norm": 0.07232653631996022,
+      "learning_rate": 6.543969511697295e-06,
+      "loss": 0.5055,
+      "step": 3235
+    },
+    {
+      "epoch": 1.6056844979520912,
+      "grad_norm": 0.07128404071201468,
+      "learning_rate": 6.542110222655076e-06,
+      "loss": 0.4794,
+      "step": 3236
+    },
+    {
+      "epoch": 1.6061809606553308,
+      "grad_norm": 0.06928059050306892,
+      "learning_rate": 6.540250697926035e-06,
+      "loss": 0.4699,
+      "step": 3237
+    },
+    {
+      "epoch": 1.6066774233585703,
+      "grad_norm": 0.06907166245498234,
+      "learning_rate": 6.5383909377943725e-06,
+      "loss": 0.4755,
+      "step": 3238
+    },
+    {
+      "epoch": 1.6071738860618097,
+      "grad_norm": 0.07215048758101493,
+      "learning_rate": 6.536530942544321e-06,
+      "loss": 0.5169,
+      "step": 3239
+    },
+    {
+      "epoch": 1.607670348765049,
+      "grad_norm": 0.07093218907508307,
+      "learning_rate": 6.534670712460151e-06,
+      "loss": 0.4936,
+      "step": 3240
+    },
+    {
+      "epoch": 1.6081668114682883,
+      "grad_norm": 0.07423867373297897,
+      "learning_rate": 6.532810247826168e-06,
+      "loss": 0.5185,
+      "step": 3241
+    },
+    {
+      "epoch": 1.6086632741715279,
+      "grad_norm": 0.07419926462973128,
+      "learning_rate": 6.530949548926716e-06,
+      "loss": 0.4977,
+      "step": 3242
+    },
+    {
+      "epoch": 1.6091597368747674,
+      "grad_norm": 0.07426546852534963,
+      "learning_rate": 6.529088616046172e-06,
+      "loss": 0.4703,
+      "step": 3243
+    },
+    {
+      "epoch": 1.6096561995780068,
+      "grad_norm": 0.07032622346168317,
+      "learning_rate": 6.527227449468951e-06,
+      "loss": 0.4602,
+      "step": 3244
+    },
+    {
+      "epoch": 1.610152662281246,
+      "grad_norm": 0.0739855696239581,
+      "learning_rate": 6.525366049479501e-06,
+      "loss": 0.4915,
+      "step": 3245
+    },
+    {
+      "epoch": 1.6106491249844854,
+      "grad_norm": 0.07666709761082861,
+      "learning_rate": 6.523504416362308e-06,
+      "loss": 0.4845,
+      "step": 3246
+    },
+    {
+      "epoch": 1.611145587687725,
+      "grad_norm": 0.07131154560337846,
+      "learning_rate": 6.521642550401894e-06,
+      "loss": 0.4763,
+      "step": 3247
+    },
+    {
+      "epoch": 1.6116420503909645,
+      "grad_norm": 0.07160752523694873,
+      "learning_rate": 6.519780451882811e-06,
+      "loss": 0.4801,
+      "step": 3248
+    },
+    {
+      "epoch": 1.6121385130942039,
+      "grad_norm": 0.07150793823658054,
+      "learning_rate": 6.517918121089656e-06,
+      "loss": 0.467,
+      "step": 3249
+    },
+    {
+      "epoch": 1.6126349757974432,
+      "grad_norm": 0.0729422427737503,
+      "learning_rate": 6.516055558307054e-06,
+      "loss": 0.5076,
+      "step": 3250
+    },
+    {
+      "epoch": 1.6131314385006825,
+      "grad_norm": 0.0727548197435697,
+      "learning_rate": 6.5141927638196665e-06,
+      "loss": 0.5082,
+      "step": 3251
+    },
+    {
+      "epoch": 1.613627901203922,
+      "grad_norm": 0.07179615875120154,
+      "learning_rate": 6.5123297379121944e-06,
+      "loss": 0.4962,
+      "step": 3252
+    },
+    {
+      "epoch": 1.6141243639071616,
+      "grad_norm": 0.07058848713675607,
+      "learning_rate": 6.510466480869369e-06,
+      "loss": 0.5126,
+      "step": 3253
+    },
+    {
+      "epoch": 1.614620826610401,
+      "grad_norm": 0.07140445583852305,
+      "learning_rate": 6.508602992975963e-06,
+      "loss": 0.4587,
+      "step": 3254
+    },
+    {
+      "epoch": 1.6151172893136403,
+      "grad_norm": 0.07630979470934252,
+      "learning_rate": 6.506739274516777e-06,
+      "loss": 0.5185,
+      "step": 3255
+    },
+    {
+      "epoch": 1.6156137520168796,
+      "grad_norm": 0.07288163814992277,
+      "learning_rate": 6.504875325776651e-06,
+      "loss": 0.5012,
+      "step": 3256
+    },
+    {
+      "epoch": 1.6161102147201192,
+      "grad_norm": 0.06951093014179835,
+      "learning_rate": 6.50301114704046e-06,
+      "loss": 0.4635,
+      "step": 3257
+    },
+    {
+      "epoch": 1.6166066774233587,
+      "grad_norm": 0.07246702259048228,
+      "learning_rate": 6.501146738593114e-06,
+      "loss": 0.495,
+      "step": 3258
+    },
+    {
+      "epoch": 1.617103140126598,
+      "grad_norm": 0.0772067141521757,
+      "learning_rate": 6.499282100719558e-06,
+      "loss": 0.5023,
+      "step": 3259
+    },
+    {
+      "epoch": 1.6175996028298374,
+      "grad_norm": 0.07720777262425242,
+      "learning_rate": 6.497417233704774e-06,
+      "loss": 0.5386,
+      "step": 3260
+    },
+    {
+      "epoch": 1.6180960655330767,
+      "grad_norm": 0.07009815130142644,
+      "learning_rate": 6.495552137833774e-06,
+      "loss": 0.4827,
+      "step": 3261
+    },
+    {
+      "epoch": 1.6185925282363163,
+      "grad_norm": 0.07666255501778249,
+      "learning_rate": 6.493686813391608e-06,
+      "loss": 0.4698,
+      "step": 3262
+    },
+    {
+      "epoch": 1.6190889909395558,
+      "grad_norm": 0.07156342643723003,
+      "learning_rate": 6.491821260663364e-06,
+      "loss": 0.5408,
+      "step": 3263
+    },
+    {
+      "epoch": 1.6195854536427952,
+      "grad_norm": 0.0681821239446925,
+      "learning_rate": 6.4899554799341576e-06,
+      "loss": 0.476,
+      "step": 3264
+    },
+    {
+      "epoch": 1.6200819163460345,
+      "grad_norm": 0.07136397273408063,
+      "learning_rate": 6.488089471489147e-06,
+      "loss": 0.5122,
+      "step": 3265
+    },
+    {
+      "epoch": 1.6205783790492738,
+      "grad_norm": 0.07370498350984535,
+      "learning_rate": 6.486223235613522e-06,
+      "loss": 0.5162,
+      "step": 3266
+    },
+    {
+      "epoch": 1.6210748417525134,
+      "grad_norm": 0.07631744016069217,
+      "learning_rate": 6.4843567725925025e-06,
+      "loss": 0.5008,
+      "step": 3267
+    },
+    {
+      "epoch": 1.621571304455753,
+      "grad_norm": 0.06930606236752354,
+      "learning_rate": 6.4824900827113506e-06,
+      "loss": 0.4766,
+      "step": 3268
+    },
+    {
+      "epoch": 1.6220677671589923,
+      "grad_norm": 0.0751654947836453,
+      "learning_rate": 6.48062316625536e-06,
+      "loss": 0.5066,
+      "step": 3269
+    },
+    {
+      "epoch": 1.6225642298622316,
+      "grad_norm": 0.06943801964539678,
+      "learning_rate": 6.478756023509859e-06,
+      "loss": 0.476,
+      "step": 3270
+    },
+    {
+      "epoch": 1.623060692565471,
+      "grad_norm": 0.07004301982561809,
+      "learning_rate": 6.47688865476021e-06,
+      "loss": 0.4733,
+      "step": 3271
+    },
+    {
+      "epoch": 1.6235571552687105,
+      "grad_norm": 0.07182265834599942,
+      "learning_rate": 6.475021060291809e-06,
+      "loss": 0.5008,
+      "step": 3272
+    },
+    {
+      "epoch": 1.6240536179719498,
+      "grad_norm": 0.07533202201536007,
+      "learning_rate": 6.47315324039009e-06,
+      "loss": 0.4876,
+      "step": 3273
+    },
+    {
+      "epoch": 1.6245500806751894,
+      "grad_norm": 0.07085573161894548,
+      "learning_rate": 6.471285195340517e-06,
+      "loss": 0.4778,
+      "step": 3274
+    },
+    {
+      "epoch": 1.6250465433784287,
+      "grad_norm": 0.07293460795213716,
+      "learning_rate": 6.469416925428593e-06,
+      "loss": 0.4879,
+      "step": 3275
+    },
+    {
+      "epoch": 1.625543006081668,
+      "grad_norm": 0.06991347889183441,
+      "learning_rate": 6.467548430939854e-06,
+      "loss": 0.4428,
+      "step": 3276
+    },
+    {
+      "epoch": 1.6260394687849076,
+      "grad_norm": 0.07308541333754964,
+      "learning_rate": 6.4656797121598655e-06,
+      "loss": 0.4782,
+      "step": 3277
+    },
+    {
+      "epoch": 1.626535931488147,
+      "grad_norm": 0.07263307742040882,
+      "learning_rate": 6.463810769374234e-06,
+      "loss": 0.4984,
+      "step": 3278
+    },
+    {
+      "epoch": 1.6270323941913865,
+      "grad_norm": 0.06995720887521895,
+      "learning_rate": 6.461941602868597e-06,
+      "loss": 0.4469,
+      "step": 3279
+    },
+    {
+      "epoch": 1.6275288568946258,
+      "grad_norm": 0.07229481627052717,
+      "learning_rate": 6.4600722129286266e-06,
+      "loss": 0.4694,
+      "step": 3280
+    },
+    {
+      "epoch": 1.6280253195978651,
+      "grad_norm": 0.07009091230720142,
+      "learning_rate": 6.458202599840028e-06,
+      "loss": 0.4849,
+      "step": 3281
+    },
+    {
+      "epoch": 1.6285217823011047,
+      "grad_norm": 0.07308347394869531,
+      "learning_rate": 6.456332763888544e-06,
+      "loss": 0.5112,
+      "step": 3282
+    },
+    {
+      "epoch": 1.629018245004344,
+      "grad_norm": 0.06933218981123568,
+      "learning_rate": 6.454462705359946e-06,
+      "loss": 0.4848,
+      "step": 3283
+    },
+    {
+      "epoch": 1.6295147077075836,
+      "grad_norm": 0.07618867420111537,
+      "learning_rate": 6.452592424540045e-06,
+      "loss": 0.506,
+      "step": 3284
+    },
+    {
+      "epoch": 1.6300111704108229,
+      "grad_norm": 0.07143326454103777,
+      "learning_rate": 6.4507219217146825e-06,
+      "loss": 0.4859,
+      "step": 3285
+    },
+    {
+      "epoch": 1.6305076331140622,
+      "grad_norm": 0.07379665108370889,
+      "learning_rate": 6.448851197169733e-06,
+      "loss": 0.4789,
+      "step": 3286
+    },
+    {
+      "epoch": 1.6310040958173018,
+      "grad_norm": 0.07383433957267552,
+      "learning_rate": 6.446980251191111e-06,
+      "loss": 0.5042,
+      "step": 3287
+    },
+    {
+      "epoch": 1.631500558520541,
+      "grad_norm": 0.0695968961937613,
+      "learning_rate": 6.445109084064758e-06,
+      "loss": 0.4803,
+      "step": 3288
+    },
+    {
+      "epoch": 1.6319970212237807,
+      "grad_norm": 0.07369256598687489,
+      "learning_rate": 6.443237696076652e-06,
+      "loss": 0.4869,
+      "step": 3289
+    },
+    {
+      "epoch": 1.63249348392702,
+      "grad_norm": 0.07215440570531192,
+      "learning_rate": 6.441366087512804e-06,
+      "loss": 0.4975,
+      "step": 3290
+    },
+    {
+      "epoch": 1.6329899466302593,
+      "grad_norm": 0.07468599079538513,
+      "learning_rate": 6.439494258659259e-06,
+      "loss": 0.5109,
+      "step": 3291
+    },
+    {
+      "epoch": 1.6334864093334989,
+      "grad_norm": 0.06772469651576586,
+      "learning_rate": 6.437622209802099e-06,
+      "loss": 0.4995,
+      "step": 3292
+    },
+    {
+      "epoch": 1.6339828720367382,
+      "grad_norm": 0.07521859544196922,
+      "learning_rate": 6.435749941227434e-06,
+      "loss": 0.5283,
+      "step": 3293
+    },
+    {
+      "epoch": 1.6344793347399778,
+      "grad_norm": 0.0692286722927468,
+      "learning_rate": 6.43387745322141e-06,
+      "loss": 0.4664,
+      "step": 3294
+    },
+    {
+      "epoch": 1.634975797443217,
+      "grad_norm": 0.06903400899827464,
+      "learning_rate": 6.432004746070209e-06,
+      "loss": 0.4662,
+      "step": 3295
+    },
+    {
+      "epoch": 1.6354722601464564,
+      "grad_norm": 0.06963374521958657,
+      "learning_rate": 6.430131820060043e-06,
+      "loss": 0.4785,
+      "step": 3296
+    },
+    {
+      "epoch": 1.635968722849696,
+      "grad_norm": 0.0702109802058609,
+      "learning_rate": 6.428258675477158e-06,
+      "loss": 0.4896,
+      "step": 3297
+    },
+    {
+      "epoch": 1.6364651855529353,
+      "grad_norm": 0.07196211846484113,
+      "learning_rate": 6.426385312607837e-06,
+      "loss": 0.5137,
+      "step": 3298
+    },
+    {
+      "epoch": 1.6369616482561749,
+      "grad_norm": 0.07225873948893603,
+      "learning_rate": 6.424511731738389e-06,
+      "loss": 0.4952,
+      "step": 3299
+    },
+    {
+      "epoch": 1.6374581109594142,
+      "grad_norm": 0.0759441267344298,
+      "learning_rate": 6.4226379331551625e-06,
+      "loss": 0.4939,
+      "step": 3300
+    },
+    {
+      "epoch": 1.6379545736626535,
+      "grad_norm": 0.07168285148556738,
+      "learning_rate": 6.420763917144539e-06,
+      "loss": 0.4928,
+      "step": 3301
+    },
+    {
+      "epoch": 1.638451036365893,
+      "grad_norm": 0.07063050009803296,
+      "learning_rate": 6.4188896839929314e-06,
+      "loss": 0.4916,
+      "step": 3302
+    },
+    {
+      "epoch": 1.6389474990691324,
+      "grad_norm": 0.07020108879688203,
+      "learning_rate": 6.417015233986786e-06,
+      "loss": 0.4688,
+      "step": 3303
+    },
+    {
+      "epoch": 1.639443961772372,
+      "grad_norm": 0.07062473758265202,
+      "learning_rate": 6.415140567412583e-06,
+      "loss": 0.4642,
+      "step": 3304
+    },
+    {
+      "epoch": 1.6399404244756113,
+      "grad_norm": 0.07243674409944932,
+      "learning_rate": 6.413265684556833e-06,
+      "loss": 0.496,
+      "step": 3305
+    },
+    {
+      "epoch": 1.6404368871788506,
+      "grad_norm": 0.0748780185500087,
+      "learning_rate": 6.4113905857060835e-06,
+      "loss": 0.5193,
+      "step": 3306
+    },
+    {
+      "epoch": 1.64093334988209,
+      "grad_norm": 0.07086725699518998,
+      "learning_rate": 6.409515271146912e-06,
+      "loss": 0.4582,
+      "step": 3307
+    },
+    {
+      "epoch": 1.6414298125853295,
+      "grad_norm": 0.07281673976046933,
+      "learning_rate": 6.4076397411659316e-06,
+      "loss": 0.4832,
+      "step": 3308
+    },
+    {
+      "epoch": 1.641926275288569,
+      "grad_norm": 0.07024856454439375,
+      "learning_rate": 6.405763996049788e-06,
+      "loss": 0.4531,
+      "step": 3309
+    },
+    {
+      "epoch": 1.6424227379918084,
+      "grad_norm": 0.07035789172828415,
+      "learning_rate": 6.403888036085155e-06,
+      "loss": 0.4604,
+      "step": 3310
+    },
+    {
+      "epoch": 1.6429192006950477,
+      "grad_norm": 0.07489507707353497,
+      "learning_rate": 6.402011861558748e-06,
+      "loss": 0.5046,
+      "step": 3311
+    },
+    {
+      "epoch": 1.643415663398287,
+      "grad_norm": 0.07344936173620556,
+      "learning_rate": 6.400135472757305e-06,
+      "loss": 0.468,
+      "step": 3312
+    },
+    {
+      "epoch": 1.6439121261015266,
+      "grad_norm": 0.07071152024432985,
+      "learning_rate": 6.398258869967606e-06,
+      "loss": 0.4573,
+      "step": 3313
+    },
+    {
+      "epoch": 1.6444085888047661,
+      "grad_norm": 0.07228608382103503,
+      "learning_rate": 6.396382053476459e-06,
+      "loss": 0.519,
+      "step": 3314
+    },
+    {
+      "epoch": 1.6449050515080055,
+      "grad_norm": 0.07024217535197713,
+      "learning_rate": 6.394505023570702e-06,
+      "loss": 0.5054,
+      "step": 3315
+    },
+    {
+      "epoch": 1.6454015142112448,
+      "grad_norm": 0.07022134255927119,
+      "learning_rate": 6.392627780537212e-06,
+      "loss": 0.4658,
+      "step": 3316
+    },
+    {
+      "epoch": 1.6458979769144841,
+      "grad_norm": 0.0717953161572256,
+      "learning_rate": 6.390750324662895e-06,
+      "loss": 0.4899,
+      "step": 3317
+    },
+    {
+      "epoch": 1.6463944396177237,
+      "grad_norm": 0.07671132428182728,
+      "learning_rate": 6.388872656234689e-06,
+      "loss": 0.5285,
+      "step": 3318
+    },
+    {
+      "epoch": 1.6468909023209632,
+      "grad_norm": 0.07014994651509567,
+      "learning_rate": 6.386994775539569e-06,
+      "loss": 0.4845,
+      "step": 3319
+    },
+    {
+      "epoch": 1.6473873650242026,
+      "grad_norm": 0.0717433473722914,
+      "learning_rate": 6.3851166828645354e-06,
+      "loss": 0.5168,
+      "step": 3320
+    },
+    {
+      "epoch": 1.647883827727442,
+      "grad_norm": 0.07534612081049234,
+      "learning_rate": 6.383238378496624e-06,
+      "loss": 0.5067,
+      "step": 3321
+    },
+    {
+      "epoch": 1.6483802904306812,
+      "grad_norm": 0.07416068493187741,
+      "learning_rate": 6.381359862722905e-06,
+      "loss": 0.5052,
+      "step": 3322
+    },
+    {
+      "epoch": 1.6488767531339208,
+      "grad_norm": 0.0738384042184749,
+      "learning_rate": 6.379481135830481e-06,
+      "loss": 0.4627,
+      "step": 3323
+    },
+    {
+      "epoch": 1.6493732158371603,
+      "grad_norm": 0.07395306527127492,
+      "learning_rate": 6.3776021981064825e-06,
+      "loss": 0.48,
+      "step": 3324
+    },
+    {
+      "epoch": 1.6498696785403997,
+      "grad_norm": 0.06933352802336462,
+      "learning_rate": 6.375723049838077e-06,
+      "loss": 0.4645,
+      "step": 3325
+    },
+    {
+      "epoch": 1.650366141243639,
+      "grad_norm": 0.07507598422114344,
+      "learning_rate": 6.37384369131246e-06,
+      "loss": 0.4641,
+      "step": 3326
+    },
+    {
+      "epoch": 1.6508626039468783,
+      "grad_norm": 0.07092097177784223,
+      "learning_rate": 6.371964122816865e-06,
+      "loss": 0.4931,
+      "step": 3327
+    },
+    {
+      "epoch": 1.651359066650118,
+      "grad_norm": 0.0717024682861258,
+      "learning_rate": 6.3700843446385495e-06,
+      "loss": 0.4808,
+      "step": 3328
+    },
+    {
+      "epoch": 1.6518555293533574,
+      "grad_norm": 0.07215914605506774,
+      "learning_rate": 6.3682043570648115e-06,
+      "loss": 0.5071,
+      "step": 3329
+    },
+    {
+      "epoch": 1.6523519920565968,
+      "grad_norm": 0.07186036631806916,
+      "learning_rate": 6.366324160382974e-06,
+      "loss": 0.5117,
+      "step": 3330
+    },
+    {
+      "epoch": 1.652848454759836,
+      "grad_norm": 0.07690482622531732,
+      "learning_rate": 6.364443754880395e-06,
+      "loss": 0.4808,
+      "step": 3331
+    },
+    {
+      "epoch": 1.6533449174630754,
+      "grad_norm": 0.0730255508982856,
+      "learning_rate": 6.362563140844465e-06,
+      "loss": 0.4823,
+      "step": 3332
+    },
+    {
+      "epoch": 1.653841380166315,
+      "grad_norm": 0.07022155162274443,
+      "learning_rate": 6.360682318562607e-06,
+      "loss": 0.4958,
+      "step": 3333
+    },
+    {
+      "epoch": 1.6543378428695545,
+      "grad_norm": 0.0742115762662353,
+      "learning_rate": 6.358801288322274e-06,
+      "loss": 0.5135,
+      "step": 3334
+    },
+    {
+      "epoch": 1.6548343055727939,
+      "grad_norm": 0.0718887982711449,
+      "learning_rate": 6.3569200504109505e-06,
+      "loss": 0.5174,
+      "step": 3335
+    },
+    {
+      "epoch": 1.6553307682760332,
+      "grad_norm": 0.07166615114113524,
+      "learning_rate": 6.355038605116155e-06,
+      "loss": 0.4916,
+      "step": 3336
+    },
+    {
+      "epoch": 1.6558272309792725,
+      "grad_norm": 0.0691153247633263,
+      "learning_rate": 6.353156952725432e-06,
+      "loss": 0.476,
+      "step": 3337
+    },
+    {
+      "epoch": 1.656323693682512,
+      "grad_norm": 0.07031968237098923,
+      "learning_rate": 6.3512750935263664e-06,
+      "loss": 0.5051,
+      "step": 3338
+    },
+    {
+      "epoch": 1.6568201563857516,
+      "grad_norm": 0.0726129063754758,
+      "learning_rate": 6.349393027806569e-06,
+      "loss": 0.4881,
+      "step": 3339
+    },
+    {
+      "epoch": 1.657316619088991,
+      "grad_norm": 0.07360865260049655,
+      "learning_rate": 6.347510755853683e-06,
+      "loss": 0.5138,
+      "step": 3340
+    },
+    {
+      "epoch": 1.6578130817922303,
+      "grad_norm": 0.07271768270838276,
+      "learning_rate": 6.345628277955384e-06,
+      "loss": 0.4731,
+      "step": 3341
+    },
+    {
+      "epoch": 1.6583095444954696,
+      "grad_norm": 0.0714115408833875,
+      "learning_rate": 6.3437455943993785e-06,
+      "loss": 0.4505,
+      "step": 3342
+    },
+    {
+      "epoch": 1.6588060071987092,
+      "grad_norm": 0.0707543409992033,
+      "learning_rate": 6.341862705473405e-06,
+      "loss": 0.5006,
+      "step": 3343
+    },
+    {
+      "epoch": 1.6593024699019487,
+      "grad_norm": 0.07128065259021685,
+      "learning_rate": 6.339979611465231e-06,
+      "loss": 0.4637,
+      "step": 3344
+    },
+    {
+      "epoch": 1.659798932605188,
+      "grad_norm": 0.06932455519727876,
+      "learning_rate": 6.338096312662658e-06,
+      "loss": 0.4929,
+      "step": 3345
+    },
+    {
+      "epoch": 1.6602953953084274,
+      "grad_norm": 0.07738642167451999,
+      "learning_rate": 6.336212809353518e-06,
+      "loss": 0.5045,
+      "step": 3346
+    },
+    {
+      "epoch": 1.6607918580116667,
+      "grad_norm": 0.07115439790368483,
+      "learning_rate": 6.334329101825676e-06,
+      "loss": 0.5263,
+      "step": 3347
+    },
+    {
+      "epoch": 1.6612883207149063,
+      "grad_norm": 0.06896941474755712,
+      "learning_rate": 6.332445190367025e-06,
+      "loss": 0.4706,
+      "step": 3348
+    },
+    {
+      "epoch": 1.6617847834181458,
+      "grad_norm": 0.06890612750608131,
+      "learning_rate": 6.330561075265489e-06,
+      "loss": 0.4599,
+      "step": 3349
+    },
+    {
+      "epoch": 1.6622812461213852,
+      "grad_norm": 0.0713066211215429,
+      "learning_rate": 6.328676756809028e-06,
+      "loss": 0.5373,
+      "step": 3350
+    },
+    {
+      "epoch": 1.6627777088246245,
+      "grad_norm": 0.07394098379135165,
+      "learning_rate": 6.326792235285628e-06,
+      "loss": 0.494,
+      "step": 3351
+    },
+    {
+      "epoch": 1.6632741715278638,
+      "grad_norm": 0.06792132504176879,
+      "learning_rate": 6.32490751098331e-06,
+      "loss": 0.4471,
+      "step": 3352
+    },
+    {
+      "epoch": 1.6637706342311034,
+      "grad_norm": 0.06981735971850177,
+      "learning_rate": 6.323022584190121e-06,
+      "loss": 0.4812,
+      "step": 3353
+    },
+    {
+      "epoch": 1.664267096934343,
+      "grad_norm": 0.07156250252199886,
+      "learning_rate": 6.321137455194142e-06,
+      "loss": 0.4741,
+      "step": 3354
+    },
+    {
+      "epoch": 1.6647635596375823,
+      "grad_norm": 0.07252274014404607,
+      "learning_rate": 6.319252124283486e-06,
+      "loss": 0.4727,
+      "step": 3355
+    },
+    {
+      "epoch": 1.6652600223408216,
+      "grad_norm": 0.07105885580438999,
+      "learning_rate": 6.317366591746296e-06,
+      "loss": 0.513,
+      "step": 3356
+    },
+    {
+      "epoch": 1.665756485044061,
+      "grad_norm": 0.06892471295198868,
+      "learning_rate": 6.315480857870746e-06,
+      "loss": 0.4579,
+      "step": 3357
+    },
+    {
+      "epoch": 1.6662529477473005,
+      "grad_norm": 0.07137113284562825,
+      "learning_rate": 6.313594922945036e-06,
+      "loss": 0.4955,
+      "step": 3358
+    },
+    {
+      "epoch": 1.66674941045054,
+      "grad_norm": 0.07259766337382742,
+      "learning_rate": 6.311708787257408e-06,
+      "loss": 0.4861,
+      "step": 3359
+    },
+    {
+      "epoch": 1.6672458731537794,
+      "grad_norm": 0.06910578629300906,
+      "learning_rate": 6.30982245109612e-06,
+      "loss": 0.4665,
+      "step": 3360
+    },
+    {
+      "epoch": 1.6677423358570187,
+      "grad_norm": 0.06959429004837796,
+      "learning_rate": 6.307935914749473e-06,
+      "loss": 0.4966,
+      "step": 3361
+    },
+    {
+      "epoch": 1.668238798560258,
+      "grad_norm": 0.07166836668037606,
+      "learning_rate": 6.306049178505793e-06,
+      "loss": 0.4742,
+      "step": 3362
+    },
+    {
+      "epoch": 1.6687352612634976,
+      "grad_norm": 0.07386543999500932,
+      "learning_rate": 6.304162242653437e-06,
+      "loss": 0.4785,
+      "step": 3363
+    },
+    {
+      "epoch": 1.6692317239667371,
+      "grad_norm": 0.07432261885864039,
+      "learning_rate": 6.302275107480792e-06,
+      "loss": 0.4758,
+      "step": 3364
+    },
+    {
+      "epoch": 1.6697281866699765,
+      "grad_norm": 0.07336508318681913,
+      "learning_rate": 6.300387773276278e-06,
+      "loss": 0.4744,
+      "step": 3365
+    },
+    {
+      "epoch": 1.6702246493732158,
+      "grad_norm": 0.07083645305442617,
+      "learning_rate": 6.298500240328342e-06,
+      "loss": 0.4725,
+      "step": 3366
+    },
+    {
+      "epoch": 1.6707211120764551,
+      "grad_norm": 0.07163096680010762,
+      "learning_rate": 6.296612508925466e-06,
+      "loss": 0.4807,
+      "step": 3367
+    },
+    {
+      "epoch": 1.6712175747796947,
+      "grad_norm": 0.06990183990931721,
+      "learning_rate": 6.294724579356157e-06,
+      "loss": 0.4683,
+      "step": 3368
+    },
+    {
+      "epoch": 1.6717140374829342,
+      "grad_norm": 0.07405777347762747,
+      "learning_rate": 6.292836451908955e-06,
+      "loss": 0.5003,
+      "step": 3369
+    },
+    {
+      "epoch": 1.6722105001861736,
+      "grad_norm": 0.07315404944387507,
+      "learning_rate": 6.290948126872429e-06,
+      "loss": 0.5086,
+      "step": 3370
+    },
+    {
+      "epoch": 1.672706962889413,
+      "grad_norm": 0.06840865988701067,
+      "learning_rate": 6.289059604535182e-06,
+      "loss": 0.489,
+      "step": 3371
+    },
+    {
+      "epoch": 1.6732034255926522,
+      "grad_norm": 0.07184124769888703,
+      "learning_rate": 6.2871708851858414e-06,
+      "loss": 0.4989,
+      "step": 3372
+    },
+    {
+      "epoch": 1.6736998882958918,
+      "grad_norm": 0.07060758468264917,
+      "learning_rate": 6.285281969113072e-06,
+      "loss": 0.4608,
+      "step": 3373
+    },
+    {
+      "epoch": 1.6741963509991313,
+      "grad_norm": 0.0693710380692579,
+      "learning_rate": 6.283392856605559e-06,
+      "loss": 0.4521,
+      "step": 3374
+    },
+    {
+      "epoch": 1.6746928137023707,
+      "grad_norm": 0.0738048416478676,
+      "learning_rate": 6.281503547952027e-06,
+      "loss": 0.4936,
+      "step": 3375
+    },
+    {
+      "epoch": 1.67518927640561,
+      "grad_norm": 0.07735478008819657,
+      "learning_rate": 6.279614043441226e-06,
+      "loss": 0.5336,
+      "step": 3376
+    },
+    {
+      "epoch": 1.6756857391088493,
+      "grad_norm": 0.06963549926932651,
+      "learning_rate": 6.277724343361933e-06,
+      "loss": 0.4604,
+      "step": 3377
+    },
+    {
+      "epoch": 1.6761822018120889,
+      "grad_norm": 0.06947653830726935,
+      "learning_rate": 6.275834448002962e-06,
+      "loss": 0.4765,
+      "step": 3378
+    },
+    {
+      "epoch": 1.6766786645153284,
+      "grad_norm": 0.06876300652452302,
+      "learning_rate": 6.273944357653152e-06,
+      "loss": 0.4757,
+      "step": 3379
+    },
+    {
+      "epoch": 1.6771751272185678,
+      "grad_norm": 0.07182670922356638,
+      "learning_rate": 6.272054072601374e-06,
+      "loss": 0.4886,
+      "step": 3380
+    },
+    {
+      "epoch": 1.677671589921807,
+      "grad_norm": 0.07003890530333579,
+      "learning_rate": 6.270163593136525e-06,
+      "loss": 0.5033,
+      "step": 3381
+    },
+    {
+      "epoch": 1.6781680526250464,
+      "grad_norm": 0.07243728795553099,
+      "learning_rate": 6.268272919547537e-06,
+      "loss": 0.5108,
+      "step": 3382
+    },
+    {
+      "epoch": 1.678664515328286,
+      "grad_norm": 0.07219023545275664,
+      "learning_rate": 6.266382052123369e-06,
+      "loss": 0.5075,
+      "step": 3383
+    },
+    {
+      "epoch": 1.6791609780315255,
+      "grad_norm": 0.07226117961469808,
+      "learning_rate": 6.26449099115301e-06,
+      "loss": 0.4954,
+      "step": 3384
+    },
+    {
+      "epoch": 1.6796574407347649,
+      "grad_norm": 0.07054829849640398,
+      "learning_rate": 6.2625997369254765e-06,
+      "loss": 0.4919,
+      "step": 3385
+    },
+    {
+      "epoch": 1.6801539034380042,
+      "grad_norm": 0.07199319774173396,
+      "learning_rate": 6.260708289729818e-06,
+      "loss": 0.503,
+      "step": 3386
+    },
+    {
+      "epoch": 1.6806503661412435,
+      "grad_norm": 0.07052930419325945,
+      "learning_rate": 6.258816649855109e-06,
+      "loss": 0.4896,
+      "step": 3387
+    },
+    {
+      "epoch": 1.681146828844483,
+      "grad_norm": 0.06905137996462095,
+      "learning_rate": 6.2569248175904615e-06,
+      "loss": 0.4785,
+      "step": 3388
+    },
+    {
+      "epoch": 1.6816432915477226,
+      "grad_norm": 0.07053248386925157,
+      "learning_rate": 6.2550327932250085e-06,
+      "loss": 0.4925,
+      "step": 3389
+    },
+    {
+      "epoch": 1.682139754250962,
+      "grad_norm": 0.07224569722130686,
+      "learning_rate": 6.2531405770479146e-06,
+      "loss": 0.4706,
+      "step": 3390
+    },
+    {
+      "epoch": 1.6826362169542013,
+      "grad_norm": 0.07274752966145347,
+      "learning_rate": 6.251248169348376e-06,
+      "loss": 0.4741,
+      "step": 3391
+    },
+    {
+      "epoch": 1.6831326796574406,
+      "grad_norm": 0.07165526373877322,
+      "learning_rate": 6.24935557041562e-06,
+      "loss": 0.4764,
+      "step": 3392
+    },
+    {
+      "epoch": 1.6836291423606802,
+      "grad_norm": 0.06999372308594969,
+      "learning_rate": 6.247462780538893e-06,
+      "loss": 0.4656,
+      "step": 3393
+    },
+    {
+      "epoch": 1.6841256050639197,
+      "grad_norm": 0.07399732812656597,
+      "learning_rate": 6.245569800007484e-06,
+      "loss": 0.5117,
+      "step": 3394
+    },
+    {
+      "epoch": 1.684622067767159,
+      "grad_norm": 0.07362880549471516,
+      "learning_rate": 6.243676629110702e-06,
+      "loss": 0.5538,
+      "step": 3395
+    },
+    {
+      "epoch": 1.6851185304703984,
+      "grad_norm": 0.0727603851572235,
+      "learning_rate": 6.241783268137888e-06,
+      "loss": 0.5069,
+      "step": 3396
+    },
+    {
+      "epoch": 1.6856149931736377,
+      "grad_norm": 0.07428177312369587,
+      "learning_rate": 6.239889717378411e-06,
+      "loss": 0.5192,
+      "step": 3397
+    },
+    {
+      "epoch": 1.6861114558768773,
+      "grad_norm": 0.07459257178624475,
+      "learning_rate": 6.2379959771216716e-06,
+      "loss": 0.4922,
+      "step": 3398
+    },
+    {
+      "epoch": 1.6866079185801168,
+      "grad_norm": 0.06924537401803085,
+      "learning_rate": 6.236102047657096e-06,
+      "loss": 0.5061,
+      "step": 3399
+    },
+    {
+      "epoch": 1.6871043812833562,
+      "grad_norm": 0.07243627093577376,
+      "learning_rate": 6.234207929274143e-06,
+      "loss": 0.4906,
+      "step": 3400
+    },
+    {
+      "epoch": 1.6876008439865955,
+      "grad_norm": 0.07172235410923561,
+      "learning_rate": 6.232313622262297e-06,
+      "loss": 0.4827,
+      "step": 3401
+    },
+    {
+      "epoch": 1.6880973066898348,
+      "grad_norm": 0.06923536972014138,
+      "learning_rate": 6.230419126911072e-06,
+      "loss": 0.4546,
+      "step": 3402
+    },
+    {
+      "epoch": 1.6885937693930744,
+      "grad_norm": 0.071292437387297,
+      "learning_rate": 6.228524443510011e-06,
+      "loss": 0.5121,
+      "step": 3403
+    },
+    {
+      "epoch": 1.689090232096314,
+      "grad_norm": 0.06941047555292386,
+      "learning_rate": 6.226629572348687e-06,
+      "loss": 0.4657,
+      "step": 3404
+    },
+    {
+      "epoch": 1.6895866947995533,
+      "grad_norm": 0.07432595643381897,
+      "learning_rate": 6.224734513716702e-06,
+      "loss": 0.4883,
+      "step": 3405
+    },
+    {
+      "epoch": 1.6900831575027926,
+      "grad_norm": 0.07122928064283822,
+      "learning_rate": 6.222839267903682e-06,
+      "loss": 0.4678,
+      "step": 3406
+    },
+    {
+      "epoch": 1.690579620206032,
+      "grad_norm": 0.06841699929757633,
+      "learning_rate": 6.220943835199286e-06,
+      "loss": 0.4791,
+      "step": 3407
+    },
+    {
+      "epoch": 1.6910760829092715,
+      "grad_norm": 0.06823369275178195,
+      "learning_rate": 6.219048215893204e-06,
+      "loss": 0.4514,
+      "step": 3408
+    },
+    {
+      "epoch": 1.691572545612511,
+      "grad_norm": 0.06835513269888316,
+      "learning_rate": 6.2171524102751454e-06,
+      "loss": 0.47,
+      "step": 3409
+    },
+    {
+      "epoch": 1.6920690083157504,
+      "grad_norm": 0.07859367473920947,
+      "learning_rate": 6.215256418634858e-06,
+      "loss": 0.5113,
+      "step": 3410
+    },
+    {
+      "epoch": 1.6925654710189897,
+      "grad_norm": 0.07080399011507223,
+      "learning_rate": 6.2133602412621116e-06,
+      "loss": 0.4849,
+      "step": 3411
+    },
+    {
+      "epoch": 1.693061933722229,
+      "grad_norm": 0.06993443237391803,
+      "learning_rate": 6.211463878446708e-06,
+      "loss": 0.4665,
+      "step": 3412
+    },
+    {
+      "epoch": 1.6935583964254686,
+      "grad_norm": 0.07103193818369047,
+      "learning_rate": 6.209567330478473e-06,
+      "loss": 0.4787,
+      "step": 3413
+    },
+    {
+      "epoch": 1.694054859128708,
+      "grad_norm": 0.0718499238641342,
+      "learning_rate": 6.207670597647266e-06,
+      "loss": 0.4775,
+      "step": 3414
+    },
+    {
+      "epoch": 1.6945513218319475,
+      "grad_norm": 0.0713450002416761,
+      "learning_rate": 6.2057736802429724e-06,
+      "loss": 0.4963,
+      "step": 3415
+    },
+    {
+      "epoch": 1.6950477845351868,
+      "grad_norm": 0.07188266897364559,
+      "learning_rate": 6.203876578555506e-06,
+      "loss": 0.477,
+      "step": 3416
+    },
+    {
+      "epoch": 1.6955442472384261,
+      "grad_norm": 0.07065734028417589,
+      "learning_rate": 6.201979292874805e-06,
+      "loss": 0.4843,
+      "step": 3417
+    },
+    {
+      "epoch": 1.6960407099416657,
+      "grad_norm": 0.07454692730874736,
+      "learning_rate": 6.200081823490842e-06,
+      "loss": 0.496,
+      "step": 3418
+    },
+    {
+      "epoch": 1.696537172644905,
+      "grad_norm": 0.07376467418188615,
+      "learning_rate": 6.198184170693615e-06,
+      "loss": 0.5022,
+      "step": 3419
+    },
+    {
+      "epoch": 1.6970336353481446,
+      "grad_norm": 0.07109089901080787,
+      "learning_rate": 6.196286334773148e-06,
+      "loss": 0.454,
+      "step": 3420
+    },
+    {
+      "epoch": 1.6975300980513839,
+      "grad_norm": 0.07441012396251873,
+      "learning_rate": 6.194388316019495e-06,
+      "loss": 0.4949,
+      "step": 3421
+    },
+    {
+      "epoch": 1.6980265607546232,
+      "grad_norm": 0.0737655799446165,
+      "learning_rate": 6.192490114722741e-06,
+      "loss": 0.5255,
+      "step": 3422
+    },
+    {
+      "epoch": 1.6985230234578628,
+      "grad_norm": 0.07027598683919967,
+      "learning_rate": 6.1905917311729915e-06,
+      "loss": 0.476,
+      "step": 3423
+    },
+    {
+      "epoch": 1.699019486161102,
+      "grad_norm": 0.07501683188473765,
+      "learning_rate": 6.188693165660387e-06,
+      "loss": 0.5079,
+      "step": 3424
+    },
+    {
+      "epoch": 1.6995159488643417,
+      "grad_norm": 0.07110659680282254,
+      "learning_rate": 6.1867944184750894e-06,
+      "loss": 0.4953,
+      "step": 3425
+    },
+    {
+      "epoch": 1.700012411567581,
+      "grad_norm": 0.0711979716541948,
+      "learning_rate": 6.184895489907293e-06,
+      "loss": 0.4677,
+      "step": 3426
+    },
+    {
+      "epoch": 1.7005088742708203,
+      "grad_norm": 0.07512243922895441,
+      "learning_rate": 6.182996380247223e-06,
+      "loss": 0.4969,
+      "step": 3427
+    },
+    {
+      "epoch": 1.7010053369740599,
+      "grad_norm": 0.07291393767148943,
+      "learning_rate": 6.181097089785121e-06,
+      "loss": 0.4645,
+      "step": 3428
+    },
+    {
+      "epoch": 1.7015017996772992,
+      "grad_norm": 0.071546062557215,
+      "learning_rate": 6.179197618811267e-06,
+      "loss": 0.4995,
+      "step": 3429
+    },
+    {
+      "epoch": 1.7019982623805388,
+      "grad_norm": 0.07167043597024016,
+      "learning_rate": 6.177297967615964e-06,
+      "loss": 0.4999,
+      "step": 3430
+    },
+    {
+      "epoch": 1.702494725083778,
+      "grad_norm": 0.07257860964521468,
+      "learning_rate": 6.175398136489542e-06,
+      "loss": 0.5015,
+      "step": 3431
+    },
+    {
+      "epoch": 1.7029911877870174,
+      "grad_norm": 0.07284246143070047,
+      "learning_rate": 6.173498125722363e-06,
+      "loss": 0.484,
+      "step": 3432
+    },
+    {
+      "epoch": 1.703487650490257,
+      "grad_norm": 0.0708350136827653,
+      "learning_rate": 6.171597935604811e-06,
+      "loss": 0.5185,
+      "step": 3433
+    },
+    {
+      "epoch": 1.7039841131934963,
+      "grad_norm": 0.07204652200212407,
+      "learning_rate": 6.1696975664273e-06,
+      "loss": 0.5097,
+      "step": 3434
+    },
+    {
+      "epoch": 1.7044805758967358,
+      "grad_norm": 0.07273453576826858,
+      "learning_rate": 6.167797018480268e-06,
+      "loss": 0.4757,
+      "step": 3435
+    },
+    {
+      "epoch": 1.7049770385999752,
+      "grad_norm": 0.07288642192421328,
+      "learning_rate": 6.1658962920541875e-06,
+      "loss": 0.5086,
+      "step": 3436
+    },
+    {
+      "epoch": 1.7054735013032145,
+      "grad_norm": 0.06952770295584065,
+      "learning_rate": 6.1639953874395534e-06,
+      "loss": 0.4691,
+      "step": 3437
+    },
+    {
+      "epoch": 1.705969964006454,
+      "grad_norm": 0.07091606836576105,
+      "learning_rate": 6.1620943049268865e-06,
+      "loss": 0.4848,
+      "step": 3438
+    },
+    {
+      "epoch": 1.7064664267096934,
+      "grad_norm": 0.07492107718253822,
+      "learning_rate": 6.160193044806738e-06,
+      "loss": 0.5146,
+      "step": 3439
+    },
+    {
+      "epoch": 1.706962889412933,
+      "grad_norm": 0.07019956876848898,
+      "learning_rate": 6.158291607369686e-06,
+      "loss": 0.5008,
+      "step": 3440
+    },
+    {
+      "epoch": 1.7074593521161723,
+      "grad_norm": 0.07279278421648618,
+      "learning_rate": 6.156389992906332e-06,
+      "loss": 0.4932,
+      "step": 3441
+    },
+    {
+      "epoch": 1.7079558148194116,
+      "grad_norm": 0.07522974421670806,
+      "learning_rate": 6.154488201707309e-06,
+      "loss": 0.4831,
+      "step": 3442
+    },
+    {
+      "epoch": 1.7084522775226512,
+      "grad_norm": 0.07153871451762317,
+      "learning_rate": 6.152586234063277e-06,
+      "loss": 0.4855,
+      "step": 3443
+    },
+    {
+      "epoch": 1.7089487402258905,
+      "grad_norm": 0.07244260575924184,
+      "learning_rate": 6.150684090264918e-06,
+      "loss": 0.502,
+      "step": 3444
+    },
+    {
+      "epoch": 1.70944520292913,
+      "grad_norm": 0.07312446444547795,
+      "learning_rate": 6.148781770602945e-06,
+      "loss": 0.4932,
+      "step": 3445
+    },
+    {
+      "epoch": 1.7099416656323694,
+      "grad_norm": 0.07183132744757643,
+      "learning_rate": 6.146879275368098e-06,
+      "loss": 0.4644,
+      "step": 3446
+    },
+    {
+      "epoch": 1.7104381283356087,
+      "grad_norm": 0.0728091776261513,
+      "learning_rate": 6.144976604851143e-06,
+      "loss": 0.4797,
+      "step": 3447
+    },
+    {
+      "epoch": 1.710934591038848,
+      "grad_norm": 0.07686865515523557,
+      "learning_rate": 6.143073759342872e-06,
+      "loss": 0.4613,
+      "step": 3448
+    },
+    {
+      "epoch": 1.7114310537420876,
+      "grad_norm": 0.06862548003626924,
+      "learning_rate": 6.141170739134107e-06,
+      "loss": 0.4723,
+      "step": 3449
+    },
+    {
+      "epoch": 1.7119275164453271,
+      "grad_norm": 0.0698538911972819,
+      "learning_rate": 6.139267544515689e-06,
+      "loss": 0.4839,
+      "step": 3450
+    },
+    {
+      "epoch": 1.7124239791485665,
+      "grad_norm": 0.07008192081244582,
+      "learning_rate": 6.1373641757784945e-06,
+      "loss": 0.468,
+      "step": 3451
+    },
+    {
+      "epoch": 1.7129204418518058,
+      "grad_norm": 0.07095298359236782,
+      "learning_rate": 6.135460633213422e-06,
+      "loss": 0.5054,
+      "step": 3452
+    },
+    {
+      "epoch": 1.7134169045550451,
+      "grad_norm": 0.07446355810920537,
+      "learning_rate": 6.133556917111396e-06,
+      "loss": 0.5204,
+      "step": 3453
+    },
+    {
+      "epoch": 1.7139133672582847,
+      "grad_norm": 0.07115588859641515,
+      "learning_rate": 6.131653027763372e-06,
+      "loss": 0.4861,
+      "step": 3454
+    },
+    {
+      "epoch": 1.7144098299615242,
+      "grad_norm": 0.0730323307012295,
+      "learning_rate": 6.129748965460327e-06,
+      "loss": 0.531,
+      "step": 3455
+    },
+    {
+      "epoch": 1.7149062926647636,
+      "grad_norm": 0.0728929348845048,
+      "learning_rate": 6.127844730493267e-06,
+      "loss": 0.5021,
+      "step": 3456
+    },
+    {
+      "epoch": 1.715402755368003,
+      "grad_norm": 0.07088824802373886,
+      "learning_rate": 6.125940323153223e-06,
+      "loss": 0.4749,
+      "step": 3457
+    },
+    {
+      "epoch": 1.7158992180712422,
+      "grad_norm": 0.07616952737021909,
+      "learning_rate": 6.1240357437312544e-06,
+      "loss": 0.4908,
+      "step": 3458
+    },
+    {
+      "epoch": 1.7163956807744818,
+      "grad_norm": 0.06961257299018885,
+      "learning_rate": 6.122130992518444e-06,
+      "loss": 0.4518,
+      "step": 3459
+    },
+    {
+      "epoch": 1.7168921434777213,
+      "grad_norm": 0.06954236191260457,
+      "learning_rate": 6.120226069805904e-06,
+      "loss": 0.4813,
+      "step": 3460
+    },
+    {
+      "epoch": 1.7173886061809607,
+      "grad_norm": 0.07114725015517909,
+      "learning_rate": 6.1183209758847715e-06,
+      "loss": 0.4787,
+      "step": 3461
+    },
+    {
+      "epoch": 1.7178850688842,
+      "grad_norm": 0.06980692059756774,
+      "learning_rate": 6.116415711046208e-06,
+      "loss": 0.4688,
+      "step": 3462
+    },
+    {
+      "epoch": 1.7183815315874393,
+      "grad_norm": 0.07553272063236754,
+      "learning_rate": 6.114510275581402e-06,
+      "loss": 0.5312,
+      "step": 3463
+    },
+    {
+      "epoch": 1.718877994290679,
+      "grad_norm": 0.07750265108845414,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.5201,
+      "step": 3464
+    },
+    {
+      "epoch": 1.7193744569939184,
+      "grad_norm": 0.07239752989184631,
+      "learning_rate": 6.1106988939379584e-06,
+      "loss": 0.4487,
+      "step": 3465
+    },
+    {
+      "epoch": 1.7198709196971578,
+      "grad_norm": 0.07267185258184378,
+      "learning_rate": 6.108792948341826e-06,
+      "loss": 0.5139,
+      "step": 3466
+    },
+    {
+      "epoch": 1.720367382400397,
+      "grad_norm": 0.06984553929988284,
+      "learning_rate": 6.106886833284469e-06,
+      "loss": 0.488,
+      "step": 3467
+    },
+    {
+      "epoch": 1.7208638451036364,
+      "grad_norm": 0.07633088437476376,
+      "learning_rate": 6.104980549057208e-06,
+      "loss": 0.5496,
+      "step": 3468
+    },
+    {
+      "epoch": 1.721360307806876,
+      "grad_norm": 0.07185447533511291,
+      "learning_rate": 6.103074095951387e-06,
+      "loss": 0.4559,
+      "step": 3469
+    },
+    {
+      "epoch": 1.7218567705101155,
+      "grad_norm": 0.07159619149542636,
+      "learning_rate": 6.101167474258377e-06,
+      "loss": 0.4655,
+      "step": 3470
+    },
+    {
+      "epoch": 1.7223532332133549,
+      "grad_norm": 0.07021731374652906,
+      "learning_rate": 6.0992606842695745e-06,
+      "loss": 0.4529,
+      "step": 3471
+    },
+    {
+      "epoch": 1.7228496959165942,
+      "grad_norm": 0.07932591092305574,
+      "learning_rate": 6.0973537262764024e-06,
+      "loss": 0.5198,
+      "step": 3472
+    },
+    {
+      "epoch": 1.7233461586198335,
+      "grad_norm": 0.07128172870853265,
+      "learning_rate": 6.095446600570306e-06,
+      "loss": 0.4945,
+      "step": 3473
+    },
+    {
+      "epoch": 1.723842621323073,
+      "grad_norm": 0.07099044484886086,
+      "learning_rate": 6.093539307442762e-06,
+      "loss": 0.4538,
+      "step": 3474
+    },
+    {
+      "epoch": 1.7243390840263126,
+      "grad_norm": 0.06842920295839011,
+      "learning_rate": 6.091631847185268e-06,
+      "loss": 0.4562,
+      "step": 3475
+    },
+    {
+      "epoch": 1.724835546729552,
+      "grad_norm": 0.0739575588141388,
+      "learning_rate": 6.089724220089351e-06,
+      "loss": 0.5191,
+      "step": 3476
+    },
+    {
+      "epoch": 1.7253320094327913,
+      "grad_norm": 0.08136158471753265,
+      "learning_rate": 6.087816426446557e-06,
+      "loss": 0.4862,
+      "step": 3477
+    },
+    {
+      "epoch": 1.7258284721360306,
+      "grad_norm": 0.07013805452078947,
+      "learning_rate": 6.0859084665484645e-06,
+      "loss": 0.4773,
+      "step": 3478
+    },
+    {
+      "epoch": 1.7263249348392702,
+      "grad_norm": 0.0739863259677037,
+      "learning_rate": 6.084000340686674e-06,
+      "loss": 0.4895,
+      "step": 3479
+    },
+    {
+      "epoch": 1.7268213975425097,
+      "grad_norm": 0.07078698868835145,
+      "learning_rate": 6.082092049152813e-06,
+      "loss": 0.4581,
+      "step": 3480
+    },
+    {
+      "epoch": 1.727317860245749,
+      "grad_norm": 0.07130648089053174,
+      "learning_rate": 6.080183592238533e-06,
+      "loss": 0.4964,
+      "step": 3481
+    },
+    {
+      "epoch": 1.7278143229489884,
+      "grad_norm": 0.07373187736910049,
+      "learning_rate": 6.078274970235509e-06,
+      "loss": 0.4881,
+      "step": 3482
+    },
+    {
+      "epoch": 1.7283107856522277,
+      "grad_norm": 0.07263393261901284,
+      "learning_rate": 6.076366183435445e-06,
+      "loss": 0.4675,
+      "step": 3483
+    },
+    {
+      "epoch": 1.7288072483554673,
+      "grad_norm": 0.07232290837907768,
+      "learning_rate": 6.074457232130067e-06,
+      "loss": 0.5158,
+      "step": 3484
+    },
+    {
+      "epoch": 1.7293037110587068,
+      "grad_norm": 0.07569196194716082,
+      "learning_rate": 6.07254811661113e-06,
+      "loss": 0.5033,
+      "step": 3485
+    },
+    {
+      "epoch": 1.7298001737619462,
+      "grad_norm": 0.0737678773736528,
+      "learning_rate": 6.0706388371704104e-06,
+      "loss": 0.5094,
+      "step": 3486
+    },
+    {
+      "epoch": 1.7302966364651855,
+      "grad_norm": 0.0767658102737874,
+      "learning_rate": 6.068729394099711e-06,
+      "loss": 0.5083,
+      "step": 3487
+    },
+    {
+      "epoch": 1.7307930991684248,
+      "grad_norm": 0.07118568963078373,
+      "learning_rate": 6.066819787690859e-06,
+      "loss": 0.4844,
+      "step": 3488
+    },
+    {
+      "epoch": 1.7312895618716644,
+      "grad_norm": 0.0733650587823233,
+      "learning_rate": 6.064910018235707e-06,
+      "loss": 0.5007,
+      "step": 3489
+    },
+    {
+      "epoch": 1.731786024574904,
+      "grad_norm": 0.07419443969678913,
+      "learning_rate": 6.063000086026134e-06,
+      "loss": 0.4936,
+      "step": 3490
+    },
+    {
+      "epoch": 1.7322824872781433,
+      "grad_norm": 0.07144300280941726,
+      "learning_rate": 6.061089991354041e-06,
+      "loss": 0.5035,
+      "step": 3491
+    },
+    {
+      "epoch": 1.7327789499813826,
+      "grad_norm": 0.07269441526800977,
+      "learning_rate": 6.059179734511357e-06,
+      "loss": 0.5006,
+      "step": 3492
+    },
+    {
+      "epoch": 1.733275412684622,
+      "grad_norm": 0.06735909256472328,
+      "learning_rate": 6.057269315790033e-06,
+      "loss": 0.4787,
+      "step": 3493
+    },
+    {
+      "epoch": 1.7337718753878615,
+      "grad_norm": 0.07101972753252063,
+      "learning_rate": 6.055358735482045e-06,
+      "loss": 0.5026,
+      "step": 3494
+    },
+    {
+      "epoch": 1.734268338091101,
+      "grad_norm": 0.07316133121592043,
+      "learning_rate": 6.053447993879397e-06,
+      "loss": 0.5288,
+      "step": 3495
+    },
+    {
+      "epoch": 1.7347648007943404,
+      "grad_norm": 0.07072090733545723,
+      "learning_rate": 6.051537091274115e-06,
+      "loss": 0.477,
+      "step": 3496
+    },
+    {
+      "epoch": 1.7352612634975797,
+      "grad_norm": 0.06911247409586789,
+      "learning_rate": 6.049626027958246e-06,
+      "loss": 0.4501,
+      "step": 3497
+    },
+    {
+      "epoch": 1.735757726200819,
+      "grad_norm": 0.07632319474987508,
+      "learning_rate": 6.047714804223871e-06,
+      "loss": 0.5038,
+      "step": 3498
+    },
+    {
+      "epoch": 1.7362541889040586,
+      "grad_norm": 0.07068492441512336,
+      "learning_rate": 6.045803420363085e-06,
+      "loss": 0.4719,
+      "step": 3499
+    },
+    {
+      "epoch": 1.7367506516072981,
+      "grad_norm": 0.08074542221028214,
+      "learning_rate": 6.043891876668015e-06,
+      "loss": 0.5156,
+      "step": 3500
+    },
+    {
+      "epoch": 1.7372471143105375,
+      "grad_norm": 0.07228894317585616,
+      "learning_rate": 6.0419801734308085e-06,
+      "loss": 0.5086,
+      "step": 3501
+    },
+    {
+      "epoch": 1.7377435770137768,
+      "grad_norm": 0.07240394720933138,
+      "learning_rate": 6.04006831094364e-06,
+      "loss": 0.4768,
+      "step": 3502
+    },
+    {
+      "epoch": 1.7382400397170161,
+      "grad_norm": 0.07257800871912748,
+      "learning_rate": 6.038156289498705e-06,
+      "loss": 0.4864,
+      "step": 3503
+    },
+    {
+      "epoch": 1.7387365024202557,
+      "grad_norm": 0.07693716933243037,
+      "learning_rate": 6.03624410938823e-06,
+      "loss": 0.5101,
+      "step": 3504
+    },
+    {
+      "epoch": 1.7392329651234952,
+      "grad_norm": 0.07042241058097551,
+      "learning_rate": 6.034331770904455e-06,
+      "loss": 0.4866,
+      "step": 3505
+    },
+    {
+      "epoch": 1.7397294278267346,
+      "grad_norm": 0.07029230350366913,
+      "learning_rate": 6.032419274339654e-06,
+      "loss": 0.5054,
+      "step": 3506
+    },
+    {
+      "epoch": 1.740225890529974,
+      "grad_norm": 0.07177352673078713,
+      "learning_rate": 6.03050661998612e-06,
+      "loss": 0.4853,
+      "step": 3507
+    },
+    {
+      "epoch": 1.7407223532332132,
+      "grad_norm": 0.07412862457534308,
+      "learning_rate": 6.028593808136173e-06,
+      "loss": 0.5043,
+      "step": 3508
+    },
+    {
+      "epoch": 1.7412188159364528,
+      "grad_norm": 0.07261567809098946,
+      "learning_rate": 6.026680839082153e-06,
+      "loss": 0.5034,
+      "step": 3509
+    },
+    {
+      "epoch": 1.7417152786396923,
+      "grad_norm": 0.07011596539977943,
+      "learning_rate": 6.024767713116429e-06,
+      "loss": 0.4814,
+      "step": 3510
+    },
+    {
+      "epoch": 1.7422117413429317,
+      "grad_norm": 0.0715432820186117,
+      "learning_rate": 6.022854430531392e-06,
+      "loss": 0.4729,
+      "step": 3511
+    },
+    {
+      "epoch": 1.742708204046171,
+      "grad_norm": 0.0713065455423351,
+      "learning_rate": 6.020940991619455e-06,
+      "loss": 0.506,
+      "step": 3512
+    },
+    {
+      "epoch": 1.7432046667494103,
+      "grad_norm": 0.07050208269041043,
+      "learning_rate": 6.019027396673058e-06,
+      "loss": 0.4828,
+      "step": 3513
+    },
+    {
+      "epoch": 1.7437011294526499,
+      "grad_norm": 0.07360098624736137,
+      "learning_rate": 6.01711364598466e-06,
+      "loss": 0.5075,
+      "step": 3514
+    },
+    {
+      "epoch": 1.7441975921558894,
+      "grad_norm": 0.07351220296533653,
+      "learning_rate": 6.015199739846751e-06,
+      "loss": 0.4904,
+      "step": 3515
+    },
+    {
+      "epoch": 1.7446940548591288,
+      "grad_norm": 0.07781887095221011,
+      "learning_rate": 6.013285678551838e-06,
+      "loss": 0.4954,
+      "step": 3516
+    },
+    {
+      "epoch": 1.745190517562368,
+      "grad_norm": 0.07444681722627283,
+      "learning_rate": 6.011371462392457e-06,
+      "loss": 0.4682,
+      "step": 3517
+    },
+    {
+      "epoch": 1.7456869802656074,
+      "grad_norm": 0.07313248542856911,
+      "learning_rate": 6.0094570916611635e-06,
+      "loss": 0.4997,
+      "step": 3518
+    },
+    {
+      "epoch": 1.746183442968847,
+      "grad_norm": 0.07056610694243752,
+      "learning_rate": 6.007542566650539e-06,
+      "loss": 0.4805,
+      "step": 3519
+    },
+    {
+      "epoch": 1.7466799056720865,
+      "grad_norm": 0.0731856933833494,
+      "learning_rate": 6.005627887653189e-06,
+      "loss": 0.4672,
+      "step": 3520
+    },
+    {
+      "epoch": 1.7471763683753259,
+      "grad_norm": 0.0733689880039657,
+      "learning_rate": 6.00371305496174e-06,
+      "loss": 0.5161,
+      "step": 3521
+    },
+    {
+      "epoch": 1.7476728310785652,
+      "grad_norm": 0.07171031605313356,
+      "learning_rate": 6.001798068868842e-06,
+      "loss": 0.4733,
+      "step": 3522
+    },
+    {
+      "epoch": 1.7481692937818045,
+      "grad_norm": 0.07327381780949192,
+      "learning_rate": 5.999882929667173e-06,
+      "loss": 0.4934,
+      "step": 3523
+    },
+    {
+      "epoch": 1.748665756485044,
+      "grad_norm": 0.06983487291449605,
+      "learning_rate": 5.997967637649431e-06,
+      "loss": 0.4889,
+      "step": 3524
+    },
+    {
+      "epoch": 1.7491622191882836,
+      "grad_norm": 0.06937374272648494,
+      "learning_rate": 5.996052193108336e-06,
+      "loss": 0.4744,
+      "step": 3525
+    },
+    {
+      "epoch": 1.749658681891523,
+      "grad_norm": 0.07137647015287153,
+      "learning_rate": 5.994136596336633e-06,
+      "loss": 0.5007,
+      "step": 3526
+    },
+    {
+      "epoch": 1.7501551445947623,
+      "grad_norm": 0.07233499968766187,
+      "learning_rate": 5.9922208476270914e-06,
+      "loss": 0.4782,
+      "step": 3527
+    },
+    {
+      "epoch": 1.7506516072980016,
+      "grad_norm": 0.07007113610888752,
+      "learning_rate": 5.990304947272503e-06,
+      "loss": 0.4633,
+      "step": 3528
+    },
+    {
+      "epoch": 1.7506516072980016,
+      "eval_loss": 0.5171855092048645,
+      "eval_runtime": 259.2374,
+      "eval_samples_per_second": 117.086,
+      "eval_steps_per_second": 14.639,
+      "step": 3528
+    },
+    {
+      "epoch": 1.7511480700012412,
+      "grad_norm": 0.0726685919128279,
+      "learning_rate": 5.988388895565681e-06,
+      "loss": 0.4911,
+      "step": 3529
+    },
+    {
+      "epoch": 1.7516445327044807,
+      "grad_norm": 0.07643078823588302,
+      "learning_rate": 5.986472692799465e-06,
+      "loss": 0.5087,
+      "step": 3530
+    },
+    {
+      "epoch": 1.75214099540772,
+      "grad_norm": 0.06621160646927267,
+      "learning_rate": 5.984556339266714e-06,
+      "loss": 0.458,
+      "step": 3531
+    },
+    {
+      "epoch": 1.7526374581109594,
+      "grad_norm": 0.06975490562081263,
+      "learning_rate": 5.9826398352603134e-06,
+      "loss": 0.4903,
+      "step": 3532
+    },
+    {
+      "epoch": 1.7531339208141987,
+      "grad_norm": 0.07050926318317503,
+      "learning_rate": 5.980723181073168e-06,
+      "loss": 0.4669,
+      "step": 3533
+    },
+    {
+      "epoch": 1.7536303835174383,
+      "grad_norm": 0.06895855481629098,
+      "learning_rate": 5.978806376998209e-06,
+      "loss": 0.4757,
+      "step": 3534
+    },
+    {
+      "epoch": 1.7541268462206778,
+      "grad_norm": 0.06917867418657715,
+      "learning_rate": 5.976889423328391e-06,
+      "loss": 0.4751,
+      "step": 3535
+    },
+    {
+      "epoch": 1.7546233089239172,
+      "grad_norm": 0.07451106891830402,
+      "learning_rate": 5.974972320356688e-06,
+      "loss": 0.5079,
+      "step": 3536
+    },
+    {
+      "epoch": 1.7551197716271565,
+      "grad_norm": 0.06940933717231007,
+      "learning_rate": 5.973055068376097e-06,
+      "loss": 0.4865,
+      "step": 3537
+    },
+    {
+      "epoch": 1.7556162343303958,
+      "grad_norm": 0.07457921708551443,
+      "learning_rate": 5.9711376676796404e-06,
+      "loss": 0.471,
+      "step": 3538
+    },
+    {
+      "epoch": 1.7561126970336354,
+      "grad_norm": 0.0743437605915209,
+      "learning_rate": 5.969220118560363e-06,
+      "loss": 0.4884,
+      "step": 3539
+    },
+    {
+      "epoch": 1.756609159736875,
+      "grad_norm": 0.07290986659316935,
+      "learning_rate": 5.967302421311331e-06,
+      "loss": 0.4943,
+      "step": 3540
+    },
+    {
+      "epoch": 1.7571056224401143,
+      "grad_norm": 0.0715462893233084,
+      "learning_rate": 5.965384576225632e-06,
+      "loss": 0.4962,
+      "step": 3541
+    },
+    {
+      "epoch": 1.7576020851433536,
+      "grad_norm": 0.07183399679200113,
+      "learning_rate": 5.96346658359638e-06,
+      "loss": 0.4726,
+      "step": 3542
+    },
+    {
+      "epoch": 1.758098547846593,
+      "grad_norm": 0.06982319209707803,
+      "learning_rate": 5.961548443716709e-06,
+      "loss": 0.4699,
+      "step": 3543
+    },
+    {
+      "epoch": 1.7585950105498325,
+      "grad_norm": 0.07251622340048915,
+      "learning_rate": 5.959630156879777e-06,
+      "loss": 0.4907,
+      "step": 3544
+    },
+    {
+      "epoch": 1.759091473253072,
+      "grad_norm": 0.07250648567166573,
+      "learning_rate": 5.957711723378759e-06,
+      "loss": 0.484,
+      "step": 3545
+    },
+    {
+      "epoch": 1.7595879359563114,
+      "grad_norm": 0.07596698660963411,
+      "learning_rate": 5.955793143506863e-06,
+      "loss": 0.4825,
+      "step": 3546
+    },
+    {
+      "epoch": 1.7600843986595507,
+      "grad_norm": 0.07171805789313515,
+      "learning_rate": 5.953874417557308e-06,
+      "loss": 0.4784,
+      "step": 3547
+    },
+    {
+      "epoch": 1.76058086136279,
+      "grad_norm": 0.07651650904469937,
+      "learning_rate": 5.951955545823342e-06,
+      "loss": 0.5252,
+      "step": 3548
+    },
+    {
+      "epoch": 1.7610773240660296,
+      "grad_norm": 0.07565542672573962,
+      "learning_rate": 5.950036528598235e-06,
+      "loss": 0.4815,
+      "step": 3549
+    },
+    {
+      "epoch": 1.7615737867692691,
+      "grad_norm": 0.07570080552039914,
+      "learning_rate": 5.948117366175278e-06,
+      "loss": 0.4925,
+      "step": 3550
+    },
+    {
+      "epoch": 1.7620702494725085,
+      "grad_norm": 0.07570945061559178,
+      "learning_rate": 5.946198058847783e-06,
+      "loss": 0.5056,
+      "step": 3551
+    },
+    {
+      "epoch": 1.7625667121757478,
+      "grad_norm": 0.0709021037213143,
+      "learning_rate": 5.944278606909086e-06,
+      "loss": 0.5034,
+      "step": 3552
+    },
+    {
+      "epoch": 1.7630631748789871,
+      "grad_norm": 0.07323276956843093,
+      "learning_rate": 5.942359010652544e-06,
+      "loss": 0.4951,
+      "step": 3553
+    },
+    {
+      "epoch": 1.7635596375822267,
+      "grad_norm": 0.07419262441731611,
+      "learning_rate": 5.940439270371538e-06,
+      "loss": 0.5276,
+      "step": 3554
+    },
+    {
+      "epoch": 1.764056100285466,
+      "grad_norm": 0.07025387603832438,
+      "learning_rate": 5.938519386359466e-06,
+      "loss": 0.4855,
+      "step": 3555
+    },
+    {
+      "epoch": 1.7645525629887056,
+      "grad_norm": 0.07550630555995716,
+      "learning_rate": 5.936599358909756e-06,
+      "loss": 0.4906,
+      "step": 3556
+    },
+    {
+      "epoch": 1.7650490256919449,
+      "grad_norm": 0.07080434868326954,
+      "learning_rate": 5.93467918831585e-06,
+      "loss": 0.4966,
+      "step": 3557
+    },
+    {
+      "epoch": 1.7655454883951842,
+      "grad_norm": 0.07189099890804855,
+      "learning_rate": 5.9327588748712165e-06,
+      "loss": 0.4963,
+      "step": 3558
+    },
+    {
+      "epoch": 1.7660419510984238,
+      "grad_norm": 0.07785509762182191,
+      "learning_rate": 5.930838418869343e-06,
+      "loss": 0.4484,
+      "step": 3559
+    },
+    {
+      "epoch": 1.766538413801663,
+      "grad_norm": 0.07505577734044602,
+      "learning_rate": 5.9289178206037456e-06,
+      "loss": 0.4751,
+      "step": 3560
+    },
+    {
+      "epoch": 1.7670348765049027,
+      "grad_norm": 0.07452023513954975,
+      "learning_rate": 5.926997080367951e-06,
+      "loss": 0.4901,
+      "step": 3561
+    },
+    {
+      "epoch": 1.767531339208142,
+      "grad_norm": 0.07419747069466309,
+      "learning_rate": 5.925076198455517e-06,
+      "loss": 0.4958,
+      "step": 3562
+    },
+    {
+      "epoch": 1.7680278019113813,
+      "grad_norm": 0.06962384440658652,
+      "learning_rate": 5.923155175160018e-06,
+      "loss": 0.4948,
+      "step": 3563
+    },
+    {
+      "epoch": 1.7685242646146209,
+      "grad_norm": 0.07400393848807546,
+      "learning_rate": 5.921234010775052e-06,
+      "loss": 0.5317,
+      "step": 3564
+    },
+    {
+      "epoch": 1.7690207273178602,
+      "grad_norm": 0.07358460565070393,
+      "learning_rate": 5.919312705594239e-06,
+      "loss": 0.5093,
+      "step": 3565
+    },
+    {
+      "epoch": 1.7695171900210998,
+      "grad_norm": 0.07520962763925339,
+      "learning_rate": 5.917391259911219e-06,
+      "loss": 0.5008,
+      "step": 3566
+    },
+    {
+      "epoch": 1.770013652724339,
+      "grad_norm": 0.07202658913675364,
+      "learning_rate": 5.915469674019654e-06,
+      "loss": 0.5081,
+      "step": 3567
+    },
+    {
+      "epoch": 1.7705101154275784,
+      "grad_norm": 0.07185311565574021,
+      "learning_rate": 5.913547948213227e-06,
+      "loss": 0.4871,
+      "step": 3568
+    },
+    {
+      "epoch": 1.771006578130818,
+      "grad_norm": 0.0735637908156275,
+      "learning_rate": 5.911626082785644e-06,
+      "loss": 0.4835,
+      "step": 3569
+    },
+    {
+      "epoch": 1.7715030408340573,
+      "grad_norm": 0.0735907052019009,
+      "learning_rate": 5.909704078030631e-06,
+      "loss": 0.4879,
+      "step": 3570
+    },
+    {
+      "epoch": 1.7719995035372968,
+      "grad_norm": 0.07008428137817518,
+      "learning_rate": 5.907781934241937e-06,
+      "loss": 0.5005,
+      "step": 3571
+    },
+    {
+      "epoch": 1.7724959662405362,
+      "grad_norm": 0.07729615402227608,
+      "learning_rate": 5.905859651713328e-06,
+      "loss": 0.4932,
+      "step": 3572
+    },
+    {
+      "epoch": 1.7729924289437755,
+      "grad_norm": 0.06997736005764132,
+      "learning_rate": 5.903937230738597e-06,
+      "loss": 0.4608,
+      "step": 3573
+    },
+    {
+      "epoch": 1.773488891647015,
+      "grad_norm": 0.0778643616820918,
+      "learning_rate": 5.902014671611553e-06,
+      "loss": 0.5364,
+      "step": 3574
+    },
+    {
+      "epoch": 1.7739853543502544,
+      "grad_norm": 0.07699448930722969,
+      "learning_rate": 5.900091974626028e-06,
+      "loss": 0.54,
+      "step": 3575
+    },
+    {
+      "epoch": 1.774481817053494,
+      "grad_norm": 0.07271609012983173,
+      "learning_rate": 5.898169140075878e-06,
+      "loss": 0.4869,
+      "step": 3576
+    },
+    {
+      "epoch": 1.7749782797567333,
+      "grad_norm": 0.07564281998394239,
+      "learning_rate": 5.896246168254976e-06,
+      "loss": 0.5295,
+      "step": 3577
+    },
+    {
+      "epoch": 1.7754747424599726,
+      "grad_norm": 0.07263423530612817,
+      "learning_rate": 5.894323059457218e-06,
+      "loss": 0.458,
+      "step": 3578
+    },
+    {
+      "epoch": 1.7759712051632122,
+      "grad_norm": 0.07281173591445123,
+      "learning_rate": 5.892399813976518e-06,
+      "loss": 0.4815,
+      "step": 3579
+    },
+    {
+      "epoch": 1.7764676678664515,
+      "grad_norm": 0.07438654268139841,
+      "learning_rate": 5.890476432106815e-06,
+      "loss": 0.4532,
+      "step": 3580
+    },
+    {
+      "epoch": 1.776964130569691,
+      "grad_norm": 0.07132895431017902,
+      "learning_rate": 5.8885529141420685e-06,
+      "loss": 0.4681,
+      "step": 3581
+    },
+    {
+      "epoch": 1.7774605932729304,
+      "grad_norm": 0.07337836587851508,
+      "learning_rate": 5.886629260376254e-06,
+      "loss": 0.5104,
+      "step": 3582
+    },
+    {
+      "epoch": 1.7779570559761697,
+      "grad_norm": 0.06888317038226811,
+      "learning_rate": 5.884705471103376e-06,
+      "loss": 0.5026,
+      "step": 3583
+    },
+    {
+      "epoch": 1.7784535186794093,
+      "grad_norm": 0.0704125385766139,
+      "learning_rate": 5.882781546617451e-06,
+      "loss": 0.5052,
+      "step": 3584
+    },
+    {
+      "epoch": 1.7789499813826486,
+      "grad_norm": 0.07913813709508663,
+      "learning_rate": 5.880857487212519e-06,
+      "loss": 0.4758,
+      "step": 3585
+    },
+    {
+      "epoch": 1.7794464440858881,
+      "grad_norm": 0.07299620364036855,
+      "learning_rate": 5.878933293182645e-06,
+      "loss": 0.4898,
+      "step": 3586
+    },
+    {
+      "epoch": 1.7799429067891275,
+      "grad_norm": 0.07113118681039178,
+      "learning_rate": 5.877008964821909e-06,
+      "loss": 0.4707,
+      "step": 3587
+    },
+    {
+      "epoch": 1.7804393694923668,
+      "grad_norm": 0.07609742154417785,
+      "learning_rate": 5.875084502424414e-06,
+      "loss": 0.5312,
+      "step": 3588
+    },
+    {
+      "epoch": 1.7809358321956061,
+      "grad_norm": 0.07232100575504447,
+      "learning_rate": 5.873159906284286e-06,
+      "loss": 0.4906,
+      "step": 3589
+    },
+    {
+      "epoch": 1.7814322948988457,
+      "grad_norm": 0.07262820041435715,
+      "learning_rate": 5.871235176695664e-06,
+      "loss": 0.4596,
+      "step": 3590
+    },
+    {
+      "epoch": 1.7819287576020852,
+      "grad_norm": 0.07247576171994086,
+      "learning_rate": 5.869310313952717e-06,
+      "loss": 0.4768,
+      "step": 3591
+    },
+    {
+      "epoch": 1.7824252203053246,
+      "grad_norm": 0.07203298175946792,
+      "learning_rate": 5.867385318349623e-06,
+      "loss": 0.477,
+      "step": 3592
+    },
+    {
+      "epoch": 1.782921683008564,
+      "grad_norm": 0.07392881293705612,
+      "learning_rate": 5.865460190180594e-06,
+      "loss": 0.5153,
+      "step": 3593
+    },
+    {
+      "epoch": 1.7834181457118032,
+      "grad_norm": 0.0756318218212652,
+      "learning_rate": 5.863534929739852e-06,
+      "loss": 0.4821,
+      "step": 3594
+    },
+    {
+      "epoch": 1.7839146084150428,
+      "grad_norm": 0.0747502853802229,
+      "learning_rate": 5.86160953732164e-06,
+      "loss": 0.4766,
+      "step": 3595
+    },
+    {
+      "epoch": 1.7844110711182823,
+      "grad_norm": 0.07297460569575788,
+      "learning_rate": 5.859684013220225e-06,
+      "loss": 0.4807,
+      "step": 3596
+    },
+    {
+      "epoch": 1.7849075338215217,
+      "grad_norm": 0.07155856829311029,
+      "learning_rate": 5.857758357729892e-06,
+      "loss": 0.4797,
+      "step": 3597
+    },
+    {
+      "epoch": 1.785403996524761,
+      "grad_norm": 0.07177962375247923,
+      "learning_rate": 5.855832571144947e-06,
+      "loss": 0.5285,
+      "step": 3598
+    },
+    {
+      "epoch": 1.7859004592280003,
+      "grad_norm": 0.0718258720492854,
+      "learning_rate": 5.853906653759718e-06,
+      "loss": 0.5025,
+      "step": 3599
+    },
+    {
+      "epoch": 1.78639692193124,
+      "grad_norm": 0.07145170610796107,
+      "learning_rate": 5.851980605868547e-06,
+      "loss": 0.5119,
+      "step": 3600
+    },
+    {
+      "epoch": 1.7868933846344794,
+      "grad_norm": 0.0711317878064843,
+      "learning_rate": 5.850054427765801e-06,
+      "loss": 0.4873,
+      "step": 3601
+    },
+    {
+      "epoch": 1.7873898473377188,
+      "grad_norm": 0.07248680936032043,
+      "learning_rate": 5.848128119745865e-06,
+      "loss": 0.4688,
+      "step": 3602
+    },
+    {
+      "epoch": 1.787886310040958,
+      "grad_norm": 0.07010165537492177,
+      "learning_rate": 5.846201682103144e-06,
+      "loss": 0.4681,
+      "step": 3603
+    },
+    {
+      "epoch": 1.7883827727441974,
+      "grad_norm": 0.0758110653402738,
+      "learning_rate": 5.844275115132064e-06,
+      "loss": 0.4998,
+      "step": 3604
+    },
+    {
+      "epoch": 1.788879235447437,
+      "grad_norm": 0.07165309101536645,
+      "learning_rate": 5.8423484191270705e-06,
+      "loss": 0.4885,
+      "step": 3605
+    },
+    {
+      "epoch": 1.7893756981506765,
+      "grad_norm": 0.0684537758607228,
+      "learning_rate": 5.840421594382627e-06,
+      "loss": 0.4421,
+      "step": 3606
+    },
+    {
+      "epoch": 1.7898721608539159,
+      "grad_norm": 0.07043974964585102,
+      "learning_rate": 5.838494641193217e-06,
+      "loss": 0.4978,
+      "step": 3607
+    },
+    {
+      "epoch": 1.7903686235571552,
+      "grad_norm": 0.07343477895002559,
+      "learning_rate": 5.836567559853346e-06,
+      "loss": 0.506,
+      "step": 3608
+    },
+    {
+      "epoch": 1.7908650862603945,
+      "grad_norm": 0.07131137027745556,
+      "learning_rate": 5.834640350657538e-06,
+      "loss": 0.5276,
+      "step": 3609
+    },
+    {
+      "epoch": 1.791361548963634,
+      "grad_norm": 0.06975660850835906,
+      "learning_rate": 5.832713013900333e-06,
+      "loss": 0.487,
+      "step": 3610
+    },
+    {
+      "epoch": 1.7918580116668736,
+      "grad_norm": 0.07013597758903344,
+      "learning_rate": 5.830785549876296e-06,
+      "loss": 0.4871,
+      "step": 3611
+    },
+    {
+      "epoch": 1.792354474370113,
+      "grad_norm": 0.07033271915719036,
+      "learning_rate": 5.828857958880008e-06,
+      "loss": 0.5121,
+      "step": 3612
+    },
+    {
+      "epoch": 1.7928509370733523,
+      "grad_norm": 0.07292135205875179,
+      "learning_rate": 5.826930241206071e-06,
+      "loss": 0.4784,
+      "step": 3613
+    },
+    {
+      "epoch": 1.7933473997765916,
+      "grad_norm": 0.07427049941543128,
+      "learning_rate": 5.825002397149105e-06,
+      "loss": 0.5119,
+      "step": 3614
+    },
+    {
+      "epoch": 1.7938438624798312,
+      "grad_norm": 0.07167706019268992,
+      "learning_rate": 5.823074427003752e-06,
+      "loss": 0.5509,
+      "step": 3615
+    },
+    {
+      "epoch": 1.7943403251830707,
+      "grad_norm": 0.07024605078269222,
+      "learning_rate": 5.821146331064669e-06,
+      "loss": 0.4837,
+      "step": 3616
+    },
+    {
+      "epoch": 1.79483678788631,
+      "grad_norm": 0.07174070828931835,
+      "learning_rate": 5.8192181096265355e-06,
+      "loss": 0.4731,
+      "step": 3617
+    },
+    {
+      "epoch": 1.7953332505895494,
+      "grad_norm": 0.07256450684278515,
+      "learning_rate": 5.817289762984048e-06,
+      "loss": 0.5094,
+      "step": 3618
+    },
+    {
+      "epoch": 1.7958297132927887,
+      "grad_norm": 0.07124042333943453,
+      "learning_rate": 5.8153612914319255e-06,
+      "loss": 0.479,
+      "step": 3619
+    },
+    {
+      "epoch": 1.7963261759960283,
+      "grad_norm": 0.07289082282917042,
+      "learning_rate": 5.813432695264903e-06,
+      "loss": 0.511,
+      "step": 3620
+    },
+    {
+      "epoch": 1.7968226386992678,
+      "grad_norm": 0.0685871218786059,
+      "learning_rate": 5.811503974777736e-06,
+      "loss": 0.4704,
+      "step": 3621
+    },
+    {
+      "epoch": 1.7973191014025072,
+      "grad_norm": 0.07621899432181643,
+      "learning_rate": 5.809575130265196e-06,
+      "loss": 0.5003,
+      "step": 3622
+    },
+    {
+      "epoch": 1.7978155641057465,
+      "grad_norm": 0.07243606226840418,
+      "learning_rate": 5.807646162022078e-06,
+      "loss": 0.4806,
+      "step": 3623
+    },
+    {
+      "epoch": 1.7983120268089858,
+      "grad_norm": 0.07334689387257744,
+      "learning_rate": 5.805717070343195e-06,
+      "loss": 0.502,
+      "step": 3624
+    },
+    {
+      "epoch": 1.7988084895122254,
+      "grad_norm": 0.07236811202655051,
+      "learning_rate": 5.803787855523377e-06,
+      "loss": 0.5212,
+      "step": 3625
+    },
+    {
+      "epoch": 1.799304952215465,
+      "grad_norm": 0.07021878117951692,
+      "learning_rate": 5.8018585178574714e-06,
+      "loss": 0.4777,
+      "step": 3626
+    },
+    {
+      "epoch": 1.7998014149187043,
+      "grad_norm": 0.07026592126708119,
+      "learning_rate": 5.79992905764035e-06,
+      "loss": 0.5112,
+      "step": 3627
+    },
+    {
+      "epoch": 1.8002978776219436,
+      "grad_norm": 0.07219620537052175,
+      "learning_rate": 5.797999475166897e-06,
+      "loss": 0.5062,
+      "step": 3628
+    },
+    {
+      "epoch": 1.800794340325183,
+      "grad_norm": 0.07311395206059088,
+      "learning_rate": 5.796069770732019e-06,
+      "loss": 0.4729,
+      "step": 3629
+    },
+    {
+      "epoch": 1.8012908030284225,
+      "grad_norm": 0.0728361404160052,
+      "learning_rate": 5.79413994463064e-06,
+      "loss": 0.4749,
+      "step": 3630
+    },
+    {
+      "epoch": 1.801787265731662,
+      "grad_norm": 0.07109436647193143,
+      "learning_rate": 5.792209997157705e-06,
+      "loss": 0.4888,
+      "step": 3631
+    },
+    {
+      "epoch": 1.8022837284349014,
+      "grad_norm": 0.07019275246724026,
+      "learning_rate": 5.790279928608173e-06,
+      "loss": 0.5022,
+      "step": 3632
+    },
+    {
+      "epoch": 1.8027801911381407,
+      "grad_norm": 0.07247375032569164,
+      "learning_rate": 5.788349739277026e-06,
+      "loss": 0.5117,
+      "step": 3633
+    },
+    {
+      "epoch": 1.80327665384138,
+      "grad_norm": 0.0699804846033475,
+      "learning_rate": 5.7864194294592615e-06,
+      "loss": 0.491,
+      "step": 3634
+    },
+    {
+      "epoch": 1.8037731165446196,
+      "grad_norm": 0.06993980001240863,
+      "learning_rate": 5.7844889994498955e-06,
+      "loss": 0.4906,
+      "step": 3635
+    },
+    {
+      "epoch": 1.8042695792478591,
+      "grad_norm": 0.07775193298527062,
+      "learning_rate": 5.782558449543964e-06,
+      "loss": 0.5035,
+      "step": 3636
+    },
+    {
+      "epoch": 1.8047660419510985,
+      "grad_norm": 0.0754723195092835,
+      "learning_rate": 5.780627780036523e-06,
+      "loss": 0.5056,
+      "step": 3637
+    },
+    {
+      "epoch": 1.8052625046543378,
+      "grad_norm": 0.08173656319877305,
+      "learning_rate": 5.77869699122264e-06,
+      "loss": 0.476,
+      "step": 3638
+    },
+    {
+      "epoch": 1.8057589673575771,
+      "grad_norm": 0.07211057900448045,
+      "learning_rate": 5.776766083397409e-06,
+      "loss": 0.5001,
+      "step": 3639
+    },
+    {
+      "epoch": 1.8062554300608167,
+      "grad_norm": 0.07357175060489121,
+      "learning_rate": 5.774835056855934e-06,
+      "loss": 0.4846,
+      "step": 3640
+    },
+    {
+      "epoch": 1.8067518927640562,
+      "grad_norm": 0.07392807260783969,
+      "learning_rate": 5.7729039118933476e-06,
+      "loss": 0.5086,
+      "step": 3641
+    },
+    {
+      "epoch": 1.8072483554672956,
+      "grad_norm": 0.0725775134465148,
+      "learning_rate": 5.770972648804789e-06,
+      "loss": 0.5274,
+      "step": 3642
+    },
+    {
+      "epoch": 1.807744818170535,
+      "grad_norm": 0.07163353524333979,
+      "learning_rate": 5.769041267885424e-06,
+      "loss": 0.4717,
+      "step": 3643
+    },
+    {
+      "epoch": 1.8082412808737742,
+      "grad_norm": 0.07105769685417251,
+      "learning_rate": 5.767109769430429e-06,
+      "loss": 0.4874,
+      "step": 3644
+    },
+    {
+      "epoch": 1.8087377435770138,
+      "grad_norm": 0.06975242191548883,
+      "learning_rate": 5.765178153735007e-06,
+      "loss": 0.4722,
+      "step": 3645
+    },
+    {
+      "epoch": 1.8092342062802533,
+      "grad_norm": 0.07347601381891362,
+      "learning_rate": 5.763246421094373e-06,
+      "loss": 0.496,
+      "step": 3646
+    },
+    {
+      "epoch": 1.8097306689834927,
+      "grad_norm": 0.07590220268232373,
+      "learning_rate": 5.761314571803761e-06,
+      "loss": 0.5011,
+      "step": 3647
+    },
+    {
+      "epoch": 1.810227131686732,
+      "grad_norm": 0.07242839670144252,
+      "learning_rate": 5.759382606158423e-06,
+      "loss": 0.478,
+      "step": 3648
+    },
+    {
+      "epoch": 1.8107235943899713,
+      "grad_norm": 0.07593439548121685,
+      "learning_rate": 5.757450524453632e-06,
+      "loss": 0.497,
+      "step": 3649
+    },
+    {
+      "epoch": 1.8112200570932109,
+      "grad_norm": 0.07167454149194932,
+      "learning_rate": 5.755518326984671e-06,
+      "loss": 0.4845,
+      "step": 3650
+    },
+    {
+      "epoch": 1.8117165197964504,
+      "grad_norm": 0.07071788965467532,
+      "learning_rate": 5.753586014046847e-06,
+      "loss": 0.4751,
+      "step": 3651
+    },
+    {
+      "epoch": 1.8122129824996898,
+      "grad_norm": 0.07152437833377714,
+      "learning_rate": 5.7516535859354835e-06,
+      "loss": 0.515,
+      "step": 3652
+    },
+    {
+      "epoch": 1.812709445202929,
+      "grad_norm": 0.07164234780108387,
+      "learning_rate": 5.749721042945924e-06,
+      "loss": 0.4768,
+      "step": 3653
+    },
+    {
+      "epoch": 1.8132059079061684,
+      "grad_norm": 0.06924108464675693,
+      "learning_rate": 5.747788385373522e-06,
+      "loss": 0.4778,
+      "step": 3654
+    },
+    {
+      "epoch": 1.813702370609408,
+      "grad_norm": 0.07367629715089329,
+      "learning_rate": 5.7458556135136545e-06,
+      "loss": 0.4942,
+      "step": 3655
+    },
+    {
+      "epoch": 1.8141988333126475,
+      "grad_norm": 0.07021653950442595,
+      "learning_rate": 5.743922727661716e-06,
+      "loss": 0.5148,
+      "step": 3656
+    },
+    {
+      "epoch": 1.8146952960158869,
+      "grad_norm": 0.07417466927168129,
+      "learning_rate": 5.7419897281131164e-06,
+      "loss": 0.5172,
+      "step": 3657
+    },
+    {
+      "epoch": 1.8151917587191262,
+      "grad_norm": 0.0678717013795942,
+      "learning_rate": 5.740056615163284e-06,
+      "loss": 0.4566,
+      "step": 3658
+    },
+    {
+      "epoch": 1.8156882214223655,
+      "grad_norm": 0.06873133916950154,
+      "learning_rate": 5.738123389107665e-06,
+      "loss": 0.4481,
+      "step": 3659
+    },
+    {
+      "epoch": 1.816184684125605,
+      "grad_norm": 0.07135619019377491,
+      "learning_rate": 5.736190050241719e-06,
+      "loss": 0.4767,
+      "step": 3660
+    },
+    {
+      "epoch": 1.8166811468288446,
+      "grad_norm": 0.0728194684169538,
+      "learning_rate": 5.7342565988609275e-06,
+      "loss": 0.4569,
+      "step": 3661
+    },
+    {
+      "epoch": 1.817177609532084,
+      "grad_norm": 0.07250024973531223,
+      "learning_rate": 5.732323035260789e-06,
+      "loss": 0.4853,
+      "step": 3662
+    },
+    {
+      "epoch": 1.8176740722353233,
+      "grad_norm": 0.06919852772378363,
+      "learning_rate": 5.730389359736816e-06,
+      "loss": 0.493,
+      "step": 3663
+    },
+    {
+      "epoch": 1.8181705349385626,
+      "grad_norm": 0.07065617648477643,
+      "learning_rate": 5.7284555725845405e-06,
+      "loss": 0.4756,
+      "step": 3664
+    },
+    {
+      "epoch": 1.8186669976418022,
+      "grad_norm": 0.07232159318569668,
+      "learning_rate": 5.726521674099511e-06,
+      "loss": 0.5141,
+      "step": 3665
+    },
+    {
+      "epoch": 1.8191634603450417,
+      "grad_norm": 0.07221720276118149,
+      "learning_rate": 5.724587664577292e-06,
+      "loss": 0.4652,
+      "step": 3666
+    },
+    {
+      "epoch": 1.819659923048281,
+      "grad_norm": 0.07053346570081506,
+      "learning_rate": 5.722653544313467e-06,
+      "loss": 0.5112,
+      "step": 3667
+    },
+    {
+      "epoch": 1.8201563857515204,
+      "grad_norm": 0.07023432041629968,
+      "learning_rate": 5.720719313603633e-06,
+      "loss": 0.4875,
+      "step": 3668
+    },
+    {
+      "epoch": 1.8206528484547597,
+      "grad_norm": 0.07095615034595218,
+      "learning_rate": 5.71878497274341e-06,
+      "loss": 0.4999,
+      "step": 3669
+    },
+    {
+      "epoch": 1.8211493111579993,
+      "grad_norm": 0.07274255009602926,
+      "learning_rate": 5.7168505220284266e-06,
+      "loss": 0.507,
+      "step": 3670
+    },
+    {
+      "epoch": 1.8216457738612388,
+      "grad_norm": 0.07247280332754438,
+      "learning_rate": 5.714915961754335e-06,
+      "loss": 0.4726,
+      "step": 3671
+    },
+    {
+      "epoch": 1.8221422365644782,
+      "grad_norm": 0.06954349998822817,
+      "learning_rate": 5.712981292216803e-06,
+      "loss": 0.4388,
+      "step": 3672
+    },
+    {
+      "epoch": 1.8226386992677175,
+      "grad_norm": 0.07008709511894415,
+      "learning_rate": 5.711046513711512e-06,
+      "loss": 0.4896,
+      "step": 3673
+    },
+    {
+      "epoch": 1.8231351619709568,
+      "grad_norm": 0.07146176222719076,
+      "learning_rate": 5.709111626534161e-06,
+      "loss": 0.471,
+      "step": 3674
+    },
+    {
+      "epoch": 1.8236316246741964,
+      "grad_norm": 0.06966762407120558,
+      "learning_rate": 5.707176630980469e-06,
+      "loss": 0.4756,
+      "step": 3675
+    },
+    {
+      "epoch": 1.824128087377436,
+      "grad_norm": 0.06975242002832147,
+      "learning_rate": 5.705241527346166e-06,
+      "loss": 0.4964,
+      "step": 3676
+    },
+    {
+      "epoch": 1.8246245500806753,
+      "grad_norm": 0.07114687927174249,
+      "learning_rate": 5.703306315927004e-06,
+      "loss": 0.4552,
+      "step": 3677
+    },
+    {
+      "epoch": 1.8251210127839146,
+      "grad_norm": 0.07157645828419164,
+      "learning_rate": 5.701370997018748e-06,
+      "loss": 0.5142,
+      "step": 3678
+    },
+    {
+      "epoch": 1.825617475487154,
+      "grad_norm": 0.06880701494843049,
+      "learning_rate": 5.69943557091718e-06,
+      "loss": 0.4483,
+      "step": 3679
+    },
+    {
+      "epoch": 1.8261139381903935,
+      "grad_norm": 0.07089399642352692,
+      "learning_rate": 5.6975000379181025e-06,
+      "loss": 0.4823,
+      "step": 3680
+    },
+    {
+      "epoch": 1.826610400893633,
+      "grad_norm": 0.07140090532891165,
+      "learning_rate": 5.695564398317326e-06,
+      "loss": 0.4902,
+      "step": 3681
+    },
+    {
+      "epoch": 1.8271068635968724,
+      "grad_norm": 0.07649533474683105,
+      "learning_rate": 5.693628652410683e-06,
+      "loss": 0.49,
+      "step": 3682
+    },
+    {
+      "epoch": 1.8276033263001117,
+      "grad_norm": 0.07226480459106793,
+      "learning_rate": 5.691692800494023e-06,
+      "loss": 0.5069,
+      "step": 3683
+    },
+    {
+      "epoch": 1.828099789003351,
+      "grad_norm": 0.07464166753150865,
+      "learning_rate": 5.689756842863208e-06,
+      "loss": 0.4629,
+      "step": 3684
+    },
+    {
+      "epoch": 1.8285962517065906,
+      "grad_norm": 0.07013676743303268,
+      "learning_rate": 5.687820779814119e-06,
+      "loss": 0.4533,
+      "step": 3685
+    },
+    {
+      "epoch": 1.8290927144098301,
+      "grad_norm": 0.07186995524898125,
+      "learning_rate": 5.6858846116426535e-06,
+      "loss": 0.4833,
+      "step": 3686
+    },
+    {
+      "epoch": 1.8295891771130695,
+      "grad_norm": 0.07153345606512074,
+      "learning_rate": 5.683948338644721e-06,
+      "loss": 0.4836,
+      "step": 3687
+    },
+    {
+      "epoch": 1.8300856398163088,
+      "grad_norm": 0.07438658965655393,
+      "learning_rate": 5.6820119611162515e-06,
+      "loss": 0.511,
+      "step": 3688
+    },
+    {
+      "epoch": 1.8305821025195481,
+      "grad_norm": 0.07184279444597692,
+      "learning_rate": 5.68007547935319e-06,
+      "loss": 0.4793,
+      "step": 3689
+    },
+    {
+      "epoch": 1.8310785652227877,
+      "grad_norm": 0.0723669994788689,
+      "learning_rate": 5.678138893651495e-06,
+      "loss": 0.5094,
+      "step": 3690
+    },
+    {
+      "epoch": 1.8315750279260272,
+      "grad_norm": 0.07299871934038199,
+      "learning_rate": 5.676202204307144e-06,
+      "loss": 0.4871,
+      "step": 3691
+    },
+    {
+      "epoch": 1.8320714906292666,
+      "grad_norm": 0.06980833513308857,
+      "learning_rate": 5.674265411616127e-06,
+      "loss": 0.4917,
+      "step": 3692
+    },
+    {
+      "epoch": 1.8325679533325059,
+      "grad_norm": 0.07292298451496881,
+      "learning_rate": 5.672328515874452e-06,
+      "loss": 0.4529,
+      "step": 3693
+    },
+    {
+      "epoch": 1.8330644160357452,
+      "grad_norm": 0.07138317703404169,
+      "learning_rate": 5.670391517378145e-06,
+      "loss": 0.5169,
+      "step": 3694
+    },
+    {
+      "epoch": 1.8335608787389848,
+      "grad_norm": 0.07162547405202699,
+      "learning_rate": 5.668454416423243e-06,
+      "loss": 0.4893,
+      "step": 3695
+    },
+    {
+      "epoch": 1.834057341442224,
+      "grad_norm": 0.07316600393252785,
+      "learning_rate": 5.666517213305802e-06,
+      "loss": 0.505,
+      "step": 3696
+    },
+    {
+      "epoch": 1.8345538041454637,
+      "grad_norm": 0.06988135104934977,
+      "learning_rate": 5.6645799083218915e-06,
+      "loss": 0.5128,
+      "step": 3697
+    },
+    {
+      "epoch": 1.835050266848703,
+      "grad_norm": 0.07303979907617231,
+      "learning_rate": 5.662642501767597e-06,
+      "loss": 0.4678,
+      "step": 3698
+    },
+    {
+      "epoch": 1.8355467295519423,
+      "grad_norm": 0.075149991058931,
+      "learning_rate": 5.66070499393902e-06,
+      "loss": 0.4879,
+      "step": 3699
+    },
+    {
+      "epoch": 1.8360431922551819,
+      "grad_norm": 0.07133476017601827,
+      "learning_rate": 5.65876738513228e-06,
+      "loss": 0.4976,
+      "step": 3700
+    },
+    {
+      "epoch": 1.8365396549584212,
+      "grad_norm": 0.07100746394861222,
+      "learning_rate": 5.656829675643506e-06,
+      "loss": 0.481,
+      "step": 3701
+    },
+    {
+      "epoch": 1.8370361176616608,
+      "grad_norm": 0.06893065620889177,
+      "learning_rate": 5.65489186576885e-06,
+      "loss": 0.4677,
+      "step": 3702
+    },
+    {
+      "epoch": 1.8375325803649,
+      "grad_norm": 0.07043388719856646,
+      "learning_rate": 5.652953955804471e-06,
+      "loss": 0.486,
+      "step": 3703
+    },
+    {
+      "epoch": 1.8380290430681394,
+      "grad_norm": 0.07186042040863846,
+      "learning_rate": 5.6510159460465485e-06,
+      "loss": 0.5171,
+      "step": 3704
+    },
+    {
+      "epoch": 1.838525505771379,
+      "grad_norm": 0.06942013042023604,
+      "learning_rate": 5.649077836791279e-06,
+      "loss": 0.4661,
+      "step": 3705
+    },
+    {
+      "epoch": 1.8390219684746183,
+      "grad_norm": 0.07320920107605887,
+      "learning_rate": 5.6471396283348676e-06,
+      "loss": 0.5008,
+      "step": 3706
+    },
+    {
+      "epoch": 1.8395184311778578,
+      "grad_norm": 0.06922693618002337,
+      "learning_rate": 5.645201320973541e-06,
+      "loss": 0.475,
+      "step": 3707
+    },
+    {
+      "epoch": 1.8400148938810972,
+      "grad_norm": 0.07336728917335142,
+      "learning_rate": 5.643262915003538e-06,
+      "loss": 0.5215,
+      "step": 3708
+    },
+    {
+      "epoch": 1.8405113565843365,
+      "grad_norm": 0.0716664034498282,
+      "learning_rate": 5.64132441072111e-06,
+      "loss": 0.5243,
+      "step": 3709
+    },
+    {
+      "epoch": 1.841007819287576,
+      "grad_norm": 0.07202044980774482,
+      "learning_rate": 5.6393858084225305e-06,
+      "loss": 0.4886,
+      "step": 3710
+    },
+    {
+      "epoch": 1.8415042819908154,
+      "grad_norm": 0.07266065385472024,
+      "learning_rate": 5.637447108404082e-06,
+      "loss": 0.4862,
+      "step": 3711
+    },
+    {
+      "epoch": 1.842000744694055,
+      "grad_norm": 0.0706724033050316,
+      "learning_rate": 5.635508310962064e-06,
+      "loss": 0.483,
+      "step": 3712
+    },
+    {
+      "epoch": 1.8424972073972943,
+      "grad_norm": 0.07233950101249319,
+      "learning_rate": 5.63356941639279e-06,
+      "loss": 0.4836,
+      "step": 3713
+    },
+    {
+      "epoch": 1.8429936701005336,
+      "grad_norm": 0.07601798497496463,
+      "learning_rate": 5.631630424992588e-06,
+      "loss": 0.5224,
+      "step": 3714
+    },
+    {
+      "epoch": 1.8434901328037732,
+      "grad_norm": 0.07146589515735953,
+      "learning_rate": 5.629691337057803e-06,
+      "loss": 0.5011,
+      "step": 3715
+    },
+    {
+      "epoch": 1.8439865955070125,
+      "grad_norm": 0.0700451438479701,
+      "learning_rate": 5.627752152884794e-06,
+      "loss": 0.4834,
+      "step": 3716
+    },
+    {
+      "epoch": 1.844483058210252,
+      "grad_norm": 0.06743704292605564,
+      "learning_rate": 5.625812872769935e-06,
+      "loss": 0.4562,
+      "step": 3717
+    },
+    {
+      "epoch": 1.8449795209134914,
+      "grad_norm": 0.06997430470081448,
+      "learning_rate": 5.623873497009612e-06,
+      "loss": 0.5186,
+      "step": 3718
+    },
+    {
+      "epoch": 1.8454759836167307,
+      "grad_norm": 0.07339075549752148,
+      "learning_rate": 5.621934025900226e-06,
+      "loss": 0.4918,
+      "step": 3719
+    },
+    {
+      "epoch": 1.8459724463199703,
+      "grad_norm": 0.0710906084904975,
+      "learning_rate": 5.619994459738198e-06,
+      "loss": 0.4829,
+      "step": 3720
+    },
+    {
+      "epoch": 1.8464689090232096,
+      "grad_norm": 0.07016695505135577,
+      "learning_rate": 5.6180547988199586e-06,
+      "loss": 0.4903,
+      "step": 3721
+    },
+    {
+      "epoch": 1.8469653717264491,
+      "grad_norm": 0.07234070412472808,
+      "learning_rate": 5.616115043441951e-06,
+      "loss": 0.478,
+      "step": 3722
+    },
+    {
+      "epoch": 1.8474618344296885,
+      "grad_norm": 0.07055140091964066,
+      "learning_rate": 5.614175193900639e-06,
+      "loss": 0.4875,
+      "step": 3723
+    },
+    {
+      "epoch": 1.8479582971329278,
+      "grad_norm": 0.0765106043318054,
+      "learning_rate": 5.612235250492495e-06,
+      "loss": 0.4839,
+      "step": 3724
+    },
+    {
+      "epoch": 1.8484547598361674,
+      "grad_norm": 0.07073929439268867,
+      "learning_rate": 5.61029521351401e-06,
+      "loss": 0.4836,
+      "step": 3725
+    },
+    {
+      "epoch": 1.8489512225394067,
+      "grad_norm": 0.07189512081581409,
+      "learning_rate": 5.608355083261686e-06,
+      "loss": 0.4724,
+      "step": 3726
+    },
+    {
+      "epoch": 1.8494476852426462,
+      "grad_norm": 0.07079012203165752,
+      "learning_rate": 5.606414860032042e-06,
+      "loss": 0.512,
+      "step": 3727
+    },
+    {
+      "epoch": 1.8499441479458856,
+      "grad_norm": 0.07272943202601717,
+      "learning_rate": 5.604474544121612e-06,
+      "loss": 0.4789,
+      "step": 3728
+    },
+    {
+      "epoch": 1.850440610649125,
+      "grad_norm": 0.06777523021078664,
+      "learning_rate": 5.602534135826939e-06,
+      "loss": 0.469,
+      "step": 3729
+    },
+    {
+      "epoch": 1.8509370733523642,
+      "grad_norm": 0.073353001049987,
+      "learning_rate": 5.600593635444583e-06,
+      "loss": 0.4993,
+      "step": 3730
+    },
+    {
+      "epoch": 1.8514335360556038,
+      "grad_norm": 0.07248977863476927,
+      "learning_rate": 5.5986530432711195e-06,
+      "loss": 0.4717,
+      "step": 3731
+    },
+    {
+      "epoch": 1.8519299987588433,
+      "grad_norm": 0.07234349061832071,
+      "learning_rate": 5.596712359603138e-06,
+      "loss": 0.4963,
+      "step": 3732
+    },
+    {
+      "epoch": 1.8524264614620827,
+      "grad_norm": 0.06920179818135419,
+      "learning_rate": 5.5947715847372385e-06,
+      "loss": 0.4582,
+      "step": 3733
+    },
+    {
+      "epoch": 1.852922924165322,
+      "grad_norm": 0.07187699072861928,
+      "learning_rate": 5.5928307189700415e-06,
+      "loss": 0.4571,
+      "step": 3734
+    },
+    {
+      "epoch": 1.8534193868685613,
+      "grad_norm": 0.07080879521086189,
+      "learning_rate": 5.590889762598171e-06,
+      "loss": 0.5281,
+      "step": 3735
+    },
+    {
+      "epoch": 1.853915849571801,
+      "grad_norm": 0.07195862744667644,
+      "learning_rate": 5.588948715918277e-06,
+      "loss": 0.527,
+      "step": 3736
+    },
+    {
+      "epoch": 1.8544123122750404,
+      "grad_norm": 0.07215211385422904,
+      "learning_rate": 5.587007579227014e-06,
+      "loss": 0.5411,
+      "step": 3737
+    },
+    {
+      "epoch": 1.8549087749782798,
+      "grad_norm": 0.07117465488308353,
+      "learning_rate": 5.5850663528210545e-06,
+      "loss": 0.4886,
+      "step": 3738
+    },
+    {
+      "epoch": 1.855405237681519,
+      "grad_norm": 0.07165570397884258,
+      "learning_rate": 5.583125036997084e-06,
+      "loss": 0.4918,
+      "step": 3739
+    },
+    {
+      "epoch": 1.8559017003847584,
+      "grad_norm": 0.06806064422601031,
+      "learning_rate": 5.581183632051801e-06,
+      "loss": 0.4638,
+      "step": 3740
+    },
+    {
+      "epoch": 1.856398163087998,
+      "grad_norm": 0.06938300711407791,
+      "learning_rate": 5.579242138281918e-06,
+      "loss": 0.488,
+      "step": 3741
+    },
+    {
+      "epoch": 1.8568946257912375,
+      "grad_norm": 0.07263388401977534,
+      "learning_rate": 5.577300555984162e-06,
+      "loss": 0.5041,
+      "step": 3742
+    },
+    {
+      "epoch": 1.8573910884944769,
+      "grad_norm": 0.07291712802999431,
+      "learning_rate": 5.5753588854552724e-06,
+      "loss": 0.4973,
+      "step": 3743
+    },
+    {
+      "epoch": 1.8578875511977162,
+      "grad_norm": 0.06984653551222911,
+      "learning_rate": 5.573417126992004e-06,
+      "loss": 0.4702,
+      "step": 3744
+    },
+    {
+      "epoch": 1.8583840139009555,
+      "grad_norm": 0.07593350841767885,
+      "learning_rate": 5.57147528089112e-06,
+      "loss": 0.4966,
+      "step": 3745
+    },
+    {
+      "epoch": 1.858880476604195,
+      "grad_norm": 0.07160799166246536,
+      "learning_rate": 5.5695333474494015e-06,
+      "loss": 0.4811,
+      "step": 3746
+    },
+    {
+      "epoch": 1.8593769393074346,
+      "grad_norm": 0.06904328199683979,
+      "learning_rate": 5.567591326963644e-06,
+      "loss": 0.4753,
+      "step": 3747
+    },
+    {
+      "epoch": 1.859873402010674,
+      "grad_norm": 0.06929606497626255,
+      "learning_rate": 5.565649219730651e-06,
+      "loss": 0.5292,
+      "step": 3748
+    },
+    {
+      "epoch": 1.8603698647139133,
+      "grad_norm": 0.07444165806160696,
+      "learning_rate": 5.563707026047246e-06,
+      "loss": 0.4787,
+      "step": 3749
+    },
+    {
+      "epoch": 1.8608663274171526,
+      "grad_norm": 0.06887681183296859,
+      "learning_rate": 5.561764746210261e-06,
+      "loss": 0.4982,
+      "step": 3750
+    },
+    {
+      "epoch": 1.8613627901203922,
+      "grad_norm": 0.07082675162940927,
+      "learning_rate": 5.559822380516539e-06,
+      "loss": 0.4702,
+      "step": 3751
+    },
+    {
+      "epoch": 1.8618592528236317,
+      "grad_norm": 0.07144978926454551,
+      "learning_rate": 5.5578799292629446e-06,
+      "loss": 0.501,
+      "step": 3752
+    },
+    {
+      "epoch": 1.862355715526871,
+      "grad_norm": 0.07049843632887456,
+      "learning_rate": 5.5559373927463476e-06,
+      "loss": 0.5032,
+      "step": 3753
+    },
+    {
+      "epoch": 1.8628521782301104,
+      "grad_norm": 0.07256530112167905,
+      "learning_rate": 5.553994771263633e-06,
+      "loss": 0.4928,
+      "step": 3754
+    },
+    {
+      "epoch": 1.8633486409333497,
+      "grad_norm": 0.07088436352644081,
+      "learning_rate": 5.5520520651117014e-06,
+      "loss": 0.4669,
+      "step": 3755
+    },
+    {
+      "epoch": 1.8638451036365893,
+      "grad_norm": 0.07330689752559463,
+      "learning_rate": 5.550109274587463e-06,
+      "loss": 0.4836,
+      "step": 3756
+    },
+    {
+      "epoch": 1.8643415663398288,
+      "grad_norm": 0.07300119259335101,
+      "learning_rate": 5.548166399987842e-06,
+      "loss": 0.5022,
+      "step": 3757
+    },
+    {
+      "epoch": 1.8648380290430682,
+      "grad_norm": 0.07139240282413961,
+      "learning_rate": 5.546223441609775e-06,
+      "loss": 0.45,
+      "step": 3758
+    },
+    {
+      "epoch": 1.8653344917463075,
+      "grad_norm": 0.07306394150942447,
+      "learning_rate": 5.544280399750214e-06,
+      "loss": 0.4938,
+      "step": 3759
+    },
+    {
+      "epoch": 1.8658309544495468,
+      "grad_norm": 0.07654037658961671,
+      "learning_rate": 5.54233727470612e-06,
+      "loss": 0.4927,
+      "step": 3760
+    },
+    {
+      "epoch": 1.8663274171527864,
+      "grad_norm": 0.0723075236426715,
+      "learning_rate": 5.540394066774471e-06,
+      "loss": 0.4759,
+      "step": 3761
+    },
+    {
+      "epoch": 1.866823879856026,
+      "grad_norm": 0.07004287361289634,
+      "learning_rate": 5.538450776252252e-06,
+      "loss": 0.4718,
+      "step": 3762
+    },
+    {
+      "epoch": 1.8673203425592653,
+      "grad_norm": 0.07222653031062394,
+      "learning_rate": 5.536507403436465e-06,
+      "loss": 0.4792,
+      "step": 3763
+    },
+    {
+      "epoch": 1.8678168052625046,
+      "grad_norm": 0.07278259090790781,
+      "learning_rate": 5.534563948624124e-06,
+      "loss": 0.4775,
+      "step": 3764
+    },
+    {
+      "epoch": 1.868313267965744,
+      "grad_norm": 0.07010260861928222,
+      "learning_rate": 5.532620412112255e-06,
+      "loss": 0.483,
+      "step": 3765
+    },
+    {
+      "epoch": 1.8688097306689835,
+      "grad_norm": 0.07351482506516754,
+      "learning_rate": 5.530676794197895e-06,
+      "loss": 0.4741,
+      "step": 3766
+    },
+    {
+      "epoch": 1.869306193372223,
+      "grad_norm": 0.07191108493092994,
+      "learning_rate": 5.528733095178097e-06,
+      "loss": 0.4639,
+      "step": 3767
+    },
+    {
+      "epoch": 1.8698026560754624,
+      "grad_norm": 0.07046867812955843,
+      "learning_rate": 5.526789315349922e-06,
+      "loss": 0.474,
+      "step": 3768
+    },
+    {
+      "epoch": 1.8702991187787017,
+      "grad_norm": 0.07249463782600657,
+      "learning_rate": 5.524845455010448e-06,
+      "loss": 0.5159,
+      "step": 3769
+    },
+    {
+      "epoch": 1.870795581481941,
+      "grad_norm": 0.07291403318196778,
+      "learning_rate": 5.52290151445676e-06,
+      "loss": 0.4721,
+      "step": 3770
+    },
+    {
+      "epoch": 1.8712920441851806,
+      "grad_norm": 0.07363004384944376,
+      "learning_rate": 5.52095749398596e-06,
+      "loss": 0.5124,
+      "step": 3771
+    },
+    {
+      "epoch": 1.8717885068884201,
+      "grad_norm": 0.07402791452077123,
+      "learning_rate": 5.51901339389516e-06,
+      "loss": 0.4947,
+      "step": 3772
+    },
+    {
+      "epoch": 1.8722849695916595,
+      "grad_norm": 0.07386850215254852,
+      "learning_rate": 5.5170692144814844e-06,
+      "loss": 0.5007,
+      "step": 3773
+    },
+    {
+      "epoch": 1.8727814322948988,
+      "grad_norm": 0.07282838700658707,
+      "learning_rate": 5.51512495604207e-06,
+      "loss": 0.4936,
+      "step": 3774
+    },
+    {
+      "epoch": 1.8732778949981381,
+      "grad_norm": 0.070148810597566,
+      "learning_rate": 5.513180618874066e-06,
+      "loss": 0.4793,
+      "step": 3775
+    },
+    {
+      "epoch": 1.8737743577013777,
+      "grad_norm": 0.07261217534361884,
+      "learning_rate": 5.51123620327463e-06,
+      "loss": 0.4862,
+      "step": 3776
+    },
+    {
+      "epoch": 1.8742708204046172,
+      "grad_norm": 0.06946317885723177,
+      "learning_rate": 5.509291709540942e-06,
+      "loss": 0.497,
+      "step": 3777
+    },
+    {
+      "epoch": 1.8747672831078566,
+      "grad_norm": 0.06961836136385555,
+      "learning_rate": 5.50734713797018e-06,
+      "loss": 0.4604,
+      "step": 3778
+    },
+    {
+      "epoch": 1.875263745811096,
+      "grad_norm": 0.07178654847839235,
+      "learning_rate": 5.5054024888595415e-06,
+      "loss": 0.4991,
+      "step": 3779
+    },
+    {
+      "epoch": 1.8757602085143352,
+      "grad_norm": 0.07408092919562587,
+      "learning_rate": 5.503457762506236e-06,
+      "loss": 0.5154,
+      "step": 3780
+    },
+    {
+      "epoch": 1.8762566712175748,
+      "grad_norm": 0.07588664418284227,
+      "learning_rate": 5.501512959207485e-06,
+      "loss": 0.5019,
+      "step": 3781
+    },
+    {
+      "epoch": 1.8767531339208143,
+      "grad_norm": 0.07277845131973065,
+      "learning_rate": 5.499568079260519e-06,
+      "loss": 0.5079,
+      "step": 3782
+    },
+    {
+      "epoch": 1.8772495966240537,
+      "grad_norm": 0.0754058108393957,
+      "learning_rate": 5.497623122962582e-06,
+      "loss": 0.5014,
+      "step": 3783
+    },
+    {
+      "epoch": 1.877746059327293,
+      "grad_norm": 0.07288657771449725,
+      "learning_rate": 5.495678090610929e-06,
+      "loss": 0.4912,
+      "step": 3784
+    },
+    {
+      "epoch": 1.8782425220305323,
+      "grad_norm": 0.07547192333018429,
+      "learning_rate": 5.493732982502828e-06,
+      "loss": 0.5102,
+      "step": 3785
+    },
+    {
+      "epoch": 1.8787389847337719,
+      "grad_norm": 0.07101553044945602,
+      "learning_rate": 5.491787798935557e-06,
+      "loss": 0.516,
+      "step": 3786
+    },
+    {
+      "epoch": 1.8792354474370114,
+      "grad_norm": 0.0755631628504027,
+      "learning_rate": 5.489842540206406e-06,
+      "loss": 0.5123,
+      "step": 3787
+    },
+    {
+      "epoch": 1.8797319101402508,
+      "grad_norm": 0.06839328149157925,
+      "learning_rate": 5.487897206612678e-06,
+      "loss": 0.4766,
+      "step": 3788
+    },
+    {
+      "epoch": 1.88022837284349,
+      "grad_norm": 0.07225884897896602,
+      "learning_rate": 5.485951798451683e-06,
+      "loss": 0.4921,
+      "step": 3789
+    },
+    {
+      "epoch": 1.8807248355467294,
+      "grad_norm": 0.07247762275978963,
+      "learning_rate": 5.484006316020747e-06,
+      "loss": 0.4911,
+      "step": 3790
+    },
+    {
+      "epoch": 1.881221298249969,
+      "grad_norm": 0.07250223736563732,
+      "learning_rate": 5.482060759617207e-06,
+      "loss": 0.4837,
+      "step": 3791
+    },
+    {
+      "epoch": 1.8817177609532085,
+      "grad_norm": 0.07185564268743984,
+      "learning_rate": 5.480115129538409e-06,
+      "loss": 0.5248,
+      "step": 3792
+    },
+    {
+      "epoch": 1.8822142236564479,
+      "grad_norm": 0.0713794358085212,
+      "learning_rate": 5.478169426081712e-06,
+      "loss": 0.4907,
+      "step": 3793
+    },
+    {
+      "epoch": 1.8827106863596872,
+      "grad_norm": 0.06919626880513847,
+      "learning_rate": 5.476223649544485e-06,
+      "loss": 0.4611,
+      "step": 3794
+    },
+    {
+      "epoch": 1.8832071490629265,
+      "grad_norm": 0.07072586859293258,
+      "learning_rate": 5.474277800224109e-06,
+      "loss": 0.4829,
+      "step": 3795
+    },
+    {
+      "epoch": 1.883703611766166,
+      "grad_norm": 0.06988680449488685,
+      "learning_rate": 5.472331878417974e-06,
+      "loss": 0.4537,
+      "step": 3796
+    },
+    {
+      "epoch": 1.8842000744694056,
+      "grad_norm": 0.06903359042318398,
+      "learning_rate": 5.470385884423486e-06,
+      "loss": 0.5165,
+      "step": 3797
+    },
+    {
+      "epoch": 1.884696537172645,
+      "grad_norm": 0.07130170675257452,
+      "learning_rate": 5.468439818538057e-06,
+      "loss": 0.4838,
+      "step": 3798
+    },
+    {
+      "epoch": 1.8851929998758843,
+      "grad_norm": 0.07051483223367049,
+      "learning_rate": 5.466493681059114e-06,
+      "loss": 0.4792,
+      "step": 3799
+    },
+    {
+      "epoch": 1.8856894625791236,
+      "grad_norm": 0.06919703127980216,
+      "learning_rate": 5.464547472284091e-06,
+      "loss": 0.458,
+      "step": 3800
+    },
+    {
+      "epoch": 1.8861859252823632,
+      "grad_norm": 0.07022779374561813,
+      "learning_rate": 5.462601192510435e-06,
+      "loss": 0.5066,
+      "step": 3801
+    },
+    {
+      "epoch": 1.8866823879856027,
+      "grad_norm": 0.07212467257348683,
+      "learning_rate": 5.4606548420356046e-06,
+      "loss": 0.4953,
+      "step": 3802
+    },
+    {
+      "epoch": 1.887178850688842,
+      "grad_norm": 0.07085298955939363,
+      "learning_rate": 5.458708421157066e-06,
+      "loss": 0.5051,
+      "step": 3803
+    },
+    {
+      "epoch": 1.8876753133920814,
+      "grad_norm": 0.07090198392557463,
+      "learning_rate": 5.4567619301723015e-06,
+      "loss": 0.4989,
+      "step": 3804
+    },
+    {
+      "epoch": 1.8881717760953207,
+      "grad_norm": 0.07218297703303209,
+      "learning_rate": 5.454815369378798e-06,
+      "loss": 0.5297,
+      "step": 3805
+    },
+    {
+      "epoch": 1.8886682387985603,
+      "grad_norm": 0.07024232525534924,
+      "learning_rate": 5.452868739074059e-06,
+      "loss": 0.5068,
+      "step": 3806
+    },
+    {
+      "epoch": 1.8891647015017998,
+      "grad_norm": 0.06920823610833787,
+      "learning_rate": 5.450922039555594e-06,
+      "loss": 0.4614,
+      "step": 3807
+    },
+    {
+      "epoch": 1.8896611642050392,
+      "grad_norm": 0.07763806617888123,
+      "learning_rate": 5.448975271120925e-06,
+      "loss": 0.4939,
+      "step": 3808
+    },
+    {
+      "epoch": 1.8901576269082785,
+      "grad_norm": 0.07052981043474175,
+      "learning_rate": 5.447028434067586e-06,
+      "loss": 0.5062,
+      "step": 3809
+    },
+    {
+      "epoch": 1.8906540896115178,
+      "grad_norm": 0.07014190175993225,
+      "learning_rate": 5.445081528693118e-06,
+      "loss": 0.5001,
+      "step": 3810
+    },
+    {
+      "epoch": 1.8911505523147574,
+      "grad_norm": 0.07208974706034772,
+      "learning_rate": 5.443134555295075e-06,
+      "loss": 0.5052,
+      "step": 3811
+    },
+    {
+      "epoch": 1.891647015017997,
+      "grad_norm": 0.07343737439730191,
+      "learning_rate": 5.441187514171018e-06,
+      "loss": 0.5319,
+      "step": 3812
+    },
+    {
+      "epoch": 1.8921434777212363,
+      "grad_norm": 0.06983822109409202,
+      "learning_rate": 5.439240405618524e-06,
+      "loss": 0.5003,
+      "step": 3813
+    },
+    {
+      "epoch": 1.8926399404244756,
+      "grad_norm": 0.07100147800127123,
+      "learning_rate": 5.437293229935178e-06,
+      "loss": 0.4855,
+      "step": 3814
+    },
+    {
+      "epoch": 1.893136403127715,
+      "grad_norm": 0.06982351812557983,
+      "learning_rate": 5.4353459874185735e-06,
+      "loss": 0.4685,
+      "step": 3815
+    },
+    {
+      "epoch": 1.8936328658309545,
+      "grad_norm": 0.07161346724602752,
+      "learning_rate": 5.433398678366314e-06,
+      "loss": 0.451,
+      "step": 3816
+    },
+    {
+      "epoch": 1.894129328534194,
+      "grad_norm": 0.07198803784772664,
+      "learning_rate": 5.431451303076015e-06,
+      "loss": 0.5078,
+      "step": 3817
+    },
+    {
+      "epoch": 1.8946257912374334,
+      "grad_norm": 0.07011876526269005,
+      "learning_rate": 5.429503861845305e-06,
+      "loss": 0.4891,
+      "step": 3818
+    },
+    {
+      "epoch": 1.8951222539406727,
+      "grad_norm": 0.07097934662924628,
+      "learning_rate": 5.427556354971812e-06,
+      "loss": 0.4812,
+      "step": 3819
+    },
+    {
+      "epoch": 1.895618716643912,
+      "grad_norm": 0.07026743873520488,
+      "learning_rate": 5.425608782753188e-06,
+      "loss": 0.5061,
+      "step": 3820
+    },
+    {
+      "epoch": 1.8961151793471516,
+      "grad_norm": 0.07349726995712234,
+      "learning_rate": 5.4236611454870865e-06,
+      "loss": 0.4738,
+      "step": 3821
+    },
+    {
+      "epoch": 1.8966116420503911,
+      "grad_norm": 0.07126025580824183,
+      "learning_rate": 5.42171344347117e-06,
+      "loss": 0.4577,
+      "step": 3822
+    },
+    {
+      "epoch": 1.8971081047536305,
+      "grad_norm": 0.06904179244008007,
+      "learning_rate": 5.419765677003116e-06,
+      "loss": 0.4556,
+      "step": 3823
+    },
+    {
+      "epoch": 1.8976045674568698,
+      "grad_norm": 0.0740816990577368,
+      "learning_rate": 5.417817846380609e-06,
+      "loss": 0.4814,
+      "step": 3824
+    },
+    {
+      "epoch": 1.8981010301601091,
+      "grad_norm": 0.0730309332342059,
+      "learning_rate": 5.415869951901344e-06,
+      "loss": 0.4987,
+      "step": 3825
+    },
+    {
+      "epoch": 1.8985974928633487,
+      "grad_norm": 0.07214431782979727,
+      "learning_rate": 5.413921993863024e-06,
+      "loss": 0.4878,
+      "step": 3826
+    },
+    {
+      "epoch": 1.8990939555665882,
+      "grad_norm": 0.07279291804221866,
+      "learning_rate": 5.411973972563363e-06,
+      "loss": 0.485,
+      "step": 3827
+    },
+    {
+      "epoch": 1.8995904182698276,
+      "grad_norm": 0.07274491253000365,
+      "learning_rate": 5.4100258883000874e-06,
+      "loss": 0.4864,
+      "step": 3828
+    },
+    {
+      "epoch": 1.9000868809730669,
+      "grad_norm": 0.07047067985378323,
+      "learning_rate": 5.408077741370927e-06,
+      "loss": 0.4782,
+      "step": 3829
+    },
+    {
+      "epoch": 1.9005833436763062,
+      "grad_norm": 0.07326147656237637,
+      "learning_rate": 5.406129532073628e-06,
+      "loss": 0.4662,
+      "step": 3830
+    },
+    {
+      "epoch": 1.9010798063795458,
+      "grad_norm": 0.07648989726494484,
+      "learning_rate": 5.4041812607059444e-06,
+      "loss": 0.484,
+      "step": 3831
+    },
+    {
+      "epoch": 1.9015762690827853,
+      "grad_norm": 0.07201495853310386,
+      "learning_rate": 5.402232927565632e-06,
+      "loss": 0.4919,
+      "step": 3832
+    },
+    {
+      "epoch": 1.9020727317860247,
+      "grad_norm": 0.07217460792172264,
+      "learning_rate": 5.4002845329504675e-06,
+      "loss": 0.4848,
+      "step": 3833
+    },
+    {
+      "epoch": 1.902569194489264,
+      "grad_norm": 0.07370157153449013,
+      "learning_rate": 5.398336077158231e-06,
+      "loss": 0.4753,
+      "step": 3834
+    },
+    {
+      "epoch": 1.9030656571925033,
+      "grad_norm": 0.06977914565580395,
+      "learning_rate": 5.39638756048671e-06,
+      "loss": 0.4935,
+      "step": 3835
+    },
+    {
+      "epoch": 1.9035621198957429,
+      "grad_norm": 0.07335495629369132,
+      "learning_rate": 5.394438983233707e-06,
+      "loss": 0.4682,
+      "step": 3836
+    },
+    {
+      "epoch": 1.9040585825989822,
+      "grad_norm": 0.07093032314632058,
+      "learning_rate": 5.39249034569703e-06,
+      "loss": 0.4617,
+      "step": 3837
+    },
+    {
+      "epoch": 1.9045550453022217,
+      "grad_norm": 0.07253844538726559,
+      "learning_rate": 5.390541648174495e-06,
+      "loss": 0.5027,
+      "step": 3838
+    },
+    {
+      "epoch": 1.905051508005461,
+      "grad_norm": 0.06916797666897612,
+      "learning_rate": 5.388592890963933e-06,
+      "loss": 0.4521,
+      "step": 3839
+    },
+    {
+      "epoch": 1.9055479707087004,
+      "grad_norm": 0.07497205907931132,
+      "learning_rate": 5.386644074363176e-06,
+      "loss": 0.4871,
+      "step": 3840
+    },
+    {
+      "epoch": 1.90604443341194,
+      "grad_norm": 0.07361931403289883,
+      "learning_rate": 5.384695198670074e-06,
+      "loss": 0.5075,
+      "step": 3841
+    },
+    {
+      "epoch": 1.9065408961151793,
+      "grad_norm": 0.07108507274036159,
+      "learning_rate": 5.38274626418248e-06,
+      "loss": 0.4663,
+      "step": 3842
+    },
+    {
+      "epoch": 1.9070373588184188,
+      "grad_norm": 0.06947858409999921,
+      "learning_rate": 5.380797271198253e-06,
+      "loss": 0.4917,
+      "step": 3843
+    },
+    {
+      "epoch": 1.9075338215216582,
+      "grad_norm": 0.07108447466202403,
+      "learning_rate": 5.378848220015271e-06,
+      "loss": 0.481,
+      "step": 3844
+    },
+    {
+      "epoch": 1.9080302842248975,
+      "grad_norm": 0.07108083929393204,
+      "learning_rate": 5.3768991109314115e-06,
+      "loss": 0.5002,
+      "step": 3845
+    },
+    {
+      "epoch": 1.908526746928137,
+      "grad_norm": 0.0717715304157745,
+      "learning_rate": 5.374949944244566e-06,
+      "loss": 0.4854,
+      "step": 3846
+    },
+    {
+      "epoch": 1.9090232096313764,
+      "grad_norm": 0.0720786792030362,
+      "learning_rate": 5.373000720252635e-06,
+      "loss": 0.5155,
+      "step": 3847
+    },
+    {
+      "epoch": 1.909519672334616,
+      "grad_norm": 0.07023062512373864,
+      "learning_rate": 5.371051439253524e-06,
+      "loss": 0.491,
+      "step": 3848
+    },
+    {
+      "epoch": 1.9100161350378553,
+      "grad_norm": 0.07508624689000723,
+      "learning_rate": 5.3691021015451494e-06,
+      "loss": 0.4648,
+      "step": 3849
+    },
+    {
+      "epoch": 1.9105125977410946,
+      "grad_norm": 0.0730348579251482,
+      "learning_rate": 5.367152707425437e-06,
+      "loss": 0.4769,
+      "step": 3850
+    },
+    {
+      "epoch": 1.9110090604443342,
+      "grad_norm": 0.07212915215112067,
+      "learning_rate": 5.36520325719232e-06,
+      "loss": 0.4645,
+      "step": 3851
+    },
+    {
+      "epoch": 1.9115055231475735,
+      "grad_norm": 0.07249343057488641,
+      "learning_rate": 5.36325375114374e-06,
+      "loss": 0.5165,
+      "step": 3852
+    },
+    {
+      "epoch": 1.912001985850813,
+      "grad_norm": 0.07200471814438766,
+      "learning_rate": 5.36130418957765e-06,
+      "loss": 0.4757,
+      "step": 3853
+    },
+    {
+      "epoch": 1.9124984485540524,
+      "grad_norm": 0.07433250128939686,
+      "learning_rate": 5.359354572792006e-06,
+      "loss": 0.5093,
+      "step": 3854
+    },
+    {
+      "epoch": 1.9129949112572917,
+      "grad_norm": 0.07198352611168042,
+      "learning_rate": 5.357404901084778e-06,
+      "loss": 0.489,
+      "step": 3855
+    },
+    {
+      "epoch": 1.9134913739605313,
+      "grad_norm": 0.07463196244659961,
+      "learning_rate": 5.355455174753941e-06,
+      "loss": 0.5198,
+      "step": 3856
+    },
+    {
+      "epoch": 1.9139878366637706,
+      "grad_norm": 0.0691576326182084,
+      "learning_rate": 5.35350539409748e-06,
+      "loss": 0.4683,
+      "step": 3857
+    },
+    {
+      "epoch": 1.9144842993670101,
+      "grad_norm": 0.07564132823230729,
+      "learning_rate": 5.351555559413389e-06,
+      "loss": 0.4893,
+      "step": 3858
+    },
+    {
+      "epoch": 1.9149807620702495,
+      "grad_norm": 0.07390246796901016,
+      "learning_rate": 5.349605670999667e-06,
+      "loss": 0.4801,
+      "step": 3859
+    },
+    {
+      "epoch": 1.9154772247734888,
+      "grad_norm": 0.0717750115158868,
+      "learning_rate": 5.347655729154323e-06,
+      "loss": 0.4849,
+      "step": 3860
+    },
+    {
+      "epoch": 1.9159736874767284,
+      "grad_norm": 0.07277950154580684,
+      "learning_rate": 5.345705734175375e-06,
+      "loss": 0.4754,
+      "step": 3861
+    },
+    {
+      "epoch": 1.9164701501799677,
+      "grad_norm": 0.07131228553584106,
+      "learning_rate": 5.343755686360849e-06,
+      "loss": 0.4685,
+      "step": 3862
+    },
+    {
+      "epoch": 1.9169666128832072,
+      "grad_norm": 0.07097194395346555,
+      "learning_rate": 5.341805586008778e-06,
+      "loss": 0.4539,
+      "step": 3863
+    },
+    {
+      "epoch": 1.9174630755864466,
+      "grad_norm": 0.07101003749169538,
+      "learning_rate": 5.339855433417203e-06,
+      "loss": 0.4888,
+      "step": 3864
+    },
+    {
+      "epoch": 1.917959538289686,
+      "grad_norm": 0.07298877591361808,
+      "learning_rate": 5.337905228884174e-06,
+      "loss": 0.4787,
+      "step": 3865
+    },
+    {
+      "epoch": 1.9184560009929255,
+      "grad_norm": 0.07395924862347028,
+      "learning_rate": 5.335954972707749e-06,
+      "loss": 0.4932,
+      "step": 3866
+    },
+    {
+      "epoch": 1.9189524636961648,
+      "grad_norm": 0.07140031977681557,
+      "learning_rate": 5.334004665185994e-06,
+      "loss": 0.4982,
+      "step": 3867
+    },
+    {
+      "epoch": 1.9194489263994043,
+      "grad_norm": 0.07305491864396065,
+      "learning_rate": 5.332054306616979e-06,
+      "loss": 0.4917,
+      "step": 3868
+    },
+    {
+      "epoch": 1.9199453891026437,
+      "grad_norm": 0.07182300787383228,
+      "learning_rate": 5.330103897298791e-06,
+      "loss": 0.5242,
+      "step": 3869
+    },
+    {
+      "epoch": 1.920441851805883,
+      "grad_norm": 0.0703090063753603,
+      "learning_rate": 5.328153437529512e-06,
+      "loss": 0.496,
+      "step": 3870
+    },
+    {
+      "epoch": 1.9209383145091226,
+      "grad_norm": 0.06929915693165271,
+      "learning_rate": 5.326202927607242e-06,
+      "loss": 0.4823,
+      "step": 3871
+    },
+    {
+      "epoch": 1.9214347772123619,
+      "grad_norm": 0.07318562935704766,
+      "learning_rate": 5.324252367830085e-06,
+      "loss": 0.5222,
+      "step": 3872
+    },
+    {
+      "epoch": 1.9219312399156014,
+      "grad_norm": 0.070345300692849,
+      "learning_rate": 5.322301758496153e-06,
+      "loss": 0.4886,
+      "step": 3873
+    },
+    {
+      "epoch": 1.9224277026188408,
+      "grad_norm": 0.0733085359507376,
+      "learning_rate": 5.320351099903565e-06,
+      "loss": 0.512,
+      "step": 3874
+    },
+    {
+      "epoch": 1.92292416532208,
+      "grad_norm": 0.0719498700081074,
+      "learning_rate": 5.318400392350449e-06,
+      "loss": 0.4699,
+      "step": 3875
+    },
+    {
+      "epoch": 1.9234206280253194,
+      "grad_norm": 0.06824895945528683,
+      "learning_rate": 5.316449636134936e-06,
+      "loss": 0.4729,
+      "step": 3876
+    },
+    {
+      "epoch": 1.923917090728559,
+      "grad_norm": 0.07041042686305006,
+      "learning_rate": 5.31449883155517e-06,
+      "loss": 0.476,
+      "step": 3877
+    },
+    {
+      "epoch": 1.9244135534317985,
+      "grad_norm": 0.067579216909461,
+      "learning_rate": 5.3125479789093014e-06,
+      "loss": 0.4677,
+      "step": 3878
+    },
+    {
+      "epoch": 1.9249100161350379,
+      "grad_norm": 0.06917056104636418,
+      "learning_rate": 5.310597078495485e-06,
+      "loss": 0.5167,
+      "step": 3879
+    },
+    {
+      "epoch": 1.9254064788382772,
+      "grad_norm": 0.07520188972413051,
+      "learning_rate": 5.308646130611885e-06,
+      "loss": 0.5043,
+      "step": 3880
+    },
+    {
+      "epoch": 1.9259029415415165,
+      "grad_norm": 0.07522798508907527,
+      "learning_rate": 5.306695135556673e-06,
+      "loss": 0.5008,
+      "step": 3881
+    },
+    {
+      "epoch": 1.926399404244756,
+      "grad_norm": 0.07326055869597388,
+      "learning_rate": 5.304744093628028e-06,
+      "loss": 0.5286,
+      "step": 3882
+    },
+    {
+      "epoch": 1.9268958669479956,
+      "grad_norm": 0.0719762524355997,
+      "learning_rate": 5.302793005124134e-06,
+      "loss": 0.5473,
+      "step": 3883
+    },
+    {
+      "epoch": 1.927392329651235,
+      "grad_norm": 0.07057243925086228,
+      "learning_rate": 5.300841870343183e-06,
+      "loss": 0.4878,
+      "step": 3884
+    },
+    {
+      "epoch": 1.9278887923544743,
+      "grad_norm": 0.07456395275600818,
+      "learning_rate": 5.298890689583377e-06,
+      "loss": 0.4613,
+      "step": 3885
+    },
+    {
+      "epoch": 1.9283852550577136,
+      "grad_norm": 0.0739750197425005,
+      "learning_rate": 5.2969394631429205e-06,
+      "loss": 0.4888,
+      "step": 3886
+    },
+    {
+      "epoch": 1.9288817177609532,
+      "grad_norm": 0.06859930114929563,
+      "learning_rate": 5.294988191320029e-06,
+      "loss": 0.4502,
+      "step": 3887
+    },
+    {
+      "epoch": 1.9293781804641927,
+      "grad_norm": 0.07333117565055676,
+      "learning_rate": 5.29303687441292e-06,
+      "loss": 0.5159,
+      "step": 3888
+    },
+    {
+      "epoch": 1.929874643167432,
+      "grad_norm": 0.07881910644631798,
+      "learning_rate": 5.2910855127198255e-06,
+      "loss": 0.5053,
+      "step": 3889
+    },
+    {
+      "epoch": 1.9303711058706714,
+      "grad_norm": 0.07125205611992942,
+      "learning_rate": 5.289134106538978e-06,
+      "loss": 0.4817,
+      "step": 3890
+    },
+    {
+      "epoch": 1.9308675685739107,
+      "grad_norm": 0.07168797442151044,
+      "learning_rate": 5.287182656168618e-06,
+      "loss": 0.4976,
+      "step": 3891
+    },
+    {
+      "epoch": 1.9313640312771503,
+      "grad_norm": 0.07173357891598954,
+      "learning_rate": 5.2852311619069915e-06,
+      "loss": 0.4605,
+      "step": 3892
+    },
+    {
+      "epoch": 1.9318604939803898,
+      "grad_norm": 0.07179667762070725,
+      "learning_rate": 5.2832796240523565e-06,
+      "loss": 0.4384,
+      "step": 3893
+    },
+    {
+      "epoch": 1.9323569566836292,
+      "grad_norm": 0.0771564342106695,
+      "learning_rate": 5.281328042902973e-06,
+      "loss": 0.4836,
+      "step": 3894
+    },
+    {
+      "epoch": 1.9328534193868685,
+      "grad_norm": 0.06971413584196552,
+      "learning_rate": 5.279376418757108e-06,
+      "loss": 0.4616,
+      "step": 3895
+    },
+    {
+      "epoch": 1.9333498820901078,
+      "grad_norm": 0.07134917459980415,
+      "learning_rate": 5.27742475191304e-06,
+      "loss": 0.4853,
+      "step": 3896
+    },
+    {
+      "epoch": 1.9338463447933474,
+      "grad_norm": 0.06948966703103633,
+      "learning_rate": 5.275473042669043e-06,
+      "loss": 0.4525,
+      "step": 3897
+    },
+    {
+      "epoch": 1.934342807496587,
+      "grad_norm": 0.09123313507126177,
+      "learning_rate": 5.273521291323411e-06,
+      "loss": 0.4759,
+      "step": 3898
+    },
+    {
+      "epoch": 1.9348392701998263,
+      "grad_norm": 0.07182320974046329,
+      "learning_rate": 5.271569498174435e-06,
+      "loss": 0.4837,
+      "step": 3899
+    },
+    {
+      "epoch": 1.9353357329030656,
+      "grad_norm": 0.07167426782717073,
+      "learning_rate": 5.269617663520414e-06,
+      "loss": 0.4925,
+      "step": 3900
+    },
+    {
+      "epoch": 1.935832195606305,
+      "grad_norm": 0.06971566479492058,
+      "learning_rate": 5.2676657876596575e-06,
+      "loss": 0.4939,
+      "step": 3901
+    },
+    {
+      "epoch": 1.9363286583095445,
+      "grad_norm": 0.068449292316369,
+      "learning_rate": 5.265713870890476e-06,
+      "loss": 0.4435,
+      "step": 3902
+    },
+    {
+      "epoch": 1.936825121012784,
+      "grad_norm": 0.07115671276963907,
+      "learning_rate": 5.263761913511189e-06,
+      "loss": 0.4797,
+      "step": 3903
+    },
+    {
+      "epoch": 1.9373215837160234,
+      "grad_norm": 0.07092945923206197,
+      "learning_rate": 5.261809915820124e-06,
+      "loss": 0.4768,
+      "step": 3904
+    },
+    {
+      "epoch": 1.9378180464192627,
+      "grad_norm": 0.07288485555482782,
+      "learning_rate": 5.259857878115611e-06,
+      "loss": 0.5005,
+      "step": 3905
+    },
+    {
+      "epoch": 1.938314509122502,
+      "grad_norm": 0.06857200953594583,
+      "learning_rate": 5.257905800695988e-06,
+      "loss": 0.4786,
+      "step": 3906
+    },
+    {
+      "epoch": 1.9388109718257416,
+      "grad_norm": 0.07014718654146762,
+      "learning_rate": 5.2559536838595995e-06,
+      "loss": 0.4651,
+      "step": 3907
+    },
+    {
+      "epoch": 1.9393074345289811,
+      "grad_norm": 0.0701723829531734,
+      "learning_rate": 5.254001527904793e-06,
+      "loss": 0.4626,
+      "step": 3908
+    },
+    {
+      "epoch": 1.9398038972322205,
+      "grad_norm": 0.07084434226107171,
+      "learning_rate": 5.252049333129925e-06,
+      "loss": 0.4729,
+      "step": 3909
+    },
+    {
+      "epoch": 1.9403003599354598,
+      "grad_norm": 0.07160630955321064,
+      "learning_rate": 5.250097099833358e-06,
+      "loss": 0.484,
+      "step": 3910
+    },
+    {
+      "epoch": 1.9407968226386991,
+      "grad_norm": 0.07034163421346201,
+      "learning_rate": 5.248144828313459e-06,
+      "loss": 0.51,
+      "step": 3911
+    },
+    {
+      "epoch": 1.9412932853419387,
+      "grad_norm": 0.06976530549909565,
+      "learning_rate": 5.2461925188686035e-06,
+      "loss": 0.4295,
+      "step": 3912
+    },
+    {
+      "epoch": 1.9417897480451782,
+      "grad_norm": 0.07191014314418367,
+      "learning_rate": 5.244240171797168e-06,
+      "loss": 0.4845,
+      "step": 3913
+    },
+    {
+      "epoch": 1.9422862107484176,
+      "grad_norm": 0.07058958360248195,
+      "learning_rate": 5.2422877873975384e-06,
+      "loss": 0.4803,
+      "step": 3914
+    },
+    {
+      "epoch": 1.942782673451657,
+      "grad_norm": 0.07252410993239682,
+      "learning_rate": 5.240335365968104e-06,
+      "loss": 0.5167,
+      "step": 3915
+    },
+    {
+      "epoch": 1.9432791361548962,
+      "grad_norm": 0.07298165209234508,
+      "learning_rate": 5.2383829078072635e-06,
+      "loss": 0.5214,
+      "step": 3916
+    },
+    {
+      "epoch": 1.9437755988581358,
+      "grad_norm": 0.07366922809256125,
+      "learning_rate": 5.236430413213419e-06,
+      "loss": 0.5065,
+      "step": 3917
+    },
+    {
+      "epoch": 1.9442720615613753,
+      "grad_norm": 0.0743591532780005,
+      "learning_rate": 5.234477882484975e-06,
+      "loss": 0.5747,
+      "step": 3918
+    },
+    {
+      "epoch": 1.9447685242646147,
+      "grad_norm": 0.0752519340136121,
+      "learning_rate": 5.232525315920346e-06,
+      "loss": 0.5071,
+      "step": 3919
+    },
+    {
+      "epoch": 1.945264986967854,
+      "grad_norm": 0.07115434487269699,
+      "learning_rate": 5.230572713817951e-06,
+      "loss": 0.4723,
+      "step": 3920
+    },
+    {
+      "epoch": 1.9457614496710933,
+      "grad_norm": 0.07085415232697705,
+      "learning_rate": 5.228620076476214e-06,
+      "loss": 0.4882,
+      "step": 3921
+    },
+    {
+      "epoch": 1.9462579123743329,
+      "grad_norm": 0.07250241964423546,
+      "learning_rate": 5.226667404193564e-06,
+      "loss": 0.5011,
+      "step": 3922
+    },
+    {
+      "epoch": 1.9467543750775724,
+      "grad_norm": 0.06899533098215475,
+      "learning_rate": 5.224714697268437e-06,
+      "loss": 0.4682,
+      "step": 3923
+    },
+    {
+      "epoch": 1.9472508377808118,
+      "grad_norm": 0.07378228905532908,
+      "learning_rate": 5.222761955999269e-06,
+      "loss": 0.4722,
+      "step": 3924
+    },
+    {
+      "epoch": 1.947747300484051,
+      "grad_norm": 0.07267710646256538,
+      "learning_rate": 5.220809180684508e-06,
+      "loss": 0.4793,
+      "step": 3925
+    },
+    {
+      "epoch": 1.9482437631872904,
+      "grad_norm": 0.07188532716957385,
+      "learning_rate": 5.218856371622605e-06,
+      "loss": 0.4758,
+      "step": 3926
+    },
+    {
+      "epoch": 1.94874022589053,
+      "grad_norm": 0.0748892540261098,
+      "learning_rate": 5.216903529112015e-06,
+      "loss": 0.4768,
+      "step": 3927
+    },
+    {
+      "epoch": 1.9492366885937695,
+      "grad_norm": 0.07361587591698933,
+      "learning_rate": 5.214950653451199e-06,
+      "loss": 0.4958,
+      "step": 3928
+    },
+    {
+      "epoch": 1.9497331512970089,
+      "grad_norm": 0.0750386427247887,
+      "learning_rate": 5.21299774493862e-06,
+      "loss": 0.5139,
+      "step": 3929
+    },
+    {
+      "epoch": 1.9502296140002482,
+      "grad_norm": 0.07157581527430623,
+      "learning_rate": 5.211044803872752e-06,
+      "loss": 0.5005,
+      "step": 3930
+    },
+    {
+      "epoch": 1.9507260767034875,
+      "grad_norm": 0.07122801864538819,
+      "learning_rate": 5.20909183055207e-06,
+      "loss": 0.4836,
+      "step": 3931
+    },
+    {
+      "epoch": 1.951222539406727,
+      "grad_norm": 0.073775379701217,
+      "learning_rate": 5.207138825275053e-06,
+      "loss": 0.5205,
+      "step": 3932
+    },
+    {
+      "epoch": 1.9517190021099666,
+      "grad_norm": 0.0699335599632651,
+      "learning_rate": 5.205185788340189e-06,
+      "loss": 0.4552,
+      "step": 3933
+    },
+    {
+      "epoch": 1.952215464813206,
+      "grad_norm": 0.0747012833856705,
+      "learning_rate": 5.2032327200459665e-06,
+      "loss": 0.5261,
+      "step": 3934
+    },
+    {
+      "epoch": 1.9527119275164453,
+      "grad_norm": 0.06986301985581114,
+      "learning_rate": 5.201279620690881e-06,
+      "loss": 0.4645,
+      "step": 3935
+    },
+    {
+      "epoch": 1.9532083902196846,
+      "grad_norm": 0.07083169381180109,
+      "learning_rate": 5.199326490573433e-06,
+      "loss": 0.4753,
+      "step": 3936
+    },
+    {
+      "epoch": 1.9537048529229242,
+      "grad_norm": 0.07316821283838185,
+      "learning_rate": 5.197373329992127e-06,
+      "loss": 0.5037,
+      "step": 3937
+    },
+    {
+      "epoch": 1.9542013156261637,
+      "grad_norm": 0.0709068153288116,
+      "learning_rate": 5.195420139245472e-06,
+      "loss": 0.46,
+      "step": 3938
+    },
+    {
+      "epoch": 1.954697778329403,
+      "grad_norm": 0.07258328229639797,
+      "learning_rate": 5.193466918631984e-06,
+      "loss": 0.4768,
+      "step": 3939
+    },
+    {
+      "epoch": 1.9551942410326424,
+      "grad_norm": 0.07013691797325394,
+      "learning_rate": 5.191513668450178e-06,
+      "loss": 0.4537,
+      "step": 3940
+    },
+    {
+      "epoch": 1.9556907037358817,
+      "grad_norm": 0.0691432700044006,
+      "learning_rate": 5.189560388998578e-06,
+      "loss": 0.4689,
+      "step": 3941
+    },
+    {
+      "epoch": 1.9561871664391213,
+      "grad_norm": 0.07255706197612641,
+      "learning_rate": 5.187607080575712e-06,
+      "loss": 0.4471,
+      "step": 3942
+    },
+    {
+      "epoch": 1.9566836291423608,
+      "grad_norm": 0.07177608806087026,
+      "learning_rate": 5.185653743480112e-06,
+      "loss": 0.449,
+      "step": 3943
+    },
+    {
+      "epoch": 1.9571800918456002,
+      "grad_norm": 0.07710954560772512,
+      "learning_rate": 5.183700378010315e-06,
+      "loss": 0.4822,
+      "step": 3944
+    },
+    {
+      "epoch": 1.9576765545488395,
+      "grad_norm": 0.07169664734250977,
+      "learning_rate": 5.1817469844648585e-06,
+      "loss": 0.4972,
+      "step": 3945
+    },
+    {
+      "epoch": 1.9581730172520788,
+      "grad_norm": 0.07193769644389156,
+      "learning_rate": 5.179793563142291e-06,
+      "loss": 0.468,
+      "step": 3946
+    },
+    {
+      "epoch": 1.9586694799553184,
+      "grad_norm": 0.07188270892453336,
+      "learning_rate": 5.17784011434116e-06,
+      "loss": 0.5197,
+      "step": 3947
+    },
+    {
+      "epoch": 1.959165942658558,
+      "grad_norm": 0.07174487240214936,
+      "learning_rate": 5.1758866383600185e-06,
+      "loss": 0.49,
+      "step": 3948
+    },
+    {
+      "epoch": 1.9596624053617973,
+      "grad_norm": 0.06851866889793691,
+      "learning_rate": 5.1739331354974245e-06,
+      "loss": 0.4647,
+      "step": 3949
+    },
+    {
+      "epoch": 1.9601588680650366,
+      "grad_norm": 0.07355462858315683,
+      "learning_rate": 5.17197960605194e-06,
+      "loss": 0.4867,
+      "step": 3950
+    },
+    {
+      "epoch": 1.960655330768276,
+      "grad_norm": 0.07129939537738009,
+      "learning_rate": 5.17002605032213e-06,
+      "loss": 0.4696,
+      "step": 3951
+    },
+    {
+      "epoch": 1.9611517934715155,
+      "grad_norm": 0.07265310768420902,
+      "learning_rate": 5.168072468606564e-06,
+      "loss": 0.5057,
+      "step": 3952
+    },
+    {
+      "epoch": 1.961648256174755,
+      "grad_norm": 0.07219085002056012,
+      "learning_rate": 5.166118861203816e-06,
+      "loss": 0.4757,
+      "step": 3953
+    },
+    {
+      "epoch": 1.9621447188779944,
+      "grad_norm": 0.07188896917633172,
+      "learning_rate": 5.1641652284124645e-06,
+      "loss": 0.4623,
+      "step": 3954
+    },
+    {
+      "epoch": 1.9626411815812337,
+      "grad_norm": 0.07047459964076412,
+      "learning_rate": 5.16221157053109e-06,
+      "loss": 0.4917,
+      "step": 3955
+    },
+    {
+      "epoch": 1.963137644284473,
+      "grad_norm": 0.07207176246858632,
+      "learning_rate": 5.160257887858278e-06,
+      "loss": 0.4835,
+      "step": 3956
+    },
+    {
+      "epoch": 1.9636341069877126,
+      "grad_norm": 0.07245058141331831,
+      "learning_rate": 5.158304180692615e-06,
+      "loss": 0.5134,
+      "step": 3957
+    },
+    {
+      "epoch": 1.9641305696909521,
+      "grad_norm": 0.0709051816408688,
+      "learning_rate": 5.156350449332698e-06,
+      "loss": 0.4731,
+      "step": 3958
+    },
+    {
+      "epoch": 1.9646270323941915,
+      "grad_norm": 0.0716399263483255,
+      "learning_rate": 5.154396694077121e-06,
+      "loss": 0.477,
+      "step": 3959
+    },
+    {
+      "epoch": 1.9651234950974308,
+      "grad_norm": 0.070990095924848,
+      "learning_rate": 5.152442915224486e-06,
+      "loss": 0.4813,
+      "step": 3960
+    },
+    {
+      "epoch": 1.9656199578006701,
+      "grad_norm": 0.07019738338899009,
+      "learning_rate": 5.150489113073394e-06,
+      "loss": 0.5073,
+      "step": 3961
+    },
+    {
+      "epoch": 1.9661164205039097,
+      "grad_norm": 0.07049687491155082,
+      "learning_rate": 5.148535287922457e-06,
+      "loss": 0.4801,
+      "step": 3962
+    },
+    {
+      "epoch": 1.9666128832071492,
+      "grad_norm": 0.07129508411132399,
+      "learning_rate": 5.1465814400702804e-06,
+      "loss": 0.484,
+      "step": 3963
+    },
+    {
+      "epoch": 1.9671093459103886,
+      "grad_norm": 0.07490675464094516,
+      "learning_rate": 5.144627569815481e-06,
+      "loss": 0.4956,
+      "step": 3964
+    },
+    {
+      "epoch": 1.9676058086136279,
+      "grad_norm": 0.07152667854181811,
+      "learning_rate": 5.142673677456676e-06,
+      "loss": 0.4937,
+      "step": 3965
+    },
+    {
+      "epoch": 1.9681022713168672,
+      "grad_norm": 0.07213237931738262,
+      "learning_rate": 5.1407197632924885e-06,
+      "loss": 0.5074,
+      "step": 3966
+    },
+    {
+      "epoch": 1.9685987340201068,
+      "grad_norm": 0.07476637301318777,
+      "learning_rate": 5.138765827621541e-06,
+      "loss": 0.5058,
+      "step": 3967
+    },
+    {
+      "epoch": 1.9690951967233463,
+      "grad_norm": 0.07153425116188193,
+      "learning_rate": 5.136811870742462e-06,
+      "loss": 0.4935,
+      "step": 3968
+    },
+    {
+      "epoch": 1.9695916594265857,
+      "grad_norm": 0.07285692288676107,
+      "learning_rate": 5.134857892953881e-06,
+      "loss": 0.4776,
+      "step": 3969
+    },
+    {
+      "epoch": 1.970088122129825,
+      "grad_norm": 0.07130975929675355,
+      "learning_rate": 5.132903894554434e-06,
+      "loss": 0.4711,
+      "step": 3970
+    },
+    {
+      "epoch": 1.9705845848330643,
+      "grad_norm": 0.06927463689733188,
+      "learning_rate": 5.130949875842758e-06,
+      "loss": 0.4741,
+      "step": 3971
+    },
+    {
+      "epoch": 1.9710810475363039,
+      "grad_norm": 0.07226451812060516,
+      "learning_rate": 5.128995837117493e-06,
+      "loss": 0.4993,
+      "step": 3972
+    },
+    {
+      "epoch": 1.9715775102395434,
+      "grad_norm": 0.06814833359691802,
+      "learning_rate": 5.127041778677283e-06,
+      "loss": 0.4552,
+      "step": 3973
+    },
+    {
+      "epoch": 1.9720739729427827,
+      "grad_norm": 0.07386779362943056,
+      "learning_rate": 5.1250877008207725e-06,
+      "loss": 0.4943,
+      "step": 3974
+    },
+    {
+      "epoch": 1.972570435646022,
+      "grad_norm": 0.07245457245389018,
+      "learning_rate": 5.123133603846613e-06,
+      "loss": 0.4883,
+      "step": 3975
+    },
+    {
+      "epoch": 1.9730668983492614,
+      "grad_norm": 0.0729979245483732,
+      "learning_rate": 5.121179488053458e-06,
+      "loss": 0.4817,
+      "step": 3976
+    },
+    {
+      "epoch": 1.973563361052501,
+      "grad_norm": 0.0718325924027625,
+      "learning_rate": 5.1192253537399595e-06,
+      "loss": 0.4552,
+      "step": 3977
+    },
+    {
+      "epoch": 1.9740598237557403,
+      "grad_norm": 0.07326345383485158,
+      "learning_rate": 5.117271201204779e-06,
+      "loss": 0.5166,
+      "step": 3978
+    },
+    {
+      "epoch": 1.9745562864589798,
+      "grad_norm": 0.07098301430790274,
+      "learning_rate": 5.115317030746575e-06,
+      "loss": 0.4545,
+      "step": 3979
+    },
+    {
+      "epoch": 1.9750527491622192,
+      "grad_norm": 0.07394770312842015,
+      "learning_rate": 5.11336284266401e-06,
+      "loss": 0.4882,
+      "step": 3980
+    },
+    {
+      "epoch": 1.9755492118654585,
+      "grad_norm": 0.07205746944153327,
+      "learning_rate": 5.111408637255754e-06,
+      "loss": 0.494,
+      "step": 3981
+    },
+    {
+      "epoch": 1.976045674568698,
+      "grad_norm": 0.06869822675283942,
+      "learning_rate": 5.109454414820475e-06,
+      "loss": 0.4724,
+      "step": 3982
+    },
+    {
+      "epoch": 1.9765421372719374,
+      "grad_norm": 0.07308322700351189,
+      "learning_rate": 5.107500175656842e-06,
+      "loss": 0.4904,
+      "step": 3983
+    },
+    {
+      "epoch": 1.977038599975177,
+      "grad_norm": 0.07165097342539073,
+      "learning_rate": 5.10554592006353e-06,
+      "loss": 0.4908,
+      "step": 3984
+    },
+    {
+      "epoch": 1.9775350626784163,
+      "grad_norm": 0.07454745150610671,
+      "learning_rate": 5.103591648339218e-06,
+      "loss": 0.4706,
+      "step": 3985
+    },
+    {
+      "epoch": 1.9780315253816556,
+      "grad_norm": 0.06909233417316848,
+      "learning_rate": 5.101637360782584e-06,
+      "loss": 0.4881,
+      "step": 3986
+    },
+    {
+      "epoch": 1.9785279880848952,
+      "grad_norm": 0.07547889134109059,
+      "learning_rate": 5.0996830576923075e-06,
+      "loss": 0.4967,
+      "step": 3987
+    },
+    {
+      "epoch": 1.9790244507881345,
+      "grad_norm": 0.07184250001754491,
+      "learning_rate": 5.097728739367076e-06,
+      "loss": 0.4631,
+      "step": 3988
+    },
+    {
+      "epoch": 1.979520913491374,
+      "grad_norm": 0.07246338867423593,
+      "learning_rate": 5.095774406105572e-06,
+      "loss": 0.5352,
+      "step": 3989
+    },
+    {
+      "epoch": 1.9800173761946134,
+      "grad_norm": 0.07117800861507659,
+      "learning_rate": 5.0938200582064846e-06,
+      "loss": 0.4887,
+      "step": 3990
+    },
+    {
+      "epoch": 1.9805138388978527,
+      "grad_norm": 0.07386014244542119,
+      "learning_rate": 5.091865695968508e-06,
+      "loss": 0.4824,
+      "step": 3991
+    },
+    {
+      "epoch": 1.9810103016010923,
+      "grad_norm": 0.07463772727319021,
+      "learning_rate": 5.089911319690331e-06,
+      "loss": 0.4731,
+      "step": 3992
+    },
+    {
+      "epoch": 1.9815067643043316,
+      "grad_norm": 0.07409681863302726,
+      "learning_rate": 5.087956929670651e-06,
+      "loss": 0.5116,
+      "step": 3993
+    },
+    {
+      "epoch": 1.9820032270075711,
+      "grad_norm": 0.07102473269761052,
+      "learning_rate": 5.086002526208166e-06,
+      "loss": 0.4988,
+      "step": 3994
+    },
+    {
+      "epoch": 1.9824996897108105,
+      "grad_norm": 0.07270851135800403,
+      "learning_rate": 5.084048109601571e-06,
+      "loss": 0.5075,
+      "step": 3995
+    },
+    {
+      "epoch": 1.9829961524140498,
+      "grad_norm": 0.07287188630853728,
+      "learning_rate": 5.0820936801495716e-06,
+      "loss": 0.5128,
+      "step": 3996
+    },
+    {
+      "epoch": 1.9834926151172894,
+      "grad_norm": 0.07061601120193683,
+      "learning_rate": 5.080139238150869e-06,
+      "loss": 0.4846,
+      "step": 3997
+    },
+    {
+      "epoch": 1.9839890778205287,
+      "grad_norm": 0.07260115702119305,
+      "learning_rate": 5.07818478390417e-06,
+      "loss": 0.4818,
+      "step": 3998
+    },
+    {
+      "epoch": 1.9844855405237682,
+      "grad_norm": 0.07763074208464117,
+      "learning_rate": 5.076230317708179e-06,
+      "loss": 0.5172,
+      "step": 3999
+    },
+    {
+      "epoch": 1.9849820032270076,
+      "grad_norm": 0.06975414599004237,
+      "learning_rate": 5.074275839861606e-06,
+      "loss": 0.4785,
+      "step": 4000
+    },
+    {
+      "epoch": 1.985478465930247,
+      "grad_norm": 0.07380231976084484,
+      "learning_rate": 5.072321350663163e-06,
+      "loss": 0.4678,
+      "step": 4001
+    },
+    {
+      "epoch": 1.9859749286334865,
+      "grad_norm": 0.07237761374407202,
+      "learning_rate": 5.070366850411561e-06,
+      "loss": 0.4779,
+      "step": 4002
+    },
+    {
+      "epoch": 1.9864713913367258,
+      "grad_norm": 0.07057959775998889,
+      "learning_rate": 5.068412339405514e-06,
+      "loss": 0.4809,
+      "step": 4003
+    },
+    {
+      "epoch": 1.9869678540399653,
+      "grad_norm": 0.0715457070565392,
+      "learning_rate": 5.066457817943738e-06,
+      "loss": 0.4992,
+      "step": 4004
+    },
+    {
+      "epoch": 1.9874643167432047,
+      "grad_norm": 0.07391983194928121,
+      "learning_rate": 5.06450328632495e-06,
+      "loss": 0.5153,
+      "step": 4005
+    },
+    {
+      "epoch": 1.987960779446444,
+      "grad_norm": 0.07503637951833564,
+      "learning_rate": 5.062548744847867e-06,
+      "loss": 0.5113,
+      "step": 4006
+    },
+    {
+      "epoch": 1.9884572421496836,
+      "grad_norm": 0.07074043181554573,
+      "learning_rate": 5.0605941938112135e-06,
+      "loss": 0.4669,
+      "step": 4007
+    },
+    {
+      "epoch": 1.9889537048529229,
+      "grad_norm": 0.07169585364157281,
+      "learning_rate": 5.058639633513708e-06,
+      "loss": 0.4969,
+      "step": 4008
+    },
+    {
+      "epoch": 1.9894501675561624,
+      "grad_norm": 0.06785942658601277,
+      "learning_rate": 5.056685064254075e-06,
+      "loss": 0.4561,
+      "step": 4009
+    },
+    {
+      "epoch": 1.9899466302594018,
+      "grad_norm": 0.06918926184232858,
+      "learning_rate": 5.054730486331041e-06,
+      "loss": 0.4759,
+      "step": 4010
+    },
+    {
+      "epoch": 1.990443092962641,
+      "grad_norm": 0.0749865895276791,
+      "learning_rate": 5.052775900043326e-06,
+      "loss": 0.5023,
+      "step": 4011
+    },
+    {
+      "epoch": 1.9909395556658807,
+      "grad_norm": 0.07200649155022336,
+      "learning_rate": 5.050821305689662e-06,
+      "loss": 0.4835,
+      "step": 4012
+    },
+    {
+      "epoch": 1.99143601836912,
+      "grad_norm": 0.06885032143514215,
+      "learning_rate": 5.048866703568778e-06,
+      "loss": 0.4554,
+      "step": 4013
+    },
+    {
+      "epoch": 1.9919324810723595,
+      "grad_norm": 0.07501649054063417,
+      "learning_rate": 5.046912093979402e-06,
+      "loss": 0.5018,
+      "step": 4014
+    },
+    {
+      "epoch": 1.9924289437755989,
+      "grad_norm": 0.07143933594119112,
+      "learning_rate": 5.044957477220261e-06,
+      "loss": 0.4684,
+      "step": 4015
+    },
+    {
+      "epoch": 1.9929254064788382,
+      "grad_norm": 0.07164911960402352,
+      "learning_rate": 5.043002853590093e-06,
+      "loss": 0.4778,
+      "step": 4016
+    },
+    {
+      "epoch": 1.9934218691820775,
+      "grad_norm": 0.0724278620954663,
+      "learning_rate": 5.0410482233876275e-06,
+      "loss": 0.4862,
+      "step": 4017
+    },
+    {
+      "epoch": 1.993918331885317,
+      "grad_norm": 0.07324551513366404,
+      "learning_rate": 5.0390935869116006e-06,
+      "loss": 0.4954,
+      "step": 4018
+    },
+    {
+      "epoch": 1.9944147945885566,
+      "grad_norm": 0.07243750592972426,
+      "learning_rate": 5.0371389444607455e-06,
+      "loss": 0.4751,
+      "step": 4019
+    },
+    {
+      "epoch": 1.994911257291796,
+      "grad_norm": 0.07093125056962983,
+      "learning_rate": 5.035184296333798e-06,
+      "loss": 0.5011,
+      "step": 4020
+    },
+    {
+      "epoch": 1.9954077199950353,
+      "grad_norm": 0.07146838278764876,
+      "learning_rate": 5.033229642829494e-06,
+      "loss": 0.4697,
+      "step": 4021
+    },
+    {
+      "epoch": 1.9959041826982746,
+      "grad_norm": 0.0680395947687067,
+      "learning_rate": 5.0312749842465725e-06,
+      "loss": 0.4548,
+      "step": 4022
+    },
+    {
+      "epoch": 1.9964006454015142,
+      "grad_norm": 0.07297368724052139,
+      "learning_rate": 5.029320320883771e-06,
+      "loss": 0.5171,
+      "step": 4023
+    },
+    {
+      "epoch": 1.9968971081047537,
+      "grad_norm": 0.07341048772340283,
+      "learning_rate": 5.0273656530398285e-06,
+      "loss": 0.4846,
+      "step": 4024
+    },
+    {
+      "epoch": 1.997393570807993,
+      "grad_norm": 0.07150096630499465,
+      "learning_rate": 5.025410981013486e-06,
+      "loss": 0.4866,
+      "step": 4025
+    },
+    {
+      "epoch": 1.9978900335112324,
+      "grad_norm": 0.06923746655335533,
+      "learning_rate": 5.023456305103482e-06,
+      "loss": 0.4842,
+      "step": 4026
+    },
+    {
+      "epoch": 1.9983864962144717,
+      "grad_norm": 0.07353105275938862,
+      "learning_rate": 5.021501625608557e-06,
+      "loss": 0.4865,
+      "step": 4027
+    },
+    {
+      "epoch": 1.9988829589177113,
+      "grad_norm": 0.07262147086529945,
+      "learning_rate": 5.019546942827452e-06,
+      "loss": 0.5071,
+      "step": 4028
+    },
+    {
+      "epoch": 1.9993794216209508,
+      "grad_norm": 0.07009105495484086,
+      "learning_rate": 5.017592257058912e-06,
+      "loss": 0.4571,
+      "step": 4029
+    },
+    {
+      "epoch": 1.9998758843241902,
+      "grad_norm": 0.0717265778317302,
+      "learning_rate": 5.015637568601678e-06,
+      "loss": 0.4721,
+      "step": 4030
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0717265778317302,
+      "learning_rate": 5.013682877754491e-06,
+      "loss": 0.1317,
+      "step": 4031
+    },
+    {
+      "epoch": 2.0003723470274295,
+      "grad_norm": 0.07314792877813803,
+      "learning_rate": 5.011728184816096e-06,
+      "loss": 0.3638,
+      "step": 4032
+    },
+    {
+      "epoch": 2.0003723470274295,
+      "eval_loss": 0.5145248770713806,
+      "eval_runtime": 258.8663,
+      "eval_samples_per_second": 117.254,
+      "eval_steps_per_second": 14.66,
+      "step": 4032
+    },
+    {
+      "epoch": 2.0004964627032393,
+      "grad_norm": 0.08562217509650846,
+      "learning_rate": 5.009773490085236e-06,
+      "loss": 0.4738,
+      "step": 4033
+    },
+    {
+      "epoch": 2.0009929254064787,
+      "grad_norm": 0.08037551216927723,
+      "learning_rate": 5.007818793860656e-06,
+      "loss": 0.4554,
+      "step": 4034
+    },
+    {
+      "epoch": 2.0014893881097184,
+      "grad_norm": 0.07699272192544572,
+      "learning_rate": 5.0058640964410975e-06,
+      "loss": 0.4558,
+      "step": 4035
+    },
+    {
+      "epoch": 2.0019858508129578,
+      "grad_norm": 0.07490333477454107,
+      "learning_rate": 5.003909398125306e-06,
+      "loss": 0.4492,
+      "step": 4036
+    },
+    {
+      "epoch": 2.002482313516197,
+      "grad_norm": 0.07484042574828073,
+      "learning_rate": 5.001954699212026e-06,
+      "loss": 0.467,
+      "step": 4037
+    },
+    {
+      "epoch": 2.0029787762194364,
+      "grad_norm": 0.07985101916527011,
+      "learning_rate": 5e-06,
+      "loss": 0.4856,
+      "step": 4038
+    },
+    {
+      "epoch": 2.0034752389226758,
+      "grad_norm": 0.07776863307734547,
+      "learning_rate": 4.998045300787976e-06,
+      "loss": 0.4442,
+      "step": 4039
+    },
+    {
+      "epoch": 2.0039717016259155,
+      "grad_norm": 0.08421470369676873,
+      "learning_rate": 4.996090601874695e-06,
+      "loss": 0.4537,
+      "step": 4040
+    },
+    {
+      "epoch": 2.004468164329155,
+      "grad_norm": 0.07856954374286516,
+      "learning_rate": 4.994135903558904e-06,
+      "loss": 0.4599,
+      "step": 4041
+    },
+    {
+      "epoch": 2.004964627032394,
+      "grad_norm": 0.07287925161116053,
+      "learning_rate": 4.9921812061393454e-06,
+      "loss": 0.4298,
+      "step": 4042
+    },
+    {
+      "epoch": 2.0054610897356335,
+      "grad_norm": 0.07811975249794761,
+      "learning_rate": 4.990226509914764e-06,
+      "loss": 0.4746,
+      "step": 4043
+    },
+    {
+      "epoch": 2.005957552438873,
+      "grad_norm": 0.08367632748009121,
+      "learning_rate": 4.9882718151839045e-06,
+      "loss": 0.4859,
+      "step": 4044
+    },
+    {
+      "epoch": 2.0064540151421126,
+      "grad_norm": 0.07739267038787806,
+      "learning_rate": 4.986317122245508e-06,
+      "loss": 0.4506,
+      "step": 4045
+    },
+    {
+      "epoch": 2.006950477845352,
+      "grad_norm": 0.07603906485692118,
+      "learning_rate": 4.984362431398324e-06,
+      "loss": 0.4648,
+      "step": 4046
+    },
+    {
+      "epoch": 2.0074469405485913,
+      "grad_norm": 0.07055244361688247,
+      "learning_rate": 4.9824077429410895e-06,
+      "loss": 0.4481,
+      "step": 4047
+    },
+    {
+      "epoch": 2.0079434032518306,
+      "grad_norm": 0.07596314574372579,
+      "learning_rate": 4.98045305717255e-06,
+      "loss": 0.4705,
+      "step": 4048
+    },
+    {
+      "epoch": 2.00843986595507,
+      "grad_norm": 0.07041469030095589,
+      "learning_rate": 4.978498374391446e-06,
+      "loss": 0.4667,
+      "step": 4049
+    },
+    {
+      "epoch": 2.0089363286583097,
+      "grad_norm": 0.07212901412219706,
+      "learning_rate": 4.976543694896521e-06,
+      "loss": 0.4696,
+      "step": 4050
+    },
+    {
+      "epoch": 2.009432791361549,
+      "grad_norm": 0.07731331158034714,
+      "learning_rate": 4.974589018986516e-06,
+      "loss": 0.4541,
+      "step": 4051
+    },
+    {
+      "epoch": 2.0099292540647884,
+      "grad_norm": 0.072220754341368,
+      "learning_rate": 4.972634346960173e-06,
+      "loss": 0.4462,
+      "step": 4052
+    },
+    {
+      "epoch": 2.0104257167680277,
+      "grad_norm": 0.07403397940617826,
+      "learning_rate": 4.97067967911623e-06,
+      "loss": 0.45,
+      "step": 4053
+    },
+    {
+      "epoch": 2.010922179471267,
+      "grad_norm": 0.0732440572549554,
+      "learning_rate": 4.968725015753429e-06,
+      "loss": 0.4322,
+      "step": 4054
+    },
+    {
+      "epoch": 2.011418642174507,
+      "grad_norm": 0.07231534439947165,
+      "learning_rate": 4.966770357170507e-06,
+      "loss": 0.4715,
+      "step": 4055
+    },
+    {
+      "epoch": 2.011915104877746,
+      "grad_norm": 0.0713353452872365,
+      "learning_rate": 4.9648157036662035e-06,
+      "loss": 0.4629,
+      "step": 4056
+    },
+    {
+      "epoch": 2.0124115675809855,
+      "grad_norm": 0.07330641016780033,
+      "learning_rate": 4.962861055539256e-06,
+      "loss": 0.4704,
+      "step": 4057
+    },
+    {
+      "epoch": 2.012908030284225,
+      "grad_norm": 0.07421035285517524,
+      "learning_rate": 4.9609064130884e-06,
+      "loss": 0.4709,
+      "step": 4058
+    },
+    {
+      "epoch": 2.013404492987464,
+      "grad_norm": 0.07083522565061746,
+      "learning_rate": 4.9589517766123725e-06,
+      "loss": 0.4426,
+      "step": 4059
+    },
+    {
+      "epoch": 2.013900955690704,
+      "grad_norm": 0.07256789965761337,
+      "learning_rate": 4.956997146409907e-06,
+      "loss": 0.453,
+      "step": 4060
+    },
+    {
+      "epoch": 2.0143974183939433,
+      "grad_norm": 0.07401292502068656,
+      "learning_rate": 4.955042522779739e-06,
+      "loss": 0.4725,
+      "step": 4061
+    },
+    {
+      "epoch": 2.0148938810971826,
+      "grad_norm": 0.07293397586128852,
+      "learning_rate": 4.953087906020601e-06,
+      "loss": 0.4625,
+      "step": 4062
+    },
+    {
+      "epoch": 2.015390343800422,
+      "grad_norm": 0.07198059298195802,
+      "learning_rate": 4.951133296431224e-06,
+      "loss": 0.4573,
+      "step": 4063
+    },
+    {
+      "epoch": 2.0158868065036613,
+      "grad_norm": 0.07547173794053258,
+      "learning_rate": 4.949178694310339e-06,
+      "loss": 0.4486,
+      "step": 4064
+    },
+    {
+      "epoch": 2.016383269206901,
+      "grad_norm": 0.07199537230690754,
+      "learning_rate": 4.9472240999566755e-06,
+      "loss": 0.4355,
+      "step": 4065
+    },
+    {
+      "epoch": 2.0168797319101404,
+      "grad_norm": 0.07316013752955937,
+      "learning_rate": 4.945269513668962e-06,
+      "loss": 0.4571,
+      "step": 4066
+    },
+    {
+      "epoch": 2.0173761946133797,
+      "grad_norm": 0.07005123491908063,
+      "learning_rate": 4.943314935745925e-06,
+      "loss": 0.4537,
+      "step": 4067
+    },
+    {
+      "epoch": 2.017872657316619,
+      "grad_norm": 0.07114454027164603,
+      "learning_rate": 4.941360366486294e-06,
+      "loss": 0.4365,
+      "step": 4068
+    },
+    {
+      "epoch": 2.0183691200198584,
+      "grad_norm": 0.0718199562081555,
+      "learning_rate": 4.939405806188788e-06,
+      "loss": 0.4993,
+      "step": 4069
+    },
+    {
+      "epoch": 2.018865582723098,
+      "grad_norm": 0.07727220254428632,
+      "learning_rate": 4.9374512551521335e-06,
+      "loss": 0.4716,
+      "step": 4070
+    },
+    {
+      "epoch": 2.0193620454263375,
+      "grad_norm": 0.07394436337606848,
+      "learning_rate": 4.935496713675052e-06,
+      "loss": 0.4612,
+      "step": 4071
+    },
+    {
+      "epoch": 2.019858508129577,
+      "grad_norm": 0.0713432047748325,
+      "learning_rate": 4.9335421820562635e-06,
+      "loss": 0.5075,
+      "step": 4072
+    },
+    {
+      "epoch": 2.020354970832816,
+      "grad_norm": 0.07297875830518837,
+      "learning_rate": 4.931587660594488e-06,
+      "loss": 0.4054,
+      "step": 4073
+    },
+    {
+      "epoch": 2.0208514335360555,
+      "grad_norm": 0.0742995803324647,
+      "learning_rate": 4.929633149588441e-06,
+      "loss": 0.4802,
+      "step": 4074
+    },
+    {
+      "epoch": 2.0213478962392952,
+      "grad_norm": 0.07235310266525152,
+      "learning_rate": 4.927678649336838e-06,
+      "loss": 0.4755,
+      "step": 4075
+    },
+    {
+      "epoch": 2.0218443589425346,
+      "grad_norm": 0.07469287262717912,
+      "learning_rate": 4.925724160138394e-06,
+      "loss": 0.4663,
+      "step": 4076
+    },
+    {
+      "epoch": 2.022340821645774,
+      "grad_norm": 0.07378999311999616,
+      "learning_rate": 4.923769682291822e-06,
+      "loss": 0.4718,
+      "step": 4077
+    },
+    {
+      "epoch": 2.022837284349013,
+      "grad_norm": 0.07371766271655966,
+      "learning_rate": 4.921815216095832e-06,
+      "loss": 0.4573,
+      "step": 4078
+    },
+    {
+      "epoch": 2.0233337470522526,
+      "grad_norm": 0.0818546782159551,
+      "learning_rate": 4.919860761849132e-06,
+      "loss": 0.4857,
+      "step": 4079
+    },
+    {
+      "epoch": 2.0238302097554923,
+      "grad_norm": 0.07398217930028606,
+      "learning_rate": 4.917906319850431e-06,
+      "loss": 0.4513,
+      "step": 4080
+    },
+    {
+      "epoch": 2.0243266724587317,
+      "grad_norm": 0.07234169176030968,
+      "learning_rate": 4.915951890398431e-06,
+      "loss": 0.4686,
+      "step": 4081
+    },
+    {
+      "epoch": 2.024823135161971,
+      "grad_norm": 0.07138272648374208,
+      "learning_rate": 4.913997473791837e-06,
+      "loss": 0.4842,
+      "step": 4082
+    },
+    {
+      "epoch": 2.0253195978652103,
+      "grad_norm": 0.07389413192702614,
+      "learning_rate": 4.9120430703293504e-06,
+      "loss": 0.4866,
+      "step": 4083
+    },
+    {
+      "epoch": 2.0258160605684497,
+      "grad_norm": 0.07146788366049606,
+      "learning_rate": 4.9100886803096696e-06,
+      "loss": 0.4646,
+      "step": 4084
+    },
+    {
+      "epoch": 2.0263125232716894,
+      "grad_norm": 0.07093367028537836,
+      "learning_rate": 4.908134304031495e-06,
+      "loss": 0.4583,
+      "step": 4085
+    },
+    {
+      "epoch": 2.0268089859749288,
+      "grad_norm": 0.07429306592375522,
+      "learning_rate": 4.906179941793516e-06,
+      "loss": 0.4953,
+      "step": 4086
+    },
+    {
+      "epoch": 2.027305448678168,
+      "grad_norm": 0.07360985294937847,
+      "learning_rate": 4.90422559389443e-06,
+      "loss": 0.4464,
+      "step": 4087
+    },
+    {
+      "epoch": 2.0278019113814074,
+      "grad_norm": 0.0718166852132005,
+      "learning_rate": 4.902271260632926e-06,
+      "loss": 0.457,
+      "step": 4088
+    },
+    {
+      "epoch": 2.0282983740846467,
+      "grad_norm": 0.07030065358374502,
+      "learning_rate": 4.900316942307693e-06,
+      "loss": 0.4384,
+      "step": 4089
+    },
+    {
+      "epoch": 2.0287948367878865,
+      "grad_norm": 0.07571066072847284,
+      "learning_rate": 4.898362639217417e-06,
+      "loss": 0.4567,
+      "step": 4090
+    },
+    {
+      "epoch": 2.029291299491126,
+      "grad_norm": 0.07956518040265176,
+      "learning_rate": 4.896408351660783e-06,
+      "loss": 0.4577,
+      "step": 4091
+    },
+    {
+      "epoch": 2.029787762194365,
+      "grad_norm": 0.0742603702620764,
+      "learning_rate": 4.89445407993647e-06,
+      "loss": 0.45,
+      "step": 4092
+    },
+    {
+      "epoch": 2.0302842248976045,
+      "grad_norm": 0.07054575013845489,
+      "learning_rate": 4.892499824343159e-06,
+      "loss": 0.457,
+      "step": 4093
+    },
+    {
+      "epoch": 2.030780687600844,
+      "grad_norm": 0.07252488643464008,
+      "learning_rate": 4.890545585179527e-06,
+      "loss": 0.4622,
+      "step": 4094
+    },
+    {
+      "epoch": 2.0312771503040836,
+      "grad_norm": 0.07143595162426333,
+      "learning_rate": 4.888591362744247e-06,
+      "loss": 0.4477,
+      "step": 4095
+    },
+    {
+      "epoch": 2.031773613007323,
+      "grad_norm": 0.07159588677019353,
+      "learning_rate": 4.886637157335992e-06,
+      "loss": 0.4622,
+      "step": 4096
+    },
+    {
+      "epoch": 2.0322700757105623,
+      "grad_norm": 0.07040447114523164,
+      "learning_rate": 4.884682969253428e-06,
+      "loss": 0.4579,
+      "step": 4097
+    },
+    {
+      "epoch": 2.0327665384138016,
+      "grad_norm": 0.07411670769151289,
+      "learning_rate": 4.882728798795223e-06,
+      "loss": 0.4516,
+      "step": 4098
+    },
+    {
+      "epoch": 2.033263001117041,
+      "grad_norm": 0.06959189954785697,
+      "learning_rate": 4.880774646260041e-06,
+      "loss": 0.4221,
+      "step": 4099
+    },
+    {
+      "epoch": 2.0337594638202807,
+      "grad_norm": 0.07199309706450548,
+      "learning_rate": 4.878820511946543e-06,
+      "loss": 0.4712,
+      "step": 4100
+    },
+    {
+      "epoch": 2.03425592652352,
+      "grad_norm": 0.07325495710301079,
+      "learning_rate": 4.876866396153388e-06,
+      "loss": 0.4794,
+      "step": 4101
+    },
+    {
+      "epoch": 2.0347523892267594,
+      "grad_norm": 0.0717853143430827,
+      "learning_rate": 4.874912299179228e-06,
+      "loss": 0.4739,
+      "step": 4102
+    },
+    {
+      "epoch": 2.0352488519299987,
+      "grad_norm": 0.07407765387786804,
+      "learning_rate": 4.872958221322719e-06,
+      "loss": 0.4842,
+      "step": 4103
+    },
+    {
+      "epoch": 2.035745314633238,
+      "grad_norm": 0.0694922585009176,
+      "learning_rate": 4.871004162882508e-06,
+      "loss": 0.4404,
+      "step": 4104
+    },
+    {
+      "epoch": 2.0362417773364774,
+      "grad_norm": 0.07157290293449331,
+      "learning_rate": 4.869050124157244e-06,
+      "loss": 0.4553,
+      "step": 4105
+    },
+    {
+      "epoch": 2.036738240039717,
+      "grad_norm": 0.07266934587376413,
+      "learning_rate": 4.8670961054455666e-06,
+      "loss": 0.4485,
+      "step": 4106
+    },
+    {
+      "epoch": 2.0372347027429565,
+      "grad_norm": 0.071777351529913,
+      "learning_rate": 4.86514210704612e-06,
+      "loss": 0.4649,
+      "step": 4107
+    },
+    {
+      "epoch": 2.037731165446196,
+      "grad_norm": 0.07503740354924117,
+      "learning_rate": 4.863188129257539e-06,
+      "loss": 0.4373,
+      "step": 4108
+    },
+    {
+      "epoch": 2.038227628149435,
+      "grad_norm": 0.07386719941049365,
+      "learning_rate": 4.8612341723784586e-06,
+      "loss": 0.4776,
+      "step": 4109
+    },
+    {
+      "epoch": 2.0387240908526745,
+      "grad_norm": 0.07248900897774578,
+      "learning_rate": 4.859280236707512e-06,
+      "loss": 0.452,
+      "step": 4110
+    },
+    {
+      "epoch": 2.0392205535559143,
+      "grad_norm": 0.07179491353728072,
+      "learning_rate": 4.857326322543325e-06,
+      "loss": 0.4478,
+      "step": 4111
+    },
+    {
+      "epoch": 2.0397170162591536,
+      "grad_norm": 0.07174429454012216,
+      "learning_rate": 4.855372430184522e-06,
+      "loss": 0.4343,
+      "step": 4112
+    },
+    {
+      "epoch": 2.040213478962393,
+      "grad_norm": 0.07519278914658567,
+      "learning_rate": 4.853418559929722e-06,
+      "loss": 0.474,
+      "step": 4113
+    },
+    {
+      "epoch": 2.0407099416656322,
+      "grad_norm": 0.07138808068817071,
+      "learning_rate": 4.851464712077546e-06,
+      "loss": 0.4454,
+      "step": 4114
+    },
+    {
+      "epoch": 2.0412064043688716,
+      "grad_norm": 0.07314394649356787,
+      "learning_rate": 4.849510886926606e-06,
+      "loss": 0.4659,
+      "step": 4115
+    },
+    {
+      "epoch": 2.0417028670721113,
+      "grad_norm": 0.07387286339499641,
+      "learning_rate": 4.847557084775515e-06,
+      "loss": 0.4686,
+      "step": 4116
+    },
+    {
+      "epoch": 2.0421993297753507,
+      "grad_norm": 0.0707490447383383,
+      "learning_rate": 4.84560330592288e-06,
+      "loss": 0.5006,
+      "step": 4117
+    },
+    {
+      "epoch": 2.04269579247859,
+      "grad_norm": 0.07012577898465304,
+      "learning_rate": 4.843649550667304e-06,
+      "loss": 0.4308,
+      "step": 4118
+    },
+    {
+      "epoch": 2.0431922551818293,
+      "grad_norm": 0.07275514929361539,
+      "learning_rate": 4.841695819307386e-06,
+      "loss": 0.4929,
+      "step": 4119
+    },
+    {
+      "epoch": 2.0436887178850687,
+      "grad_norm": 0.07248163580645842,
+      "learning_rate": 4.839742112141725e-06,
+      "loss": 0.4415,
+      "step": 4120
+    },
+    {
+      "epoch": 2.0441851805883084,
+      "grad_norm": 0.07201595951270065,
+      "learning_rate": 4.8377884294689114e-06,
+      "loss": 0.4482,
+      "step": 4121
+    },
+    {
+      "epoch": 2.044681643291548,
+      "grad_norm": 0.07366828864299017,
+      "learning_rate": 4.835834771587537e-06,
+      "loss": 0.4547,
+      "step": 4122
+    },
+    {
+      "epoch": 2.045178105994787,
+      "grad_norm": 0.07590040021361757,
+      "learning_rate": 4.833881138796185e-06,
+      "loss": 0.4332,
+      "step": 4123
+    },
+    {
+      "epoch": 2.0456745686980264,
+      "grad_norm": 0.07052536062844234,
+      "learning_rate": 4.831927531393437e-06,
+      "loss": 0.4399,
+      "step": 4124
+    },
+    {
+      "epoch": 2.0461710314012658,
+      "grad_norm": 0.07317965073722368,
+      "learning_rate": 4.829973949677871e-06,
+      "loss": 0.4596,
+      "step": 4125
+    },
+    {
+      "epoch": 2.0466674941045055,
+      "grad_norm": 0.07648174917490391,
+      "learning_rate": 4.828020393948061e-06,
+      "loss": 0.4876,
+      "step": 4126
+    },
+    {
+      "epoch": 2.047163956807745,
+      "grad_norm": 0.07241753151580768,
+      "learning_rate": 4.826066864502578e-06,
+      "loss": 0.4543,
+      "step": 4127
+    },
+    {
+      "epoch": 2.047660419510984,
+      "grad_norm": 0.07405990811154906,
+      "learning_rate": 4.824113361639984e-06,
+      "loss": 0.4693,
+      "step": 4128
+    },
+    {
+      "epoch": 2.0481568822142235,
+      "grad_norm": 0.07489984157760932,
+      "learning_rate": 4.822159885658843e-06,
+      "loss": 0.432,
+      "step": 4129
+    },
+    {
+      "epoch": 2.048653344917463,
+      "grad_norm": 0.06986823081594058,
+      "learning_rate": 4.8202064368577116e-06,
+      "loss": 0.4457,
+      "step": 4130
+    },
+    {
+      "epoch": 2.0491498076207026,
+      "grad_norm": 0.07234328790911547,
+      "learning_rate": 4.818253015535143e-06,
+      "loss": 0.4661,
+      "step": 4131
+    },
+    {
+      "epoch": 2.049646270323942,
+      "grad_norm": 0.07155335491576369,
+      "learning_rate": 4.816299621989687e-06,
+      "loss": 0.4741,
+      "step": 4132
+    },
+    {
+      "epoch": 2.0501427330271813,
+      "grad_norm": 0.07599145080679516,
+      "learning_rate": 4.81434625651989e-06,
+      "loss": 0.4851,
+      "step": 4133
+    },
+    {
+      "epoch": 2.0506391957304206,
+      "grad_norm": 0.07201140956511509,
+      "learning_rate": 4.81239291942429e-06,
+      "loss": 0.4552,
+      "step": 4134
+    },
+    {
+      "epoch": 2.05113565843366,
+      "grad_norm": 0.07298263836130801,
+      "learning_rate": 4.810439611001423e-06,
+      "loss": 0.4642,
+      "step": 4135
+    },
+    {
+      "epoch": 2.0516321211368997,
+      "grad_norm": 0.07558026313940648,
+      "learning_rate": 4.808486331549824e-06,
+      "loss": 0.4708,
+      "step": 4136
+    },
+    {
+      "epoch": 2.052128583840139,
+      "grad_norm": 0.07086733181913618,
+      "learning_rate": 4.806533081368017e-06,
+      "loss": 0.4487,
+      "step": 4137
+    },
+    {
+      "epoch": 2.0526250465433784,
+      "grad_norm": 0.07450006007010589,
+      "learning_rate": 4.804579860754528e-06,
+      "loss": 0.4851,
+      "step": 4138
+    },
+    {
+      "epoch": 2.0531215092466177,
+      "grad_norm": 0.07069277786323414,
+      "learning_rate": 4.802626670007873e-06,
+      "loss": 0.434,
+      "step": 4139
+    },
+    {
+      "epoch": 2.053617971949857,
+      "grad_norm": 0.06861088663923186,
+      "learning_rate": 4.800673509426567e-06,
+      "loss": 0.4435,
+      "step": 4140
+    },
+    {
+      "epoch": 2.054114434653097,
+      "grad_norm": 0.07350117652911003,
+      "learning_rate": 4.7987203793091186e-06,
+      "loss": 0.4843,
+      "step": 4141
+    },
+    {
+      "epoch": 2.054610897356336,
+      "grad_norm": 0.07202045677553137,
+      "learning_rate": 4.796767279954034e-06,
+      "loss": 0.4352,
+      "step": 4142
+    },
+    {
+      "epoch": 2.0551073600595755,
+      "grad_norm": 0.07171344406046876,
+      "learning_rate": 4.794814211659813e-06,
+      "loss": 0.4538,
+      "step": 4143
+    },
+    {
+      "epoch": 2.055603822762815,
+      "grad_norm": 0.07280878355114719,
+      "learning_rate": 4.7928611747249495e-06,
+      "loss": 0.4429,
+      "step": 4144
+    },
+    {
+      "epoch": 2.056100285466054,
+      "grad_norm": 0.07357463543017809,
+      "learning_rate": 4.790908169447933e-06,
+      "loss": 0.4816,
+      "step": 4145
+    },
+    {
+      "epoch": 2.056596748169294,
+      "grad_norm": 0.07156764439656021,
+      "learning_rate": 4.78895519612725e-06,
+      "loss": 0.458,
+      "step": 4146
+    },
+    {
+      "epoch": 2.0570932108725333,
+      "grad_norm": 0.07225385651619044,
+      "learning_rate": 4.787002255061381e-06,
+      "loss": 0.4581,
+      "step": 4147
+    },
+    {
+      "epoch": 2.0575896735757726,
+      "grad_norm": 0.07084771506549391,
+      "learning_rate": 4.785049346548803e-06,
+      "loss": 0.4373,
+      "step": 4148
+    },
+    {
+      "epoch": 2.058086136279012,
+      "grad_norm": 0.07376134194351902,
+      "learning_rate": 4.783096470887987e-06,
+      "loss": 0.4264,
+      "step": 4149
+    },
+    {
+      "epoch": 2.0585825989822513,
+      "grad_norm": 0.07512222501568963,
+      "learning_rate": 4.781143628377396e-06,
+      "loss": 0.4394,
+      "step": 4150
+    },
+    {
+      "epoch": 2.059079061685491,
+      "grad_norm": 0.07189691667984506,
+      "learning_rate": 4.779190819315493e-06,
+      "loss": 0.4604,
+      "step": 4151
+    },
+    {
+      "epoch": 2.0595755243887304,
+      "grad_norm": 0.07219890185194544,
+      "learning_rate": 4.777238044000732e-06,
+      "loss": 0.4514,
+      "step": 4152
+    },
+    {
+      "epoch": 2.0600719870919697,
+      "grad_norm": 0.07489990717793613,
+      "learning_rate": 4.775285302731565e-06,
+      "loss": 0.4493,
+      "step": 4153
+    },
+    {
+      "epoch": 2.060568449795209,
+      "grad_norm": 0.07554161871208594,
+      "learning_rate": 4.773332595806437e-06,
+      "loss": 0.4458,
+      "step": 4154
+    },
+    {
+      "epoch": 2.0610649124984484,
+      "grad_norm": 0.0720993346206349,
+      "learning_rate": 4.771379923523786e-06,
+      "loss": 0.4399,
+      "step": 4155
+    },
+    {
+      "epoch": 2.061561375201688,
+      "grad_norm": 0.07289060496223992,
+      "learning_rate": 4.76942728618205e-06,
+      "loss": 0.4335,
+      "step": 4156
+    },
+    {
+      "epoch": 2.0620578379049275,
+      "grad_norm": 0.07282969468184067,
+      "learning_rate": 4.767474684079655e-06,
+      "loss": 0.4732,
+      "step": 4157
+    },
+    {
+      "epoch": 2.062554300608167,
+      "grad_norm": 0.07354514479733838,
+      "learning_rate": 4.765522117515026e-06,
+      "loss": 0.4434,
+      "step": 4158
+    },
+    {
+      "epoch": 2.063050763311406,
+      "grad_norm": 0.07316254525944305,
+      "learning_rate": 4.763569586786584e-06,
+      "loss": 0.4721,
+      "step": 4159
+    },
+    {
+      "epoch": 2.0635472260146455,
+      "grad_norm": 0.07388552565859352,
+      "learning_rate": 4.761617092192737e-06,
+      "loss": 0.447,
+      "step": 4160
+    },
+    {
+      "epoch": 2.0640436887178852,
+      "grad_norm": 0.07238240778799938,
+      "learning_rate": 4.759664634031897e-06,
+      "loss": 0.4706,
+      "step": 4161
+    },
+    {
+      "epoch": 2.0645401514211246,
+      "grad_norm": 0.07372646052049893,
+      "learning_rate": 4.757712212602464e-06,
+      "loss": 0.4518,
+      "step": 4162
+    },
+    {
+      "epoch": 2.065036614124364,
+      "grad_norm": 0.07457439512766961,
+      "learning_rate": 4.755759828202834e-06,
+      "loss": 0.4712,
+      "step": 4163
+    },
+    {
+      "epoch": 2.0655330768276032,
+      "grad_norm": 0.07399410722838877,
+      "learning_rate": 4.753807481131398e-06,
+      "loss": 0.4604,
+      "step": 4164
+    },
+    {
+      "epoch": 2.0660295395308426,
+      "grad_norm": 0.07082211955770341,
+      "learning_rate": 4.751855171686542e-06,
+      "loss": 0.4498,
+      "step": 4165
+    },
+    {
+      "epoch": 2.0665260022340823,
+      "grad_norm": 0.07398347450665645,
+      "learning_rate": 4.7499029001666435e-06,
+      "loss": 0.464,
+      "step": 4166
+    },
+    {
+      "epoch": 2.0670224649373217,
+      "grad_norm": 0.07175179899612245,
+      "learning_rate": 4.747950666870076e-06,
+      "loss": 0.45,
+      "step": 4167
+    },
+    {
+      "epoch": 2.067518927640561,
+      "grad_norm": 0.0725543113621323,
+      "learning_rate": 4.745998472095209e-06,
+      "loss": 0.4459,
+      "step": 4168
+    },
+    {
+      "epoch": 2.0680153903438003,
+      "grad_norm": 0.07356433723103205,
+      "learning_rate": 4.744046316140402e-06,
+      "loss": 0.4871,
+      "step": 4169
+    },
+    {
+      "epoch": 2.0685118530470397,
+      "grad_norm": 0.07077237681782449,
+      "learning_rate": 4.742094199304013e-06,
+      "loss": 0.4189,
+      "step": 4170
+    },
+    {
+      "epoch": 2.0690083157502794,
+      "grad_norm": 0.07397114345236917,
+      "learning_rate": 4.74014212188439e-06,
+      "loss": 0.4607,
+      "step": 4171
+    },
+    {
+      "epoch": 2.0695047784535188,
+      "grad_norm": 0.07633194862451866,
+      "learning_rate": 4.738190084179876e-06,
+      "loss": 0.4775,
+      "step": 4172
+    },
+    {
+      "epoch": 2.070001241156758,
+      "grad_norm": 0.0742711453109311,
+      "learning_rate": 4.73623808648881e-06,
+      "loss": 0.4769,
+      "step": 4173
+    },
+    {
+      "epoch": 2.0704977038599974,
+      "grad_norm": 0.07170333053549788,
+      "learning_rate": 4.7342861291095244e-06,
+      "loss": 0.4608,
+      "step": 4174
+    },
+    {
+      "epoch": 2.0709941665632368,
+      "grad_norm": 0.0748163728180664,
+      "learning_rate": 4.732334212340345e-06,
+      "loss": 0.4928,
+      "step": 4175
+    },
+    {
+      "epoch": 2.0714906292664765,
+      "grad_norm": 0.07541859685661392,
+      "learning_rate": 4.7303823364795874e-06,
+      "loss": 0.4681,
+      "step": 4176
+    },
+    {
+      "epoch": 2.071987091969716,
+      "grad_norm": 0.0733725681131633,
+      "learning_rate": 4.728430501825567e-06,
+      "loss": 0.4728,
+      "step": 4177
+    },
+    {
+      "epoch": 2.072483554672955,
+      "grad_norm": 0.07409349750433514,
+      "learning_rate": 4.72647870867659e-06,
+      "loss": 0.4497,
+      "step": 4178
+    },
+    {
+      "epoch": 2.0729800173761945,
+      "grad_norm": 0.07399731870180891,
+      "learning_rate": 4.7245269573309575e-06,
+      "loss": 0.4831,
+      "step": 4179
+    },
+    {
+      "epoch": 2.073476480079434,
+      "grad_norm": 0.07655395352644413,
+      "learning_rate": 4.722575248086962e-06,
+      "loss": 0.4491,
+      "step": 4180
+    },
+    {
+      "epoch": 2.0739729427826736,
+      "grad_norm": 0.0749787725080781,
+      "learning_rate": 4.720623581242893e-06,
+      "loss": 0.4489,
+      "step": 4181
+    },
+    {
+      "epoch": 2.074469405485913,
+      "grad_norm": 0.07444718116125149,
+      "learning_rate": 4.7186719570970285e-06,
+      "loss": 0.4766,
+      "step": 4182
+    },
+    {
+      "epoch": 2.0749658681891523,
+      "grad_norm": 0.07366724823253058,
+      "learning_rate": 4.716720375947644e-06,
+      "loss": 0.463,
+      "step": 4183
+    },
+    {
+      "epoch": 2.0754623308923916,
+      "grad_norm": 0.0713571217812567,
+      "learning_rate": 4.714768838093009e-06,
+      "loss": 0.449,
+      "step": 4184
+    },
+    {
+      "epoch": 2.075958793595631,
+      "grad_norm": 0.07236983736508307,
+      "learning_rate": 4.712817343831384e-06,
+      "loss": 0.4451,
+      "step": 4185
+    },
+    {
+      "epoch": 2.0764552562988707,
+      "grad_norm": 0.07472833997107589,
+      "learning_rate": 4.710865893461024e-06,
+      "loss": 0.4833,
+      "step": 4186
+    },
+    {
+      "epoch": 2.07695171900211,
+      "grad_norm": 0.07565048657385896,
+      "learning_rate": 4.708914487280175e-06,
+      "loss": 0.4785,
+      "step": 4187
+    },
+    {
+      "epoch": 2.0774481817053494,
+      "grad_norm": 0.07173844405137295,
+      "learning_rate": 4.706963125587079e-06,
+      "loss": 0.4531,
+      "step": 4188
+    },
+    {
+      "epoch": 2.0779446444085887,
+      "grad_norm": 0.07287877873442934,
+      "learning_rate": 4.705011808679972e-06,
+      "loss": 0.4379,
+      "step": 4189
+    },
+    {
+      "epoch": 2.078441107111828,
+      "grad_norm": 0.0716401851616967,
+      "learning_rate": 4.7030605368570795e-06,
+      "loss": 0.4344,
+      "step": 4190
+    },
+    {
+      "epoch": 2.078937569815068,
+      "grad_norm": 0.0713923972030575,
+      "learning_rate": 4.701109310416626e-06,
+      "loss": 0.4426,
+      "step": 4191
+    },
+    {
+      "epoch": 2.079434032518307,
+      "grad_norm": 0.07268447504285887,
+      "learning_rate": 4.699158129656818e-06,
+      "loss": 0.4577,
+      "step": 4192
+    },
+    {
+      "epoch": 2.0799304952215465,
+      "grad_norm": 0.06999453896271132,
+      "learning_rate": 4.697206994875869e-06,
+      "loss": 0.4345,
+      "step": 4193
+    },
+    {
+      "epoch": 2.080426957924786,
+      "grad_norm": 0.07212199517441456,
+      "learning_rate": 4.695255906371974e-06,
+      "loss": 0.4577,
+      "step": 4194
+    },
+    {
+      "epoch": 2.080923420628025,
+      "grad_norm": 0.07328964439731311,
+      "learning_rate": 4.693304864443328e-06,
+      "loss": 0.4832,
+      "step": 4195
+    },
+    {
+      "epoch": 2.081419883331265,
+      "grad_norm": 0.07344622598547146,
+      "learning_rate": 4.6913538693881155e-06,
+      "loss": 0.4568,
+      "step": 4196
+    },
+    {
+      "epoch": 2.0819163460345043,
+      "grad_norm": 0.07094064393577482,
+      "learning_rate": 4.689402921504516e-06,
+      "loss": 0.4739,
+      "step": 4197
+    },
+    {
+      "epoch": 2.0824128087377436,
+      "grad_norm": 0.07297562583866672,
+      "learning_rate": 4.6874520210907e-06,
+      "loss": 0.4617,
+      "step": 4198
+    },
+    {
+      "epoch": 2.082909271440983,
+      "grad_norm": 0.07116335621343138,
+      "learning_rate": 4.685501168444831e-06,
+      "loss": 0.4518,
+      "step": 4199
+    },
+    {
+      "epoch": 2.0834057341442223,
+      "grad_norm": 0.07456698856493216,
+      "learning_rate": 4.683550363865065e-06,
+      "loss": 0.4774,
+      "step": 4200
+    },
+    {
+      "epoch": 2.083902196847462,
+      "grad_norm": 0.06963389093636772,
+      "learning_rate": 4.681599607649553e-06,
+      "loss": 0.4249,
+      "step": 4201
+    },
+    {
+      "epoch": 2.0843986595507014,
+      "grad_norm": 0.07174632242610744,
+      "learning_rate": 4.679648900096436e-06,
+      "loss": 0.4459,
+      "step": 4202
+    },
+    {
+      "epoch": 2.0848951222539407,
+      "grad_norm": 0.07098206575623335,
+      "learning_rate": 4.677698241503847e-06,
+      "loss": 0.4341,
+      "step": 4203
+    },
+    {
+      "epoch": 2.08539158495718,
+      "grad_norm": 0.06976961364064482,
+      "learning_rate": 4.675747632169916e-06,
+      "loss": 0.4402,
+      "step": 4204
+    },
+    {
+      "epoch": 2.0858880476604194,
+      "grad_norm": 0.07452720372868127,
+      "learning_rate": 4.673797072392759e-06,
+      "loss": 0.4773,
+      "step": 4205
+    },
+    {
+      "epoch": 2.086384510363659,
+      "grad_norm": 0.07246580630946493,
+      "learning_rate": 4.671846562470489e-06,
+      "loss": 0.44,
+      "step": 4206
+    },
+    {
+      "epoch": 2.0868809730668985,
+      "grad_norm": 0.0757655362478522,
+      "learning_rate": 4.669896102701212e-06,
+      "loss": 0.4277,
+      "step": 4207
+    },
+    {
+      "epoch": 2.087377435770138,
+      "grad_norm": 0.07543467059805738,
+      "learning_rate": 4.667945693383022e-06,
+      "loss": 0.4902,
+      "step": 4208
+    },
+    {
+      "epoch": 2.087873898473377,
+      "grad_norm": 0.07464446756126804,
+      "learning_rate": 4.665995334814009e-06,
+      "loss": 0.4925,
+      "step": 4209
+    },
+    {
+      "epoch": 2.0883703611766165,
+      "grad_norm": 0.07342049375383967,
+      "learning_rate": 4.664045027292252e-06,
+      "loss": 0.4497,
+      "step": 4210
+    },
+    {
+      "epoch": 2.0888668238798562,
+      "grad_norm": 0.07066862656094619,
+      "learning_rate": 4.662094771115828e-06,
+      "loss": 0.4516,
+      "step": 4211
+    },
+    {
+      "epoch": 2.0893632865830956,
+      "grad_norm": 0.07338215598137313,
+      "learning_rate": 4.660144566582799e-06,
+      "loss": 0.4427,
+      "step": 4212
+    },
+    {
+      "epoch": 2.089859749286335,
+      "grad_norm": 0.07355061364313716,
+      "learning_rate": 4.658194413991224e-06,
+      "loss": 0.4551,
+      "step": 4213
+    },
+    {
+      "epoch": 2.090356211989574,
+      "grad_norm": 0.07318149318093226,
+      "learning_rate": 4.656244313639153e-06,
+      "loss": 0.4288,
+      "step": 4214
+    },
+    {
+      "epoch": 2.0908526746928136,
+      "grad_norm": 0.07415826471442223,
+      "learning_rate": 4.654294265824627e-06,
+      "loss": 0.5057,
+      "step": 4215
+    },
+    {
+      "epoch": 2.0913491373960533,
+      "grad_norm": 0.07366828306193732,
+      "learning_rate": 4.652344270845679e-06,
+      "loss": 0.4383,
+      "step": 4216
+    },
+    {
+      "epoch": 2.0918456000992927,
+      "grad_norm": 0.07147693654112407,
+      "learning_rate": 4.650394329000334e-06,
+      "loss": 0.4739,
+      "step": 4217
+    },
+    {
+      "epoch": 2.092342062802532,
+      "grad_norm": 0.07286515685522153,
+      "learning_rate": 4.648444440586612e-06,
+      "loss": 0.4832,
+      "step": 4218
+    },
+    {
+      "epoch": 2.0928385255057713,
+      "grad_norm": 0.07170150457742781,
+      "learning_rate": 4.6464946059025205e-06,
+      "loss": 0.4367,
+      "step": 4219
+    },
+    {
+      "epoch": 2.0933349882090106,
+      "grad_norm": 0.07026426443694084,
+      "learning_rate": 4.644544825246059e-06,
+      "loss": 0.4491,
+      "step": 4220
+    },
+    {
+      "epoch": 2.0938314509122504,
+      "grad_norm": 0.07262877550044473,
+      "learning_rate": 4.6425950989152225e-06,
+      "loss": 0.4325,
+      "step": 4221
+    },
+    {
+      "epoch": 2.0943279136154898,
+      "grad_norm": 0.07352269126408716,
+      "learning_rate": 4.640645427207995e-06,
+      "loss": 0.4918,
+      "step": 4222
+    },
+    {
+      "epoch": 2.094824376318729,
+      "grad_norm": 0.0768702240002514,
+      "learning_rate": 4.638695810422353e-06,
+      "loss": 0.5017,
+      "step": 4223
+    },
+    {
+      "epoch": 2.0953208390219684,
+      "grad_norm": 0.0721775418798943,
+      "learning_rate": 4.6367462488562614e-06,
+      "loss": 0.4525,
+      "step": 4224
+    },
+    {
+      "epoch": 2.0958173017252077,
+      "grad_norm": 0.07251926836651529,
+      "learning_rate": 4.634796742807683e-06,
+      "loss": 0.4445,
+      "step": 4225
+    },
+    {
+      "epoch": 2.0963137644284475,
+      "grad_norm": 0.07281659816445292,
+      "learning_rate": 4.632847292574565e-06,
+      "loss": 0.4292,
+      "step": 4226
+    },
+    {
+      "epoch": 2.096810227131687,
+      "grad_norm": 0.07524695595787657,
+      "learning_rate": 4.630897898454852e-06,
+      "loss": 0.4568,
+      "step": 4227
+    },
+    {
+      "epoch": 2.097306689834926,
+      "grad_norm": 0.07048486967770481,
+      "learning_rate": 4.628948560746477e-06,
+      "loss": 0.4311,
+      "step": 4228
+    },
+    {
+      "epoch": 2.0978031525381655,
+      "grad_norm": 0.07302922412326243,
+      "learning_rate": 4.6269992797473655e-06,
+      "loss": 0.4369,
+      "step": 4229
+    },
+    {
+      "epoch": 2.098299615241405,
+      "grad_norm": 0.07451648437833568,
+      "learning_rate": 4.6250500557554346e-06,
+      "loss": 0.4324,
+      "step": 4230
+    },
+    {
+      "epoch": 2.0987960779446446,
+      "grad_norm": 0.07491305988077178,
+      "learning_rate": 4.623100889068589e-06,
+      "loss": 0.4622,
+      "step": 4231
+    },
+    {
+      "epoch": 2.099292540647884,
+      "grad_norm": 0.07198956615781928,
+      "learning_rate": 4.6211517799847304e-06,
+      "loss": 0.447,
+      "step": 4232
+    },
+    {
+      "epoch": 2.0997890033511233,
+      "grad_norm": 0.07299965844460025,
+      "learning_rate": 4.6192027288017475e-06,
+      "loss": 0.4849,
+      "step": 4233
+    },
+    {
+      "epoch": 2.1002854660543626,
+      "grad_norm": 0.07247630851840142,
+      "learning_rate": 4.617253735817522e-06,
+      "loss": 0.4775,
+      "step": 4234
+    },
+    {
+      "epoch": 2.100781928757602,
+      "grad_norm": 0.07203360554721318,
+      "learning_rate": 4.6153048013299266e-06,
+      "loss": 0.4304,
+      "step": 4235
+    },
+    {
+      "epoch": 2.1012783914608413,
+      "grad_norm": 0.07326018253851636,
+      "learning_rate": 4.613355925636823e-06,
+      "loss": 0.4263,
+      "step": 4236
+    },
+    {
+      "epoch": 2.101774854164081,
+      "grad_norm": 0.07376985663412329,
+      "learning_rate": 4.611407109036067e-06,
+      "loss": 0.4495,
+      "step": 4237
+    },
+    {
+      "epoch": 2.1022713168673204,
+      "grad_norm": 0.0692281617188793,
+      "learning_rate": 4.609458351825505e-06,
+      "loss": 0.424,
+      "step": 4238
+    },
+    {
+      "epoch": 2.1027677795705597,
+      "grad_norm": 0.07246012599168375,
+      "learning_rate": 4.607509654302973e-06,
+      "loss": 0.4617,
+      "step": 4239
+    },
+    {
+      "epoch": 2.103264242273799,
+      "grad_norm": 0.07411607321908148,
+      "learning_rate": 4.605561016766295e-06,
+      "loss": 0.4908,
+      "step": 4240
+    },
+    {
+      "epoch": 2.103760704977039,
+      "grad_norm": 0.07717174587548606,
+      "learning_rate": 4.603612439513293e-06,
+      "loss": 0.4774,
+      "step": 4241
+    },
+    {
+      "epoch": 2.104257167680278,
+      "grad_norm": 0.07097666906897472,
+      "learning_rate": 4.6016639228417726e-06,
+      "loss": 0.4691,
+      "step": 4242
+    },
+    {
+      "epoch": 2.1047536303835175,
+      "grad_norm": 0.07450003788778216,
+      "learning_rate": 4.599715467049534e-06,
+      "loss": 0.4802,
+      "step": 4243
+    },
+    {
+      "epoch": 2.105250093086757,
+      "grad_norm": 0.0742256173702752,
+      "learning_rate": 4.597767072434369e-06,
+      "loss": 0.4748,
+      "step": 4244
+    },
+    {
+      "epoch": 2.105746555789996,
+      "grad_norm": 0.07224204359839762,
+      "learning_rate": 4.595818739294058e-06,
+      "loss": 0.4576,
+      "step": 4245
+    },
+    {
+      "epoch": 2.1062430184932355,
+      "grad_norm": 0.07318531262067612,
+      "learning_rate": 4.593870467926373e-06,
+      "loss": 0.4626,
+      "step": 4246
+    },
+    {
+      "epoch": 2.1067394811964752,
+      "grad_norm": 0.07430147251059871,
+      "learning_rate": 4.591922258629074e-06,
+      "loss": 0.4571,
+      "step": 4247
+    },
+    {
+      "epoch": 2.1072359438997146,
+      "grad_norm": 0.07254417646526728,
+      "learning_rate": 4.589974111699914e-06,
+      "loss": 0.4794,
+      "step": 4248
+    },
+    {
+      "epoch": 2.107732406602954,
+      "grad_norm": 0.07515164952717794,
+      "learning_rate": 4.5880260274366375e-06,
+      "loss": 0.4692,
+      "step": 4249
+    },
+    {
+      "epoch": 2.1082288693061932,
+      "grad_norm": 0.07383658645149892,
+      "learning_rate": 4.586078006136977e-06,
+      "loss": 0.4874,
+      "step": 4250
+    },
+    {
+      "epoch": 2.108725332009433,
+      "grad_norm": 0.07257460141396958,
+      "learning_rate": 4.584130048098658e-06,
+      "loss": 0.4819,
+      "step": 4251
+    },
+    {
+      "epoch": 2.1092217947126723,
+      "grad_norm": 0.07337974249936584,
+      "learning_rate": 4.582182153619392e-06,
+      "loss": 0.4706,
+      "step": 4252
+    },
+    {
+      "epoch": 2.1097182574159117,
+      "grad_norm": 0.07199015701301469,
+      "learning_rate": 4.5802343229968845e-06,
+      "loss": 0.4666,
+      "step": 4253
+    },
+    {
+      "epoch": 2.110214720119151,
+      "grad_norm": 0.07286868193978167,
+      "learning_rate": 4.5782865565288296e-06,
+      "loss": 0.436,
+      "step": 4254
+    },
+    {
+      "epoch": 2.1107111828223903,
+      "grad_norm": 0.06980078227341749,
+      "learning_rate": 4.576338854512916e-06,
+      "loss": 0.446,
+      "step": 4255
+    },
+    {
+      "epoch": 2.1112076455256297,
+      "grad_norm": 0.07755006219874234,
+      "learning_rate": 4.574391217246813e-06,
+      "loss": 0.4662,
+      "step": 4256
+    },
+    {
+      "epoch": 2.1117041082288694,
+      "grad_norm": 0.07399265860382301,
+      "learning_rate": 4.5724436450281895e-06,
+      "loss": 0.4495,
+      "step": 4257
+    },
+    {
+      "epoch": 2.112200570932109,
+      "grad_norm": 0.0746046076715519,
+      "learning_rate": 4.570496138154699e-06,
+      "loss": 0.4606,
+      "step": 4258
+    },
+    {
+      "epoch": 2.112697033635348,
+      "grad_norm": 0.0716888428047803,
+      "learning_rate": 4.568548696923986e-06,
+      "loss": 0.4414,
+      "step": 4259
+    },
+    {
+      "epoch": 2.1131934963385874,
+      "grad_norm": 0.06988503629826938,
+      "learning_rate": 4.566601321633688e-06,
+      "loss": 0.4656,
+      "step": 4260
+    },
+    {
+      "epoch": 2.1136899590418268,
+      "grad_norm": 0.07322255277171939,
+      "learning_rate": 4.564654012581428e-06,
+      "loss": 0.4289,
+      "step": 4261
+    },
+    {
+      "epoch": 2.1141864217450665,
+      "grad_norm": 0.07170049650607539,
+      "learning_rate": 4.562706770064824e-06,
+      "loss": 0.456,
+      "step": 4262
+    },
+    {
+      "epoch": 2.114682884448306,
+      "grad_norm": 0.07498256697277747,
+      "learning_rate": 4.560759594381477e-06,
+      "loss": 0.464,
+      "step": 4263
+    },
+    {
+      "epoch": 2.115179347151545,
+      "grad_norm": 0.07575026511841638,
+      "learning_rate": 4.558812485828983e-06,
+      "loss": 0.4769,
+      "step": 4264
+    },
+    {
+      "epoch": 2.1156758098547845,
+      "grad_norm": 0.07380441248442098,
+      "learning_rate": 4.556865444704928e-06,
+      "loss": 0.4638,
+      "step": 4265
+    },
+    {
+      "epoch": 2.116172272558024,
+      "grad_norm": 0.0696430455459803,
+      "learning_rate": 4.554918471306883e-06,
+      "loss": 0.4171,
+      "step": 4266
+    },
+    {
+      "epoch": 2.1166687352612636,
+      "grad_norm": 0.07436796696460703,
+      "learning_rate": 4.552971565932416e-06,
+      "loss": 0.4443,
+      "step": 4267
+    },
+    {
+      "epoch": 2.117165197964503,
+      "grad_norm": 0.07578969996317546,
+      "learning_rate": 4.551024728879075e-06,
+      "loss": 0.4688,
+      "step": 4268
+    },
+    {
+      "epoch": 2.1176616606677423,
+      "grad_norm": 0.07395243694691826,
+      "learning_rate": 4.549077960444407e-06,
+      "loss": 0.4541,
+      "step": 4269
+    },
+    {
+      "epoch": 2.1181581233709816,
+      "grad_norm": 0.07110739817951225,
+      "learning_rate": 4.547131260925941e-06,
+      "loss": 0.4581,
+      "step": 4270
+    },
+    {
+      "epoch": 2.118654586074221,
+      "grad_norm": 0.0741125823175704,
+      "learning_rate": 4.545184630621202e-06,
+      "loss": 0.4824,
+      "step": 4271
+    },
+    {
+      "epoch": 2.1191510487774607,
+      "grad_norm": 0.07664528627796972,
+      "learning_rate": 4.543238069827701e-06,
+      "loss": 0.4501,
+      "step": 4272
+    },
+    {
+      "epoch": 2.1196475114807,
+      "grad_norm": 0.07570433962985741,
+      "learning_rate": 4.541291578842936e-06,
+      "loss": 0.497,
+      "step": 4273
+    },
+    {
+      "epoch": 2.1201439741839394,
+      "grad_norm": 0.07617127656102472,
+      "learning_rate": 4.539345157964399e-06,
+      "loss": 0.4631,
+      "step": 4274
+    },
+    {
+      "epoch": 2.1206404368871787,
+      "grad_norm": 0.07313940894695849,
+      "learning_rate": 4.537398807489568e-06,
+      "loss": 0.4678,
+      "step": 4275
+    },
+    {
+      "epoch": 2.121136899590418,
+      "grad_norm": 0.07300910218402296,
+      "learning_rate": 4.535452527715911e-06,
+      "loss": 0.4542,
+      "step": 4276
+    },
+    {
+      "epoch": 2.121633362293658,
+      "grad_norm": 0.07492994635211063,
+      "learning_rate": 4.533506318940888e-06,
+      "loss": 0.4396,
+      "step": 4277
+    },
+    {
+      "epoch": 2.122129824996897,
+      "grad_norm": 0.07652124159341604,
+      "learning_rate": 4.531560181461944e-06,
+      "loss": 0.4225,
+      "step": 4278
+    },
+    {
+      "epoch": 2.1226262877001365,
+      "grad_norm": 0.07259064899112683,
+      "learning_rate": 4.529614115576515e-06,
+      "loss": 0.4752,
+      "step": 4279
+    },
+    {
+      "epoch": 2.123122750403376,
+      "grad_norm": 0.07304094179424345,
+      "learning_rate": 4.5276681215820266e-06,
+      "loss": 0.4583,
+      "step": 4280
+    },
+    {
+      "epoch": 2.123619213106615,
+      "grad_norm": 0.07441873823770657,
+      "learning_rate": 4.525722199775893e-06,
+      "loss": 0.4626,
+      "step": 4281
+    },
+    {
+      "epoch": 2.124115675809855,
+      "grad_norm": 0.07328536701423397,
+      "learning_rate": 4.523776350455516e-06,
+      "loss": 0.4899,
+      "step": 4282
+    },
+    {
+      "epoch": 2.1246121385130943,
+      "grad_norm": 0.0703197526894117,
+      "learning_rate": 4.521830573918289e-06,
+      "loss": 0.4233,
+      "step": 4283
+    },
+    {
+      "epoch": 2.1251086012163336,
+      "grad_norm": 0.07408961544325211,
+      "learning_rate": 4.5198848704615915e-06,
+      "loss": 0.4929,
+      "step": 4284
+    },
+    {
+      "epoch": 2.125605063919573,
+      "grad_norm": 0.07399250331444931,
+      "learning_rate": 4.517939240382793e-06,
+      "loss": 0.4982,
+      "step": 4285
+    },
+    {
+      "epoch": 2.1261015266228123,
+      "grad_norm": 0.07313344981416467,
+      "learning_rate": 4.515993683979253e-06,
+      "loss": 0.4701,
+      "step": 4286
+    },
+    {
+      "epoch": 2.126597989326052,
+      "grad_norm": 0.07221818239629586,
+      "learning_rate": 4.514048201548318e-06,
+      "loss": 0.4547,
+      "step": 4287
+    },
+    {
+      "epoch": 2.1270944520292914,
+      "grad_norm": 0.07303252754167958,
+      "learning_rate": 4.512102793387325e-06,
+      "loss": 0.4578,
+      "step": 4288
+    },
+    {
+      "epoch": 2.1275909147325307,
+      "grad_norm": 0.11487612831188249,
+      "learning_rate": 4.510157459793596e-06,
+      "loss": 0.4151,
+      "step": 4289
+    },
+    {
+      "epoch": 2.12808737743577,
+      "grad_norm": 0.07265820335361244,
+      "learning_rate": 4.508212201064446e-06,
+      "loss": 0.4341,
+      "step": 4290
+    },
+    {
+      "epoch": 2.1285838401390094,
+      "grad_norm": 0.074447871083182,
+      "learning_rate": 4.506267017497174e-06,
+      "loss": 0.481,
+      "step": 4291
+    },
+    {
+      "epoch": 2.129080302842249,
+      "grad_norm": 0.07330142691315944,
+      "learning_rate": 4.504321909389072e-06,
+      "loss": 0.4345,
+      "step": 4292
+    },
+    {
+      "epoch": 2.1295767655454885,
+      "grad_norm": 0.07169395950659548,
+      "learning_rate": 4.5023768770374195e-06,
+      "loss": 0.4803,
+      "step": 4293
+    },
+    {
+      "epoch": 2.130073228248728,
+      "grad_norm": 0.07144159301453944,
+      "learning_rate": 4.500431920739482e-06,
+      "loss": 0.4379,
+      "step": 4294
+    },
+    {
+      "epoch": 2.130569690951967,
+      "grad_norm": 0.07462658435580564,
+      "learning_rate": 4.498487040792517e-06,
+      "loss": 0.4455,
+      "step": 4295
+    },
+    {
+      "epoch": 2.1310661536552065,
+      "grad_norm": 0.07373163804771965,
+      "learning_rate": 4.496542237493765e-06,
+      "loss": 0.4269,
+      "step": 4296
+    },
+    {
+      "epoch": 2.1315626163584462,
+      "grad_norm": 0.07622904870198172,
+      "learning_rate": 4.49459751114046e-06,
+      "loss": 0.5137,
+      "step": 4297
+    },
+    {
+      "epoch": 2.1320590790616856,
+      "grad_norm": 0.07372742998907207,
+      "learning_rate": 4.492652862029822e-06,
+      "loss": 0.5211,
+      "step": 4298
+    },
+    {
+      "epoch": 2.132555541764925,
+      "grad_norm": 0.07370616974666123,
+      "learning_rate": 4.490708290459061e-06,
+      "loss": 0.4474,
+      "step": 4299
+    },
+    {
+      "epoch": 2.1330520044681642,
+      "grad_norm": 0.07458809789142895,
+      "learning_rate": 4.488763796725369e-06,
+      "loss": 0.4831,
+      "step": 4300
+    },
+    {
+      "epoch": 2.1335484671714036,
+      "grad_norm": 0.07398203891052577,
+      "learning_rate": 4.486819381125935e-06,
+      "loss": 0.4833,
+      "step": 4301
+    },
+    {
+      "epoch": 2.1340449298746433,
+      "grad_norm": 0.07415054973135388,
+      "learning_rate": 4.48487504395793e-06,
+      "loss": 0.4858,
+      "step": 4302
+    },
+    {
+      "epoch": 2.1345413925778827,
+      "grad_norm": 0.07379822859689968,
+      "learning_rate": 4.4829307855185155e-06,
+      "loss": 0.4328,
+      "step": 4303
+    },
+    {
+      "epoch": 2.135037855281122,
+      "grad_norm": 0.07379863650798167,
+      "learning_rate": 4.480986606104842e-06,
+      "loss": 0.4874,
+      "step": 4304
+    },
+    {
+      "epoch": 2.1355343179843613,
+      "grad_norm": 0.07413874016115828,
+      "learning_rate": 4.479042506014042e-06,
+      "loss": 0.4744,
+      "step": 4305
+    },
+    {
+      "epoch": 2.1360307806876007,
+      "grad_norm": 0.07332637014689809,
+      "learning_rate": 4.477098485543242e-06,
+      "loss": 0.4842,
+      "step": 4306
+    },
+    {
+      "epoch": 2.1365272433908404,
+      "grad_norm": 0.07626723329610915,
+      "learning_rate": 4.475154544989554e-06,
+      "loss": 0.4637,
+      "step": 4307
+    },
+    {
+      "epoch": 2.1370237060940798,
+      "grad_norm": 0.07124366540358651,
+      "learning_rate": 4.473210684650079e-06,
+      "loss": 0.4553,
+      "step": 4308
+    },
+    {
+      "epoch": 2.137520168797319,
+      "grad_norm": 0.07106600874663206,
+      "learning_rate": 4.471266904821904e-06,
+      "loss": 0.4579,
+      "step": 4309
+    },
+    {
+      "epoch": 2.1380166315005584,
+      "grad_norm": 0.07320126920930652,
+      "learning_rate": 4.469323205802105e-06,
+      "loss": 0.4761,
+      "step": 4310
+    },
+    {
+      "epoch": 2.1385130942037978,
+      "grad_norm": 0.0748936232395386,
+      "learning_rate": 4.467379587887747e-06,
+      "loss": 0.49,
+      "step": 4311
+    },
+    {
+      "epoch": 2.1390095569070375,
+      "grad_norm": 0.070655034353901,
+      "learning_rate": 4.465436051375877e-06,
+      "loss": 0.4706,
+      "step": 4312
+    },
+    {
+      "epoch": 2.139506019610277,
+      "grad_norm": 0.07154275743223767,
+      "learning_rate": 4.463492596563536e-06,
+      "loss": 0.4721,
+      "step": 4313
+    },
+    {
+      "epoch": 2.140002482313516,
+      "grad_norm": 0.06944606811288602,
+      "learning_rate": 4.46154922374775e-06,
+      "loss": 0.4293,
+      "step": 4314
+    },
+    {
+      "epoch": 2.1404989450167555,
+      "grad_norm": 0.0695554372896349,
+      "learning_rate": 4.459605933225531e-06,
+      "loss": 0.4095,
+      "step": 4315
+    },
+    {
+      "epoch": 2.140995407719995,
+      "grad_norm": 0.07387663952272498,
+      "learning_rate": 4.4576627252938805e-06,
+      "loss": 0.5105,
+      "step": 4316
+    },
+    {
+      "epoch": 2.1414918704232346,
+      "grad_norm": 0.07296540945926891,
+      "learning_rate": 4.455719600249787e-06,
+      "loss": 0.4748,
+      "step": 4317
+    },
+    {
+      "epoch": 2.141988333126474,
+      "grad_norm": 0.07444996417793436,
+      "learning_rate": 4.453776558390225e-06,
+      "loss": 0.4476,
+      "step": 4318
+    },
+    {
+      "epoch": 2.1424847958297133,
+      "grad_norm": 0.07412269708771786,
+      "learning_rate": 4.4518336000121585e-06,
+      "loss": 0.4835,
+      "step": 4319
+    },
+    {
+      "epoch": 2.1429812585329526,
+      "grad_norm": 0.07223272136559325,
+      "learning_rate": 4.4498907254125394e-06,
+      "loss": 0.4209,
+      "step": 4320
+    },
+    {
+      "epoch": 2.143477721236192,
+      "grad_norm": 0.07511790139542897,
+      "learning_rate": 4.4479479348883e-06,
+      "loss": 0.4588,
+      "step": 4321
+    },
+    {
+      "epoch": 2.1439741839394317,
+      "grad_norm": 0.07007837954759942,
+      "learning_rate": 4.446005228736368e-06,
+      "loss": 0.4573,
+      "step": 4322
+    },
+    {
+      "epoch": 2.144470646642671,
+      "grad_norm": 0.0728602451708517,
+      "learning_rate": 4.444062607253655e-06,
+      "loss": 0.5192,
+      "step": 4323
+    },
+    {
+      "epoch": 2.1449671093459104,
+      "grad_norm": 0.07341402010785382,
+      "learning_rate": 4.442120070737057e-06,
+      "loss": 0.459,
+      "step": 4324
+    },
+    {
+      "epoch": 2.1454635720491497,
+      "grad_norm": 0.07380914908331342,
+      "learning_rate": 4.4401776194834615e-06,
+      "loss": 0.4823,
+      "step": 4325
+    },
+    {
+      "epoch": 2.145960034752389,
+      "grad_norm": 0.07298131712761911,
+      "learning_rate": 4.438235253789741e-06,
+      "loss": 0.4405,
+      "step": 4326
+    },
+    {
+      "epoch": 2.146456497455629,
+      "grad_norm": 0.07533361099651531,
+      "learning_rate": 4.436292973952755e-06,
+      "loss": 0.4602,
+      "step": 4327
+    },
+    {
+      "epoch": 2.146952960158868,
+      "grad_norm": 0.07118691927368932,
+      "learning_rate": 4.43435078026935e-06,
+      "loss": 0.4753,
+      "step": 4328
+    },
+    {
+      "epoch": 2.1474494228621075,
+      "grad_norm": 0.07142698648583826,
+      "learning_rate": 4.432408673036358e-06,
+      "loss": 0.4595,
+      "step": 4329
+    },
+    {
+      "epoch": 2.147945885565347,
+      "grad_norm": 0.07637953793097572,
+      "learning_rate": 4.430466652550599e-06,
+      "loss": 0.4579,
+      "step": 4330
+    },
+    {
+      "epoch": 2.148442348268586,
+      "grad_norm": 0.07424465808567507,
+      "learning_rate": 4.4285247191088815e-06,
+      "loss": 0.4672,
+      "step": 4331
+    },
+    {
+      "epoch": 2.148938810971826,
+      "grad_norm": 0.07278996829833347,
+      "learning_rate": 4.426582873007999e-06,
+      "loss": 0.474,
+      "step": 4332
+    },
+    {
+      "epoch": 2.1494352736750653,
+      "grad_norm": 0.0717818181339654,
+      "learning_rate": 4.4246411145447275e-06,
+      "loss": 0.4381,
+      "step": 4333
+    },
+    {
+      "epoch": 2.1499317363783046,
+      "grad_norm": 0.0730160088959314,
+      "learning_rate": 4.422699444015838e-06,
+      "loss": 0.4499,
+      "step": 4334
+    },
+    {
+      "epoch": 2.150428199081544,
+      "grad_norm": 0.07393221348452107,
+      "learning_rate": 4.420757861718082e-06,
+      "loss": 0.4661,
+      "step": 4335
+    },
+    {
+      "epoch": 2.1509246617847833,
+      "grad_norm": 0.07274849802941666,
+      "learning_rate": 4.418816367948201e-06,
+      "loss": 0.4393,
+      "step": 4336
+    },
+    {
+      "epoch": 2.151421124488023,
+      "grad_norm": 0.07902231214358871,
+      "learning_rate": 4.416874963002918e-06,
+      "loss": 0.4608,
+      "step": 4337
+    },
+    {
+      "epoch": 2.1519175871912624,
+      "grad_norm": 0.07283494180291966,
+      "learning_rate": 4.414933647178948e-06,
+      "loss": 0.4765,
+      "step": 4338
+    },
+    {
+      "epoch": 2.1524140498945017,
+      "grad_norm": 0.0701883235014208,
+      "learning_rate": 4.412992420772988e-06,
+      "loss": 0.4504,
+      "step": 4339
+    },
+    {
+      "epoch": 2.152910512597741,
+      "grad_norm": 0.07421158965760369,
+      "learning_rate": 4.411051284081725e-06,
+      "loss": 0.4522,
+      "step": 4340
+    },
+    {
+      "epoch": 2.1534069753009804,
+      "grad_norm": 0.07571375423025403,
+      "learning_rate": 4.4091102374018295e-06,
+      "loss": 0.4634,
+      "step": 4341
+    },
+    {
+      "epoch": 2.15390343800422,
+      "grad_norm": 0.07236772001970997,
+      "learning_rate": 4.40716928102996e-06,
+      "loss": 0.4478,
+      "step": 4342
+    },
+    {
+      "epoch": 2.1543999007074595,
+      "grad_norm": 0.07188417221152126,
+      "learning_rate": 4.405228415262763e-06,
+      "loss": 0.4645,
+      "step": 4343
+    },
+    {
+      "epoch": 2.154896363410699,
+      "grad_norm": 0.07382921539610335,
+      "learning_rate": 4.403287640396864e-06,
+      "loss": 0.4807,
+      "step": 4344
+    },
+    {
+      "epoch": 2.155392826113938,
+      "grad_norm": 0.06982201525129672,
+      "learning_rate": 4.401346956728881e-06,
+      "loss": 0.4393,
+      "step": 4345
+    },
+    {
+      "epoch": 2.1558892888171775,
+      "grad_norm": 0.07054004972844154,
+      "learning_rate": 4.3994063645554185e-06,
+      "loss": 0.4309,
+      "step": 4346
+    },
+    {
+      "epoch": 2.1563857515204172,
+      "grad_norm": 0.07132879012840733,
+      "learning_rate": 4.397465864173063e-06,
+      "loss": 0.4381,
+      "step": 4347
+    },
+    {
+      "epoch": 2.1568822142236566,
+      "grad_norm": 0.07432312035402415,
+      "learning_rate": 4.39552545587839e-06,
+      "loss": 0.4612,
+      "step": 4348
+    },
+    {
+      "epoch": 2.157378676926896,
+      "grad_norm": 0.07411331173872626,
+      "learning_rate": 4.393585139967958e-06,
+      "loss": 0.4989,
+      "step": 4349
+    },
+    {
+      "epoch": 2.157875139630135,
+      "grad_norm": 0.07267337288227944,
+      "learning_rate": 4.391644916738314e-06,
+      "loss": 0.4461,
+      "step": 4350
+    },
+    {
+      "epoch": 2.1583716023333746,
+      "grad_norm": 0.07331622047891374,
+      "learning_rate": 4.38970478648599e-06,
+      "loss": 0.459,
+      "step": 4351
+    },
+    {
+      "epoch": 2.1588680650366143,
+      "grad_norm": 0.07453181757687845,
+      "learning_rate": 4.387764749507507e-06,
+      "loss": 0.4597,
+      "step": 4352
+    },
+    {
+      "epoch": 2.1593645277398537,
+      "grad_norm": 0.07315319571236073,
+      "learning_rate": 4.385824806099364e-06,
+      "loss": 0.4769,
+      "step": 4353
+    },
+    {
+      "epoch": 2.159860990443093,
+      "grad_norm": 0.07693153678377337,
+      "learning_rate": 4.383884956558051e-06,
+      "loss": 0.4861,
+      "step": 4354
+    },
+    {
+      "epoch": 2.1603574531463323,
+      "grad_norm": 0.07417595641227936,
+      "learning_rate": 4.381945201180045e-06,
+      "loss": 0.4691,
+      "step": 4355
+    },
+    {
+      "epoch": 2.1608539158495716,
+      "grad_norm": 0.07529427254296801,
+      "learning_rate": 4.380005540261803e-06,
+      "loss": 0.4748,
+      "step": 4356
+    },
+    {
+      "epoch": 2.1613503785528114,
+      "grad_norm": 0.07267164548374914,
+      "learning_rate": 4.378065974099775e-06,
+      "loss": 0.4566,
+      "step": 4357
+    },
+    {
+      "epoch": 2.1618468412560508,
+      "grad_norm": 0.07203923664865745,
+      "learning_rate": 4.37612650299039e-06,
+      "loss": 0.4685,
+      "step": 4358
+    },
+    {
+      "epoch": 2.16234330395929,
+      "grad_norm": 0.07187394235177122,
+      "learning_rate": 4.374187127230068e-06,
+      "loss": 0.4557,
+      "step": 4359
+    },
+    {
+      "epoch": 2.1628397666625294,
+      "grad_norm": 0.07011681787618959,
+      "learning_rate": 4.3722478471152065e-06,
+      "loss": 0.4321,
+      "step": 4360
+    },
+    {
+      "epoch": 2.1633362293657687,
+      "grad_norm": 0.07272195445241587,
+      "learning_rate": 4.370308662942198e-06,
+      "loss": 0.4518,
+      "step": 4361
+    },
+    {
+      "epoch": 2.1638326920690085,
+      "grad_norm": 0.07460252071373111,
+      "learning_rate": 4.368369575007413e-06,
+      "loss": 0.4907,
+      "step": 4362
+    },
+    {
+      "epoch": 2.164329154772248,
+      "grad_norm": 0.07561379198675501,
+      "learning_rate": 4.3664305836072116e-06,
+      "loss": 0.4796,
+      "step": 4363
+    },
+    {
+      "epoch": 2.164825617475487,
+      "grad_norm": 0.07136976585613097,
+      "learning_rate": 4.364491689037938e-06,
+      "loss": 0.439,
+      "step": 4364
+    },
+    {
+      "epoch": 2.1653220801787265,
+      "grad_norm": 0.07162067869532027,
+      "learning_rate": 4.362552891595919e-06,
+      "loss": 0.4359,
+      "step": 4365
+    },
+    {
+      "epoch": 2.165818542881966,
+      "grad_norm": 0.0733281484806096,
+      "learning_rate": 4.3606141915774695e-06,
+      "loss": 0.4771,
+      "step": 4366
+    },
+    {
+      "epoch": 2.166315005585205,
+      "grad_norm": 0.07224633811997479,
+      "learning_rate": 4.3586755892788896e-06,
+      "loss": 0.4372,
+      "step": 4367
+    },
+    {
+      "epoch": 2.166811468288445,
+      "grad_norm": 0.07028756349782972,
+      "learning_rate": 4.356737084996465e-06,
+      "loss": 0.448,
+      "step": 4368
+    },
+    {
+      "epoch": 2.1673079309916843,
+      "grad_norm": 0.07108239096848315,
+      "learning_rate": 4.354798679026461e-06,
+      "loss": 0.4479,
+      "step": 4369
+    },
+    {
+      "epoch": 2.1678043936949236,
+      "grad_norm": 0.07372307764488349,
+      "learning_rate": 4.352860371665134e-06,
+      "loss": 0.4601,
+      "step": 4370
+    },
+    {
+      "epoch": 2.168300856398163,
+      "grad_norm": 0.06975696761984575,
+      "learning_rate": 4.350922163208724e-06,
+      "loss": 0.4494,
+      "step": 4371
+    },
+    {
+      "epoch": 2.1687973191014027,
+      "grad_norm": 0.07546779879045096,
+      "learning_rate": 4.348984053953453e-06,
+      "loss": 0.4524,
+      "step": 4372
+    },
+    {
+      "epoch": 2.169293781804642,
+      "grad_norm": 0.07280842926844215,
+      "learning_rate": 4.3470460441955306e-06,
+      "loss": 0.4536,
+      "step": 4373
+    },
+    {
+      "epoch": 2.1697902445078814,
+      "grad_norm": 0.07703084007865943,
+      "learning_rate": 4.345108134231152e-06,
+      "loss": 0.4747,
+      "step": 4374
+    },
+    {
+      "epoch": 2.1702867072111207,
+      "grad_norm": 0.07110900628117096,
+      "learning_rate": 4.343170324356495e-06,
+      "loss": 0.4499,
+      "step": 4375
+    },
+    {
+      "epoch": 2.17078316991436,
+      "grad_norm": 0.07369518298980911,
+      "learning_rate": 4.341232614867722e-06,
+      "loss": 0.457,
+      "step": 4376
+    },
+    {
+      "epoch": 2.1712796326175994,
+      "grad_norm": 0.07157372517615651,
+      "learning_rate": 4.3392950060609804e-06,
+      "loss": 0.4473,
+      "step": 4377
+    },
+    {
+      "epoch": 2.171776095320839,
+      "grad_norm": 0.0717701752371705,
+      "learning_rate": 4.337357498232405e-06,
+      "loss": 0.4516,
+      "step": 4378
+    },
+    {
+      "epoch": 2.1722725580240785,
+      "grad_norm": 0.07316611230340865,
+      "learning_rate": 4.335420091678109e-06,
+      "loss": 0.4419,
+      "step": 4379
+    },
+    {
+      "epoch": 2.172769020727318,
+      "grad_norm": 0.0756994196452662,
+      "learning_rate": 4.3334827866941995e-06,
+      "loss": 0.475,
+      "step": 4380
+    },
+    {
+      "epoch": 2.173265483430557,
+      "grad_norm": 0.0785324379774332,
+      "learning_rate": 4.331545583576758e-06,
+      "loss": 0.4879,
+      "step": 4381
+    },
+    {
+      "epoch": 2.173761946133797,
+      "grad_norm": 0.07380761788530164,
+      "learning_rate": 4.329608482621855e-06,
+      "loss": 0.4705,
+      "step": 4382
+    },
+    {
+      "epoch": 2.1742584088370362,
+      "grad_norm": 0.07386854839808037,
+      "learning_rate": 4.327671484125548e-06,
+      "loss": 0.4466,
+      "step": 4383
+    },
+    {
+      "epoch": 2.1747548715402756,
+      "grad_norm": 0.07355308329216982,
+      "learning_rate": 4.325734588383876e-06,
+      "loss": 0.4671,
+      "step": 4384
+    },
+    {
+      "epoch": 2.175251334243515,
+      "grad_norm": 0.07717137637860379,
+      "learning_rate": 4.323797795692859e-06,
+      "loss": 0.4768,
+      "step": 4385
+    },
+    {
+      "epoch": 2.1757477969467542,
+      "grad_norm": 0.10530308493803685,
+      "learning_rate": 4.321861106348507e-06,
+      "loss": 0.4548,
+      "step": 4386
+    },
+    {
+      "epoch": 2.1762442596499936,
+      "grad_norm": 0.07624838294925372,
+      "learning_rate": 4.319924520646812e-06,
+      "loss": 0.464,
+      "step": 4387
+    },
+    {
+      "epoch": 2.1767407223532333,
+      "grad_norm": 0.07398353032038894,
+      "learning_rate": 4.31798803888375e-06,
+      "loss": 0.4362,
+      "step": 4388
+    },
+    {
+      "epoch": 2.1772371850564727,
+      "grad_norm": 0.07362322859625167,
+      "learning_rate": 4.31605166135528e-06,
+      "loss": 0.4702,
+      "step": 4389
+    },
+    {
+      "epoch": 2.177733647759712,
+      "grad_norm": 0.07241274976591916,
+      "learning_rate": 4.314115388357348e-06,
+      "loss": 0.45,
+      "step": 4390
+    },
+    {
+      "epoch": 2.1782301104629513,
+      "grad_norm": 0.06990581661486783,
+      "learning_rate": 4.312179220185883e-06,
+      "loss": 0.4299,
+      "step": 4391
+    },
+    {
+      "epoch": 2.178726573166191,
+      "grad_norm": 0.07375544071049109,
+      "learning_rate": 4.310243157136794e-06,
+      "loss": 0.4975,
+      "step": 4392
+    },
+    {
+      "epoch": 2.1792230358694304,
+      "grad_norm": 0.07164232045346458,
+      "learning_rate": 4.308307199505979e-06,
+      "loss": 0.4429,
+      "step": 4393
+    },
+    {
+      "epoch": 2.17971949857267,
+      "grad_norm": 0.07286003848759594,
+      "learning_rate": 4.306371347589318e-06,
+      "loss": 0.4513,
+      "step": 4394
+    },
+    {
+      "epoch": 2.180215961275909,
+      "grad_norm": 0.07478821375829875,
+      "learning_rate": 4.304435601682675e-06,
+      "loss": 0.4706,
+      "step": 4395
+    },
+    {
+      "epoch": 2.1807124239791484,
+      "grad_norm": 0.07402377964070475,
+      "learning_rate": 4.3024999620819e-06,
+      "loss": 0.4735,
+      "step": 4396
+    },
+    {
+      "epoch": 2.1812088866823878,
+      "grad_norm": 0.07372832051354254,
+      "learning_rate": 4.30056442908282e-06,
+      "loss": 0.4672,
+      "step": 4397
+    },
+    {
+      "epoch": 2.1817053493856275,
+      "grad_norm": 0.0723792667046002,
+      "learning_rate": 4.298629002981253e-06,
+      "loss": 0.4732,
+      "step": 4398
+    },
+    {
+      "epoch": 2.182201812088867,
+      "grad_norm": 0.07305135775888415,
+      "learning_rate": 4.296693684072997e-06,
+      "loss": 0.4385,
+      "step": 4399
+    },
+    {
+      "epoch": 2.182698274792106,
+      "grad_norm": 0.07387898291563187,
+      "learning_rate": 4.294758472653837e-06,
+      "loss": 0.4795,
+      "step": 4400
+    },
+    {
+      "epoch": 2.1831947374953455,
+      "grad_norm": 0.07342139584048227,
+      "learning_rate": 4.2928233690195345e-06,
+      "loss": 0.4317,
+      "step": 4401
+    },
+    {
+      "epoch": 2.1836912001985853,
+      "grad_norm": 0.07323848425271148,
+      "learning_rate": 4.290888373465841e-06,
+      "loss": 0.4771,
+      "step": 4402
+    },
+    {
+      "epoch": 2.1841876629018246,
+      "grad_norm": 0.07217563028765697,
+      "learning_rate": 4.288953486288491e-06,
+      "loss": 0.4176,
+      "step": 4403
+    },
+    {
+      "epoch": 2.184684125605064,
+      "grad_norm": 0.07588623975529438,
+      "learning_rate": 4.2870187077832e-06,
+      "loss": 0.4761,
+      "step": 4404
+    },
+    {
+      "epoch": 2.1851805883083033,
+      "grad_norm": 0.07679164359797928,
+      "learning_rate": 4.2850840382456656e-06,
+      "loss": 0.4809,
+      "step": 4405
+    },
+    {
+      "epoch": 2.1856770510115426,
+      "grad_norm": 0.07357731906381813,
+      "learning_rate": 4.283149477971575e-06,
+      "loss": 0.4922,
+      "step": 4406
+    },
+    {
+      "epoch": 2.186173513714782,
+      "grad_norm": 0.07414935840899083,
+      "learning_rate": 4.281215027256592e-06,
+      "loss": 0.4418,
+      "step": 4407
+    },
+    {
+      "epoch": 2.1866699764180217,
+      "grad_norm": 0.07168473465744193,
+      "learning_rate": 4.2792806863963685e-06,
+      "loss": 0.4318,
+      "step": 4408
+    },
+    {
+      "epoch": 2.187166439121261,
+      "grad_norm": 0.07499491492610341,
+      "learning_rate": 4.277346455686535e-06,
+      "loss": 0.4517,
+      "step": 4409
+    },
+    {
+      "epoch": 2.1876629018245004,
+      "grad_norm": 0.07423494586098568,
+      "learning_rate": 4.275412335422709e-06,
+      "loss": 0.4415,
+      "step": 4410
+    },
+    {
+      "epoch": 2.1881593645277397,
+      "grad_norm": 0.0729238681152025,
+      "learning_rate": 4.27347832590049e-06,
+      "loss": 0.4583,
+      "step": 4411
+    },
+    {
+      "epoch": 2.188655827230979,
+      "grad_norm": 0.07542756084684298,
+      "learning_rate": 4.27154442741546e-06,
+      "loss": 0.4458,
+      "step": 4412
+    },
+    {
+      "epoch": 2.189152289934219,
+      "grad_norm": 0.07608446572149048,
+      "learning_rate": 4.269610640263185e-06,
+      "loss": 0.4653,
+      "step": 4413
+    },
+    {
+      "epoch": 2.189648752637458,
+      "grad_norm": 0.07361868469969725,
+      "learning_rate": 4.267676964739211e-06,
+      "loss": 0.4667,
+      "step": 4414
+    },
+    {
+      "epoch": 2.1901452153406975,
+      "grad_norm": 0.0687170502926834,
+      "learning_rate": 4.2657434011390725e-06,
+      "loss": 0.4419,
+      "step": 4415
+    },
+    {
+      "epoch": 2.190641678043937,
+      "grad_norm": 0.07000188477823314,
+      "learning_rate": 4.2638099497582835e-06,
+      "loss": 0.4715,
+      "step": 4416
+    },
+    {
+      "epoch": 2.191138140747176,
+      "grad_norm": 0.07273493307479256,
+      "learning_rate": 4.2618766108923385e-06,
+      "loss": 0.4671,
+      "step": 4417
+    },
+    {
+      "epoch": 2.191634603450416,
+      "grad_norm": 0.07397448831392173,
+      "learning_rate": 4.259943384836717e-06,
+      "loss": 0.4562,
+      "step": 4418
+    },
+    {
+      "epoch": 2.1921310661536553,
+      "grad_norm": 0.07195426602600087,
+      "learning_rate": 4.258010271886885e-06,
+      "loss": 0.4525,
+      "step": 4419
+    },
+    {
+      "epoch": 2.1926275288568946,
+      "grad_norm": 0.07397602097862697,
+      "learning_rate": 4.256077272338286e-06,
+      "loss": 0.4471,
+      "step": 4420
+    },
+    {
+      "epoch": 2.193123991560134,
+      "grad_norm": 0.07350928801561125,
+      "learning_rate": 4.254144386486347e-06,
+      "loss": 0.4875,
+      "step": 4421
+    },
+    {
+      "epoch": 2.1936204542633733,
+      "grad_norm": 0.07269756927451913,
+      "learning_rate": 4.2522116146264805e-06,
+      "loss": 0.4569,
+      "step": 4422
+    },
+    {
+      "epoch": 2.194116916966613,
+      "grad_norm": 0.07206227980041602,
+      "learning_rate": 4.2502789570540775e-06,
+      "loss": 0.447,
+      "step": 4423
+    },
+    {
+      "epoch": 2.1946133796698524,
+      "grad_norm": 0.07262743407923547,
+      "learning_rate": 4.248346414064517e-06,
+      "loss": 0.4778,
+      "step": 4424
+    },
+    {
+      "epoch": 2.1951098423730917,
+      "grad_norm": 0.07271820913507389,
+      "learning_rate": 4.246413985953154e-06,
+      "loss": 0.491,
+      "step": 4425
+    },
+    {
+      "epoch": 2.195606305076331,
+      "grad_norm": 0.0704878390287762,
+      "learning_rate": 4.24448167301533e-06,
+      "loss": 0.4448,
+      "step": 4426
+    },
+    {
+      "epoch": 2.1961027677795704,
+      "grad_norm": 0.07215568890408974,
+      "learning_rate": 4.242549475546369e-06,
+      "loss": 0.4624,
+      "step": 4427
+    },
+    {
+      "epoch": 2.19659923048281,
+      "grad_norm": 0.07656441422337223,
+      "learning_rate": 4.240617393841578e-06,
+      "loss": 0.4868,
+      "step": 4428
+    },
+    {
+      "epoch": 2.1970956931860495,
+      "grad_norm": 0.07123197731052923,
+      "learning_rate": 4.23868542819624e-06,
+      "loss": 0.4561,
+      "step": 4429
+    },
+    {
+      "epoch": 2.197592155889289,
+      "grad_norm": 0.07713749872276374,
+      "learning_rate": 4.236753578905627e-06,
+      "loss": 0.4506,
+      "step": 4430
+    },
+    {
+      "epoch": 2.198088618592528,
+      "grad_norm": 0.07255814303647512,
+      "learning_rate": 4.234821846264993e-06,
+      "loss": 0.4292,
+      "step": 4431
+    },
+    {
+      "epoch": 2.1985850812957675,
+      "grad_norm": 0.0736169066359428,
+      "learning_rate": 4.232890230569573e-06,
+      "loss": 0.5202,
+      "step": 4432
+    },
+    {
+      "epoch": 2.1990815439990072,
+      "grad_norm": 0.07229802672521726,
+      "learning_rate": 4.2309587321145795e-06,
+      "loss": 0.4574,
+      "step": 4433
+    },
+    {
+      "epoch": 2.1995780067022466,
+      "grad_norm": 0.0753428203681146,
+      "learning_rate": 4.229027351195213e-06,
+      "loss": 0.4579,
+      "step": 4434
+    },
+    {
+      "epoch": 2.200074469405486,
+      "grad_norm": 0.07479409648984664,
+      "learning_rate": 4.227096088106655e-06,
+      "loss": 0.4518,
+      "step": 4435
+    },
+    {
+      "epoch": 2.2005709321087252,
+      "grad_norm": 0.07571046025431778,
+      "learning_rate": 4.225164943144067e-06,
+      "loss": 0.469,
+      "step": 4436
+    },
+    {
+      "epoch": 2.2010673948119646,
+      "grad_norm": 0.07471281694591714,
+      "learning_rate": 4.223233916602593e-06,
+      "loss": 0.4571,
+      "step": 4437
+    },
+    {
+      "epoch": 2.2015638575152043,
+      "grad_norm": 0.07239478998206747,
+      "learning_rate": 4.221303008777361e-06,
+      "loss": 0.467,
+      "step": 4438
+    },
+    {
+      "epoch": 2.2020603202184437,
+      "grad_norm": 0.07068125404733483,
+      "learning_rate": 4.219372219963479e-06,
+      "loss": 0.4512,
+      "step": 4439
+    },
+    {
+      "epoch": 2.202556782921683,
+      "grad_norm": 0.07725181283625165,
+      "learning_rate": 4.217441550456037e-06,
+      "loss": 0.4959,
+      "step": 4440
+    },
+    {
+      "epoch": 2.2030532456249223,
+      "grad_norm": 0.07637029431018237,
+      "learning_rate": 4.215511000550106e-06,
+      "loss": 0.4788,
+      "step": 4441
+    },
+    {
+      "epoch": 2.2035497083281617,
+      "grad_norm": 0.07344370158757503,
+      "learning_rate": 4.21358057054074e-06,
+      "loss": 0.4706,
+      "step": 4442
+    },
+    {
+      "epoch": 2.2040461710314014,
+      "grad_norm": 0.07304870738294844,
+      "learning_rate": 4.211650260722975e-06,
+      "loss": 0.461,
+      "step": 4443
+    },
+    {
+      "epoch": 2.2045426337346408,
+      "grad_norm": 0.07258620076213448,
+      "learning_rate": 4.209720071391827e-06,
+      "loss": 0.4306,
+      "step": 4444
+    },
+    {
+      "epoch": 2.20503909643788,
+      "grad_norm": 0.0732693591865333,
+      "learning_rate": 4.207790002842296e-06,
+      "loss": 0.4573,
+      "step": 4445
+    },
+    {
+      "epoch": 2.2055355591411194,
+      "grad_norm": 0.07384740915845746,
+      "learning_rate": 4.20586005536936e-06,
+      "loss": 0.4423,
+      "step": 4446
+    },
+    {
+      "epoch": 2.2060320218443588,
+      "grad_norm": 0.07183291018678276,
+      "learning_rate": 4.203930229267982e-06,
+      "loss": 0.4744,
+      "step": 4447
+    },
+    {
+      "epoch": 2.2065284845475985,
+      "grad_norm": 0.07160026517347787,
+      "learning_rate": 4.2020005248331056e-06,
+      "loss": 0.4652,
+      "step": 4448
+    },
+    {
+      "epoch": 2.207024947250838,
+      "grad_norm": 0.073264422607452,
+      "learning_rate": 4.2000709423596515e-06,
+      "loss": 0.476,
+      "step": 4449
+    },
+    {
+      "epoch": 2.207521409954077,
+      "grad_norm": 0.06959340873010447,
+      "learning_rate": 4.198141482142529e-06,
+      "loss": 0.4654,
+      "step": 4450
+    },
+    {
+      "epoch": 2.2080178726573165,
+      "grad_norm": 0.07551089938013718,
+      "learning_rate": 4.196212144476626e-06,
+      "loss": 0.4698,
+      "step": 4451
+    },
+    {
+      "epoch": 2.208514335360556,
+      "grad_norm": 0.07608020844071971,
+      "learning_rate": 4.194282929656806e-06,
+      "loss": 0.4578,
+      "step": 4452
+    },
+    {
+      "epoch": 2.2090107980637956,
+      "grad_norm": 0.07294714292734182,
+      "learning_rate": 4.1923538379779224e-06,
+      "loss": 0.4295,
+      "step": 4453
+    },
+    {
+      "epoch": 2.209507260767035,
+      "grad_norm": 0.07450752572030765,
+      "learning_rate": 4.190424869734805e-06,
+      "loss": 0.4917,
+      "step": 4454
+    },
+    {
+      "epoch": 2.2100037234702743,
+      "grad_norm": 0.08921788801344878,
+      "learning_rate": 4.188496025222266e-06,
+      "loss": 0.481,
+      "step": 4455
+    },
+    {
+      "epoch": 2.2105001861735136,
+      "grad_norm": 0.07214432985792273,
+      "learning_rate": 4.186567304735099e-06,
+      "loss": 0.489,
+      "step": 4456
+    },
+    {
+      "epoch": 2.210996648876753,
+      "grad_norm": 0.07181054002878612,
+      "learning_rate": 4.184638708568075e-06,
+      "loss": 0.4494,
+      "step": 4457
+    },
+    {
+      "epoch": 2.2114931115799927,
+      "grad_norm": 0.07328512025912555,
+      "learning_rate": 4.1827102370159526e-06,
+      "loss": 0.4832,
+      "step": 4458
+    },
+    {
+      "epoch": 2.211989574283232,
+      "grad_norm": 0.07298126138675862,
+      "learning_rate": 4.180781890373465e-06,
+      "loss": 0.4687,
+      "step": 4459
+    },
+    {
+      "epoch": 2.2124860369864714,
+      "grad_norm": 0.0726929787084046,
+      "learning_rate": 4.178853668935332e-06,
+      "loss": 0.4734,
+      "step": 4460
+    },
+    {
+      "epoch": 2.2129824996897107,
+      "grad_norm": 0.07150065845892105,
+      "learning_rate": 4.1769255729962494e-06,
+      "loss": 0.4627,
+      "step": 4461
+    },
+    {
+      "epoch": 2.21347896239295,
+      "grad_norm": 0.077463177549223,
+      "learning_rate": 4.174997602850895e-06,
+      "loss": 0.4447,
+      "step": 4462
+    },
+    {
+      "epoch": 2.21397542509619,
+      "grad_norm": 0.0724247437106297,
+      "learning_rate": 4.173069758793929e-06,
+      "loss": 0.4667,
+      "step": 4463
+    },
+    {
+      "epoch": 2.214471887799429,
+      "grad_norm": 0.07246835564653938,
+      "learning_rate": 4.171142041119994e-06,
+      "loss": 0.4517,
+      "step": 4464
+    },
+    {
+      "epoch": 2.2149683505026685,
+      "grad_norm": 0.06890663228644153,
+      "learning_rate": 4.169214450123706e-06,
+      "loss": 0.4339,
+      "step": 4465
+    },
+    {
+      "epoch": 2.215464813205908,
+      "grad_norm": 0.07297905706760326,
+      "learning_rate": 4.167286986099669e-06,
+      "loss": 0.4085,
+      "step": 4466
+    },
+    {
+      "epoch": 2.215961275909147,
+      "grad_norm": 0.07651788554616279,
+      "learning_rate": 4.165359649342464e-06,
+      "loss": 0.4529,
+      "step": 4467
+    },
+    {
+      "epoch": 2.216457738612387,
+      "grad_norm": 0.071610709788606,
+      "learning_rate": 4.163432440146655e-06,
+      "loss": 0.4552,
+      "step": 4468
+    },
+    {
+      "epoch": 2.2169542013156263,
+      "grad_norm": 0.07536894363343133,
+      "learning_rate": 4.161505358806784e-06,
+      "loss": 0.4496,
+      "step": 4469
+    },
+    {
+      "epoch": 2.2174506640188656,
+      "grad_norm": 0.07399735928163335,
+      "learning_rate": 4.159578405617375e-06,
+      "loss": 0.4601,
+      "step": 4470
+    },
+    {
+      "epoch": 2.217947126722105,
+      "grad_norm": 0.07202784194770909,
+      "learning_rate": 4.15765158087293e-06,
+      "loss": 0.4554,
+      "step": 4471
+    },
+    {
+      "epoch": 2.2184435894253443,
+      "grad_norm": 0.075485992762671,
+      "learning_rate": 4.155724884867937e-06,
+      "loss": 0.4814,
+      "step": 4472
+    },
+    {
+      "epoch": 2.218940052128584,
+      "grad_norm": 0.07469834057898289,
+      "learning_rate": 4.153798317896857e-06,
+      "loss": 0.4916,
+      "step": 4473
+    },
+    {
+      "epoch": 2.2194365148318234,
+      "grad_norm": 0.07480518347076377,
+      "learning_rate": 4.151871880254137e-06,
+      "loss": 0.4586,
+      "step": 4474
+    },
+    {
+      "epoch": 2.2199329775350627,
+      "grad_norm": 0.07250194926970677,
+      "learning_rate": 4.1499455722342e-06,
+      "loss": 0.4629,
+      "step": 4475
+    },
+    {
+      "epoch": 2.220429440238302,
+      "grad_norm": 0.07107799191740055,
+      "learning_rate": 4.148019394131454e-06,
+      "loss": 0.453,
+      "step": 4476
+    },
+    {
+      "epoch": 2.2209259029415414,
+      "grad_norm": 0.07251330487830691,
+      "learning_rate": 4.146093346240284e-06,
+      "loss": 0.465,
+      "step": 4477
+    },
+    {
+      "epoch": 2.221422365644781,
+      "grad_norm": 0.07090740682014357,
+      "learning_rate": 4.144167428855053e-06,
+      "loss": 0.4513,
+      "step": 4478
+    },
+    {
+      "epoch": 2.2219188283480205,
+      "grad_norm": 0.07642691152132507,
+      "learning_rate": 4.142241642270109e-06,
+      "loss": 0.4221,
+      "step": 4479
+    },
+    {
+      "epoch": 2.22241529105126,
+      "grad_norm": 0.07441365762455417,
+      "learning_rate": 4.140315986779778e-06,
+      "loss": 0.4571,
+      "step": 4480
+    },
+    {
+      "epoch": 2.222911753754499,
+      "grad_norm": 0.07042826712428749,
+      "learning_rate": 4.138390462678364e-06,
+      "loss": 0.4605,
+      "step": 4481
+    },
+    {
+      "epoch": 2.2234082164577385,
+      "grad_norm": 0.07180591373029381,
+      "learning_rate": 4.136465070260151e-06,
+      "loss": 0.4278,
+      "step": 4482
+    },
+    {
+      "epoch": 2.2239046791609782,
+      "grad_norm": 0.07131430714390916,
+      "learning_rate": 4.134539809819407e-06,
+      "loss": 0.4639,
+      "step": 4483
+    },
+    {
+      "epoch": 2.2244011418642176,
+      "grad_norm": 0.07681011768308688,
+      "learning_rate": 4.1326146816503776e-06,
+      "loss": 0.4602,
+      "step": 4484
+    },
+    {
+      "epoch": 2.224897604567457,
+      "grad_norm": 0.07260171427884389,
+      "learning_rate": 4.130689686047286e-06,
+      "loss": 0.4388,
+      "step": 4485
+    },
+    {
+      "epoch": 2.225394067270696,
+      "grad_norm": 0.07201246409421869,
+      "learning_rate": 4.1287648233043366e-06,
+      "loss": 0.4125,
+      "step": 4486
+    },
+    {
+      "epoch": 2.2258905299739355,
+      "grad_norm": 0.07254989653192677,
+      "learning_rate": 4.126840093715715e-06,
+      "loss": 0.4525,
+      "step": 4487
+    },
+    {
+      "epoch": 2.2263869926771753,
+      "grad_norm": 0.07166160294326122,
+      "learning_rate": 4.124915497575587e-06,
+      "loss": 0.4568,
+      "step": 4488
+    },
+    {
+      "epoch": 2.2268834553804147,
+      "grad_norm": 0.07122292370884728,
+      "learning_rate": 4.122991035178093e-06,
+      "loss": 0.4559,
+      "step": 4489
+    },
+    {
+      "epoch": 2.227379918083654,
+      "grad_norm": 0.07258420031929565,
+      "learning_rate": 4.121066706817357e-06,
+      "loss": 0.4711,
+      "step": 4490
+    },
+    {
+      "epoch": 2.2278763807868933,
+      "grad_norm": 0.07465745398445925,
+      "learning_rate": 4.1191425127874824e-06,
+      "loss": 0.491,
+      "step": 4491
+    },
+    {
+      "epoch": 2.2283728434901326,
+      "grad_norm": 0.07335909603079206,
+      "learning_rate": 4.117218453382551e-06,
+      "loss": 0.4785,
+      "step": 4492
+    },
+    {
+      "epoch": 2.2288693061933724,
+      "grad_norm": 0.07387175642761908,
+      "learning_rate": 4.115294528896627e-06,
+      "loss": 0.4666,
+      "step": 4493
+    },
+    {
+      "epoch": 2.2293657688966118,
+      "grad_norm": 0.07207293863663537,
+      "learning_rate": 4.113370739623746e-06,
+      "loss": 0.451,
+      "step": 4494
+    },
+    {
+      "epoch": 2.229862231599851,
+      "grad_norm": 0.07313205287182717,
+      "learning_rate": 4.111447085857932e-06,
+      "loss": 0.4912,
+      "step": 4495
+    },
+    {
+      "epoch": 2.2303586943030904,
+      "grad_norm": 0.07288468015896155,
+      "learning_rate": 4.109523567893185e-06,
+      "loss": 0.4716,
+      "step": 4496
+    },
+    {
+      "epoch": 2.2308551570063297,
+      "grad_norm": 0.07035718648695975,
+      "learning_rate": 4.107600186023485e-06,
+      "loss": 0.4427,
+      "step": 4497
+    },
+    {
+      "epoch": 2.2313516197095695,
+      "grad_norm": 0.07110839782679326,
+      "learning_rate": 4.105676940542785e-06,
+      "loss": 0.4382,
+      "step": 4498
+    },
+    {
+      "epoch": 2.231848082412809,
+      "grad_norm": 0.0702316925680538,
+      "learning_rate": 4.103753831745025e-06,
+      "loss": 0.4661,
+      "step": 4499
+    },
+    {
+      "epoch": 2.232344545116048,
+      "grad_norm": 0.07315427281933387,
+      "learning_rate": 4.1018308599241245e-06,
+      "loss": 0.4681,
+      "step": 4500
+    },
+    {
+      "epoch": 2.2328410078192875,
+      "grad_norm": 0.0709859756454207,
+      "learning_rate": 4.099908025373973e-06,
+      "loss": 0.462,
+      "step": 4501
+    },
+    {
+      "epoch": 2.233337470522527,
+      "grad_norm": 0.07390214333546574,
+      "learning_rate": 4.097985328388449e-06,
+      "loss": 0.4674,
+      "step": 4502
+    },
+    {
+      "epoch": 2.2338339332257666,
+      "grad_norm": 0.07310495770803771,
+      "learning_rate": 4.096062769261405e-06,
+      "loss": 0.444,
+      "step": 4503
+    },
+    {
+      "epoch": 2.234330395929006,
+      "grad_norm": 0.07318819058228206,
+      "learning_rate": 4.094140348286674e-06,
+      "loss": 0.4579,
+      "step": 4504
+    },
+    {
+      "epoch": 2.2348268586322453,
+      "grad_norm": 0.07154890926200361,
+      "learning_rate": 4.092218065758065e-06,
+      "loss": 0.4827,
+      "step": 4505
+    },
+    {
+      "epoch": 2.2353233213354846,
+      "grad_norm": 0.07123253786534825,
+      "learning_rate": 4.0902959219693695e-06,
+      "loss": 0.4058,
+      "step": 4506
+    },
+    {
+      "epoch": 2.235819784038724,
+      "grad_norm": 0.0743483487279238,
+      "learning_rate": 4.088373917214357e-06,
+      "loss": 0.467,
+      "step": 4507
+    },
+    {
+      "epoch": 2.2363162467419633,
+      "grad_norm": 0.0739892303014404,
+      "learning_rate": 4.086452051786774e-06,
+      "loss": 0.4671,
+      "step": 4508
+    },
+    {
+      "epoch": 2.236812709445203,
+      "grad_norm": 0.07372586400995458,
+      "learning_rate": 4.0845303259803485e-06,
+      "loss": 0.4632,
+      "step": 4509
+    },
+    {
+      "epoch": 2.2373091721484424,
+      "grad_norm": 0.07350804599337754,
+      "learning_rate": 4.082608740088782e-06,
+      "loss": 0.4676,
+      "step": 4510
+    },
+    {
+      "epoch": 2.2378056348516817,
+      "grad_norm": 0.07078113633327661,
+      "learning_rate": 4.080687294405761e-06,
+      "loss": 0.456,
+      "step": 4511
+    },
+    {
+      "epoch": 2.238302097554921,
+      "grad_norm": 0.07305462569174823,
+      "learning_rate": 4.078765989224947e-06,
+      "loss": 0.4667,
+      "step": 4512
+    },
+    {
+      "epoch": 2.238798560258161,
+      "grad_norm": 0.07116321137722682,
+      "learning_rate": 4.0768448248399836e-06,
+      "loss": 0.4246,
+      "step": 4513
+    },
+    {
+      "epoch": 2.2392950229614,
+      "grad_norm": 0.07662697234502294,
+      "learning_rate": 4.074923801544485e-06,
+      "loss": 0.489,
+      "step": 4514
+    },
+    {
+      "epoch": 2.2397914856646395,
+      "grad_norm": 0.07701510571760695,
+      "learning_rate": 4.07300291963205e-06,
+      "loss": 0.4552,
+      "step": 4515
+    },
+    {
+      "epoch": 2.240287948367879,
+      "grad_norm": 0.07124722345033124,
+      "learning_rate": 4.071082179396257e-06,
+      "loss": 0.4634,
+      "step": 4516
+    },
+    {
+      "epoch": 2.240784411071118,
+      "grad_norm": 0.07170234298992674,
+      "learning_rate": 4.069161581130658e-06,
+      "loss": 0.4505,
+      "step": 4517
+    },
+    {
+      "epoch": 2.2412808737743575,
+      "grad_norm": 0.06941445386725237,
+      "learning_rate": 4.067241125128786e-06,
+      "loss": 0.4254,
+      "step": 4518
+    },
+    {
+      "epoch": 2.2417773364775972,
+      "grad_norm": 0.07459034297591921,
+      "learning_rate": 4.065320811684152e-06,
+      "loss": 0.5102,
+      "step": 4519
+    },
+    {
+      "epoch": 2.2422737991808366,
+      "grad_norm": 0.07136337958486601,
+      "learning_rate": 4.063400641090246e-06,
+      "loss": 0.4246,
+      "step": 4520
+    },
+    {
+      "epoch": 2.242770261884076,
+      "grad_norm": 0.07180395198419941,
+      "learning_rate": 4.0614806136405355e-06,
+      "loss": 0.4437,
+      "step": 4521
+    },
+    {
+      "epoch": 2.2432667245873152,
+      "grad_norm": 0.06959895930282148,
+      "learning_rate": 4.059560729628465e-06,
+      "loss": 0.4395,
+      "step": 4522
+    },
+    {
+      "epoch": 2.243763187290555,
+      "grad_norm": 0.07319596333666036,
+      "learning_rate": 4.0576409893474566e-06,
+      "loss": 0.4562,
+      "step": 4523
+    },
+    {
+      "epoch": 2.2442596499937943,
+      "grad_norm": 0.07184203088659276,
+      "learning_rate": 4.0557213930909146e-06,
+      "loss": 0.4495,
+      "step": 4524
+    },
+    {
+      "epoch": 2.2447561126970337,
+      "grad_norm": 0.07533243438464896,
+      "learning_rate": 4.053801941152218e-06,
+      "loss": 0.4814,
+      "step": 4525
+    },
+    {
+      "epoch": 2.245252575400273,
+      "grad_norm": 0.0744110741286213,
+      "learning_rate": 4.051882633824723e-06,
+      "loss": 0.4598,
+      "step": 4526
+    },
+    {
+      "epoch": 2.2457490381035123,
+      "grad_norm": 0.07628997779462912,
+      "learning_rate": 4.049963471401765e-06,
+      "loss": 0.4827,
+      "step": 4527
+    },
+    {
+      "epoch": 2.2462455008067517,
+      "grad_norm": 0.07467086146241933,
+      "learning_rate": 4.048044454176658e-06,
+      "loss": 0.4991,
+      "step": 4528
+    },
+    {
+      "epoch": 2.2467419635099914,
+      "grad_norm": 0.07426614710662187,
+      "learning_rate": 4.046125582442695e-06,
+      "loss": 0.4705,
+      "step": 4529
+    },
+    {
+      "epoch": 2.2472384262132308,
+      "grad_norm": 0.07379210496441937,
+      "learning_rate": 4.04420685649314e-06,
+      "loss": 0.4504,
+      "step": 4530
+    },
+    {
+      "epoch": 2.24773488891647,
+      "grad_norm": 0.07265582803024827,
+      "learning_rate": 4.0422882766212416e-06,
+      "loss": 0.4326,
+      "step": 4531
+    },
+    {
+      "epoch": 2.2482313516197094,
+      "grad_norm": 0.07594842599950401,
+      "learning_rate": 4.040369843120226e-06,
+      "loss": 0.458,
+      "step": 4532
+    },
+    {
+      "epoch": 2.248727814322949,
+      "grad_norm": 0.06816504846270663,
+      "learning_rate": 4.038451556283292e-06,
+      "loss": 0.4504,
+      "step": 4533
+    },
+    {
+      "epoch": 2.2492242770261885,
+      "grad_norm": 0.07282920440283609,
+      "learning_rate": 4.036533416403621e-06,
+      "loss": 0.4792,
+      "step": 4534
+    },
+    {
+      "epoch": 2.249720739729428,
+      "grad_norm": 0.07628329549507812,
+      "learning_rate": 4.034615423774369e-06,
+      "loss": 0.498,
+      "step": 4535
+    },
+    {
+      "epoch": 2.250217202432667,
+      "grad_norm": 0.07438066243638336,
+      "learning_rate": 4.03269757868867e-06,
+      "loss": 0.4489,
+      "step": 4536
+    },
+    {
+      "epoch": 2.250217202432667,
+      "eval_loss": 0.5172906517982483,
+      "eval_runtime": 258.8938,
+      "eval_samples_per_second": 117.241,
+      "eval_steps_per_second": 14.659,
+      "step": 4536
+    },
+    {
+      "epoch": 2.2507136651359065,
+      "grad_norm": 0.07331941505145612,
+      "learning_rate": 4.030779881439639e-06,
+      "loss": 0.4721,
+      "step": 4537
+    },
+    {
+      "epoch": 2.251210127839146,
+      "grad_norm": 0.07216350901232152,
+      "learning_rate": 4.02886233232036e-06,
+      "loss": 0.4654,
+      "step": 4538
+    },
+    {
+      "epoch": 2.2517065905423856,
+      "grad_norm": 0.07347245654808018,
+      "learning_rate": 4.026944931623905e-06,
+      "loss": 0.4839,
+      "step": 4539
+    },
+    {
+      "epoch": 2.252203053245625,
+      "grad_norm": 0.07355019456418468,
+      "learning_rate": 4.025027679643314e-06,
+      "loss": 0.4442,
+      "step": 4540
+    },
+    {
+      "epoch": 2.2526995159488643,
+      "grad_norm": 0.07176085614756975,
+      "learning_rate": 4.02311057667161e-06,
+      "loss": 0.4591,
+      "step": 4541
+    },
+    {
+      "epoch": 2.2531959786521036,
+      "grad_norm": 0.07193387754107909,
+      "learning_rate": 4.0211936230017915e-06,
+      "loss": 0.4232,
+      "step": 4542
+    },
+    {
+      "epoch": 2.2536924413553434,
+      "grad_norm": 0.07306749615179373,
+      "learning_rate": 4.019276818926833e-06,
+      "loss": 0.4709,
+      "step": 4543
+    },
+    {
+      "epoch": 2.2541889040585827,
+      "grad_norm": 0.07275645118232545,
+      "learning_rate": 4.017360164739687e-06,
+      "loss": 0.4711,
+      "step": 4544
+    },
+    {
+      "epoch": 2.254685366761822,
+      "grad_norm": 0.07207487833978099,
+      "learning_rate": 4.015443660733288e-06,
+      "loss": 0.463,
+      "step": 4545
+    },
+    {
+      "epoch": 2.2551818294650614,
+      "grad_norm": 0.07244408374466355,
+      "learning_rate": 4.0135273072005374e-06,
+      "loss": 0.4418,
+      "step": 4546
+    },
+    {
+      "epoch": 2.2556782921683007,
+      "grad_norm": 0.07541937868644752,
+      "learning_rate": 4.01161110443432e-06,
+      "loss": 0.4724,
+      "step": 4547
+    },
+    {
+      "epoch": 2.25617475487154,
+      "grad_norm": 0.07245802199577502,
+      "learning_rate": 4.009695052727499e-06,
+      "loss": 0.456,
+      "step": 4548
+    },
+    {
+      "epoch": 2.25667121757478,
+      "grad_norm": 0.07409608483995006,
+      "learning_rate": 4.00777915237291e-06,
+      "loss": 0.5128,
+      "step": 4549
+    },
+    {
+      "epoch": 2.257167680278019,
+      "grad_norm": 0.0720396749298917,
+      "learning_rate": 4.005863403663368e-06,
+      "loss": 0.4677,
+      "step": 4550
+    },
+    {
+      "epoch": 2.2576641429812585,
+      "grad_norm": 0.07528407483916275,
+      "learning_rate": 4.0039478068916655e-06,
+      "loss": 0.4787,
+      "step": 4551
+    },
+    {
+      "epoch": 2.258160605684498,
+      "grad_norm": 0.07836436675954314,
+      "learning_rate": 4.002032362350571e-06,
+      "loss": 0.5044,
+      "step": 4552
+    },
+    {
+      "epoch": 2.2586570683877376,
+      "grad_norm": 0.07308686718367485,
+      "learning_rate": 4.000117070332828e-06,
+      "loss": 0.4483,
+      "step": 4553
+    },
+    {
+      "epoch": 2.259153531090977,
+      "grad_norm": 0.07185200327263667,
+      "learning_rate": 3.998201931131159e-06,
+      "loss": 0.4353,
+      "step": 4554
+    },
+    {
+      "epoch": 2.2596499937942163,
+      "grad_norm": 0.07071000554552216,
+      "learning_rate": 3.996286945038263e-06,
+      "loss": 0.4598,
+      "step": 4555
+    },
+    {
+      "epoch": 2.2601464564974556,
+      "grad_norm": 0.07354071205498476,
+      "learning_rate": 3.9943721123468124e-06,
+      "loss": 0.4366,
+      "step": 4556
+    },
+    {
+      "epoch": 2.260642919200695,
+      "grad_norm": 0.07360684253852036,
+      "learning_rate": 3.992457433349461e-06,
+      "loss": 0.4447,
+      "step": 4557
+    },
+    {
+      "epoch": 2.2611393819039343,
+      "grad_norm": 0.07349793093279754,
+      "learning_rate": 3.990542908338837e-06,
+      "loss": 0.4723,
+      "step": 4558
+    },
+    {
+      "epoch": 2.261635844607174,
+      "grad_norm": 0.07447147137326113,
+      "learning_rate": 3.988628537607544e-06,
+      "loss": 0.4786,
+      "step": 4559
+    },
+    {
+      "epoch": 2.2621323073104134,
+      "grad_norm": 0.07332903082275616,
+      "learning_rate": 3.986714321448162e-06,
+      "loss": 0.444,
+      "step": 4560
+    },
+    {
+      "epoch": 2.2626287700136527,
+      "grad_norm": 0.0732381531192602,
+      "learning_rate": 3.984800260153251e-06,
+      "loss": 0.4744,
+      "step": 4561
+    },
+    {
+      "epoch": 2.263125232716892,
+      "grad_norm": 0.07108143720306151,
+      "learning_rate": 3.982886354015341e-06,
+      "loss": 0.4677,
+      "step": 4562
+    },
+    {
+      "epoch": 2.263621695420132,
+      "grad_norm": 0.07661093044858573,
+      "learning_rate": 3.980972603326945e-06,
+      "loss": 0.4656,
+      "step": 4563
+    },
+    {
+      "epoch": 2.264118158123371,
+      "grad_norm": 0.07662723476502775,
+      "learning_rate": 3.979059008380547e-06,
+      "loss": 0.4559,
+      "step": 4564
+    },
+    {
+      "epoch": 2.2646146208266105,
+      "grad_norm": 0.06964828076274196,
+      "learning_rate": 3.97714556946861e-06,
+      "loss": 0.4762,
+      "step": 4565
+    },
+    {
+      "epoch": 2.26511108352985,
+      "grad_norm": 0.07254551987290445,
+      "learning_rate": 3.9752322868835715e-06,
+      "loss": 0.4519,
+      "step": 4566
+    },
+    {
+      "epoch": 2.265607546233089,
+      "grad_norm": 0.07699102489984336,
+      "learning_rate": 3.973319160917848e-06,
+      "loss": 0.4897,
+      "step": 4567
+    },
+    {
+      "epoch": 2.2661040089363285,
+      "grad_norm": 0.07175368111713772,
+      "learning_rate": 3.971406191863829e-06,
+      "loss": 0.4541,
+      "step": 4568
+    },
+    {
+      "epoch": 2.2666004716395682,
+      "grad_norm": 0.07336946290089652,
+      "learning_rate": 3.969493380013882e-06,
+      "loss": 0.4348,
+      "step": 4569
+    },
+    {
+      "epoch": 2.2670969343428076,
+      "grad_norm": 0.07404276252315657,
+      "learning_rate": 3.967580725660348e-06,
+      "loss": 0.4694,
+      "step": 4570
+    },
+    {
+      "epoch": 2.267593397046047,
+      "grad_norm": 0.07209603130099349,
+      "learning_rate": 3.965668229095546e-06,
+      "loss": 0.4415,
+      "step": 4571
+    },
+    {
+      "epoch": 2.2680898597492862,
+      "grad_norm": 0.07126832127112002,
+      "learning_rate": 3.963755890611772e-06,
+      "loss": 0.4545,
+      "step": 4572
+    },
+    {
+      "epoch": 2.2685863224525256,
+      "grad_norm": 0.07424826158725521,
+      "learning_rate": 3.961843710501294e-06,
+      "loss": 0.4582,
+      "step": 4573
+    },
+    {
+      "epoch": 2.2690827851557653,
+      "grad_norm": 0.07626947433796435,
+      "learning_rate": 3.959931689056362e-06,
+      "loss": 0.4714,
+      "step": 4574
+    },
+    {
+      "epoch": 2.2695792478590047,
+      "grad_norm": 0.07459916034486681,
+      "learning_rate": 3.958019826569192e-06,
+      "loss": 0.4339,
+      "step": 4575
+    },
+    {
+      "epoch": 2.270075710562244,
+      "grad_norm": 0.07557272071979547,
+      "learning_rate": 3.956108123331986e-06,
+      "loss": 0.483,
+      "step": 4576
+    },
+    {
+      "epoch": 2.2705721732654833,
+      "grad_norm": 0.07299559261994196,
+      "learning_rate": 3.954196579636918e-06,
+      "loss": 0.4832,
+      "step": 4577
+    },
+    {
+      "epoch": 2.2710686359687227,
+      "grad_norm": 0.07190351112626157,
+      "learning_rate": 3.952285195776132e-06,
+      "loss": 0.4256,
+      "step": 4578
+    },
+    {
+      "epoch": 2.2715650986719624,
+      "grad_norm": 0.0746426816115579,
+      "learning_rate": 3.950373972041755e-06,
+      "loss": 0.4581,
+      "step": 4579
+    },
+    {
+      "epoch": 2.2720615613752018,
+      "grad_norm": 0.0758770115143761,
+      "learning_rate": 3.948462908725888e-06,
+      "loss": 0.4602,
+      "step": 4580
+    },
+    {
+      "epoch": 2.272558024078441,
+      "grad_norm": 0.07214617535370865,
+      "learning_rate": 3.946552006120604e-06,
+      "loss": 0.4662,
+      "step": 4581
+    },
+    {
+      "epoch": 2.2730544867816804,
+      "grad_norm": 0.0734472665053787,
+      "learning_rate": 3.944641264517956e-06,
+      "loss": 0.4376,
+      "step": 4582
+    },
+    {
+      "epoch": 2.2735509494849198,
+      "grad_norm": 0.07316103188106508,
+      "learning_rate": 3.942730684209969e-06,
+      "loss": 0.4521,
+      "step": 4583
+    },
+    {
+      "epoch": 2.2740474121881595,
+      "grad_norm": 0.07210985009452146,
+      "learning_rate": 3.940820265488644e-06,
+      "loss": 0.4538,
+      "step": 4584
+    },
+    {
+      "epoch": 2.274543874891399,
+      "grad_norm": 0.07222754057585111,
+      "learning_rate": 3.9389100086459604e-06,
+      "loss": 0.4665,
+      "step": 4585
+    },
+    {
+      "epoch": 2.275040337594638,
+      "grad_norm": 0.07140562198443162,
+      "learning_rate": 3.936999913973868e-06,
+      "loss": 0.4496,
+      "step": 4586
+    },
+    {
+      "epoch": 2.2755368002978775,
+      "grad_norm": 0.07177302930612432,
+      "learning_rate": 3.9350899817642946e-06,
+      "loss": 0.4618,
+      "step": 4587
+    },
+    {
+      "epoch": 2.276033263001117,
+      "grad_norm": 0.07207407681313742,
+      "learning_rate": 3.933180212309143e-06,
+      "loss": 0.4576,
+      "step": 4588
+    },
+    {
+      "epoch": 2.2765297257043566,
+      "grad_norm": 0.07314155056855802,
+      "learning_rate": 3.93127060590029e-06,
+      "loss": 0.4324,
+      "step": 4589
+    },
+    {
+      "epoch": 2.277026188407596,
+      "grad_norm": 0.0750859987810042,
+      "learning_rate": 3.929361162829591e-06,
+      "loss": 0.4596,
+      "step": 4590
+    },
+    {
+      "epoch": 2.2775226511108353,
+      "grad_norm": 0.07556163426963661,
+      "learning_rate": 3.9274518833888704e-06,
+      "loss": 0.4658,
+      "step": 4591
+    },
+    {
+      "epoch": 2.2780191138140746,
+      "grad_norm": 0.07079600144616138,
+      "learning_rate": 3.925542767869933e-06,
+      "loss": 0.4845,
+      "step": 4592
+    },
+    {
+      "epoch": 2.278515576517314,
+      "grad_norm": 0.07579154104447022,
+      "learning_rate": 3.923633816564558e-06,
+      "loss": 0.4973,
+      "step": 4593
+    },
+    {
+      "epoch": 2.2790120392205537,
+      "grad_norm": 0.07561556923014369,
+      "learning_rate": 3.921725029764494e-06,
+      "loss": 0.4806,
+      "step": 4594
+    },
+    {
+      "epoch": 2.279508501923793,
+      "grad_norm": 0.07106848910421479,
+      "learning_rate": 3.91981640776147e-06,
+      "loss": 0.4363,
+      "step": 4595
+    },
+    {
+      "epoch": 2.2800049646270324,
+      "grad_norm": 0.07222499569491124,
+      "learning_rate": 3.9179079508471876e-06,
+      "loss": 0.4437,
+      "step": 4596
+    },
+    {
+      "epoch": 2.2805014273302717,
+      "grad_norm": 0.07272568112911992,
+      "learning_rate": 3.915999659313328e-06,
+      "loss": 0.462,
+      "step": 4597
+    },
+    {
+      "epoch": 2.280997890033511,
+      "grad_norm": 0.0749925291982409,
+      "learning_rate": 3.914091533451537e-06,
+      "loss": 0.4757,
+      "step": 4598
+    },
+    {
+      "epoch": 2.281494352736751,
+      "grad_norm": 0.0733452428212777,
+      "learning_rate": 3.9121835735534446e-06,
+      "loss": 0.4628,
+      "step": 4599
+    },
+    {
+      "epoch": 2.28199081543999,
+      "grad_norm": 0.07469474773325638,
+      "learning_rate": 3.910275779910651e-06,
+      "loss": 0.4348,
+      "step": 4600
+    },
+    {
+      "epoch": 2.2824872781432295,
+      "grad_norm": 0.06954130209101946,
+      "learning_rate": 3.908368152814733e-06,
+      "loss": 0.4267,
+      "step": 4601
+    },
+    {
+      "epoch": 2.282983740846469,
+      "grad_norm": 0.07280494243859224,
+      "learning_rate": 3.906460692557239e-06,
+      "loss": 0.464,
+      "step": 4602
+    },
+    {
+      "epoch": 2.283480203549708,
+      "grad_norm": 0.07491485368077401,
+      "learning_rate": 3.904553399429695e-06,
+      "loss": 0.4673,
+      "step": 4603
+    },
+    {
+      "epoch": 2.283976666252948,
+      "grad_norm": 0.0725255163654333,
+      "learning_rate": 3.902646273723599e-06,
+      "loss": 0.4582,
+      "step": 4604
+    },
+    {
+      "epoch": 2.2844731289561873,
+      "grad_norm": 0.07376370932884879,
+      "learning_rate": 3.900739315730426e-06,
+      "loss": 0.4774,
+      "step": 4605
+    },
+    {
+      "epoch": 2.2849695916594266,
+      "grad_norm": 0.07147727373362511,
+      "learning_rate": 3.898832525741624e-06,
+      "loss": 0.4488,
+      "step": 4606
+    },
+    {
+      "epoch": 2.285466054362666,
+      "grad_norm": 0.0707745677288137,
+      "learning_rate": 3.896925904048614e-06,
+      "loss": 0.4299,
+      "step": 4607
+    },
+    {
+      "epoch": 2.2859625170659053,
+      "grad_norm": 0.07233717904719753,
+      "learning_rate": 3.895019450942793e-06,
+      "loss": 0.4501,
+      "step": 4608
+    },
+    {
+      "epoch": 2.286458979769145,
+      "grad_norm": 0.07081364354190414,
+      "learning_rate": 3.893113166715533e-06,
+      "loss": 0.4609,
+      "step": 4609
+    },
+    {
+      "epoch": 2.2869554424723844,
+      "grad_norm": 0.0724291518918608,
+      "learning_rate": 3.891207051658177e-06,
+      "loss": 0.4768,
+      "step": 4610
+    },
+    {
+      "epoch": 2.2874519051756237,
+      "grad_norm": 0.07451520468130642,
+      "learning_rate": 3.889301106062044e-06,
+      "loss": 0.4979,
+      "step": 4611
+    },
+    {
+      "epoch": 2.287948367878863,
+      "grad_norm": 0.07300620972050165,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.4579,
+      "step": 4612
+    },
+    {
+      "epoch": 2.2884448305821024,
+      "grad_norm": 0.06996007226973054,
+      "learning_rate": 3.885489724418599e-06,
+      "loss": 0.4366,
+      "step": 4613
+    },
+    {
+      "epoch": 2.288941293285342,
+      "grad_norm": 0.07132885315218875,
+      "learning_rate": 3.883584288953794e-06,
+      "loss": 0.4526,
+      "step": 4614
+    },
+    {
+      "epoch": 2.2894377559885815,
+      "grad_norm": 0.06908714335809077,
+      "learning_rate": 3.88167902411523e-06,
+      "loss": 0.4333,
+      "step": 4615
+    },
+    {
+      "epoch": 2.289934218691821,
+      "grad_norm": 0.07239860976789113,
+      "learning_rate": 3.8797739301940965e-06,
+      "loss": 0.4796,
+      "step": 4616
+    },
+    {
+      "epoch": 2.29043068139506,
+      "grad_norm": 0.07372826754791378,
+      "learning_rate": 3.877869007481557e-06,
+      "loss": 0.471,
+      "step": 4617
+    },
+    {
+      "epoch": 2.2909271440982995,
+      "grad_norm": 0.07371280520150718,
+      "learning_rate": 3.875964256268747e-06,
+      "loss": 0.4527,
+      "step": 4618
+    },
+    {
+      "epoch": 2.2914236068015392,
+      "grad_norm": 0.07959721066448272,
+      "learning_rate": 3.874059676846778e-06,
+      "loss": 0.4963,
+      "step": 4619
+    },
+    {
+      "epoch": 2.2919200695047786,
+      "grad_norm": 0.07356143469580767,
+      "learning_rate": 3.8721552695067334e-06,
+      "loss": 0.4834,
+      "step": 4620
+    },
+    {
+      "epoch": 2.292416532208018,
+      "grad_norm": 0.07043793579089817,
+      "learning_rate": 3.8702510345396736e-06,
+      "loss": 0.4365,
+      "step": 4621
+    },
+    {
+      "epoch": 2.292912994911257,
+      "grad_norm": 0.07389661566696645,
+      "learning_rate": 3.868346972236629e-06,
+      "loss": 0.4515,
+      "step": 4622
+    },
+    {
+      "epoch": 2.2934094576144965,
+      "grad_norm": 0.07056702907409622,
+      "learning_rate": 3.866443082888605e-06,
+      "loss": 0.4476,
+      "step": 4623
+    },
+    {
+      "epoch": 2.2939059203177363,
+      "grad_norm": 0.07431132634179353,
+      "learning_rate": 3.864539366786579e-06,
+      "loss": 0.4317,
+      "step": 4624
+    },
+    {
+      "epoch": 2.2944023830209757,
+      "grad_norm": 0.07411892259856963,
+      "learning_rate": 3.862635824221508e-06,
+      "loss": 0.4711,
+      "step": 4625
+    },
+    {
+      "epoch": 2.294898845724215,
+      "grad_norm": 0.07390766032523606,
+      "learning_rate": 3.860732455484314e-06,
+      "loss": 0.4745,
+      "step": 4626
+    },
+    {
+      "epoch": 2.2953953084274543,
+      "grad_norm": 0.07129476349377506,
+      "learning_rate": 3.8588292608658964e-06,
+      "loss": 0.4645,
+      "step": 4627
+    },
+    {
+      "epoch": 2.2958917711306936,
+      "grad_norm": 0.07595256250205015,
+      "learning_rate": 3.856926240657129e-06,
+      "loss": 0.4909,
+      "step": 4628
+    },
+    {
+      "epoch": 2.296388233833933,
+      "grad_norm": 0.07254923696850328,
+      "learning_rate": 3.855023395148858e-06,
+      "loss": 0.4317,
+      "step": 4629
+    },
+    {
+      "epoch": 2.2968846965371728,
+      "grad_norm": 0.07350447502657023,
+      "learning_rate": 3.853120724631903e-06,
+      "loss": 0.4751,
+      "step": 4630
+    },
+    {
+      "epoch": 2.297381159240412,
+      "grad_norm": 0.07131503869721514,
+      "learning_rate": 3.8512182293970565e-06,
+      "loss": 0.4662,
+      "step": 4631
+    },
+    {
+      "epoch": 2.2978776219436514,
+      "grad_norm": 0.06930825093142902,
+      "learning_rate": 3.849315909735084e-06,
+      "loss": 0.4226,
+      "step": 4632
+    },
+    {
+      "epoch": 2.2983740846468907,
+      "grad_norm": 0.07264603579286186,
+      "learning_rate": 3.847413765936724e-06,
+      "loss": 0.4282,
+      "step": 4633
+    },
+    {
+      "epoch": 2.2988705473501305,
+      "grad_norm": 0.07438691342521027,
+      "learning_rate": 3.845511798292692e-06,
+      "loss": 0.461,
+      "step": 4634
+    },
+    {
+      "epoch": 2.29936701005337,
+      "grad_norm": 0.07688398707415792,
+      "learning_rate": 3.843610007093669e-06,
+      "loss": 0.4229,
+      "step": 4635
+    },
+    {
+      "epoch": 2.299863472756609,
+      "grad_norm": 0.0719622667679788,
+      "learning_rate": 3.841708392630315e-06,
+      "loss": 0.4517,
+      "step": 4636
+    },
+    {
+      "epoch": 2.3003599354598485,
+      "grad_norm": 0.07299102365171561,
+      "learning_rate": 3.839806955193262e-06,
+      "loss": 0.4466,
+      "step": 4637
+    },
+    {
+      "epoch": 2.300856398163088,
+      "grad_norm": 0.07375205625152867,
+      "learning_rate": 3.837905695073114e-06,
+      "loss": 0.447,
+      "step": 4638
+    },
+    {
+      "epoch": 2.301352860866327,
+      "grad_norm": 0.06837142772677776,
+      "learning_rate": 3.836004612560447e-06,
+      "loss": 0.4203,
+      "step": 4639
+    },
+    {
+      "epoch": 2.301849323569567,
+      "grad_norm": 0.07062241886373752,
+      "learning_rate": 3.8341037079458125e-06,
+      "loss": 0.4407,
+      "step": 4640
+    },
+    {
+      "epoch": 2.3023457862728063,
+      "grad_norm": 0.07010432830866563,
+      "learning_rate": 3.8322029815197335e-06,
+      "loss": 0.4299,
+      "step": 4641
+    },
+    {
+      "epoch": 2.3028422489760456,
+      "grad_norm": 0.07444368732307707,
+      "learning_rate": 3.830302433572704e-06,
+      "loss": 0.472,
+      "step": 4642
+    },
+    {
+      "epoch": 2.303338711679285,
+      "grad_norm": 0.07268106792030828,
+      "learning_rate": 3.828402064395191e-06,
+      "loss": 0.4365,
+      "step": 4643
+    },
+    {
+      "epoch": 2.3038351743825247,
+      "grad_norm": 0.07458663611375892,
+      "learning_rate": 3.8265018742776374e-06,
+      "loss": 0.4595,
+      "step": 4644
+    },
+    {
+      "epoch": 2.304331637085764,
+      "grad_norm": 0.07467767197871673,
+      "learning_rate": 3.824601863510459e-06,
+      "loss": 0.4958,
+      "step": 4645
+    },
+    {
+      "epoch": 2.3048280997890034,
+      "grad_norm": 0.07201811203621303,
+      "learning_rate": 3.822702032384038e-06,
+      "loss": 0.4437,
+      "step": 4646
+    },
+    {
+      "epoch": 2.3053245624922427,
+      "grad_norm": 0.0692424696265978,
+      "learning_rate": 3.820802381188735e-06,
+      "loss": 0.4488,
+      "step": 4647
+    },
+    {
+      "epoch": 2.305821025195482,
+      "grad_norm": 0.07300687048478505,
+      "learning_rate": 3.818902910214881e-06,
+      "loss": 0.4693,
+      "step": 4648
+    },
+    {
+      "epoch": 2.3063174878987214,
+      "grad_norm": 0.07596628725546059,
+      "learning_rate": 3.817003619752779e-06,
+      "loss": 0.4634,
+      "step": 4649
+    },
+    {
+      "epoch": 2.306813950601961,
+      "grad_norm": 0.07268075534926753,
+      "learning_rate": 3.8151045100927075e-06,
+      "loss": 0.442,
+      "step": 4650
+    },
+    {
+      "epoch": 2.3073104133052005,
+      "grad_norm": 0.07167150254029588,
+      "learning_rate": 3.8132055815249127e-06,
+      "loss": 0.4477,
+      "step": 4651
+    },
+    {
+      "epoch": 2.30780687600844,
+      "grad_norm": 0.0713520151906273,
+      "learning_rate": 3.811306834339615e-06,
+      "loss": 0.4492,
+      "step": 4652
+    },
+    {
+      "epoch": 2.308303338711679,
+      "grad_norm": 0.07299337925654024,
+      "learning_rate": 3.809408268827009e-06,
+      "loss": 0.4426,
+      "step": 4653
+    },
+    {
+      "epoch": 2.308799801414919,
+      "grad_norm": 0.07345951629738362,
+      "learning_rate": 3.8075098852772607e-06,
+      "loss": 0.4836,
+      "step": 4654
+    },
+    {
+      "epoch": 2.3092962641181582,
+      "grad_norm": 0.07194786674780375,
+      "learning_rate": 3.8056116839805048e-06,
+      "loss": 0.4386,
+      "step": 4655
+    },
+    {
+      "epoch": 2.3097927268213976,
+      "grad_norm": 0.07183541026450214,
+      "learning_rate": 3.8037136652268524e-06,
+      "loss": 0.4489,
+      "step": 4656
+    },
+    {
+      "epoch": 2.310289189524637,
+      "grad_norm": 0.07199156509588549,
+      "learning_rate": 3.801815829306388e-06,
+      "loss": 0.4482,
+      "step": 4657
+    },
+    {
+      "epoch": 2.3107856522278762,
+      "grad_norm": 0.07300167018565622,
+      "learning_rate": 3.7999181765091597e-06,
+      "loss": 0.4555,
+      "step": 4658
+    },
+    {
+      "epoch": 2.3112821149311156,
+      "grad_norm": 0.0727637096854206,
+      "learning_rate": 3.7980207071251967e-06,
+      "loss": 0.4541,
+      "step": 4659
+    },
+    {
+      "epoch": 2.3117785776343553,
+      "grad_norm": 0.07124656239632847,
+      "learning_rate": 3.7961234214444963e-06,
+      "loss": 0.4468,
+      "step": 4660
+    },
+    {
+      "epoch": 2.3122750403375947,
+      "grad_norm": 0.07460627268561644,
+      "learning_rate": 3.7942263197570297e-06,
+      "loss": 0.4975,
+      "step": 4661
+    },
+    {
+      "epoch": 2.312771503040834,
+      "grad_norm": 0.06854605918792946,
+      "learning_rate": 3.7923294023527353e-06,
+      "loss": 0.4466,
+      "step": 4662
+    },
+    {
+      "epoch": 2.3132679657440733,
+      "grad_norm": 0.07300630381478836,
+      "learning_rate": 3.7904326695215283e-06,
+      "loss": 0.4615,
+      "step": 4663
+    },
+    {
+      "epoch": 2.313764428447313,
+      "grad_norm": 0.07456192765025539,
+      "learning_rate": 3.788536121553294e-06,
+      "loss": 0.4685,
+      "step": 4664
+    },
+    {
+      "epoch": 2.3142608911505524,
+      "grad_norm": 0.07439993755686335,
+      "learning_rate": 3.786639758737889e-06,
+      "loss": 0.5119,
+      "step": 4665
+    },
+    {
+      "epoch": 2.3147573538537918,
+      "grad_norm": 0.0767387688734663,
+      "learning_rate": 3.7847435813651436e-06,
+      "loss": 0.5161,
+      "step": 4666
+    },
+    {
+      "epoch": 2.315253816557031,
+      "grad_norm": 0.07332142670152268,
+      "learning_rate": 3.782847589724855e-06,
+      "loss": 0.4265,
+      "step": 4667
+    },
+    {
+      "epoch": 2.3157502792602704,
+      "grad_norm": 0.07089426764347762,
+      "learning_rate": 3.7809517841067976e-06,
+      "loss": 0.4339,
+      "step": 4668
+    },
+    {
+      "epoch": 2.3162467419635098,
+      "grad_norm": 0.07230696778867758,
+      "learning_rate": 3.7790561648007136e-06,
+      "loss": 0.4289,
+      "step": 4669
+    },
+    {
+      "epoch": 2.3167432046667495,
+      "grad_norm": 0.07350340322008529,
+      "learning_rate": 3.777160732096318e-06,
+      "loss": 0.454,
+      "step": 4670
+    },
+    {
+      "epoch": 2.317239667369989,
+      "grad_norm": 0.07118246729742177,
+      "learning_rate": 3.7752654862832995e-06,
+      "loss": 0.4501,
+      "step": 4671
+    },
+    {
+      "epoch": 2.317736130073228,
+      "grad_norm": 0.07416497221248512,
+      "learning_rate": 3.773370427651313e-06,
+      "loss": 0.4617,
+      "step": 4672
+    },
+    {
+      "epoch": 2.3182325927764675,
+      "grad_norm": 0.07372197345098633,
+      "learning_rate": 3.7714755564899913e-06,
+      "loss": 0.4706,
+      "step": 4673
+    },
+    {
+      "epoch": 2.3187290554797073,
+      "grad_norm": 0.07601035801539498,
+      "learning_rate": 3.7695808730889307e-06,
+      "loss": 0.4793,
+      "step": 4674
+    },
+    {
+      "epoch": 2.3192255181829466,
+      "grad_norm": 0.07121752187002783,
+      "learning_rate": 3.7676863777377055e-06,
+      "loss": 0.4519,
+      "step": 4675
+    },
+    {
+      "epoch": 2.319721980886186,
+      "grad_norm": 0.0754180647916415,
+      "learning_rate": 3.7657920707258588e-06,
+      "loss": 0.4667,
+      "step": 4676
+    },
+    {
+      "epoch": 2.3202184435894253,
+      "grad_norm": 0.06998138945634584,
+      "learning_rate": 3.763897952342906e-06,
+      "loss": 0.4435,
+      "step": 4677
+    },
+    {
+      "epoch": 2.3207149062926646,
+      "grad_norm": 0.07131847650966808,
+      "learning_rate": 3.7620040228783305e-06,
+      "loss": 0.4423,
+      "step": 4678
+    },
+    {
+      "epoch": 2.321211368995904,
+      "grad_norm": 0.07090715276736156,
+      "learning_rate": 3.760110282621591e-06,
+      "loss": 0.4384,
+      "step": 4679
+    },
+    {
+      "epoch": 2.3217078316991437,
+      "grad_norm": 0.0734856641174945,
+      "learning_rate": 3.7582167318621136e-06,
+      "loss": 0.4365,
+      "step": 4680
+    },
+    {
+      "epoch": 2.322204294402383,
+      "grad_norm": 0.07982924966824814,
+      "learning_rate": 3.7563233708892993e-06,
+      "loss": 0.4773,
+      "step": 4681
+    },
+    {
+      "epoch": 2.3227007571056224,
+      "grad_norm": 0.07623726513167152,
+      "learning_rate": 3.7544301999925176e-06,
+      "loss": 0.4559,
+      "step": 4682
+    },
+    {
+      "epoch": 2.3231972198088617,
+      "grad_norm": 0.07411457548761642,
+      "learning_rate": 3.7525372194611075e-06,
+      "loss": 0.4538,
+      "step": 4683
+    },
+    {
+      "epoch": 2.3236936825121015,
+      "grad_norm": 0.07073415295298938,
+      "learning_rate": 3.750644429584382e-06,
+      "loss": 0.4418,
+      "step": 4684
+    },
+    {
+      "epoch": 2.324190145215341,
+      "grad_norm": 0.07123935814247413,
+      "learning_rate": 3.7487518306516237e-06,
+      "loss": 0.4542,
+      "step": 4685
+    },
+    {
+      "epoch": 2.32468660791858,
+      "grad_norm": 0.0755939983672755,
+      "learning_rate": 3.7468594229520854e-06,
+      "loss": 0.4836,
+      "step": 4686
+    },
+    {
+      "epoch": 2.3251830706218195,
+      "grad_norm": 0.07368993868071555,
+      "learning_rate": 3.744967206774993e-06,
+      "loss": 0.4825,
+      "step": 4687
+    },
+    {
+      "epoch": 2.325679533325059,
+      "grad_norm": 0.06850787814380067,
+      "learning_rate": 3.743075182409539e-06,
+      "loss": 0.442,
+      "step": 4688
+    },
+    {
+      "epoch": 2.326175996028298,
+      "grad_norm": 0.07061990343981624,
+      "learning_rate": 3.7411833501448924e-06,
+      "loss": 0.4516,
+      "step": 4689
+    },
+    {
+      "epoch": 2.326672458731538,
+      "grad_norm": 0.07427819657272731,
+      "learning_rate": 3.7392917102701854e-06,
+      "loss": 0.4821,
+      "step": 4690
+    },
+    {
+      "epoch": 2.3271689214347773,
+      "grad_norm": 0.07667376209001088,
+      "learning_rate": 3.737400263074526e-06,
+      "loss": 0.4857,
+      "step": 4691
+    },
+    {
+      "epoch": 2.3276653841380166,
+      "grad_norm": 0.07378650706390087,
+      "learning_rate": 3.7355090088469924e-06,
+      "loss": 0.473,
+      "step": 4692
+    },
+    {
+      "epoch": 2.328161846841256,
+      "grad_norm": 0.07232148035439803,
+      "learning_rate": 3.733617947876633e-06,
+      "loss": 0.4687,
+      "step": 4693
+    },
+    {
+      "epoch": 2.3286583095444957,
+      "grad_norm": 0.07035460992891325,
+      "learning_rate": 3.731727080452464e-06,
+      "loss": 0.4407,
+      "step": 4694
+    },
+    {
+      "epoch": 2.329154772247735,
+      "grad_norm": 0.07141269813437955,
+      "learning_rate": 3.7298364068634764e-06,
+      "loss": 0.4538,
+      "step": 4695
+    },
+    {
+      "epoch": 2.3296512349509744,
+      "grad_norm": 0.07038667356360034,
+      "learning_rate": 3.727945927398628e-06,
+      "loss": 0.4754,
+      "step": 4696
+    },
+    {
+      "epoch": 2.3301476976542137,
+      "grad_norm": 0.07429587927666312,
+      "learning_rate": 3.7260556423468486e-06,
+      "loss": 0.4659,
+      "step": 4697
+    },
+    {
+      "epoch": 2.330644160357453,
+      "grad_norm": 0.07302518698343577,
+      "learning_rate": 3.72416555199704e-06,
+      "loss": 0.4832,
+      "step": 4698
+    },
+    {
+      "epoch": 2.3311406230606924,
+      "grad_norm": 0.07168300385035588,
+      "learning_rate": 3.722275656638068e-06,
+      "loss": 0.483,
+      "step": 4699
+    },
+    {
+      "epoch": 2.331637085763932,
+      "grad_norm": 0.07072808980615392,
+      "learning_rate": 3.7203859565587765e-06,
+      "loss": 0.438,
+      "step": 4700
+    },
+    {
+      "epoch": 2.3321335484671715,
+      "grad_norm": 0.0735926019372795,
+      "learning_rate": 3.7184964520479737e-06,
+      "loss": 0.4593,
+      "step": 4701
+    },
+    {
+      "epoch": 2.332630011170411,
+      "grad_norm": 0.07033503029697981,
+      "learning_rate": 3.7166071433944407e-06,
+      "loss": 0.4483,
+      "step": 4702
+    },
+    {
+      "epoch": 2.33312647387365,
+      "grad_norm": 0.07182229172907455,
+      "learning_rate": 3.7147180308869296e-06,
+      "loss": 0.4705,
+      "step": 4703
+    },
+    {
+      "epoch": 2.33362293657689,
+      "grad_norm": 0.07143612225361773,
+      "learning_rate": 3.712829114814158e-06,
+      "loss": 0.4254,
+      "step": 4704
+    },
+    {
+      "epoch": 2.3341193992801292,
+      "grad_norm": 0.07400848281236902,
+      "learning_rate": 3.7109403954648208e-06,
+      "loss": 0.4588,
+      "step": 4705
+    },
+    {
+      "epoch": 2.3346158619833686,
+      "grad_norm": 0.0727514120789118,
+      "learning_rate": 3.7090518731275738e-06,
+      "loss": 0.4498,
+      "step": 4706
+    },
+    {
+      "epoch": 2.335112324686608,
+      "grad_norm": 0.07610809200395814,
+      "learning_rate": 3.7071635480910486e-06,
+      "loss": 0.4679,
+      "step": 4707
+    },
+    {
+      "epoch": 2.3356087873898472,
+      "grad_norm": 0.07723757088222308,
+      "learning_rate": 3.7052754206438455e-06,
+      "loss": 0.4844,
+      "step": 4708
+    },
+    {
+      "epoch": 2.3361052500930866,
+      "grad_norm": 0.07401073014952221,
+      "learning_rate": 3.703387491074536e-06,
+      "loss": 0.4456,
+      "step": 4709
+    },
+    {
+      "epoch": 2.3366017127963263,
+      "grad_norm": 0.07439980480315013,
+      "learning_rate": 3.7014997596716596e-06,
+      "loss": 0.4761,
+      "step": 4710
+    },
+    {
+      "epoch": 2.3370981754995657,
+      "grad_norm": 0.07226783163233316,
+      "learning_rate": 3.699612226723724e-06,
+      "loss": 0.4856,
+      "step": 4711
+    },
+    {
+      "epoch": 2.337594638202805,
+      "grad_norm": 0.07470804406103008,
+      "learning_rate": 3.697724892519209e-06,
+      "loss": 0.4688,
+      "step": 4712
+    },
+    {
+      "epoch": 2.3380911009060443,
+      "grad_norm": 0.07314284196363455,
+      "learning_rate": 3.6958377573465643e-06,
+      "loss": 0.4721,
+      "step": 4713
+    },
+    {
+      "epoch": 2.3385875636092837,
+      "grad_norm": 0.07246523023838439,
+      "learning_rate": 3.693950821494209e-06,
+      "loss": 0.4408,
+      "step": 4714
+    },
+    {
+      "epoch": 2.3390840263125234,
+      "grad_norm": 0.07405405668151005,
+      "learning_rate": 3.692064085250528e-06,
+      "loss": 0.4592,
+      "step": 4715
+    },
+    {
+      "epoch": 2.3395804890157628,
+      "grad_norm": 0.07165590250964264,
+      "learning_rate": 3.6901775489038804e-06,
+      "loss": 0.4557,
+      "step": 4716
+    },
+    {
+      "epoch": 2.340076951719002,
+      "grad_norm": 0.07382760497377183,
+      "learning_rate": 3.688291212742594e-06,
+      "loss": 0.4756,
+      "step": 4717
+    },
+    {
+      "epoch": 2.3405734144222414,
+      "grad_norm": 0.07176344836460424,
+      "learning_rate": 3.686405077054963e-06,
+      "loss": 0.4409,
+      "step": 4718
+    },
+    {
+      "epoch": 2.3410698771254808,
+      "grad_norm": 0.07038888718316055,
+      "learning_rate": 3.6845191421292558e-06,
+      "loss": 0.448,
+      "step": 4719
+    },
+    {
+      "epoch": 2.3415663398287205,
+      "grad_norm": 0.07348014482842789,
+      "learning_rate": 3.682633408253704e-06,
+      "loss": 0.4652,
+      "step": 4720
+    },
+    {
+      "epoch": 2.34206280253196,
+      "grad_norm": 0.07286941695899553,
+      "learning_rate": 3.680747875716514e-06,
+      "loss": 0.4535,
+      "step": 4721
+    },
+    {
+      "epoch": 2.342559265235199,
+      "grad_norm": 0.07332682139319671,
+      "learning_rate": 3.6788625448058605e-06,
+      "loss": 0.4762,
+      "step": 4722
+    },
+    {
+      "epoch": 2.3430557279384385,
+      "grad_norm": 0.07381530635235646,
+      "learning_rate": 3.676977415809882e-06,
+      "loss": 0.4707,
+      "step": 4723
+    },
+    {
+      "epoch": 2.343552190641678,
+      "grad_norm": 0.06990799813327832,
+      "learning_rate": 3.675092489016693e-06,
+      "loss": 0.4018,
+      "step": 4724
+    },
+    {
+      "epoch": 2.3440486533449176,
+      "grad_norm": 0.07394454106338585,
+      "learning_rate": 3.673207764714373e-06,
+      "loss": 0.4408,
+      "step": 4725
+    },
+    {
+      "epoch": 2.344545116048157,
+      "grad_norm": 0.07301506376321407,
+      "learning_rate": 3.671323243190974e-06,
+      "loss": 0.4544,
+      "step": 4726
+    },
+    {
+      "epoch": 2.3450415787513963,
+      "grad_norm": 0.07588801616599444,
+      "learning_rate": 3.669438924734512e-06,
+      "loss": 0.4637,
+      "step": 4727
+    },
+    {
+      "epoch": 2.3455380414546356,
+      "grad_norm": 0.07434133158405588,
+      "learning_rate": 3.6675548096329773e-06,
+      "loss": 0.4667,
+      "step": 4728
+    },
+    {
+      "epoch": 2.346034504157875,
+      "grad_norm": 0.07567857678764008,
+      "learning_rate": 3.665670898174325e-06,
+      "loss": 0.4717,
+      "step": 4729
+    },
+    {
+      "epoch": 2.3465309668611147,
+      "grad_norm": 0.07114717020994527,
+      "learning_rate": 3.6637871906464834e-06,
+      "loss": 0.4274,
+      "step": 4730
+    },
+    {
+      "epoch": 2.347027429564354,
+      "grad_norm": 0.07074817565632668,
+      "learning_rate": 3.6619036873373435e-06,
+      "loss": 0.4247,
+      "step": 4731
+    },
+    {
+      "epoch": 2.3475238922675934,
+      "grad_norm": 0.0722416760901901,
+      "learning_rate": 3.660020388534771e-06,
+      "loss": 0.448,
+      "step": 4732
+    },
+    {
+      "epoch": 2.3480203549708327,
+      "grad_norm": 0.07185437341829799,
+      "learning_rate": 3.658137294526596e-06,
+      "loss": 0.4598,
+      "step": 4733
+    },
+    {
+      "epoch": 2.348516817674072,
+      "grad_norm": 0.07141278433821945,
+      "learning_rate": 3.656254405600621e-06,
+      "loss": 0.4486,
+      "step": 4734
+    },
+    {
+      "epoch": 2.349013280377312,
+      "grad_norm": 0.07133360394274396,
+      "learning_rate": 3.654371722044616e-06,
+      "loss": 0.4357,
+      "step": 4735
+    },
+    {
+      "epoch": 2.349509743080551,
+      "grad_norm": 0.07194937115926209,
+      "learning_rate": 3.6524892441463166e-06,
+      "loss": 0.4415,
+      "step": 4736
+    },
+    {
+      "epoch": 2.3500062057837905,
+      "grad_norm": 0.07480819749333578,
+      "learning_rate": 3.650606972193431e-06,
+      "loss": 0.5018,
+      "step": 4737
+    },
+    {
+      "epoch": 2.35050266848703,
+      "grad_norm": 0.07296178767261839,
+      "learning_rate": 3.6487249064736352e-06,
+      "loss": 0.4458,
+      "step": 4738
+    },
+    {
+      "epoch": 2.350999131190269,
+      "grad_norm": 0.07008863773692527,
+      "learning_rate": 3.64684304727457e-06,
+      "loss": 0.4185,
+      "step": 4739
+    },
+    {
+      "epoch": 2.351495593893509,
+      "grad_norm": 0.07221727150834073,
+      "learning_rate": 3.644961394883848e-06,
+      "loss": 0.4826,
+      "step": 4740
+    },
+    {
+      "epoch": 2.3519920565967483,
+      "grad_norm": 0.07125064634096756,
+      "learning_rate": 3.643079949589051e-06,
+      "loss": 0.4761,
+      "step": 4741
+    },
+    {
+      "epoch": 2.3524885192999876,
+      "grad_norm": 0.0740146726475471,
+      "learning_rate": 3.641198711677728e-06,
+      "loss": 0.4661,
+      "step": 4742
+    },
+    {
+      "epoch": 2.352984982003227,
+      "grad_norm": 0.07190469581219823,
+      "learning_rate": 3.6393176814373944e-06,
+      "loss": 0.485,
+      "step": 4743
+    },
+    {
+      "epoch": 2.3534814447064663,
+      "grad_norm": 0.06876822776109015,
+      "learning_rate": 3.6374368591555352e-06,
+      "loss": 0.4461,
+      "step": 4744
+    },
+    {
+      "epoch": 2.353977907409706,
+      "grad_norm": 0.07565380640590723,
+      "learning_rate": 3.6355562451196065e-06,
+      "loss": 0.4843,
+      "step": 4745
+    },
+    {
+      "epoch": 2.3544743701129454,
+      "grad_norm": 0.07404096345374254,
+      "learning_rate": 3.633675839617028e-06,
+      "loss": 0.4652,
+      "step": 4746
+    },
+    {
+      "epoch": 2.3549708328161847,
+      "grad_norm": 0.07180746556237937,
+      "learning_rate": 3.6317956429351906e-06,
+      "loss": 0.4339,
+      "step": 4747
+    },
+    {
+      "epoch": 2.355467295519424,
+      "grad_norm": 0.0738081781163325,
+      "learning_rate": 3.6299156553614513e-06,
+      "loss": 0.444,
+      "step": 4748
+    },
+    {
+      "epoch": 2.3559637582226634,
+      "grad_norm": 0.07143636633873711,
+      "learning_rate": 3.628035877183136e-06,
+      "loss": 0.4452,
+      "step": 4749
+    },
+    {
+      "epoch": 2.356460220925903,
+      "grad_norm": 0.0750219048239024,
+      "learning_rate": 3.6261563086875396e-06,
+      "loss": 0.4937,
+      "step": 4750
+    },
+    {
+      "epoch": 2.3569566836291425,
+      "grad_norm": 0.07145869095204205,
+      "learning_rate": 3.6242769501619245e-06,
+      "loss": 0.4502,
+      "step": 4751
+    },
+    {
+      "epoch": 2.357453146332382,
+      "grad_norm": 0.07262263133026953,
+      "learning_rate": 3.622397801893518e-06,
+      "loss": 0.4621,
+      "step": 4752
+    },
+    {
+      "epoch": 2.357949609035621,
+      "grad_norm": 0.07474341812925704,
+      "learning_rate": 3.62051886416952e-06,
+      "loss": 0.4569,
+      "step": 4753
+    },
+    {
+      "epoch": 2.3584460717388605,
+      "grad_norm": 0.07221546383431686,
+      "learning_rate": 3.618640137277097e-06,
+      "loss": 0.4749,
+      "step": 4754
+    },
+    {
+      "epoch": 2.3589425344421002,
+      "grad_norm": 0.07333306403627074,
+      "learning_rate": 3.6167616215033784e-06,
+      "loss": 0.4638,
+      "step": 4755
+    },
+    {
+      "epoch": 2.3594389971453396,
+      "grad_norm": 0.07300394997026594,
+      "learning_rate": 3.614883317135467e-06,
+      "loss": 0.4449,
+      "step": 4756
+    },
+    {
+      "epoch": 2.359935459848579,
+      "grad_norm": 0.07595971084123176,
+      "learning_rate": 3.613005224460433e-06,
+      "loss": 0.4845,
+      "step": 4757
+    },
+    {
+      "epoch": 2.360431922551818,
+      "grad_norm": 0.0744674579731717,
+      "learning_rate": 3.6111273437653114e-06,
+      "loss": 0.4364,
+      "step": 4758
+    },
+    {
+      "epoch": 2.3609283852550575,
+      "grad_norm": 0.07356470176802757,
+      "learning_rate": 3.6092496753371064e-06,
+      "loss": 0.4598,
+      "step": 4759
+    },
+    {
+      "epoch": 2.3614248479582973,
+      "grad_norm": 0.07331618173536777,
+      "learning_rate": 3.6073722194627893e-06,
+      "loss": 0.4691,
+      "step": 4760
+    },
+    {
+      "epoch": 2.3619213106615367,
+      "grad_norm": 0.07269515877497877,
+      "learning_rate": 3.6054949764292996e-06,
+      "loss": 0.4351,
+      "step": 4761
+    },
+    {
+      "epoch": 2.362417773364776,
+      "grad_norm": 0.07255848749377546,
+      "learning_rate": 3.6036179465235432e-06,
+      "loss": 0.4852,
+      "step": 4762
+    },
+    {
+      "epoch": 2.3629142360680153,
+      "grad_norm": 0.07436795161560912,
+      "learning_rate": 3.6017411300323957e-06,
+      "loss": 0.4641,
+      "step": 4763
+    },
+    {
+      "epoch": 2.3634106987712546,
+      "grad_norm": 0.07223596806015442,
+      "learning_rate": 3.599864527242696e-06,
+      "loss": 0.4643,
+      "step": 4764
+    },
+    {
+      "epoch": 2.3639071614744944,
+      "grad_norm": 0.07363739183083648,
+      "learning_rate": 3.5979881384412534e-06,
+      "loss": 0.4593,
+      "step": 4765
+    },
+    {
+      "epoch": 2.3644036241777338,
+      "grad_norm": 0.07133242037314486,
+      "learning_rate": 3.5961119639148443e-06,
+      "loss": 0.457,
+      "step": 4766
+    },
+    {
+      "epoch": 2.364900086880973,
+      "grad_norm": 0.07048145352143227,
+      "learning_rate": 3.5942360039502135e-06,
+      "loss": 0.4448,
+      "step": 4767
+    },
+    {
+      "epoch": 2.3653965495842124,
+      "grad_norm": 0.07333678955240325,
+      "learning_rate": 3.592360258834069e-06,
+      "loss": 0.4491,
+      "step": 4768
+    },
+    {
+      "epoch": 2.3658930122874517,
+      "grad_norm": 0.07318665298320112,
+      "learning_rate": 3.5904847288530882e-06,
+      "loss": 0.4669,
+      "step": 4769
+    },
+    {
+      "epoch": 2.366389474990691,
+      "grad_norm": 0.07357706600034569,
+      "learning_rate": 3.5886094142939195e-06,
+      "loss": 0.479,
+      "step": 4770
+    },
+    {
+      "epoch": 2.366885937693931,
+      "grad_norm": 0.07131078314109707,
+      "learning_rate": 3.5867343154431693e-06,
+      "loss": 0.4538,
+      "step": 4771
+    },
+    {
+      "epoch": 2.36738240039717,
+      "grad_norm": 0.07233993020557973,
+      "learning_rate": 3.584859432587419e-06,
+      "loss": 0.4595,
+      "step": 4772
+    },
+    {
+      "epoch": 2.3678788631004095,
+      "grad_norm": 0.07227522944783686,
+      "learning_rate": 3.582984766013215e-06,
+      "loss": 0.4681,
+      "step": 4773
+    },
+    {
+      "epoch": 2.368375325803649,
+      "grad_norm": 0.07378966999244926,
+      "learning_rate": 3.58111031600707e-06,
+      "loss": 0.4467,
+      "step": 4774
+    },
+    {
+      "epoch": 2.3688717885068886,
+      "grad_norm": 0.07255376065435246,
+      "learning_rate": 3.5792360828554615e-06,
+      "loss": 0.4649,
+      "step": 4775
+    },
+    {
+      "epoch": 2.369368251210128,
+      "grad_norm": 0.07349043423441821,
+      "learning_rate": 3.5773620668448384e-06,
+      "loss": 0.4427,
+      "step": 4776
+    },
+    {
+      "epoch": 2.3698647139133673,
+      "grad_norm": 0.0737917842996249,
+      "learning_rate": 3.575488268261613e-06,
+      "loss": 0.4484,
+      "step": 4777
+    },
+    {
+      "epoch": 2.3703611766166066,
+      "grad_norm": 0.07131005263256235,
+      "learning_rate": 3.5736146873921652e-06,
+      "loss": 0.4636,
+      "step": 4778
+    },
+    {
+      "epoch": 2.370857639319846,
+      "grad_norm": 0.07285168554855122,
+      "learning_rate": 3.5717413245228434e-06,
+      "loss": 0.4796,
+      "step": 4779
+    },
+    {
+      "epoch": 2.3713541020230853,
+      "grad_norm": 0.07242742265876065,
+      "learning_rate": 3.569868179939958e-06,
+      "loss": 0.4462,
+      "step": 4780
+    },
+    {
+      "epoch": 2.371850564726325,
+      "grad_norm": 0.07264450598228168,
+      "learning_rate": 3.567995253929792e-06,
+      "loss": 0.4966,
+      "step": 4781
+    },
+    {
+      "epoch": 2.3723470274295644,
+      "grad_norm": 0.07568634838040776,
+      "learning_rate": 3.56612254677859e-06,
+      "loss": 0.4577,
+      "step": 4782
+    },
+    {
+      "epoch": 2.3728434901328037,
+      "grad_norm": 0.07437246189255993,
+      "learning_rate": 3.564250058772567e-06,
+      "loss": 0.5073,
+      "step": 4783
+    },
+    {
+      "epoch": 2.373339952836043,
+      "grad_norm": 0.07202710745369477,
+      "learning_rate": 3.562377790197903e-06,
+      "loss": 0.4652,
+      "step": 4784
+    },
+    {
+      "epoch": 2.373836415539283,
+      "grad_norm": 0.073800591321293,
+      "learning_rate": 3.560505741340742e-06,
+      "loss": 0.441,
+      "step": 4785
+    },
+    {
+      "epoch": 2.374332878242522,
+      "grad_norm": 0.0712780849538718,
+      "learning_rate": 3.5586339124871993e-06,
+      "loss": 0.447,
+      "step": 4786
+    },
+    {
+      "epoch": 2.3748293409457615,
+      "grad_norm": 0.07491310671013832,
+      "learning_rate": 3.556762303923351e-06,
+      "loss": 0.4816,
+      "step": 4787
+    },
+    {
+      "epoch": 2.375325803649001,
+      "grad_norm": 0.07467968796444399,
+      "learning_rate": 3.554890915935244e-06,
+      "loss": 0.4778,
+      "step": 4788
+    },
+    {
+      "epoch": 2.37582226635224,
+      "grad_norm": 0.07237031006567743,
+      "learning_rate": 3.5530197488088904e-06,
+      "loss": 0.4571,
+      "step": 4789
+    },
+    {
+      "epoch": 2.3763187290554795,
+      "grad_norm": 0.07262263860714888,
+      "learning_rate": 3.5511488028302676e-06,
+      "loss": 0.4385,
+      "step": 4790
+    },
+    {
+      "epoch": 2.3768151917587192,
+      "grad_norm": 0.07185368446839807,
+      "learning_rate": 3.5492780782853196e-06,
+      "loss": 0.4311,
+      "step": 4791
+    },
+    {
+      "epoch": 2.3773116544619586,
+      "grad_norm": 0.07324880246011775,
+      "learning_rate": 3.547407575459957e-06,
+      "loss": 0.4472,
+      "step": 4792
+    },
+    {
+      "epoch": 2.377808117165198,
+      "grad_norm": 0.07446233139953379,
+      "learning_rate": 3.545537294640055e-06,
+      "loss": 0.5045,
+      "step": 4793
+    },
+    {
+      "epoch": 2.3783045798684372,
+      "grad_norm": 0.070140983659222,
+      "learning_rate": 3.543667236111458e-06,
+      "loss": 0.4607,
+      "step": 4794
+    },
+    {
+      "epoch": 2.378801042571677,
+      "grad_norm": 0.07250620197395204,
+      "learning_rate": 3.541797400159973e-06,
+      "loss": 0.4479,
+      "step": 4795
+    },
+    {
+      "epoch": 2.3792975052749163,
+      "grad_norm": 0.07540106712813605,
+      "learning_rate": 3.539927787071375e-06,
+      "loss": 0.5026,
+      "step": 4796
+    },
+    {
+      "epoch": 2.3797939679781557,
+      "grad_norm": 0.07498674621009987,
+      "learning_rate": 3.5380583971314043e-06,
+      "loss": 0.457,
+      "step": 4797
+    },
+    {
+      "epoch": 2.380290430681395,
+      "grad_norm": 0.07174711237052699,
+      "learning_rate": 3.5361892306257666e-06,
+      "loss": 0.4536,
+      "step": 4798
+    },
+    {
+      "epoch": 2.3807868933846343,
+      "grad_norm": 0.0716157727005352,
+      "learning_rate": 3.534320287840135e-06,
+      "loss": 0.4606,
+      "step": 4799
+    },
+    {
+      "epoch": 2.3812833560878737,
+      "grad_norm": 0.07398435347159586,
+      "learning_rate": 3.532451569060148e-06,
+      "loss": 0.4672,
+      "step": 4800
+    },
+    {
+      "epoch": 2.3817798187911134,
+      "grad_norm": 0.07071056978522963,
+      "learning_rate": 3.530583074571407e-06,
+      "loss": 0.4429,
+      "step": 4801
+    },
+    {
+      "epoch": 2.3822762814943528,
+      "grad_norm": 0.07240856332363192,
+      "learning_rate": 3.5287148046594847e-06,
+      "loss": 0.4532,
+      "step": 4802
+    },
+    {
+      "epoch": 2.382772744197592,
+      "grad_norm": 0.0719354437460463,
+      "learning_rate": 3.5268467596099126e-06,
+      "loss": 0.4594,
+      "step": 4803
+    },
+    {
+      "epoch": 2.3832692069008314,
+      "grad_norm": 0.07040324938389426,
+      "learning_rate": 3.524978939708193e-06,
+      "loss": 0.439,
+      "step": 4804
+    },
+    {
+      "epoch": 2.383765669604071,
+      "grad_norm": 0.07230438616729792,
+      "learning_rate": 3.5231113452397927e-06,
+      "loss": 0.4264,
+      "step": 4805
+    },
+    {
+      "epoch": 2.3842621323073105,
+      "grad_norm": 0.07399897104250126,
+      "learning_rate": 3.5212439764901425e-06,
+      "loss": 0.4619,
+      "step": 4806
+    },
+    {
+      "epoch": 2.38475859501055,
+      "grad_norm": 0.07318270307419919,
+      "learning_rate": 3.5193768337446418e-06,
+      "loss": 0.4589,
+      "step": 4807
+    },
+    {
+      "epoch": 2.385255057713789,
+      "grad_norm": 0.07315629477062484,
+      "learning_rate": 3.5175099172886507e-06,
+      "loss": 0.4932,
+      "step": 4808
+    },
+    {
+      "epoch": 2.3857515204170285,
+      "grad_norm": 0.07225843453148965,
+      "learning_rate": 3.515643227407499e-06,
+      "loss": 0.456,
+      "step": 4809
+    },
+    {
+      "epoch": 2.386247983120268,
+      "grad_norm": 0.07421180142061967,
+      "learning_rate": 3.51377676438648e-06,
+      "loss": 0.4508,
+      "step": 4810
+    },
+    {
+      "epoch": 2.3867444458235076,
+      "grad_norm": 0.07108575490410428,
+      "learning_rate": 3.511910528510854e-06,
+      "loss": 0.4356,
+      "step": 4811
+    },
+    {
+      "epoch": 2.387240908526747,
+      "grad_norm": 0.07724965177245846,
+      "learning_rate": 3.510044520065843e-06,
+      "loss": 0.4847,
+      "step": 4812
+    },
+    {
+      "epoch": 2.3877373712299863,
+      "grad_norm": 0.07282260233134959,
+      "learning_rate": 3.5081787393366374e-06,
+      "loss": 0.4386,
+      "step": 4813
+    },
+    {
+      "epoch": 2.3882338339332256,
+      "grad_norm": 0.07152279900165055,
+      "learning_rate": 3.506313186608392e-06,
+      "loss": 0.4293,
+      "step": 4814
+    },
+    {
+      "epoch": 2.3887302966364654,
+      "grad_norm": 0.0746752837500685,
+      "learning_rate": 3.504447862166227e-06,
+      "loss": 0.4697,
+      "step": 4815
+    },
+    {
+      "epoch": 2.3892267593397047,
+      "grad_norm": 0.07436156575975729,
+      "learning_rate": 3.502582766295227e-06,
+      "loss": 0.4888,
+      "step": 4816
+    },
+    {
+      "epoch": 2.389723222042944,
+      "grad_norm": 0.07478976244885759,
+      "learning_rate": 3.500717899280442e-06,
+      "loss": 0.4683,
+      "step": 4817
+    },
+    {
+      "epoch": 2.3902196847461834,
+      "grad_norm": 0.07037325001387294,
+      "learning_rate": 3.498853261406888e-06,
+      "loss": 0.4378,
+      "step": 4818
+    },
+    {
+      "epoch": 2.3907161474494227,
+      "grad_norm": 0.07631373189207717,
+      "learning_rate": 3.4969888529595426e-06,
+      "loss": 0.479,
+      "step": 4819
+    },
+    {
+      "epoch": 2.391212610152662,
+      "grad_norm": 0.07426025126427815,
+      "learning_rate": 3.4951246742233517e-06,
+      "loss": 0.4594,
+      "step": 4820
+    },
+    {
+      "epoch": 2.391709072855902,
+      "grad_norm": 0.0727151200637013,
+      "learning_rate": 3.4932607254832257e-06,
+      "loss": 0.4694,
+      "step": 4821
+    },
+    {
+      "epoch": 2.392205535559141,
+      "grad_norm": 0.07184130765612347,
+      "learning_rate": 3.4913970070240388e-06,
+      "loss": 0.4627,
+      "step": 4822
+    },
+    {
+      "epoch": 2.3927019982623805,
+      "grad_norm": 0.07136506615380843,
+      "learning_rate": 3.4895335191306323e-06,
+      "loss": 0.4406,
+      "step": 4823
+    },
+    {
+      "epoch": 2.39319846096562,
+      "grad_norm": 0.0831197941937164,
+      "learning_rate": 3.4876702620878072e-06,
+      "loss": 0.5137,
+      "step": 4824
+    },
+    {
+      "epoch": 2.3936949236688596,
+      "grad_norm": 0.07098730544237608,
+      "learning_rate": 3.4858072361803347e-06,
+      "loss": 0.4703,
+      "step": 4825
+    },
+    {
+      "epoch": 2.394191386372099,
+      "grad_norm": 0.07435224540563783,
+      "learning_rate": 3.483944441692948e-06,
+      "loss": 0.432,
+      "step": 4826
+    },
+    {
+      "epoch": 2.3946878490753383,
+      "grad_norm": 0.07156278419670799,
+      "learning_rate": 3.482081878910346e-06,
+      "loss": 0.4451,
+      "step": 4827
+    },
+    {
+      "epoch": 2.3951843117785776,
+      "grad_norm": 0.07422050116299996,
+      "learning_rate": 3.4802195481171895e-06,
+      "loss": 0.4613,
+      "step": 4828
+    },
+    {
+      "epoch": 2.395680774481817,
+      "grad_norm": 0.07199435924853591,
+      "learning_rate": 3.4783574495981075e-06,
+      "loss": 0.4442,
+      "step": 4829
+    },
+    {
+      "epoch": 2.3961772371850563,
+      "grad_norm": 0.07585944049032202,
+      "learning_rate": 3.4764955836376924e-06,
+      "loss": 0.4979,
+      "step": 4830
+    },
+    {
+      "epoch": 2.396673699888296,
+      "grad_norm": 0.07129212048060686,
+      "learning_rate": 3.4746339505204986e-06,
+      "loss": 0.4557,
+      "step": 4831
+    },
+    {
+      "epoch": 2.3971701625915354,
+      "grad_norm": 0.072693573519986,
+      "learning_rate": 3.4727725505310496e-06,
+      "loss": 0.4352,
+      "step": 4832
+    },
+    {
+      "epoch": 2.3976666252947747,
+      "grad_norm": 0.07448848764186855,
+      "learning_rate": 3.470911383953828e-06,
+      "loss": 0.44,
+      "step": 4833
+    },
+    {
+      "epoch": 2.398163087998014,
+      "grad_norm": 0.07497358791166973,
+      "learning_rate": 3.469050451073287e-06,
+      "loss": 0.4587,
+      "step": 4834
+    },
+    {
+      "epoch": 2.398659550701254,
+      "grad_norm": 0.07184294064328557,
+      "learning_rate": 3.467189752173835e-06,
+      "loss": 0.453,
+      "step": 4835
+    },
+    {
+      "epoch": 2.399156013404493,
+      "grad_norm": 0.07433389489274306,
+      "learning_rate": 3.4653292875398523e-06,
+      "loss": 0.4821,
+      "step": 4836
+    },
+    {
+      "epoch": 2.3996524761077325,
+      "grad_norm": 0.07068190313163356,
+      "learning_rate": 3.4634690574556815e-06,
+      "loss": 0.4549,
+      "step": 4837
+    },
+    {
+      "epoch": 2.400148938810972,
+      "grad_norm": 0.07465144485609781,
+      "learning_rate": 3.4616090622056296e-06,
+      "loss": 0.48,
+      "step": 4838
+    },
+    {
+      "epoch": 2.400645401514211,
+      "grad_norm": 0.07169302316248427,
+      "learning_rate": 3.459749302073967e-06,
+      "loss": 0.4451,
+      "step": 4839
+    },
+    {
+      "epoch": 2.4011418642174505,
+      "grad_norm": 0.07085649592361846,
+      "learning_rate": 3.457889777344926e-06,
+      "loss": 0.4607,
+      "step": 4840
+    },
+    {
+      "epoch": 2.4016383269206902,
+      "grad_norm": 0.07329969355902644,
+      "learning_rate": 3.4560304883027072e-06,
+      "loss": 0.4545,
+      "step": 4841
+    },
+    {
+      "epoch": 2.4021347896239296,
+      "grad_norm": 0.07310247900538014,
+      "learning_rate": 3.4541714352314726e-06,
+      "loss": 0.4726,
+      "step": 4842
+    },
+    {
+      "epoch": 2.402631252327169,
+      "grad_norm": 0.07103939876575813,
+      "learning_rate": 3.4523126184153483e-06,
+      "loss": 0.4501,
+      "step": 4843
+    },
+    {
+      "epoch": 2.4031277150304082,
+      "grad_norm": 0.08263769021042183,
+      "learning_rate": 3.4504540381384265e-06,
+      "loss": 0.4879,
+      "step": 4844
+    },
+    {
+      "epoch": 2.403624177733648,
+      "grad_norm": 0.07030234205937556,
+      "learning_rate": 3.448595694684758e-06,
+      "loss": 0.4631,
+      "step": 4845
+    },
+    {
+      "epoch": 2.4041206404368873,
+      "grad_norm": 0.06845809783951995,
+      "learning_rate": 3.4467375883383638e-06,
+      "loss": 0.4174,
+      "step": 4846
+    },
+    {
+      "epoch": 2.4046171031401267,
+      "grad_norm": 0.07368759949851977,
+      "learning_rate": 3.444879719383224e-06,
+      "loss": 0.4655,
+      "step": 4847
+    },
+    {
+      "epoch": 2.405113565843366,
+      "grad_norm": 0.06867761594775768,
+      "learning_rate": 3.4430220881032855e-06,
+      "loss": 0.4207,
+      "step": 4848
+    },
+    {
+      "epoch": 2.4056100285466053,
+      "grad_norm": 0.07176053293602179,
+      "learning_rate": 3.441164694782456e-06,
+      "loss": 0.4497,
+      "step": 4849
+    },
+    {
+      "epoch": 2.4061064912498447,
+      "grad_norm": 0.07308211559699186,
+      "learning_rate": 3.4393075397046105e-06,
+      "loss": 0.4453,
+      "step": 4850
+    },
+    {
+      "epoch": 2.4066029539530844,
+      "grad_norm": 0.0729020237538028,
+      "learning_rate": 3.437450623153582e-06,
+      "loss": 0.461,
+      "step": 4851
+    },
+    {
+      "epoch": 2.4070994166563238,
+      "grad_norm": 0.06998983752353476,
+      "learning_rate": 3.4355939454131722e-06,
+      "loss": 0.4848,
+      "step": 4852
+    },
+    {
+      "epoch": 2.407595879359563,
+      "grad_norm": 0.07186086298195375,
+      "learning_rate": 3.433737506767144e-06,
+      "loss": 0.4565,
+      "step": 4853
+    },
+    {
+      "epoch": 2.4080923420628024,
+      "grad_norm": 0.07480359801968427,
+      "learning_rate": 3.4318813074992253e-06,
+      "loss": 0.453,
+      "step": 4854
+    },
+    {
+      "epoch": 2.4085888047660418,
+      "grad_norm": 0.07144531652703363,
+      "learning_rate": 3.430025347893107e-06,
+      "loss": 0.4567,
+      "step": 4855
+    },
+    {
+      "epoch": 2.4090852674692815,
+      "grad_norm": 0.07124218226500924,
+      "learning_rate": 3.4281696282324402e-06,
+      "loss": 0.4288,
+      "step": 4856
+    },
+    {
+      "epoch": 2.409581730172521,
+      "grad_norm": 0.07058037892685524,
+      "learning_rate": 3.426314148800843e-06,
+      "loss": 0.4511,
+      "step": 4857
+    },
+    {
+      "epoch": 2.41007819287576,
+      "grad_norm": 0.06929868897535611,
+      "learning_rate": 3.424458909881897e-06,
+      "loss": 0.4511,
+      "step": 4858
+    },
+    {
+      "epoch": 2.4105746555789995,
+      "grad_norm": 0.07525160754840045,
+      "learning_rate": 3.4226039117591443e-06,
+      "loss": 0.5151,
+      "step": 4859
+    },
+    {
+      "epoch": 2.411071118282239,
+      "grad_norm": 0.07513115447244399,
+      "learning_rate": 3.420749154716093e-06,
+      "loss": 0.4584,
+      "step": 4860
+    },
+    {
+      "epoch": 2.4115675809854786,
+      "grad_norm": 0.07269003207123895,
+      "learning_rate": 3.418894639036211e-06,
+      "loss": 0.4729,
+      "step": 4861
+    },
+    {
+      "epoch": 2.412064043688718,
+      "grad_norm": 0.07404592017298095,
+      "learning_rate": 3.4170403650029327e-06,
+      "loss": 0.4696,
+      "step": 4862
+    },
+    {
+      "epoch": 2.4125605063919573,
+      "grad_norm": 0.07014295136455598,
+      "learning_rate": 3.415186332899653e-06,
+      "loss": 0.4162,
+      "step": 4863
+    },
+    {
+      "epoch": 2.4130569690951966,
+      "grad_norm": 0.07251567829613642,
+      "learning_rate": 3.4133325430097337e-06,
+      "loss": 0.482,
+      "step": 4864
+    },
+    {
+      "epoch": 2.413553431798436,
+      "grad_norm": 0.07182370674066571,
+      "learning_rate": 3.411478995616493e-06,
+      "loss": 0.4331,
+      "step": 4865
+    },
+    {
+      "epoch": 2.4140498945016757,
+      "grad_norm": 0.07201533809263864,
+      "learning_rate": 3.409625691003221e-06,
+      "loss": 0.4801,
+      "step": 4866
+    },
+    {
+      "epoch": 2.414546357204915,
+      "grad_norm": 0.07336794806212467,
+      "learning_rate": 3.407772629453159e-06,
+      "loss": 0.4623,
+      "step": 4867
+    },
+    {
+      "epoch": 2.4150428199081544,
+      "grad_norm": 0.07361076825262522,
+      "learning_rate": 3.405919811249522e-06,
+      "loss": 0.4716,
+      "step": 4868
+    },
+    {
+      "epoch": 2.4155392826113937,
+      "grad_norm": 0.07468347922530201,
+      "learning_rate": 3.404067236675483e-06,
+      "loss": 0.4764,
+      "step": 4869
+    },
+    {
+      "epoch": 2.416035745314633,
+      "grad_norm": 0.07232813832383798,
+      "learning_rate": 3.4022149060141775e-06,
+      "loss": 0.4794,
+      "step": 4870
+    },
+    {
+      "epoch": 2.416532208017873,
+      "grad_norm": 0.07258532390558077,
+      "learning_rate": 3.400362819548706e-06,
+      "loss": 0.4761,
+      "step": 4871
+    },
+    {
+      "epoch": 2.417028670721112,
+      "grad_norm": 0.07485777438080334,
+      "learning_rate": 3.3985109775621284e-06,
+      "loss": 0.5034,
+      "step": 4872
+    },
+    {
+      "epoch": 2.4175251334243515,
+      "grad_norm": 0.07137979655942545,
+      "learning_rate": 3.3966593803374703e-06,
+      "loss": 0.4739,
+      "step": 4873
+    },
+    {
+      "epoch": 2.418021596127591,
+      "grad_norm": 0.06859686509924706,
+      "learning_rate": 3.394808028157718e-06,
+      "loss": 0.4281,
+      "step": 4874
+    },
+    {
+      "epoch": 2.41851805883083,
+      "grad_norm": 0.07187345846385193,
+      "learning_rate": 3.392956921305821e-06,
+      "loss": 0.4418,
+      "step": 4875
+    },
+    {
+      "epoch": 2.41901452153407,
+      "grad_norm": 0.07594593901365239,
+      "learning_rate": 3.3911060600646934e-06,
+      "loss": 0.4575,
+      "step": 4876
+    },
+    {
+      "epoch": 2.4195109842373093,
+      "grad_norm": 0.07446272844243572,
+      "learning_rate": 3.3892554447172066e-06,
+      "loss": 0.4414,
+      "step": 4877
+    },
+    {
+      "epoch": 2.4200074469405486,
+      "grad_norm": 0.07141489930382675,
+      "learning_rate": 3.3874050755461984e-06,
+      "loss": 0.4506,
+      "step": 4878
+    },
+    {
+      "epoch": 2.420503909643788,
+      "grad_norm": 0.07426444294024223,
+      "learning_rate": 3.385554952834469e-06,
+      "loss": 0.4685,
+      "step": 4879
+    },
+    {
+      "epoch": 2.4210003723470273,
+      "grad_norm": 0.06977047817721685,
+      "learning_rate": 3.3837050768647784e-06,
+      "loss": 0.4509,
+      "step": 4880
+    },
+    {
+      "epoch": 2.421496835050267,
+      "grad_norm": 0.07200114942417052,
+      "learning_rate": 3.3818554479198532e-06,
+      "loss": 0.461,
+      "step": 4881
+    },
+    {
+      "epoch": 2.4219932977535064,
+      "grad_norm": 0.07146768375177204,
+      "learning_rate": 3.380006066282378e-06,
+      "loss": 0.4449,
+      "step": 4882
+    },
+    {
+      "epoch": 2.4224897604567457,
+      "grad_norm": 0.07020862149032203,
+      "learning_rate": 3.3781569322350006e-06,
+      "loss": 0.433,
+      "step": 4883
+    },
+    {
+      "epoch": 2.422986223159985,
+      "grad_norm": 0.07643393327996241,
+      "learning_rate": 3.3763080460603307e-06,
+      "loss": 0.4977,
+      "step": 4884
+    },
+    {
+      "epoch": 2.4234826858632244,
+      "grad_norm": 0.07396500457066017,
+      "learning_rate": 3.374459408040942e-06,
+      "loss": 0.452,
+      "step": 4885
+    },
+    {
+      "epoch": 2.423979148566464,
+      "grad_norm": 0.07520521684618664,
+      "learning_rate": 3.3726110184593697e-06,
+      "loss": 0.4555,
+      "step": 4886
+    },
+    {
+      "epoch": 2.4244756112697035,
+      "grad_norm": 0.07150715320587676,
+      "learning_rate": 3.3707628775981106e-06,
+      "loss": 0.4581,
+      "step": 4887
+    },
+    {
+      "epoch": 2.424972073972943,
+      "grad_norm": 0.07074504126181777,
+      "learning_rate": 3.368914985739622e-06,
+      "loss": 0.4827,
+      "step": 4888
+    },
+    {
+      "epoch": 2.425468536676182,
+      "grad_norm": 0.0711238360588336,
+      "learning_rate": 3.367067343166326e-06,
+      "loss": 0.4586,
+      "step": 4889
+    },
+    {
+      "epoch": 2.4259649993794214,
+      "grad_norm": 0.07267839426440299,
+      "learning_rate": 3.365219950160603e-06,
+      "loss": 0.4937,
+      "step": 4890
+    },
+    {
+      "epoch": 2.4264614620826612,
+      "grad_norm": 0.07170663157707102,
+      "learning_rate": 3.3633728070048e-06,
+      "loss": 0.4666,
+      "step": 4891
+    },
+    {
+      "epoch": 2.4269579247859006,
+      "grad_norm": 0.07319404492664315,
+      "learning_rate": 3.3615259139812227e-06,
+      "loss": 0.4579,
+      "step": 4892
+    },
+    {
+      "epoch": 2.42745438748914,
+      "grad_norm": 0.07268583391352942,
+      "learning_rate": 3.359679271372138e-06,
+      "loss": 0.4534,
+      "step": 4893
+    },
+    {
+      "epoch": 2.427950850192379,
+      "grad_norm": 0.07320504663664425,
+      "learning_rate": 3.357832879459776e-06,
+      "loss": 0.4474,
+      "step": 4894
+    },
+    {
+      "epoch": 2.4284473128956185,
+      "grad_norm": 0.07329344087726802,
+      "learning_rate": 3.3559867385263277e-06,
+      "loss": 0.4982,
+      "step": 4895
+    },
+    {
+      "epoch": 2.4289437755988583,
+      "grad_norm": 0.0741107539071744,
+      "learning_rate": 3.3541408488539474e-06,
+      "loss": 0.4815,
+      "step": 4896
+    },
+    {
+      "epoch": 2.4294402383020977,
+      "grad_norm": 0.07340753028917309,
+      "learning_rate": 3.3522952107247496e-06,
+      "loss": 0.4671,
+      "step": 4897
+    },
+    {
+      "epoch": 2.429936701005337,
+      "grad_norm": 0.07199561910628434,
+      "learning_rate": 3.35044982442081e-06,
+      "loss": 0.4743,
+      "step": 4898
+    },
+    {
+      "epoch": 2.4304331637085763,
+      "grad_norm": 0.07489732948889798,
+      "learning_rate": 3.3486046902241663e-06,
+      "loss": 0.4826,
+      "step": 4899
+    },
+    {
+      "epoch": 2.4309296264118156,
+      "grad_norm": 0.08229476417910944,
+      "learning_rate": 3.346759808416816e-06,
+      "loss": 0.4673,
+      "step": 4900
+    },
+    {
+      "epoch": 2.4314260891150554,
+      "grad_norm": 0.0734959805508446,
+      "learning_rate": 3.344915179280722e-06,
+      "loss": 0.4691,
+      "step": 4901
+    },
+    {
+      "epoch": 2.4319225518182948,
+      "grad_norm": 0.07588738560643025,
+      "learning_rate": 3.3430708030978055e-06,
+      "loss": 0.4871,
+      "step": 4902
+    },
+    {
+      "epoch": 2.432419014521534,
+      "grad_norm": 0.0723640039745302,
+      "learning_rate": 3.3412266801499503e-06,
+      "loss": 0.4723,
+      "step": 4903
+    },
+    {
+      "epoch": 2.4329154772247734,
+      "grad_norm": 0.07436716244661329,
+      "learning_rate": 3.339382810719001e-06,
+      "loss": 0.4694,
+      "step": 4904
+    },
+    {
+      "epoch": 2.4334119399280127,
+      "grad_norm": 0.07110548824172484,
+      "learning_rate": 3.337539195086762e-06,
+      "loss": 0.4439,
+      "step": 4905
+    },
+    {
+      "epoch": 2.4339084026312525,
+      "grad_norm": 0.07318853233485054,
+      "learning_rate": 3.335695833535001e-06,
+      "loss": 0.4558,
+      "step": 4906
+    },
+    {
+      "epoch": 2.434404865334492,
+      "grad_norm": 0.07070760691238002,
+      "learning_rate": 3.3338527263454478e-06,
+      "loss": 0.4672,
+      "step": 4907
+    },
+    {
+      "epoch": 2.434901328037731,
+      "grad_norm": 0.07408841815287856,
+      "learning_rate": 3.3320098737997915e-06,
+      "loss": 0.4727,
+      "step": 4908
+    },
+    {
+      "epoch": 2.4353977907409705,
+      "grad_norm": 0.07050033637736432,
+      "learning_rate": 3.3301672761796805e-06,
+      "loss": 0.438,
+      "step": 4909
+    },
+    {
+      "epoch": 2.43589425344421,
+      "grad_norm": 0.07223651097731444,
+      "learning_rate": 3.328324933766728e-06,
+      "loss": 0.4651,
+      "step": 4910
+    },
+    {
+      "epoch": 2.436390716147449,
+      "grad_norm": 0.07259102222113285,
+      "learning_rate": 3.326482846842506e-06,
+      "loss": 0.442,
+      "step": 4911
+    },
+    {
+      "epoch": 2.436887178850689,
+      "grad_norm": 0.07238593049477171,
+      "learning_rate": 3.3246410156885477e-06,
+      "loss": 0.4697,
+      "step": 4912
+    },
+    {
+      "epoch": 2.4373836415539283,
+      "grad_norm": 0.07343801139786234,
+      "learning_rate": 3.322799440586349e-06,
+      "loss": 0.4807,
+      "step": 4913
+    },
+    {
+      "epoch": 2.4378801042571676,
+      "grad_norm": 0.0704308581928645,
+      "learning_rate": 3.3209581218173636e-06,
+      "loss": 0.4603,
+      "step": 4914
+    },
+    {
+      "epoch": 2.438376566960407,
+      "grad_norm": 0.07078467248069595,
+      "learning_rate": 3.3191170596630085e-06,
+      "loss": 0.4174,
+      "step": 4915
+    },
+    {
+      "epoch": 2.4388730296636467,
+      "grad_norm": 0.07345596267263887,
+      "learning_rate": 3.317276254404659e-06,
+      "loss": 0.4572,
+      "step": 4916
+    },
+    {
+      "epoch": 2.439369492366886,
+      "grad_norm": 0.07196161644566479,
+      "learning_rate": 3.315435706323653e-06,
+      "loss": 0.4454,
+      "step": 4917
+    },
+    {
+      "epoch": 2.4398659550701254,
+      "grad_norm": 0.07141470733605967,
+      "learning_rate": 3.3135954157012894e-06,
+      "loss": 0.4449,
+      "step": 4918
+    },
+    {
+      "epoch": 2.4403624177733647,
+      "grad_norm": 0.07559725227479616,
+      "learning_rate": 3.3117553828188275e-06,
+      "loss": 0.4462,
+      "step": 4919
+    },
+    {
+      "epoch": 2.440858880476604,
+      "grad_norm": 0.07448387570855905,
+      "learning_rate": 3.309915607957487e-06,
+      "loss": 0.4733,
+      "step": 4920
+    },
+    {
+      "epoch": 2.4413553431798434,
+      "grad_norm": 0.07351317002445866,
+      "learning_rate": 3.3080760913984468e-06,
+      "loss": 0.4597,
+      "step": 4921
+    },
+    {
+      "epoch": 2.441851805883083,
+      "grad_norm": 0.06980086573237342,
+      "learning_rate": 3.306236833422848e-06,
+      "loss": 0.4514,
+      "step": 4922
+    },
+    {
+      "epoch": 2.4423482685863225,
+      "grad_norm": 0.07197898130409723,
+      "learning_rate": 3.3043978343117916e-06,
+      "loss": 0.4644,
+      "step": 4923
+    },
+    {
+      "epoch": 2.442844731289562,
+      "grad_norm": 0.07488286290152556,
+      "learning_rate": 3.3025590943463403e-06,
+      "loss": 0.4903,
+      "step": 4924
+    },
+    {
+      "epoch": 2.443341193992801,
+      "grad_norm": 0.07248650394814932,
+      "learning_rate": 3.3007206138075143e-06,
+      "loss": 0.4519,
+      "step": 4925
+    },
+    {
+      "epoch": 2.443837656696041,
+      "grad_norm": 0.0690974071655818,
+      "learning_rate": 3.2988823929762965e-06,
+      "loss": 0.4323,
+      "step": 4926
+    },
+    {
+      "epoch": 2.4443341193992802,
+      "grad_norm": 0.07407912445276958,
+      "learning_rate": 3.2970444321336294e-06,
+      "loss": 0.4667,
+      "step": 4927
+    },
+    {
+      "epoch": 2.4448305821025196,
+      "grad_norm": 0.07280384197326245,
+      "learning_rate": 3.2952067315604162e-06,
+      "loss": 0.4606,
+      "step": 4928
+    },
+    {
+      "epoch": 2.445327044805759,
+      "grad_norm": 0.07026914227270502,
+      "learning_rate": 3.2933692915375205e-06,
+      "loss": 0.4274,
+      "step": 4929
+    },
+    {
+      "epoch": 2.4458235075089982,
+      "grad_norm": 0.07225020966833783,
+      "learning_rate": 3.2915321123457654e-06,
+      "loss": 0.4325,
+      "step": 4930
+    },
+    {
+      "epoch": 2.4463199702122376,
+      "grad_norm": 0.07631532903043024,
+      "learning_rate": 3.2896951942659334e-06,
+      "loss": 0.4743,
+      "step": 4931
+    },
+    {
+      "epoch": 2.4468164329154773,
+      "grad_norm": 0.07656256169012282,
+      "learning_rate": 3.2878585375787676e-06,
+      "loss": 0.4711,
+      "step": 4932
+    },
+    {
+      "epoch": 2.4473128956187167,
+      "grad_norm": 0.0757992804891011,
+      "learning_rate": 3.2860221425649714e-06,
+      "loss": 0.4798,
+      "step": 4933
+    },
+    {
+      "epoch": 2.447809358321956,
+      "grad_norm": 0.07241109508407907,
+      "learning_rate": 3.2841860095052096e-06,
+      "loss": 0.4587,
+      "step": 4934
+    },
+    {
+      "epoch": 2.4483058210251953,
+      "grad_norm": 0.07277584878584002,
+      "learning_rate": 3.2823501386801055e-06,
+      "loss": 0.4628,
+      "step": 4935
+    },
+    {
+      "epoch": 2.448802283728435,
+      "grad_norm": 0.07268028763143215,
+      "learning_rate": 3.2805145303702433e-06,
+      "loss": 0.4536,
+      "step": 4936
+    },
+    {
+      "epoch": 2.4492987464316744,
+      "grad_norm": 0.07500003929320131,
+      "learning_rate": 3.278679184856164e-06,
+      "loss": 0.4681,
+      "step": 4937
+    },
+    {
+      "epoch": 2.4497952091349138,
+      "grad_norm": 0.07799652555267876,
+      "learning_rate": 3.276844102418372e-06,
+      "loss": 0.51,
+      "step": 4938
+    },
+    {
+      "epoch": 2.450291671838153,
+      "grad_norm": 0.07138560918989682,
+      "learning_rate": 3.2750092833373303e-06,
+      "loss": 0.4368,
+      "step": 4939
+    },
+    {
+      "epoch": 2.4507881345413924,
+      "grad_norm": 0.07259687502633978,
+      "learning_rate": 3.273174727893463e-06,
+      "loss": 0.4782,
+      "step": 4940
+    },
+    {
+      "epoch": 2.4512845972446318,
+      "grad_norm": 0.07348721712699163,
+      "learning_rate": 3.27134043636715e-06,
+      "loss": 0.4433,
+      "step": 4941
+    },
+    {
+      "epoch": 2.4517810599478715,
+      "grad_norm": 0.07367500408800275,
+      "learning_rate": 3.2695064090387328e-06,
+      "loss": 0.4806,
+      "step": 4942
+    },
+    {
+      "epoch": 2.452277522651111,
+      "grad_norm": 0.0718678944614559,
+      "learning_rate": 3.2676726461885167e-06,
+      "loss": 0.4614,
+      "step": 4943
+    },
+    {
+      "epoch": 2.45277398535435,
+      "grad_norm": 0.07364739250647794,
+      "learning_rate": 3.2658391480967594e-06,
+      "loss": 0.4864,
+      "step": 4944
+    },
+    {
+      "epoch": 2.4532704480575895,
+      "grad_norm": 0.07305171136004614,
+      "learning_rate": 3.264005915043685e-06,
+      "loss": 0.479,
+      "step": 4945
+    },
+    {
+      "epoch": 2.4537669107608293,
+      "grad_norm": 0.07653717036474185,
+      "learning_rate": 3.2621729473094704e-06,
+      "loss": 0.4646,
+      "step": 4946
+    },
+    {
+      "epoch": 2.4542633734640686,
+      "grad_norm": 0.07221997080472961,
+      "learning_rate": 3.2603402451742594e-06,
+      "loss": 0.4405,
+      "step": 4947
+    },
+    {
+      "epoch": 2.454759836167308,
+      "grad_norm": 0.0737897302199099,
+      "learning_rate": 3.258507808918146e-06,
+      "loss": 0.4587,
+      "step": 4948
+    },
+    {
+      "epoch": 2.4552562988705473,
+      "grad_norm": 0.07212895440920399,
+      "learning_rate": 3.2566756388211917e-06,
+      "loss": 0.4522,
+      "step": 4949
+    },
+    {
+      "epoch": 2.4557527615737866,
+      "grad_norm": 0.07275303783957517,
+      "learning_rate": 3.254843735163414e-06,
+      "loss": 0.435,
+      "step": 4950
+    },
+    {
+      "epoch": 2.456249224277026,
+      "grad_norm": 0.07110446769175349,
+      "learning_rate": 3.253012098224789e-06,
+      "loss": 0.4625,
+      "step": 4951
+    },
+    {
+      "epoch": 2.4567456869802657,
+      "grad_norm": 0.07533616154433118,
+      "learning_rate": 3.2511807282852564e-06,
+      "loss": 0.476,
+      "step": 4952
+    },
+    {
+      "epoch": 2.457242149683505,
+      "grad_norm": 0.07121508664898368,
+      "learning_rate": 3.2493496256247074e-06,
+      "loss": 0.437,
+      "step": 4953
+    },
+    {
+      "epoch": 2.4577386123867444,
+      "grad_norm": 0.07253166936136166,
+      "learning_rate": 3.247518790522999e-06,
+      "loss": 0.4483,
+      "step": 4954
+    },
+    {
+      "epoch": 2.4582350750899837,
+      "grad_norm": 0.07591020227064676,
+      "learning_rate": 3.245688223259944e-06,
+      "loss": 0.4883,
+      "step": 4955
+    },
+    {
+      "epoch": 2.4587315377932235,
+      "grad_norm": 0.07434210263708153,
+      "learning_rate": 3.2438579241153166e-06,
+      "loss": 0.4957,
+      "step": 4956
+    },
+    {
+      "epoch": 2.459228000496463,
+      "grad_norm": 0.0736921894882178,
+      "learning_rate": 3.242027893368849e-06,
+      "loss": 0.4487,
+      "step": 4957
+    },
+    {
+      "epoch": 2.459724463199702,
+      "grad_norm": 0.07417743880067373,
+      "learning_rate": 3.240198131300229e-06,
+      "loss": 0.4582,
+      "step": 4958
+    },
+    {
+      "epoch": 2.4602209259029415,
+      "grad_norm": 0.07213475102302708,
+      "learning_rate": 3.2383686381891087e-06,
+      "loss": 0.4626,
+      "step": 4959
+    },
+    {
+      "epoch": 2.460717388606181,
+      "grad_norm": 0.07342311279959657,
+      "learning_rate": 3.236539414315096e-06,
+      "loss": 0.4572,
+      "step": 4960
+    },
+    {
+      "epoch": 2.46121385130942,
+      "grad_norm": 0.07185612590850873,
+      "learning_rate": 3.234710459957761e-06,
+      "loss": 0.4919,
+      "step": 4961
+    },
+    {
+      "epoch": 2.46171031401266,
+      "grad_norm": 0.07275605720944686,
+      "learning_rate": 3.232881775396626e-06,
+      "loss": 0.4511,
+      "step": 4962
+    },
+    {
+      "epoch": 2.4622067767158993,
+      "grad_norm": 0.07622583368006376,
+      "learning_rate": 3.2310533609111805e-06,
+      "loss": 0.4777,
+      "step": 4963
+    },
+    {
+      "epoch": 2.4627032394191386,
+      "grad_norm": 0.07466851152744547,
+      "learning_rate": 3.229225216780864e-06,
+      "loss": 0.4793,
+      "step": 4964
+    },
+    {
+      "epoch": 2.463199702122378,
+      "grad_norm": 0.07152900402073481,
+      "learning_rate": 3.227397343285081e-06,
+      "loss": 0.4191,
+      "step": 4965
+    },
+    {
+      "epoch": 2.4636961648256177,
+      "grad_norm": 0.07188262340628947,
+      "learning_rate": 3.2255697407031924e-06,
+      "loss": 0.4366,
+      "step": 4966
+    },
+    {
+      "epoch": 2.464192627528857,
+      "grad_norm": 0.07348716146826911,
+      "learning_rate": 3.2237424093145175e-06,
+      "loss": 0.4817,
+      "step": 4967
+    },
+    {
+      "epoch": 2.4646890902320964,
+      "grad_norm": 0.0713609536000802,
+      "learning_rate": 3.221915349398337e-06,
+      "loss": 0.4474,
+      "step": 4968
+    },
+    {
+      "epoch": 2.4651855529353357,
+      "grad_norm": 0.07659482845157203,
+      "learning_rate": 3.2200885612338846e-06,
+      "loss": 0.4926,
+      "step": 4969
+    },
+    {
+      "epoch": 2.465682015638575,
+      "grad_norm": 0.07235133973202737,
+      "learning_rate": 3.2182620451003565e-06,
+      "loss": 0.4976,
+      "step": 4970
+    },
+    {
+      "epoch": 2.4661784783418144,
+      "grad_norm": 0.07497447521649205,
+      "learning_rate": 3.216435801276907e-06,
+      "loss": 0.4966,
+      "step": 4971
+    },
+    {
+      "epoch": 2.466674941045054,
+      "grad_norm": 0.07042342843861166,
+      "learning_rate": 3.2146098300426485e-06,
+      "loss": 0.4481,
+      "step": 4972
+    },
+    {
+      "epoch": 2.4671714037482935,
+      "grad_norm": 0.07471585111012378,
+      "learning_rate": 3.2127841316766515e-06,
+      "loss": 0.4472,
+      "step": 4973
+    },
+    {
+      "epoch": 2.467667866451533,
+      "grad_norm": 0.07143622672924964,
+      "learning_rate": 3.210958706457944e-06,
+      "loss": 0.4297,
+      "step": 4974
+    },
+    {
+      "epoch": 2.468164329154772,
+      "grad_norm": 0.07066114201680801,
+      "learning_rate": 3.2091335546655124e-06,
+      "loss": 0.458,
+      "step": 4975
+    },
+    {
+      "epoch": 2.468660791858012,
+      "grad_norm": 0.07504532464434141,
+      "learning_rate": 3.2073086765783023e-06,
+      "loss": 0.4729,
+      "step": 4976
+    },
+    {
+      "epoch": 2.4691572545612512,
+      "grad_norm": 0.07239769361863543,
+      "learning_rate": 3.2054840724752184e-06,
+      "loss": 0.4961,
+      "step": 4977
+    },
+    {
+      "epoch": 2.4696537172644906,
+      "grad_norm": 0.07219230294929461,
+      "learning_rate": 3.2036597426351203e-06,
+      "loss": 0.4435,
+      "step": 4978
+    },
+    {
+      "epoch": 2.47015017996773,
+      "grad_norm": 0.0758055553305713,
+      "learning_rate": 3.2018356873368307e-06,
+      "loss": 0.4634,
+      "step": 4979
+    },
+    {
+      "epoch": 2.4706466426709692,
+      "grad_norm": 0.07221304089912467,
+      "learning_rate": 3.2000119068591227e-06,
+      "loss": 0.4689,
+      "step": 4980
+    },
+    {
+      "epoch": 2.4711431053742086,
+      "grad_norm": 0.07256877504850277,
+      "learning_rate": 3.198188401480734e-06,
+      "loss": 0.4632,
+      "step": 4981
+    },
+    {
+      "epoch": 2.4716395680774483,
+      "grad_norm": 0.07368479644516852,
+      "learning_rate": 3.196365171480359e-06,
+      "loss": 0.4585,
+      "step": 4982
+    },
+    {
+      "epoch": 2.4721360307806877,
+      "grad_norm": 0.07170466723428409,
+      "learning_rate": 3.1945422171366482e-06,
+      "loss": 0.4643,
+      "step": 4983
+    },
+    {
+      "epoch": 2.472632493483927,
+      "grad_norm": 0.0736886755768893,
+      "learning_rate": 3.192719538728212e-06,
+      "loss": 0.4404,
+      "step": 4984
+    },
+    {
+      "epoch": 2.4731289561871663,
+      "grad_norm": 0.07409447169142401,
+      "learning_rate": 3.190897136533615e-06,
+      "loss": 0.4486,
+      "step": 4985
+    },
+    {
+      "epoch": 2.473625418890406,
+      "grad_norm": 0.07350626854049384,
+      "learning_rate": 3.189075010831385e-06,
+      "loss": 0.4565,
+      "step": 4986
+    },
+    {
+      "epoch": 2.4741218815936454,
+      "grad_norm": 0.07724817239470792,
+      "learning_rate": 3.1872531619000024e-06,
+      "loss": 0.4575,
+      "step": 4987
+    },
+    {
+      "epoch": 2.4746183442968848,
+      "grad_norm": 0.07394461490386524,
+      "learning_rate": 3.18543159001791e-06,
+      "loss": 0.465,
+      "step": 4988
+    },
+    {
+      "epoch": 2.475114807000124,
+      "grad_norm": 0.07219693102311857,
+      "learning_rate": 3.183610295463505e-06,
+      "loss": 0.4636,
+      "step": 4989
+    },
+    {
+      "epoch": 2.4756112697033634,
+      "grad_norm": 0.0755170770173934,
+      "learning_rate": 3.1817892785151426e-06,
+      "loss": 0.4921,
+      "step": 4990
+    },
+    {
+      "epoch": 2.4761077324066028,
+      "grad_norm": 0.07125952744942823,
+      "learning_rate": 3.179968539451135e-06,
+      "loss": 0.4562,
+      "step": 4991
+    },
+    {
+      "epoch": 2.4766041951098425,
+      "grad_norm": 0.0710903052738868,
+      "learning_rate": 3.1781480785497555e-06,
+      "loss": 0.4521,
+      "step": 4992
+    },
+    {
+      "epoch": 2.477100657813082,
+      "grad_norm": 0.07147477701855874,
+      "learning_rate": 3.17632789608923e-06,
+      "loss": 0.4319,
+      "step": 4993
+    },
+    {
+      "epoch": 2.477597120516321,
+      "grad_norm": 0.06860598579714546,
+      "learning_rate": 3.174507992347746e-06,
+      "loss": 0.4218,
+      "step": 4994
+    },
+    {
+      "epoch": 2.4780935832195605,
+      "grad_norm": 0.06928050623670558,
+      "learning_rate": 3.172688367603447e-06,
+      "loss": 0.4201,
+      "step": 4995
+    },
+    {
+      "epoch": 2.4785900459228,
+      "grad_norm": 0.07393072454941838,
+      "learning_rate": 3.170869022134432e-06,
+      "loss": 0.5016,
+      "step": 4996
+    },
+    {
+      "epoch": 2.4790865086260396,
+      "grad_norm": 0.07611053154211594,
+      "learning_rate": 3.1690499562187573e-06,
+      "loss": 0.4742,
+      "step": 4997
+    },
+    {
+      "epoch": 2.479582971329279,
+      "grad_norm": 0.07172895151999717,
+      "learning_rate": 3.1672311701344404e-06,
+      "loss": 0.44,
+      "step": 4998
+    },
+    {
+      "epoch": 2.4800794340325183,
+      "grad_norm": 0.0717638559387684,
+      "learning_rate": 3.165412664159453e-06,
+      "loss": 0.4627,
+      "step": 4999
+    },
+    {
+      "epoch": 2.4805758967357576,
+      "grad_norm": 0.07041680276845996,
+      "learning_rate": 3.163594438571725e-06,
+      "loss": 0.4697,
+      "step": 5000
+    },
+    {
+      "epoch": 2.481072359438997,
+      "grad_norm": 0.06961263950820218,
+      "learning_rate": 3.161776493649141e-06,
+      "loss": 0.44,
+      "step": 5001
+    },
+    {
+      "epoch": 2.4815688221422367,
+      "grad_norm": 0.07677040647683987,
+      "learning_rate": 3.1599588296695476e-06,
+      "loss": 0.4688,
+      "step": 5002
+    },
+    {
+      "epoch": 2.482065284845476,
+      "grad_norm": 0.07432966041252718,
+      "learning_rate": 3.158141446910744e-06,
+      "loss": 0.4706,
+      "step": 5003
+    },
+    {
+      "epoch": 2.4825617475487154,
+      "grad_norm": 0.070305882460339,
+      "learning_rate": 3.1563243456504877e-06,
+      "loss": 0.4812,
+      "step": 5004
+    },
+    {
+      "epoch": 2.4830582102519547,
+      "grad_norm": 0.07137322616898634,
+      "learning_rate": 3.1545075261664954e-06,
+      "loss": 0.4549,
+      "step": 5005
+    },
+    {
+      "epoch": 2.483554672955194,
+      "grad_norm": 0.07262431628108891,
+      "learning_rate": 3.1526909887364365e-06,
+      "loss": 0.4325,
+      "step": 5006
+    },
+    {
+      "epoch": 2.484051135658434,
+      "grad_norm": 0.07360856338067412,
+      "learning_rate": 3.1508747336379407e-06,
+      "loss": 0.4459,
+      "step": 5007
+    },
+    {
+      "epoch": 2.484547598361673,
+      "grad_norm": 0.07116264212561499,
+      "learning_rate": 3.1490587611485936e-06,
+      "loss": 0.4599,
+      "step": 5008
+    },
+    {
+      "epoch": 2.4850440610649125,
+      "grad_norm": 0.07434591832468768,
+      "learning_rate": 3.1472430715459366e-06,
+      "loss": 0.4683,
+      "step": 5009
+    },
+    {
+      "epoch": 2.485540523768152,
+      "grad_norm": 0.0726460286968446,
+      "learning_rate": 3.145427665107471e-06,
+      "loss": 0.4627,
+      "step": 5010
+    },
+    {
+      "epoch": 2.486036986471391,
+      "grad_norm": 0.07457139906557984,
+      "learning_rate": 3.1436125421106507e-06,
+      "loss": 0.4867,
+      "step": 5011
+    },
+    {
+      "epoch": 2.486533449174631,
+      "grad_norm": 0.07471436808388297,
+      "learning_rate": 3.1417977028328884e-06,
+      "loss": 0.5004,
+      "step": 5012
+    },
+    {
+      "epoch": 2.4870299118778703,
+      "grad_norm": 0.07351308940818314,
+      "learning_rate": 3.139983147551552e-06,
+      "loss": 0.4538,
+      "step": 5013
+    },
+    {
+      "epoch": 2.4875263745811096,
+      "grad_norm": 0.07325613680703755,
+      "learning_rate": 3.138168876543969e-06,
+      "loss": 0.4745,
+      "step": 5014
+    },
+    {
+      "epoch": 2.488022837284349,
+      "grad_norm": 0.07606605052849479,
+      "learning_rate": 3.136354890087421e-06,
+      "loss": 0.4661,
+      "step": 5015
+    },
+    {
+      "epoch": 2.4885192999875883,
+      "grad_norm": 0.07037890181768297,
+      "learning_rate": 3.134541188459147e-06,
+      "loss": 0.4309,
+      "step": 5016
+    },
+    {
+      "epoch": 2.489015762690828,
+      "grad_norm": 0.07154622198629479,
+      "learning_rate": 3.1327277719363413e-06,
+      "loss": 0.4731,
+      "step": 5017
+    },
+    {
+      "epoch": 2.4895122253940674,
+      "grad_norm": 0.07199847674452635,
+      "learning_rate": 3.1309146407961565e-06,
+      "loss": 0.4684,
+      "step": 5018
+    },
+    {
+      "epoch": 2.4900086880973067,
+      "grad_norm": 0.0688715415526427,
+      "learning_rate": 3.1291017953157003e-06,
+      "loss": 0.4405,
+      "step": 5019
+    },
+    {
+      "epoch": 2.490505150800546,
+      "grad_norm": 0.07243291775387756,
+      "learning_rate": 3.1272892357720376e-06,
+      "loss": 0.466,
+      "step": 5020
+    },
+    {
+      "epoch": 2.4910016135037854,
+      "grad_norm": 0.072838356671563,
+      "learning_rate": 3.125476962442189e-06,
+      "loss": 0.456,
+      "step": 5021
+    },
+    {
+      "epoch": 2.491498076207025,
+      "grad_norm": 0.0726349062067118,
+      "learning_rate": 3.12366497560313e-06,
+      "loss": 0.461,
+      "step": 5022
+    },
+    {
+      "epoch": 2.4919945389102645,
+      "grad_norm": 0.07215929842189774,
+      "learning_rate": 3.121853275531794e-06,
+      "loss": 0.441,
+      "step": 5023
+    },
+    {
+      "epoch": 2.492491001613504,
+      "grad_norm": 0.07542042515887382,
+      "learning_rate": 3.120041862505072e-06,
+      "loss": 0.5146,
+      "step": 5024
+    },
+    {
+      "epoch": 2.492987464316743,
+      "grad_norm": 0.2562024226795915,
+      "learning_rate": 3.118230736799809e-06,
+      "loss": 0.4881,
+      "step": 5025
+    },
+    {
+      "epoch": 2.4934839270199824,
+      "grad_norm": 0.07217630690547586,
+      "learning_rate": 3.1164198986928064e-06,
+      "loss": 0.482,
+      "step": 5026
+    },
+    {
+      "epoch": 2.4939803897232222,
+      "grad_norm": 0.07135677306049457,
+      "learning_rate": 3.114609348460821e-06,
+      "loss": 0.4663,
+      "step": 5027
+    },
+    {
+      "epoch": 2.4944768524264616,
+      "grad_norm": 0.07086569894661765,
+      "learning_rate": 3.1127990863805668e-06,
+      "loss": 0.4597,
+      "step": 5028
+    },
+    {
+      "epoch": 2.494973315129701,
+      "grad_norm": 0.07288462245322094,
+      "learning_rate": 3.110989112728713e-06,
+      "loss": 0.4716,
+      "step": 5029
+    },
+    {
+      "epoch": 2.49546977783294,
+      "grad_norm": 0.07127975550983165,
+      "learning_rate": 3.1091794277818845e-06,
+      "loss": 0.4759,
+      "step": 5030
+    },
+    {
+      "epoch": 2.4959662405361795,
+      "grad_norm": 0.07251358966959501,
+      "learning_rate": 3.1073700318166638e-06,
+      "loss": 0.47,
+      "step": 5031
+    },
+    {
+      "epoch": 2.4964627032394193,
+      "grad_norm": 0.07208491301981003,
+      "learning_rate": 3.1055609251095874e-06,
+      "loss": 0.4532,
+      "step": 5032
+    },
+    {
+      "epoch": 2.4969591659426587,
+      "grad_norm": 0.07166247839743241,
+      "learning_rate": 3.1037521079371503e-06,
+      "loss": 0.4669,
+      "step": 5033
+    },
+    {
+      "epoch": 2.497455628645898,
+      "grad_norm": 0.07111482822718003,
+      "learning_rate": 3.101943580575798e-06,
+      "loss": 0.4285,
+      "step": 5034
+    },
+    {
+      "epoch": 2.4979520913491373,
+      "grad_norm": 0.0749071795087903,
+      "learning_rate": 3.1001353433019365e-06,
+      "loss": 0.466,
+      "step": 5035
+    },
+    {
+      "epoch": 2.4984485540523766,
+      "grad_norm": 0.07281000376215524,
+      "learning_rate": 3.098327396391926e-06,
+      "loss": 0.4883,
+      "step": 5036
+    },
+    {
+      "epoch": 2.4989450167556164,
+      "grad_norm": 0.07552457928843555,
+      "learning_rate": 3.0965197401220824e-06,
+      "loss": 0.4741,
+      "step": 5037
+    },
+    {
+      "epoch": 2.4994414794588558,
+      "grad_norm": 0.07389154079442611,
+      "learning_rate": 3.0947123747686756e-06,
+      "loss": 0.4884,
+      "step": 5038
+    },
+    {
+      "epoch": 2.499937942162095,
+      "grad_norm": 0.0730632967561261,
+      "learning_rate": 3.0929053006079336e-06,
+      "loss": 0.4806,
+      "step": 5039
+    },
+    {
+      "epoch": 2.5004344048653344,
+      "grad_norm": 0.07248646669142482,
+      "learning_rate": 3.091098517916039e-06,
+      "loss": 0.4496,
+      "step": 5040
+    },
+    {
+      "epoch": 2.5004344048653344,
+      "eval_loss": 0.5154539942741394,
+      "eval_runtime": 259.0671,
+      "eval_samples_per_second": 117.163,
+      "eval_steps_per_second": 14.649,
+      "step": 5040
+    },
+    {
+      "epoch": 2.5009308675685737,
+      "grad_norm": 0.07096380398662457,
+      "learning_rate": 3.0892920269691284e-06,
+      "loss": 0.4433,
+      "step": 5041
+    },
+    {
+      "epoch": 2.501427330271813,
+      "grad_norm": 0.0711112506139062,
+      "learning_rate": 3.087485828043296e-06,
+      "loss": 0.4585,
+      "step": 5042
+    },
+    {
+      "epoch": 2.501923792975053,
+      "grad_norm": 0.07417792998843728,
+      "learning_rate": 3.085679921414591e-06,
+      "loss": 0.4804,
+      "step": 5043
+    },
+    {
+      "epoch": 2.502420255678292,
+      "grad_norm": 0.07232246786213962,
+      "learning_rate": 3.083874307359016e-06,
+      "loss": 0.4652,
+      "step": 5044
+    },
+    {
+      "epoch": 2.5029167183815315,
+      "grad_norm": 0.07231894396958254,
+      "learning_rate": 3.0820689861525295e-06,
+      "loss": 0.46,
+      "step": 5045
+    },
+    {
+      "epoch": 2.503413181084771,
+      "grad_norm": 0.0741298988460473,
+      "learning_rate": 3.0802639580710465e-06,
+      "loss": 0.4631,
+      "step": 5046
+    },
+    {
+      "epoch": 2.5039096437880106,
+      "grad_norm": 0.07126734306557868,
+      "learning_rate": 3.0784592233904363e-06,
+      "loss": 0.458,
+      "step": 5047
+    },
+    {
+      "epoch": 2.50440610649125,
+      "grad_norm": 0.07541124816238452,
+      "learning_rate": 3.0766547823865255e-06,
+      "loss": 0.4984,
+      "step": 5048
+    },
+    {
+      "epoch": 2.5049025691944893,
+      "grad_norm": 0.07292424807759758,
+      "learning_rate": 3.0748506353350928e-06,
+      "loss": 0.4524,
+      "step": 5049
+    },
+    {
+      "epoch": 2.5053990318977286,
+      "grad_norm": 0.07190778953502837,
+      "learning_rate": 3.0730467825118727e-06,
+      "loss": 0.4614,
+      "step": 5050
+    },
+    {
+      "epoch": 2.505895494600968,
+      "grad_norm": 0.07507039186827946,
+      "learning_rate": 3.0712432241925547e-06,
+      "loss": 0.4662,
+      "step": 5051
+    },
+    {
+      "epoch": 2.5063919573042073,
+      "grad_norm": 0.07226146094163524,
+      "learning_rate": 3.0694399606527853e-06,
+      "loss": 0.4409,
+      "step": 5052
+    },
+    {
+      "epoch": 2.506888420007447,
+      "grad_norm": 0.074729432072801,
+      "learning_rate": 3.067636992168165e-06,
+      "loss": 0.4838,
+      "step": 5053
+    },
+    {
+      "epoch": 2.5073848827106864,
+      "grad_norm": 0.0733553965599642,
+      "learning_rate": 3.0658343190142454e-06,
+      "loss": 0.453,
+      "step": 5054
+    },
+    {
+      "epoch": 2.5078813454139257,
+      "grad_norm": 0.06986856930273888,
+      "learning_rate": 3.064031941466539e-06,
+      "loss": 0.405,
+      "step": 5055
+    },
+    {
+      "epoch": 2.508377808117165,
+      "grad_norm": 0.0714472505987162,
+      "learning_rate": 3.0622298598005085e-06,
+      "loss": 0.4381,
+      "step": 5056
+    },
+    {
+      "epoch": 2.508874270820405,
+      "grad_norm": 0.07511575062773596,
+      "learning_rate": 3.060428074291575e-06,
+      "loss": 0.4821,
+      "step": 5057
+    },
+    {
+      "epoch": 2.509370733523644,
+      "grad_norm": 0.07157582915674207,
+      "learning_rate": 3.058626585215112e-06,
+      "loss": 0.4321,
+      "step": 5058
+    },
+    {
+      "epoch": 2.5098671962268835,
+      "grad_norm": 0.07366666603107923,
+      "learning_rate": 3.056825392846449e-06,
+      "loss": 0.4608,
+      "step": 5059
+    },
+    {
+      "epoch": 2.510363658930123,
+      "grad_norm": 0.07254565589833509,
+      "learning_rate": 3.0550244974608675e-06,
+      "loss": 0.473,
+      "step": 5060
+    },
+    {
+      "epoch": 2.510860121633362,
+      "grad_norm": 0.07520433230268554,
+      "learning_rate": 3.053223899333605e-06,
+      "loss": 0.493,
+      "step": 5061
+    },
+    {
+      "epoch": 2.5113565843366015,
+      "grad_norm": 0.0724140826027767,
+      "learning_rate": 3.0514235987398553e-06,
+      "loss": 0.4469,
+      "step": 5062
+    },
+    {
+      "epoch": 2.5118530470398412,
+      "grad_norm": 0.07411108457650745,
+      "learning_rate": 3.049623595954766e-06,
+      "loss": 0.4481,
+      "step": 5063
+    },
+    {
+      "epoch": 2.5123495097430806,
+      "grad_norm": 0.07471077841339425,
+      "learning_rate": 3.047823891253438e-06,
+      "loss": 0.5006,
+      "step": 5064
+    },
+    {
+      "epoch": 2.51284597244632,
+      "grad_norm": 0.07343300102872428,
+      "learning_rate": 3.046024484910929e-06,
+      "loss": 0.4542,
+      "step": 5065
+    },
+    {
+      "epoch": 2.5133424351495592,
+      "grad_norm": 0.0731642469946916,
+      "learning_rate": 3.0442253772022457e-06,
+      "loss": 0.4634,
+      "step": 5066
+    },
+    {
+      "epoch": 2.513838897852799,
+      "grad_norm": 0.07152879351921768,
+      "learning_rate": 3.0424265684023556e-06,
+      "loss": 0.4507,
+      "step": 5067
+    },
+    {
+      "epoch": 2.5143353605560383,
+      "grad_norm": 0.07093497332741829,
+      "learning_rate": 3.0406280587861775e-06,
+      "loss": 0.453,
+      "step": 5068
+    },
+    {
+      "epoch": 2.5148318232592777,
+      "grad_norm": 0.07409077210272415,
+      "learning_rate": 3.038829848628584e-06,
+      "loss": 0.4536,
+      "step": 5069
+    },
+    {
+      "epoch": 2.515328285962517,
+      "grad_norm": 0.07212986037709085,
+      "learning_rate": 3.0370319382044046e-06,
+      "loss": 0.474,
+      "step": 5070
+    },
+    {
+      "epoch": 2.5158247486657563,
+      "grad_norm": 0.07102190391920636,
+      "learning_rate": 3.035234327788418e-06,
+      "loss": 0.4156,
+      "step": 5071
+    },
+    {
+      "epoch": 2.5163212113689957,
+      "grad_norm": 0.0707281924452623,
+      "learning_rate": 3.033437017655363e-06,
+      "loss": 0.433,
+      "step": 5072
+    },
+    {
+      "epoch": 2.5168176740722354,
+      "grad_norm": 0.0693389163609712,
+      "learning_rate": 3.031640008079927e-06,
+      "loss": 0.4283,
+      "step": 5073
+    },
+    {
+      "epoch": 2.5173141367754748,
+      "grad_norm": 0.07143182707566005,
+      "learning_rate": 3.0298432993367577e-06,
+      "loss": 0.458,
+      "step": 5074
+    },
+    {
+      "epoch": 2.517810599478714,
+      "grad_norm": 0.07233793461023176,
+      "learning_rate": 3.02804689170045e-06,
+      "loss": 0.4539,
+      "step": 5075
+    },
+    {
+      "epoch": 2.5183070621819534,
+      "grad_norm": 0.07295793425095794,
+      "learning_rate": 3.026250785445558e-06,
+      "loss": 0.4411,
+      "step": 5076
+    },
+    {
+      "epoch": 2.518803524885193,
+      "grad_norm": 0.0727848842428524,
+      "learning_rate": 3.024454980846585e-06,
+      "loss": 0.465,
+      "step": 5077
+    },
+    {
+      "epoch": 2.5192999875884325,
+      "grad_norm": 0.07399648656532462,
+      "learning_rate": 3.0226594781779926e-06,
+      "loss": 0.4643,
+      "step": 5078
+    },
+    {
+      "epoch": 2.519796450291672,
+      "grad_norm": 0.07215847850860996,
+      "learning_rate": 3.0208642777141954e-06,
+      "loss": 0.4373,
+      "step": 5079
+    },
+    {
+      "epoch": 2.520292912994911,
+      "grad_norm": 0.07261890968416367,
+      "learning_rate": 3.01906937972956e-06,
+      "loss": 0.4551,
+      "step": 5080
+    },
+    {
+      "epoch": 2.5207893756981505,
+      "grad_norm": 0.06909668282131737,
+      "learning_rate": 3.0172747844984098e-06,
+      "loss": 0.4226,
+      "step": 5081
+    },
+    {
+      "epoch": 2.52128583840139,
+      "grad_norm": 0.0706356833724616,
+      "learning_rate": 3.0154804922950166e-06,
+      "loss": 0.4314,
+      "step": 5082
+    },
+    {
+      "epoch": 2.5217823011046296,
+      "grad_norm": 0.07115984795525102,
+      "learning_rate": 3.01368650339361e-06,
+      "loss": 0.4088,
+      "step": 5083
+    },
+    {
+      "epoch": 2.522278763807869,
+      "grad_norm": 0.07209812169556237,
+      "learning_rate": 3.011892818068374e-06,
+      "loss": 0.4479,
+      "step": 5084
+    },
+    {
+      "epoch": 2.5227752265111083,
+      "grad_norm": 0.07197949076439182,
+      "learning_rate": 3.0100994365934443e-06,
+      "loss": 0.4529,
+      "step": 5085
+    },
+    {
+      "epoch": 2.5232716892143476,
+      "grad_norm": 0.07294575988417028,
+      "learning_rate": 3.0083063592429108e-06,
+      "loss": 0.456,
+      "step": 5086
+    },
+    {
+      "epoch": 2.5237681519175874,
+      "grad_norm": 0.07085387809339508,
+      "learning_rate": 3.0065135862908147e-06,
+      "loss": 0.4451,
+      "step": 5087
+    },
+    {
+      "epoch": 2.5242646146208267,
+      "grad_norm": 0.0737491970059168,
+      "learning_rate": 3.0047211180111537e-06,
+      "loss": 0.448,
+      "step": 5088
+    },
+    {
+      "epoch": 2.524761077324066,
+      "grad_norm": 0.07316982337451786,
+      "learning_rate": 3.0029289546778782e-06,
+      "loss": 0.4675,
+      "step": 5089
+    },
+    {
+      "epoch": 2.5252575400273054,
+      "grad_norm": 0.07173406586479004,
+      "learning_rate": 3.0011370965648925e-06,
+      "loss": 0.4927,
+      "step": 5090
+    },
+    {
+      "epoch": 2.5257540027305447,
+      "grad_norm": 0.06955527651479629,
+      "learning_rate": 2.999345543946052e-06,
+      "loss": 0.4526,
+      "step": 5091
+    },
+    {
+      "epoch": 2.526250465433784,
+      "grad_norm": 0.06966454912794548,
+      "learning_rate": 2.997554297095167e-06,
+      "loss": 0.4378,
+      "step": 5092
+    },
+    {
+      "epoch": 2.526746928137024,
+      "grad_norm": 0.0720482880545221,
+      "learning_rate": 2.995763356286e-06,
+      "loss": 0.4619,
+      "step": 5093
+    },
+    {
+      "epoch": 2.527243390840263,
+      "grad_norm": 0.07265439796787254,
+      "learning_rate": 2.9939727217922685e-06,
+      "loss": 0.4676,
+      "step": 5094
+    },
+    {
+      "epoch": 2.5277398535435025,
+      "grad_norm": 0.07214286064337516,
+      "learning_rate": 2.9921823938876426e-06,
+      "loss": 0.4884,
+      "step": 5095
+    },
+    {
+      "epoch": 2.528236316246742,
+      "grad_norm": 0.07252007490926463,
+      "learning_rate": 2.990392372845744e-06,
+      "loss": 0.4385,
+      "step": 5096
+    },
+    {
+      "epoch": 2.5287327789499816,
+      "grad_norm": 0.07497006099995851,
+      "learning_rate": 2.9886026589401517e-06,
+      "loss": 0.4551,
+      "step": 5097
+    },
+    {
+      "epoch": 2.529229241653221,
+      "grad_norm": 0.074559241799246,
+      "learning_rate": 2.986813252444391e-06,
+      "loss": 0.4865,
+      "step": 5098
+    },
+    {
+      "epoch": 2.5297257043564603,
+      "grad_norm": 0.07273738797586408,
+      "learning_rate": 2.985024153631946e-06,
+      "loss": 0.4199,
+      "step": 5099
+    },
+    {
+      "epoch": 2.5302221670596996,
+      "grad_norm": 0.07055498664081801,
+      "learning_rate": 2.9832353627762513e-06,
+      "loss": 0.4251,
+      "step": 5100
+    },
+    {
+      "epoch": 2.530718629762939,
+      "grad_norm": 0.07257439997911404,
+      "learning_rate": 2.9814468801506945e-06,
+      "loss": 0.4616,
+      "step": 5101
+    },
+    {
+      "epoch": 2.5312150924661783,
+      "grad_norm": 0.07045107768696802,
+      "learning_rate": 2.979658706028619e-06,
+      "loss": 0.4351,
+      "step": 5102
+    },
+    {
+      "epoch": 2.531711555169418,
+      "grad_norm": 0.06993957099507159,
+      "learning_rate": 2.977870840683315e-06,
+      "loss": 0.3987,
+      "step": 5103
+    },
+    {
+      "epoch": 2.5322080178726574,
+      "grad_norm": 0.07233526039800486,
+      "learning_rate": 2.976083284388031e-06,
+      "loss": 0.4892,
+      "step": 5104
+    },
+    {
+      "epoch": 2.5327044805758967,
+      "grad_norm": 0.07852328921692033,
+      "learning_rate": 2.9742960374159656e-06,
+      "loss": 0.4635,
+      "step": 5105
+    },
+    {
+      "epoch": 2.533200943279136,
+      "grad_norm": 0.0724494414963727,
+      "learning_rate": 2.9725091000402716e-06,
+      "loss": 0.4802,
+      "step": 5106
+    },
+    {
+      "epoch": 2.533697405982376,
+      "grad_norm": 0.0686244582602942,
+      "learning_rate": 2.9707224725340543e-06,
+      "loss": 0.4251,
+      "step": 5107
+    },
+    {
+      "epoch": 2.534193868685615,
+      "grad_norm": 0.06821451041156537,
+      "learning_rate": 2.9689361551703693e-06,
+      "loss": 0.4142,
+      "step": 5108
+    },
+    {
+      "epoch": 2.5346903313888545,
+      "grad_norm": 0.07016181940774685,
+      "learning_rate": 2.9671501482222277e-06,
+      "loss": 0.4622,
+      "step": 5109
+    },
+    {
+      "epoch": 2.535186794092094,
+      "grad_norm": 0.07390854121226326,
+      "learning_rate": 2.9653644519625915e-06,
+      "loss": 0.4571,
+      "step": 5110
+    },
+    {
+      "epoch": 2.535683256795333,
+      "grad_norm": 0.07218976475871981,
+      "learning_rate": 2.963579066664375e-06,
+      "loss": 0.4771,
+      "step": 5111
+    },
+    {
+      "epoch": 2.5361797194985725,
+      "grad_norm": 0.07368470243297218,
+      "learning_rate": 2.961793992600447e-06,
+      "loss": 0.459,
+      "step": 5112
+    },
+    {
+      "epoch": 2.5366761822018122,
+      "grad_norm": 0.07320081493889653,
+      "learning_rate": 2.960009230043628e-06,
+      "loss": 0.4496,
+      "step": 5113
+    },
+    {
+      "epoch": 2.5371726449050516,
+      "grad_norm": 0.07304793240187735,
+      "learning_rate": 2.9582247792666876e-06,
+      "loss": 0.4598,
+      "step": 5114
+    },
+    {
+      "epoch": 2.537669107608291,
+      "grad_norm": 0.07328829880376828,
+      "learning_rate": 2.956440640542353e-06,
+      "loss": 0.4597,
+      "step": 5115
+    },
+    {
+      "epoch": 2.5381655703115302,
+      "grad_norm": 0.07674094919345029,
+      "learning_rate": 2.9546568141433007e-06,
+      "loss": 0.4669,
+      "step": 5116
+    },
+    {
+      "epoch": 2.53866203301477,
+      "grad_norm": 0.07259058089747297,
+      "learning_rate": 2.9528733003421597e-06,
+      "loss": 0.4718,
+      "step": 5117
+    },
+    {
+      "epoch": 2.5391584957180093,
+      "grad_norm": 0.07046795106722405,
+      "learning_rate": 2.9510900994115125e-06,
+      "loss": 0.4421,
+      "step": 5118
+    },
+    {
+      "epoch": 2.5396549584212487,
+      "grad_norm": 0.07287748922294839,
+      "learning_rate": 2.949307211623891e-06,
+      "loss": 0.5054,
+      "step": 5119
+    },
+    {
+      "epoch": 2.540151421124488,
+      "grad_norm": 0.07125903898864844,
+      "learning_rate": 2.947524637251782e-06,
+      "loss": 0.4689,
+      "step": 5120
+    },
+    {
+      "epoch": 2.5406478838277273,
+      "grad_norm": 0.07329623501191143,
+      "learning_rate": 2.945742376567623e-06,
+      "loss": 0.4722,
+      "step": 5121
+    },
+    {
+      "epoch": 2.5411443465309667,
+      "grad_norm": 0.07112535061046521,
+      "learning_rate": 2.943960429843804e-06,
+      "loss": 0.417,
+      "step": 5122
+    },
+    {
+      "epoch": 2.5416408092342064,
+      "grad_norm": 0.07322676770404164,
+      "learning_rate": 2.9421787973526694e-06,
+      "loss": 0.4622,
+      "step": 5123
+    },
+    {
+      "epoch": 2.5421372719374458,
+      "grad_norm": 0.07091383942713003,
+      "learning_rate": 2.940397479366509e-06,
+      "loss": 0.4664,
+      "step": 5124
+    },
+    {
+      "epoch": 2.542633734640685,
+      "grad_norm": 0.07158178260639342,
+      "learning_rate": 2.9386164761575716e-06,
+      "loss": 0.4424,
+      "step": 5125
+    },
+    {
+      "epoch": 2.5431301973439244,
+      "grad_norm": 0.07217103078505158,
+      "learning_rate": 2.936835787998053e-06,
+      "loss": 0.4537,
+      "step": 5126
+    },
+    {
+      "epoch": 2.543626660047164,
+      "grad_norm": 0.07171812937440548,
+      "learning_rate": 2.935055415160104e-06,
+      "loss": 0.4806,
+      "step": 5127
+    },
+    {
+      "epoch": 2.5441231227504035,
+      "grad_norm": 0.07193834390014926,
+      "learning_rate": 2.933275357915826e-06,
+      "loss": 0.4415,
+      "step": 5128
+    },
+    {
+      "epoch": 2.544619585453643,
+      "grad_norm": 0.07409938519274292,
+      "learning_rate": 2.9314956165372726e-06,
+      "loss": 0.487,
+      "step": 5129
+    },
+    {
+      "epoch": 2.545116048156882,
+      "grad_norm": 0.07535215883909349,
+      "learning_rate": 2.9297161912964476e-06,
+      "loss": 0.4376,
+      "step": 5130
+    },
+    {
+      "epoch": 2.5456125108601215,
+      "grad_norm": 0.0740990586903259,
+      "learning_rate": 2.9279370824653087e-06,
+      "loss": 0.4427,
+      "step": 5131
+    },
+    {
+      "epoch": 2.546108973563361,
+      "grad_norm": 0.07393544908024881,
+      "learning_rate": 2.926158290315764e-06,
+      "loss": 0.4564,
+      "step": 5132
+    },
+    {
+      "epoch": 2.5466054362666006,
+      "grad_norm": 0.07023920417474953,
+      "learning_rate": 2.924379815119672e-06,
+      "loss": 0.4595,
+      "step": 5133
+    },
+    {
+      "epoch": 2.54710189896984,
+      "grad_norm": 0.07061916856504286,
+      "learning_rate": 2.9226016571488467e-06,
+      "loss": 0.4635,
+      "step": 5134
+    },
+    {
+      "epoch": 2.5475983616730793,
+      "grad_norm": 0.07356084383044202,
+      "learning_rate": 2.9208238166750485e-06,
+      "loss": 0.4317,
+      "step": 5135
+    },
+    {
+      "epoch": 2.5480948243763186,
+      "grad_norm": 0.07020373559944719,
+      "learning_rate": 2.9190462939699925e-06,
+      "loss": 0.4249,
+      "step": 5136
+    },
+    {
+      "epoch": 2.5485912870795584,
+      "grad_norm": 0.07407304637604707,
+      "learning_rate": 2.917269089305347e-06,
+      "loss": 0.4646,
+      "step": 5137
+    },
+    {
+      "epoch": 2.5490877497827977,
+      "grad_norm": 0.07443631533106954,
+      "learning_rate": 2.915492202952724e-06,
+      "loss": 0.4656,
+      "step": 5138
+    },
+    {
+      "epoch": 2.549584212486037,
+      "grad_norm": 0.07372474395212635,
+      "learning_rate": 2.9137156351837005e-06,
+      "loss": 0.4526,
+      "step": 5139
+    },
+    {
+      "epoch": 2.5500806751892764,
+      "grad_norm": 0.07683958008697943,
+      "learning_rate": 2.911939386269786e-06,
+      "loss": 0.4832,
+      "step": 5140
+    },
+    {
+      "epoch": 2.5505771378925157,
+      "grad_norm": 0.07350654884183572,
+      "learning_rate": 2.9101634564824586e-06,
+      "loss": 0.5004,
+      "step": 5141
+    },
+    {
+      "epoch": 2.551073600595755,
+      "grad_norm": 0.07041706999389266,
+      "learning_rate": 2.908387846093138e-06,
+      "loss": 0.4626,
+      "step": 5142
+    },
+    {
+      "epoch": 2.551570063298995,
+      "grad_norm": 0.07167233151165785,
+      "learning_rate": 2.9066125553732003e-06,
+      "loss": 0.4389,
+      "step": 5143
+    },
+    {
+      "epoch": 2.552066526002234,
+      "grad_norm": 0.07262148199601952,
+      "learning_rate": 2.904837584593968e-06,
+      "loss": 0.4583,
+      "step": 5144
+    },
+    {
+      "epoch": 2.5525629887054735,
+      "grad_norm": 0.07309536918324455,
+      "learning_rate": 2.9030629340267165e-06,
+      "loss": 0.4663,
+      "step": 5145
+    },
+    {
+      "epoch": 2.553059451408713,
+      "grad_norm": 0.0719409786309271,
+      "learning_rate": 2.9012886039426747e-06,
+      "loss": 0.4438,
+      "step": 5146
+    },
+    {
+      "epoch": 2.5535559141119526,
+      "grad_norm": 0.07427692712025832,
+      "learning_rate": 2.8995145946130182e-06,
+      "loss": 0.4592,
+      "step": 5147
+    },
+    {
+      "epoch": 2.554052376815192,
+      "grad_norm": 0.07254799565282642,
+      "learning_rate": 2.897740906308879e-06,
+      "loss": 0.4138,
+      "step": 5148
+    },
+    {
+      "epoch": 2.5545488395184313,
+      "grad_norm": 0.07232628351087324,
+      "learning_rate": 2.8959675393013353e-06,
+      "loss": 0.4817,
+      "step": 5149
+    },
+    {
+      "epoch": 2.5550453022216706,
+      "grad_norm": 0.07168098229006624,
+      "learning_rate": 2.894194493861415e-06,
+      "loss": 0.4674,
+      "step": 5150
+    },
+    {
+      "epoch": 2.55554176492491,
+      "grad_norm": 0.07361419756507165,
+      "learning_rate": 2.8924217702601048e-06,
+      "loss": 0.4692,
+      "step": 5151
+    },
+    {
+      "epoch": 2.5560382276281493,
+      "grad_norm": 0.07536609973844947,
+      "learning_rate": 2.8906493687683324e-06,
+      "loss": 0.4618,
+      "step": 5152
+    },
+    {
+      "epoch": 2.5565346903313886,
+      "grad_norm": 0.07145853364222705,
+      "learning_rate": 2.888877289656985e-06,
+      "loss": 0.424,
+      "step": 5153
+    },
+    {
+      "epoch": 2.5570311530346284,
+      "grad_norm": 0.07113786320279357,
+      "learning_rate": 2.887105533196895e-06,
+      "loss": 0.4493,
+      "step": 5154
+    },
+    {
+      "epoch": 2.5575276157378677,
+      "grad_norm": 0.07575964297661865,
+      "learning_rate": 2.885334099658844e-06,
+      "loss": 0.4769,
+      "step": 5155
+    },
+    {
+      "epoch": 2.558024078441107,
+      "grad_norm": 0.07264729662949794,
+      "learning_rate": 2.8835629893135747e-06,
+      "loss": 0.4821,
+      "step": 5156
+    },
+    {
+      "epoch": 2.558520541144347,
+      "grad_norm": 0.07052412346155278,
+      "learning_rate": 2.8817922024317636e-06,
+      "loss": 0.4328,
+      "step": 5157
+    },
+    {
+      "epoch": 2.559017003847586,
+      "grad_norm": 0.07054010099118446,
+      "learning_rate": 2.880021739284053e-06,
+      "loss": 0.4454,
+      "step": 5158
+    },
+    {
+      "epoch": 2.5595134665508255,
+      "grad_norm": 0.07035637400898442,
+      "learning_rate": 2.8782516001410287e-06,
+      "loss": 0.4498,
+      "step": 5159
+    },
+    {
+      "epoch": 2.560009929254065,
+      "grad_norm": 0.07225924124383622,
+      "learning_rate": 2.876481785273225e-06,
+      "loss": 0.4541,
+      "step": 5160
+    },
+    {
+      "epoch": 2.560506391957304,
+      "grad_norm": 0.0699158749891402,
+      "learning_rate": 2.874712294951135e-06,
+      "loss": 0.426,
+      "step": 5161
+    },
+    {
+      "epoch": 2.5610028546605434,
+      "grad_norm": 0.07178418597986615,
+      "learning_rate": 2.8729431294451926e-06,
+      "loss": 0.4484,
+      "step": 5162
+    },
+    {
+      "epoch": 2.561499317363783,
+      "grad_norm": 0.07431620471359664,
+      "learning_rate": 2.87117428902579e-06,
+      "loss": 0.4625,
+      "step": 5163
+    },
+    {
+      "epoch": 2.5619957800670226,
+      "grad_norm": 0.07142951600063215,
+      "learning_rate": 2.869405773963264e-06,
+      "loss": 0.4613,
+      "step": 5164
+    },
+    {
+      "epoch": 2.562492242770262,
+      "grad_norm": 0.07148690845215873,
+      "learning_rate": 2.8676375845279013e-06,
+      "loss": 0.4236,
+      "step": 5165
+    },
+    {
+      "epoch": 2.562988705473501,
+      "grad_norm": 0.0706622462248769,
+      "learning_rate": 2.8658697209899467e-06,
+      "loss": 0.4556,
+      "step": 5166
+    },
+    {
+      "epoch": 2.563485168176741,
+      "grad_norm": 0.07418442942380814,
+      "learning_rate": 2.8641021836195853e-06,
+      "loss": 0.4574,
+      "step": 5167
+    },
+    {
+      "epoch": 2.5639816308799803,
+      "grad_norm": 0.07327595477365036,
+      "learning_rate": 2.8623349726869606e-06,
+      "loss": 0.4472,
+      "step": 5168
+    },
+    {
+      "epoch": 2.5644780935832197,
+      "grad_norm": 0.07382844367057993,
+      "learning_rate": 2.86056808846216e-06,
+      "loss": 0.4572,
+      "step": 5169
+    },
+    {
+      "epoch": 2.564974556286459,
+      "grad_norm": 0.07016839471992978,
+      "learning_rate": 2.8588015312152215e-06,
+      "loss": 0.4435,
+      "step": 5170
+    },
+    {
+      "epoch": 2.5654710189896983,
+      "grad_norm": 0.07083415871411265,
+      "learning_rate": 2.85703530121614e-06,
+      "loss": 0.4339,
+      "step": 5171
+    },
+    {
+      "epoch": 2.5659674816929376,
+      "grad_norm": 0.0714983644379608,
+      "learning_rate": 2.8552693987348533e-06,
+      "loss": 0.4402,
+      "step": 5172
+    },
+    {
+      "epoch": 2.566463944396177,
+      "grad_norm": 0.07685230780477147,
+      "learning_rate": 2.8535038240412503e-06,
+      "loss": 0.4853,
+      "step": 5173
+    },
+    {
+      "epoch": 2.5669604070994168,
+      "grad_norm": 0.07383361729449742,
+      "learning_rate": 2.85173857740517e-06,
+      "loss": 0.4739,
+      "step": 5174
+    },
+    {
+      "epoch": 2.567456869802656,
+      "grad_norm": 0.07358388092931081,
+      "learning_rate": 2.8499736590964043e-06,
+      "loss": 0.4462,
+      "step": 5175
+    },
+    {
+      "epoch": 2.5679533325058954,
+      "grad_norm": 0.07380900887304771,
+      "learning_rate": 2.8482090693846926e-06,
+      "loss": 0.492,
+      "step": 5176
+    },
+    {
+      "epoch": 2.568449795209135,
+      "grad_norm": 0.07617068375564527,
+      "learning_rate": 2.8464448085397212e-06,
+      "loss": 0.449,
+      "step": 5177
+    },
+    {
+      "epoch": 2.5689462579123745,
+      "grad_norm": 0.07348344159954795,
+      "learning_rate": 2.844680876831133e-06,
+      "loss": 0.4512,
+      "step": 5178
+    },
+    {
+      "epoch": 2.569442720615614,
+      "grad_norm": 0.07192285134930039,
+      "learning_rate": 2.8429172745285127e-06,
+      "loss": 0.4458,
+      "step": 5179
+    },
+    {
+      "epoch": 2.569939183318853,
+      "grad_norm": 0.0749431146540679,
+      "learning_rate": 2.8411540019014026e-06,
+      "loss": 0.5048,
+      "step": 5180
+    },
+    {
+      "epoch": 2.5704356460220925,
+      "grad_norm": 0.07407244688469924,
+      "learning_rate": 2.8393910592192898e-06,
+      "loss": 0.4891,
+      "step": 5181
+    },
+    {
+      "epoch": 2.570932108725332,
+      "grad_norm": 0.07213297514571944,
+      "learning_rate": 2.837628446751608e-06,
+      "loss": 0.4697,
+      "step": 5182
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.0699179683327418,
+      "learning_rate": 2.8358661647677497e-06,
+      "loss": 0.4355,
+      "step": 5183
+    },
+    {
+      "epoch": 2.571925034131811,
+      "grad_norm": 0.07765505093200081,
+      "learning_rate": 2.834104213537047e-06,
+      "loss": 0.4939,
+      "step": 5184
+    },
+    {
+      "epoch": 2.5724214968350503,
+      "grad_norm": 0.08172733691964389,
+      "learning_rate": 2.8323425933287883e-06,
+      "loss": 0.4986,
+      "step": 5185
+    },
+    {
+      "epoch": 2.5729179595382896,
+      "grad_norm": 0.07907186060467485,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.4528,
+      "step": 5186
+    },
+    {
+      "epoch": 2.573414422241529,
+      "grad_norm": 0.07227752346027236,
+      "learning_rate": 2.828820347056493e-06,
+      "loss": 0.4412,
+      "step": 5187
+    },
+    {
+      "epoch": 2.5739108849447687,
+      "grad_norm": 0.07282996319859109,
+      "learning_rate": 2.827059721530777e-06,
+      "loss": 0.4652,
+      "step": 5188
+    },
+    {
+      "epoch": 2.574407347648008,
+      "grad_norm": 0.07180952079077647,
+      "learning_rate": 2.8252994281041392e-06,
+      "loss": 0.4461,
+      "step": 5189
+    },
+    {
+      "epoch": 2.5749038103512474,
+      "grad_norm": 0.07342289227262047,
+      "learning_rate": 2.8235394670456164e-06,
+      "loss": 0.4255,
+      "step": 5190
+    },
+    {
+      "epoch": 2.5754002730544867,
+      "grad_norm": 0.07426475934548302,
+      "learning_rate": 2.821779838624188e-06,
+      "loss": 0.4461,
+      "step": 5191
+    },
+    {
+      "epoch": 2.575896735757726,
+      "grad_norm": 0.0724320917076261,
+      "learning_rate": 2.8200205431087868e-06,
+      "loss": 0.4669,
+      "step": 5192
+    },
+    {
+      "epoch": 2.5763931984609654,
+      "grad_norm": 0.07369682353577538,
+      "learning_rate": 2.8182615807682933e-06,
+      "loss": 0.4788,
+      "step": 5193
+    },
+    {
+      "epoch": 2.576889661164205,
+      "grad_norm": 0.07439988616448766,
+      "learning_rate": 2.8165029518715337e-06,
+      "loss": 0.4428,
+      "step": 5194
+    },
+    {
+      "epoch": 2.5773861238674445,
+      "grad_norm": 0.07255177311200152,
+      "learning_rate": 2.8147446566872894e-06,
+      "loss": 0.4723,
+      "step": 5195
+    },
+    {
+      "epoch": 2.577882586570684,
+      "grad_norm": 0.07390677701238471,
+      "learning_rate": 2.812986695484287e-06,
+      "loss": 0.4552,
+      "step": 5196
+    },
+    {
+      "epoch": 2.578379049273923,
+      "grad_norm": 0.07207378742571481,
+      "learning_rate": 2.8112290685312005e-06,
+      "loss": 0.4666,
+      "step": 5197
+    },
+    {
+      "epoch": 2.578875511977163,
+      "grad_norm": 0.07190111822941408,
+      "learning_rate": 2.8094717760966584e-06,
+      "loss": 0.427,
+      "step": 5198
+    },
+    {
+      "epoch": 2.5793719746804022,
+      "grad_norm": 0.07009977758785436,
+      "learning_rate": 2.80771481844923e-06,
+      "loss": 0.4448,
+      "step": 5199
+    },
+    {
+      "epoch": 2.5798684373836416,
+      "grad_norm": 0.0723752660969619,
+      "learning_rate": 2.8059581958574434e-06,
+      "loss": 0.4441,
+      "step": 5200
+    },
+    {
+      "epoch": 2.580364900086881,
+      "grad_norm": 0.0751303317079359,
+      "learning_rate": 2.804201908589768e-06,
+      "loss": 0.4811,
+      "step": 5201
+    },
+    {
+      "epoch": 2.5808613627901202,
+      "grad_norm": 0.07485829073586375,
+      "learning_rate": 2.8024459569146223e-06,
+      "loss": 0.4675,
+      "step": 5202
+    },
+    {
+      "epoch": 2.5813578254933596,
+      "grad_norm": 0.07364181111902501,
+      "learning_rate": 2.800690341100378e-06,
+      "loss": 0.4505,
+      "step": 5203
+    },
+    {
+      "epoch": 2.5818542881965993,
+      "grad_norm": 0.07357147393693832,
+      "learning_rate": 2.7989350614153532e-06,
+      "loss": 0.4468,
+      "step": 5204
+    },
+    {
+      "epoch": 2.5823507508998387,
+      "grad_norm": 0.07244406818148143,
+      "learning_rate": 2.7971801181278115e-06,
+      "loss": 0.4375,
+      "step": 5205
+    },
+    {
+      "epoch": 2.582847213603078,
+      "grad_norm": 0.07097629774725632,
+      "learning_rate": 2.795425511505968e-06,
+      "loss": 0.4429,
+      "step": 5206
+    },
+    {
+      "epoch": 2.5833436763063173,
+      "grad_norm": 0.07113223350037606,
+      "learning_rate": 2.793671241817989e-06,
+      "loss": 0.452,
+      "step": 5207
+    },
+    {
+      "epoch": 2.583840139009557,
+      "grad_norm": 0.06911189314520064,
+      "learning_rate": 2.791917309331985e-06,
+      "loss": 0.4245,
+      "step": 5208
+    },
+    {
+      "epoch": 2.5843366017127964,
+      "grad_norm": 0.07235990203667617,
+      "learning_rate": 2.7901637143160143e-06,
+      "loss": 0.4572,
+      "step": 5209
+    },
+    {
+      "epoch": 2.5848330644160358,
+      "grad_norm": 0.07286253379182837,
+      "learning_rate": 2.7884104570380906e-06,
+      "loss": 0.4801,
+      "step": 5210
+    },
+    {
+      "epoch": 2.585329527119275,
+      "grad_norm": 0.0729800350875364,
+      "learning_rate": 2.7866575377661654e-06,
+      "loss": 0.4913,
+      "step": 5211
+    },
+    {
+      "epoch": 2.5858259898225144,
+      "grad_norm": 0.0706958205684523,
+      "learning_rate": 2.7849049567681496e-06,
+      "loss": 0.4358,
+      "step": 5212
+    },
+    {
+      "epoch": 2.5863224525257538,
+      "grad_norm": 0.07357005050554628,
+      "learning_rate": 2.7831527143118954e-06,
+      "loss": 0.4699,
+      "step": 5213
+    },
+    {
+      "epoch": 2.5868189152289935,
+      "grad_norm": 0.07152756428761167,
+      "learning_rate": 2.781400810665201e-06,
+      "loss": 0.4756,
+      "step": 5214
+    },
+    {
+      "epoch": 2.587315377932233,
+      "grad_norm": 0.07534545721814909,
+      "learning_rate": 2.7796492460958237e-06,
+      "loss": 0.4545,
+      "step": 5215
+    },
+    {
+      "epoch": 2.587811840635472,
+      "grad_norm": 0.0717316509368772,
+      "learning_rate": 2.7778980208714556e-06,
+      "loss": 0.4268,
+      "step": 5216
+    },
+    {
+      "epoch": 2.5883083033387115,
+      "grad_norm": 0.07247442996879212,
+      "learning_rate": 2.7761471352597486e-06,
+      "loss": 0.4506,
+      "step": 5217
+    },
+    {
+      "epoch": 2.5888047660419513,
+      "grad_norm": 0.07364335888385559,
+      "learning_rate": 2.7743965895282956e-06,
+      "loss": 0.456,
+      "step": 5218
+    },
+    {
+      "epoch": 2.5893012287451906,
+      "grad_norm": 0.07489070055107547,
+      "learning_rate": 2.772646383944636e-06,
+      "loss": 0.4884,
+      "step": 5219
+    },
+    {
+      "epoch": 2.58979769144843,
+      "grad_norm": 0.07236646941891323,
+      "learning_rate": 2.7708965187762683e-06,
+      "loss": 0.4646,
+      "step": 5220
+    },
+    {
+      "epoch": 2.5902941541516693,
+      "grad_norm": 0.0726014680663615,
+      "learning_rate": 2.769146994290623e-06,
+      "loss": 0.4401,
+      "step": 5221
+    },
+    {
+      "epoch": 2.5907906168549086,
+      "grad_norm": 0.07146054387083108,
+      "learning_rate": 2.7673978107550925e-06,
+      "loss": 0.49,
+      "step": 5222
+    },
+    {
+      "epoch": 2.591287079558148,
+      "grad_norm": 0.07183695866470052,
+      "learning_rate": 2.7656489684370068e-06,
+      "loss": 0.465,
+      "step": 5223
+    },
+    {
+      "epoch": 2.5917835422613877,
+      "grad_norm": 0.07319204412500135,
+      "learning_rate": 2.763900467603654e-06,
+      "loss": 0.4441,
+      "step": 5224
+    },
+    {
+      "epoch": 2.592280004964627,
+      "grad_norm": 0.07224797748649779,
+      "learning_rate": 2.7621523085222612e-06,
+      "loss": 0.4376,
+      "step": 5225
+    },
+    {
+      "epoch": 2.5927764676678664,
+      "grad_norm": 0.07293475246622837,
+      "learning_rate": 2.760404491460006e-06,
+      "loss": 0.4719,
+      "step": 5226
+    },
+    {
+      "epoch": 2.5932729303711057,
+      "grad_norm": 0.07161942837682458,
+      "learning_rate": 2.7586570166840154e-06,
+      "loss": 0.4566,
+      "step": 5227
+    },
+    {
+      "epoch": 2.5937693930743455,
+      "grad_norm": 0.07225675300278564,
+      "learning_rate": 2.7569098844613616e-06,
+      "loss": 0.4366,
+      "step": 5228
+    },
+    {
+      "epoch": 2.594265855777585,
+      "grad_norm": 0.06933097869876324,
+      "learning_rate": 2.7551630950590686e-06,
+      "loss": 0.44,
+      "step": 5229
+    },
+    {
+      "epoch": 2.594762318480824,
+      "grad_norm": 0.07459117930805313,
+      "learning_rate": 2.753416648744103e-06,
+      "loss": 0.4643,
+      "step": 5230
+    },
+    {
+      "epoch": 2.5952587811840635,
+      "grad_norm": 0.07090794968370165,
+      "learning_rate": 2.75167054578338e-06,
+      "loss": 0.4572,
+      "step": 5231
+    },
+    {
+      "epoch": 2.595755243887303,
+      "grad_norm": 0.06958101880273615,
+      "learning_rate": 2.749924786443766e-06,
+      "loss": 0.4524,
+      "step": 5232
+    },
+    {
+      "epoch": 2.596251706590542,
+      "grad_norm": 0.0709070300995474,
+      "learning_rate": 2.7481793709920722e-06,
+      "loss": 0.4306,
+      "step": 5233
+    },
+    {
+      "epoch": 2.596748169293782,
+      "grad_norm": 0.07236911324643573,
+      "learning_rate": 2.7464342996950537e-06,
+      "loss": 0.452,
+      "step": 5234
+    },
+    {
+      "epoch": 2.5972446319970213,
+      "grad_norm": 0.07405711913344337,
+      "learning_rate": 2.744689572819421e-06,
+      "loss": 0.4733,
+      "step": 5235
+    },
+    {
+      "epoch": 2.5977410947002606,
+      "grad_norm": 0.07253420026733598,
+      "learning_rate": 2.742945190631827e-06,
+      "loss": 0.4638,
+      "step": 5236
+    },
+    {
+      "epoch": 2.5982375574035,
+      "grad_norm": 0.07081417321177245,
+      "learning_rate": 2.7412011533988707e-06,
+      "loss": 0.4249,
+      "step": 5237
+    },
+    {
+      "epoch": 2.5987340201067397,
+      "grad_norm": 0.07403116256998661,
+      "learning_rate": 2.7394574613870995e-06,
+      "loss": 0.4637,
+      "step": 5238
+    },
+    {
+      "epoch": 2.599230482809979,
+      "grad_norm": 0.07481669438242297,
+      "learning_rate": 2.7377141148630116e-06,
+      "loss": 0.4659,
+      "step": 5239
+    },
+    {
+      "epoch": 2.5997269455132184,
+      "grad_norm": 0.07377666582825643,
+      "learning_rate": 2.735971114093049e-06,
+      "loss": 0.4768,
+      "step": 5240
+    },
+    {
+      "epoch": 2.6002234082164577,
+      "grad_norm": 0.06995115542811843,
+      "learning_rate": 2.734228459343598e-06,
+      "loss": 0.4562,
+      "step": 5241
+    },
+    {
+      "epoch": 2.600719870919697,
+      "grad_norm": 0.0710589217991799,
+      "learning_rate": 2.7324861508810007e-06,
+      "loss": 0.4635,
+      "step": 5242
+    },
+    {
+      "epoch": 2.6012163336229364,
+      "grad_norm": 0.07548066876254467,
+      "learning_rate": 2.730744188971536e-06,
+      "loss": 0.4843,
+      "step": 5243
+    },
+    {
+      "epoch": 2.601712796326176,
+      "grad_norm": 0.07286147196013953,
+      "learning_rate": 2.72900257388144e-06,
+      "loss": 0.4366,
+      "step": 5244
+    },
+    {
+      "epoch": 2.6022092590294155,
+      "grad_norm": 0.07419676476869297,
+      "learning_rate": 2.7272613058768865e-06,
+      "loss": 0.4753,
+      "step": 5245
+    },
+    {
+      "epoch": 2.602705721732655,
+      "grad_norm": 0.07254724545442845,
+      "learning_rate": 2.725520385224001e-06,
+      "loss": 0.436,
+      "step": 5246
+    },
+    {
+      "epoch": 2.603202184435894,
+      "grad_norm": 0.07388891093945382,
+      "learning_rate": 2.723779812188857e-06,
+      "loss": 0.4783,
+      "step": 5247
+    },
+    {
+      "epoch": 2.603698647139134,
+      "grad_norm": 0.07329201727578473,
+      "learning_rate": 2.7220395870374715e-06,
+      "loss": 0.4403,
+      "step": 5248
+    },
+    {
+      "epoch": 2.6041951098423732,
+      "grad_norm": 0.07432559254448866,
+      "learning_rate": 2.7202997100358117e-06,
+      "loss": 0.482,
+      "step": 5249
+    },
+    {
+      "epoch": 2.6046915725456126,
+      "grad_norm": 0.07208369927428669,
+      "learning_rate": 2.7185601814497897e-06,
+      "loss": 0.4623,
+      "step": 5250
+    },
+    {
+      "epoch": 2.605188035248852,
+      "grad_norm": 0.07212394690380779,
+      "learning_rate": 2.7168210015452625e-06,
+      "loss": 0.4224,
+      "step": 5251
+    },
+    {
+      "epoch": 2.6056844979520912,
+      "grad_norm": 0.07141617635703298,
+      "learning_rate": 2.7150821705880403e-06,
+      "loss": 0.4435,
+      "step": 5252
+    },
+    {
+      "epoch": 2.6061809606553306,
+      "grad_norm": 0.07306470846553159,
+      "learning_rate": 2.7133436888438684e-06,
+      "loss": 0.4748,
+      "step": 5253
+    },
+    {
+      "epoch": 2.6066774233585703,
+      "grad_norm": 0.07049611258940773,
+      "learning_rate": 2.711605556578452e-06,
+      "loss": 0.4371,
+      "step": 5254
+    },
+    {
+      "epoch": 2.6071738860618097,
+      "grad_norm": 0.07231778349554253,
+      "learning_rate": 2.709867774057433e-06,
+      "loss": 0.4408,
+      "step": 5255
+    },
+    {
+      "epoch": 2.607670348765049,
+      "grad_norm": 0.07306994518093786,
+      "learning_rate": 2.708130341546407e-06,
+      "loss": 0.4745,
+      "step": 5256
+    },
+    {
+      "epoch": 2.6081668114682883,
+      "grad_norm": 0.07139666450342631,
+      "learning_rate": 2.706393259310911e-06,
+      "loss": 0.4202,
+      "step": 5257
+    },
+    {
+      "epoch": 2.608663274171528,
+      "grad_norm": 0.07431580209413276,
+      "learning_rate": 2.7046565276164283e-06,
+      "loss": 0.4874,
+      "step": 5258
+    },
+    {
+      "epoch": 2.6091597368747674,
+      "grad_norm": 0.07024347861273901,
+      "learning_rate": 2.7029201467283937e-06,
+      "loss": 0.4098,
+      "step": 5259
+    },
+    {
+      "epoch": 2.6096561995780068,
+      "grad_norm": 0.07203456244316435,
+      "learning_rate": 2.7011841169121825e-06,
+      "loss": 0.4504,
+      "step": 5260
+    },
+    {
+      "epoch": 2.610152662281246,
+      "grad_norm": 0.07237420715731252,
+      "learning_rate": 2.699448438433122e-06,
+      "loss": 0.4364,
+      "step": 5261
+    },
+    {
+      "epoch": 2.6106491249844854,
+      "grad_norm": 0.07079548912149816,
+      "learning_rate": 2.6977131115564814e-06,
+      "loss": 0.4437,
+      "step": 5262
+    },
+    {
+      "epoch": 2.6111455876877248,
+      "grad_norm": 0.07229082269715763,
+      "learning_rate": 2.695978136547476e-06,
+      "loss": 0.4594,
+      "step": 5263
+    },
+    {
+      "epoch": 2.6116420503909645,
+      "grad_norm": 0.0728641580489102,
+      "learning_rate": 2.694243513671271e-06,
+      "loss": 0.4638,
+      "step": 5264
+    },
+    {
+      "epoch": 2.612138513094204,
+      "grad_norm": 0.06905678657408329,
+      "learning_rate": 2.6925092431929734e-06,
+      "loss": 0.4335,
+      "step": 5265
+    },
+    {
+      "epoch": 2.612634975797443,
+      "grad_norm": 0.07285435311798948,
+      "learning_rate": 2.690775325377642e-06,
+      "loss": 0.4708,
+      "step": 5266
+    },
+    {
+      "epoch": 2.6131314385006825,
+      "grad_norm": 0.07043222760531233,
+      "learning_rate": 2.6890417604902765e-06,
+      "loss": 0.4315,
+      "step": 5267
+    },
+    {
+      "epoch": 2.6136279012039223,
+      "grad_norm": 0.0746756557661863,
+      "learning_rate": 2.687308548795825e-06,
+      "loss": 0.4816,
+      "step": 5268
+    },
+    {
+      "epoch": 2.6141243639071616,
+      "grad_norm": 0.07081456837335927,
+      "learning_rate": 2.68557569055918e-06,
+      "loss": 0.4362,
+      "step": 5269
+    },
+    {
+      "epoch": 2.614620826610401,
+      "grad_norm": 0.07330130227050775,
+      "learning_rate": 2.6838431860451797e-06,
+      "loss": 0.4465,
+      "step": 5270
+    },
+    {
+      "epoch": 2.6151172893136403,
+      "grad_norm": 0.0716011816724277,
+      "learning_rate": 2.682111035518614e-06,
+      "loss": 0.4539,
+      "step": 5271
+    },
+    {
+      "epoch": 2.6156137520168796,
+      "grad_norm": 0.06965330881909364,
+      "learning_rate": 2.6803792392442123e-06,
+      "loss": 0.4139,
+      "step": 5272
+    },
+    {
+      "epoch": 2.616110214720119,
+      "grad_norm": 0.07124158785595132,
+      "learning_rate": 2.6786477974866494e-06,
+      "loss": 0.437,
+      "step": 5273
+    },
+    {
+      "epoch": 2.6166066774233587,
+      "grad_norm": 0.0723706784290423,
+      "learning_rate": 2.676916710510552e-06,
+      "loss": 0.4659,
+      "step": 5274
+    },
+    {
+      "epoch": 2.617103140126598,
+      "grad_norm": 0.06963126779958274,
+      "learning_rate": 2.675185978580487e-06,
+      "loss": 0.4257,
+      "step": 5275
+    },
+    {
+      "epoch": 2.6175996028298374,
+      "grad_norm": 0.07338333374455405,
+      "learning_rate": 2.6734556019609704e-06,
+      "loss": 0.4481,
+      "step": 5276
+    },
+    {
+      "epoch": 2.6180960655330767,
+      "grad_norm": 0.07251896116734101,
+      "learning_rate": 2.6717255809164615e-06,
+      "loss": 0.4472,
+      "step": 5277
+    },
+    {
+      "epoch": 2.6185925282363165,
+      "grad_norm": 0.07067789143415738,
+      "learning_rate": 2.6699959157113653e-06,
+      "loss": 0.4428,
+      "step": 5278
+    },
+    {
+      "epoch": 2.619088990939556,
+      "grad_norm": 0.07525851620413397,
+      "learning_rate": 2.668266606610036e-06,
+      "loss": 0.4604,
+      "step": 5279
+    },
+    {
+      "epoch": 2.619585453642795,
+      "grad_norm": 0.07533316106282821,
+      "learning_rate": 2.6665376538767684e-06,
+      "loss": 0.4754,
+      "step": 5280
+    },
+    {
+      "epoch": 2.6200819163460345,
+      "grad_norm": 0.07008424279012973,
+      "learning_rate": 2.664809057775807e-06,
+      "loss": 0.4828,
+      "step": 5281
+    },
+    {
+      "epoch": 2.620578379049274,
+      "grad_norm": 0.06972061411105103,
+      "learning_rate": 2.66308081857134e-06,
+      "loss": 0.4342,
+      "step": 5282
+    },
+    {
+      "epoch": 2.621074841752513,
+      "grad_norm": 0.07186982715753283,
+      "learning_rate": 2.6613529365274974e-06,
+      "loss": 0.4165,
+      "step": 5283
+    },
+    {
+      "epoch": 2.621571304455753,
+      "grad_norm": 0.07285331263904996,
+      "learning_rate": 2.659625411908366e-06,
+      "loss": 0.4799,
+      "step": 5284
+    },
+    {
+      "epoch": 2.6220677671589923,
+      "grad_norm": 0.07501163105738619,
+      "learning_rate": 2.657898244977961e-06,
+      "loss": 0.4717,
+      "step": 5285
+    },
+    {
+      "epoch": 2.6225642298622316,
+      "grad_norm": 0.07160123786520918,
+      "learning_rate": 2.656171436000258e-06,
+      "loss": 0.4443,
+      "step": 5286
+    },
+    {
+      "epoch": 2.623060692565471,
+      "grad_norm": 0.06884269309644556,
+      "learning_rate": 2.6544449852391695e-06,
+      "loss": 0.4373,
+      "step": 5287
+    },
+    {
+      "epoch": 2.6235571552687107,
+      "grad_norm": 0.07252049270899412,
+      "learning_rate": 2.652718892958558e-06,
+      "loss": 0.4457,
+      "step": 5288
+    },
+    {
+      "epoch": 2.62405361797195,
+      "grad_norm": 0.07301113966245568,
+      "learning_rate": 2.650993159422228e-06,
+      "loss": 0.4817,
+      "step": 5289
+    },
+    {
+      "epoch": 2.6245500806751894,
+      "grad_norm": 0.07392352396215351,
+      "learning_rate": 2.649267784893929e-06,
+      "loss": 0.4418,
+      "step": 5290
+    },
+    {
+      "epoch": 2.6250465433784287,
+      "grad_norm": 0.07452490832212791,
+      "learning_rate": 2.6475427696373598e-06,
+      "loss": 0.4644,
+      "step": 5291
+    },
+    {
+      "epoch": 2.625543006081668,
+      "grad_norm": 0.07118282301353918,
+      "learning_rate": 2.6458181139161564e-06,
+      "loss": 0.463,
+      "step": 5292
+    },
+    {
+      "epoch": 2.6260394687849073,
+      "grad_norm": 0.07152366375897204,
+      "learning_rate": 2.644093817993911e-06,
+      "loss": 0.4285,
+      "step": 5293
+    },
+    {
+      "epoch": 2.6265359314881467,
+      "grad_norm": 0.07181298428188997,
+      "learning_rate": 2.642369882134151e-06,
+      "loss": 0.4416,
+      "step": 5294
+    },
+    {
+      "epoch": 2.6270323941913865,
+      "grad_norm": 0.0707996904302843,
+      "learning_rate": 2.6406463066003505e-06,
+      "loss": 0.4639,
+      "step": 5295
+    },
+    {
+      "epoch": 2.627528856894626,
+      "grad_norm": 0.07691379853284944,
+      "learning_rate": 2.638923091655935e-06,
+      "loss": 0.4563,
+      "step": 5296
+    },
+    {
+      "epoch": 2.628025319597865,
+      "grad_norm": 0.07191388401411938,
+      "learning_rate": 2.6372002375642657e-06,
+      "loss": 0.483,
+      "step": 5297
+    },
+    {
+      "epoch": 2.628521782301105,
+      "grad_norm": 0.07283482258711875,
+      "learning_rate": 2.635477744588658e-06,
+      "loss": 0.4473,
+      "step": 5298
+    },
+    {
+      "epoch": 2.6290182450043442,
+      "grad_norm": 0.07144510285173247,
+      "learning_rate": 2.6337556129923648e-06,
+      "loss": 0.4561,
+      "step": 5299
+    },
+    {
+      "epoch": 2.6295147077075836,
+      "grad_norm": 0.07225771158425691,
+      "learning_rate": 2.6320338430385857e-06,
+      "loss": 0.4326,
+      "step": 5300
+    },
+    {
+      "epoch": 2.630011170410823,
+      "grad_norm": 0.07092103282667439,
+      "learning_rate": 2.630312434990466e-06,
+      "loss": 0.4432,
+      "step": 5301
+    },
+    {
+      "epoch": 2.630507633114062,
+      "grad_norm": 0.0701907480460393,
+      "learning_rate": 2.628591389111095e-06,
+      "loss": 0.4169,
+      "step": 5302
+    },
+    {
+      "epoch": 2.6310040958173015,
+      "grad_norm": 0.07194177966806387,
+      "learning_rate": 2.6268707056635077e-06,
+      "loss": 0.4447,
+      "step": 5303
+    },
+    {
+      "epoch": 2.631500558520541,
+      "grad_norm": 0.07281610835524266,
+      "learning_rate": 2.625150384910682e-06,
+      "loss": 0.4744,
+      "step": 5304
+    },
+    {
+      "epoch": 2.6319970212237807,
+      "grad_norm": 0.07556119554107671,
+      "learning_rate": 2.6234304271155443e-06,
+      "loss": 0.4868,
+      "step": 5305
+    },
+    {
+      "epoch": 2.63249348392702,
+      "grad_norm": 0.07372561099256456,
+      "learning_rate": 2.6217108325409594e-06,
+      "loss": 0.4733,
+      "step": 5306
+    },
+    {
+      "epoch": 2.6329899466302593,
+      "grad_norm": 0.07215826218897892,
+      "learning_rate": 2.6199916014497396e-06,
+      "loss": 0.4485,
+      "step": 5307
+    },
+    {
+      "epoch": 2.633486409333499,
+      "grad_norm": 0.07287009510752426,
+      "learning_rate": 2.618272734104645e-06,
+      "loss": 0.4489,
+      "step": 5308
+    },
+    {
+      "epoch": 2.6339828720367384,
+      "grad_norm": 0.07013210868053496,
+      "learning_rate": 2.6165542307683744e-06,
+      "loss": 0.4987,
+      "step": 5309
+    },
+    {
+      "epoch": 2.6344793347399778,
+      "grad_norm": 0.07198909527337301,
+      "learning_rate": 2.614836091703572e-06,
+      "loss": 0.4406,
+      "step": 5310
+    },
+    {
+      "epoch": 2.634975797443217,
+      "grad_norm": 0.07119081849050292,
+      "learning_rate": 2.6131183171728323e-06,
+      "loss": 0.4628,
+      "step": 5311
+    },
+    {
+      "epoch": 2.6354722601464564,
+      "grad_norm": 0.07010495374139804,
+      "learning_rate": 2.611400907438685e-06,
+      "loss": 0.4388,
+      "step": 5312
+    },
+    {
+      "epoch": 2.6359687228496957,
+      "grad_norm": 0.07035383373874375,
+      "learning_rate": 2.6096838627636124e-06,
+      "loss": 0.451,
+      "step": 5313
+    },
+    {
+      "epoch": 2.636465185552935,
+      "grad_norm": 0.07252865320160067,
+      "learning_rate": 2.6079671834100354e-06,
+      "loss": 0.4461,
+      "step": 5314
+    },
+    {
+      "epoch": 2.636961648256175,
+      "grad_norm": 0.07337124382294172,
+      "learning_rate": 2.60625086964032e-06,
+      "loss": 0.4631,
+      "step": 5315
+    },
+    {
+      "epoch": 2.637458110959414,
+      "grad_norm": 0.07423342378313234,
+      "learning_rate": 2.6045349217167815e-06,
+      "loss": 0.4961,
+      "step": 5316
+    },
+    {
+      "epoch": 2.6379545736626535,
+      "grad_norm": 0.07250274236591756,
+      "learning_rate": 2.6028193399016677e-06,
+      "loss": 0.4608,
+      "step": 5317
+    },
+    {
+      "epoch": 2.6384510363658933,
+      "grad_norm": 0.07249481412053288,
+      "learning_rate": 2.6011041244571844e-06,
+      "loss": 0.4269,
+      "step": 5318
+    },
+    {
+      "epoch": 2.6389474990691326,
+      "grad_norm": 0.06832436591372723,
+      "learning_rate": 2.5993892756454702e-06,
+      "loss": 0.4358,
+      "step": 5319
+    },
+    {
+      "epoch": 2.639443961772372,
+      "grad_norm": 0.07458496665240169,
+      "learning_rate": 2.597674793728616e-06,
+      "loss": 0.4899,
+      "step": 5320
+    },
+    {
+      "epoch": 2.6399404244756113,
+      "grad_norm": 0.07045333283556546,
+      "learning_rate": 2.595960678968652e-06,
+      "loss": 0.4481,
+      "step": 5321
+    },
+    {
+      "epoch": 2.6404368871788506,
+      "grad_norm": 0.07350913336828889,
+      "learning_rate": 2.59424693162755e-06,
+      "loss": 0.4553,
+      "step": 5322
+    },
+    {
+      "epoch": 2.64093334988209,
+      "grad_norm": 0.07608854710093517,
+      "learning_rate": 2.5925335519672333e-06,
+      "loss": 0.4687,
+      "step": 5323
+    },
+    {
+      "epoch": 2.6414298125853293,
+      "grad_norm": 0.06943444544262166,
+      "learning_rate": 2.5908205402495603e-06,
+      "loss": 0.4295,
+      "step": 5324
+    },
+    {
+      "epoch": 2.641926275288569,
+      "grad_norm": 0.0723083110709931,
+      "learning_rate": 2.5891078967363416e-06,
+      "loss": 0.4454,
+      "step": 5325
+    },
+    {
+      "epoch": 2.6424227379918084,
+      "grad_norm": 0.07429140563455172,
+      "learning_rate": 2.587395621689325e-06,
+      "loss": 0.4728,
+      "step": 5326
+    },
+    {
+      "epoch": 2.6429192006950477,
+      "grad_norm": 0.07137582757890498,
+      "learning_rate": 2.585683715370202e-06,
+      "loss": 0.4362,
+      "step": 5327
+    },
+    {
+      "epoch": 2.643415663398287,
+      "grad_norm": 0.07321803169563242,
+      "learning_rate": 2.5839721780406146e-06,
+      "loss": 0.4516,
+      "step": 5328
+    },
+    {
+      "epoch": 2.643912126101527,
+      "grad_norm": 0.0709602832592665,
+      "learning_rate": 2.5822610099621402e-06,
+      "loss": 0.4278,
+      "step": 5329
+    },
+    {
+      "epoch": 2.644408588804766,
+      "grad_norm": 0.07032601586797163,
+      "learning_rate": 2.5805502113963066e-06,
+      "loss": 0.4457,
+      "step": 5330
+    },
+    {
+      "epoch": 2.6449050515080055,
+      "grad_norm": 0.07003931206303413,
+      "learning_rate": 2.5788397826045807e-06,
+      "loss": 0.4407,
+      "step": 5331
+    },
+    {
+      "epoch": 2.645401514211245,
+      "grad_norm": 0.07218328091870577,
+      "learning_rate": 2.577129723848373e-06,
+      "loss": 0.4537,
+      "step": 5332
+    },
+    {
+      "epoch": 2.645897976914484,
+      "grad_norm": 0.0725217289580609,
+      "learning_rate": 2.57542003538904e-06,
+      "loss": 0.4565,
+      "step": 5333
+    },
+    {
+      "epoch": 2.6463944396177235,
+      "grad_norm": 0.07169708777564429,
+      "learning_rate": 2.5737107174878773e-06,
+      "loss": 0.4279,
+      "step": 5334
+    },
+    {
+      "epoch": 2.6468909023209632,
+      "grad_norm": 0.071467454593286,
+      "learning_rate": 2.5720017704061307e-06,
+      "loss": 0.4613,
+      "step": 5335
+    },
+    {
+      "epoch": 2.6473873650242026,
+      "grad_norm": 0.07225220351631556,
+      "learning_rate": 2.5702931944049816e-06,
+      "loss": 0.4536,
+      "step": 5336
+    },
+    {
+      "epoch": 2.647883827727442,
+      "grad_norm": 0.07089284434717677,
+      "learning_rate": 2.5685849897455617e-06,
+      "loss": 0.4357,
+      "step": 5337
+    },
+    {
+      "epoch": 2.6483802904306812,
+      "grad_norm": 0.0759453690608335,
+      "learning_rate": 2.5668771566889415e-06,
+      "loss": 0.4604,
+      "step": 5338
+    },
+    {
+      "epoch": 2.648876753133921,
+      "grad_norm": 0.07174183489715347,
+      "learning_rate": 2.565169695496134e-06,
+      "loss": 0.4383,
+      "step": 5339
+    },
+    {
+      "epoch": 2.6493732158371603,
+      "grad_norm": 0.07130619769369462,
+      "learning_rate": 2.563462606428101e-06,
+      "loss": 0.4544,
+      "step": 5340
+    },
+    {
+      "epoch": 2.6498696785403997,
+      "grad_norm": 0.07324873611482309,
+      "learning_rate": 2.5617558897457402e-06,
+      "loss": 0.4446,
+      "step": 5341
+    },
+    {
+      "epoch": 2.650366141243639,
+      "grad_norm": 0.07448900607353087,
+      "learning_rate": 2.5600495457098984e-06,
+      "loss": 0.453,
+      "step": 5342
+    },
+    {
+      "epoch": 2.6508626039468783,
+      "grad_norm": 0.07312292007238767,
+      "learning_rate": 2.5583435745813624e-06,
+      "loss": 0.4763,
+      "step": 5343
+    },
+    {
+      "epoch": 2.6513590666501177,
+      "grad_norm": 0.07299329196320681,
+      "learning_rate": 2.5566379766208602e-06,
+      "loss": 0.4385,
+      "step": 5344
+    },
+    {
+      "epoch": 2.6518555293533574,
+      "grad_norm": 0.07218798632859258,
+      "learning_rate": 2.5549327520890686e-06,
+      "loss": 0.4301,
+      "step": 5345
+    },
+    {
+      "epoch": 2.6523519920565968,
+      "grad_norm": 0.08102474836920227,
+      "learning_rate": 2.5532279012466025e-06,
+      "loss": 0.4467,
+      "step": 5346
+    },
+    {
+      "epoch": 2.652848454759836,
+      "grad_norm": 0.07313090190592135,
+      "learning_rate": 2.5515234243540186e-06,
+      "loss": 0.4355,
+      "step": 5347
+    },
+    {
+      "epoch": 2.6533449174630754,
+      "grad_norm": 0.07443189436453763,
+      "learning_rate": 2.549819321671825e-06,
+      "loss": 0.4887,
+      "step": 5348
+    },
+    {
+      "epoch": 2.653841380166315,
+      "grad_norm": 0.07160561955915772,
+      "learning_rate": 2.5481155934604585e-06,
+      "loss": 0.4512,
+      "step": 5349
+    },
+    {
+      "epoch": 2.6543378428695545,
+      "grad_norm": 0.07063791790927199,
+      "learning_rate": 2.5464122399803126e-06,
+      "loss": 0.4499,
+      "step": 5350
+    },
+    {
+      "epoch": 2.654834305572794,
+      "grad_norm": 0.0737680394360274,
+      "learning_rate": 2.5447092614917128e-06,
+      "loss": 0.4767,
+      "step": 5351
+    },
+    {
+      "epoch": 2.655330768276033,
+      "grad_norm": 0.07062001938640819,
+      "learning_rate": 2.5430066582549373e-06,
+      "loss": 0.4576,
+      "step": 5352
+    },
+    {
+      "epoch": 2.6558272309792725,
+      "grad_norm": 0.07382855500342596,
+      "learning_rate": 2.5413044305301993e-06,
+      "loss": 0.4667,
+      "step": 5353
+    },
+    {
+      "epoch": 2.656323693682512,
+      "grad_norm": 0.07193888010833445,
+      "learning_rate": 2.5396025785776545e-06,
+      "loss": 0.4515,
+      "step": 5354
+    },
+    {
+      "epoch": 2.6568201563857516,
+      "grad_norm": 0.07084732634769446,
+      "learning_rate": 2.5379011026574084e-06,
+      "loss": 0.4633,
+      "step": 5355
+    },
+    {
+      "epoch": 2.657316619088991,
+      "grad_norm": 0.071847207505773,
+      "learning_rate": 2.536200003029501e-06,
+      "loss": 0.4683,
+      "step": 5356
+    },
+    {
+      "epoch": 2.6578130817922303,
+      "grad_norm": 0.07187666934920998,
+      "learning_rate": 2.5344992799539193e-06,
+      "loss": 0.4625,
+      "step": 5357
+    },
+    {
+      "epoch": 2.6583095444954696,
+      "grad_norm": 0.07160584760597262,
+      "learning_rate": 2.5327989336905923e-06,
+      "loss": 0.4439,
+      "step": 5358
+    },
+    {
+      "epoch": 2.6588060071987094,
+      "grad_norm": 0.07208708458309498,
+      "learning_rate": 2.5310989644993876e-06,
+      "loss": 0.4971,
+      "step": 5359
+    },
+    {
+      "epoch": 2.6593024699019487,
+      "grad_norm": 0.07247788488979347,
+      "learning_rate": 2.5293993726401224e-06,
+      "loss": 0.4371,
+      "step": 5360
+    },
+    {
+      "epoch": 2.659798932605188,
+      "grad_norm": 0.07266963378543632,
+      "learning_rate": 2.527700158372548e-06,
+      "loss": 0.4542,
+      "step": 5361
+    },
+    {
+      "epoch": 2.6602953953084274,
+      "grad_norm": 0.0731112273892814,
+      "learning_rate": 2.5260013219563663e-06,
+      "loss": 0.4674,
+      "step": 5362
+    },
+    {
+      "epoch": 2.6607918580116667,
+      "grad_norm": 0.0728325130956322,
+      "learning_rate": 2.5243028636512146e-06,
+      "loss": 0.4788,
+      "step": 5363
+    },
+    {
+      "epoch": 2.661288320714906,
+      "grad_norm": 0.07168492677060803,
+      "learning_rate": 2.5226047837166757e-06,
+      "loss": 0.4502,
+      "step": 5364
+    },
+    {
+      "epoch": 2.661784783418146,
+      "grad_norm": 0.07164432402739039,
+      "learning_rate": 2.5209070824122733e-06,
+      "loss": 0.4511,
+      "step": 5365
+    },
+    {
+      "epoch": 2.662281246121385,
+      "grad_norm": 0.07246429922456672,
+      "learning_rate": 2.519209759997472e-06,
+      "loss": 0.4567,
+      "step": 5366
+    },
+    {
+      "epoch": 2.6627777088246245,
+      "grad_norm": 0.07313252363944685,
+      "learning_rate": 2.5175128167316848e-06,
+      "loss": 0.4645,
+      "step": 5367
+    },
+    {
+      "epoch": 2.663274171527864,
+      "grad_norm": 0.07246259249132068,
+      "learning_rate": 2.515816252874258e-06,
+      "loss": 0.468,
+      "step": 5368
+    },
+    {
+      "epoch": 2.6637706342311036,
+      "grad_norm": 0.07090541004756543,
+      "learning_rate": 2.514120068684488e-06,
+      "loss": 0.4514,
+      "step": 5369
+    },
+    {
+      "epoch": 2.664267096934343,
+      "grad_norm": 0.0702345762551842,
+      "learning_rate": 2.5124242644216066e-06,
+      "loss": 0.4468,
+      "step": 5370
+    },
+    {
+      "epoch": 2.6647635596375823,
+      "grad_norm": 0.07252724362767601,
+      "learning_rate": 2.5107288403447906e-06,
+      "loss": 0.4414,
+      "step": 5371
+    },
+    {
+      "epoch": 2.6652600223408216,
+      "grad_norm": 0.07389080736112623,
+      "learning_rate": 2.50903379671316e-06,
+      "loss": 0.4709,
+      "step": 5372
+    },
+    {
+      "epoch": 2.665756485044061,
+      "grad_norm": 0.07191391269126059,
+      "learning_rate": 2.5073391337857722e-06,
+      "loss": 0.4651,
+      "step": 5373
+    },
+    {
+      "epoch": 2.6662529477473003,
+      "grad_norm": 0.074543193027082,
+      "learning_rate": 2.505644851821633e-06,
+      "loss": 0.4822,
+      "step": 5374
+    },
+    {
+      "epoch": 2.66674941045054,
+      "grad_norm": 0.07255691804503774,
+      "learning_rate": 2.5039509510796843e-06,
+      "loss": 0.4574,
+      "step": 5375
+    },
+    {
+      "epoch": 2.6672458731537794,
+      "grad_norm": 0.07225389782589921,
+      "learning_rate": 2.50225743181881e-06,
+      "loss": 0.4651,
+      "step": 5376
+    },
+    {
+      "epoch": 2.6677423358570187,
+      "grad_norm": 0.07329724006761641,
+      "learning_rate": 2.50056429429784e-06,
+      "loss": 0.4445,
+      "step": 5377
+    },
+    {
+      "epoch": 2.668238798560258,
+      "grad_norm": 0.0704766031668799,
+      "learning_rate": 2.4988715387755415e-06,
+      "loss": 0.4381,
+      "step": 5378
+    },
+    {
+      "epoch": 2.668735261263498,
+      "grad_norm": 0.07337203960969434,
+      "learning_rate": 2.4971791655106263e-06,
+      "loss": 0.466,
+      "step": 5379
+    },
+    {
+      "epoch": 2.669231723966737,
+      "grad_norm": 0.07181577127392384,
+      "learning_rate": 2.4954871747617472e-06,
+      "loss": 0.4701,
+      "step": 5380
+    },
+    {
+      "epoch": 2.6697281866699765,
+      "grad_norm": 0.07358688312804215,
+      "learning_rate": 2.493795566787496e-06,
+      "loss": 0.4855,
+      "step": 5381
+    },
+    {
+      "epoch": 2.670224649373216,
+      "grad_norm": 0.07195123164241551,
+      "learning_rate": 2.4921043418464085e-06,
+      "loss": 0.4245,
+      "step": 5382
+    },
+    {
+      "epoch": 2.670721112076455,
+      "grad_norm": 0.0734381724103889,
+      "learning_rate": 2.4904135001969595e-06,
+      "loss": 0.451,
+      "step": 5383
+    },
+    {
+      "epoch": 2.6712175747796945,
+      "grad_norm": 0.07307924081621357,
+      "learning_rate": 2.4887230420975705e-06,
+      "loss": 0.4627,
+      "step": 5384
+    },
+    {
+      "epoch": 2.6717140374829342,
+      "grad_norm": 0.07196229922624982,
+      "learning_rate": 2.4870329678065997e-06,
+      "loss": 0.4464,
+      "step": 5385
+    },
+    {
+      "epoch": 2.6722105001861736,
+      "grad_norm": 0.07175602703078268,
+      "learning_rate": 2.4853432775823457e-06,
+      "loss": 0.4415,
+      "step": 5386
+    },
+    {
+      "epoch": 2.672706962889413,
+      "grad_norm": 0.07118605838847018,
+      "learning_rate": 2.4836539716830533e-06,
+      "loss": 0.4573,
+      "step": 5387
+    },
+    {
+      "epoch": 2.6732034255926522,
+      "grad_norm": 0.06966429924987237,
+      "learning_rate": 2.4819650503669035e-06,
+      "loss": 0.4597,
+      "step": 5388
+    },
+    {
+      "epoch": 2.673699888295892,
+      "grad_norm": 0.0715539394926452,
+      "learning_rate": 2.4802765138920236e-06,
+      "loss": 0.4165,
+      "step": 5389
+    },
+    {
+      "epoch": 2.6741963509991313,
+      "grad_norm": 0.07073444357197714,
+      "learning_rate": 2.478588362516478e-06,
+      "loss": 0.4528,
+      "step": 5390
+    },
+    {
+      "epoch": 2.6746928137023707,
+      "grad_norm": 0.0727857459617948,
+      "learning_rate": 2.4769005964982718e-06,
+      "loss": 0.4782,
+      "step": 5391
+    },
+    {
+      "epoch": 2.67518927640561,
+      "grad_norm": 0.07476771403087601,
+      "learning_rate": 2.475213216095356e-06,
+      "loss": 0.4949,
+      "step": 5392
+    },
+    {
+      "epoch": 2.6756857391088493,
+      "grad_norm": 0.0747474889538692,
+      "learning_rate": 2.473526221565617e-06,
+      "loss": 0.4991,
+      "step": 5393
+    },
+    {
+      "epoch": 2.6761822018120887,
+      "grad_norm": 0.072689971266639,
+      "learning_rate": 2.4718396131668877e-06,
+      "loss": 0.485,
+      "step": 5394
+    },
+    {
+      "epoch": 2.6766786645153284,
+      "grad_norm": 0.07221036618857145,
+      "learning_rate": 2.4701533911569375e-06,
+      "loss": 0.4839,
+      "step": 5395
+    },
+    {
+      "epoch": 2.6771751272185678,
+      "grad_norm": 0.07272155353905288,
+      "learning_rate": 2.4684675557934766e-06,
+      "loss": 0.434,
+      "step": 5396
+    },
+    {
+      "epoch": 2.677671589921807,
+      "grad_norm": 0.07156056790757943,
+      "learning_rate": 2.4667821073341636e-06,
+      "loss": 0.4596,
+      "step": 5397
+    },
+    {
+      "epoch": 2.6781680526250464,
+      "grad_norm": 0.07222255501510809,
+      "learning_rate": 2.4650970460365846e-06,
+      "loss": 0.439,
+      "step": 5398
+    },
+    {
+      "epoch": 2.678664515328286,
+      "grad_norm": 0.0757415187469245,
+      "learning_rate": 2.4634123721582804e-06,
+      "loss": 0.4942,
+      "step": 5399
+    },
+    {
+      "epoch": 2.6791609780315255,
+      "grad_norm": 0.07333904857592394,
+      "learning_rate": 2.461728085956722e-06,
+      "loss": 0.4576,
+      "step": 5400
+    },
+    {
+      "epoch": 2.679657440734765,
+      "grad_norm": 0.07286665550490455,
+      "learning_rate": 2.460044187689328e-06,
+      "loss": 0.4352,
+      "step": 5401
+    },
+    {
+      "epoch": 2.680153903438004,
+      "grad_norm": 0.07097365341929482,
+      "learning_rate": 2.458360677613457e-06,
+      "loss": 0.447,
+      "step": 5402
+    },
+    {
+      "epoch": 2.6806503661412435,
+      "grad_norm": 0.07406683798365062,
+      "learning_rate": 2.456677555986401e-06,
+      "loss": 0.4562,
+      "step": 5403
+    },
+    {
+      "epoch": 2.681146828844483,
+      "grad_norm": 0.0715410713991149,
+      "learning_rate": 2.4549948230654034e-06,
+      "loss": 0.4626,
+      "step": 5404
+    },
+    {
+      "epoch": 2.6816432915477226,
+      "grad_norm": 0.07133636588398236,
+      "learning_rate": 2.4533124791076396e-06,
+      "loss": 0.4677,
+      "step": 5405
+    },
+    {
+      "epoch": 2.682139754250962,
+      "grad_norm": 0.07290857535557271,
+      "learning_rate": 2.451630524370232e-06,
+      "loss": 0.4533,
+      "step": 5406
+    },
+    {
+      "epoch": 2.6826362169542013,
+      "grad_norm": 0.07127985534713971,
+      "learning_rate": 2.4499489591102395e-06,
+      "loss": 0.4893,
+      "step": 5407
+    },
+    {
+      "epoch": 2.6831326796574406,
+      "grad_norm": 0.07140839768994581,
+      "learning_rate": 2.448267783584659e-06,
+      "loss": 0.4489,
+      "step": 5408
+    },
+    {
+      "epoch": 2.6836291423606804,
+      "grad_norm": 0.07109980383685978,
+      "learning_rate": 2.446586998050436e-06,
+      "loss": 0.4484,
+      "step": 5409
+    },
+    {
+      "epoch": 2.6841256050639197,
+      "grad_norm": 0.07442132086297334,
+      "learning_rate": 2.4449066027644473e-06,
+      "loss": 0.485,
+      "step": 5410
+    },
+    {
+      "epoch": 2.684622067767159,
+      "grad_norm": 0.07286157873915275,
+      "learning_rate": 2.4432265979835183e-06,
+      "loss": 0.455,
+      "step": 5411
+    },
+    {
+      "epoch": 2.6851185304703984,
+      "grad_norm": 0.06958316341672992,
+      "learning_rate": 2.4415469839644094e-06,
+      "loss": 0.4441,
+      "step": 5412
+    },
+    {
+      "epoch": 2.6856149931736377,
+      "grad_norm": 0.07471228524829766,
+      "learning_rate": 2.4398677609638228e-06,
+      "loss": 0.4566,
+      "step": 5413
+    },
+    {
+      "epoch": 2.686111455876877,
+      "grad_norm": 0.07131159705902516,
+      "learning_rate": 2.4381889292383997e-06,
+      "loss": 0.4298,
+      "step": 5414
+    },
+    {
+      "epoch": 2.686607918580117,
+      "grad_norm": 0.0742376202400464,
+      "learning_rate": 2.4365104890447218e-06,
+      "loss": 0.4762,
+      "step": 5415
+    },
+    {
+      "epoch": 2.687104381283356,
+      "grad_norm": 0.07242206286086188,
+      "learning_rate": 2.434832440639315e-06,
+      "loss": 0.444,
+      "step": 5416
+    },
+    {
+      "epoch": 2.6876008439865955,
+      "grad_norm": 0.07382874049757732,
+      "learning_rate": 2.433154784278638e-06,
+      "loss": 0.5035,
+      "step": 5417
+    },
+    {
+      "epoch": 2.688097306689835,
+      "grad_norm": 0.07053618916196587,
+      "learning_rate": 2.4314775202190983e-06,
+      "loss": 0.4455,
+      "step": 5418
+    },
+    {
+      "epoch": 2.6885937693930746,
+      "grad_norm": 0.0731789208347564,
+      "learning_rate": 2.429800648717036e-06,
+      "loss": 0.4613,
+      "step": 5419
+    },
+    {
+      "epoch": 2.689090232096314,
+      "grad_norm": 0.07283239329695157,
+      "learning_rate": 2.4281241700287334e-06,
+      "loss": 0.4591,
+      "step": 5420
+    },
+    {
+      "epoch": 2.6895866947995533,
+      "grad_norm": 0.0725960512541698,
+      "learning_rate": 2.426448084410416e-06,
+      "loss": 0.4716,
+      "step": 5421
+    },
+    {
+      "epoch": 2.6900831575027926,
+      "grad_norm": 0.07360807080088544,
+      "learning_rate": 2.424772392118245e-06,
+      "loss": 0.4685,
+      "step": 5422
+    },
+    {
+      "epoch": 2.690579620206032,
+      "grad_norm": 0.07326003572886362,
+      "learning_rate": 2.4230970934083216e-06,
+      "loss": 0.4794,
+      "step": 5423
+    },
+    {
+      "epoch": 2.6910760829092712,
+      "grad_norm": 0.07094010720894862,
+      "learning_rate": 2.4214221885366918e-06,
+      "loss": 0.4601,
+      "step": 5424
+    },
+    {
+      "epoch": 2.691572545612511,
+      "grad_norm": 0.07394289944310807,
+      "learning_rate": 2.4197476777593336e-06,
+      "loss": 0.4576,
+      "step": 5425
+    },
+    {
+      "epoch": 2.6920690083157504,
+      "grad_norm": 0.0727641219406632,
+      "learning_rate": 2.4180735613321745e-06,
+      "loss": 0.4528,
+      "step": 5426
+    },
+    {
+      "epoch": 2.6925654710189897,
+      "grad_norm": 0.07508411585218325,
+      "learning_rate": 2.4163998395110732e-06,
+      "loss": 0.4533,
+      "step": 5427
+    },
+    {
+      "epoch": 2.693061933722229,
+      "grad_norm": 0.07197259385452393,
+      "learning_rate": 2.4147265125518292e-06,
+      "loss": 0.4684,
+      "step": 5428
+    },
+    {
+      "epoch": 2.693558396425469,
+      "grad_norm": 0.07182235273910088,
+      "learning_rate": 2.4130535807101905e-06,
+      "loss": 0.431,
+      "step": 5429
+    },
+    {
+      "epoch": 2.694054859128708,
+      "grad_norm": 0.07238059271016718,
+      "learning_rate": 2.4113810442418293e-06,
+      "loss": 0.4487,
+      "step": 5430
+    },
+    {
+      "epoch": 2.6945513218319475,
+      "grad_norm": 0.07061423607281134,
+      "learning_rate": 2.4097089034023726e-06,
+      "loss": 0.464,
+      "step": 5431
+    },
+    {
+      "epoch": 2.695047784535187,
+      "grad_norm": 0.0696942313119842,
+      "learning_rate": 2.408037158447375e-06,
+      "loss": 0.434,
+      "step": 5432
+    },
+    {
+      "epoch": 2.695544247238426,
+      "grad_norm": 0.07009626749102861,
+      "learning_rate": 2.406365809632341e-06,
+      "loss": 0.4455,
+      "step": 5433
+    },
+    {
+      "epoch": 2.6960407099416654,
+      "grad_norm": 0.07490099364686856,
+      "learning_rate": 2.4046948572127077e-06,
+      "loss": 0.459,
+      "step": 5434
+    },
+    {
+      "epoch": 2.696537172644905,
+      "grad_norm": 0.07357840038096851,
+      "learning_rate": 2.403024301443851e-06,
+      "loss": 0.4497,
+      "step": 5435
+    },
+    {
+      "epoch": 2.6970336353481446,
+      "grad_norm": 0.07322032132576278,
+      "learning_rate": 2.4013541425810916e-06,
+      "loss": 0.4502,
+      "step": 5436
+    },
+    {
+      "epoch": 2.697530098051384,
+      "grad_norm": 0.07315206853573321,
+      "learning_rate": 2.3996843808796845e-06,
+      "loss": 0.4657,
+      "step": 5437
+    },
+    {
+      "epoch": 2.698026560754623,
+      "grad_norm": 0.07619258559402037,
+      "learning_rate": 2.398015016594828e-06,
+      "loss": 0.501,
+      "step": 5438
+    },
+    {
+      "epoch": 2.698523023457863,
+      "grad_norm": 0.07015291482571168,
+      "learning_rate": 2.3963460499816564e-06,
+      "loss": 0.4231,
+      "step": 5439
+    },
+    {
+      "epoch": 2.6990194861611023,
+      "grad_norm": 0.07276475840477924,
+      "learning_rate": 2.394677481295243e-06,
+      "loss": 0.4425,
+      "step": 5440
+    },
+    {
+      "epoch": 2.6995159488643417,
+      "grad_norm": 0.07276518401473396,
+      "learning_rate": 2.393009310790606e-06,
+      "loss": 0.4847,
+      "step": 5441
+    },
+    {
+      "epoch": 2.700012411567581,
+      "grad_norm": 0.07071737375167014,
+      "learning_rate": 2.3913415387226936e-06,
+      "loss": 0.4427,
+      "step": 5442
+    },
+    {
+      "epoch": 2.7005088742708203,
+      "grad_norm": 0.07287583155324498,
+      "learning_rate": 2.389674165346402e-06,
+      "loss": 0.4741,
+      "step": 5443
+    },
+    {
+      "epoch": 2.7010053369740596,
+      "grad_norm": 0.07055392236578718,
+      "learning_rate": 2.3880071909165607e-06,
+      "loss": 0.4644,
+      "step": 5444
+    },
+    {
+      "epoch": 2.701501799677299,
+      "grad_norm": 0.07323683960653447,
+      "learning_rate": 2.386340615687941e-06,
+      "loss": 0.4754,
+      "step": 5445
+    },
+    {
+      "epoch": 2.7019982623805388,
+      "grad_norm": 0.07564816505304253,
+      "learning_rate": 2.3846744399152504e-06,
+      "loss": 0.4702,
+      "step": 5446
+    },
+    {
+      "epoch": 2.702494725083778,
+      "grad_norm": 0.06768172491094766,
+      "learning_rate": 2.3830086638531367e-06,
+      "loss": 0.4209,
+      "step": 5447
+    },
+    {
+      "epoch": 2.7029911877870174,
+      "grad_norm": 0.0717282008086792,
+      "learning_rate": 2.3813432877561903e-06,
+      "loss": 0.4452,
+      "step": 5448
+    },
+    {
+      "epoch": 2.703487650490257,
+      "grad_norm": 0.07271166896844942,
+      "learning_rate": 2.3796783118789335e-06,
+      "loss": 0.4541,
+      "step": 5449
+    },
+    {
+      "epoch": 2.7039841131934965,
+      "grad_norm": 0.07021717573627075,
+      "learning_rate": 2.378013736475835e-06,
+      "loss": 0.4052,
+      "step": 5450
+    },
+    {
+      "epoch": 2.704480575896736,
+      "grad_norm": 0.07301532107170802,
+      "learning_rate": 2.3763495618012967e-06,
+      "loss": 0.5041,
+      "step": 5451
+    },
+    {
+      "epoch": 2.704977038599975,
+      "grad_norm": 0.07150478795652813,
+      "learning_rate": 2.3746857881096586e-06,
+      "loss": 0.4766,
+      "step": 5452
+    },
+    {
+      "epoch": 2.7054735013032145,
+      "grad_norm": 0.07278421297777359,
+      "learning_rate": 2.3730224156552063e-06,
+      "loss": 0.4369,
+      "step": 5453
+    },
+    {
+      "epoch": 2.705969964006454,
+      "grad_norm": 0.07163528857638941,
+      "learning_rate": 2.3713594446921552e-06,
+      "loss": 0.4496,
+      "step": 5454
+    },
+    {
+      "epoch": 2.706466426709693,
+      "grad_norm": 0.07379949986235082,
+      "learning_rate": 2.3696968754746672e-06,
+      "loss": 0.4663,
+      "step": 5455
+    },
+    {
+      "epoch": 2.706962889412933,
+      "grad_norm": 0.06835946522496905,
+      "learning_rate": 2.3680347082568396e-06,
+      "loss": 0.4316,
+      "step": 5456
+    },
+    {
+      "epoch": 2.7074593521161723,
+      "grad_norm": 0.07097554521865745,
+      "learning_rate": 2.3663729432927034e-06,
+      "loss": 0.4601,
+      "step": 5457
+    },
+    {
+      "epoch": 2.7079558148194116,
+      "grad_norm": 0.07192575583390655,
+      "learning_rate": 2.364711580836238e-06,
+      "loss": 0.4741,
+      "step": 5458
+    },
+    {
+      "epoch": 2.7084522775226514,
+      "grad_norm": 0.07413193566513956,
+      "learning_rate": 2.363050621141354e-06,
+      "loss": 0.4797,
+      "step": 5459
+    },
+    {
+      "epoch": 2.7089487402258907,
+      "grad_norm": 0.07397341842935361,
+      "learning_rate": 2.3613900644619005e-06,
+      "loss": 0.4601,
+      "step": 5460
+    },
+    {
+      "epoch": 2.70944520292913,
+      "grad_norm": 0.07104537353838479,
+      "learning_rate": 2.3597299110516718e-06,
+      "loss": 0.4395,
+      "step": 5461
+    },
+    {
+      "epoch": 2.7099416656323694,
+      "grad_norm": 0.07220429705497064,
+      "learning_rate": 2.3580701611643896e-06,
+      "loss": 0.4491,
+      "step": 5462
+    },
+    {
+      "epoch": 2.7104381283356087,
+      "grad_norm": 0.07237243617685923,
+      "learning_rate": 2.356410815053725e-06,
+      "loss": 0.4477,
+      "step": 5463
+    },
+    {
+      "epoch": 2.710934591038848,
+      "grad_norm": 0.07142769983441732,
+      "learning_rate": 2.3547518729732788e-06,
+      "loss": 0.434,
+      "step": 5464
+    },
+    {
+      "epoch": 2.7114310537420874,
+      "grad_norm": 0.06938918876377276,
+      "learning_rate": 2.3530933351765967e-06,
+      "loss": 0.4308,
+      "step": 5465
+    },
+    {
+      "epoch": 2.711927516445327,
+      "grad_norm": 0.07652269457501713,
+      "learning_rate": 2.351435201917159e-06,
+      "loss": 0.4627,
+      "step": 5466
+    },
+    {
+      "epoch": 2.7124239791485665,
+      "grad_norm": 0.0704284583828208,
+      "learning_rate": 2.3497774734483827e-06,
+      "loss": 0.4459,
+      "step": 5467
+    },
+    {
+      "epoch": 2.712920441851806,
+      "grad_norm": 0.07163313771334788,
+      "learning_rate": 2.348120150023627e-06,
+      "loss": 0.4592,
+      "step": 5468
+    },
+    {
+      "epoch": 2.713416904555045,
+      "grad_norm": 0.07290632848977743,
+      "learning_rate": 2.346463231896186e-06,
+      "loss": 0.4463,
+      "step": 5469
+    },
+    {
+      "epoch": 2.713913367258285,
+      "grad_norm": 0.07016113247477145,
+      "learning_rate": 2.3448067193192953e-06,
+      "loss": 0.4487,
+      "step": 5470
+    },
+    {
+      "epoch": 2.7144098299615242,
+      "grad_norm": 0.07214476803834971,
+      "learning_rate": 2.3431506125461243e-06,
+      "loss": 0.4741,
+      "step": 5471
+    },
+    {
+      "epoch": 2.7149062926647636,
+      "grad_norm": 0.07456185081547756,
+      "learning_rate": 2.341494911829782e-06,
+      "loss": 0.4622,
+      "step": 5472
+    },
+    {
+      "epoch": 2.715402755368003,
+      "grad_norm": 0.07237360880431196,
+      "learning_rate": 2.339839617423318e-06,
+      "loss": 0.4471,
+      "step": 5473
+    },
+    {
+      "epoch": 2.7158992180712422,
+      "grad_norm": 0.07839016208893054,
+      "learning_rate": 2.338184729579714e-06,
+      "loss": 0.4548,
+      "step": 5474
+    },
+    {
+      "epoch": 2.7163956807744816,
+      "grad_norm": 0.07457366522480684,
+      "learning_rate": 2.3365302485518966e-06,
+      "loss": 0.4692,
+      "step": 5475
+    },
+    {
+      "epoch": 2.7168921434777213,
+      "grad_norm": 0.07175463069898462,
+      "learning_rate": 2.3348761745927258e-06,
+      "loss": 0.4495,
+      "step": 5476
+    },
+    {
+      "epoch": 2.7173886061809607,
+      "grad_norm": 0.07097627701034127,
+      "learning_rate": 2.3332225079549995e-06,
+      "loss": 0.4255,
+      "step": 5477
+    },
+    {
+      "epoch": 2.7178850688842,
+      "grad_norm": 0.07196217651367964,
+      "learning_rate": 2.3315692488914544e-06,
+      "loss": 0.49,
+      "step": 5478
+    },
+    {
+      "epoch": 2.7183815315874393,
+      "grad_norm": 0.07703244837469461,
+      "learning_rate": 2.329916397654763e-06,
+      "loss": 0.4936,
+      "step": 5479
+    },
+    {
+      "epoch": 2.718877994290679,
+      "grad_norm": 0.07217612227351226,
+      "learning_rate": 2.32826395449754e-06,
+      "loss": 0.483,
+      "step": 5480
+    },
+    {
+      "epoch": 2.7193744569939184,
+      "grad_norm": 0.07172214175782687,
+      "learning_rate": 2.326611919672332e-06,
+      "loss": 0.463,
+      "step": 5481
+    },
+    {
+      "epoch": 2.7198709196971578,
+      "grad_norm": 0.07239940889748474,
+      "learning_rate": 2.324960293431629e-06,
+      "loss": 0.4364,
+      "step": 5482
+    },
+    {
+      "epoch": 2.720367382400397,
+      "grad_norm": 0.07164091885322517,
+      "learning_rate": 2.3233090760278544e-06,
+      "loss": 0.4499,
+      "step": 5483
+    },
+    {
+      "epoch": 2.7208638451036364,
+      "grad_norm": 0.07414212769490823,
+      "learning_rate": 2.3216582677133682e-06,
+      "loss": 0.4659,
+      "step": 5484
+    },
+    {
+      "epoch": 2.7213603078068758,
+      "grad_norm": 0.07135247244737274,
+      "learning_rate": 2.3200078687404736e-06,
+      "loss": 0.4467,
+      "step": 5485
+    },
+    {
+      "epoch": 2.7218567705101155,
+      "grad_norm": 0.07245678162225738,
+      "learning_rate": 2.3183578793614043e-06,
+      "loss": 0.4637,
+      "step": 5486
+    },
+    {
+      "epoch": 2.722353233213355,
+      "grad_norm": 0.0713600463717167,
+      "learning_rate": 2.316708299828338e-06,
+      "loss": 0.4504,
+      "step": 5487
+    },
+    {
+      "epoch": 2.722849695916594,
+      "grad_norm": 0.0736838370568765,
+      "learning_rate": 2.3150591303933852e-06,
+      "loss": 0.459,
+      "step": 5488
+    },
+    {
+      "epoch": 2.7233461586198335,
+      "grad_norm": 0.07028199554059375,
+      "learning_rate": 2.313410371308592e-06,
+      "loss": 0.4556,
+      "step": 5489
+    },
+    {
+      "epoch": 2.7238426213230733,
+      "grad_norm": 0.07112815035167214,
+      "learning_rate": 2.311762022825949e-06,
+      "loss": 0.4603,
+      "step": 5490
+    },
+    {
+      "epoch": 2.7243390840263126,
+      "grad_norm": 0.07077727556976685,
+      "learning_rate": 2.3101140851973768e-06,
+      "loss": 0.4589,
+      "step": 5491
+    },
+    {
+      "epoch": 2.724835546729552,
+      "grad_norm": 0.07497765247200448,
+      "learning_rate": 2.3084665586747397e-06,
+      "loss": 0.4635,
+      "step": 5492
+    },
+    {
+      "epoch": 2.7253320094327913,
+      "grad_norm": 0.07385568530723025,
+      "learning_rate": 2.3068194435098334e-06,
+      "loss": 0.4556,
+      "step": 5493
+    },
+    {
+      "epoch": 2.7258284721360306,
+      "grad_norm": 0.07406628874892704,
+      "learning_rate": 2.3051727399543934e-06,
+      "loss": 0.4707,
+      "step": 5494
+    },
+    {
+      "epoch": 2.72632493483927,
+      "grad_norm": 0.07091425812135041,
+      "learning_rate": 2.3035264482600915e-06,
+      "loss": 0.4367,
+      "step": 5495
+    },
+    {
+      "epoch": 2.7268213975425097,
+      "grad_norm": 0.06863687149117234,
+      "learning_rate": 2.3018805686785362e-06,
+      "loss": 0.4276,
+      "step": 5496
+    },
+    {
+      "epoch": 2.727317860245749,
+      "grad_norm": 0.07051110974771434,
+      "learning_rate": 2.300235101461276e-06,
+      "loss": 0.4584,
+      "step": 5497
+    },
+    {
+      "epoch": 2.7278143229489884,
+      "grad_norm": 0.071099105480711,
+      "learning_rate": 2.298590046859793e-06,
+      "loss": 0.4293,
+      "step": 5498
+    },
+    {
+      "epoch": 2.7283107856522277,
+      "grad_norm": 0.07455777915040125,
+      "learning_rate": 2.2969454051255064e-06,
+      "loss": 0.459,
+      "step": 5499
+    },
+    {
+      "epoch": 2.7288072483554675,
+      "grad_norm": 0.07307833101340502,
+      "learning_rate": 2.295301176509776e-06,
+      "loss": 0.4612,
+      "step": 5500
+    },
+    {
+      "epoch": 2.729303711058707,
+      "grad_norm": 0.07217219742126983,
+      "learning_rate": 2.2936573612638922e-06,
+      "loss": 0.4683,
+      "step": 5501
+    },
+    {
+      "epoch": 2.729800173761946,
+      "grad_norm": 0.0744157919215635,
+      "learning_rate": 2.29201395963909e-06,
+      "loss": 0.4659,
+      "step": 5502
+    },
+    {
+      "epoch": 2.7302966364651855,
+      "grad_norm": 0.07289841722012438,
+      "learning_rate": 2.2903709718865347e-06,
+      "loss": 0.445,
+      "step": 5503
+    },
+    {
+      "epoch": 2.730793099168425,
+      "grad_norm": 0.07298563895816411,
+      "learning_rate": 2.2887283982573287e-06,
+      "loss": 0.4453,
+      "step": 5504
+    },
+    {
+      "epoch": 2.731289561871664,
+      "grad_norm": 0.1200957041826658,
+      "learning_rate": 2.2870862390025172e-06,
+      "loss": 0.4756,
+      "step": 5505
+    },
+    {
+      "epoch": 2.731786024574904,
+      "grad_norm": 0.0694534964088227,
+      "learning_rate": 2.2854444943730735e-06,
+      "loss": 0.4292,
+      "step": 5506
+    },
+    {
+      "epoch": 2.7322824872781433,
+      "grad_norm": 0.07376327414716859,
+      "learning_rate": 2.2838031646199164e-06,
+      "loss": 0.4563,
+      "step": 5507
+    },
+    {
+      "epoch": 2.7327789499813826,
+      "grad_norm": 0.07381798743005258,
+      "learning_rate": 2.282162249993895e-06,
+      "loss": 0.4722,
+      "step": 5508
+    },
+    {
+      "epoch": 2.733275412684622,
+      "grad_norm": 0.07069931901928059,
+      "learning_rate": 2.280521750745796e-06,
+      "loss": 0.4527,
+      "step": 5509
+    },
+    {
+      "epoch": 2.7337718753878617,
+      "grad_norm": 0.07238019601026131,
+      "learning_rate": 2.2788816671263443e-06,
+      "loss": 0.4562,
+      "step": 5510
+    },
+    {
+      "epoch": 2.734268338091101,
+      "grad_norm": 0.07535935212905417,
+      "learning_rate": 2.277241999386198e-06,
+      "loss": 0.4729,
+      "step": 5511
+    },
+    {
+      "epoch": 2.7347648007943404,
+      "grad_norm": 0.07148930959040346,
+      "learning_rate": 2.2756027477759573e-06,
+      "loss": 0.4691,
+      "step": 5512
+    },
+    {
+      "epoch": 2.7352612634975797,
+      "grad_norm": 0.07042681073402034,
+      "learning_rate": 2.2739639125461526e-06,
+      "loss": 0.4502,
+      "step": 5513
+    },
+    {
+      "epoch": 2.735757726200819,
+      "grad_norm": 0.0825610958308842,
+      "learning_rate": 2.272325493947257e-06,
+      "loss": 0.446,
+      "step": 5514
+    },
+    {
+      "epoch": 2.7362541889040584,
+      "grad_norm": 0.07122332767066362,
+      "learning_rate": 2.2706874922296756e-06,
+      "loss": 0.4481,
+      "step": 5515
+    },
+    {
+      "epoch": 2.736750651607298,
+      "grad_norm": 0.07202439783077769,
+      "learning_rate": 2.2690499076437472e-06,
+      "loss": 0.4452,
+      "step": 5516
+    },
+    {
+      "epoch": 2.7372471143105375,
+      "grad_norm": 0.0759937883079161,
+      "learning_rate": 2.267412740439755e-06,
+      "loss": 0.4989,
+      "step": 5517
+    },
+    {
+      "epoch": 2.737743577013777,
+      "grad_norm": 0.07122852007942947,
+      "learning_rate": 2.2657759908679093e-06,
+      "loss": 0.4392,
+      "step": 5518
+    },
+    {
+      "epoch": 2.738240039717016,
+      "grad_norm": 0.0722685401508018,
+      "learning_rate": 2.264139659178366e-06,
+      "loss": 0.4496,
+      "step": 5519
+    },
+    {
+      "epoch": 2.738736502420256,
+      "grad_norm": 0.07136737189013136,
+      "learning_rate": 2.2625037456212096e-06,
+      "loss": 0.4319,
+      "step": 5520
+    },
+    {
+      "epoch": 2.7392329651234952,
+      "grad_norm": 0.07323565448937264,
+      "learning_rate": 2.2608682504464614e-06,
+      "loss": 0.4462,
+      "step": 5521
+    },
+    {
+      "epoch": 2.7397294278267346,
+      "grad_norm": 0.07192415827088411,
+      "learning_rate": 2.259233173904084e-06,
+      "loss": 0.4269,
+      "step": 5522
+    },
+    {
+      "epoch": 2.740225890529974,
+      "grad_norm": 0.07138721363420222,
+      "learning_rate": 2.257598516243969e-06,
+      "loss": 0.4625,
+      "step": 5523
+    },
+    {
+      "epoch": 2.7407223532332132,
+      "grad_norm": 0.07603992677823823,
+      "learning_rate": 2.2559642777159525e-06,
+      "loss": 0.4848,
+      "step": 5524
+    },
+    {
+      "epoch": 2.7412188159364526,
+      "grad_norm": 0.07173016922467354,
+      "learning_rate": 2.2543304585697977e-06,
+      "loss": 0.436,
+      "step": 5525
+    },
+    {
+      "epoch": 2.7417152786396923,
+      "grad_norm": 0.0718564409263651,
+      "learning_rate": 2.25269705905521e-06,
+      "loss": 0.4443,
+      "step": 5526
+    },
+    {
+      "epoch": 2.7422117413429317,
+      "grad_norm": 0.07145531667680151,
+      "learning_rate": 2.2510640794218264e-06,
+      "loss": 0.4714,
+      "step": 5527
+    },
+    {
+      "epoch": 2.742708204046171,
+      "grad_norm": 0.0703115660351647,
+      "learning_rate": 2.2494315199192206e-06,
+      "loss": 0.4541,
+      "step": 5528
+    },
+    {
+      "epoch": 2.7432046667494103,
+      "grad_norm": 0.07513605107381305,
+      "learning_rate": 2.2477993807969074e-06,
+      "loss": 0.4958,
+      "step": 5529
+    },
+    {
+      "epoch": 2.74370112945265,
+      "grad_norm": 0.07026907153785616,
+      "learning_rate": 2.2461676623043283e-06,
+      "loss": 0.4617,
+      "step": 5530
+    },
+    {
+      "epoch": 2.7441975921558894,
+      "grad_norm": 0.0731908821401188,
+      "learning_rate": 2.2445363646908698e-06,
+      "loss": 0.5116,
+      "step": 5531
+    },
+    {
+      "epoch": 2.7446940548591288,
+      "grad_norm": 0.07507413862164442,
+      "learning_rate": 2.242905488205848e-06,
+      "loss": 0.4828,
+      "step": 5532
+    },
+    {
+      "epoch": 2.745190517562368,
+      "grad_norm": 0.07326962669308193,
+      "learning_rate": 2.241275033098513e-06,
+      "loss": 0.4768,
+      "step": 5533
+    },
+    {
+      "epoch": 2.7456869802656074,
+      "grad_norm": 0.07446257400543148,
+      "learning_rate": 2.239644999618058e-06,
+      "loss": 0.4586,
+      "step": 5534
+    },
+    {
+      "epoch": 2.7461834429688468,
+      "grad_norm": 0.06965894292069011,
+      "learning_rate": 2.2380153880136062e-06,
+      "loss": 0.4433,
+      "step": 5535
+    },
+    {
+      "epoch": 2.7466799056720865,
+      "grad_norm": 0.06927299902017811,
+      "learning_rate": 2.2363861985342156e-06,
+      "loss": 0.4458,
+      "step": 5536
+    },
+    {
+      "epoch": 2.747176368375326,
+      "grad_norm": 0.06964886413539237,
+      "learning_rate": 2.234757431428885e-06,
+      "loss": 0.4247,
+      "step": 5537
+    },
+    {
+      "epoch": 2.747672831078565,
+      "grad_norm": 0.07289524357337542,
+      "learning_rate": 2.233129086946542e-06,
+      "loss": 0.4678,
+      "step": 5538
+    },
+    {
+      "epoch": 2.7481692937818045,
+      "grad_norm": 0.07102667782964327,
+      "learning_rate": 2.231501165336057e-06,
+      "loss": 0.456,
+      "step": 5539
+    },
+    {
+      "epoch": 2.7486657564850443,
+      "grad_norm": 0.07411113072941954,
+      "learning_rate": 2.229873666846229e-06,
+      "loss": 0.475,
+      "step": 5540
+    },
+    {
+      "epoch": 2.7491622191882836,
+      "grad_norm": 0.07161052155330794,
+      "learning_rate": 2.2282465917257952e-06,
+      "loss": 0.4589,
+      "step": 5541
+    },
+    {
+      "epoch": 2.749658681891523,
+      "grad_norm": 0.07736293002603781,
+      "learning_rate": 2.2266199402234286e-06,
+      "loss": 0.4569,
+      "step": 5542
+    },
+    {
+      "epoch": 2.7501551445947623,
+      "grad_norm": 0.07242961476602337,
+      "learning_rate": 2.224993712587734e-06,
+      "loss": 0.4766,
+      "step": 5543
+    },
+    {
+      "epoch": 2.7506516072980016,
+      "grad_norm": 0.07153571234095471,
+      "learning_rate": 2.2233679090672584e-06,
+      "loss": 0.4473,
+      "step": 5544
+    },
+    {
+      "epoch": 2.7506516072980016,
+      "eval_loss": 0.5140753388404846,
+      "eval_runtime": 259.2366,
+      "eval_samples_per_second": 117.086,
+      "eval_steps_per_second": 14.639,
+      "step": 5544
+    },
+    {
+      "epoch": 2.751148070001241,
+      "grad_norm": 0.07403899048266355,
+      "learning_rate": 2.221742529910477e-06,
+      "loss": 0.4497,
+      "step": 5545
+    },
+    {
+      "epoch": 2.7516445327044807,
+      "grad_norm": 0.06701646326960832,
+      "learning_rate": 2.2201175753658048e-06,
+      "loss": 0.4053,
+      "step": 5546
+    },
+    {
+      "epoch": 2.75214099540772,
+      "grad_norm": 0.0674570629487927,
+      "learning_rate": 2.2184930456815897e-06,
+      "loss": 0.4052,
+      "step": 5547
+    },
+    {
+      "epoch": 2.7526374581109594,
+      "grad_norm": 0.07291171901110415,
+      "learning_rate": 2.2168689411061123e-06,
+      "loss": 0.4813,
+      "step": 5548
+    },
+    {
+      "epoch": 2.7531339208141987,
+      "grad_norm": 0.07132406878589202,
+      "learning_rate": 2.2152452618875954e-06,
+      "loss": 0.4408,
+      "step": 5549
+    },
+    {
+      "epoch": 2.7536303835174385,
+      "grad_norm": 0.07134727618995322,
+      "learning_rate": 2.2136220082741876e-06,
+      "loss": 0.4308,
+      "step": 5550
+    },
+    {
+      "epoch": 2.754126846220678,
+      "grad_norm": 0.07343701614477756,
+      "learning_rate": 2.2119991805139823e-06,
+      "loss": 0.4781,
+      "step": 5551
+    },
+    {
+      "epoch": 2.754623308923917,
+      "grad_norm": 0.07166201620344755,
+      "learning_rate": 2.2103767788549996e-06,
+      "loss": 0.4588,
+      "step": 5552
+    },
+    {
+      "epoch": 2.7551197716271565,
+      "grad_norm": 0.07609433226158972,
+      "learning_rate": 2.208754803545196e-06,
+      "loss": 0.4684,
+      "step": 5553
+    },
+    {
+      "epoch": 2.755616234330396,
+      "grad_norm": 0.072588971915586,
+      "learning_rate": 2.2071332548324688e-06,
+      "loss": 0.487,
+      "step": 5554
+    },
+    {
+      "epoch": 2.756112697033635,
+      "grad_norm": 0.0728012780360345,
+      "learning_rate": 2.2055121329646416e-06,
+      "loss": 0.4359,
+      "step": 5555
+    },
+    {
+      "epoch": 2.756609159736875,
+      "grad_norm": 0.0714482252676178,
+      "learning_rate": 2.20389143818948e-06,
+      "loss": 0.4607,
+      "step": 5556
+    },
+    {
+      "epoch": 2.7571056224401143,
+      "grad_norm": 0.07300698715457024,
+      "learning_rate": 2.20227117075468e-06,
+      "loss": 0.4526,
+      "step": 5557
+    },
+    {
+      "epoch": 2.7576020851433536,
+      "grad_norm": 0.07058202943928703,
+      "learning_rate": 2.200651330907874e-06,
+      "loss": 0.4424,
+      "step": 5558
+    },
+    {
+      "epoch": 2.758098547846593,
+      "grad_norm": 0.07194090382598976,
+      "learning_rate": 2.1990319188966276e-06,
+      "loss": 0.4564,
+      "step": 5559
+    },
+    {
+      "epoch": 2.7585950105498327,
+      "grad_norm": 0.07413470061933844,
+      "learning_rate": 2.19741293496844e-06,
+      "loss": 0.4785,
+      "step": 5560
+    },
+    {
+      "epoch": 2.759091473253072,
+      "grad_norm": 0.07339601606666014,
+      "learning_rate": 2.1957943793707517e-06,
+      "loss": 0.4804,
+      "step": 5561
+    },
+    {
+      "epoch": 2.7595879359563114,
+      "grad_norm": 0.07682989147253744,
+      "learning_rate": 2.1941762523509282e-06,
+      "loss": 0.4757,
+      "step": 5562
+    },
+    {
+      "epoch": 2.7600843986595507,
+      "grad_norm": 0.07177187759459464,
+      "learning_rate": 2.192558554156278e-06,
+      "loss": 0.4583,
+      "step": 5563
+    },
+    {
+      "epoch": 2.76058086136279,
+      "grad_norm": 0.07238024054866415,
+      "learning_rate": 2.1909412850340395e-06,
+      "loss": 0.4201,
+      "step": 5564
+    },
+    {
+      "epoch": 2.7610773240660293,
+      "grad_norm": 0.07107079343533929,
+      "learning_rate": 2.189324445231384e-06,
+      "loss": 0.4708,
+      "step": 5565
+    },
+    {
+      "epoch": 2.761573786769269,
+      "grad_norm": 0.07580931353151424,
+      "learning_rate": 2.187708034995423e-06,
+      "loss": 0.4555,
+      "step": 5566
+    },
+    {
+      "epoch": 2.7620702494725085,
+      "grad_norm": 0.074047343219353,
+      "learning_rate": 2.186092054573195e-06,
+      "loss": 0.4661,
+      "step": 5567
+    },
+    {
+      "epoch": 2.762566712175748,
+      "grad_norm": 0.07350297127509982,
+      "learning_rate": 2.184476504211681e-06,
+      "loss": 0.4955,
+      "step": 5568
+    },
+    {
+      "epoch": 2.763063174878987,
+      "grad_norm": 0.07236753011939805,
+      "learning_rate": 2.18286138415779e-06,
+      "loss": 0.465,
+      "step": 5569
+    },
+    {
+      "epoch": 2.763559637582227,
+      "grad_norm": 0.07229715160125913,
+      "learning_rate": 2.1812466946583654e-06,
+      "loss": 0.4349,
+      "step": 5570
+    },
+    {
+      "epoch": 2.764056100285466,
+      "grad_norm": 0.07448114359837037,
+      "learning_rate": 2.1796324359601896e-06,
+      "loss": 0.4787,
+      "step": 5571
+    },
+    {
+      "epoch": 2.7645525629887056,
+      "grad_norm": 0.06942310726010976,
+      "learning_rate": 2.1780186083099746e-06,
+      "loss": 0.4285,
+      "step": 5572
+    },
+    {
+      "epoch": 2.765049025691945,
+      "grad_norm": 0.07393983006350734,
+      "learning_rate": 2.176405211954369e-06,
+      "loss": 0.4731,
+      "step": 5573
+    },
+    {
+      "epoch": 2.765545488395184,
+      "grad_norm": 0.0713507331251344,
+      "learning_rate": 2.1747922471399517e-06,
+      "loss": 0.4254,
+      "step": 5574
+    },
+    {
+      "epoch": 2.7660419510984235,
+      "grad_norm": 0.07469750472023376,
+      "learning_rate": 2.1731797141132425e-06,
+      "loss": 0.5091,
+      "step": 5575
+    },
+    {
+      "epoch": 2.766538413801663,
+      "grad_norm": 0.07529947136126779,
+      "learning_rate": 2.1715676131206893e-06,
+      "loss": 0.498,
+      "step": 5576
+    },
+    {
+      "epoch": 2.7670348765049027,
+      "grad_norm": 0.0714992330808803,
+      "learning_rate": 2.169955944408674e-06,
+      "loss": 0.4412,
+      "step": 5577
+    },
+    {
+      "epoch": 2.767531339208142,
+      "grad_norm": 0.0730209938522385,
+      "learning_rate": 2.168344708223519e-06,
+      "loss": 0.4711,
+      "step": 5578
+    },
+    {
+      "epoch": 2.7680278019113813,
+      "grad_norm": 0.07410312426064489,
+      "learning_rate": 2.166733904811472e-06,
+      "loss": 0.5138,
+      "step": 5579
+    },
+    {
+      "epoch": 2.768524264614621,
+      "grad_norm": 0.07265453900182571,
+      "learning_rate": 2.1651235344187183e-06,
+      "loss": 0.4711,
+      "step": 5580
+    },
+    {
+      "epoch": 2.7690207273178604,
+      "grad_norm": 0.07219486082807354,
+      "learning_rate": 2.163513597291381e-06,
+      "loss": 0.449,
+      "step": 5581
+    },
+    {
+      "epoch": 2.7695171900210998,
+      "grad_norm": 0.07169161843650458,
+      "learning_rate": 2.1619040936755083e-06,
+      "loss": 0.4685,
+      "step": 5582
+    },
+    {
+      "epoch": 2.770013652724339,
+      "grad_norm": 0.0716278342968845,
+      "learning_rate": 2.160295023817091e-06,
+      "loss": 0.4806,
+      "step": 5583
+    },
+    {
+      "epoch": 2.7705101154275784,
+      "grad_norm": 0.07458597427259725,
+      "learning_rate": 2.1586863879620486e-06,
+      "loss": 0.4651,
+      "step": 5584
+    },
+    {
+      "epoch": 2.7710065781308177,
+      "grad_norm": 0.07458008902961828,
+      "learning_rate": 2.1570781863562328e-06,
+      "loss": 0.4538,
+      "step": 5585
+    },
+    {
+      "epoch": 2.771503040834057,
+      "grad_norm": 0.07663571921112627,
+      "learning_rate": 2.155470419245435e-06,
+      "loss": 0.4871,
+      "step": 5586
+    },
+    {
+      "epoch": 2.771999503537297,
+      "grad_norm": 0.07248567735016598,
+      "learning_rate": 2.153863086875374e-06,
+      "loss": 0.4953,
+      "step": 5587
+    },
+    {
+      "epoch": 2.772495966240536,
+      "grad_norm": 0.07208045569274636,
+      "learning_rate": 2.152256189491707e-06,
+      "loss": 0.4117,
+      "step": 5588
+    },
+    {
+      "epoch": 2.7729924289437755,
+      "grad_norm": 0.07040926609012249,
+      "learning_rate": 2.1506497273400218e-06,
+      "loss": 0.4512,
+      "step": 5589
+    },
+    {
+      "epoch": 2.7734888916470153,
+      "grad_norm": 0.07146556046724373,
+      "learning_rate": 2.1490437006658393e-06,
+      "loss": 0.4704,
+      "step": 5590
+    },
+    {
+      "epoch": 2.7739853543502546,
+      "grad_norm": 0.072885103918037,
+      "learning_rate": 2.1474381097146163e-06,
+      "loss": 0.4435,
+      "step": 5591
+    },
+    {
+      "epoch": 2.774481817053494,
+      "grad_norm": 0.07419544353699468,
+      "learning_rate": 2.1458329547317384e-06,
+      "loss": 0.4622,
+      "step": 5592
+    },
+    {
+      "epoch": 2.7749782797567333,
+      "grad_norm": 0.07108922480782838,
+      "learning_rate": 2.144228235962533e-06,
+      "loss": 0.4458,
+      "step": 5593
+    },
+    {
+      "epoch": 2.7754747424599726,
+      "grad_norm": 0.07248340212697033,
+      "learning_rate": 2.1426239536522497e-06,
+      "loss": 0.455,
+      "step": 5594
+    },
+    {
+      "epoch": 2.775971205163212,
+      "grad_norm": 0.07119818086272339,
+      "learning_rate": 2.1410201080460837e-06,
+      "loss": 0.4766,
+      "step": 5595
+    },
+    {
+      "epoch": 2.7764676678664513,
+      "grad_norm": 0.07272794995230875,
+      "learning_rate": 2.139416699389153e-06,
+      "loss": 0.4557,
+      "step": 5596
+    },
+    {
+      "epoch": 2.776964130569691,
+      "grad_norm": 0.07336170551958457,
+      "learning_rate": 2.1378137279265126e-06,
+      "loss": 0.5142,
+      "step": 5597
+    },
+    {
+      "epoch": 2.7774605932729304,
+      "grad_norm": 0.07038873630375103,
+      "learning_rate": 2.1362111939031538e-06,
+      "loss": 0.4593,
+      "step": 5598
+    },
+    {
+      "epoch": 2.7779570559761697,
+      "grad_norm": 0.07121169984717571,
+      "learning_rate": 2.134609097563995e-06,
+      "loss": 0.4536,
+      "step": 5599
+    },
+    {
+      "epoch": 2.7784535186794095,
+      "grad_norm": 0.07275418181501159,
+      "learning_rate": 2.133007439153894e-06,
+      "loss": 0.4929,
+      "step": 5600
+    },
+    {
+      "epoch": 2.778949981382649,
+      "grad_norm": 0.0711687503533935,
+      "learning_rate": 2.131406218917637e-06,
+      "loss": 0.4441,
+      "step": 5601
+    },
+    {
+      "epoch": 2.779446444085888,
+      "grad_norm": 0.0717822196418287,
+      "learning_rate": 2.129805437099944e-06,
+      "loss": 0.4597,
+      "step": 5602
+    },
+    {
+      "epoch": 2.7799429067891275,
+      "grad_norm": 0.07293237039541688,
+      "learning_rate": 2.1282050939454713e-06,
+      "loss": 0.4644,
+      "step": 5603
+    },
+    {
+      "epoch": 2.780439369492367,
+      "grad_norm": 0.07441525762705241,
+      "learning_rate": 2.126605189698803e-06,
+      "loss": 0.4539,
+      "step": 5604
+    },
+    {
+      "epoch": 2.780935832195606,
+      "grad_norm": 0.07075570422249783,
+      "learning_rate": 2.125005724604461e-06,
+      "loss": 0.447,
+      "step": 5605
+    },
+    {
+      "epoch": 2.7814322948988455,
+      "grad_norm": 0.07113131933005146,
+      "learning_rate": 2.1234066989068972e-06,
+      "loss": 0.4424,
+      "step": 5606
+    },
+    {
+      "epoch": 2.7819287576020852,
+      "grad_norm": 0.07083162859588368,
+      "learning_rate": 2.121808112850497e-06,
+      "loss": 0.4477,
+      "step": 5607
+    },
+    {
+      "epoch": 2.7824252203053246,
+      "grad_norm": 0.07373674057526543,
+      "learning_rate": 2.1202099666795783e-06,
+      "loss": 0.4429,
+      "step": 5608
+    },
+    {
+      "epoch": 2.782921683008564,
+      "grad_norm": 0.07349052401745303,
+      "learning_rate": 2.118612260638391e-06,
+      "loss": 0.5113,
+      "step": 5609
+    },
+    {
+      "epoch": 2.7834181457118032,
+      "grad_norm": 0.07150161174098543,
+      "learning_rate": 2.117014994971121e-06,
+      "loss": 0.4765,
+      "step": 5610
+    },
+    {
+      "epoch": 2.783914608415043,
+      "grad_norm": 0.06954490397678757,
+      "learning_rate": 2.115418169921883e-06,
+      "loss": 0.4373,
+      "step": 5611
+    },
+    {
+      "epoch": 2.7844110711182823,
+      "grad_norm": 0.06960572970474305,
+      "learning_rate": 2.1138217857347284e-06,
+      "loss": 0.4309,
+      "step": 5612
+    },
+    {
+      "epoch": 2.7849075338215217,
+      "grad_norm": 0.07441963347632186,
+      "learning_rate": 2.1122258426536373e-06,
+      "loss": 0.498,
+      "step": 5613
+    },
+    {
+      "epoch": 2.785403996524761,
+      "grad_norm": 0.07185784460169722,
+      "learning_rate": 2.1106303409225228e-06,
+      "loss": 0.4486,
+      "step": 5614
+    },
+    {
+      "epoch": 2.7859004592280003,
+      "grad_norm": 0.07171485254229928,
+      "learning_rate": 2.1090352807852344e-06,
+      "loss": 0.4587,
+      "step": 5615
+    },
+    {
+      "epoch": 2.7863969219312397,
+      "grad_norm": 0.07395003793923444,
+      "learning_rate": 2.1074406624855505e-06,
+      "loss": 0.4222,
+      "step": 5616
+    },
+    {
+      "epoch": 2.7868933846344794,
+      "grad_norm": 0.06817105944314727,
+      "learning_rate": 2.1058464862671805e-06,
+      "loss": 0.4311,
+      "step": 5617
+    },
+    {
+      "epoch": 2.7873898473377188,
+      "grad_norm": 0.07146224716442358,
+      "learning_rate": 2.1042527523737732e-06,
+      "loss": 0.4587,
+      "step": 5618
+    },
+    {
+      "epoch": 2.787886310040958,
+      "grad_norm": 0.07262673582585218,
+      "learning_rate": 2.1026594610489013e-06,
+      "loss": 0.4652,
+      "step": 5619
+    },
+    {
+      "epoch": 2.7883827727441974,
+      "grad_norm": 0.07216345014974743,
+      "learning_rate": 2.1010666125360767e-06,
+      "loss": 0.471,
+      "step": 5620
+    },
+    {
+      "epoch": 2.788879235447437,
+      "grad_norm": 0.07400589154296823,
+      "learning_rate": 2.09947420707874e-06,
+      "loss": 0.4607,
+      "step": 5621
+    },
+    {
+      "epoch": 2.7893756981506765,
+      "grad_norm": 0.0758128860095602,
+      "learning_rate": 2.097882244920264e-06,
+      "loss": 0.4726,
+      "step": 5622
+    },
+    {
+      "epoch": 2.789872160853916,
+      "grad_norm": 0.0716427119065037,
+      "learning_rate": 2.096290726303955e-06,
+      "loss": 0.4512,
+      "step": 5623
+    },
+    {
+      "epoch": 2.790368623557155,
+      "grad_norm": 0.07378414988935275,
+      "learning_rate": 2.0946996514730494e-06,
+      "loss": 0.4646,
+      "step": 5624
+    },
+    {
+      "epoch": 2.7908650862603945,
+      "grad_norm": 0.07321208031026308,
+      "learning_rate": 2.093109020670721e-06,
+      "loss": 0.4619,
+      "step": 5625
+    },
+    {
+      "epoch": 2.791361548963634,
+      "grad_norm": 0.07314239115958235,
+      "learning_rate": 2.0915188341400685e-06,
+      "loss": 0.4998,
+      "step": 5626
+    },
+    {
+      "epoch": 2.7918580116668736,
+      "grad_norm": 0.07090958314562905,
+      "learning_rate": 2.08992909212413e-06,
+      "loss": 0.4523,
+      "step": 5627
+    },
+    {
+      "epoch": 2.792354474370113,
+      "grad_norm": 0.07124752970334998,
+      "learning_rate": 2.0883397948658702e-06,
+      "loss": 0.4463,
+      "step": 5628
+    },
+    {
+      "epoch": 2.7928509370733523,
+      "grad_norm": 0.07221734657858735,
+      "learning_rate": 2.086750942608186e-06,
+      "loss": 0.4478,
+      "step": 5629
+    },
+    {
+      "epoch": 2.7933473997765916,
+      "grad_norm": 0.07331998158164842,
+      "learning_rate": 2.0851625355939117e-06,
+      "loss": 0.488,
+      "step": 5630
+    },
+    {
+      "epoch": 2.7938438624798314,
+      "grad_norm": 0.07110673575019735,
+      "learning_rate": 2.0835745740658057e-06,
+      "loss": 0.4409,
+      "step": 5631
+    },
+    {
+      "epoch": 2.7943403251830707,
+      "grad_norm": 0.06747812385303419,
+      "learning_rate": 2.0819870582665676e-06,
+      "loss": 0.4219,
+      "step": 5632
+    },
+    {
+      "epoch": 2.79483678788631,
+      "grad_norm": 0.07053503688375928,
+      "learning_rate": 2.08039998843882e-06,
+      "loss": 0.4516,
+      "step": 5633
+    },
+    {
+      "epoch": 2.7953332505895494,
+      "grad_norm": 0.06946889062359742,
+      "learning_rate": 2.0788133648251207e-06,
+      "loss": 0.4344,
+      "step": 5634
+    },
+    {
+      "epoch": 2.7958297132927887,
+      "grad_norm": 0.07207456844077999,
+      "learning_rate": 2.0772271876679624e-06,
+      "loss": 0.4659,
+      "step": 5635
+    },
+    {
+      "epoch": 2.796326175996028,
+      "grad_norm": 0.07027972391444388,
+      "learning_rate": 2.0756414572097635e-06,
+      "loss": 0.4347,
+      "step": 5636
+    },
+    {
+      "epoch": 2.796822638699268,
+      "grad_norm": 0.07229760993502869,
+      "learning_rate": 2.074056173692881e-06,
+      "loss": 0.4492,
+      "step": 5637
+    },
+    {
+      "epoch": 2.797319101402507,
+      "grad_norm": 0.07510172712807424,
+      "learning_rate": 2.072471337359599e-06,
+      "loss": 0.467,
+      "step": 5638
+    },
+    {
+      "epoch": 2.7978155641057465,
+      "grad_norm": 0.06966540788364817,
+      "learning_rate": 2.070886948452133e-06,
+      "loss": 0.4342,
+      "step": 5639
+    },
+    {
+      "epoch": 2.798312026808986,
+      "grad_norm": 0.06957085396117565,
+      "learning_rate": 2.069303007212633e-06,
+      "loss": 0.4387,
+      "step": 5640
+    },
+    {
+      "epoch": 2.7988084895122256,
+      "grad_norm": 0.07003352401136151,
+      "learning_rate": 2.067719513883176e-06,
+      "loss": 0.4369,
+      "step": 5641
+    },
+    {
+      "epoch": 2.799304952215465,
+      "grad_norm": 0.07038614205771057,
+      "learning_rate": 2.0661364687057772e-06,
+      "loss": 0.4383,
+      "step": 5642
+    },
+    {
+      "epoch": 2.7998014149187043,
+      "grad_norm": 0.07017671240403632,
+      "learning_rate": 2.0645538719223767e-06,
+      "loss": 0.4514,
+      "step": 5643
+    },
+    {
+      "epoch": 2.8002978776219436,
+      "grad_norm": 0.0747377193940271,
+      "learning_rate": 2.0629717237748526e-06,
+      "loss": 0.4993,
+      "step": 5644
+    },
+    {
+      "epoch": 2.800794340325183,
+      "grad_norm": 0.07037477944829895,
+      "learning_rate": 2.0613900245050083e-06,
+      "loss": 0.4285,
+      "step": 5645
+    },
+    {
+      "epoch": 2.8012908030284223,
+      "grad_norm": 0.0710517615274243,
+      "learning_rate": 2.0598087743545807e-06,
+      "loss": 0.438,
+      "step": 5646
+    },
+    {
+      "epoch": 2.801787265731662,
+      "grad_norm": 0.07301424456740002,
+      "learning_rate": 2.058227973565241e-06,
+      "loss": 0.4335,
+      "step": 5647
+    },
+    {
+      "epoch": 2.8022837284349014,
+      "grad_norm": 0.07214885003404842,
+      "learning_rate": 2.0566476223785857e-06,
+      "loss": 0.4445,
+      "step": 5648
+    },
+    {
+      "epoch": 2.8027801911381407,
+      "grad_norm": 0.0708736347174934,
+      "learning_rate": 2.0550677210361502e-06,
+      "loss": 0.4649,
+      "step": 5649
+    },
+    {
+      "epoch": 2.80327665384138,
+      "grad_norm": 0.07880190272797606,
+      "learning_rate": 2.0534882697793957e-06,
+      "loss": 0.462,
+      "step": 5650
+    },
+    {
+      "epoch": 2.80377311654462,
+      "grad_norm": 0.06952515190967841,
+      "learning_rate": 2.0519092688497133e-06,
+      "loss": 0.4126,
+      "step": 5651
+    },
+    {
+      "epoch": 2.804269579247859,
+      "grad_norm": 0.07423937996352062,
+      "learning_rate": 2.0503307184884313e-06,
+      "loss": 0.4581,
+      "step": 5652
+    },
+    {
+      "epoch": 2.8047660419510985,
+      "grad_norm": 0.07246555572886494,
+      "learning_rate": 2.0487526189368058e-06,
+      "loss": 0.4463,
+      "step": 5653
+    },
+    {
+      "epoch": 2.805262504654338,
+      "grad_norm": 0.07132292131720794,
+      "learning_rate": 2.0471749704360218e-06,
+      "loss": 0.4423,
+      "step": 5654
+    },
+    {
+      "epoch": 2.805758967357577,
+      "grad_norm": 0.07368385174282674,
+      "learning_rate": 2.045597773227199e-06,
+      "loss": 0.4742,
+      "step": 5655
+    },
+    {
+      "epoch": 2.8062554300608165,
+      "grad_norm": 0.07227760152949314,
+      "learning_rate": 2.0440210275513845e-06,
+      "loss": 0.4569,
+      "step": 5656
+    },
+    {
+      "epoch": 2.8067518927640562,
+      "grad_norm": 0.07272838553171634,
+      "learning_rate": 2.042444733649562e-06,
+      "loss": 0.4589,
+      "step": 5657
+    },
+    {
+      "epoch": 2.8072483554672956,
+      "grad_norm": 0.07416783495308385,
+      "learning_rate": 2.0408688917626402e-06,
+      "loss": 0.4945,
+      "step": 5658
+    },
+    {
+      "epoch": 2.807744818170535,
+      "grad_norm": 0.07326226133632513,
+      "learning_rate": 2.039293502131463e-06,
+      "loss": 0.4667,
+      "step": 5659
+    },
+    {
+      "epoch": 2.8082412808737742,
+      "grad_norm": 0.06972396959337102,
+      "learning_rate": 2.0377185649968036e-06,
+      "loss": 0.467,
+      "step": 5660
+    },
+    {
+      "epoch": 2.808737743577014,
+      "grad_norm": 0.07338530096769981,
+      "learning_rate": 2.0361440805993627e-06,
+      "loss": 0.4593,
+      "step": 5661
+    },
+    {
+      "epoch": 2.8092342062802533,
+      "grad_norm": 0.07264115187947948,
+      "learning_rate": 2.0345700491797786e-06,
+      "loss": 0.4579,
+      "step": 5662
+    },
+    {
+      "epoch": 2.8097306689834927,
+      "grad_norm": 0.07200847901130374,
+      "learning_rate": 2.0329964709786144e-06,
+      "loss": 0.473,
+      "step": 5663
+    },
+    {
+      "epoch": 2.810227131686732,
+      "grad_norm": 0.07334938393925998,
+      "learning_rate": 2.0314233462363687e-06,
+      "loss": 0.4704,
+      "step": 5664
+    },
+    {
+      "epoch": 2.8107235943899713,
+      "grad_norm": 0.06806632390588159,
+      "learning_rate": 2.029850675193467e-06,
+      "loss": 0.4235,
+      "step": 5665
+    },
+    {
+      "epoch": 2.8112200570932107,
+      "grad_norm": 0.07441274494262383,
+      "learning_rate": 2.0282784580902655e-06,
+      "loss": 0.4518,
+      "step": 5666
+    },
+    {
+      "epoch": 2.8117165197964504,
+      "grad_norm": 0.07318598053775449,
+      "learning_rate": 2.026706695167055e-06,
+      "loss": 0.4687,
+      "step": 5667
+    },
+    {
+      "epoch": 2.8122129824996898,
+      "grad_norm": 0.07264327682527112,
+      "learning_rate": 2.0251353866640515e-06,
+      "loss": 0.4364,
+      "step": 5668
+    },
+    {
+      "epoch": 2.812709445202929,
+      "grad_norm": 0.07139704661863898,
+      "learning_rate": 2.0235645328214077e-06,
+      "loss": 0.4434,
+      "step": 5669
+    },
+    {
+      "epoch": 2.8132059079061684,
+      "grad_norm": 0.07453629708077483,
+      "learning_rate": 2.0219941338792016e-06,
+      "loss": 0.4633,
+      "step": 5670
+    },
+    {
+      "epoch": 2.813702370609408,
+      "grad_norm": 0.07224548451159646,
+      "learning_rate": 2.0204241900774434e-06,
+      "loss": 0.4775,
+      "step": 5671
+    },
+    {
+      "epoch": 2.8141988333126475,
+      "grad_norm": 0.07199251613284192,
+      "learning_rate": 2.0188547016560738e-06,
+      "loss": 0.4794,
+      "step": 5672
+    },
+    {
+      "epoch": 2.814695296015887,
+      "grad_norm": 0.07081110183530961,
+      "learning_rate": 2.017285668854962e-06,
+      "loss": 0.4782,
+      "step": 5673
+    },
+    {
+      "epoch": 2.815191758719126,
+      "grad_norm": 0.073617482185421,
+      "learning_rate": 2.0157170919139137e-06,
+      "loss": 0.462,
+      "step": 5674
+    },
+    {
+      "epoch": 2.8156882214223655,
+      "grad_norm": 0.07112186816229596,
+      "learning_rate": 2.0141489710726566e-06,
+      "loss": 0.4551,
+      "step": 5675
+    },
+    {
+      "epoch": 2.816184684125605,
+      "grad_norm": 0.07135177337572862,
+      "learning_rate": 2.0125813065708568e-06,
+      "loss": 0.4374,
+      "step": 5676
+    },
+    {
+      "epoch": 2.8166811468288446,
+      "grad_norm": 0.07214896219901143,
+      "learning_rate": 2.0110140986481043e-06,
+      "loss": 0.4553,
+      "step": 5677
+    },
+    {
+      "epoch": 2.817177609532084,
+      "grad_norm": 0.07258899423829851,
+      "learning_rate": 2.00944734754392e-06,
+      "loss": 0.4495,
+      "step": 5678
+    },
+    {
+      "epoch": 2.8176740722353233,
+      "grad_norm": 0.07313560055809436,
+      "learning_rate": 2.007881053497761e-06,
+      "loss": 0.4505,
+      "step": 5679
+    },
+    {
+      "epoch": 2.8181705349385626,
+      "grad_norm": 0.07223163063288415,
+      "learning_rate": 2.006315216749006e-06,
+      "loss": 0.4493,
+      "step": 5680
+    },
+    {
+      "epoch": 2.8186669976418024,
+      "grad_norm": 0.07237609101988172,
+      "learning_rate": 2.004749837536972e-06,
+      "loss": 0.446,
+      "step": 5681
+    },
+    {
+      "epoch": 2.8191634603450417,
+      "grad_norm": 0.07157138426508139,
+      "learning_rate": 2.0031849161009003e-06,
+      "loss": 0.4389,
+      "step": 5682
+    },
+    {
+      "epoch": 2.819659923048281,
+      "grad_norm": 0.06975936253696055,
+      "learning_rate": 2.001620452679962e-06,
+      "loss": 0.4585,
+      "step": 5683
+    },
+    {
+      "epoch": 2.8201563857515204,
+      "grad_norm": 0.07346619156749418,
+      "learning_rate": 2.000056447513264e-06,
+      "loss": 0.4805,
+      "step": 5684
+    },
+    {
+      "epoch": 2.8206528484547597,
+      "grad_norm": 0.07014287296263343,
+      "learning_rate": 1.998492900839836e-06,
+      "loss": 0.4613,
+      "step": 5685
+    },
+    {
+      "epoch": 2.821149311157999,
+      "grad_norm": 0.0706945389906741,
+      "learning_rate": 1.9969298128986457e-06,
+      "loss": 0.4497,
+      "step": 5686
+    },
+    {
+      "epoch": 2.821645773861239,
+      "grad_norm": 0.07091160174834128,
+      "learning_rate": 1.9953671839285794e-06,
+      "loss": 0.4453,
+      "step": 5687
+    },
+    {
+      "epoch": 2.822142236564478,
+      "grad_norm": 0.0726675916967775,
+      "learning_rate": 1.993805014168465e-06,
+      "loss": 0.5102,
+      "step": 5688
+    },
+    {
+      "epoch": 2.8226386992677175,
+      "grad_norm": 0.0717704906392301,
+      "learning_rate": 1.9922433038570544e-06,
+      "loss": 0.4428,
+      "step": 5689
+    },
+    {
+      "epoch": 2.823135161970957,
+      "grad_norm": 0.07092607030772675,
+      "learning_rate": 1.9906820532330262e-06,
+      "loss": 0.4959,
+      "step": 5690
+    },
+    {
+      "epoch": 2.8236316246741966,
+      "grad_norm": 0.07007383114953025,
+      "learning_rate": 1.9891212625349983e-06,
+      "loss": 0.4435,
+      "step": 5691
+    },
+    {
+      "epoch": 2.824128087377436,
+      "grad_norm": 0.0728017778027581,
+      "learning_rate": 1.987560932001509e-06,
+      "loss": 0.4779,
+      "step": 5692
+    },
+    {
+      "epoch": 2.8246245500806753,
+      "grad_norm": 0.07280756722503544,
+      "learning_rate": 1.9860010618710286e-06,
+      "loss": 0.4541,
+      "step": 5693
+    },
+    {
+      "epoch": 2.8251210127839146,
+      "grad_norm": 0.07308516502105063,
+      "learning_rate": 1.9844416523819622e-06,
+      "loss": 0.431,
+      "step": 5694
+    },
+    {
+      "epoch": 2.825617475487154,
+      "grad_norm": 0.07145369691673427,
+      "learning_rate": 1.982882703772636e-06,
+      "loss": 0.4234,
+      "step": 5695
+    },
+    {
+      "epoch": 2.8261139381903932,
+      "grad_norm": 0.07453301298395182,
+      "learning_rate": 1.981324216281315e-06,
+      "loss": 0.4628,
+      "step": 5696
+    },
+    {
+      "epoch": 2.826610400893633,
+      "grad_norm": 0.07203730077033106,
+      "learning_rate": 1.979766190146187e-06,
+      "loss": 0.48,
+      "step": 5697
+    },
+    {
+      "epoch": 2.8271068635968724,
+      "grad_norm": 0.07398813444664835,
+      "learning_rate": 1.978208625605369e-06,
+      "loss": 0.4809,
+      "step": 5698
+    },
+    {
+      "epoch": 2.8276033263001117,
+      "grad_norm": 0.06817625019096056,
+      "learning_rate": 1.9766515228969136e-06,
+      "loss": 0.4405,
+      "step": 5699
+    },
+    {
+      "epoch": 2.828099789003351,
+      "grad_norm": 0.07373964121100986,
+      "learning_rate": 1.9750948822587955e-06,
+      "loss": 0.4928,
+      "step": 5700
+    },
+    {
+      "epoch": 2.828596251706591,
+      "grad_norm": 0.07050400870015364,
+      "learning_rate": 1.973538703928926e-06,
+      "loss": 0.4548,
+      "step": 5701
+    },
+    {
+      "epoch": 2.82909271440983,
+      "grad_norm": 0.0746664124741609,
+      "learning_rate": 1.9719829881451396e-06,
+      "loss": 0.4307,
+      "step": 5702
+    },
+    {
+      "epoch": 2.8295891771130695,
+      "grad_norm": 0.07130220874136327,
+      "learning_rate": 1.9704277351452028e-06,
+      "loss": 0.4581,
+      "step": 5703
+    },
+    {
+      "epoch": 2.830085639816309,
+      "grad_norm": 0.07260529178342427,
+      "learning_rate": 1.9688729451668116e-06,
+      "loss": 0.4329,
+      "step": 5704
+    },
+    {
+      "epoch": 2.830582102519548,
+      "grad_norm": 0.07492902420199271,
+      "learning_rate": 1.9673186184475885e-06,
+      "loss": 0.4686,
+      "step": 5705
+    },
+    {
+      "epoch": 2.8310785652227874,
+      "grad_norm": 0.0714214093617678,
+      "learning_rate": 1.965764755225091e-06,
+      "loss": 0.4387,
+      "step": 5706
+    },
+    {
+      "epoch": 2.831575027926027,
+      "grad_norm": 0.07067957272379431,
+      "learning_rate": 1.964211355736798e-06,
+      "loss": 0.4279,
+      "step": 5707
+    },
+    {
+      "epoch": 2.8320714906292666,
+      "grad_norm": 0.07015666137280017,
+      "learning_rate": 1.9626584202201267e-06,
+      "loss": 0.4546,
+      "step": 5708
+    },
+    {
+      "epoch": 2.832567953332506,
+      "grad_norm": 0.07454452830903732,
+      "learning_rate": 1.961105948912415e-06,
+      "loss": 0.4727,
+      "step": 5709
+    },
+    {
+      "epoch": 2.833064416035745,
+      "grad_norm": 0.07120708361201979,
+      "learning_rate": 1.9595539420509328e-06,
+      "loss": 0.4443,
+      "step": 5710
+    },
+    {
+      "epoch": 2.833560878738985,
+      "grad_norm": 0.07118528149786242,
+      "learning_rate": 1.9580023998728823e-06,
+      "loss": 0.46,
+      "step": 5711
+    },
+    {
+      "epoch": 2.8340573414422243,
+      "grad_norm": 0.07069768707561298,
+      "learning_rate": 1.956451322615389e-06,
+      "loss": 0.4881,
+      "step": 5712
+    },
+    {
+      "epoch": 2.8345538041454637,
+      "grad_norm": 0.07300010654314026,
+      "learning_rate": 1.9549007105155127e-06,
+      "loss": 0.4538,
+      "step": 5713
+    },
+    {
+      "epoch": 2.835050266848703,
+      "grad_norm": 0.07363656559015999,
+      "learning_rate": 1.9533505638102384e-06,
+      "loss": 0.4644,
+      "step": 5714
+    },
+    {
+      "epoch": 2.8355467295519423,
+      "grad_norm": 0.07306036415060045,
+      "learning_rate": 1.951800882736479e-06,
+      "loss": 0.4213,
+      "step": 5715
+    },
+    {
+      "epoch": 2.8360431922551816,
+      "grad_norm": 0.07289854460238754,
+      "learning_rate": 1.9502516675310836e-06,
+      "loss": 0.4724,
+      "step": 5716
+    },
+    {
+      "epoch": 2.836539654958421,
+      "grad_norm": 0.07420132974049734,
+      "learning_rate": 1.9487029184308186e-06,
+      "loss": 0.4756,
+      "step": 5717
+    },
+    {
+      "epoch": 2.8370361176616608,
+      "grad_norm": 0.07423329737411025,
+      "learning_rate": 1.947154635672393e-06,
+      "loss": 0.4725,
+      "step": 5718
+    },
+    {
+      "epoch": 2.8375325803649,
+      "grad_norm": 0.07415959326109783,
+      "learning_rate": 1.945606819492429e-06,
+      "loss": 0.4674,
+      "step": 5719
+    },
+    {
+      "epoch": 2.8380290430681394,
+      "grad_norm": 0.07260651034602289,
+      "learning_rate": 1.9440594701274906e-06,
+      "loss": 0.4796,
+      "step": 5720
+    },
+    {
+      "epoch": 2.838525505771379,
+      "grad_norm": 0.07364464260727976,
+      "learning_rate": 1.9425125878140644e-06,
+      "loss": 0.4597,
+      "step": 5721
+    },
+    {
+      "epoch": 2.8390219684746185,
+      "grad_norm": 0.07212719334038142,
+      "learning_rate": 1.9409661727885638e-06,
+      "loss": 0.4396,
+      "step": 5722
+    },
+    {
+      "epoch": 2.839518431177858,
+      "grad_norm": 0.07636798983029885,
+      "learning_rate": 1.9394202252873377e-06,
+      "loss": 0.4647,
+      "step": 5723
+    },
+    {
+      "epoch": 2.840014893881097,
+      "grad_norm": 0.0703000889908455,
+      "learning_rate": 1.9378747455466563e-06,
+      "loss": 0.4367,
+      "step": 5724
+    },
+    {
+      "epoch": 2.8405113565843365,
+      "grad_norm": 0.07379106009795416,
+      "learning_rate": 1.9363297338027236e-06,
+      "loss": 0.4513,
+      "step": 5725
+    },
+    {
+      "epoch": 2.841007819287576,
+      "grad_norm": 0.07225481677416651,
+      "learning_rate": 1.9347851902916694e-06,
+      "loss": 0.4421,
+      "step": 5726
+    },
+    {
+      "epoch": 2.841504281990815,
+      "grad_norm": 0.07090370027965347,
+      "learning_rate": 1.93324111524955e-06,
+      "loss": 0.4678,
+      "step": 5727
+    },
+    {
+      "epoch": 2.842000744694055,
+      "grad_norm": 0.07410037463754408,
+      "learning_rate": 1.9316975089123556e-06,
+      "loss": 0.462,
+      "step": 5728
+    },
+    {
+      "epoch": 2.8424972073972943,
+      "grad_norm": 0.0731944774344661,
+      "learning_rate": 1.9301543715160014e-06,
+      "loss": 0.4861,
+      "step": 5729
+    },
+    {
+      "epoch": 2.8429936701005336,
+      "grad_norm": 0.07017095928272804,
+      "learning_rate": 1.928611703296328e-06,
+      "loss": 0.4635,
+      "step": 5730
+    },
+    {
+      "epoch": 2.8434901328037734,
+      "grad_norm": 0.0730360434590913,
+      "learning_rate": 1.927069504489112e-06,
+      "loss": 0.503,
+      "step": 5731
+    },
+    {
+      "epoch": 2.8439865955070127,
+      "grad_norm": 0.06964490329233675,
+      "learning_rate": 1.925527775330049e-06,
+      "loss": 0.4186,
+      "step": 5732
+    },
+    {
+      "epoch": 2.844483058210252,
+      "grad_norm": 0.07140910322648998,
+      "learning_rate": 1.923986516054772e-06,
+      "loss": 0.4371,
+      "step": 5733
+    },
+    {
+      "epoch": 2.8449795209134914,
+      "grad_norm": 0.07106555542516287,
+      "learning_rate": 1.9224457268988367e-06,
+      "loss": 0.4386,
+      "step": 5734
+    },
+    {
+      "epoch": 2.8454759836167307,
+      "grad_norm": 0.07498041184595833,
+      "learning_rate": 1.9209054080977262e-06,
+      "loss": 0.449,
+      "step": 5735
+    },
+    {
+      "epoch": 2.84597244631997,
+      "grad_norm": 0.0730042564864519,
+      "learning_rate": 1.9193655598868557e-06,
+      "loss": 0.4714,
+      "step": 5736
+    },
+    {
+      "epoch": 2.8464689090232094,
+      "grad_norm": 0.07411614717666876,
+      "learning_rate": 1.9178261825015625e-06,
+      "loss": 0.4519,
+      "step": 5737
+    },
+    {
+      "epoch": 2.846965371726449,
+      "grad_norm": 0.06889493988420331,
+      "learning_rate": 1.9162872761771207e-06,
+      "loss": 0.4145,
+      "step": 5738
+    },
+    {
+      "epoch": 2.8474618344296885,
+      "grad_norm": 0.07156478169309566,
+      "learning_rate": 1.9147488411487226e-06,
+      "loss": 0.465,
+      "step": 5739
+    },
+    {
+      "epoch": 2.847958297132928,
+      "grad_norm": 0.07360117429747257,
+      "learning_rate": 1.9132108776514985e-06,
+      "loss": 0.4448,
+      "step": 5740
+    },
+    {
+      "epoch": 2.8484547598361676,
+      "grad_norm": 0.07013544132089725,
+      "learning_rate": 1.9116733859204984e-06,
+      "loss": 0.4219,
+      "step": 5741
+    },
+    {
+      "epoch": 2.848951222539407,
+      "grad_norm": 0.07207103402288535,
+      "learning_rate": 1.910136366190702e-06,
+      "loss": 0.4669,
+      "step": 5742
+    },
+    {
+      "epoch": 2.8494476852426462,
+      "grad_norm": 0.07340381981525523,
+      "learning_rate": 1.9085998186970215e-06,
+      "loss": 0.4436,
+      "step": 5743
+    },
+    {
+      "epoch": 2.8499441479458856,
+      "grad_norm": 0.07207559538977827,
+      "learning_rate": 1.9070637436742905e-06,
+      "loss": 0.471,
+      "step": 5744
+    },
+    {
+      "epoch": 2.850440610649125,
+      "grad_norm": 0.07266945487920531,
+      "learning_rate": 1.9055281413572763e-06,
+      "loss": 0.4404,
+      "step": 5745
+    },
+    {
+      "epoch": 2.8509370733523642,
+      "grad_norm": 0.07519234782264786,
+      "learning_rate": 1.9039930119806698e-06,
+      "loss": 0.4387,
+      "step": 5746
+    },
+    {
+      "epoch": 2.8514335360556036,
+      "grad_norm": 0.07715864442431582,
+      "learning_rate": 1.9024583557790889e-06,
+      "loss": 0.4538,
+      "step": 5747
+    },
+    {
+      "epoch": 2.8519299987588433,
+      "grad_norm": 0.07373136234905418,
+      "learning_rate": 1.9009241729870842e-06,
+      "loss": 0.4618,
+      "step": 5748
+    },
+    {
+      "epoch": 2.8524264614620827,
+      "grad_norm": 0.07179951947194625,
+      "learning_rate": 1.8993904638391286e-06,
+      "loss": 0.4335,
+      "step": 5749
+    },
+    {
+      "epoch": 2.852922924165322,
+      "grad_norm": 0.07371238617587633,
+      "learning_rate": 1.8978572285696296e-06,
+      "loss": 0.4597,
+      "step": 5750
+    },
+    {
+      "epoch": 2.8534193868685613,
+      "grad_norm": 0.07073156813998753,
+      "learning_rate": 1.8963244674129104e-06,
+      "loss": 0.4488,
+      "step": 5751
+    },
+    {
+      "epoch": 2.853915849571801,
+      "grad_norm": 0.07732984470628916,
+      "learning_rate": 1.894792180603235e-06,
+      "loss": 0.4575,
+      "step": 5752
+    },
+    {
+      "epoch": 2.8544123122750404,
+      "grad_norm": 0.07187498679301367,
+      "learning_rate": 1.8932603683747858e-06,
+      "loss": 0.4303,
+      "step": 5753
+    },
+    {
+      "epoch": 2.8549087749782798,
+      "grad_norm": 0.07386629483634839,
+      "learning_rate": 1.8917290309616754e-06,
+      "loss": 0.4666,
+      "step": 5754
+    },
+    {
+      "epoch": 2.855405237681519,
+      "grad_norm": 0.0722961968794769,
+      "learning_rate": 1.8901981685979464e-06,
+      "loss": 0.4656,
+      "step": 5755
+    },
+    {
+      "epoch": 2.8559017003847584,
+      "grad_norm": 0.07270464651085981,
+      "learning_rate": 1.8886677815175642e-06,
+      "loss": 0.4698,
+      "step": 5756
+    },
+    {
+      "epoch": 2.8563981630879978,
+      "grad_norm": 0.07533072554757832,
+      "learning_rate": 1.887137869954427e-06,
+      "loss": 0.4578,
+      "step": 5757
+    },
+    {
+      "epoch": 2.8568946257912375,
+      "grad_norm": 0.06979861982970094,
+      "learning_rate": 1.8856084341423552e-06,
+      "loss": 0.503,
+      "step": 5758
+    },
+    {
+      "epoch": 2.857391088494477,
+      "grad_norm": 0.07449849243741927,
+      "learning_rate": 1.884079474315097e-06,
+      "loss": 0.4635,
+      "step": 5759
+    },
+    {
+      "epoch": 2.857887551197716,
+      "grad_norm": 0.07483026512075892,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.4863,
+      "step": 5760
+    },
+    {
+      "epoch": 2.8583840139009555,
+      "grad_norm": 0.07179229556946186,
+      "learning_rate": 1.8810229835496635e-06,
+      "loss": 0.4537,
+      "step": 5761
+    },
+    {
+      "epoch": 2.8588804766041953,
+      "grad_norm": 0.07270191976343439,
+      "learning_rate": 1.8794954530786242e-06,
+      "loss": 0.4632,
+      "step": 5762
+    },
+    {
+      "epoch": 2.8593769393074346,
+      "grad_norm": 0.07087539190876527,
+      "learning_rate": 1.8779683995266712e-06,
+      "loss": 0.4599,
+      "step": 5763
+    },
+    {
+      "epoch": 2.859873402010674,
+      "grad_norm": 0.07380883554497368,
+      "learning_rate": 1.8764418231271885e-06,
+      "loss": 0.4351,
+      "step": 5764
+    },
+    {
+      "epoch": 2.8603698647139133,
+      "grad_norm": 0.07207762992746394,
+      "learning_rate": 1.8749157241134924e-06,
+      "loss": 0.4574,
+      "step": 5765
+    },
+    {
+      "epoch": 2.8608663274171526,
+      "grad_norm": 0.07121802723137254,
+      "learning_rate": 1.8733901027188212e-06,
+      "loss": 0.4715,
+      "step": 5766
+    },
+    {
+      "epoch": 2.861362790120392,
+      "grad_norm": 0.07084904663872903,
+      "learning_rate": 1.8718649591763415e-06,
+      "loss": 0.4782,
+      "step": 5767
+    },
+    {
+      "epoch": 2.8618592528236317,
+      "grad_norm": 0.07167843014443905,
+      "learning_rate": 1.8703402937191467e-06,
+      "loss": 0.4913,
+      "step": 5768
+    },
+    {
+      "epoch": 2.862355715526871,
+      "grad_norm": 0.07151970972114965,
+      "learning_rate": 1.8688161065802563e-06,
+      "loss": 0.4573,
+      "step": 5769
+    },
+    {
+      "epoch": 2.8628521782301104,
+      "grad_norm": 0.07035903364506926,
+      "learning_rate": 1.8672923979926205e-06,
+      "loss": 0.4413,
+      "step": 5770
+    },
+    {
+      "epoch": 2.8633486409333497,
+      "grad_norm": 0.07578753526967076,
+      "learning_rate": 1.865769168189111e-06,
+      "loss": 0.4756,
+      "step": 5771
+    },
+    {
+      "epoch": 2.8638451036365895,
+      "grad_norm": 0.07075700344737111,
+      "learning_rate": 1.8642464174025327e-06,
+      "loss": 0.4449,
+      "step": 5772
+    },
+    {
+      "epoch": 2.864341566339829,
+      "grad_norm": 0.07203092781562426,
+      "learning_rate": 1.862724145865611e-06,
+      "loss": 0.4492,
+      "step": 5773
+    },
+    {
+      "epoch": 2.864838029043068,
+      "grad_norm": 0.06914479180823499,
+      "learning_rate": 1.8612023538109998e-06,
+      "loss": 0.4561,
+      "step": 5774
+    },
+    {
+      "epoch": 2.8653344917463075,
+      "grad_norm": 0.07273419996837144,
+      "learning_rate": 1.8596810414712835e-06,
+      "loss": 0.4722,
+      "step": 5775
+    },
+    {
+      "epoch": 2.865830954449547,
+      "grad_norm": 0.07111294951471243,
+      "learning_rate": 1.8581602090789674e-06,
+      "loss": 0.4595,
+      "step": 5776
+    },
+    {
+      "epoch": 2.866327417152786,
+      "grad_norm": 0.07008046197731649,
+      "learning_rate": 1.8566398568664896e-06,
+      "loss": 0.4373,
+      "step": 5777
+    },
+    {
+      "epoch": 2.866823879856026,
+      "grad_norm": 0.07270960522246615,
+      "learning_rate": 1.8551199850662094e-06,
+      "loss": 0.4753,
+      "step": 5778
+    },
+    {
+      "epoch": 2.8673203425592653,
+      "grad_norm": 0.07023865205596509,
+      "learning_rate": 1.8536005939104135e-06,
+      "loss": 0.4712,
+      "step": 5779
+    },
+    {
+      "epoch": 2.8678168052625046,
+      "grad_norm": 0.07396879722713033,
+      "learning_rate": 1.8520816836313195e-06,
+      "loss": 0.4589,
+      "step": 5780
+    },
+    {
+      "epoch": 2.868313267965744,
+      "grad_norm": 0.07038011700790384,
+      "learning_rate": 1.8505632544610657e-06,
+      "loss": 0.4226,
+      "step": 5781
+    },
+    {
+      "epoch": 2.8688097306689837,
+      "grad_norm": 0.06978356890739894,
+      "learning_rate": 1.8490453066317248e-06,
+      "loss": 0.4428,
+      "step": 5782
+    },
+    {
+      "epoch": 2.869306193372223,
+      "grad_norm": 0.07286599599417821,
+      "learning_rate": 1.8475278403752833e-06,
+      "loss": 0.4756,
+      "step": 5783
+    },
+    {
+      "epoch": 2.8698026560754624,
+      "grad_norm": 0.07421525777471157,
+      "learning_rate": 1.8460108559236673e-06,
+      "loss": 0.4417,
+      "step": 5784
+    },
+    {
+      "epoch": 2.8702991187787017,
+      "grad_norm": 0.07028522021448058,
+      "learning_rate": 1.8444943535087217e-06,
+      "loss": 0.4358,
+      "step": 5785
+    },
+    {
+      "epoch": 2.870795581481941,
+      "grad_norm": 0.0764537827742573,
+      "learning_rate": 1.8429783333622176e-06,
+      "loss": 0.4842,
+      "step": 5786
+    },
+    {
+      "epoch": 2.8712920441851804,
+      "grad_norm": 0.07188922291547878,
+      "learning_rate": 1.8414627957158577e-06,
+      "loss": 0.4432,
+      "step": 5787
+    },
+    {
+      "epoch": 2.87178850688842,
+      "grad_norm": 0.07264028901682451,
+      "learning_rate": 1.8399477408012645e-06,
+      "loss": 0.4695,
+      "step": 5788
+    },
+    {
+      "epoch": 2.8722849695916595,
+      "grad_norm": 0.07189418837208039,
+      "learning_rate": 1.8384331688499934e-06,
+      "loss": 0.4258,
+      "step": 5789
+    },
+    {
+      "epoch": 2.872781432294899,
+      "grad_norm": 0.07198764033425765,
+      "learning_rate": 1.836919080093521e-06,
+      "loss": 0.4401,
+      "step": 5790
+    },
+    {
+      "epoch": 2.873277894998138,
+      "grad_norm": 0.07460868295659728,
+      "learning_rate": 1.8354054747632489e-06,
+      "loss": 0.464,
+      "step": 5791
+    },
+    {
+      "epoch": 2.873774357701378,
+      "grad_norm": 0.07080197418491792,
+      "learning_rate": 1.8338923530905112e-06,
+      "loss": 0.4311,
+      "step": 5792
+    },
+    {
+      "epoch": 2.8742708204046172,
+      "grad_norm": 0.07043644767035809,
+      "learning_rate": 1.832379715306561e-06,
+      "loss": 0.4416,
+      "step": 5793
+    },
+    {
+      "epoch": 2.8747672831078566,
+      "grad_norm": 0.07550125283727059,
+      "learning_rate": 1.8308675616425843e-06,
+      "loss": 0.4471,
+      "step": 5794
+    },
+    {
+      "epoch": 2.875263745811096,
+      "grad_norm": 0.07238050144322952,
+      "learning_rate": 1.8293558923296873e-06,
+      "loss": 0.4458,
+      "step": 5795
+    },
+    {
+      "epoch": 2.8757602085143352,
+      "grad_norm": 0.07071205447465201,
+      "learning_rate": 1.8278447075989037e-06,
+      "loss": 0.4369,
+      "step": 5796
+    },
+    {
+      "epoch": 2.8762566712175746,
+      "grad_norm": 0.07135347695106668,
+      "learning_rate": 1.8263340076811958e-06,
+      "loss": 0.4327,
+      "step": 5797
+    },
+    {
+      "epoch": 2.8767531339208143,
+      "grad_norm": 0.07025016632242698,
+      "learning_rate": 1.8248237928074492e-06,
+      "loss": 0.4237,
+      "step": 5798
+    },
+    {
+      "epoch": 2.8772495966240537,
+      "grad_norm": 0.07701605183732108,
+      "learning_rate": 1.823314063208476e-06,
+      "loss": 0.4931,
+      "step": 5799
+    },
+    {
+      "epoch": 2.877746059327293,
+      "grad_norm": 0.07220406200650448,
+      "learning_rate": 1.8218048191150123e-06,
+      "loss": 0.4396,
+      "step": 5800
+    },
+    {
+      "epoch": 2.8782425220305323,
+      "grad_norm": 0.07287019392934202,
+      "learning_rate": 1.8202960607577246e-06,
+      "loss": 0.4656,
+      "step": 5801
+    },
+    {
+      "epoch": 2.878738984733772,
+      "grad_norm": 0.07524685673051981,
+      "learning_rate": 1.8187877883672024e-06,
+      "loss": 0.5075,
+      "step": 5802
+    },
+    {
+      "epoch": 2.8792354474370114,
+      "grad_norm": 0.07106511484488788,
+      "learning_rate": 1.8172800021739573e-06,
+      "loss": 0.4338,
+      "step": 5803
+    },
+    {
+      "epoch": 2.8797319101402508,
+      "grad_norm": 0.07288049698231516,
+      "learning_rate": 1.8157727024084348e-06,
+      "loss": 0.455,
+      "step": 5804
+    },
+    {
+      "epoch": 2.88022837284349,
+      "grad_norm": 0.07187959245629297,
+      "learning_rate": 1.8142658893009995e-06,
+      "loss": 0.4552,
+      "step": 5805
+    },
+    {
+      "epoch": 2.8807248355467294,
+      "grad_norm": 0.07247759393249832,
+      "learning_rate": 1.8127595630819422e-06,
+      "loss": 0.4909,
+      "step": 5806
+    },
+    {
+      "epoch": 2.8812212982499688,
+      "grad_norm": 0.0721473018210314,
+      "learning_rate": 1.8112537239814836e-06,
+      "loss": 0.4628,
+      "step": 5807
+    },
+    {
+      "epoch": 2.8817177609532085,
+      "grad_norm": 0.0719175610043289,
+      "learning_rate": 1.8097483722297644e-06,
+      "loss": 0.4486,
+      "step": 5808
+    },
+    {
+      "epoch": 2.882214223656448,
+      "grad_norm": 0.07175367367263759,
+      "learning_rate": 1.8082435080568556e-06,
+      "loss": 0.426,
+      "step": 5809
+    },
+    {
+      "epoch": 2.882710686359687,
+      "grad_norm": 0.0746276008814608,
+      "learning_rate": 1.8067391316927514e-06,
+      "loss": 0.4519,
+      "step": 5810
+    },
+    {
+      "epoch": 2.8832071490629265,
+      "grad_norm": 0.06967624512108889,
+      "learning_rate": 1.8052352433673687e-06,
+      "loss": 0.463,
+      "step": 5811
+    },
+    {
+      "epoch": 2.8837036117661663,
+      "grad_norm": 0.07322094442708088,
+      "learning_rate": 1.8037318433105566e-06,
+      "loss": 0.4662,
+      "step": 5812
+    },
+    {
+      "epoch": 2.8842000744694056,
+      "grad_norm": 0.07187306447088909,
+      "learning_rate": 1.8022289317520826e-06,
+      "loss": 0.4726,
+      "step": 5813
+    },
+    {
+      "epoch": 2.884696537172645,
+      "grad_norm": 0.07162376268349631,
+      "learning_rate": 1.800726508921647e-06,
+      "loss": 0.4581,
+      "step": 5814
+    },
+    {
+      "epoch": 2.8851929998758843,
+      "grad_norm": 0.0727599031863759,
+      "learning_rate": 1.799224575048865e-06,
+      "loss": 0.4741,
+      "step": 5815
+    },
+    {
+      "epoch": 2.8856894625791236,
+      "grad_norm": 0.06846531021709305,
+      "learning_rate": 1.797723130363288e-06,
+      "loss": 0.4339,
+      "step": 5816
+    },
+    {
+      "epoch": 2.886185925282363,
+      "grad_norm": 0.0761370981096178,
+      "learning_rate": 1.7962221750943859e-06,
+      "loss": 0.459,
+      "step": 5817
+    },
+    {
+      "epoch": 2.8866823879856027,
+      "grad_norm": 0.07505223732737246,
+      "learning_rate": 1.7947217094715536e-06,
+      "loss": 0.4905,
+      "step": 5818
+    },
+    {
+      "epoch": 2.887178850688842,
+      "grad_norm": 0.07104653840701323,
+      "learning_rate": 1.793221733724117e-06,
+      "loss": 0.4611,
+      "step": 5819
+    },
+    {
+      "epoch": 2.8876753133920814,
+      "grad_norm": 0.07272648271449884,
+      "learning_rate": 1.7917222480813202e-06,
+      "loss": 0.4638,
+      "step": 5820
+    },
+    {
+      "epoch": 2.8881717760953207,
+      "grad_norm": 0.07331371305136444,
+      "learning_rate": 1.7902232527723385e-06,
+      "loss": 0.446,
+      "step": 5821
+    },
+    {
+      "epoch": 2.8886682387985605,
+      "grad_norm": 0.07126558933581785,
+      "learning_rate": 1.7887247480262677e-06,
+      "loss": 0.4548,
+      "step": 5822
+    },
+    {
+      "epoch": 2.8891647015018,
+      "grad_norm": 0.07465013992994197,
+      "learning_rate": 1.7872267340721289e-06,
+      "loss": 0.4363,
+      "step": 5823
+    },
+    {
+      "epoch": 2.889661164205039,
+      "grad_norm": 0.0740872483391922,
+      "learning_rate": 1.7857292111388724e-06,
+      "loss": 0.4726,
+      "step": 5824
+    },
+    {
+      "epoch": 2.8901576269082785,
+      "grad_norm": 0.07407082014718436,
+      "learning_rate": 1.7842321794553674e-06,
+      "loss": 0.4612,
+      "step": 5825
+    },
+    {
+      "epoch": 2.890654089611518,
+      "grad_norm": 0.07602916638072348,
+      "learning_rate": 1.7827356392504142e-06,
+      "loss": 0.4958,
+      "step": 5826
+    },
+    {
+      "epoch": 2.891150552314757,
+      "grad_norm": 0.06968809240417105,
+      "learning_rate": 1.781239590752734e-06,
+      "loss": 0.4284,
+      "step": 5827
+    },
+    {
+      "epoch": 2.891647015017997,
+      "grad_norm": 0.07346959924100413,
+      "learning_rate": 1.7797440341909716e-06,
+      "loss": 0.4841,
+      "step": 5828
+    },
+    {
+      "epoch": 2.8921434777212363,
+      "grad_norm": 0.08088666560812105,
+      "learning_rate": 1.7782489697937027e-06,
+      "loss": 0.5473,
+      "step": 5829
+    },
+    {
+      "epoch": 2.8926399404244756,
+      "grad_norm": 0.07151855750187917,
+      "learning_rate": 1.7767543977894198e-06,
+      "loss": 0.4897,
+      "step": 5830
+    },
+    {
+      "epoch": 2.893136403127715,
+      "grad_norm": 0.07415823417761293,
+      "learning_rate": 1.7752603184065498e-06,
+      "loss": 0.4661,
+      "step": 5831
+    },
+    {
+      "epoch": 2.8936328658309547,
+      "grad_norm": 0.06945222643401203,
+      "learning_rate": 1.7737667318734326e-06,
+      "loss": 0.4221,
+      "step": 5832
+    },
+    {
+      "epoch": 2.894129328534194,
+      "grad_norm": 0.07058770550666034,
+      "learning_rate": 1.7722736384183426e-06,
+      "loss": 0.4423,
+      "step": 5833
+    },
+    {
+      "epoch": 2.8946257912374334,
+      "grad_norm": 0.07479340645321704,
+      "learning_rate": 1.7707810382694745e-06,
+      "loss": 0.4731,
+      "step": 5834
+    },
+    {
+      "epoch": 2.8951222539406727,
+      "grad_norm": 0.07147594140636955,
+      "learning_rate": 1.7692889316549465e-06,
+      "loss": 0.4379,
+      "step": 5835
+    },
+    {
+      "epoch": 2.895618716643912,
+      "grad_norm": 0.07253687881245385,
+      "learning_rate": 1.7677973188028069e-06,
+      "loss": 0.4715,
+      "step": 5836
+    },
+    {
+      "epoch": 2.8961151793471513,
+      "grad_norm": 0.07241853848143014,
+      "learning_rate": 1.7663061999410209e-06,
+      "loss": 0.4785,
+      "step": 5837
+    },
+    {
+      "epoch": 2.896611642050391,
+      "grad_norm": 0.07371362598436333,
+      "learning_rate": 1.7648155752974848e-06,
+      "loss": 0.4683,
+      "step": 5838
+    },
+    {
+      "epoch": 2.8971081047536305,
+      "grad_norm": 0.07216697012999135,
+      "learning_rate": 1.7633254451000164e-06,
+      "loss": 0.447,
+      "step": 5839
+    },
+    {
+      "epoch": 2.89760456745687,
+      "grad_norm": 0.07059782522077165,
+      "learning_rate": 1.761835809576356e-06,
+      "loss": 0.4441,
+      "step": 5840
+    },
+    {
+      "epoch": 2.898101030160109,
+      "grad_norm": 0.07011250857472014,
+      "learning_rate": 1.7603466689541737e-06,
+      "loss": 0.435,
+      "step": 5841
+    },
+    {
+      "epoch": 2.898597492863349,
+      "grad_norm": 0.07215203526460655,
+      "learning_rate": 1.7588580234610592e-06,
+      "loss": 0.4473,
+      "step": 5842
+    },
+    {
+      "epoch": 2.899093955566588,
+      "grad_norm": 0.07211326609340975,
+      "learning_rate": 1.7573698733245258e-06,
+      "loss": 0.4371,
+      "step": 5843
+    },
+    {
+      "epoch": 2.8995904182698276,
+      "grad_norm": 0.07631968305239617,
+      "learning_rate": 1.755882218772018e-06,
+      "loss": 0.4498,
+      "step": 5844
+    },
+    {
+      "epoch": 2.900086880973067,
+      "grad_norm": 0.07312916715153597,
+      "learning_rate": 1.7543950600308957e-06,
+      "loss": 0.4595,
+      "step": 5845
+    },
+    {
+      "epoch": 2.900583343676306,
+      "grad_norm": 0.07114981311857127,
+      "learning_rate": 1.7529083973284506e-06,
+      "loss": 0.4413,
+      "step": 5846
+    },
+    {
+      "epoch": 2.9010798063795455,
+      "grad_norm": 0.07136003851732914,
+      "learning_rate": 1.7514222308918944e-06,
+      "loss": 0.4493,
+      "step": 5847
+    },
+    {
+      "epoch": 2.9015762690827853,
+      "grad_norm": 0.07570826101502601,
+      "learning_rate": 1.7499365609483627e-06,
+      "loss": 0.4815,
+      "step": 5848
+    },
+    {
+      "epoch": 2.9020727317860247,
+      "grad_norm": 0.07242856991285997,
+      "learning_rate": 1.748451387724917e-06,
+      "loss": 0.4437,
+      "step": 5849
+    },
+    {
+      "epoch": 2.902569194489264,
+      "grad_norm": 0.07259034528913118,
+      "learning_rate": 1.74696671144854e-06,
+      "loss": 0.4513,
+      "step": 5850
+    },
+    {
+      "epoch": 2.9030656571925033,
+      "grad_norm": 0.07113950495000526,
+      "learning_rate": 1.745482532346145e-06,
+      "loss": 0.4478,
+      "step": 5851
+    },
+    {
+      "epoch": 2.903562119895743,
+      "grad_norm": 0.0720227909235229,
+      "learning_rate": 1.743998850644561e-06,
+      "loss": 0.4631,
+      "step": 5852
+    },
+    {
+      "epoch": 2.9040585825989824,
+      "grad_norm": 0.07017499253087572,
+      "learning_rate": 1.7425156665705478e-06,
+      "loss": 0.4327,
+      "step": 5853
+    },
+    {
+      "epoch": 2.9045550453022217,
+      "grad_norm": 0.07243409627224551,
+      "learning_rate": 1.741032980350786e-06,
+      "loss": 0.4664,
+      "step": 5854
+    },
+    {
+      "epoch": 2.905051508005461,
+      "grad_norm": 0.06974170123821136,
+      "learning_rate": 1.739550792211877e-06,
+      "loss": 0.4444,
+      "step": 5855
+    },
+    {
+      "epoch": 2.9055479707087004,
+      "grad_norm": 0.07346891443058372,
+      "learning_rate": 1.7380691023803543e-06,
+      "loss": 0.4645,
+      "step": 5856
+    },
+    {
+      "epoch": 2.9060444334119397,
+      "grad_norm": 0.06952419588660884,
+      "learning_rate": 1.7365879110826667e-06,
+      "loss": 0.4535,
+      "step": 5857
+    },
+    {
+      "epoch": 2.906540896115179,
+      "grad_norm": 0.0722978135792198,
+      "learning_rate": 1.7351072185451934e-06,
+      "loss": 0.4418,
+      "step": 5858
+    },
+    {
+      "epoch": 2.907037358818419,
+      "grad_norm": 0.07169213101467879,
+      "learning_rate": 1.7336270249942333e-06,
+      "loss": 0.4442,
+      "step": 5859
+    },
+    {
+      "epoch": 2.907533821521658,
+      "grad_norm": 0.07027625231493947,
+      "learning_rate": 1.7321473306560082e-06,
+      "loss": 0.4422,
+      "step": 5860
+    },
+    {
+      "epoch": 2.9080302842248975,
+      "grad_norm": 0.0724082992248681,
+      "learning_rate": 1.7306681357566695e-06,
+      "loss": 0.4418,
+      "step": 5861
+    },
+    {
+      "epoch": 2.9085267469281373,
+      "grad_norm": 0.07361412412474029,
+      "learning_rate": 1.7291894405222847e-06,
+      "loss": 0.4546,
+      "step": 5862
+    },
+    {
+      "epoch": 2.9090232096313766,
+      "grad_norm": 0.07628467050428009,
+      "learning_rate": 1.7277112451788542e-06,
+      "loss": 0.4687,
+      "step": 5863
+    },
+    {
+      "epoch": 2.909519672334616,
+      "grad_norm": 0.07302537711850288,
+      "learning_rate": 1.7262335499522886e-06,
+      "loss": 0.4471,
+      "step": 5864
+    },
+    {
+      "epoch": 2.9100161350378553,
+      "grad_norm": 0.07516702388732484,
+      "learning_rate": 1.7247563550684366e-06,
+      "loss": 0.4561,
+      "step": 5865
+    },
+    {
+      "epoch": 2.9105125977410946,
+      "grad_norm": 0.07184057747732295,
+      "learning_rate": 1.7232796607530606e-06,
+      "loss": 0.4588,
+      "step": 5866
+    },
+    {
+      "epoch": 2.911009060444334,
+      "grad_norm": 0.06890078743814726,
+      "learning_rate": 1.7218034672318485e-06,
+      "loss": 0.4434,
+      "step": 5867
+    },
+    {
+      "epoch": 2.9115055231475733,
+      "grad_norm": 0.07510715431067853,
+      "learning_rate": 1.7203277747304164e-06,
+      "loss": 0.4912,
+      "step": 5868
+    },
+    {
+      "epoch": 2.912001985850813,
+      "grad_norm": 0.07157713937648363,
+      "learning_rate": 1.718852583474297e-06,
+      "loss": 0.4378,
+      "step": 5869
+    },
+    {
+      "epoch": 2.9124984485540524,
+      "grad_norm": 0.06864218506193251,
+      "learning_rate": 1.7173778936889523e-06,
+      "loss": 0.4438,
+      "step": 5870
+    },
+    {
+      "epoch": 2.9129949112572917,
+      "grad_norm": 0.07437592986907665,
+      "learning_rate": 1.715903705599764e-06,
+      "loss": 0.4704,
+      "step": 5871
+    },
+    {
+      "epoch": 2.9134913739605315,
+      "grad_norm": 0.0733699303510351,
+      "learning_rate": 1.7144300194320357e-06,
+      "loss": 0.4618,
+      "step": 5872
+    },
+    {
+      "epoch": 2.913987836663771,
+      "grad_norm": 0.0718682272644226,
+      "learning_rate": 1.712956835411001e-06,
+      "loss": 0.4473,
+      "step": 5873
+    },
+    {
+      "epoch": 2.91448429936701,
+      "grad_norm": 0.07602899319234606,
+      "learning_rate": 1.7114841537618081e-06,
+      "loss": 0.4771,
+      "step": 5874
+    },
+    {
+      "epoch": 2.9149807620702495,
+      "grad_norm": 0.07309674636622461,
+      "learning_rate": 1.7100119747095372e-06,
+      "loss": 0.4348,
+      "step": 5875
+    },
+    {
+      "epoch": 2.915477224773489,
+      "grad_norm": 0.07057419622476488,
+      "learning_rate": 1.7085402984791848e-06,
+      "loss": 0.4342,
+      "step": 5876
+    },
+    {
+      "epoch": 2.915973687476728,
+      "grad_norm": 0.07101692739372809,
+      "learning_rate": 1.707069125295671e-06,
+      "loss": 0.4413,
+      "step": 5877
+    },
+    {
+      "epoch": 2.9164701501799675,
+      "grad_norm": 0.06989251553843452,
+      "learning_rate": 1.7055984553838455e-06,
+      "loss": 0.4089,
+      "step": 5878
+    },
+    {
+      "epoch": 2.9169666128832072,
+      "grad_norm": 0.06969315355799537,
+      "learning_rate": 1.7041282889684746e-06,
+      "loss": 0.4518,
+      "step": 5879
+    },
+    {
+      "epoch": 2.9174630755864466,
+      "grad_norm": 0.07050586361392111,
+      "learning_rate": 1.702658626274249e-06,
+      "loss": 0.4643,
+      "step": 5880
+    },
+    {
+      "epoch": 2.917959538289686,
+      "grad_norm": 0.0703534380162176,
+      "learning_rate": 1.701189467525784e-06,
+      "loss": 0.4701,
+      "step": 5881
+    },
+    {
+      "epoch": 2.9184560009929257,
+      "grad_norm": 0.07285478802365085,
+      "learning_rate": 1.6997208129476144e-06,
+      "loss": 0.4818,
+      "step": 5882
+    },
+    {
+      "epoch": 2.918952463696165,
+      "grad_norm": 0.0713570994512701,
+      "learning_rate": 1.6982526627642043e-06,
+      "loss": 0.4396,
+      "step": 5883
+    },
+    {
+      "epoch": 2.9194489263994043,
+      "grad_norm": 0.07084619268097336,
+      "learning_rate": 1.6967850171999334e-06,
+      "loss": 0.4483,
+      "step": 5884
+    },
+    {
+      "epoch": 2.9199453891026437,
+      "grad_norm": 0.07312699322762967,
+      "learning_rate": 1.6953178764791116e-06,
+      "loss": 0.5011,
+      "step": 5885
+    },
+    {
+      "epoch": 2.920441851805883,
+      "grad_norm": 0.06913757556808983,
+      "learning_rate": 1.6938512408259655e-06,
+      "loss": 0.4439,
+      "step": 5886
+    },
+    {
+      "epoch": 2.9209383145091223,
+      "grad_norm": 0.07112579876663329,
+      "learning_rate": 1.6923851104646461e-06,
+      "loss": 0.4515,
+      "step": 5887
+    },
+    {
+      "epoch": 2.9214347772123617,
+      "grad_norm": 0.07114918399378936,
+      "learning_rate": 1.69091948561923e-06,
+      "loss": 0.4366,
+      "step": 5888
+    },
+    {
+      "epoch": 2.9219312399156014,
+      "grad_norm": 0.07312432271511711,
+      "learning_rate": 1.689454366513712e-06,
+      "loss": 0.4458,
+      "step": 5889
+    },
+    {
+      "epoch": 2.9224277026188408,
+      "grad_norm": 0.07083664164054572,
+      "learning_rate": 1.6879897533720151e-06,
+      "loss": 0.4653,
+      "step": 5890
+    },
+    {
+      "epoch": 2.92292416532208,
+      "grad_norm": 0.07591038722612382,
+      "learning_rate": 1.6865256464179808e-06,
+      "loss": 0.468,
+      "step": 5891
+    },
+    {
+      "epoch": 2.9234206280253194,
+      "grad_norm": 0.07074166751616634,
+      "learning_rate": 1.685062045875372e-06,
+      "loss": 0.4627,
+      "step": 5892
+    },
+    {
+      "epoch": 2.923917090728559,
+      "grad_norm": 0.07313498112061012,
+      "learning_rate": 1.6835989519678802e-06,
+      "loss": 0.4496,
+      "step": 5893
+    },
+    {
+      "epoch": 2.9244135534317985,
+      "grad_norm": 0.0703811251728174,
+      "learning_rate": 1.682136364919112e-06,
+      "loss": 0.4339,
+      "step": 5894
+    },
+    {
+      "epoch": 2.924910016135038,
+      "grad_norm": 0.07137372235534829,
+      "learning_rate": 1.6806742849526064e-06,
+      "loss": 0.435,
+      "step": 5895
+    },
+    {
+      "epoch": 2.925406478838277,
+      "grad_norm": 0.07235141275487113,
+      "learning_rate": 1.6792127122918116e-06,
+      "loss": 0.5001,
+      "step": 5896
+    },
+    {
+      "epoch": 2.9259029415415165,
+      "grad_norm": 0.07145659685400398,
+      "learning_rate": 1.6777516471601103e-06,
+      "loss": 0.4395,
+      "step": 5897
+    },
+    {
+      "epoch": 2.926399404244756,
+      "grad_norm": 0.0739168525849856,
+      "learning_rate": 1.6762910897808017e-06,
+      "loss": 0.4632,
+      "step": 5898
+    },
+    {
+      "epoch": 2.9268958669479956,
+      "grad_norm": 0.07021567224806549,
+      "learning_rate": 1.6748310403771067e-06,
+      "loss": 0.4623,
+      "step": 5899
+    },
+    {
+      "epoch": 2.927392329651235,
+      "grad_norm": 0.07010595995431612,
+      "learning_rate": 1.6733714991721738e-06,
+      "loss": 0.4528,
+      "step": 5900
+    },
+    {
+      "epoch": 2.9278887923544743,
+      "grad_norm": 0.07479842189450096,
+      "learning_rate": 1.6719124663890674e-06,
+      "loss": 0.4643,
+      "step": 5901
+    },
+    {
+      "epoch": 2.9283852550577136,
+      "grad_norm": 0.07251917648131366,
+      "learning_rate": 1.6704539422507803e-06,
+      "loss": 0.4624,
+      "step": 5902
+    },
+    {
+      "epoch": 2.9288817177609534,
+      "grad_norm": 0.07173529330807134,
+      "learning_rate": 1.668995926980223e-06,
+      "loss": 0.4721,
+      "step": 5903
+    },
+    {
+      "epoch": 2.9293781804641927,
+      "grad_norm": 0.07271552046391146,
+      "learning_rate": 1.6675384208002275e-06,
+      "loss": 0.473,
+      "step": 5904
+    },
+    {
+      "epoch": 2.929874643167432,
+      "grad_norm": 0.07226439840476027,
+      "learning_rate": 1.666081423933555e-06,
+      "loss": 0.4586,
+      "step": 5905
+    },
+    {
+      "epoch": 2.9303711058706714,
+      "grad_norm": 0.0714342130930368,
+      "learning_rate": 1.6646249366028788e-06,
+      "loss": 0.4581,
+      "step": 5906
+    },
+    {
+      "epoch": 2.9308675685739107,
+      "grad_norm": 0.06945596117253568,
+      "learning_rate": 1.6631689590308049e-06,
+      "loss": 0.4604,
+      "step": 5907
+    },
+    {
+      "epoch": 2.93136403127715,
+      "grad_norm": 0.07326856612947807,
+      "learning_rate": 1.661713491439853e-06,
+      "loss": 0.4557,
+      "step": 5908
+    },
+    {
+      "epoch": 2.93186049398039,
+      "grad_norm": 0.07050725043012718,
+      "learning_rate": 1.6602585340524669e-06,
+      "loss": 0.4416,
+      "step": 5909
+    },
+    {
+      "epoch": 2.932356956683629,
+      "grad_norm": 0.07050986823414582,
+      "learning_rate": 1.658804087091017e-06,
+      "loss": 0.4476,
+      "step": 5910
+    },
+    {
+      "epoch": 2.9328534193868685,
+      "grad_norm": 0.07159943014698252,
+      "learning_rate": 1.6573501507777906e-06,
+      "loss": 0.4698,
+      "step": 5911
+    },
+    {
+      "epoch": 2.933349882090108,
+      "grad_norm": 0.07083310107065469,
+      "learning_rate": 1.6558967253349983e-06,
+      "loss": 0.464,
+      "step": 5912
+    },
+    {
+      "epoch": 2.9338463447933476,
+      "grad_norm": 0.06952388035745817,
+      "learning_rate": 1.654443810984771e-06,
+      "loss": 0.4147,
+      "step": 5913
+    },
+    {
+      "epoch": 2.934342807496587,
+      "grad_norm": 0.07059058804372653,
+      "learning_rate": 1.652991407949167e-06,
+      "loss": 0.4576,
+      "step": 5914
+    },
+    {
+      "epoch": 2.9348392701998263,
+      "grad_norm": 0.07134194094863748,
+      "learning_rate": 1.6515395164501613e-06,
+      "loss": 0.4631,
+      "step": 5915
+    },
+    {
+      "epoch": 2.9353357329030656,
+      "grad_norm": 0.07335501990274129,
+      "learning_rate": 1.6500881367096506e-06,
+      "loss": 0.4772,
+      "step": 5916
+    },
+    {
+      "epoch": 2.935832195606305,
+      "grad_norm": 0.07235343050752824,
+      "learning_rate": 1.6486372689494573e-06,
+      "loss": 0.4827,
+      "step": 5917
+    },
+    {
+      "epoch": 2.9363286583095443,
+      "grad_norm": 0.07111464655511497,
+      "learning_rate": 1.6471869133913232e-06,
+      "loss": 0.4351,
+      "step": 5918
+    },
+    {
+      "epoch": 2.936825121012784,
+      "grad_norm": 0.07387778721142053,
+      "learning_rate": 1.6457370702569093e-06,
+      "loss": 0.4706,
+      "step": 5919
+    },
+    {
+      "epoch": 2.9373215837160234,
+      "grad_norm": 0.07393542980680054,
+      "learning_rate": 1.6442877397678042e-06,
+      "loss": 0.4569,
+      "step": 5920
+    },
+    {
+      "epoch": 2.9378180464192627,
+      "grad_norm": 0.0720364009150225,
+      "learning_rate": 1.6428389221455115e-06,
+      "loss": 0.454,
+      "step": 5921
+    },
+    {
+      "epoch": 2.938314509122502,
+      "grad_norm": 0.07412971708528018,
+      "learning_rate": 1.6413906176114636e-06,
+      "loss": 0.4657,
+      "step": 5922
+    },
+    {
+      "epoch": 2.938810971825742,
+      "grad_norm": 0.07287513599920468,
+      "learning_rate": 1.6399428263870082e-06,
+      "loss": 0.4298,
+      "step": 5923
+    },
+    {
+      "epoch": 2.939307434528981,
+      "grad_norm": 0.07213379859066761,
+      "learning_rate": 1.6384955486934157e-06,
+      "loss": 0.4532,
+      "step": 5924
+    },
+    {
+      "epoch": 2.9398038972322205,
+      "grad_norm": 0.07156534659604881,
+      "learning_rate": 1.6370487847518829e-06,
+      "loss": 0.4687,
+      "step": 5925
+    },
+    {
+      "epoch": 2.94030035993546,
+      "grad_norm": 0.07205669107853233,
+      "learning_rate": 1.6356025347835209e-06,
+      "loss": 0.4455,
+      "step": 5926
+    },
+    {
+      "epoch": 2.940796822638699,
+      "grad_norm": 0.07571569336289337,
+      "learning_rate": 1.6341567990093704e-06,
+      "loss": 0.463,
+      "step": 5927
+    },
+    {
+      "epoch": 2.9412932853419385,
+      "grad_norm": 0.07239860011491428,
+      "learning_rate": 1.6327115776503833e-06,
+      "loss": 0.4341,
+      "step": 5928
+    },
+    {
+      "epoch": 2.9417897480451782,
+      "grad_norm": 0.0741988580326704,
+      "learning_rate": 1.631266870927442e-06,
+      "loss": 0.4464,
+      "step": 5929
+    },
+    {
+      "epoch": 2.9422862107484176,
+      "grad_norm": 0.07056138848741017,
+      "learning_rate": 1.6298226790613464e-06,
+      "loss": 0.459,
+      "step": 5930
+    },
+    {
+      "epoch": 2.942782673451657,
+      "grad_norm": 0.07345289330859334,
+      "learning_rate": 1.6283790022728164e-06,
+      "loss": 0.4449,
+      "step": 5931
+    },
+    {
+      "epoch": 2.9432791361548962,
+      "grad_norm": 0.07104898966124364,
+      "learning_rate": 1.626935840782497e-06,
+      "loss": 0.4536,
+      "step": 5932
+    },
+    {
+      "epoch": 2.943775598858136,
+      "grad_norm": 0.07431978745376211,
+      "learning_rate": 1.6254931948109498e-06,
+      "loss": 0.4765,
+      "step": 5933
+    },
+    {
+      "epoch": 2.9442720615613753,
+      "grad_norm": 0.07430155293124209,
+      "learning_rate": 1.6240510645786639e-06,
+      "loss": 0.5009,
+      "step": 5934
+    },
+    {
+      "epoch": 2.9447685242646147,
+      "grad_norm": 0.07254532970085129,
+      "learning_rate": 1.622609450306043e-06,
+      "loss": 0.4661,
+      "step": 5935
+    },
+    {
+      "epoch": 2.945264986967854,
+      "grad_norm": 0.07190075269617335,
+      "learning_rate": 1.6211683522134136e-06,
+      "loss": 0.4431,
+      "step": 5936
+    },
+    {
+      "epoch": 2.9457614496710933,
+      "grad_norm": 0.07219291405916406,
+      "learning_rate": 1.6197277705210278e-06,
+      "loss": 0.4528,
+      "step": 5937
+    },
+    {
+      "epoch": 2.9462579123743327,
+      "grad_norm": 0.07069487209787771,
+      "learning_rate": 1.6182877054490526e-06,
+      "loss": 0.4334,
+      "step": 5938
+    },
+    {
+      "epoch": 2.9467543750775724,
+      "grad_norm": 0.07109599185929143,
+      "learning_rate": 1.6168481572175814e-06,
+      "loss": 0.454,
+      "step": 5939
+    },
+    {
+      "epoch": 2.9472508377808118,
+      "grad_norm": 0.07409010851854753,
+      "learning_rate": 1.6154091260466242e-06,
+      "loss": 0.4723,
+      "step": 5940
+    },
+    {
+      "epoch": 2.947747300484051,
+      "grad_norm": 0.07691475767500006,
+      "learning_rate": 1.6139706121561133e-06,
+      "loss": 0.4499,
+      "step": 5941
+    },
+    {
+      "epoch": 2.9482437631872904,
+      "grad_norm": 0.07057759341189006,
+      "learning_rate": 1.6125326157659048e-06,
+      "loss": 0.4875,
+      "step": 5942
+    },
+    {
+      "epoch": 2.94874022589053,
+      "grad_norm": 0.07207202143271935,
+      "learning_rate": 1.6110951370957723e-06,
+      "loss": 0.464,
+      "step": 5943
+    },
+    {
+      "epoch": 2.9492366885937695,
+      "grad_norm": 0.07171710857679257,
+      "learning_rate": 1.6096581763654106e-06,
+      "loss": 0.4791,
+      "step": 5944
+    },
+    {
+      "epoch": 2.949733151297009,
+      "grad_norm": 0.07025891311039467,
+      "learning_rate": 1.6082217337944357e-06,
+      "loss": 0.4558,
+      "step": 5945
+    },
+    {
+      "epoch": 2.950229614000248,
+      "grad_norm": 0.07092578727672147,
+      "learning_rate": 1.6067858096023869e-06,
+      "loss": 0.4495,
+      "step": 5946
+    },
+    {
+      "epoch": 2.9507260767034875,
+      "grad_norm": 0.0701539238256779,
+      "learning_rate": 1.6053504040087208e-06,
+      "loss": 0.4527,
+      "step": 5947
+    },
+    {
+      "epoch": 2.951222539406727,
+      "grad_norm": 0.07164960976189151,
+      "learning_rate": 1.6039155172328153e-06,
+      "loss": 0.4314,
+      "step": 5948
+    },
+    {
+      "epoch": 2.9517190021099666,
+      "grad_norm": 0.07069326099764968,
+      "learning_rate": 1.6024811494939723e-06,
+      "loss": 0.454,
+      "step": 5949
+    },
+    {
+      "epoch": 2.952215464813206,
+      "grad_norm": 0.07281057872718284,
+      "learning_rate": 1.601047301011409e-06,
+      "loss": 0.4546,
+      "step": 5950
+    },
+    {
+      "epoch": 2.9527119275164453,
+      "grad_norm": 0.07337024104844138,
+      "learning_rate": 1.5996139720042692e-06,
+      "loss": 0.4763,
+      "step": 5951
+    },
+    {
+      "epoch": 2.9532083902196846,
+      "grad_norm": 0.07255991546812811,
+      "learning_rate": 1.5981811626916126e-06,
+      "loss": 0.4956,
+      "step": 5952
+    },
+    {
+      "epoch": 2.9537048529229244,
+      "grad_norm": 0.07118427601670174,
+      "learning_rate": 1.5967488732924202e-06,
+      "loss": 0.4625,
+      "step": 5953
+    },
+    {
+      "epoch": 2.9542013156261637,
+      "grad_norm": 0.0705626273726107,
+      "learning_rate": 1.5953171040255965e-06,
+      "loss": 0.4526,
+      "step": 5954
+    },
+    {
+      "epoch": 2.954697778329403,
+      "grad_norm": 0.07193505828742014,
+      "learning_rate": 1.5938858551099639e-06,
+      "loss": 0.4391,
+      "step": 5955
+    },
+    {
+      "epoch": 2.9551942410326424,
+      "grad_norm": 0.07037932237490879,
+      "learning_rate": 1.5924551267642641e-06,
+      "loss": 0.4617,
+      "step": 5956
+    },
+    {
+      "epoch": 2.9556907037358817,
+      "grad_norm": 0.07160562539684742,
+      "learning_rate": 1.5910249192071637e-06,
+      "loss": 0.4643,
+      "step": 5957
+    },
+    {
+      "epoch": 2.956187166439121,
+      "grad_norm": 0.07359941219396972,
+      "learning_rate": 1.5895952326572438e-06,
+      "loss": 0.5014,
+      "step": 5958
+    },
+    {
+      "epoch": 2.956683629142361,
+      "grad_norm": 0.07216612036347472,
+      "learning_rate": 1.5881660673330141e-06,
+      "loss": 0.4807,
+      "step": 5959
+    },
+    {
+      "epoch": 2.9571800918456,
+      "grad_norm": 0.07016235803755881,
+      "learning_rate": 1.5867374234528938e-06,
+      "loss": 0.4389,
+      "step": 5960
+    },
+    {
+      "epoch": 2.9576765545488395,
+      "grad_norm": 0.0743924650931678,
+      "learning_rate": 1.5853093012352317e-06,
+      "loss": 0.4759,
+      "step": 5961
+    },
+    {
+      "epoch": 2.958173017252079,
+      "grad_norm": 0.07307681745982117,
+      "learning_rate": 1.5838817008982927e-06,
+      "loss": 0.4666,
+      "step": 5962
+    },
+    {
+      "epoch": 2.9586694799553186,
+      "grad_norm": 0.07148307237166242,
+      "learning_rate": 1.5824546226602611e-06,
+      "loss": 0.4297,
+      "step": 5963
+    },
+    {
+      "epoch": 2.959165942658558,
+      "grad_norm": 0.0733250074435015,
+      "learning_rate": 1.5810280667392458e-06,
+      "loss": 0.4792,
+      "step": 5964
+    },
+    {
+      "epoch": 2.9596624053617973,
+      "grad_norm": 0.07349016981862934,
+      "learning_rate": 1.5796020333532696e-06,
+      "loss": 0.4583,
+      "step": 5965
+    },
+    {
+      "epoch": 2.9601588680650366,
+      "grad_norm": 0.07165327067178426,
+      "learning_rate": 1.5781765227202822e-06,
+      "loss": 0.4659,
+      "step": 5966
+    },
+    {
+      "epoch": 2.960655330768276,
+      "grad_norm": 0.07070030003800495,
+      "learning_rate": 1.5767515350581492e-06,
+      "loss": 0.4535,
+      "step": 5967
+    },
+    {
+      "epoch": 2.9611517934715152,
+      "grad_norm": 0.06980187457227569,
+      "learning_rate": 1.575327070584654e-06,
+      "loss": 0.3996,
+      "step": 5968
+    },
+    {
+      "epoch": 2.961648256174755,
+      "grad_norm": 0.07050217628886313,
+      "learning_rate": 1.5739031295175078e-06,
+      "loss": 0.457,
+      "step": 5969
+    },
+    {
+      "epoch": 2.9621447188779944,
+      "grad_norm": 0.07083558784751204,
+      "learning_rate": 1.572479712074333e-06,
+      "loss": 0.4465,
+      "step": 5970
+    },
+    {
+      "epoch": 2.9626411815812337,
+      "grad_norm": 0.07305312265469574,
+      "learning_rate": 1.5710568184726799e-06,
+      "loss": 0.4616,
+      "step": 5971
+    },
+    {
+      "epoch": 2.963137644284473,
+      "grad_norm": 0.07297673470614084,
+      "learning_rate": 1.569634448930013e-06,
+      "loss": 0.465,
+      "step": 5972
+    },
+    {
+      "epoch": 2.963634106987713,
+      "grad_norm": 0.07431334323820772,
+      "learning_rate": 1.5682126036637174e-06,
+      "loss": 0.4777,
+      "step": 5973
+    },
+    {
+      "epoch": 2.964130569690952,
+      "grad_norm": 0.07221414404814905,
+      "learning_rate": 1.5667912828911025e-06,
+      "loss": 0.4685,
+      "step": 5974
+    },
+    {
+      "epoch": 2.9646270323941915,
+      "grad_norm": 0.0709090572935716,
+      "learning_rate": 1.5653704868293928e-06,
+      "loss": 0.4347,
+      "step": 5975
+    },
+    {
+      "epoch": 2.965123495097431,
+      "grad_norm": 0.07158548900446969,
+      "learning_rate": 1.5639502156957337e-06,
+      "loss": 0.457,
+      "step": 5976
+    },
+    {
+      "epoch": 2.96561995780067,
+      "grad_norm": 0.06958496316397475,
+      "learning_rate": 1.5625304697071897e-06,
+      "loss": 0.4617,
+      "step": 5977
+    },
+    {
+      "epoch": 2.9661164205039094,
+      "grad_norm": 0.06999282640269491,
+      "learning_rate": 1.5611112490807496e-06,
+      "loss": 0.4293,
+      "step": 5978
+    },
+    {
+      "epoch": 2.966612883207149,
+      "grad_norm": 0.06936000279838299,
+      "learning_rate": 1.559692554033317e-06,
+      "loss": 0.4229,
+      "step": 5979
+    },
+    {
+      "epoch": 2.9671093459103886,
+      "grad_norm": 0.07260037421289114,
+      "learning_rate": 1.5582743847817138e-06,
+      "loss": 0.4839,
+      "step": 5980
+    },
+    {
+      "epoch": 2.967605808613628,
+      "grad_norm": 0.07236131524189705,
+      "learning_rate": 1.5568567415426893e-06,
+      "loss": 0.4806,
+      "step": 5981
+    },
+    {
+      "epoch": 2.968102271316867,
+      "grad_norm": 0.07341692924894432,
+      "learning_rate": 1.555439624532904e-06,
+      "loss": 0.4724,
+      "step": 5982
+    },
+    {
+      "epoch": 2.968598734020107,
+      "grad_norm": 0.07241636608277653,
+      "learning_rate": 1.5540230339689437e-06,
+      "loss": 0.4436,
+      "step": 5983
+    },
+    {
+      "epoch": 2.9690951967233463,
+      "grad_norm": 0.07406946952048966,
+      "learning_rate": 1.5526069700673108e-06,
+      "loss": 0.4581,
+      "step": 5984
+    },
+    {
+      "epoch": 2.9695916594265857,
+      "grad_norm": 0.07302094613595563,
+      "learning_rate": 1.551191433044426e-06,
+      "loss": 0.4616,
+      "step": 5985
+    },
+    {
+      "epoch": 2.970088122129825,
+      "grad_norm": 0.07019743543659121,
+      "learning_rate": 1.549776423116635e-06,
+      "loss": 0.4505,
+      "step": 5986
+    },
+    {
+      "epoch": 2.9705845848330643,
+      "grad_norm": 0.07316166105045384,
+      "learning_rate": 1.5483619405001965e-06,
+      "loss": 0.4403,
+      "step": 5987
+    },
+    {
+      "epoch": 2.9710810475363036,
+      "grad_norm": 0.07297545044125849,
+      "learning_rate": 1.5469479854112934e-06,
+      "loss": 0.4729,
+      "step": 5988
+    },
+    {
+      "epoch": 2.9715775102395434,
+      "grad_norm": 0.07282454195737302,
+      "learning_rate": 1.5455345580660259e-06,
+      "loss": 0.439,
+      "step": 5989
+    },
+    {
+      "epoch": 2.9720739729427827,
+      "grad_norm": 0.07479071932810989,
+      "learning_rate": 1.544121658680411e-06,
+      "loss": 0.4816,
+      "step": 5990
+    },
+    {
+      "epoch": 2.972570435646022,
+      "grad_norm": 0.07439592386649715,
+      "learning_rate": 1.542709287470393e-06,
+      "loss": 0.4629,
+      "step": 5991
+    },
+    {
+      "epoch": 2.9730668983492614,
+      "grad_norm": 0.07067142551209074,
+      "learning_rate": 1.5412974446518243e-06,
+      "loss": 0.4465,
+      "step": 5992
+    },
+    {
+      "epoch": 2.973563361052501,
+      "grad_norm": 0.07345460653832109,
+      "learning_rate": 1.539886130440486e-06,
+      "loss": 0.4693,
+      "step": 5993
+    },
+    {
+      "epoch": 2.9740598237557405,
+      "grad_norm": 0.07332983971617145,
+      "learning_rate": 1.5384753450520739e-06,
+      "loss": 0.4377,
+      "step": 5994
+    },
+    {
+      "epoch": 2.97455628645898,
+      "grad_norm": 0.07165864927196039,
+      "learning_rate": 1.537065088702203e-06,
+      "loss": 0.446,
+      "step": 5995
+    },
+    {
+      "epoch": 2.975052749162219,
+      "grad_norm": 0.07460898613291474,
+      "learning_rate": 1.5356553616064107e-06,
+      "loss": 0.4678,
+      "step": 5996
+    },
+    {
+      "epoch": 2.9755492118654585,
+      "grad_norm": 0.07235821467864646,
+      "learning_rate": 1.5342461639801481e-06,
+      "loss": 0.4672,
+      "step": 5997
+    },
+    {
+      "epoch": 2.976045674568698,
+      "grad_norm": 0.07151717029488126,
+      "learning_rate": 1.532837496038792e-06,
+      "loss": 0.4513,
+      "step": 5998
+    },
+    {
+      "epoch": 2.976542137271937,
+      "grad_norm": 0.07072519813246403,
+      "learning_rate": 1.531429357997633e-06,
+      "loss": 0.4461,
+      "step": 5999
+    },
+    {
+      "epoch": 2.977038599975177,
+      "grad_norm": 0.07209317328064743,
+      "learning_rate": 1.5300217500718806e-06,
+      "loss": 0.4505,
+      "step": 6000
+    },
+    {
+      "epoch": 2.9775350626784163,
+      "grad_norm": 0.07019272634029788,
+      "learning_rate": 1.5286146724766681e-06,
+      "loss": 0.441,
+      "step": 6001
+    },
+    {
+      "epoch": 2.9780315253816556,
+      "grad_norm": 0.07047003562021524,
+      "learning_rate": 1.5272081254270421e-06,
+      "loss": 0.4696,
+      "step": 6002
+    },
+    {
+      "epoch": 2.9785279880848954,
+      "grad_norm": 0.0714175852136517,
+      "learning_rate": 1.5258021091379738e-06,
+      "loss": 0.4593,
+      "step": 6003
+    },
+    {
+      "epoch": 2.9790244507881347,
+      "grad_norm": 0.06983063937057955,
+      "learning_rate": 1.5243966238243484e-06,
+      "loss": 0.4244,
+      "step": 6004
+    },
+    {
+      "epoch": 2.979520913491374,
+      "grad_norm": 0.07233901190332044,
+      "learning_rate": 1.5229916697009706e-06,
+      "loss": 0.4704,
+      "step": 6005
+    },
+    {
+      "epoch": 2.9800173761946134,
+      "grad_norm": 0.0710352377551852,
+      "learning_rate": 1.5215872469825682e-06,
+      "loss": 0.4429,
+      "step": 6006
+    },
+    {
+      "epoch": 2.9805138388978527,
+      "grad_norm": 0.07279079307995248,
+      "learning_rate": 1.520183355883783e-06,
+      "loss": 0.4928,
+      "step": 6007
+    },
+    {
+      "epoch": 2.981010301601092,
+      "grad_norm": 0.072720773924886,
+      "learning_rate": 1.5187799966191769e-06,
+      "loss": 0.453,
+      "step": 6008
+    },
+    {
+      "epoch": 2.9815067643043314,
+      "grad_norm": 0.07541142329483538,
+      "learning_rate": 1.5173771694032296e-06,
+      "loss": 0.4541,
+      "step": 6009
+    },
+    {
+      "epoch": 2.982003227007571,
+      "grad_norm": 0.07134543270914809,
+      "learning_rate": 1.5159748744503444e-06,
+      "loss": 0.465,
+      "step": 6010
+    },
+    {
+      "epoch": 2.9824996897108105,
+      "grad_norm": 0.07153087313411305,
+      "learning_rate": 1.5145731119748376e-06,
+      "loss": 0.4563,
+      "step": 6011
+    },
+    {
+      "epoch": 2.98299615241405,
+      "grad_norm": 0.07466590611705792,
+      "learning_rate": 1.5131718821909435e-06,
+      "loss": 0.4896,
+      "step": 6012
+    },
+    {
+      "epoch": 2.9834926151172896,
+      "grad_norm": 0.07143816309455199,
+      "learning_rate": 1.5117711853128225e-06,
+      "loss": 0.4649,
+      "step": 6013
+    },
+    {
+      "epoch": 2.983989077820529,
+      "grad_norm": 0.07168287010692334,
+      "learning_rate": 1.5103710215545448e-06,
+      "loss": 0.4644,
+      "step": 6014
+    },
+    {
+      "epoch": 2.9844855405237682,
+      "grad_norm": 0.07015268650999319,
+      "learning_rate": 1.5089713911301063e-06,
+      "loss": 0.4485,
+      "step": 6015
+    },
+    {
+      "epoch": 2.9849820032270076,
+      "grad_norm": 0.06921333182373515,
+      "learning_rate": 1.5075722942534154e-06,
+      "loss": 0.4314,
+      "step": 6016
+    },
+    {
+      "epoch": 2.985478465930247,
+      "grad_norm": 0.07178123073223433,
+      "learning_rate": 1.5061737311383018e-06,
+      "loss": 0.4686,
+      "step": 6017
+    },
+    {
+      "epoch": 2.9859749286334862,
+      "grad_norm": 0.07114264232083874,
+      "learning_rate": 1.5047757019985155e-06,
+      "loss": 0.4572,
+      "step": 6018
+    },
+    {
+      "epoch": 2.9864713913367256,
+      "grad_norm": 0.07295002675156483,
+      "learning_rate": 1.5033782070477192e-06,
+      "loss": 0.4481,
+      "step": 6019
+    },
+    {
+      "epoch": 2.9869678540399653,
+      "grad_norm": 0.07224442006644777,
+      "learning_rate": 1.5019812464995027e-06,
+      "loss": 0.4652,
+      "step": 6020
+    },
+    {
+      "epoch": 2.9874643167432047,
+      "grad_norm": 0.07070678180071457,
+      "learning_rate": 1.5005848205673652e-06,
+      "loss": 0.4544,
+      "step": 6021
+    },
+    {
+      "epoch": 2.987960779446444,
+      "grad_norm": 0.06968067830115295,
+      "learning_rate": 1.4991889294647277e-06,
+      "loss": 0.4659,
+      "step": 6022
+    },
+    {
+      "epoch": 2.988457242149684,
+      "grad_norm": 0.07194371180524783,
+      "learning_rate": 1.4977935734049342e-06,
+      "loss": 0.4687,
+      "step": 6023
+    },
+    {
+      "epoch": 2.988953704852923,
+      "grad_norm": 0.07353135929655469,
+      "learning_rate": 1.4963987526012368e-06,
+      "loss": 0.4295,
+      "step": 6024
+    },
+    {
+      "epoch": 2.9894501675561624,
+      "grad_norm": 0.07387419105574315,
+      "learning_rate": 1.495004467266815e-06,
+      "loss": 0.4508,
+      "step": 6025
+    },
+    {
+      "epoch": 2.9899466302594018,
+      "grad_norm": 0.07221393034988988,
+      "learning_rate": 1.4936107176147606e-06,
+      "loss": 0.4572,
+      "step": 6026
+    },
+    {
+      "epoch": 2.990443092962641,
+      "grad_norm": 0.07335610921440719,
+      "learning_rate": 1.4922175038580894e-06,
+      "loss": 0.4533,
+      "step": 6027
+    },
+    {
+      "epoch": 2.9909395556658804,
+      "grad_norm": 0.07038270855175356,
+      "learning_rate": 1.4908248262097292e-06,
+      "loss": 0.4344,
+      "step": 6028
+    },
+    {
+      "epoch": 2.9914360183691198,
+      "grad_norm": 0.07300052489851201,
+      "learning_rate": 1.4894326848825275e-06,
+      "loss": 0.4814,
+      "step": 6029
+    },
+    {
+      "epoch": 2.9919324810723595,
+      "grad_norm": 0.07050113627752637,
+      "learning_rate": 1.4880410800892541e-06,
+      "loss": 0.4727,
+      "step": 6030
+    },
+    {
+      "epoch": 2.992428943775599,
+      "grad_norm": 0.07210413625778277,
+      "learning_rate": 1.4866500120425914e-06,
+      "loss": 0.4464,
+      "step": 6031
+    },
+    {
+      "epoch": 2.992925406478838,
+      "grad_norm": 0.07003570060025452,
+      "learning_rate": 1.4852594809551402e-06,
+      "loss": 0.4552,
+      "step": 6032
+    },
+    {
+      "epoch": 2.9934218691820775,
+      "grad_norm": 0.06897753877390084,
+      "learning_rate": 1.483869487039425e-06,
+      "loss": 0.4197,
+      "step": 6033
+    },
+    {
+      "epoch": 2.9939183318853173,
+      "grad_norm": 0.0721561684411484,
+      "learning_rate": 1.4824800305078797e-06,
+      "loss": 0.4576,
+      "step": 6034
+    },
+    {
+      "epoch": 2.9944147945885566,
+      "grad_norm": 0.0706978250471,
+      "learning_rate": 1.4810911115728644e-06,
+      "loss": 0.4262,
+      "step": 6035
+    },
+    {
+      "epoch": 2.994911257291796,
+      "grad_norm": 0.07334245092152264,
+      "learning_rate": 1.479702730446651e-06,
+      "loss": 0.4694,
+      "step": 6036
+    },
+    {
+      "epoch": 2.9954077199950353,
+      "grad_norm": 0.06934060094770478,
+      "learning_rate": 1.4783148873414305e-06,
+      "loss": 0.4424,
+      "step": 6037
+    },
+    {
+      "epoch": 2.9959041826982746,
+      "grad_norm": 0.07370151222123784,
+      "learning_rate": 1.4769275824693146e-06,
+      "loss": 0.4715,
+      "step": 6038
+    },
+    {
+      "epoch": 2.996400645401514,
+      "grad_norm": 0.07250380480661225,
+      "learning_rate": 1.4755408160423302e-06,
+      "loss": 0.4568,
+      "step": 6039
+    },
+    {
+      "epoch": 2.9968971081047537,
+      "grad_norm": 0.07043634822680875,
+      "learning_rate": 1.4741545882724213e-06,
+      "loss": 0.4634,
+      "step": 6040
+    },
+    {
+      "epoch": 2.997393570807993,
+      "grad_norm": 0.0733358803979804,
+      "learning_rate": 1.4727688993714494e-06,
+      "loss": 0.4536,
+      "step": 6041
+    },
+    {
+      "epoch": 2.9978900335112324,
+      "grad_norm": 0.07071774708399711,
+      "learning_rate": 1.4713837495511978e-06,
+      "loss": 0.4465,
+      "step": 6042
+    },
+    {
+      "epoch": 2.9983864962144717,
+      "grad_norm": 0.0717978688509802,
+      "learning_rate": 1.4699991390233631e-06,
+      "loss": 0.4446,
+      "step": 6043
+    },
+    {
+      "epoch": 2.9988829589177115,
+      "grad_norm": 0.07146905900228022,
+      "learning_rate": 1.4686150679995592e-06,
+      "loss": 0.4412,
+      "step": 6044
+    },
+    {
+      "epoch": 2.999379421620951,
+      "grad_norm": 0.07219369507453394,
+      "learning_rate": 1.467231536691322e-06,
+      "loss": 0.4552,
+      "step": 6045
+    },
+    {
+      "epoch": 2.99987588432419,
+      "grad_norm": 0.0713535462878617,
+      "learning_rate": 1.4658485453100996e-06,
+      "loss": 0.4784,
+      "step": 6046
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0713535462878617,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.1017,
+      "step": 6047
+    },
+    {
+      "epoch": 3.0003723470274295,
+      "grad_norm": 0.06977470998602367,
+      "learning_rate": 1.4630841831740955e-06,
+      "loss": 0.335,
+      "step": 6048
+    },
+    {
+      "epoch": 3.0003723470274295,
+      "eval_loss": 0.5128746032714844,
+      "eval_runtime": 258.8736,
+      "eval_samples_per_second": 117.25,
+      "eval_steps_per_second": 14.66,
+      "step": 6048
+    },
+    {
+      "epoch": 3.0004964627032393,
+      "grad_norm": 0.07879870082790116,
+      "learning_rate": 1.4617028128417993e-06,
+      "loss": 0.4138,
+      "step": 6049
+    },
+    {
+      "epoch": 3.0009929254064787,
+      "grad_norm": 0.08185603013604671,
+      "learning_rate": 1.4603219832814968e-06,
+      "loss": 0.4423,
+      "step": 6050
+    },
+    {
+      "epoch": 3.0014893881097184,
+      "grad_norm": 0.08009656994151068,
+      "learning_rate": 1.4589416947042234e-06,
+      "loss": 0.4033,
+      "step": 6051
+    },
+    {
+      "epoch": 3.0019858508129578,
+      "grad_norm": 0.07847564767768977,
+      "learning_rate": 1.4575619473209373e-06,
+      "loss": 0.4638,
+      "step": 6052
+    },
+    {
+      "epoch": 3.002482313516197,
+      "grad_norm": 0.07147341894384611,
+      "learning_rate": 1.4561827413425089e-06,
+      "loss": 0.4149,
+      "step": 6053
+    },
+    {
+      "epoch": 3.0029787762194364,
+      "grad_norm": 0.07426846370280647,
+      "learning_rate": 1.4548040769797255e-06,
+      "loss": 0.4317,
+      "step": 6054
+    },
+    {
+      "epoch": 3.0034752389226758,
+      "grad_norm": 0.07080595435948296,
+      "learning_rate": 1.4534259544432983e-06,
+      "loss": 0.4281,
+      "step": 6055
+    },
+    {
+      "epoch": 3.0039717016259155,
+      "grad_norm": 0.07189277655550798,
+      "learning_rate": 1.452048373943849e-06,
+      "loss": 0.4378,
+      "step": 6056
+    },
+    {
+      "epoch": 3.004468164329155,
+      "grad_norm": 0.0735231757191823,
+      "learning_rate": 1.4506713356919184e-06,
+      "loss": 0.4592,
+      "step": 6057
+    },
+    {
+      "epoch": 3.004964627032394,
+      "grad_norm": 0.07177815486814404,
+      "learning_rate": 1.4492948398979634e-06,
+      "loss": 0.4519,
+      "step": 6058
+    },
+    {
+      "epoch": 3.0054610897356335,
+      "grad_norm": 0.07308116284527004,
+      "learning_rate": 1.447918886772362e-06,
+      "loss": 0.4372,
+      "step": 6059
+    },
+    {
+      "epoch": 3.005957552438873,
+      "grad_norm": 0.07352080350431428,
+      "learning_rate": 1.446543476525406e-06,
+      "loss": 0.4313,
+      "step": 6060
+    },
+    {
+      "epoch": 3.0064540151421126,
+      "grad_norm": 0.07591613815457766,
+      "learning_rate": 1.4451686093673028e-06,
+      "loss": 0.4694,
+      "step": 6061
+    },
+    {
+      "epoch": 3.006950477845352,
+      "grad_norm": 0.07713518741085477,
+      "learning_rate": 1.4437942855081816e-06,
+      "loss": 0.4749,
+      "step": 6062
+    },
+    {
+      "epoch": 3.0074469405485913,
+      "grad_norm": 0.07522268953804548,
+      "learning_rate": 1.4424205051580831e-06,
+      "loss": 0.4296,
+      "step": 6063
+    },
+    {
+      "epoch": 3.0079434032518306,
+      "grad_norm": 0.07371488261024522,
+      "learning_rate": 1.4410472685269699e-06,
+      "loss": 0.4359,
+      "step": 6064
+    },
+    {
+      "epoch": 3.00843986595507,
+      "grad_norm": 0.07325013732702464,
+      "learning_rate": 1.4396745758247189e-06,
+      "loss": 0.4266,
+      "step": 6065
+    },
+    {
+      "epoch": 3.0089363286583097,
+      "grad_norm": 0.07395652356856376,
+      "learning_rate": 1.4383024272611217e-06,
+      "loss": 0.4729,
+      "step": 6066
+    },
+    {
+      "epoch": 3.009432791361549,
+      "grad_norm": 0.07352284430323344,
+      "learning_rate": 1.4369308230458927e-06,
+      "loss": 0.4499,
+      "step": 6067
+    },
+    {
+      "epoch": 3.0099292540647884,
+      "grad_norm": 0.0757223936029539,
+      "learning_rate": 1.4355597633886576e-06,
+      "loss": 0.4423,
+      "step": 6068
+    },
+    {
+      "epoch": 3.0104257167680277,
+      "grad_norm": 0.07337655222505507,
+      "learning_rate": 1.43418924849896e-06,
+      "loss": 0.4469,
+      "step": 6069
+    },
+    {
+      "epoch": 3.010922179471267,
+      "grad_norm": 0.07487651129076654,
+      "learning_rate": 1.4328192785862638e-06,
+      "loss": 0.4367,
+      "step": 6070
+    },
+    {
+      "epoch": 3.011418642174507,
+      "grad_norm": 0.07221681940111216,
+      "learning_rate": 1.4314498538599437e-06,
+      "loss": 0.4398,
+      "step": 6071
+    },
+    {
+      "epoch": 3.011915104877746,
+      "grad_norm": 0.07034556104531757,
+      "learning_rate": 1.4300809745292993e-06,
+      "loss": 0.4346,
+      "step": 6072
+    },
+    {
+      "epoch": 3.0124115675809855,
+      "grad_norm": 0.0735456721483014,
+      "learning_rate": 1.4287126408035356e-06,
+      "loss": 0.4478,
+      "step": 6073
+    },
+    {
+      "epoch": 3.012908030284225,
+      "grad_norm": 0.07182254899050108,
+      "learning_rate": 1.4273448528917854e-06,
+      "loss": 0.4519,
+      "step": 6074
+    },
+    {
+      "epoch": 3.013404492987464,
+      "grad_norm": 0.069424652232987,
+      "learning_rate": 1.425977611003091e-06,
+      "loss": 0.4125,
+      "step": 6075
+    },
+    {
+      "epoch": 3.013900955690704,
+      "grad_norm": 0.07186620818532258,
+      "learning_rate": 1.424610915346412e-06,
+      "loss": 0.4109,
+      "step": 6076
+    },
+    {
+      "epoch": 3.0143974183939433,
+      "grad_norm": 0.0750254821896234,
+      "learning_rate": 1.4232447661306292e-06,
+      "loss": 0.444,
+      "step": 6077
+    },
+    {
+      "epoch": 3.0148938810971826,
+      "grad_norm": 0.07316063569686765,
+      "learning_rate": 1.4218791635645335e-06,
+      "loss": 0.4644,
+      "step": 6078
+    },
+    {
+      "epoch": 3.015390343800422,
+      "grad_norm": 0.07325720344215891,
+      "learning_rate": 1.4205141078568384e-06,
+      "loss": 0.4305,
+      "step": 6079
+    },
+    {
+      "epoch": 3.0158868065036613,
+      "grad_norm": 0.07186913073023993,
+      "learning_rate": 1.419149599216169e-06,
+      "loss": 0.4305,
+      "step": 6080
+    },
+    {
+      "epoch": 3.016383269206901,
+      "grad_norm": 0.07259629560798145,
+      "learning_rate": 1.4177856378510675e-06,
+      "loss": 0.446,
+      "step": 6081
+    },
+    {
+      "epoch": 3.0168797319101404,
+      "grad_norm": 0.07347257396391449,
+      "learning_rate": 1.416422223969996e-06,
+      "loss": 0.4521,
+      "step": 6082
+    },
+    {
+      "epoch": 3.0173761946133797,
+      "grad_norm": 0.07406415018966424,
+      "learning_rate": 1.4150593577813282e-06,
+      "loss": 0.4268,
+      "step": 6083
+    },
+    {
+      "epoch": 3.017872657316619,
+      "grad_norm": 0.07238241289397684,
+      "learning_rate": 1.4136970394933586e-06,
+      "loss": 0.4731,
+      "step": 6084
+    },
+    {
+      "epoch": 3.0183691200198584,
+      "grad_norm": 0.06964934760379549,
+      "learning_rate": 1.4123352693142945e-06,
+      "loss": 0.455,
+      "step": 6085
+    },
+    {
+      "epoch": 3.018865582723098,
+      "grad_norm": 0.07462793481034907,
+      "learning_rate": 1.4109740474522594e-06,
+      "loss": 0.4637,
+      "step": 6086
+    },
+    {
+      "epoch": 3.0193620454263375,
+      "grad_norm": 0.07060949249255187,
+      "learning_rate": 1.4096133741152967e-06,
+      "loss": 0.4375,
+      "step": 6087
+    },
+    {
+      "epoch": 3.019858508129577,
+      "grad_norm": 0.07329812020775885,
+      "learning_rate": 1.4082532495113627e-06,
+      "loss": 0.4302,
+      "step": 6088
+    },
+    {
+      "epoch": 3.020354970832816,
+      "grad_norm": 0.07481253449274504,
+      "learning_rate": 1.4068936738483302e-06,
+      "loss": 0.4526,
+      "step": 6089
+    },
+    {
+      "epoch": 3.0208514335360555,
+      "grad_norm": 0.07216798863248236,
+      "learning_rate": 1.4055346473339865e-06,
+      "loss": 0.4164,
+      "step": 6090
+    },
+    {
+      "epoch": 3.0213478962392952,
+      "grad_norm": 0.07267376251676638,
+      "learning_rate": 1.4041761701760414e-06,
+      "loss": 0.4244,
+      "step": 6091
+    },
+    {
+      "epoch": 3.0218443589425346,
+      "grad_norm": 0.07297391019769917,
+      "learning_rate": 1.4028182425821135e-06,
+      "loss": 0.4361,
+      "step": 6092
+    },
+    {
+      "epoch": 3.022340821645774,
+      "grad_norm": 0.07239707684815393,
+      "learning_rate": 1.4014608647597394e-06,
+      "loss": 0.4369,
+      "step": 6093
+    },
+    {
+      "epoch": 3.022837284349013,
+      "grad_norm": 0.07207550735039364,
+      "learning_rate": 1.4001040369163755e-06,
+      "loss": 0.4504,
+      "step": 6094
+    },
+    {
+      "epoch": 3.0233337470522526,
+      "grad_norm": 0.07489432295105117,
+      "learning_rate": 1.398747759259388e-06,
+      "loss": 0.5066,
+      "step": 6095
+    },
+    {
+      "epoch": 3.0238302097554923,
+      "grad_norm": 0.07117973295591797,
+      "learning_rate": 1.3973920319960654e-06,
+      "loss": 0.416,
+      "step": 6096
+    },
+    {
+      "epoch": 3.0243266724587317,
+      "grad_norm": 0.07233810281882866,
+      "learning_rate": 1.3960368553336073e-06,
+      "loss": 0.4223,
+      "step": 6097
+    },
+    {
+      "epoch": 3.024823135161971,
+      "grad_norm": 0.07396695918371478,
+      "learning_rate": 1.394682229479129e-06,
+      "loss": 0.452,
+      "step": 6098
+    },
+    {
+      "epoch": 3.0253195978652103,
+      "grad_norm": 0.0706555544795235,
+      "learning_rate": 1.3933281546396665e-06,
+      "loss": 0.4223,
+      "step": 6099
+    },
+    {
+      "epoch": 3.0258160605684497,
+      "grad_norm": 0.07178200800744759,
+      "learning_rate": 1.391974631022166e-06,
+      "loss": 0.4497,
+      "step": 6100
+    },
+    {
+      "epoch": 3.0263125232716894,
+      "grad_norm": 0.07286657811058043,
+      "learning_rate": 1.3906216588334936e-06,
+      "loss": 0.4361,
+      "step": 6101
+    },
+    {
+      "epoch": 3.0268089859749288,
+      "grad_norm": 0.07299593785310476,
+      "learning_rate": 1.3892692382804295e-06,
+      "loss": 0.4534,
+      "step": 6102
+    },
+    {
+      "epoch": 3.027305448678168,
+      "grad_norm": 0.07099933881802412,
+      "learning_rate": 1.3879173695696668e-06,
+      "loss": 0.4137,
+      "step": 6103
+    },
+    {
+      "epoch": 3.0278019113814074,
+      "grad_norm": 0.07139228360005229,
+      "learning_rate": 1.3865660529078218e-06,
+      "loss": 0.439,
+      "step": 6104
+    },
+    {
+      "epoch": 3.0282983740846467,
+      "grad_norm": 0.07235298810056374,
+      "learning_rate": 1.3852152885014152e-06,
+      "loss": 0.4395,
+      "step": 6105
+    },
+    {
+      "epoch": 3.0287948367878865,
+      "grad_norm": 0.0756061099029309,
+      "learning_rate": 1.383865076556895e-06,
+      "loss": 0.4624,
+      "step": 6106
+    },
+    {
+      "epoch": 3.029291299491126,
+      "grad_norm": 0.07318085394261264,
+      "learning_rate": 1.3825154172806176e-06,
+      "loss": 0.4427,
+      "step": 6107
+    },
+    {
+      "epoch": 3.029787762194365,
+      "grad_norm": 0.06982982435111475,
+      "learning_rate": 1.3811663108788553e-06,
+      "loss": 0.4132,
+      "step": 6108
+    },
+    {
+      "epoch": 3.0302842248976045,
+      "grad_norm": 0.07128249031773394,
+      "learning_rate": 1.3798177575577998e-06,
+      "loss": 0.4325,
+      "step": 6109
+    },
+    {
+      "epoch": 3.030780687600844,
+      "grad_norm": 0.06998626113303408,
+      "learning_rate": 1.378469757523554e-06,
+      "loss": 0.4446,
+      "step": 6110
+    },
+    {
+      "epoch": 3.0312771503040836,
+      "grad_norm": 0.07236540853128076,
+      "learning_rate": 1.3771223109821402e-06,
+      "loss": 0.411,
+      "step": 6111
+    },
+    {
+      "epoch": 3.031773613007323,
+      "grad_norm": 0.07183106915540166,
+      "learning_rate": 1.3757754181394921e-06,
+      "loss": 0.4344,
+      "step": 6112
+    },
+    {
+      "epoch": 3.0322700757105623,
+      "grad_norm": 0.07249512228100471,
+      "learning_rate": 1.374429079201461e-06,
+      "loss": 0.4312,
+      "step": 6113
+    },
+    {
+      "epoch": 3.0327665384138016,
+      "grad_norm": 0.07091446563916892,
+      "learning_rate": 1.3730832943738143e-06,
+      "loss": 0.4271,
+      "step": 6114
+    },
+    {
+      "epoch": 3.033263001117041,
+      "grad_norm": 0.07431020907428121,
+      "learning_rate": 1.3717380638622313e-06,
+      "loss": 0.4429,
+      "step": 6115
+    },
+    {
+      "epoch": 3.0337594638202807,
+      "grad_norm": 0.07269213005302484,
+      "learning_rate": 1.3703933878723119e-06,
+      "loss": 0.4787,
+      "step": 6116
+    },
+    {
+      "epoch": 3.03425592652352,
+      "grad_norm": 0.06960826643523638,
+      "learning_rate": 1.3690492666095672e-06,
+      "loss": 0.4291,
+      "step": 6117
+    },
+    {
+      "epoch": 3.0347523892267594,
+      "grad_norm": 0.07072163570955363,
+      "learning_rate": 1.3677057002794226e-06,
+      "loss": 0.403,
+      "step": 6118
+    },
+    {
+      "epoch": 3.0352488519299987,
+      "grad_norm": 0.07224535948018528,
+      "learning_rate": 1.3663626890872239e-06,
+      "loss": 0.414,
+      "step": 6119
+    },
+    {
+      "epoch": 3.035745314633238,
+      "grad_norm": 0.07340254247977736,
+      "learning_rate": 1.3650202332382273e-06,
+      "loss": 0.4506,
+      "step": 6120
+    },
+    {
+      "epoch": 3.0362417773364774,
+      "grad_norm": 0.07281206895696977,
+      "learning_rate": 1.3636783329376053e-06,
+      "loss": 0.4593,
+      "step": 6121
+    },
+    {
+      "epoch": 3.036738240039717,
+      "grad_norm": 0.07520592685480292,
+      "learning_rate": 1.3623369883904447e-06,
+      "loss": 0.4624,
+      "step": 6122
+    },
+    {
+      "epoch": 3.0372347027429565,
+      "grad_norm": 0.07005595449616403,
+      "learning_rate": 1.3609961998017519e-06,
+      "loss": 0.4288,
+      "step": 6123
+    },
+    {
+      "epoch": 3.037731165446196,
+      "grad_norm": 0.07100814556301673,
+      "learning_rate": 1.3596559673764421e-06,
+      "loss": 0.4118,
+      "step": 6124
+    },
+    {
+      "epoch": 3.038227628149435,
+      "grad_norm": 0.07300186961223747,
+      "learning_rate": 1.3583162913193483e-06,
+      "loss": 0.4358,
+      "step": 6125
+    },
+    {
+      "epoch": 3.0387240908526745,
+      "grad_norm": 0.07392695568371571,
+      "learning_rate": 1.3569771718352208e-06,
+      "loss": 0.433,
+      "step": 6126
+    },
+    {
+      "epoch": 3.0392205535559143,
+      "grad_norm": 0.07423043383981226,
+      "learning_rate": 1.3556386091287193e-06,
+      "loss": 0.447,
+      "step": 6127
+    },
+    {
+      "epoch": 3.0397170162591536,
+      "grad_norm": 0.07263636650400927,
+      "learning_rate": 1.3543006034044255e-06,
+      "loss": 0.4255,
+      "step": 6128
+    },
+    {
+      "epoch": 3.040213478962393,
+      "grad_norm": 0.07583843572765442,
+      "learning_rate": 1.3529631548668298e-06,
+      "loss": 0.4454,
+      "step": 6129
+    },
+    {
+      "epoch": 3.0407099416656322,
+      "grad_norm": 0.0720704890024341,
+      "learning_rate": 1.3516262637203392e-06,
+      "loss": 0.4425,
+      "step": 6130
+    },
+    {
+      "epoch": 3.0412064043688716,
+      "grad_norm": 0.07381962200231772,
+      "learning_rate": 1.350289930169278e-06,
+      "loss": 0.4614,
+      "step": 6131
+    },
+    {
+      "epoch": 3.0417028670721113,
+      "grad_norm": 0.07144967882171453,
+      "learning_rate": 1.3489541544178808e-06,
+      "loss": 0.4375,
+      "step": 6132
+    },
+    {
+      "epoch": 3.0421993297753507,
+      "grad_norm": 0.07133068638464211,
+      "learning_rate": 1.3476189366703024e-06,
+      "loss": 0.4284,
+      "step": 6133
+    },
+    {
+      "epoch": 3.04269579247859,
+      "grad_norm": 0.0729834506663787,
+      "learning_rate": 1.3462842771306084e-06,
+      "loss": 0.4435,
+      "step": 6134
+    },
+    {
+      "epoch": 3.0431922551818293,
+      "grad_norm": 0.07057614017895736,
+      "learning_rate": 1.3449501760027778e-06,
+      "loss": 0.434,
+      "step": 6135
+    },
+    {
+      "epoch": 3.0436887178850687,
+      "grad_norm": 0.0731597201840205,
+      "learning_rate": 1.3436166334907118e-06,
+      "loss": 0.4133,
+      "step": 6136
+    },
+    {
+      "epoch": 3.0441851805883084,
+      "grad_norm": 0.07246949898925738,
+      "learning_rate": 1.342283649798215e-06,
+      "loss": 0.4195,
+      "step": 6137
+    },
+    {
+      "epoch": 3.044681643291548,
+      "grad_norm": 0.06981040620325808,
+      "learning_rate": 1.3409512251290164e-06,
+      "loss": 0.4159,
+      "step": 6138
+    },
+    {
+      "epoch": 3.045178105994787,
+      "grad_norm": 0.07238922242350312,
+      "learning_rate": 1.3396193596867534e-06,
+      "loss": 0.4276,
+      "step": 6139
+    },
+    {
+      "epoch": 3.0456745686980264,
+      "grad_norm": 0.0719935930840651,
+      "learning_rate": 1.3382880536749831e-06,
+      "loss": 0.4257,
+      "step": 6140
+    },
+    {
+      "epoch": 3.0461710314012658,
+      "grad_norm": 0.07403638629611554,
+      "learning_rate": 1.3369573072971725e-06,
+      "loss": 0.4716,
+      "step": 6141
+    },
+    {
+      "epoch": 3.0466674941045055,
+      "grad_norm": 0.07230713050437415,
+      "learning_rate": 1.3356271207567033e-06,
+      "loss": 0.437,
+      "step": 6142
+    },
+    {
+      "epoch": 3.047163956807745,
+      "grad_norm": 0.07508824326437552,
+      "learning_rate": 1.334297494256877e-06,
+      "loss": 0.4708,
+      "step": 6143
+    },
+    {
+      "epoch": 3.047660419510984,
+      "grad_norm": 0.07437845757042565,
+      "learning_rate": 1.3329684280009032e-06,
+      "loss": 0.4236,
+      "step": 6144
+    },
+    {
+      "epoch": 3.0481568822142235,
+      "grad_norm": 0.07369128333228318,
+      "learning_rate": 1.3316399221919075e-06,
+      "loss": 0.4321,
+      "step": 6145
+    },
+    {
+      "epoch": 3.048653344917463,
+      "grad_norm": 0.07308115787987783,
+      "learning_rate": 1.3303119770329336e-06,
+      "loss": 0.4463,
+      "step": 6146
+    },
+    {
+      "epoch": 3.0491498076207026,
+      "grad_norm": 0.07129520875002196,
+      "learning_rate": 1.3289845927269335e-06,
+      "loss": 0.4389,
+      "step": 6147
+    },
+    {
+      "epoch": 3.049646270323942,
+      "grad_norm": 0.07374682289602172,
+      "learning_rate": 1.3276577694767794e-06,
+      "loss": 0.454,
+      "step": 6148
+    },
+    {
+      "epoch": 3.0501427330271813,
+      "grad_norm": 0.07117621386936727,
+      "learning_rate": 1.326331507485254e-06,
+      "loss": 0.4338,
+      "step": 6149
+    },
+    {
+      "epoch": 3.0506391957304206,
+      "grad_norm": 0.07589034699881031,
+      "learning_rate": 1.325005806955053e-06,
+      "loss": 0.4371,
+      "step": 6150
+    },
+    {
+      "epoch": 3.05113565843366,
+      "grad_norm": 0.07142220044539167,
+      "learning_rate": 1.323680668088792e-06,
+      "loss": 0.4004,
+      "step": 6151
+    },
+    {
+      "epoch": 3.0516321211368997,
+      "grad_norm": 0.07036115905012849,
+      "learning_rate": 1.322356091088996e-06,
+      "loss": 0.4205,
+      "step": 6152
+    },
+    {
+      "epoch": 3.052128583840139,
+      "grad_norm": 0.07329646557855506,
+      "learning_rate": 1.3210320761581047e-06,
+      "loss": 0.4436,
+      "step": 6153
+    },
+    {
+      "epoch": 3.0526250465433784,
+      "grad_norm": 0.07196844227110943,
+      "learning_rate": 1.3197086234984707e-06,
+      "loss": 0.402,
+      "step": 6154
+    },
+    {
+      "epoch": 3.0531215092466177,
+      "grad_norm": 0.07401728202452947,
+      "learning_rate": 1.3183857333123667e-06,
+      "loss": 0.4541,
+      "step": 6155
+    },
+    {
+      "epoch": 3.053617971949857,
+      "grad_norm": 0.07027866583372665,
+      "learning_rate": 1.3170634058019733e-06,
+      "loss": 0.4249,
+      "step": 6156
+    },
+    {
+      "epoch": 3.054114434653097,
+      "grad_norm": 0.07119927820007768,
+      "learning_rate": 1.3157416411693851e-06,
+      "loss": 0.4274,
+      "step": 6157
+    },
+    {
+      "epoch": 3.054610897356336,
+      "grad_norm": 0.07180601201705737,
+      "learning_rate": 1.314420439616616e-06,
+      "loss": 0.4176,
+      "step": 6158
+    },
+    {
+      "epoch": 3.0551073600595755,
+      "grad_norm": 0.07158720435499417,
+      "learning_rate": 1.3130998013455875e-06,
+      "loss": 0.4651,
+      "step": 6159
+    },
+    {
+      "epoch": 3.055603822762815,
+      "grad_norm": 0.07084365083214748,
+      "learning_rate": 1.3117797265581412e-06,
+      "loss": 0.4343,
+      "step": 6160
+    },
+    {
+      "epoch": 3.056100285466054,
+      "grad_norm": 0.07440074991217603,
+      "learning_rate": 1.3104602154560275e-06,
+      "loss": 0.4307,
+      "step": 6161
+    },
+    {
+      "epoch": 3.056596748169294,
+      "grad_norm": 0.07341912115141237,
+      "learning_rate": 1.3091412682409104e-06,
+      "loss": 0.4348,
+      "step": 6162
+    },
+    {
+      "epoch": 3.0570932108725333,
+      "grad_norm": 0.07307954047079919,
+      "learning_rate": 1.3078228851143743e-06,
+      "loss": 0.4696,
+      "step": 6163
+    },
+    {
+      "epoch": 3.0575896735757726,
+      "grad_norm": 0.07098981514671952,
+      "learning_rate": 1.3065050662779088e-06,
+      "loss": 0.4488,
+      "step": 6164
+    },
+    {
+      "epoch": 3.058086136279012,
+      "grad_norm": 0.07061819183409725,
+      "learning_rate": 1.3051878119329248e-06,
+      "loss": 0.4005,
+      "step": 6165
+    },
+    {
+      "epoch": 3.0585825989822513,
+      "grad_norm": 0.07396280603570082,
+      "learning_rate": 1.303871122280742e-06,
+      "loss": 0.445,
+      "step": 6166
+    },
+    {
+      "epoch": 3.059079061685491,
+      "grad_norm": 0.07506077026161759,
+      "learning_rate": 1.3025549975225936e-06,
+      "loss": 0.4275,
+      "step": 6167
+    },
+    {
+      "epoch": 3.0595755243887304,
+      "grad_norm": 0.07159324115577584,
+      "learning_rate": 1.3012394378596333e-06,
+      "loss": 0.4265,
+      "step": 6168
+    },
+    {
+      "epoch": 3.0600719870919697,
+      "grad_norm": 0.0732539848540184,
+      "learning_rate": 1.2999244434929159e-06,
+      "loss": 0.4668,
+      "step": 6169
+    },
+    {
+      "epoch": 3.060568449795209,
+      "grad_norm": 0.07326446643050033,
+      "learning_rate": 1.298610014623423e-06,
+      "loss": 0.4576,
+      "step": 6170
+    },
+    {
+      "epoch": 3.0610649124984484,
+      "grad_norm": 0.07468402165767048,
+      "learning_rate": 1.2972961514520411e-06,
+      "loss": 0.4438,
+      "step": 6171
+    },
+    {
+      "epoch": 3.061561375201688,
+      "grad_norm": 0.0731658754962425,
+      "learning_rate": 1.295982854179575e-06,
+      "loss": 0.4258,
+      "step": 6172
+    },
+    {
+      "epoch": 3.0620578379049275,
+      "grad_norm": 0.07494630711068176,
+      "learning_rate": 1.2946701230067405e-06,
+      "loss": 0.4398,
+      "step": 6173
+    },
+    {
+      "epoch": 3.062554300608167,
+      "grad_norm": 0.07269504657824281,
+      "learning_rate": 1.293357958134166e-06,
+      "loss": 0.4357,
+      "step": 6174
+    },
+    {
+      "epoch": 3.063050763311406,
+      "grad_norm": 0.07279870347394367,
+      "learning_rate": 1.2920463597623972e-06,
+      "loss": 0.4661,
+      "step": 6175
+    },
+    {
+      "epoch": 3.0635472260146455,
+      "grad_norm": 0.07318599816002383,
+      "learning_rate": 1.2907353280918883e-06,
+      "loss": 0.4121,
+      "step": 6176
+    },
+    {
+      "epoch": 3.0640436887178852,
+      "grad_norm": 0.07288523031937426,
+      "learning_rate": 1.2894248633230128e-06,
+      "loss": 0.4415,
+      "step": 6177
+    },
+    {
+      "epoch": 3.0645401514211246,
+      "grad_norm": 0.07467560919192302,
+      "learning_rate": 1.2881149656560522e-06,
+      "loss": 0.4657,
+      "step": 6178
+    },
+    {
+      "epoch": 3.065036614124364,
+      "grad_norm": 0.07437833595978252,
+      "learning_rate": 1.2868056352912018e-06,
+      "loss": 0.4361,
+      "step": 6179
+    },
+    {
+      "epoch": 3.0655330768276032,
+      "grad_norm": 0.06973346690351331,
+      "learning_rate": 1.2854968724285755e-06,
+      "loss": 0.4131,
+      "step": 6180
+    },
+    {
+      "epoch": 3.0660295395308426,
+      "grad_norm": 0.07142243005484775,
+      "learning_rate": 1.2841886772681944e-06,
+      "loss": 0.4384,
+      "step": 6181
+    },
+    {
+      "epoch": 3.0665260022340823,
+      "grad_norm": 0.07484904761118684,
+      "learning_rate": 1.2828810500099936e-06,
+      "loss": 0.4676,
+      "step": 6182
+    },
+    {
+      "epoch": 3.0670224649373217,
+      "grad_norm": 0.07121129131372053,
+      "learning_rate": 1.2815739908538272e-06,
+      "loss": 0.4256,
+      "step": 6183
+    },
+    {
+      "epoch": 3.067518927640561,
+      "grad_norm": 0.0737911004124479,
+      "learning_rate": 1.2802674999994553e-06,
+      "loss": 0.4595,
+      "step": 6184
+    },
+    {
+      "epoch": 3.0680153903438003,
+      "grad_norm": 0.0733653209074709,
+      "learning_rate": 1.2789615776465547e-06,
+      "loss": 0.4608,
+      "step": 6185
+    },
+    {
+      "epoch": 3.0685118530470397,
+      "grad_norm": 0.07226851556953984,
+      "learning_rate": 1.2776562239947133e-06,
+      "loss": 0.4365,
+      "step": 6186
+    },
+    {
+      "epoch": 3.0690083157502794,
+      "grad_norm": 0.07582035370251462,
+      "learning_rate": 1.276351439243436e-06,
+      "loss": 0.4519,
+      "step": 6187
+    },
+    {
+      "epoch": 3.0695047784535188,
+      "grad_norm": 0.074998501917356,
+      "learning_rate": 1.2750472235921374e-06,
+      "loss": 0.4604,
+      "step": 6188
+    },
+    {
+      "epoch": 3.070001241156758,
+      "grad_norm": 0.07647818404479714,
+      "learning_rate": 1.273743577240144e-06,
+      "loss": 0.4659,
+      "step": 6189
+    },
+    {
+      "epoch": 3.0704977038599974,
+      "grad_norm": 0.07188538726427558,
+      "learning_rate": 1.2724405003867002e-06,
+      "loss": 0.4307,
+      "step": 6190
+    },
+    {
+      "epoch": 3.0709941665632368,
+      "grad_norm": 0.07189098099018701,
+      "learning_rate": 1.2711379932309576e-06,
+      "loss": 0.4576,
+      "step": 6191
+    },
+    {
+      "epoch": 3.0714906292664765,
+      "grad_norm": 0.07098272142750643,
+      "learning_rate": 1.2698360559719863e-06,
+      "loss": 0.4266,
+      "step": 6192
+    },
+    {
+      "epoch": 3.071987091969716,
+      "grad_norm": 0.07501841044922372,
+      "learning_rate": 1.2685346888087657e-06,
+      "loss": 0.4335,
+      "step": 6193
+    },
+    {
+      "epoch": 3.072483554672955,
+      "grad_norm": 0.0699826251700936,
+      "learning_rate": 1.2672338919401866e-06,
+      "loss": 0.4272,
+      "step": 6194
+    },
+    {
+      "epoch": 3.0729800173761945,
+      "grad_norm": 0.07246590479610392,
+      "learning_rate": 1.2659336655650583e-06,
+      "loss": 0.4406,
+      "step": 6195
+    },
+    {
+      "epoch": 3.073476480079434,
+      "grad_norm": 0.07362225122051128,
+      "learning_rate": 1.2646340098820969e-06,
+      "loss": 0.4413,
+      "step": 6196
+    },
+    {
+      "epoch": 3.0739729427826736,
+      "grad_norm": 0.07196618201196071,
+      "learning_rate": 1.2633349250899363e-06,
+      "loss": 0.4322,
+      "step": 6197
+    },
+    {
+      "epoch": 3.074469405485913,
+      "grad_norm": 0.07255732105207222,
+      "learning_rate": 1.2620364113871193e-06,
+      "loss": 0.4427,
+      "step": 6198
+    },
+    {
+      "epoch": 3.0749658681891523,
+      "grad_norm": 0.07340975032145909,
+      "learning_rate": 1.2607384689721014e-06,
+      "loss": 0.4338,
+      "step": 6199
+    },
+    {
+      "epoch": 3.0754623308923916,
+      "grad_norm": 0.07278936088395885,
+      "learning_rate": 1.2594410980432575e-06,
+      "loss": 0.4512,
+      "step": 6200
+    },
+    {
+      "epoch": 3.075958793595631,
+      "grad_norm": 0.07149143967747107,
+      "learning_rate": 1.2581442987988635e-06,
+      "loss": 0.4305,
+      "step": 6201
+    },
+    {
+      "epoch": 3.0764552562988707,
+      "grad_norm": 0.07230038614740367,
+      "learning_rate": 1.2568480714371183e-06,
+      "loss": 0.445,
+      "step": 6202
+    },
+    {
+      "epoch": 3.07695171900211,
+      "grad_norm": 0.07130612387394257,
+      "learning_rate": 1.2555524161561277e-06,
+      "loss": 0.4493,
+      "step": 6203
+    },
+    {
+      "epoch": 3.0774481817053494,
+      "grad_norm": 0.0729240502747821,
+      "learning_rate": 1.2542573331539136e-06,
+      "loss": 0.4409,
+      "step": 6204
+    },
+    {
+      "epoch": 3.0779446444085887,
+      "grad_norm": 0.07394585409694572,
+      "learning_rate": 1.252962822628408e-06,
+      "loss": 0.4311,
+      "step": 6205
+    },
+    {
+      "epoch": 3.078441107111828,
+      "grad_norm": 0.07284506441536161,
+      "learning_rate": 1.2516688847774545e-06,
+      "loss": 0.4271,
+      "step": 6206
+    },
+    {
+      "epoch": 3.078937569815068,
+      "grad_norm": 0.07289169893543256,
+      "learning_rate": 1.2503755197988132e-06,
+      "loss": 0.448,
+      "step": 6207
+    },
+    {
+      "epoch": 3.079434032518307,
+      "grad_norm": 0.07072808271669016,
+      "learning_rate": 1.2490827278901513e-06,
+      "loss": 0.4373,
+      "step": 6208
+    },
+    {
+      "epoch": 3.0799304952215465,
+      "grad_norm": 0.07481156352728258,
+      "learning_rate": 1.247790509249055e-06,
+      "loss": 0.4403,
+      "step": 6209
+    },
+    {
+      "epoch": 3.080426957924786,
+      "grad_norm": 0.07547011321762705,
+      "learning_rate": 1.246498864073017e-06,
+      "loss": 0.4545,
+      "step": 6210
+    },
+    {
+      "epoch": 3.080923420628025,
+      "grad_norm": 0.06958365489123573,
+      "learning_rate": 1.2452077925594435e-06,
+      "loss": 0.4153,
+      "step": 6211
+    },
+    {
+      "epoch": 3.081419883331265,
+      "grad_norm": 0.07301974008845732,
+      "learning_rate": 1.2439172949056566e-06,
+      "loss": 0.4474,
+      "step": 6212
+    },
+    {
+      "epoch": 3.0819163460345043,
+      "grad_norm": 0.07440526375197373,
+      "learning_rate": 1.242627371308886e-06,
+      "loss": 0.431,
+      "step": 6213
+    },
+    {
+      "epoch": 3.0824128087377436,
+      "grad_norm": 0.07377167465451313,
+      "learning_rate": 1.2413380219662779e-06,
+      "loss": 0.4352,
+      "step": 6214
+    },
+    {
+      "epoch": 3.082909271440983,
+      "grad_norm": 0.0719765132891477,
+      "learning_rate": 1.2400492470748877e-06,
+      "loss": 0.4196,
+      "step": 6215
+    },
+    {
+      "epoch": 3.0834057341442223,
+      "grad_norm": 0.07128437563213033,
+      "learning_rate": 1.2387610468316835e-06,
+      "loss": 0.4619,
+      "step": 6216
+    },
+    {
+      "epoch": 3.083902196847462,
+      "grad_norm": 0.07342401754003233,
+      "learning_rate": 1.237473421433547e-06,
+      "loss": 0.439,
+      "step": 6217
+    },
+    {
+      "epoch": 3.0843986595507014,
+      "grad_norm": 0.07135736719422998,
+      "learning_rate": 1.2361863710772686e-06,
+      "loss": 0.4189,
+      "step": 6218
+    },
+    {
+      "epoch": 3.0848951222539407,
+      "grad_norm": 0.07493974809013262,
+      "learning_rate": 1.234899895959557e-06,
+      "loss": 0.4636,
+      "step": 6219
+    },
+    {
+      "epoch": 3.08539158495718,
+      "grad_norm": 0.07140821273764963,
+      "learning_rate": 1.233613996277027e-06,
+      "loss": 0.4014,
+      "step": 6220
+    },
+    {
+      "epoch": 3.0858880476604194,
+      "grad_norm": 0.07335384354682381,
+      "learning_rate": 1.2323286722262074e-06,
+      "loss": 0.4369,
+      "step": 6221
+    },
+    {
+      "epoch": 3.086384510363659,
+      "grad_norm": 0.07320122746466778,
+      "learning_rate": 1.2310439240035415e-06,
+      "loss": 0.4643,
+      "step": 6222
+    },
+    {
+      "epoch": 3.0868809730668985,
+      "grad_norm": 0.07273563727400321,
+      "learning_rate": 1.22975975180538e-06,
+      "loss": 0.4441,
+      "step": 6223
+    },
+    {
+      "epoch": 3.087377435770138,
+      "grad_norm": 0.07338840600727937,
+      "learning_rate": 1.2284761558279901e-06,
+      "loss": 0.452,
+      "step": 6224
+    },
+    {
+      "epoch": 3.087873898473377,
+      "grad_norm": 0.07432857851656079,
+      "learning_rate": 1.2271931362675482e-06,
+      "loss": 0.4367,
+      "step": 6225
+    },
+    {
+      "epoch": 3.0883703611766165,
+      "grad_norm": 0.06959785661299726,
+      "learning_rate": 1.225910693320142e-06,
+      "loss": 0.3849,
+      "step": 6226
+    },
+    {
+      "epoch": 3.0888668238798562,
+      "grad_norm": 0.069439748035569,
+      "learning_rate": 1.224628827181774e-06,
+      "loss": 0.4129,
+      "step": 6227
+    },
+    {
+      "epoch": 3.0893632865830956,
+      "grad_norm": 0.07167422324818025,
+      "learning_rate": 1.2233475380483557e-06,
+      "loss": 0.4383,
+      "step": 6228
+    },
+    {
+      "epoch": 3.089859749286335,
+      "grad_norm": 0.07409177756432055,
+      "learning_rate": 1.2220668261157132e-06,
+      "loss": 0.4442,
+      "step": 6229
+    },
+    {
+      "epoch": 3.090356211989574,
+      "grad_norm": 0.07229680378414288,
+      "learning_rate": 1.2207866915795818e-06,
+      "loss": 0.4375,
+      "step": 6230
+    },
+    {
+      "epoch": 3.0908526746928136,
+      "grad_norm": 0.0703528873093919,
+      "learning_rate": 1.2195071346356086e-06,
+      "loss": 0.4116,
+      "step": 6231
+    },
+    {
+      "epoch": 3.0913491373960533,
+      "grad_norm": 0.07187634392529822,
+      "learning_rate": 1.2182281554793567e-06,
+      "loss": 0.439,
+      "step": 6232
+    },
+    {
+      "epoch": 3.0918456000992927,
+      "grad_norm": 0.07010903570551587,
+      "learning_rate": 1.2169497543062924e-06,
+      "loss": 0.3927,
+      "step": 6233
+    },
+    {
+      "epoch": 3.092342062802532,
+      "grad_norm": 0.07399038526407921,
+      "learning_rate": 1.2156719313118026e-06,
+      "loss": 0.4438,
+      "step": 6234
+    },
+    {
+      "epoch": 3.0928385255057713,
+      "grad_norm": 0.0717854306834454,
+      "learning_rate": 1.21439468669118e-06,
+      "loss": 0.4301,
+      "step": 6235
+    },
+    {
+      "epoch": 3.0933349882090106,
+      "grad_norm": 0.07238415883742463,
+      "learning_rate": 1.2131180206396331e-06,
+      "loss": 0.4317,
+      "step": 6236
+    },
+    {
+      "epoch": 3.0938314509122504,
+      "grad_norm": 0.07298178970303208,
+      "learning_rate": 1.211841933352279e-06,
+      "loss": 0.4735,
+      "step": 6237
+    },
+    {
+      "epoch": 3.0943279136154898,
+      "grad_norm": 0.07236548395027288,
+      "learning_rate": 1.2105664250241455e-06,
+      "loss": 0.4372,
+      "step": 6238
+    },
+    {
+      "epoch": 3.094824376318729,
+      "grad_norm": 0.0703106987227702,
+      "learning_rate": 1.209291495850176e-06,
+      "loss": 0.4474,
+      "step": 6239
+    },
+    {
+      "epoch": 3.0953208390219684,
+      "grad_norm": 0.07531809414647281,
+      "learning_rate": 1.208017146025221e-06,
+      "loss": 0.4375,
+      "step": 6240
+    },
+    {
+      "epoch": 3.0958173017252077,
+      "grad_norm": 0.06888357618016823,
+      "learning_rate": 1.2067433757440466e-06,
+      "loss": 0.4312,
+      "step": 6241
+    },
+    {
+      "epoch": 3.0963137644284475,
+      "grad_norm": 0.07044369668372993,
+      "learning_rate": 1.2054701852013267e-06,
+      "loss": 0.4318,
+      "step": 6242
+    },
+    {
+      "epoch": 3.096810227131687,
+      "grad_norm": 0.07449678303052629,
+      "learning_rate": 1.2041975745916474e-06,
+      "loss": 0.4308,
+      "step": 6243
+    },
+    {
+      "epoch": 3.097306689834926,
+      "grad_norm": 0.07327395073971159,
+      "learning_rate": 1.202925544109509e-06,
+      "loss": 0.4371,
+      "step": 6244
+    },
+    {
+      "epoch": 3.0978031525381655,
+      "grad_norm": 0.07065465741904894,
+      "learning_rate": 1.2016540939493182e-06,
+      "loss": 0.4227,
+      "step": 6245
+    },
+    {
+      "epoch": 3.098299615241405,
+      "grad_norm": 0.07317786063065264,
+      "learning_rate": 1.2003832243053987e-06,
+      "loss": 0.4681,
+      "step": 6246
+    },
+    {
+      "epoch": 3.0987960779446446,
+      "grad_norm": 0.07502800326692527,
+      "learning_rate": 1.1991129353719816e-06,
+      "loss": 0.4739,
+      "step": 6247
+    },
+    {
+      "epoch": 3.099292540647884,
+      "grad_norm": 0.07359521511173439,
+      "learning_rate": 1.1978432273432095e-06,
+      "loss": 0.45,
+      "step": 6248
+    },
+    {
+      "epoch": 3.0997890033511233,
+      "grad_norm": 0.07096005437948809,
+      "learning_rate": 1.1965741004131365e-06,
+      "loss": 0.4318,
+      "step": 6249
+    },
+    {
+      "epoch": 3.1002854660543626,
+      "grad_norm": 0.07356071866870935,
+      "learning_rate": 1.195305554775728e-06,
+      "loss": 0.4426,
+      "step": 6250
+    },
+    {
+      "epoch": 3.100781928757602,
+      "grad_norm": 0.07342378306599293,
+      "learning_rate": 1.1940375906248635e-06,
+      "loss": 0.4719,
+      "step": 6251
+    },
+    {
+      "epoch": 3.1012783914608413,
+      "grad_norm": 0.07146026379036559,
+      "learning_rate": 1.1927702081543279e-06,
+      "loss": 0.4047,
+      "step": 6252
+    },
+    {
+      "epoch": 3.101774854164081,
+      "grad_norm": 0.07286765177483547,
+      "learning_rate": 1.191503407557823e-06,
+      "loss": 0.4729,
+      "step": 6253
+    },
+    {
+      "epoch": 3.1022713168673204,
+      "grad_norm": 0.07231918414330475,
+      "learning_rate": 1.190237189028957e-06,
+      "loss": 0.433,
+      "step": 6254
+    },
+    {
+      "epoch": 3.1027677795705597,
+      "grad_norm": 0.0737948461976323,
+      "learning_rate": 1.188971552761251e-06,
+      "loss": 0.4387,
+      "step": 6255
+    },
+    {
+      "epoch": 3.103264242273799,
+      "grad_norm": 0.07951141015270875,
+      "learning_rate": 1.1877064989481396e-06,
+      "loss": 0.5079,
+      "step": 6256
+    },
+    {
+      "epoch": 3.103760704977039,
+      "grad_norm": 0.0752100583948817,
+      "learning_rate": 1.186442027782964e-06,
+      "loss": 0.4767,
+      "step": 6257
+    },
+    {
+      "epoch": 3.104257167680278,
+      "grad_norm": 0.07290817664959637,
+      "learning_rate": 1.1851781394589774e-06,
+      "loss": 0.4312,
+      "step": 6258
+    },
+    {
+      "epoch": 3.1047536303835175,
+      "grad_norm": 0.07437746217306612,
+      "learning_rate": 1.1839148341693473e-06,
+      "loss": 0.4496,
+      "step": 6259
+    },
+    {
+      "epoch": 3.105250093086757,
+      "grad_norm": 0.07676748643472597,
+      "learning_rate": 1.1826521121071476e-06,
+      "loss": 0.4541,
+      "step": 6260
+    },
+    {
+      "epoch": 3.105746555789996,
+      "grad_norm": 0.074831957192136,
+      "learning_rate": 1.1813899734653673e-06,
+      "loss": 0.4548,
+      "step": 6261
+    },
+    {
+      "epoch": 3.1062430184932355,
+      "grad_norm": 0.07346740038768987,
+      "learning_rate": 1.1801284184369022e-06,
+      "loss": 0.426,
+      "step": 6262
+    },
+    {
+      "epoch": 3.1067394811964752,
+      "grad_norm": 0.07200241111962989,
+      "learning_rate": 1.1788674472145607e-06,
+      "loss": 0.426,
+      "step": 6263
+    },
+    {
+      "epoch": 3.1072359438997146,
+      "grad_norm": 0.0697525499088635,
+      "learning_rate": 1.177607059991065e-06,
+      "loss": 0.4152,
+      "step": 6264
+    },
+    {
+      "epoch": 3.107732406602954,
+      "grad_norm": 0.0706362300791092,
+      "learning_rate": 1.1763472569590405e-06,
+      "loss": 0.4381,
+      "step": 6265
+    },
+    {
+      "epoch": 3.1082288693061932,
+      "grad_norm": 0.07258543705591866,
+      "learning_rate": 1.1750880383110313e-06,
+      "loss": 0.4223,
+      "step": 6266
+    },
+    {
+      "epoch": 3.108725332009433,
+      "grad_norm": 0.0730392166405372,
+      "learning_rate": 1.1738294042394859e-06,
+      "loss": 0.4464,
+      "step": 6267
+    },
+    {
+      "epoch": 3.1092217947126723,
+      "grad_norm": 0.07034950376141262,
+      "learning_rate": 1.17257135493677e-06,
+      "loss": 0.4271,
+      "step": 6268
+    },
+    {
+      "epoch": 3.1097182574159117,
+      "grad_norm": 0.07318720180760736,
+      "learning_rate": 1.1713138905951538e-06,
+      "loss": 0.4745,
+      "step": 6269
+    },
+    {
+      "epoch": 3.110214720119151,
+      "grad_norm": 0.07215315623619104,
+      "learning_rate": 1.1700570114068204e-06,
+      "loss": 0.4306,
+      "step": 6270
+    },
+    {
+      "epoch": 3.1107111828223903,
+      "grad_norm": 0.07331765254976616,
+      "learning_rate": 1.1688007175638655e-06,
+      "loss": 0.4265,
+      "step": 6271
+    },
+    {
+      "epoch": 3.1112076455256297,
+      "grad_norm": 0.0729633099311388,
+      "learning_rate": 1.1675450092582908e-06,
+      "loss": 0.4234,
+      "step": 6272
+    },
+    {
+      "epoch": 3.1117041082288694,
+      "grad_norm": 0.07376419714140751,
+      "learning_rate": 1.1662898866820139e-06,
+      "loss": 0.4218,
+      "step": 6273
+    },
+    {
+      "epoch": 3.112200570932109,
+      "grad_norm": 0.076356924762706,
+      "learning_rate": 1.1650353500268592e-06,
+      "loss": 0.4411,
+      "step": 6274
+    },
+    {
+      "epoch": 3.112697033635348,
+      "grad_norm": 0.0722139900522232,
+      "learning_rate": 1.1637813994845604e-06,
+      "loss": 0.4335,
+      "step": 6275
+    },
+    {
+      "epoch": 3.1131934963385874,
+      "grad_norm": 0.0735799273371477,
+      "learning_rate": 1.1625280352467676e-06,
+      "loss": 0.4767,
+      "step": 6276
+    },
+    {
+      "epoch": 3.1136899590418268,
+      "grad_norm": 0.07136714993097293,
+      "learning_rate": 1.161275257505034e-06,
+      "loss": 0.4209,
+      "step": 6277
+    },
+    {
+      "epoch": 3.1141864217450665,
+      "grad_norm": 0.0728243696761246,
+      "learning_rate": 1.1600230664508288e-06,
+      "loss": 0.4442,
+      "step": 6278
+    },
+    {
+      "epoch": 3.114682884448306,
+      "grad_norm": 0.07277821478264992,
+      "learning_rate": 1.158771462275529e-06,
+      "loss": 0.4466,
+      "step": 6279
+    },
+    {
+      "epoch": 3.115179347151545,
+      "grad_norm": 0.07147557567766659,
+      "learning_rate": 1.1575204451704208e-06,
+      "loss": 0.4269,
+      "step": 6280
+    },
+    {
+      "epoch": 3.1156758098547845,
+      "grad_norm": 0.07178507916184856,
+      "learning_rate": 1.1562700153267053e-06,
+      "loss": 0.4215,
+      "step": 6281
+    },
+    {
+      "epoch": 3.116172272558024,
+      "grad_norm": 0.07484274630335912,
+      "learning_rate": 1.155020172935486e-06,
+      "loss": 0.4544,
+      "step": 6282
+    },
+    {
+      "epoch": 3.1166687352612636,
+      "grad_norm": 0.07133263940129107,
+      "learning_rate": 1.153770918187785e-06,
+      "loss": 0.4419,
+      "step": 6283
+    },
+    {
+      "epoch": 3.117165197964503,
+      "grad_norm": 0.07243435975635203,
+      "learning_rate": 1.1525222512745277e-06,
+      "loss": 0.4372,
+      "step": 6284
+    },
+    {
+      "epoch": 3.1176616606677423,
+      "grad_norm": 0.07267053715381483,
+      "learning_rate": 1.1512741723865562e-06,
+      "loss": 0.4248,
+      "step": 6285
+    },
+    {
+      "epoch": 3.1181581233709816,
+      "grad_norm": 0.0705129375128535,
+      "learning_rate": 1.1500266817146183e-06,
+      "loss": 0.4301,
+      "step": 6286
+    },
+    {
+      "epoch": 3.118654586074221,
+      "grad_norm": 0.07294580347031818,
+      "learning_rate": 1.1487797794493704e-06,
+      "loss": 0.4394,
+      "step": 6287
+    },
+    {
+      "epoch": 3.1191510487774607,
+      "grad_norm": 0.07004685048181185,
+      "learning_rate": 1.1475334657813858e-06,
+      "loss": 0.4043,
+      "step": 6288
+    },
+    {
+      "epoch": 3.1196475114807,
+      "grad_norm": 0.07101860635293346,
+      "learning_rate": 1.1462877409011396e-06,
+      "loss": 0.4585,
+      "step": 6289
+    },
+    {
+      "epoch": 3.1201439741839394,
+      "grad_norm": 0.07307878307525968,
+      "learning_rate": 1.1450426049990237e-06,
+      "loss": 0.4548,
+      "step": 6290
+    },
+    {
+      "epoch": 3.1206404368871787,
+      "grad_norm": 0.07427427511786734,
+      "learning_rate": 1.1437980582653364e-06,
+      "loss": 0.4561,
+      "step": 6291
+    },
+    {
+      "epoch": 3.121136899590418,
+      "grad_norm": 0.07270067758658703,
+      "learning_rate": 1.1425541008902852e-06,
+      "loss": 0.4385,
+      "step": 6292
+    },
+    {
+      "epoch": 3.121633362293658,
+      "grad_norm": 0.0730383047366194,
+      "learning_rate": 1.141310733063991e-06,
+      "loss": 0.4318,
+      "step": 6293
+    },
+    {
+      "epoch": 3.122129824996897,
+      "grad_norm": 0.07194331961927128,
+      "learning_rate": 1.1400679549764826e-06,
+      "loss": 0.4502,
+      "step": 6294
+    },
+    {
+      "epoch": 3.1226262877001365,
+      "grad_norm": 0.07276370816193148,
+      "learning_rate": 1.138825766817696e-06,
+      "loss": 0.4557,
+      "step": 6295
+    },
+    {
+      "epoch": 3.123122750403376,
+      "grad_norm": 0.07479660195719427,
+      "learning_rate": 1.1375841687774836e-06,
+      "loss": 0.4412,
+      "step": 6296
+    },
+    {
+      "epoch": 3.123619213106615,
+      "grad_norm": 0.07232842889066814,
+      "learning_rate": 1.1363431610456015e-06,
+      "loss": 0.4302,
+      "step": 6297
+    },
+    {
+      "epoch": 3.124115675809855,
+      "grad_norm": 0.07297903528073248,
+      "learning_rate": 1.1351027438117185e-06,
+      "loss": 0.4387,
+      "step": 6298
+    },
+    {
+      "epoch": 3.1246121385130943,
+      "grad_norm": 0.07194249745343922,
+      "learning_rate": 1.133862917265411e-06,
+      "loss": 0.4487,
+      "step": 6299
+    },
+    {
+      "epoch": 3.1251086012163336,
+      "grad_norm": 0.07472724689638927,
+      "learning_rate": 1.1326236815961683e-06,
+      "loss": 0.438,
+      "step": 6300
+    },
+    {
+      "epoch": 3.125605063919573,
+      "grad_norm": 0.07327862212617026,
+      "learning_rate": 1.1313850369933875e-06,
+      "loss": 0.4267,
+      "step": 6301
+    },
+    {
+      "epoch": 3.1261015266228123,
+      "grad_norm": 0.0717658959881318,
+      "learning_rate": 1.1301469836463747e-06,
+      "loss": 0.4662,
+      "step": 6302
+    },
+    {
+      "epoch": 3.126597989326052,
+      "grad_norm": 0.06887025184343006,
+      "learning_rate": 1.128909521744348e-06,
+      "loss": 0.416,
+      "step": 6303
+    },
+    {
+      "epoch": 3.1270944520292914,
+      "grad_norm": 0.0734880042440317,
+      "learning_rate": 1.1276726514764309e-06,
+      "loss": 0.4409,
+      "step": 6304
+    },
+    {
+      "epoch": 3.1275909147325307,
+      "grad_norm": 0.07515049650490668,
+      "learning_rate": 1.1264363730316623e-06,
+      "loss": 0.4584,
+      "step": 6305
+    },
+    {
+      "epoch": 3.12808737743577,
+      "grad_norm": 0.0760084427465557,
+      "learning_rate": 1.1252006865989868e-06,
+      "loss": 0.4372,
+      "step": 6306
+    },
+    {
+      "epoch": 3.1285838401390094,
+      "grad_norm": 0.07086822307829199,
+      "learning_rate": 1.123965592367257e-06,
+      "loss": 0.419,
+      "step": 6307
+    },
+    {
+      "epoch": 3.129080302842249,
+      "grad_norm": 0.06959338016658051,
+      "learning_rate": 1.1227310905252402e-06,
+      "loss": 0.3947,
+      "step": 6308
+    },
+    {
+      "epoch": 3.1295767655454885,
+      "grad_norm": 0.07331027924562893,
+      "learning_rate": 1.1214971812616083e-06,
+      "loss": 0.4441,
+      "step": 6309
+    },
+    {
+      "epoch": 3.130073228248728,
+      "grad_norm": 0.07242813224198356,
+      "learning_rate": 1.1202638647649456e-06,
+      "loss": 0.4346,
+      "step": 6310
+    },
+    {
+      "epoch": 3.130569690951967,
+      "grad_norm": 0.07372477735829013,
+      "learning_rate": 1.1190311412237448e-06,
+      "loss": 0.4342,
+      "step": 6311
+    },
+    {
+      "epoch": 3.1310661536552065,
+      "grad_norm": 0.07320995587740181,
+      "learning_rate": 1.117799010826406e-06,
+      "loss": 0.4417,
+      "step": 6312
+    },
+    {
+      "epoch": 3.1315626163584462,
+      "grad_norm": 0.07220257005066855,
+      "learning_rate": 1.1165674737612447e-06,
+      "loss": 0.4315,
+      "step": 6313
+    },
+    {
+      "epoch": 3.1320590790616856,
+      "grad_norm": 0.07047700088670927,
+      "learning_rate": 1.1153365302164765e-06,
+      "loss": 0.3936,
+      "step": 6314
+    },
+    {
+      "epoch": 3.132555541764925,
+      "grad_norm": 0.07480885332612296,
+      "learning_rate": 1.1141061803802344e-06,
+      "loss": 0.485,
+      "step": 6315
+    },
+    {
+      "epoch": 3.1330520044681642,
+      "grad_norm": 0.07306069618694938,
+      "learning_rate": 1.1128764244405564e-06,
+      "loss": 0.4415,
+      "step": 6316
+    },
+    {
+      "epoch": 3.1335484671714036,
+      "grad_norm": 0.07185953235470903,
+      "learning_rate": 1.111647262585393e-06,
+      "loss": 0.4144,
+      "step": 6317
+    },
+    {
+      "epoch": 3.1340449298746433,
+      "grad_norm": 0.07214525618658277,
+      "learning_rate": 1.1104186950026003e-06,
+      "loss": 0.445,
+      "step": 6318
+    },
+    {
+      "epoch": 3.1345413925778827,
+      "grad_norm": 0.07463149967030125,
+      "learning_rate": 1.1091907218799442e-06,
+      "loss": 0.4548,
+      "step": 6319
+    },
+    {
+      "epoch": 3.135037855281122,
+      "grad_norm": 0.07160401781801154,
+      "learning_rate": 1.107963343405103e-06,
+      "loss": 0.4191,
+      "step": 6320
+    },
+    {
+      "epoch": 3.1355343179843613,
+      "grad_norm": 0.07409967039536204,
+      "learning_rate": 1.1067365597656592e-06,
+      "loss": 0.4395,
+      "step": 6321
+    },
+    {
+      "epoch": 3.1360307806876007,
+      "grad_norm": 0.07520117080543616,
+      "learning_rate": 1.10551037114911e-06,
+      "loss": 0.4616,
+      "step": 6322
+    },
+    {
+      "epoch": 3.1365272433908404,
+      "grad_norm": 0.07411725758901368,
+      "learning_rate": 1.1042847777428573e-06,
+      "loss": 0.4527,
+      "step": 6323
+    },
+    {
+      "epoch": 3.1370237060940798,
+      "grad_norm": 0.07384442107466406,
+      "learning_rate": 1.103059779734212e-06,
+      "loss": 0.4346,
+      "step": 6324
+    },
+    {
+      "epoch": 3.137520168797319,
+      "grad_norm": 0.07447981354299785,
+      "learning_rate": 1.1018353773103979e-06,
+      "loss": 0.4372,
+      "step": 6325
+    },
+    {
+      "epoch": 3.1380166315005584,
+      "grad_norm": 0.07263836411793177,
+      "learning_rate": 1.100611570658543e-06,
+      "loss": 0.441,
+      "step": 6326
+    },
+    {
+      "epoch": 3.1385130942037978,
+      "grad_norm": 0.07281104769870578,
+      "learning_rate": 1.0993883599656885e-06,
+      "loss": 0.44,
+      "step": 6327
+    },
+    {
+      "epoch": 3.1390095569070375,
+      "grad_norm": 0.07305992366538848,
+      "learning_rate": 1.0981657454187816e-06,
+      "loss": 0.4222,
+      "step": 6328
+    },
+    {
+      "epoch": 3.139506019610277,
+      "grad_norm": 0.07490868994563034,
+      "learning_rate": 1.0969437272046795e-06,
+      "loss": 0.439,
+      "step": 6329
+    },
+    {
+      "epoch": 3.140002482313516,
+      "grad_norm": 0.07418711798138108,
+      "learning_rate": 1.0957223055101485e-06,
+      "loss": 0.4421,
+      "step": 6330
+    },
+    {
+      "epoch": 3.1404989450167555,
+      "grad_norm": 0.07387558875548761,
+      "learning_rate": 1.0945014805218607e-06,
+      "loss": 0.4777,
+      "step": 6331
+    },
+    {
+      "epoch": 3.140995407719995,
+      "grad_norm": 0.07110874455248554,
+      "learning_rate": 1.093281252426403e-06,
+      "loss": 0.431,
+      "step": 6332
+    },
+    {
+      "epoch": 3.1414918704232346,
+      "grad_norm": 0.07177204619436062,
+      "learning_rate": 1.0920616214102669e-06,
+      "loss": 0.4225,
+      "step": 6333
+    },
+    {
+      "epoch": 3.141988333126474,
+      "grad_norm": 0.07159994867603817,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 0.4401,
+      "step": 6334
+    },
+    {
+      "epoch": 3.1424847958297133,
+      "grad_norm": 0.07431659814634801,
+      "learning_rate": 1.0896241513614691e-06,
+      "loss": 0.4616,
+      "step": 6335
+    },
+    {
+      "epoch": 3.1429812585329526,
+      "grad_norm": 0.07415439237796943,
+      "learning_rate": 1.0884063127013355e-06,
+      "loss": 0.4235,
+      "step": 6336
+    },
+    {
+      "epoch": 3.143477721236192,
+      "grad_norm": 0.07277972879207999,
+      "learning_rate": 1.0871890718655815e-06,
+      "loss": 0.437,
+      "step": 6337
+    },
+    {
+      "epoch": 3.1439741839394317,
+      "grad_norm": 0.07089605624142527,
+      "learning_rate": 1.08597242904024e-06,
+      "loss": 0.3935,
+      "step": 6338
+    },
+    {
+      "epoch": 3.144470646642671,
+      "grad_norm": 0.07232687962567792,
+      "learning_rate": 1.0847563844112552e-06,
+      "loss": 0.4183,
+      "step": 6339
+    },
+    {
+      "epoch": 3.1449671093459104,
+      "grad_norm": 0.07480668654229129,
+      "learning_rate": 1.0835409381644819e-06,
+      "loss": 0.4317,
+      "step": 6340
+    },
+    {
+      "epoch": 3.1454635720491497,
+      "grad_norm": 0.0718756198950488,
+      "learning_rate": 1.0823260904856791e-06,
+      "loss": 0.4562,
+      "step": 6341
+    },
+    {
+      "epoch": 3.145960034752389,
+      "grad_norm": 0.07311229140682408,
+      "learning_rate": 1.0811118415605198e-06,
+      "loss": 0.4423,
+      "step": 6342
+    },
+    {
+      "epoch": 3.146456497455629,
+      "grad_norm": 0.07167006067349847,
+      "learning_rate": 1.079898191574581e-06,
+      "loss": 0.4406,
+      "step": 6343
+    },
+    {
+      "epoch": 3.146952960158868,
+      "grad_norm": 0.07363214240251828,
+      "learning_rate": 1.078685140713348e-06,
+      "loss": 0.4644,
+      "step": 6344
+    },
+    {
+      "epoch": 3.1474494228621075,
+      "grad_norm": 0.07392910711508574,
+      "learning_rate": 1.0774726891622206e-06,
+      "loss": 0.4424,
+      "step": 6345
+    },
+    {
+      "epoch": 3.147945885565347,
+      "grad_norm": 0.06997889628056078,
+      "learning_rate": 1.076260837106497e-06,
+      "loss": 0.4269,
+      "step": 6346
+    },
+    {
+      "epoch": 3.148442348268586,
+      "grad_norm": 0.07117551469703765,
+      "learning_rate": 1.0750495847313936e-06,
+      "loss": 0.3957,
+      "step": 6347
+    },
+    {
+      "epoch": 3.148938810971826,
+      "grad_norm": 0.0733274547222186,
+      "learning_rate": 1.0738389322220276e-06,
+      "loss": 0.4594,
+      "step": 6348
+    },
+    {
+      "epoch": 3.1494352736750653,
+      "grad_norm": 0.07420284307941621,
+      "learning_rate": 1.0726288797634316e-06,
+      "loss": 0.4519,
+      "step": 6349
+    },
+    {
+      "epoch": 3.1499317363783046,
+      "grad_norm": 0.06961884503816268,
+      "learning_rate": 1.0714194275405399e-06,
+      "loss": 0.4011,
+      "step": 6350
+    },
+    {
+      "epoch": 3.150428199081544,
+      "grad_norm": 0.07483189324920517,
+      "learning_rate": 1.0702105757381982e-06,
+      "loss": 0.4443,
+      "step": 6351
+    },
+    {
+      "epoch": 3.1509246617847833,
+      "grad_norm": 0.07206434301062702,
+      "learning_rate": 1.0690023245411613e-06,
+      "loss": 0.452,
+      "step": 6352
+    },
+    {
+      "epoch": 3.151421124488023,
+      "grad_norm": 0.07105045953167007,
+      "learning_rate": 1.0677946741340888e-06,
+      "loss": 0.4057,
+      "step": 6353
+    },
+    {
+      "epoch": 3.1519175871912624,
+      "grad_norm": 0.07108778081638781,
+      "learning_rate": 1.0665876247015545e-06,
+      "loss": 0.4222,
+      "step": 6354
+    },
+    {
+      "epoch": 3.1524140498945017,
+      "grad_norm": 0.07334785786034573,
+      "learning_rate": 1.0653811764280336e-06,
+      "loss": 0.4428,
+      "step": 6355
+    },
+    {
+      "epoch": 3.152910512597741,
+      "grad_norm": 0.07542948877894473,
+      "learning_rate": 1.064175329497912e-06,
+      "loss": 0.4917,
+      "step": 6356
+    },
+    {
+      "epoch": 3.1534069753009804,
+      "grad_norm": 0.072007413616065,
+      "learning_rate": 1.0629700840954866e-06,
+      "loss": 0.428,
+      "step": 6357
+    },
+    {
+      "epoch": 3.15390343800422,
+      "grad_norm": 0.07136862999002559,
+      "learning_rate": 1.0617654404049566e-06,
+      "loss": 0.4179,
+      "step": 6358
+    },
+    {
+      "epoch": 3.1543999007074595,
+      "grad_norm": 0.07166078367580378,
+      "learning_rate": 1.0605613986104357e-06,
+      "loss": 0.4425,
+      "step": 6359
+    },
+    {
+      "epoch": 3.154896363410699,
+      "grad_norm": 0.07290498579597421,
+      "learning_rate": 1.0593579588959412e-06,
+      "loss": 0.431,
+      "step": 6360
+    },
+    {
+      "epoch": 3.155392826113938,
+      "grad_norm": 0.07492878534177091,
+      "learning_rate": 1.058155121445399e-06,
+      "loss": 0.4495,
+      "step": 6361
+    },
+    {
+      "epoch": 3.1558892888171775,
+      "grad_norm": 0.073171418580753,
+      "learning_rate": 1.0569528864426444e-06,
+      "loss": 0.4395,
+      "step": 6362
+    },
+    {
+      "epoch": 3.1563857515204172,
+      "grad_norm": 0.07486368150671195,
+      "learning_rate": 1.055751254071417e-06,
+      "loss": 0.4636,
+      "step": 6363
+    },
+    {
+      "epoch": 3.1568822142236566,
+      "grad_norm": 0.07233176790796679,
+      "learning_rate": 1.054550224515371e-06,
+      "loss": 0.4491,
+      "step": 6364
+    },
+    {
+      "epoch": 3.157378676926896,
+      "grad_norm": 0.07546816185177538,
+      "learning_rate": 1.053349797958061e-06,
+      "loss": 0.4631,
+      "step": 6365
+    },
+    {
+      "epoch": 3.157875139630135,
+      "grad_norm": 0.07337190052275312,
+      "learning_rate": 1.052149974582956e-06,
+      "loss": 0.4368,
+      "step": 6366
+    },
+    {
+      "epoch": 3.1583716023333746,
+      "grad_norm": 0.07122430623194338,
+      "learning_rate": 1.0509507545734289e-06,
+      "loss": 0.4028,
+      "step": 6367
+    },
+    {
+      "epoch": 3.1588680650366143,
+      "grad_norm": 0.0737020536658153,
+      "learning_rate": 1.0497521381127595e-06,
+      "loss": 0.4069,
+      "step": 6368
+    },
+    {
+      "epoch": 3.1593645277398537,
+      "grad_norm": 0.07323902157198689,
+      "learning_rate": 1.0485541253841391e-06,
+      "loss": 0.4547,
+      "step": 6369
+    },
+    {
+      "epoch": 3.159860990443093,
+      "grad_norm": 0.07346998905499097,
+      "learning_rate": 1.0473567165706643e-06,
+      "loss": 0.4655,
+      "step": 6370
+    },
+    {
+      "epoch": 3.1603574531463323,
+      "grad_norm": 0.07434490171632929,
+      "learning_rate": 1.0461599118553383e-06,
+      "loss": 0.431,
+      "step": 6371
+    },
+    {
+      "epoch": 3.1608539158495716,
+      "grad_norm": 0.07278054746974955,
+      "learning_rate": 1.0449637114210765e-06,
+      "loss": 0.4377,
+      "step": 6372
+    },
+    {
+      "epoch": 3.1613503785528114,
+      "grad_norm": 0.076400465350522,
+      "learning_rate": 1.0437681154506951e-06,
+      "loss": 0.4499,
+      "step": 6373
+    },
+    {
+      "epoch": 3.1618468412560508,
+      "grad_norm": 0.07364498414420374,
+      "learning_rate": 1.0425731241269255e-06,
+      "loss": 0.4249,
+      "step": 6374
+    },
+    {
+      "epoch": 3.16234330395929,
+      "grad_norm": 0.07177693357710595,
+      "learning_rate": 1.041378737632402e-06,
+      "loss": 0.4007,
+      "step": 6375
+    },
+    {
+      "epoch": 3.1628397666625294,
+      "grad_norm": 0.075074506947561,
+      "learning_rate": 1.0401849561496647e-06,
+      "loss": 0.4399,
+      "step": 6376
+    },
+    {
+      "epoch": 3.1633362293657687,
+      "grad_norm": 0.07142224911160262,
+      "learning_rate": 1.0389917798611687e-06,
+      "loss": 0.4377,
+      "step": 6377
+    },
+    {
+      "epoch": 3.1638326920690085,
+      "grad_norm": 0.07646993575423702,
+      "learning_rate": 1.0377992089492666e-06,
+      "loss": 0.4689,
+      "step": 6378
+    },
+    {
+      "epoch": 3.164329154772248,
+      "grad_norm": 0.07544046356558812,
+      "learning_rate": 1.0366072435962283e-06,
+      "loss": 0.4547,
+      "step": 6379
+    },
+    {
+      "epoch": 3.164825617475487,
+      "grad_norm": 0.07212582877183446,
+      "learning_rate": 1.0354158839842226e-06,
+      "loss": 0.4336,
+      "step": 6380
+    },
+    {
+      "epoch": 3.1653220801787265,
+      "grad_norm": 0.0717380986515421,
+      "learning_rate": 1.0342251302953332e-06,
+      "loss": 0.4163,
+      "step": 6381
+    },
+    {
+      "epoch": 3.165818542881966,
+      "grad_norm": 0.07452498123015722,
+      "learning_rate": 1.0330349827115466e-06,
+      "loss": 0.461,
+      "step": 6382
+    },
+    {
+      "epoch": 3.166315005585205,
+      "grad_norm": 0.07269605501628315,
+      "learning_rate": 1.031845441414756e-06,
+      "loss": 0.4466,
+      "step": 6383
+    },
+    {
+      "epoch": 3.166811468288445,
+      "grad_norm": 0.07389670081610103,
+      "learning_rate": 1.0306565065867663e-06,
+      "loss": 0.4551,
+      "step": 6384
+    },
+    {
+      "epoch": 3.1673079309916843,
+      "grad_norm": 0.07430456948686814,
+      "learning_rate": 1.0294681784092847e-06,
+      "loss": 0.4422,
+      "step": 6385
+    },
+    {
+      "epoch": 3.1678043936949236,
+      "grad_norm": 0.07299404342467042,
+      "learning_rate": 1.028280457063931e-06,
+      "loss": 0.4246,
+      "step": 6386
+    },
+    {
+      "epoch": 3.168300856398163,
+      "grad_norm": 0.0739483810668814,
+      "learning_rate": 1.0270933427322277e-06,
+      "loss": 0.4374,
+      "step": 6387
+    },
+    {
+      "epoch": 3.1687973191014027,
+      "grad_norm": 0.07241633132347361,
+      "learning_rate": 1.0259068355956047e-06,
+      "loss": 0.4226,
+      "step": 6388
+    },
+    {
+      "epoch": 3.169293781804642,
+      "grad_norm": 0.07554269802453953,
+      "learning_rate": 1.0247209358354038e-06,
+      "loss": 0.4836,
+      "step": 6389
+    },
+    {
+      "epoch": 3.1697902445078814,
+      "grad_norm": 0.07350081967710687,
+      "learning_rate": 1.0235356436328675e-06,
+      "loss": 0.4304,
+      "step": 6390
+    },
+    {
+      "epoch": 3.1702867072111207,
+      "grad_norm": 0.07696261227487028,
+      "learning_rate": 1.0223509591691517e-06,
+      "loss": 0.4475,
+      "step": 6391
+    },
+    {
+      "epoch": 3.17078316991436,
+      "grad_norm": 0.07744577811871206,
+      "learning_rate": 1.0211668826253147e-06,
+      "loss": 0.5039,
+      "step": 6392
+    },
+    {
+      "epoch": 3.1712796326175994,
+      "grad_norm": 0.07504917661298924,
+      "learning_rate": 1.0199834141823244e-06,
+      "loss": 0.4789,
+      "step": 6393
+    },
+    {
+      "epoch": 3.171776095320839,
+      "grad_norm": 0.07305119846164813,
+      "learning_rate": 1.0188005540210545e-06,
+      "loss": 0.4571,
+      "step": 6394
+    },
+    {
+      "epoch": 3.1722725580240785,
+      "grad_norm": 0.07491863925128799,
+      "learning_rate": 1.0176183023222847e-06,
+      "loss": 0.4501,
+      "step": 6395
+    },
+    {
+      "epoch": 3.172769020727318,
+      "grad_norm": 0.07452081497636626,
+      "learning_rate": 1.0164366592667063e-06,
+      "loss": 0.4654,
+      "step": 6396
+    },
+    {
+      "epoch": 3.173265483430557,
+      "grad_norm": 0.07618022845511371,
+      "learning_rate": 1.015255625034911e-06,
+      "loss": 0.4584,
+      "step": 6397
+    },
+    {
+      "epoch": 3.173761946133797,
+      "grad_norm": 0.07043046185576388,
+      "learning_rate": 1.014075199807405e-06,
+      "loss": 0.3949,
+      "step": 6398
+    },
+    {
+      "epoch": 3.1742584088370362,
+      "grad_norm": 0.07227920540362784,
+      "learning_rate": 1.012895383764595e-06,
+      "loss": 0.4514,
+      "step": 6399
+    },
+    {
+      "epoch": 3.1747548715402756,
+      "grad_norm": 0.07305956247134797,
+      "learning_rate": 1.0117161770867962e-06,
+      "loss": 0.4134,
+      "step": 6400
+    },
+    {
+      "epoch": 3.175251334243515,
+      "grad_norm": 0.07062457812145931,
+      "learning_rate": 1.0105375799542334e-06,
+      "loss": 0.4259,
+      "step": 6401
+    },
+    {
+      "epoch": 3.1757477969467542,
+      "grad_norm": 0.07382671423583863,
+      "learning_rate": 1.009359592547034e-06,
+      "loss": 0.4587,
+      "step": 6402
+    },
+    {
+      "epoch": 3.1762442596499936,
+      "grad_norm": 0.07164535630823299,
+      "learning_rate": 1.008182215045237e-06,
+      "loss": 0.4309,
+      "step": 6403
+    },
+    {
+      "epoch": 3.1767407223532333,
+      "grad_norm": 0.07310659452909928,
+      "learning_rate": 1.007005447628785e-06,
+      "loss": 0.4711,
+      "step": 6404
+    },
+    {
+      "epoch": 3.1772371850564727,
+      "grad_norm": 0.07450978246926393,
+      "learning_rate": 1.0058292904775257e-06,
+      "loss": 0.4181,
+      "step": 6405
+    },
+    {
+      "epoch": 3.177733647759712,
+      "grad_norm": 0.0716823391899861,
+      "learning_rate": 1.0046537437712196e-06,
+      "loss": 0.4086,
+      "step": 6406
+    },
+    {
+      "epoch": 3.1782301104629513,
+      "grad_norm": 0.07328278411569006,
+      "learning_rate": 1.003478807689528e-06,
+      "loss": 0.44,
+      "step": 6407
+    },
+    {
+      "epoch": 3.178726573166191,
+      "grad_norm": 0.07225344228354333,
+      "learning_rate": 1.0023044824120198e-06,
+      "loss": 0.4034,
+      "step": 6408
+    },
+    {
+      "epoch": 3.1792230358694304,
+      "grad_norm": 0.07361715251773729,
+      "learning_rate": 1.001130768118176e-06,
+      "loss": 0.4183,
+      "step": 6409
+    },
+    {
+      "epoch": 3.17971949857267,
+      "grad_norm": 0.07047270273695008,
+      "learning_rate": 9.999576649873744e-07,
+      "loss": 0.4054,
+      "step": 6410
+    },
+    {
+      "epoch": 3.180215961275909,
+      "grad_norm": 0.07264361495431505,
+      "learning_rate": 9.987851731989096e-07,
+      "loss": 0.4336,
+      "step": 6411
+    },
+    {
+      "epoch": 3.1807124239791484,
+      "grad_norm": 0.0726626866759461,
+      "learning_rate": 9.976132929319755e-07,
+      "loss": 0.4264,
+      "step": 6412
+    },
+    {
+      "epoch": 3.1812088866823878,
+      "grad_norm": 0.0768203689765329,
+      "learning_rate": 9.96442024365677e-07,
+      "loss": 0.4593,
+      "step": 6413
+    },
+    {
+      "epoch": 3.1817053493856275,
+      "grad_norm": 0.07293971639410052,
+      "learning_rate": 9.952713676790227e-07,
+      "loss": 0.4796,
+      "step": 6414
+    },
+    {
+      "epoch": 3.182201812088867,
+      "grad_norm": 0.07490695058582275,
+      "learning_rate": 9.941013230509278e-07,
+      "loss": 0.4807,
+      "step": 6415
+    },
+    {
+      "epoch": 3.182698274792106,
+      "grad_norm": 0.06927060845757176,
+      "learning_rate": 9.929318906602176e-07,
+      "loss": 0.4096,
+      "step": 6416
+    },
+    {
+      "epoch": 3.1831947374953455,
+      "grad_norm": 0.07276995387218554,
+      "learning_rate": 9.91763070685618e-07,
+      "loss": 0.4309,
+      "step": 6417
+    },
+    {
+      "epoch": 3.1836912001985853,
+      "grad_norm": 0.07277977198244508,
+      "learning_rate": 9.905948633057666e-07,
+      "loss": 0.4338,
+      "step": 6418
+    },
+    {
+      "epoch": 3.1841876629018246,
+      "grad_norm": 0.07314677971605135,
+      "learning_rate": 9.894272686992052e-07,
+      "loss": 0.4528,
+      "step": 6419
+    },
+    {
+      "epoch": 3.184684125605064,
+      "grad_norm": 0.07423666704428677,
+      "learning_rate": 9.882602870443796e-07,
+      "loss": 0.4465,
+      "step": 6420
+    },
+    {
+      "epoch": 3.1851805883083033,
+      "grad_norm": 0.0732412581872415,
+      "learning_rate": 9.87093918519647e-07,
+      "loss": 0.4428,
+      "step": 6421
+    },
+    {
+      "epoch": 3.1856770510115426,
+      "grad_norm": 0.06905045477934818,
+      "learning_rate": 9.859281633032653e-07,
+      "loss": 0.4127,
+      "step": 6422
+    },
+    {
+      "epoch": 3.186173513714782,
+      "grad_norm": 0.0720047625731592,
+      "learning_rate": 9.84763021573405e-07,
+      "loss": 0.4335,
+      "step": 6423
+    },
+    {
+      "epoch": 3.1866699764180217,
+      "grad_norm": 0.07284181108689597,
+      "learning_rate": 9.835984935081371e-07,
+      "loss": 0.4232,
+      "step": 6424
+    },
+    {
+      "epoch": 3.187166439121261,
+      "grad_norm": 0.07034437883600633,
+      "learning_rate": 9.82434579285441e-07,
+      "loss": 0.4335,
+      "step": 6425
+    },
+    {
+      "epoch": 3.1876629018245004,
+      "grad_norm": 0.07541412625398493,
+      "learning_rate": 9.812712790832035e-07,
+      "loss": 0.4843,
+      "step": 6426
+    },
+    {
+      "epoch": 3.1881593645277397,
+      "grad_norm": 0.07461764739658933,
+      "learning_rate": 9.801085930792138e-07,
+      "loss": 0.4556,
+      "step": 6427
+    },
+    {
+      "epoch": 3.188655827230979,
+      "grad_norm": 0.07150957407049094,
+      "learning_rate": 9.789465214511729e-07,
+      "loss": 0.423,
+      "step": 6428
+    },
+    {
+      "epoch": 3.189152289934219,
+      "grad_norm": 0.07748019008724999,
+      "learning_rate": 9.777850643766823e-07,
+      "loss": 0.4611,
+      "step": 6429
+    },
+    {
+      "epoch": 3.189648752637458,
+      "grad_norm": 0.07366850775212323,
+      "learning_rate": 9.766242220332544e-07,
+      "loss": 0.441,
+      "step": 6430
+    },
+    {
+      "epoch": 3.1901452153406975,
+      "grad_norm": 0.07468096275004382,
+      "learning_rate": 9.754639945983041e-07,
+      "loss": 0.4764,
+      "step": 6431
+    },
+    {
+      "epoch": 3.190641678043937,
+      "grad_norm": 0.07281699571452858,
+      "learning_rate": 9.743043822491528e-07,
+      "loss": 0.4225,
+      "step": 6432
+    },
+    {
+      "epoch": 3.191138140747176,
+      "grad_norm": 0.07269377638605191,
+      "learning_rate": 9.731453851630308e-07,
+      "loss": 0.4481,
+      "step": 6433
+    },
+    {
+      "epoch": 3.191634603450416,
+      "grad_norm": 0.07479476219760563,
+      "learning_rate": 9.719870035170697e-07,
+      "loss": 0.4536,
+      "step": 6434
+    },
+    {
+      "epoch": 3.1921310661536553,
+      "grad_norm": 0.06903390753037493,
+      "learning_rate": 9.708292374883121e-07,
+      "loss": 0.4096,
+      "step": 6435
+    },
+    {
+      "epoch": 3.1926275288568946,
+      "grad_norm": 0.0752804219093628,
+      "learning_rate": 9.696720872537023e-07,
+      "loss": 0.4711,
+      "step": 6436
+    },
+    {
+      "epoch": 3.193123991560134,
+      "grad_norm": 0.07274574408985673,
+      "learning_rate": 9.68515552990092e-07,
+      "loss": 0.4552,
+      "step": 6437
+    },
+    {
+      "epoch": 3.1936204542633733,
+      "grad_norm": 0.07134757154327871,
+      "learning_rate": 9.673596348742404e-07,
+      "loss": 0.4354,
+      "step": 6438
+    },
+    {
+      "epoch": 3.194116916966613,
+      "grad_norm": 0.07197174710606297,
+      "learning_rate": 9.662043330828086e-07,
+      "loss": 0.4185,
+      "step": 6439
+    },
+    {
+      "epoch": 3.1946133796698524,
+      "grad_norm": 0.0757707799900384,
+      "learning_rate": 9.650496477923687e-07,
+      "loss": 0.4806,
+      "step": 6440
+    },
+    {
+      "epoch": 3.1951098423730917,
+      "grad_norm": 0.07346692458731068,
+      "learning_rate": 9.638955791793952e-07,
+      "loss": 0.4359,
+      "step": 6441
+    },
+    {
+      "epoch": 3.195606305076331,
+      "grad_norm": 0.07060145600998394,
+      "learning_rate": 9.62742127420268e-07,
+      "loss": 0.4094,
+      "step": 6442
+    },
+    {
+      "epoch": 3.1961027677795704,
+      "grad_norm": 0.07566580712909188,
+      "learning_rate": 9.615892926912745e-07,
+      "loss": 0.4724,
+      "step": 6443
+    },
+    {
+      "epoch": 3.19659923048281,
+      "grad_norm": 0.07197491092931253,
+      "learning_rate": 9.60437075168605e-07,
+      "loss": 0.4514,
+      "step": 6444
+    },
+    {
+      "epoch": 3.1970956931860495,
+      "grad_norm": 0.07526582352172888,
+      "learning_rate": 9.592854750283604e-07,
+      "loss": 0.451,
+      "step": 6445
+    },
+    {
+      "epoch": 3.197592155889289,
+      "grad_norm": 0.07443960349293295,
+      "learning_rate": 9.58134492446543e-07,
+      "loss": 0.4488,
+      "step": 6446
+    },
+    {
+      "epoch": 3.198088618592528,
+      "grad_norm": 0.07285278760288882,
+      "learning_rate": 9.569841275990611e-07,
+      "loss": 0.4734,
+      "step": 6447
+    },
+    {
+      "epoch": 3.1985850812957675,
+      "grad_norm": 0.07308057042372908,
+      "learning_rate": 9.558343806617316e-07,
+      "loss": 0.4407,
+      "step": 6448
+    },
+    {
+      "epoch": 3.1990815439990072,
+      "grad_norm": 0.07175580604720955,
+      "learning_rate": 9.546852518102723e-07,
+      "loss": 0.4589,
+      "step": 6449
+    },
+    {
+      "epoch": 3.1995780067022466,
+      "grad_norm": 0.07329134585274771,
+      "learning_rate": 9.535367412203117e-07,
+      "loss": 0.4387,
+      "step": 6450
+    },
+    {
+      "epoch": 3.200074469405486,
+      "grad_norm": 0.07397527035561834,
+      "learning_rate": 9.523888490673805e-07,
+      "loss": 0.4383,
+      "step": 6451
+    },
+    {
+      "epoch": 3.2005709321087252,
+      "grad_norm": 0.06990958524922075,
+      "learning_rate": 9.512415755269139e-07,
+      "loss": 0.4369,
+      "step": 6452
+    },
+    {
+      "epoch": 3.2010673948119646,
+      "grad_norm": 0.07557786165555054,
+      "learning_rate": 9.500949207742566e-07,
+      "loss": 0.4607,
+      "step": 6453
+    },
+    {
+      "epoch": 3.2015638575152043,
+      "grad_norm": 0.07303988553622057,
+      "learning_rate": 9.48948884984654e-07,
+      "loss": 0.4558,
+      "step": 6454
+    },
+    {
+      "epoch": 3.2020603202184437,
+      "grad_norm": 0.07264926895563321,
+      "learning_rate": 9.478034683332621e-07,
+      "loss": 0.4355,
+      "step": 6455
+    },
+    {
+      "epoch": 3.202556782921683,
+      "grad_norm": 0.07321784380890624,
+      "learning_rate": 9.466586709951381e-07,
+      "loss": 0.4615,
+      "step": 6456
+    },
+    {
+      "epoch": 3.2030532456249223,
+      "grad_norm": 0.07183308431742852,
+      "learning_rate": 9.455144931452459e-07,
+      "loss": 0.4633,
+      "step": 6457
+    },
+    {
+      "epoch": 3.2035497083281617,
+      "grad_norm": 0.0760389707040024,
+      "learning_rate": 9.443709349584546e-07,
+      "loss": 0.4849,
+      "step": 6458
+    },
+    {
+      "epoch": 3.2040461710314014,
+      "grad_norm": 0.07351781761787028,
+      "learning_rate": 9.432279966095376e-07,
+      "loss": 0.4174,
+      "step": 6459
+    },
+    {
+      "epoch": 3.2045426337346408,
+      "grad_norm": 0.07417225150428665,
+      "learning_rate": 9.420856782731774e-07,
+      "loss": 0.4439,
+      "step": 6460
+    },
+    {
+      "epoch": 3.20503909643788,
+      "grad_norm": 0.0723067891275983,
+      "learning_rate": 9.409439801239561e-07,
+      "loss": 0.4309,
+      "step": 6461
+    },
+    {
+      "epoch": 3.2055355591411194,
+      "grad_norm": 0.07521119417390677,
+      "learning_rate": 9.398029023363664e-07,
+      "loss": 0.4372,
+      "step": 6462
+    },
+    {
+      "epoch": 3.2060320218443588,
+      "grad_norm": 0.07288334819556669,
+      "learning_rate": 9.386624450848031e-07,
+      "loss": 0.4368,
+      "step": 6463
+    },
+    {
+      "epoch": 3.2065284845475985,
+      "grad_norm": 0.07580814940598336,
+      "learning_rate": 9.375226085435652e-07,
+      "loss": 0.4715,
+      "step": 6464
+    },
+    {
+      "epoch": 3.207024947250838,
+      "grad_norm": 0.07768050645172682,
+      "learning_rate": 9.36383392886861e-07,
+      "loss": 0.4533,
+      "step": 6465
+    },
+    {
+      "epoch": 3.207521409954077,
+      "grad_norm": 0.07101000923713795,
+      "learning_rate": 9.352447982887986e-07,
+      "loss": 0.4064,
+      "step": 6466
+    },
+    {
+      "epoch": 3.2080178726573165,
+      "grad_norm": 0.07276121017150415,
+      "learning_rate": 9.341068249233964e-07,
+      "loss": 0.4469,
+      "step": 6467
+    },
+    {
+      "epoch": 3.208514335360556,
+      "grad_norm": 0.07130837808449741,
+      "learning_rate": 9.32969472964575e-07,
+      "loss": 0.4168,
+      "step": 6468
+    },
+    {
+      "epoch": 3.2090107980637956,
+      "grad_norm": 0.06987465664506198,
+      "learning_rate": 9.318327425861584e-07,
+      "loss": 0.4124,
+      "step": 6469
+    },
+    {
+      "epoch": 3.209507260767035,
+      "grad_norm": 0.07393044314992846,
+      "learning_rate": 9.3069663396188e-07,
+      "loss": 0.4618,
+      "step": 6470
+    },
+    {
+      "epoch": 3.2100037234702743,
+      "grad_norm": 0.07386630528663508,
+      "learning_rate": 9.295611472653737e-07,
+      "loss": 0.4777,
+      "step": 6471
+    },
+    {
+      "epoch": 3.2105001861735136,
+      "grad_norm": 0.07340938735988556,
+      "learning_rate": 9.284262826701823e-07,
+      "loss": 0.4472,
+      "step": 6472
+    },
+    {
+      "epoch": 3.210996648876753,
+      "grad_norm": 0.07400222313268992,
+      "learning_rate": 9.272920403497515e-07,
+      "loss": 0.4424,
+      "step": 6473
+    },
+    {
+      "epoch": 3.2114931115799927,
+      "grad_norm": 0.07318685134351284,
+      "learning_rate": 9.26158420477431e-07,
+      "loss": 0.4171,
+      "step": 6474
+    },
+    {
+      "epoch": 3.211989574283232,
+      "grad_norm": 0.07338316230220633,
+      "learning_rate": 9.250254232264772e-07,
+      "loss": 0.4351,
+      "step": 6475
+    },
+    {
+      "epoch": 3.2124860369864714,
+      "grad_norm": 0.0741471470218761,
+      "learning_rate": 9.238930487700487e-07,
+      "loss": 0.4822,
+      "step": 6476
+    },
+    {
+      "epoch": 3.2129824996897107,
+      "grad_norm": 0.07200053996890503,
+      "learning_rate": 9.227612972812139e-07,
+      "loss": 0.4385,
+      "step": 6477
+    },
+    {
+      "epoch": 3.21347896239295,
+      "grad_norm": 0.073457939378401,
+      "learning_rate": 9.216301689329393e-07,
+      "loss": 0.4168,
+      "step": 6478
+    },
+    {
+      "epoch": 3.21397542509619,
+      "grad_norm": 0.0756181528793698,
+      "learning_rate": 9.204996638981034e-07,
+      "loss": 0.4703,
+      "step": 6479
+    },
+    {
+      "epoch": 3.214471887799429,
+      "grad_norm": 0.0714661891520631,
+      "learning_rate": 9.193697823494846e-07,
+      "loss": 0.4305,
+      "step": 6480
+    },
+    {
+      "epoch": 3.2149683505026685,
+      "grad_norm": 0.07422591700487427,
+      "learning_rate": 9.182405244597647e-07,
+      "loss": 0.4465,
+      "step": 6481
+    },
+    {
+      "epoch": 3.215464813205908,
+      "grad_norm": 0.07131454314173444,
+      "learning_rate": 9.171118904015358e-07,
+      "loss": 0.4495,
+      "step": 6482
+    },
+    {
+      "epoch": 3.215961275909147,
+      "grad_norm": 0.07187530227156332,
+      "learning_rate": 9.159838803472904e-07,
+      "loss": 0.4511,
+      "step": 6483
+    },
+    {
+      "epoch": 3.216457738612387,
+      "grad_norm": 0.07481112874301267,
+      "learning_rate": 9.148564944694255e-07,
+      "loss": 0.4203,
+      "step": 6484
+    },
+    {
+      "epoch": 3.2169542013156263,
+      "grad_norm": 0.07372505230900529,
+      "learning_rate": 9.137297329402467e-07,
+      "loss": 0.4524,
+      "step": 6485
+    },
+    {
+      "epoch": 3.2174506640188656,
+      "grad_norm": 0.07173024709136377,
+      "learning_rate": 9.126035959319579e-07,
+      "loss": 0.464,
+      "step": 6486
+    },
+    {
+      "epoch": 3.217947126722105,
+      "grad_norm": 0.0724764831646147,
+      "learning_rate": 9.114780836166748e-07,
+      "loss": 0.4441,
+      "step": 6487
+    },
+    {
+      "epoch": 3.2184435894253443,
+      "grad_norm": 0.07239183478303986,
+      "learning_rate": 9.10353196166412e-07,
+      "loss": 0.4182,
+      "step": 6488
+    },
+    {
+      "epoch": 3.218940052128584,
+      "grad_norm": 0.0747413146612994,
+      "learning_rate": 9.092289337530907e-07,
+      "loss": 0.462,
+      "step": 6489
+    },
+    {
+      "epoch": 3.2194365148318234,
+      "grad_norm": 0.07302171673062939,
+      "learning_rate": 9.081052965485365e-07,
+      "loss": 0.4306,
+      "step": 6490
+    },
+    {
+      "epoch": 3.2199329775350627,
+      "grad_norm": 0.07440786317465062,
+      "learning_rate": 9.06982284724478e-07,
+      "loss": 0.4445,
+      "step": 6491
+    },
+    {
+      "epoch": 3.220429440238302,
+      "grad_norm": 0.0737803698740409,
+      "learning_rate": 9.058598984525518e-07,
+      "loss": 0.4701,
+      "step": 6492
+    },
+    {
+      "epoch": 3.2209259029415414,
+      "grad_norm": 0.07497800702697728,
+      "learning_rate": 9.047381379042941e-07,
+      "loss": 0.5053,
+      "step": 6493
+    },
+    {
+      "epoch": 3.221422365644781,
+      "grad_norm": 0.07190727725351215,
+      "learning_rate": 9.03617003251151e-07,
+      "loss": 0.4145,
+      "step": 6494
+    },
+    {
+      "epoch": 3.2219188283480205,
+      "grad_norm": 0.07583370749987378,
+      "learning_rate": 9.024964946644682e-07,
+      "loss": 0.4384,
+      "step": 6495
+    },
+    {
+      "epoch": 3.22241529105126,
+      "grad_norm": 0.07157005347045454,
+      "learning_rate": 9.013766123154965e-07,
+      "loss": 0.4445,
+      "step": 6496
+    },
+    {
+      "epoch": 3.222911753754499,
+      "grad_norm": 0.07665603143646456,
+      "learning_rate": 9.002573563753947e-07,
+      "loss": 0.4546,
+      "step": 6497
+    },
+    {
+      "epoch": 3.2234082164577385,
+      "grad_norm": 0.07106104303117836,
+      "learning_rate": 8.991387270152202e-07,
+      "loss": 0.4151,
+      "step": 6498
+    },
+    {
+      "epoch": 3.2239046791609782,
+      "grad_norm": 0.07238514243593716,
+      "learning_rate": 8.980207244059402e-07,
+      "loss": 0.4171,
+      "step": 6499
+    },
+    {
+      "epoch": 3.2244011418642176,
+      "grad_norm": 0.07173738237603448,
+      "learning_rate": 8.969033487184225e-07,
+      "loss": 0.4273,
+      "step": 6500
+    },
+    {
+      "epoch": 3.224897604567457,
+      "grad_norm": 0.07179915383513251,
+      "learning_rate": 8.957866001234383e-07,
+      "loss": 0.4591,
+      "step": 6501
+    },
+    {
+      "epoch": 3.225394067270696,
+      "grad_norm": 0.07446357660033671,
+      "learning_rate": 8.946704787916676e-07,
+      "loss": 0.4613,
+      "step": 6502
+    },
+    {
+      "epoch": 3.2258905299739355,
+      "grad_norm": 0.07259418059023226,
+      "learning_rate": 8.935549848936887e-07,
+      "loss": 0.4708,
+      "step": 6503
+    },
+    {
+      "epoch": 3.2263869926771753,
+      "grad_norm": 0.07650140470117987,
+      "learning_rate": 8.924401185999904e-07,
+      "loss": 0.4464,
+      "step": 6504
+    },
+    {
+      "epoch": 3.2268834553804147,
+      "grad_norm": 0.07349494700194908,
+      "learning_rate": 8.913258800809598e-07,
+      "loss": 0.4133,
+      "step": 6505
+    },
+    {
+      "epoch": 3.227379918083654,
+      "grad_norm": 0.0741847627217174,
+      "learning_rate": 8.902122695068905e-07,
+      "loss": 0.4239,
+      "step": 6506
+    },
+    {
+      "epoch": 3.2278763807868933,
+      "grad_norm": 0.07361488605938472,
+      "learning_rate": 8.890992870479809e-07,
+      "loss": 0.4447,
+      "step": 6507
+    },
+    {
+      "epoch": 3.2283728434901326,
+      "grad_norm": 0.07135406931538264,
+      "learning_rate": 8.879869328743306e-07,
+      "loss": 0.4152,
+      "step": 6508
+    },
+    {
+      "epoch": 3.2288693061933724,
+      "grad_norm": 0.07102552704377421,
+      "learning_rate": 8.868752071559478e-07,
+      "loss": 0.3938,
+      "step": 6509
+    },
+    {
+      "epoch": 3.2293657688966118,
+      "grad_norm": 0.07214553650346917,
+      "learning_rate": 8.857641100627395e-07,
+      "loss": 0.453,
+      "step": 6510
+    },
+    {
+      "epoch": 3.229862231599851,
+      "grad_norm": 0.07070501925694196,
+      "learning_rate": 8.846536417645213e-07,
+      "loss": 0.4183,
+      "step": 6511
+    },
+    {
+      "epoch": 3.2303586943030904,
+      "grad_norm": 0.07286313246206055,
+      "learning_rate": 8.835438024310095e-07,
+      "loss": 0.4319,
+      "step": 6512
+    },
+    {
+      "epoch": 3.2308551570063297,
+      "grad_norm": 0.07248086505232414,
+      "learning_rate": 8.824345922318234e-07,
+      "loss": 0.4465,
+      "step": 6513
+    },
+    {
+      "epoch": 3.2313516197095695,
+      "grad_norm": 0.07204426946399531,
+      "learning_rate": 8.813260113364913e-07,
+      "loss": 0.4534,
+      "step": 6514
+    },
+    {
+      "epoch": 3.231848082412809,
+      "grad_norm": 0.07261340576260167,
+      "learning_rate": 8.802180599144394e-07,
+      "loss": 0.4506,
+      "step": 6515
+    },
+    {
+      "epoch": 3.232344545116048,
+      "grad_norm": 0.07080102187886099,
+      "learning_rate": 8.791107381350028e-07,
+      "loss": 0.4398,
+      "step": 6516
+    },
+    {
+      "epoch": 3.2328410078192875,
+      "grad_norm": 0.0729790380657279,
+      "learning_rate": 8.780040461674161e-07,
+      "loss": 0.4562,
+      "step": 6517
+    },
+    {
+      "epoch": 3.233337470522527,
+      "grad_norm": 0.07468382100067729,
+      "learning_rate": 8.768979841808184e-07,
+      "loss": 0.4425,
+      "step": 6518
+    },
+    {
+      "epoch": 3.2338339332257666,
+      "grad_norm": 0.0726797100674497,
+      "learning_rate": 8.757925523442562e-07,
+      "loss": 0.4553,
+      "step": 6519
+    },
+    {
+      "epoch": 3.234330395929006,
+      "grad_norm": 0.0717963078697041,
+      "learning_rate": 8.74687750826676e-07,
+      "loss": 0.4297,
+      "step": 6520
+    },
+    {
+      "epoch": 3.2348268586322453,
+      "grad_norm": 0.07134789754728933,
+      "learning_rate": 8.735835797969272e-07,
+      "loss": 0.4024,
+      "step": 6521
+    },
+    {
+      "epoch": 3.2353233213354846,
+      "grad_norm": 0.07157494654779836,
+      "learning_rate": 8.724800394237675e-07,
+      "loss": 0.436,
+      "step": 6522
+    },
+    {
+      "epoch": 3.235819784038724,
+      "grad_norm": 0.07397407102075705,
+      "learning_rate": 8.713771298758539e-07,
+      "loss": 0.4178,
+      "step": 6523
+    },
+    {
+      "epoch": 3.2363162467419633,
+      "grad_norm": 0.07244233820873923,
+      "learning_rate": 8.702748513217491e-07,
+      "loss": 0.4178,
+      "step": 6524
+    },
+    {
+      "epoch": 3.236812709445203,
+      "grad_norm": 0.07271891990618194,
+      "learning_rate": 8.691732039299167e-07,
+      "loss": 0.4409,
+      "step": 6525
+    },
+    {
+      "epoch": 3.2373091721484424,
+      "grad_norm": 0.07477147839968602,
+      "learning_rate": 8.680721878687281e-07,
+      "loss": 0.4558,
+      "step": 6526
+    },
+    {
+      "epoch": 3.2378056348516817,
+      "grad_norm": 0.07443087408775124,
+      "learning_rate": 8.669718033064556e-07,
+      "loss": 0.4303,
+      "step": 6527
+    },
+    {
+      "epoch": 3.238302097554921,
+      "grad_norm": 0.07495800672885247,
+      "learning_rate": 8.658720504112733e-07,
+      "loss": 0.4674,
+      "step": 6528
+    },
+    {
+      "epoch": 3.238798560258161,
+      "grad_norm": 0.07210207630859929,
+      "learning_rate": 8.647729293512642e-07,
+      "loss": 0.4576,
+      "step": 6529
+    },
+    {
+      "epoch": 3.2392950229614,
+      "grad_norm": 0.07586349592109952,
+      "learning_rate": 8.636744402944075e-07,
+      "loss": 0.45,
+      "step": 6530
+    },
+    {
+      "epoch": 3.2397914856646395,
+      "grad_norm": 0.06898888811284278,
+      "learning_rate": 8.625765834085936e-07,
+      "loss": 0.3799,
+      "step": 6531
+    },
+    {
+      "epoch": 3.240287948367879,
+      "grad_norm": 0.07370951952665843,
+      "learning_rate": 8.6147935886161e-07,
+      "loss": 0.441,
+      "step": 6532
+    },
+    {
+      "epoch": 3.240784411071118,
+      "grad_norm": 0.0726414034311996,
+      "learning_rate": 8.60382766821149e-07,
+      "loss": 0.4383,
+      "step": 6533
+    },
+    {
+      "epoch": 3.2412808737743575,
+      "grad_norm": 0.07261925823331064,
+      "learning_rate": 8.5928680745481e-07,
+      "loss": 0.421,
+      "step": 6534
+    },
+    {
+      "epoch": 3.2417773364775972,
+      "grad_norm": 0.0731272020451477,
+      "learning_rate": 8.581914809300895e-07,
+      "loss": 0.4459,
+      "step": 6535
+    },
+    {
+      "epoch": 3.2422737991808366,
+      "grad_norm": 0.07293597481858168,
+      "learning_rate": 8.570967874143937e-07,
+      "loss": 0.4536,
+      "step": 6536
+    },
+    {
+      "epoch": 3.242770261884076,
+      "grad_norm": 0.07281065453153535,
+      "learning_rate": 8.560027270750276e-07,
+      "loss": 0.4517,
+      "step": 6537
+    },
+    {
+      "epoch": 3.2432667245873152,
+      "grad_norm": 0.07396847550314199,
+      "learning_rate": 8.549093000792008e-07,
+      "loss": 0.4249,
+      "step": 6538
+    },
+    {
+      "epoch": 3.243763187290555,
+      "grad_norm": 0.07359192677743913,
+      "learning_rate": 8.538165065940263e-07,
+      "loss": 0.4451,
+      "step": 6539
+    },
+    {
+      "epoch": 3.2442596499937943,
+      "grad_norm": 0.07186329185364042,
+      "learning_rate": 8.527243467865176e-07,
+      "loss": 0.4144,
+      "step": 6540
+    },
+    {
+      "epoch": 3.2447561126970337,
+      "grad_norm": 0.07531998855534346,
+      "learning_rate": 8.51632820823598e-07,
+      "loss": 0.4572,
+      "step": 6541
+    },
+    {
+      "epoch": 3.245252575400273,
+      "grad_norm": 0.07640602680421152,
+      "learning_rate": 8.505419288720862e-07,
+      "loss": 0.4686,
+      "step": 6542
+    },
+    {
+      "epoch": 3.2457490381035123,
+      "grad_norm": 0.072281263654525,
+      "learning_rate": 8.494516710987105e-07,
+      "loss": 0.4187,
+      "step": 6543
+    },
+    {
+      "epoch": 3.2462455008067517,
+      "grad_norm": 0.07397887512980252,
+      "learning_rate": 8.483620476700977e-07,
+      "loss": 0.471,
+      "step": 6544
+    },
+    {
+      "epoch": 3.2467419635099914,
+      "grad_norm": 0.07094087678701722,
+      "learning_rate": 8.472730587527783e-07,
+      "loss": 0.4385,
+      "step": 6545
+    },
+    {
+      "epoch": 3.2472384262132308,
+      "grad_norm": 0.07255957063372506,
+      "learning_rate": 8.461847045131894e-07,
+      "loss": 0.4559,
+      "step": 6546
+    },
+    {
+      "epoch": 3.24773488891647,
+      "grad_norm": 0.07083546706731952,
+      "learning_rate": 8.450969851176655e-07,
+      "loss": 0.4517,
+      "step": 6547
+    },
+    {
+      "epoch": 3.2482313516197094,
+      "grad_norm": 0.06990490031823457,
+      "learning_rate": 8.440099007324498e-07,
+      "loss": 0.4226,
+      "step": 6548
+    },
+    {
+      "epoch": 3.248727814322949,
+      "grad_norm": 0.07250659990226761,
+      "learning_rate": 8.429234515236845e-07,
+      "loss": 0.4105,
+      "step": 6549
+    },
+    {
+      "epoch": 3.2492242770261885,
+      "grad_norm": 0.07359684298440573,
+      "learning_rate": 8.41837637657415e-07,
+      "loss": 0.4406,
+      "step": 6550
+    },
+    {
+      "epoch": 3.249720739729428,
+      "grad_norm": 0.0715269597858152,
+      "learning_rate": 8.40752459299593e-07,
+      "loss": 0.4393,
+      "step": 6551
+    },
+    {
+      "epoch": 3.250217202432667,
+      "grad_norm": 0.07355218307599753,
+      "learning_rate": 8.396679166160676e-07,
+      "loss": 0.4328,
+      "step": 6552
+    },
+    {
+      "epoch": 3.250217202432667,
+      "eval_loss": 0.5165188908576965,
+      "eval_runtime": 258.8306,
+      "eval_samples_per_second": 117.27,
+      "eval_steps_per_second": 14.662,
+      "step": 6552
+    },
+    {
+      "epoch": 3.2507136651359065,
+      "grad_norm": 0.07283223774832169,
+      "learning_rate": 8.385840097725961e-07,
+      "loss": 0.4602,
+      "step": 6553
+    },
+    {
+      "epoch": 3.251210127839146,
+      "grad_norm": 0.07311451943838647,
+      "learning_rate": 8.375007389348361e-07,
+      "loss": 0.4147,
+      "step": 6554
+    },
+    {
+      "epoch": 3.2517065905423856,
+      "grad_norm": 0.07258731207752103,
+      "learning_rate": 8.364181042683472e-07,
+      "loss": 0.4504,
+      "step": 6555
+    },
+    {
+      "epoch": 3.252203053245625,
+      "grad_norm": 0.07437859093988324,
+      "learning_rate": 8.353361059385934e-07,
+      "loss": 0.4295,
+      "step": 6556
+    },
+    {
+      "epoch": 3.2526995159488643,
+      "grad_norm": 0.07547480599085805,
+      "learning_rate": 8.342547441109389e-07,
+      "loss": 0.468,
+      "step": 6557
+    },
+    {
+      "epoch": 3.2531959786521036,
+      "grad_norm": 0.0736194297795164,
+      "learning_rate": 8.33174018950656e-07,
+      "loss": 0.4362,
+      "step": 6558
+    },
+    {
+      "epoch": 3.2536924413553434,
+      "grad_norm": 0.07181393467415445,
+      "learning_rate": 8.320939306229125e-07,
+      "loss": 0.4473,
+      "step": 6559
+    },
+    {
+      "epoch": 3.2541889040585827,
+      "grad_norm": 0.07282434869419582,
+      "learning_rate": 8.310144792927855e-07,
+      "loss": 0.441,
+      "step": 6560
+    },
+    {
+      "epoch": 3.254685366761822,
+      "grad_norm": 0.07474077075724365,
+      "learning_rate": 8.299356651252511e-07,
+      "loss": 0.4461,
+      "step": 6561
+    },
+    {
+      "epoch": 3.2551818294650614,
+      "grad_norm": 0.0722199749212488,
+      "learning_rate": 8.288574882851874e-07,
+      "loss": 0.4277,
+      "step": 6562
+    },
+    {
+      "epoch": 3.2556782921683007,
+      "grad_norm": 0.06972358362863357,
+      "learning_rate": 8.277799489373784e-07,
+      "loss": 0.4003,
+      "step": 6563
+    },
+    {
+      "epoch": 3.25617475487154,
+      "grad_norm": 0.07348155810926192,
+      "learning_rate": 8.267030472465077e-07,
+      "loss": 0.4195,
+      "step": 6564
+    },
+    {
+      "epoch": 3.25667121757478,
+      "grad_norm": 0.0723048472609885,
+      "learning_rate": 8.256267833771608e-07,
+      "loss": 0.4733,
+      "step": 6565
+    },
+    {
+      "epoch": 3.257167680278019,
+      "grad_norm": 0.0700181266923299,
+      "learning_rate": 8.245511574938309e-07,
+      "loss": 0.4238,
+      "step": 6566
+    },
+    {
+      "epoch": 3.2576641429812585,
+      "grad_norm": 0.07302398518997803,
+      "learning_rate": 8.234761697609073e-07,
+      "loss": 0.4402,
+      "step": 6567
+    },
+    {
+      "epoch": 3.258160605684498,
+      "grad_norm": 0.07275318432187972,
+      "learning_rate": 8.224018203426864e-07,
+      "loss": 0.4229,
+      "step": 6568
+    },
+    {
+      "epoch": 3.2586570683877376,
+      "grad_norm": 0.07250942596803592,
+      "learning_rate": 8.213281094033648e-07,
+      "loss": 0.4626,
+      "step": 6569
+    },
+    {
+      "epoch": 3.259153531090977,
+      "grad_norm": 0.07289631800331221,
+      "learning_rate": 8.202550371070417e-07,
+      "loss": 0.4587,
+      "step": 6570
+    },
+    {
+      "epoch": 3.2596499937942163,
+      "grad_norm": 0.07463255498980362,
+      "learning_rate": 8.191826036177191e-07,
+      "loss": 0.4722,
+      "step": 6571
+    },
+    {
+      "epoch": 3.2601464564974556,
+      "grad_norm": 0.07140120891728631,
+      "learning_rate": 8.181108090993001e-07,
+      "loss": 0.4533,
+      "step": 6572
+    },
+    {
+      "epoch": 3.260642919200695,
+      "grad_norm": 0.07450462444078626,
+      "learning_rate": 8.170396537155934e-07,
+      "loss": 0.4234,
+      "step": 6573
+    },
+    {
+      "epoch": 3.2611393819039343,
+      "grad_norm": 0.0720324864948931,
+      "learning_rate": 8.159691376303059e-07,
+      "loss": 0.4616,
+      "step": 6574
+    },
+    {
+      "epoch": 3.261635844607174,
+      "grad_norm": 0.07382569910297018,
+      "learning_rate": 8.148992610070511e-07,
+      "loss": 0.4639,
+      "step": 6575
+    },
+    {
+      "epoch": 3.2621323073104134,
+      "grad_norm": 0.07158104231727033,
+      "learning_rate": 8.138300240093411e-07,
+      "loss": 0.4013,
+      "step": 6576
+    },
+    {
+      "epoch": 3.2626287700136527,
+      "grad_norm": 0.07524488381765479,
+      "learning_rate": 8.127614268005907e-07,
+      "loss": 0.455,
+      "step": 6577
+    },
+    {
+      "epoch": 3.263125232716892,
+      "grad_norm": 0.07346798365093564,
+      "learning_rate": 8.116934695441203e-07,
+      "loss": 0.4535,
+      "step": 6578
+    },
+    {
+      "epoch": 3.263621695420132,
+      "grad_norm": 0.07102589429860762,
+      "learning_rate": 8.106261524031473e-07,
+      "loss": 0.4505,
+      "step": 6579
+    },
+    {
+      "epoch": 3.264118158123371,
+      "grad_norm": 0.07197763258149469,
+      "learning_rate": 8.095594755407971e-07,
+      "loss": 0.4594,
+      "step": 6580
+    },
+    {
+      "epoch": 3.2646146208266105,
+      "grad_norm": 0.07320604094259688,
+      "learning_rate": 8.084934391200916e-07,
+      "loss": 0.4655,
+      "step": 6581
+    },
+    {
+      "epoch": 3.26511108352985,
+      "grad_norm": 0.07415582580125131,
+      "learning_rate": 8.074280433039577e-07,
+      "loss": 0.4337,
+      "step": 6582
+    },
+    {
+      "epoch": 3.265607546233089,
+      "grad_norm": 0.07681223797920941,
+      "learning_rate": 8.063632882552258e-07,
+      "loss": 0.4595,
+      "step": 6583
+    },
+    {
+      "epoch": 3.2661040089363285,
+      "grad_norm": 0.07285440928148708,
+      "learning_rate": 8.05299174136624e-07,
+      "loss": 0.4656,
+      "step": 6584
+    },
+    {
+      "epoch": 3.2666004716395682,
+      "grad_norm": 0.07396212889199635,
+      "learning_rate": 8.042357011107877e-07,
+      "loss": 0.437,
+      "step": 6585
+    },
+    {
+      "epoch": 3.2670969343428076,
+      "grad_norm": 0.07276458104314697,
+      "learning_rate": 8.031728693402502e-07,
+      "loss": 0.4507,
+      "step": 6586
+    },
+    {
+      "epoch": 3.267593397046047,
+      "grad_norm": 0.07299764124445621,
+      "learning_rate": 8.021106789874494e-07,
+      "loss": 0.461,
+      "step": 6587
+    },
+    {
+      "epoch": 3.2680898597492862,
+      "grad_norm": 0.069975440469783,
+      "learning_rate": 8.010491302147227e-07,
+      "loss": 0.4335,
+      "step": 6588
+    },
+    {
+      "epoch": 3.2685863224525256,
+      "grad_norm": 0.07246995485880185,
+      "learning_rate": 7.999882231843104e-07,
+      "loss": 0.4561,
+      "step": 6589
+    },
+    {
+      "epoch": 3.2690827851557653,
+      "grad_norm": 0.07541022380946706,
+      "learning_rate": 7.989279580583569e-07,
+      "loss": 0.4674,
+      "step": 6590
+    },
+    {
+      "epoch": 3.2695792478590047,
+      "grad_norm": 0.07260758175017704,
+      "learning_rate": 7.978683349989052e-07,
+      "loss": 0.4461,
+      "step": 6591
+    },
+    {
+      "epoch": 3.270075710562244,
+      "grad_norm": 0.07168818674294791,
+      "learning_rate": 7.968093541679039e-07,
+      "loss": 0.4314,
+      "step": 6592
+    },
+    {
+      "epoch": 3.2705721732654833,
+      "grad_norm": 0.07247063626781816,
+      "learning_rate": 7.957510157271991e-07,
+      "loss": 0.4335,
+      "step": 6593
+    },
+    {
+      "epoch": 3.2710686359687227,
+      "grad_norm": 0.07147894182938196,
+      "learning_rate": 7.946933198385409e-07,
+      "loss": 0.4468,
+      "step": 6594
+    },
+    {
+      "epoch": 3.2715650986719624,
+      "grad_norm": 0.06996799665630328,
+      "learning_rate": 7.936362666635827e-07,
+      "loss": 0.4263,
+      "step": 6595
+    },
+    {
+      "epoch": 3.2720615613752018,
+      "grad_norm": 0.07277142907800285,
+      "learning_rate": 7.92579856363876e-07,
+      "loss": 0.4432,
+      "step": 6596
+    },
+    {
+      "epoch": 3.272558024078441,
+      "grad_norm": 0.0718025553116523,
+      "learning_rate": 7.915240891008785e-07,
+      "loss": 0.4427,
+      "step": 6597
+    },
+    {
+      "epoch": 3.2730544867816804,
+      "grad_norm": 0.07453027337364937,
+      "learning_rate": 7.904689650359465e-07,
+      "loss": 0.4538,
+      "step": 6598
+    },
+    {
+      "epoch": 3.2735509494849198,
+      "grad_norm": 0.07212902370180194,
+      "learning_rate": 7.894144843303375e-07,
+      "loss": 0.4375,
+      "step": 6599
+    },
+    {
+      "epoch": 3.2740474121881595,
+      "grad_norm": 0.07253003013199788,
+      "learning_rate": 7.883606471452138e-07,
+      "loss": 0.4248,
+      "step": 6600
+    },
+    {
+      "epoch": 3.274543874891399,
+      "grad_norm": 0.07345711876328244,
+      "learning_rate": 7.873074536416365e-07,
+      "loss": 0.4647,
+      "step": 6601
+    },
+    {
+      "epoch": 3.275040337594638,
+      "grad_norm": 0.0746637762429317,
+      "learning_rate": 7.8625490398057e-07,
+      "loss": 0.4463,
+      "step": 6602
+    },
+    {
+      "epoch": 3.2755368002978775,
+      "grad_norm": 0.07196118986595847,
+      "learning_rate": 7.852029983228787e-07,
+      "loss": 0.4165,
+      "step": 6603
+    },
+    {
+      "epoch": 3.276033263001117,
+      "grad_norm": 0.0758385444525904,
+      "learning_rate": 7.841517368293289e-07,
+      "loss": 0.4912,
+      "step": 6604
+    },
+    {
+      "epoch": 3.2765297257043566,
+      "grad_norm": 0.07158036061870192,
+      "learning_rate": 7.831011196605915e-07,
+      "loss": 0.4542,
+      "step": 6605
+    },
+    {
+      "epoch": 3.277026188407596,
+      "grad_norm": 0.07296484306255716,
+      "learning_rate": 7.820511469772341e-07,
+      "loss": 0.4123,
+      "step": 6606
+    },
+    {
+      "epoch": 3.2775226511108353,
+      "grad_norm": 0.07231519525271822,
+      "learning_rate": 7.810018189397301e-07,
+      "loss": 0.432,
+      "step": 6607
+    },
+    {
+      "epoch": 3.2780191138140746,
+      "grad_norm": 0.07706888568761443,
+      "learning_rate": 7.799531357084517e-07,
+      "loss": 0.4663,
+      "step": 6608
+    },
+    {
+      "epoch": 3.278515576517314,
+      "grad_norm": 0.07134553846073277,
+      "learning_rate": 7.789050974436719e-07,
+      "loss": 0.4302,
+      "step": 6609
+    },
+    {
+      "epoch": 3.2790120392205537,
+      "grad_norm": 0.07451304099114202,
+      "learning_rate": 7.77857704305569e-07,
+      "loss": 0.4416,
+      "step": 6610
+    },
+    {
+      "epoch": 3.279508501923793,
+      "grad_norm": 0.07403930544580954,
+      "learning_rate": 7.768109564542181e-07,
+      "loss": 0.4684,
+      "step": 6611
+    },
+    {
+      "epoch": 3.2800049646270324,
+      "grad_norm": 0.07093526312200696,
+      "learning_rate": 7.757648540495999e-07,
+      "loss": 0.413,
+      "step": 6612
+    },
+    {
+      "epoch": 3.2805014273302717,
+      "grad_norm": 0.07084022920084229,
+      "learning_rate": 7.747193972515932e-07,
+      "loss": 0.417,
+      "step": 6613
+    },
+    {
+      "epoch": 3.280997890033511,
+      "grad_norm": 0.07213204436676163,
+      "learning_rate": 7.736745862199785e-07,
+      "loss": 0.434,
+      "step": 6614
+    },
+    {
+      "epoch": 3.281494352736751,
+      "grad_norm": 0.07241724518413502,
+      "learning_rate": 7.726304211144403e-07,
+      "loss": 0.4143,
+      "step": 6615
+    },
+    {
+      "epoch": 3.28199081543999,
+      "grad_norm": 0.07182507666799981,
+      "learning_rate": 7.715869020945604e-07,
+      "loss": 0.4547,
+      "step": 6616
+    },
+    {
+      "epoch": 3.2824872781432295,
+      "grad_norm": 0.07260704595384614,
+      "learning_rate": 7.705440293198263e-07,
+      "loss": 0.4552,
+      "step": 6617
+    },
+    {
+      "epoch": 3.282983740846469,
+      "grad_norm": 0.0723972933923348,
+      "learning_rate": 7.695018029496232e-07,
+      "loss": 0.4297,
+      "step": 6618
+    },
+    {
+      "epoch": 3.283480203549708,
+      "grad_norm": 0.0724527393243901,
+      "learning_rate": 7.684602231432381e-07,
+      "loss": 0.4312,
+      "step": 6619
+    },
+    {
+      "epoch": 3.283976666252948,
+      "grad_norm": 0.0716458404206346,
+      "learning_rate": 7.674192900598609e-07,
+      "loss": 0.4388,
+      "step": 6620
+    },
+    {
+      "epoch": 3.2844731289561873,
+      "grad_norm": 0.07041881628193564,
+      "learning_rate": 7.663790038585794e-07,
+      "loss": 0.4363,
+      "step": 6621
+    },
+    {
+      "epoch": 3.2849695916594266,
+      "grad_norm": 0.07313003855013313,
+      "learning_rate": 7.653393646983875e-07,
+      "loss": 0.4093,
+      "step": 6622
+    },
+    {
+      "epoch": 3.285466054362666,
+      "grad_norm": 0.07334277170570097,
+      "learning_rate": 7.643003727381754e-07,
+      "loss": 0.4477,
+      "step": 6623
+    },
+    {
+      "epoch": 3.2859625170659053,
+      "grad_norm": 0.07465207208298735,
+      "learning_rate": 7.632620281367376e-07,
+      "loss": 0.4392,
+      "step": 6624
+    },
+    {
+      "epoch": 3.286458979769145,
+      "grad_norm": 0.07036024314880787,
+      "learning_rate": 7.622243310527678e-07,
+      "loss": 0.4083,
+      "step": 6625
+    },
+    {
+      "epoch": 3.2869554424723844,
+      "grad_norm": 0.070643471682643,
+      "learning_rate": 7.611872816448606e-07,
+      "loss": 0.4264,
+      "step": 6626
+    },
+    {
+      "epoch": 3.2874519051756237,
+      "grad_norm": 0.0721032209785513,
+      "learning_rate": 7.601508800715146e-07,
+      "loss": 0.4184,
+      "step": 6627
+    },
+    {
+      "epoch": 3.287948367878863,
+      "grad_norm": 0.07275339367069343,
+      "learning_rate": 7.59115126491124e-07,
+      "loss": 0.4532,
+      "step": 6628
+    },
+    {
+      "epoch": 3.2884448305821024,
+      "grad_norm": 0.07564767357943082,
+      "learning_rate": 7.580800210619904e-07,
+      "loss": 0.4665,
+      "step": 6629
+    },
+    {
+      "epoch": 3.288941293285342,
+      "grad_norm": 0.07434048177123104,
+      "learning_rate": 7.570455639423119e-07,
+      "loss": 0.4294,
+      "step": 6630
+    },
+    {
+      "epoch": 3.2894377559885815,
+      "grad_norm": 0.07498662516913521,
+      "learning_rate": 7.560117552901863e-07,
+      "loss": 0.4346,
+      "step": 6631
+    },
+    {
+      "epoch": 3.289934218691821,
+      "grad_norm": 0.07379359019856084,
+      "learning_rate": 7.549785952636185e-07,
+      "loss": 0.4166,
+      "step": 6632
+    },
+    {
+      "epoch": 3.29043068139506,
+      "grad_norm": 0.07508251711812729,
+      "learning_rate": 7.539460840205076e-07,
+      "loss": 0.4472,
+      "step": 6633
+    },
+    {
+      "epoch": 3.2909271440982995,
+      "grad_norm": 0.07424590033382815,
+      "learning_rate": 7.529142217186596e-07,
+      "loss": 0.4505,
+      "step": 6634
+    },
+    {
+      "epoch": 3.2914236068015392,
+      "grad_norm": 0.07437842299007032,
+      "learning_rate": 7.518830085157735e-07,
+      "loss": 0.4588,
+      "step": 6635
+    },
+    {
+      "epoch": 3.2919200695047786,
+      "grad_norm": 0.07079216645025961,
+      "learning_rate": 7.508524445694577e-07,
+      "loss": 0.4155,
+      "step": 6636
+    },
+    {
+      "epoch": 3.292416532208018,
+      "grad_norm": 0.07371760019391377,
+      "learning_rate": 7.498225300372152e-07,
+      "loss": 0.4622,
+      "step": 6637
+    },
+    {
+      "epoch": 3.292912994911257,
+      "grad_norm": 0.07168229019197096,
+      "learning_rate": 7.487932650764523e-07,
+      "loss": 0.4252,
+      "step": 6638
+    },
+    {
+      "epoch": 3.2934094576144965,
+      "grad_norm": 0.07461181078661946,
+      "learning_rate": 7.477646498444762e-07,
+      "loss": 0.4851,
+      "step": 6639
+    },
+    {
+      "epoch": 3.2939059203177363,
+      "grad_norm": 0.07593723177468553,
+      "learning_rate": 7.467366844984946e-07,
+      "loss": 0.4457,
+      "step": 6640
+    },
+    {
+      "epoch": 3.2944023830209757,
+      "grad_norm": 0.07516771486988753,
+      "learning_rate": 7.457093691956136e-07,
+      "loss": 0.4604,
+      "step": 6641
+    },
+    {
+      "epoch": 3.294898845724215,
+      "grad_norm": 0.0717665501625767,
+      "learning_rate": 7.446827040928439e-07,
+      "loss": 0.4588,
+      "step": 6642
+    },
+    {
+      "epoch": 3.2953953084274543,
+      "grad_norm": 0.07372662891217222,
+      "learning_rate": 7.436566893470937e-07,
+      "loss": 0.4408,
+      "step": 6643
+    },
+    {
+      "epoch": 3.2958917711306936,
+      "grad_norm": 0.0732164428085213,
+      "learning_rate": 7.426313251151734e-07,
+      "loss": 0.4303,
+      "step": 6644
+    },
+    {
+      "epoch": 3.296388233833933,
+      "grad_norm": 0.07271523459017934,
+      "learning_rate": 7.41606611553794e-07,
+      "loss": 0.4416,
+      "step": 6645
+    },
+    {
+      "epoch": 3.2968846965371728,
+      "grad_norm": 0.07027489236127155,
+      "learning_rate": 7.405825488195645e-07,
+      "loss": 0.4408,
+      "step": 6646
+    },
+    {
+      "epoch": 3.297381159240412,
+      "grad_norm": 0.0703499925933336,
+      "learning_rate": 7.395591370689992e-07,
+      "loss": 0.4136,
+      "step": 6647
+    },
+    {
+      "epoch": 3.2978776219436514,
+      "grad_norm": 0.0732883393369235,
+      "learning_rate": 7.385363764585074e-07,
+      "loss": 0.4365,
+      "step": 6648
+    },
+    {
+      "epoch": 3.2983740846468907,
+      "grad_norm": 0.07330766129833696,
+      "learning_rate": 7.375142671444046e-07,
+      "loss": 0.4408,
+      "step": 6649
+    },
+    {
+      "epoch": 3.2988705473501305,
+      "grad_norm": 0.07290035642087749,
+      "learning_rate": 7.364928092829021e-07,
+      "loss": 0.4456,
+      "step": 6650
+    },
+    {
+      "epoch": 3.29936701005337,
+      "grad_norm": 0.07259967466068204,
+      "learning_rate": 7.354720030301138e-07,
+      "loss": 0.4303,
+      "step": 6651
+    },
+    {
+      "epoch": 3.299863472756609,
+      "grad_norm": 0.07300187700640386,
+      "learning_rate": 7.344518485420526e-07,
+      "loss": 0.461,
+      "step": 6652
+    },
+    {
+      "epoch": 3.3003599354598485,
+      "grad_norm": 0.07125456429321311,
+      "learning_rate": 7.334323459746329e-07,
+      "loss": 0.4008,
+      "step": 6653
+    },
+    {
+      "epoch": 3.300856398163088,
+      "grad_norm": 0.07285745627014652,
+      "learning_rate": 7.32413495483671e-07,
+      "loss": 0.4441,
+      "step": 6654
+    },
+    {
+      "epoch": 3.301352860866327,
+      "grad_norm": 0.07707300575756339,
+      "learning_rate": 7.313952972248795e-07,
+      "loss": 0.4527,
+      "step": 6655
+    },
+    {
+      "epoch": 3.301849323569567,
+      "grad_norm": 0.0759425918137467,
+      "learning_rate": 7.303777513538762e-07,
+      "loss": 0.4993,
+      "step": 6656
+    },
+    {
+      "epoch": 3.3023457862728063,
+      "grad_norm": 0.07405805280796507,
+      "learning_rate": 7.293608580261757e-07,
+      "loss": 0.4367,
+      "step": 6657
+    },
+    {
+      "epoch": 3.3028422489760456,
+      "grad_norm": 0.07201013691773817,
+      "learning_rate": 7.28344617397192e-07,
+      "loss": 0.4382,
+      "step": 6658
+    },
+    {
+      "epoch": 3.303338711679285,
+      "grad_norm": 0.07353768250813451,
+      "learning_rate": 7.27329029622244e-07,
+      "loss": 0.4627,
+      "step": 6659
+    },
+    {
+      "epoch": 3.3038351743825247,
+      "grad_norm": 0.07316616724213289,
+      "learning_rate": 7.263140948565456e-07,
+      "loss": 0.4491,
+      "step": 6660
+    },
+    {
+      "epoch": 3.304331637085764,
+      "grad_norm": 0.07104927273550239,
+      "learning_rate": 7.252998132552158e-07,
+      "loss": 0.4536,
+      "step": 6661
+    },
+    {
+      "epoch": 3.3048280997890034,
+      "grad_norm": 0.07102527141998924,
+      "learning_rate": 7.242861849732696e-07,
+      "loss": 0.4445,
+      "step": 6662
+    },
+    {
+      "epoch": 3.3053245624922427,
+      "grad_norm": 0.07093892808583452,
+      "learning_rate": 7.232732101656231e-07,
+      "loss": 0.4407,
+      "step": 6663
+    },
+    {
+      "epoch": 3.305821025195482,
+      "grad_norm": 0.07291337911424048,
+      "learning_rate": 7.222608889870958e-07,
+      "loss": 0.4391,
+      "step": 6664
+    },
+    {
+      "epoch": 3.3063174878987214,
+      "grad_norm": 0.07466482337371662,
+      "learning_rate": 7.212492215924016e-07,
+      "loss": 0.4799,
+      "step": 6665
+    },
+    {
+      "epoch": 3.306813950601961,
+      "grad_norm": 0.07150248934206467,
+      "learning_rate": 7.202382081361619e-07,
+      "loss": 0.4479,
+      "step": 6666
+    },
+    {
+      "epoch": 3.3073104133052005,
+      "grad_norm": 0.0732130705782928,
+      "learning_rate": 7.192278487728893e-07,
+      "loss": 0.4144,
+      "step": 6667
+    },
+    {
+      "epoch": 3.30780687600844,
+      "grad_norm": 0.07400423447607553,
+      "learning_rate": 7.182181436570041e-07,
+      "loss": 0.4268,
+      "step": 6668
+    },
+    {
+      "epoch": 3.308303338711679,
+      "grad_norm": 0.07357958751155688,
+      "learning_rate": 7.172090929428221e-07,
+      "loss": 0.4483,
+      "step": 6669
+    },
+    {
+      "epoch": 3.308799801414919,
+      "grad_norm": 0.07229389098614548,
+      "learning_rate": 7.162006967845602e-07,
+      "loss": 0.4661,
+      "step": 6670
+    },
+    {
+      "epoch": 3.3092962641181582,
+      "grad_norm": 0.07090210708677912,
+      "learning_rate": 7.151929553363368e-07,
+      "loss": 0.4369,
+      "step": 6671
+    },
+    {
+      "epoch": 3.3097927268213976,
+      "grad_norm": 0.07275196546922465,
+      "learning_rate": 7.14185868752168e-07,
+      "loss": 0.4386,
+      "step": 6672
+    },
+    {
+      "epoch": 3.310289189524637,
+      "grad_norm": 0.07239941321768839,
+      "learning_rate": 7.131794371859724e-07,
+      "loss": 0.4629,
+      "step": 6673
+    },
+    {
+      "epoch": 3.3107856522278762,
+      "grad_norm": 0.07150491620273815,
+      "learning_rate": 7.121736607915658e-07,
+      "loss": 0.4281,
+      "step": 6674
+    },
+    {
+      "epoch": 3.3112821149311156,
+      "grad_norm": 0.07010195635506701,
+      "learning_rate": 7.111685397226642e-07,
+      "loss": 0.4399,
+      "step": 6675
+    },
+    {
+      "epoch": 3.3117785776343553,
+      "grad_norm": 0.07081236027352858,
+      "learning_rate": 7.101640741328858e-07,
+      "loss": 0.4311,
+      "step": 6676
+    },
+    {
+      "epoch": 3.3122750403375947,
+      "grad_norm": 0.0705597997075449,
+      "learning_rate": 7.091602641757467e-07,
+      "loss": 0.397,
+      "step": 6677
+    },
+    {
+      "epoch": 3.312771503040834,
+      "grad_norm": 0.07351201791505935,
+      "learning_rate": 7.081571100046613e-07,
+      "loss": 0.4216,
+      "step": 6678
+    },
+    {
+      "epoch": 3.3132679657440733,
+      "grad_norm": 0.07162297978284737,
+      "learning_rate": 7.071546117729489e-07,
+      "loss": 0.4418,
+      "step": 6679
+    },
+    {
+      "epoch": 3.313764428447313,
+      "grad_norm": 0.07072229240363065,
+      "learning_rate": 7.061527696338221e-07,
+      "loss": 0.4229,
+      "step": 6680
+    },
+    {
+      "epoch": 3.3142608911505524,
+      "grad_norm": 0.07178441225873806,
+      "learning_rate": 7.051515837403989e-07,
+      "loss": 0.4321,
+      "step": 6681
+    },
+    {
+      "epoch": 3.3147573538537918,
+      "grad_norm": 0.07357930531007792,
+      "learning_rate": 7.041510542456936e-07,
+      "loss": 0.4377,
+      "step": 6682
+    },
+    {
+      "epoch": 3.315253816557031,
+      "grad_norm": 0.07120147170563958,
+      "learning_rate": 7.03151181302621e-07,
+      "loss": 0.4562,
+      "step": 6683
+    },
+    {
+      "epoch": 3.3157502792602704,
+      "grad_norm": 0.07540081844652202,
+      "learning_rate": 7.021519650639952e-07,
+      "loss": 0.4627,
+      "step": 6684
+    },
+    {
+      "epoch": 3.3162467419635098,
+      "grad_norm": 0.07113171542314269,
+      "learning_rate": 7.011534056825303e-07,
+      "loss": 0.408,
+      "step": 6685
+    },
+    {
+      "epoch": 3.3167432046667495,
+      "grad_norm": 0.07734922712339089,
+      "learning_rate": 7.001555033108414e-07,
+      "loss": 0.4541,
+      "step": 6686
+    },
+    {
+      "epoch": 3.317239667369989,
+      "grad_norm": 0.0719354552889049,
+      "learning_rate": 6.991582581014394e-07,
+      "loss": 0.4084,
+      "step": 6687
+    },
+    {
+      "epoch": 3.317736130073228,
+      "grad_norm": 0.07473471116857953,
+      "learning_rate": 6.981616702067406e-07,
+      "loss": 0.4333,
+      "step": 6688
+    },
+    {
+      "epoch": 3.3182325927764675,
+      "grad_norm": 0.07355981521959422,
+      "learning_rate": 6.97165739779056e-07,
+      "loss": 0.4495,
+      "step": 6689
+    },
+    {
+      "epoch": 3.3187290554797073,
+      "grad_norm": 0.0738486895010106,
+      "learning_rate": 6.96170466970596e-07,
+      "loss": 0.4207,
+      "step": 6690
+    },
+    {
+      "epoch": 3.3192255181829466,
+      "grad_norm": 0.07518710674642788,
+      "learning_rate": 6.951758519334745e-07,
+      "loss": 0.4186,
+      "step": 6691
+    },
+    {
+      "epoch": 3.319721980886186,
+      "grad_norm": 0.07191266899444419,
+      "learning_rate": 6.941818948197005e-07,
+      "loss": 0.4359,
+      "step": 6692
+    },
+    {
+      "epoch": 3.3202184435894253,
+      "grad_norm": 0.07226725932523494,
+      "learning_rate": 6.931885957811862e-07,
+      "loss": 0.4553,
+      "step": 6693
+    },
+    {
+      "epoch": 3.3207149062926646,
+      "grad_norm": 0.07825433273516179,
+      "learning_rate": 6.921959549697404e-07,
+      "loss": 0.4912,
+      "step": 6694
+    },
+    {
+      "epoch": 3.321211368995904,
+      "grad_norm": 0.07284332065755789,
+      "learning_rate": 6.912039725370717e-07,
+      "loss": 0.4222,
+      "step": 6695
+    },
+    {
+      "epoch": 3.3217078316991437,
+      "grad_norm": 0.07377691811710664,
+      "learning_rate": 6.902126486347904e-07,
+      "loss": 0.4481,
+      "step": 6696
+    },
+    {
+      "epoch": 3.322204294402383,
+      "grad_norm": 0.0753173718581925,
+      "learning_rate": 6.892219834144032e-07,
+      "loss": 0.4612,
+      "step": 6697
+    },
+    {
+      "epoch": 3.3227007571056224,
+      "grad_norm": 0.06991257631606883,
+      "learning_rate": 6.882319770273193e-07,
+      "loss": 0.422,
+      "step": 6698
+    },
+    {
+      "epoch": 3.3231972198088617,
+      "grad_norm": 0.07307525817105578,
+      "learning_rate": 6.872426296248413e-07,
+      "loss": 0.4445,
+      "step": 6699
+    },
+    {
+      "epoch": 3.3236936825121015,
+      "grad_norm": 0.07331646472148728,
+      "learning_rate": 6.862539413581792e-07,
+      "loss": 0.4462,
+      "step": 6700
+    },
+    {
+      "epoch": 3.324190145215341,
+      "grad_norm": 0.07309253081404553,
+      "learning_rate": 6.85265912378436e-07,
+      "loss": 0.466,
+      "step": 6701
+    },
+    {
+      "epoch": 3.32468660791858,
+      "grad_norm": 0.07396548326621871,
+      "learning_rate": 6.842785428366161e-07,
+      "loss": 0.4405,
+      "step": 6702
+    },
+    {
+      "epoch": 3.3251830706218195,
+      "grad_norm": 0.0730798787215561,
+      "learning_rate": 6.832918328836247e-07,
+      "loss": 0.453,
+      "step": 6703
+    },
+    {
+      "epoch": 3.325679533325059,
+      "grad_norm": 0.0728092690289243,
+      "learning_rate": 6.823057826702617e-07,
+      "loss": 0.4485,
+      "step": 6704
+    },
+    {
+      "epoch": 3.326175996028298,
+      "grad_norm": 0.07346187407672358,
+      "learning_rate": 6.813203923472328e-07,
+      "loss": 0.44,
+      "step": 6705
+    },
+    {
+      "epoch": 3.326672458731538,
+      "grad_norm": 0.07255358778733134,
+      "learning_rate": 6.803356620651364e-07,
+      "loss": 0.4154,
+      "step": 6706
+    },
+    {
+      "epoch": 3.3271689214347773,
+      "grad_norm": 0.0705595758422318,
+      "learning_rate": 6.793515919744725e-07,
+      "loss": 0.435,
+      "step": 6707
+    },
+    {
+      "epoch": 3.3276653841380166,
+      "grad_norm": 0.07369430638371367,
+      "learning_rate": 6.783681822256433e-07,
+      "loss": 0.4277,
+      "step": 6708
+    },
+    {
+      "epoch": 3.328161846841256,
+      "grad_norm": 0.07246376987908641,
+      "learning_rate": 6.773854329689433e-07,
+      "loss": 0.4301,
+      "step": 6709
+    },
+    {
+      "epoch": 3.3286583095444957,
+      "grad_norm": 0.07242937121960161,
+      "learning_rate": 6.764033443545737e-07,
+      "loss": 0.4405,
+      "step": 6710
+    },
+    {
+      "epoch": 3.329154772247735,
+      "grad_norm": 0.07134622113216685,
+      "learning_rate": 6.754219165326293e-07,
+      "loss": 0.4774,
+      "step": 6711
+    },
+    {
+      "epoch": 3.3296512349509744,
+      "grad_norm": 0.07254229774617119,
+      "learning_rate": 6.744411496531045e-07,
+      "loss": 0.4304,
+      "step": 6712
+    },
+    {
+      "epoch": 3.3301476976542137,
+      "grad_norm": 0.07269788593095719,
+      "learning_rate": 6.734610438658957e-07,
+      "loss": 0.4559,
+      "step": 6713
+    },
+    {
+      "epoch": 3.330644160357453,
+      "grad_norm": 0.07285441368630612,
+      "learning_rate": 6.724815993207956e-07,
+      "loss": 0.4203,
+      "step": 6714
+    },
+    {
+      "epoch": 3.3311406230606924,
+      "grad_norm": 0.07401827820352684,
+      "learning_rate": 6.71502816167497e-07,
+      "loss": 0.449,
+      "step": 6715
+    },
+    {
+      "epoch": 3.331637085763932,
+      "grad_norm": 0.06939161478859503,
+      "learning_rate": 6.705246945555905e-07,
+      "loss": 0.4281,
+      "step": 6716
+    },
+    {
+      "epoch": 3.3321335484671715,
+      "grad_norm": 0.07253069731870333,
+      "learning_rate": 6.695472346345655e-07,
+      "loss": 0.4227,
+      "step": 6717
+    },
+    {
+      "epoch": 3.332630011170411,
+      "grad_norm": 0.07181321712064559,
+      "learning_rate": 6.685704365538132e-07,
+      "loss": 0.4217,
+      "step": 6718
+    },
+    {
+      "epoch": 3.33312647387365,
+      "grad_norm": 0.07691835372690169,
+      "learning_rate": 6.67594300462619e-07,
+      "loss": 0.4493,
+      "step": 6719
+    },
+    {
+      "epoch": 3.33362293657689,
+      "grad_norm": 0.07082709018052989,
+      "learning_rate": 6.666188265101725e-07,
+      "loss": 0.4322,
+      "step": 6720
+    },
+    {
+      "epoch": 3.3341193992801292,
+      "grad_norm": 0.07958434335917278,
+      "learning_rate": 6.656440148455584e-07,
+      "loss": 0.5023,
+      "step": 6721
+    },
+    {
+      "epoch": 3.3346158619833686,
+      "grad_norm": 0.07347018470497847,
+      "learning_rate": 6.646698656177591e-07,
+      "loss": 0.4587,
+      "step": 6722
+    },
+    {
+      "epoch": 3.335112324686608,
+      "grad_norm": 0.06977330862340675,
+      "learning_rate": 6.636963789756601e-07,
+      "loss": 0.4125,
+      "step": 6723
+    },
+    {
+      "epoch": 3.3356087873898472,
+      "grad_norm": 0.06947659199276247,
+      "learning_rate": 6.627235550680411e-07,
+      "loss": 0.4247,
+      "step": 6724
+    },
+    {
+      "epoch": 3.3361052500930866,
+      "grad_norm": 0.07559550092142363,
+      "learning_rate": 6.617513940435849e-07,
+      "loss": 0.4608,
+      "step": 6725
+    },
+    {
+      "epoch": 3.3366017127963263,
+      "grad_norm": 0.07477526832332422,
+      "learning_rate": 6.607798960508693e-07,
+      "loss": 0.442,
+      "step": 6726
+    },
+    {
+      "epoch": 3.3370981754995657,
+      "grad_norm": 0.07208237172360653,
+      "learning_rate": 6.598090612383723e-07,
+      "loss": 0.4387,
+      "step": 6727
+    },
+    {
+      "epoch": 3.337594638202805,
+      "grad_norm": 0.07072698804456641,
+      "learning_rate": 6.588388897544707e-07,
+      "loss": 0.4371,
+      "step": 6728
+    },
+    {
+      "epoch": 3.3380911009060443,
+      "grad_norm": 0.07388009610832842,
+      "learning_rate": 6.578693817474391e-07,
+      "loss": 0.4517,
+      "step": 6729
+    },
+    {
+      "epoch": 3.3385875636092837,
+      "grad_norm": 0.07622744923989354,
+      "learning_rate": 6.569005373654524e-07,
+      "loss": 0.4459,
+      "step": 6730
+    },
+    {
+      "epoch": 3.3390840263125234,
+      "grad_norm": 0.07231725065360844,
+      "learning_rate": 6.559323567565828e-07,
+      "loss": 0.4494,
+      "step": 6731
+    },
+    {
+      "epoch": 3.3395804890157628,
+      "grad_norm": 0.07271324266775285,
+      "learning_rate": 6.549648400688003e-07,
+      "loss": 0.4261,
+      "step": 6732
+    },
+    {
+      "epoch": 3.340076951719002,
+      "grad_norm": 0.07026922793424965,
+      "learning_rate": 6.539979874499747e-07,
+      "loss": 0.4192,
+      "step": 6733
+    },
+    {
+      "epoch": 3.3405734144222414,
+      "grad_norm": 0.07098739030810841,
+      "learning_rate": 6.530317990478729e-07,
+      "loss": 0.44,
+      "step": 6734
+    },
+    {
+      "epoch": 3.3410698771254808,
+      "grad_norm": 0.07210641928798285,
+      "learning_rate": 6.52066275010163e-07,
+      "loss": 0.4289,
+      "step": 6735
+    },
+    {
+      "epoch": 3.3415663398287205,
+      "grad_norm": 0.07414660760570124,
+      "learning_rate": 6.511014154844081e-07,
+      "loss": 0.4556,
+      "step": 6736
+    },
+    {
+      "epoch": 3.34206280253196,
+      "grad_norm": 0.07313210011289042,
+      "learning_rate": 6.50137220618074e-07,
+      "loss": 0.4128,
+      "step": 6737
+    },
+    {
+      "epoch": 3.342559265235199,
+      "grad_norm": 0.07093195725850568,
+      "learning_rate": 6.491736905585211e-07,
+      "loss": 0.4447,
+      "step": 6738
+    },
+    {
+      "epoch": 3.3430557279384385,
+      "grad_norm": 0.07546538508603158,
+      "learning_rate": 6.482108254530078e-07,
+      "loss": 0.4196,
+      "step": 6739
+    },
+    {
+      "epoch": 3.343552190641678,
+      "grad_norm": 0.07474759213803385,
+      "learning_rate": 6.472486254486954e-07,
+      "loss": 0.4545,
+      "step": 6740
+    },
+    {
+      "epoch": 3.3440486533449176,
+      "grad_norm": 0.07306879951905888,
+      "learning_rate": 6.462870906926389e-07,
+      "loss": 0.4343,
+      "step": 6741
+    },
+    {
+      "epoch": 3.344545116048157,
+      "grad_norm": 0.0713618723632236,
+      "learning_rate": 6.453262213317946e-07,
+      "loss": 0.4405,
+      "step": 6742
+    },
+    {
+      "epoch": 3.3450415787513963,
+      "grad_norm": 0.07294681200393098,
+      "learning_rate": 6.443660175130157e-07,
+      "loss": 0.4634,
+      "step": 6743
+    },
+    {
+      "epoch": 3.3455380414546356,
+      "grad_norm": 0.07476448744900865,
+      "learning_rate": 6.43406479383053e-07,
+      "loss": 0.4443,
+      "step": 6744
+    },
+    {
+      "epoch": 3.346034504157875,
+      "grad_norm": 0.07228777016156072,
+      "learning_rate": 6.424476070885582e-07,
+      "loss": 0.4204,
+      "step": 6745
+    },
+    {
+      "epoch": 3.3465309668611147,
+      "grad_norm": 0.07164394178278337,
+      "learning_rate": 6.414894007760769e-07,
+      "loss": 0.4323,
+      "step": 6746
+    },
+    {
+      "epoch": 3.347027429564354,
+      "grad_norm": 0.07392067999596949,
+      "learning_rate": 6.405318605920602e-07,
+      "loss": 0.442,
+      "step": 6747
+    },
+    {
+      "epoch": 3.3475238922675934,
+      "grad_norm": 0.07339914027298029,
+      "learning_rate": 6.395749866828477e-07,
+      "loss": 0.4214,
+      "step": 6748
+    },
+    {
+      "epoch": 3.3480203549708327,
+      "grad_norm": 0.0715534750654056,
+      "learning_rate": 6.386187791946852e-07,
+      "loss": 0.4309,
+      "step": 6749
+    },
+    {
+      "epoch": 3.348516817674072,
+      "grad_norm": 0.08010573872207297,
+      "learning_rate": 6.376632382737125e-07,
+      "loss": 0.4804,
+      "step": 6750
+    },
+    {
+      "epoch": 3.349013280377312,
+      "grad_norm": 0.07490044375536065,
+      "learning_rate": 6.367083640659682e-07,
+      "loss": 0.4469,
+      "step": 6751
+    },
+    {
+      "epoch": 3.349509743080551,
+      "grad_norm": 0.07309406245651992,
+      "learning_rate": 6.35754156717392e-07,
+      "loss": 0.4459,
+      "step": 6752
+    },
+    {
+      "epoch": 3.3500062057837905,
+      "grad_norm": 0.07124688102133844,
+      "learning_rate": 6.348006163738174e-07,
+      "loss": 0.4321,
+      "step": 6753
+    },
+    {
+      "epoch": 3.35050266848703,
+      "grad_norm": 0.07098792326562689,
+      "learning_rate": 6.338477431809764e-07,
+      "loss": 0.4355,
+      "step": 6754
+    },
+    {
+      "epoch": 3.350999131190269,
+      "grad_norm": 0.0716445990956709,
+      "learning_rate": 6.328955372845036e-07,
+      "loss": 0.4388,
+      "step": 6755
+    },
+    {
+      "epoch": 3.351495593893509,
+      "grad_norm": 0.0722516039969341,
+      "learning_rate": 6.319439988299253e-07,
+      "loss": 0.418,
+      "step": 6756
+    },
+    {
+      "epoch": 3.3519920565967483,
+      "grad_norm": 0.07210491335036923,
+      "learning_rate": 6.309931279626713e-07,
+      "loss": 0.4405,
+      "step": 6757
+    },
+    {
+      "epoch": 3.3524885192999876,
+      "grad_norm": 0.07130023795309107,
+      "learning_rate": 6.300429248280659e-07,
+      "loss": 0.4413,
+      "step": 6758
+    },
+    {
+      "epoch": 3.352984982003227,
+      "grad_norm": 0.07101581035496832,
+      "learning_rate": 6.29093389571332e-07,
+      "loss": 0.4202,
+      "step": 6759
+    },
+    {
+      "epoch": 3.3534814447064663,
+      "grad_norm": 0.07383445946068722,
+      "learning_rate": 6.281445223375921e-07,
+      "loss": 0.4532,
+      "step": 6760
+    },
+    {
+      "epoch": 3.353977907409706,
+      "grad_norm": 0.07163606670276983,
+      "learning_rate": 6.271963232718631e-07,
+      "loss": 0.439,
+      "step": 6761
+    },
+    {
+      "epoch": 3.3544743701129454,
+      "grad_norm": 0.07111746615527417,
+      "learning_rate": 6.262487925190653e-07,
+      "loss": 0.4341,
+      "step": 6762
+    },
+    {
+      "epoch": 3.3549708328161847,
+      "grad_norm": 0.07433150143658057,
+      "learning_rate": 6.253019302240115e-07,
+      "loss": 0.4579,
+      "step": 6763
+    },
+    {
+      "epoch": 3.355467295519424,
+      "grad_norm": 0.07299509800751336,
+      "learning_rate": 6.243557365314146e-07,
+      "loss": 0.4544,
+      "step": 6764
+    },
+    {
+      "epoch": 3.3559637582226634,
+      "grad_norm": 0.07241430620074621,
+      "learning_rate": 6.234102115858853e-07,
+      "loss": 0.4524,
+      "step": 6765
+    },
+    {
+      "epoch": 3.356460220925903,
+      "grad_norm": 0.0735593480259739,
+      "learning_rate": 6.224653555319309e-07,
+      "loss": 0.4345,
+      "step": 6766
+    },
+    {
+      "epoch": 3.3569566836291425,
+      "grad_norm": 0.07549715209388864,
+      "learning_rate": 6.215211685139594e-07,
+      "loss": 0.4439,
+      "step": 6767
+    },
+    {
+      "epoch": 3.357453146332382,
+      "grad_norm": 0.07636832298287011,
+      "learning_rate": 6.205776506762729e-07,
+      "loss": 0.4379,
+      "step": 6768
+    },
+    {
+      "epoch": 3.357949609035621,
+      "grad_norm": 0.074032413581356,
+      "learning_rate": 6.196348021630749e-07,
+      "loss": 0.439,
+      "step": 6769
+    },
+    {
+      "epoch": 3.3584460717388605,
+      "grad_norm": 0.07225979545657661,
+      "learning_rate": 6.186926231184631e-07,
+      "loss": 0.4016,
+      "step": 6770
+    },
+    {
+      "epoch": 3.3589425344421002,
+      "grad_norm": 0.07537400284700459,
+      "learning_rate": 6.17751113686434e-07,
+      "loss": 0.4623,
+      "step": 6771
+    },
+    {
+      "epoch": 3.3594389971453396,
+      "grad_norm": 0.0738458214689033,
+      "learning_rate": 6.168102740108844e-07,
+      "loss": 0.45,
+      "step": 6772
+    },
+    {
+      "epoch": 3.359935459848579,
+      "grad_norm": 0.0728896823511883,
+      "learning_rate": 6.158701042356046e-07,
+      "loss": 0.4309,
+      "step": 6773
+    },
+    {
+      "epoch": 3.360431922551818,
+      "grad_norm": 0.07260783854191073,
+      "learning_rate": 6.149306045042858e-07,
+      "loss": 0.4526,
+      "step": 6774
+    },
+    {
+      "epoch": 3.3609283852550575,
+      "grad_norm": 0.07493342650916683,
+      "learning_rate": 6.139917749605151e-07,
+      "loss": 0.4618,
+      "step": 6775
+    },
+    {
+      "epoch": 3.3614248479582973,
+      "grad_norm": 0.07315163191067617,
+      "learning_rate": 6.130536157477757e-07,
+      "loss": 0.4478,
+      "step": 6776
+    },
+    {
+      "epoch": 3.3619213106615367,
+      "grad_norm": 0.07513373087127617,
+      "learning_rate": 6.121161270094533e-07,
+      "loss": 0.4531,
+      "step": 6777
+    },
+    {
+      "epoch": 3.362417773364776,
+      "grad_norm": 0.07382894483836297,
+      "learning_rate": 6.111793088888257e-07,
+      "loss": 0.4439,
+      "step": 6778
+    },
+    {
+      "epoch": 3.3629142360680153,
+      "grad_norm": 0.07150084158217984,
+      "learning_rate": 6.102431615290727e-07,
+      "loss": 0.4357,
+      "step": 6779
+    },
+    {
+      "epoch": 3.3634106987712546,
+      "grad_norm": 0.07094916622144724,
+      "learning_rate": 6.093076850732665e-07,
+      "loss": 0.4007,
+      "step": 6780
+    },
+    {
+      "epoch": 3.3639071614744944,
+      "grad_norm": 0.07021291401826622,
+      "learning_rate": 6.083728796643823e-07,
+      "loss": 0.4385,
+      "step": 6781
+    },
+    {
+      "epoch": 3.3644036241777338,
+      "grad_norm": 0.07429863866638267,
+      "learning_rate": 6.074387454452891e-07,
+      "loss": 0.4475,
+      "step": 6782
+    },
+    {
+      "epoch": 3.364900086880973,
+      "grad_norm": 0.07453176224524409,
+      "learning_rate": 6.06505282558753e-07,
+      "loss": 0.4496,
+      "step": 6783
+    },
+    {
+      "epoch": 3.3653965495842124,
+      "grad_norm": 0.07369756215281187,
+      "learning_rate": 6.055724911474415e-07,
+      "loss": 0.41,
+      "step": 6784
+    },
+    {
+      "epoch": 3.3658930122874517,
+      "grad_norm": 0.0729390433187085,
+      "learning_rate": 6.04640371353914e-07,
+      "loss": 0.4471,
+      "step": 6785
+    },
+    {
+      "epoch": 3.366389474990691,
+      "grad_norm": 0.07105795342109092,
+      "learning_rate": 6.037089233206328e-07,
+      "loss": 0.4204,
+      "step": 6786
+    },
+    {
+      "epoch": 3.366885937693931,
+      "grad_norm": 0.07336405749120571,
+      "learning_rate": 6.027781471899535e-07,
+      "loss": 0.4408,
+      "step": 6787
+    },
+    {
+      "epoch": 3.36738240039717,
+      "grad_norm": 0.0710225486742565,
+      "learning_rate": 6.01848043104129e-07,
+      "loss": 0.4284,
+      "step": 6788
+    },
+    {
+      "epoch": 3.3678788631004095,
+      "grad_norm": 0.07099142741394315,
+      "learning_rate": 6.009186112053134e-07,
+      "loss": 0.4189,
+      "step": 6789
+    },
+    {
+      "epoch": 3.368375325803649,
+      "grad_norm": 0.07195270358324399,
+      "learning_rate": 5.99989851635554e-07,
+      "loss": 0.447,
+      "step": 6790
+    },
+    {
+      "epoch": 3.3688717885068886,
+      "grad_norm": 0.0718929631105409,
+      "learning_rate": 5.990617645367963e-07,
+      "loss": 0.4543,
+      "step": 6791
+    },
+    {
+      "epoch": 3.369368251210128,
+      "grad_norm": 0.0717920961659943,
+      "learning_rate": 5.981343500508846e-07,
+      "loss": 0.4456,
+      "step": 6792
+    },
+    {
+      "epoch": 3.3698647139133673,
+      "grad_norm": 0.07208811889172369,
+      "learning_rate": 5.972076083195583e-07,
+      "loss": 0.4165,
+      "step": 6793
+    },
+    {
+      "epoch": 3.3703611766166066,
+      "grad_norm": 0.07661792816181558,
+      "learning_rate": 5.962815394844567e-07,
+      "loss": 0.465,
+      "step": 6794
+    },
+    {
+      "epoch": 3.370857639319846,
+      "grad_norm": 0.0732824276079872,
+      "learning_rate": 5.953561436871135e-07,
+      "loss": 0.4335,
+      "step": 6795
+    },
+    {
+      "epoch": 3.3713541020230853,
+      "grad_norm": 0.07402146099841271,
+      "learning_rate": 5.944314210689611e-07,
+      "loss": 0.4125,
+      "step": 6796
+    },
+    {
+      "epoch": 3.371850564726325,
+      "grad_norm": 0.07417228224981427,
+      "learning_rate": 5.935073717713274e-07,
+      "loss": 0.4319,
+      "step": 6797
+    },
+    {
+      "epoch": 3.3723470274295644,
+      "grad_norm": 0.07261266000053443,
+      "learning_rate": 5.925839959354384e-07,
+      "loss": 0.437,
+      "step": 6798
+    },
+    {
+      "epoch": 3.3728434901328037,
+      "grad_norm": 0.07326393118590568,
+      "learning_rate": 5.916612937024191e-07,
+      "loss": 0.4382,
+      "step": 6799
+    },
+    {
+      "epoch": 3.373339952836043,
+      "grad_norm": 0.07171944920636655,
+      "learning_rate": 5.907392652132876e-07,
+      "loss": 0.4327,
+      "step": 6800
+    },
+    {
+      "epoch": 3.373836415539283,
+      "grad_norm": 0.07266217367510201,
+      "learning_rate": 5.898179106089635e-07,
+      "loss": 0.4184,
+      "step": 6801
+    },
+    {
+      "epoch": 3.374332878242522,
+      "grad_norm": 0.07068059522225369,
+      "learning_rate": 5.888972300302598e-07,
+      "loss": 0.4373,
+      "step": 6802
+    },
+    {
+      "epoch": 3.3748293409457615,
+      "grad_norm": 0.07446209367229231,
+      "learning_rate": 5.879772236178871e-07,
+      "loss": 0.4597,
+      "step": 6803
+    },
+    {
+      "epoch": 3.375325803649001,
+      "grad_norm": 0.07286519517357345,
+      "learning_rate": 5.870578915124547e-07,
+      "loss": 0.4295,
+      "step": 6804
+    },
+    {
+      "epoch": 3.37582226635224,
+      "grad_norm": 0.07239552432886055,
+      "learning_rate": 5.861392338544669e-07,
+      "loss": 0.4179,
+      "step": 6805
+    },
+    {
+      "epoch": 3.3763187290554795,
+      "grad_norm": 0.07165189079916912,
+      "learning_rate": 5.852212507843274e-07,
+      "loss": 0.4039,
+      "step": 6806
+    },
+    {
+      "epoch": 3.3768151917587192,
+      "grad_norm": 0.07292560249677038,
+      "learning_rate": 5.843039424423341e-07,
+      "loss": 0.4455,
+      "step": 6807
+    },
+    {
+      "epoch": 3.3773116544619586,
+      "grad_norm": 0.07363371915212126,
+      "learning_rate": 5.833873089686815e-07,
+      "loss": 0.4304,
+      "step": 6808
+    },
+    {
+      "epoch": 3.377808117165198,
+      "grad_norm": 0.07100216585939323,
+      "learning_rate": 5.824713505034651e-07,
+      "loss": 0.4214,
+      "step": 6809
+    },
+    {
+      "epoch": 3.3783045798684372,
+      "grad_norm": 0.07409743627324886,
+      "learning_rate": 5.815560671866721e-07,
+      "loss": 0.4566,
+      "step": 6810
+    },
+    {
+      "epoch": 3.378801042571677,
+      "grad_norm": 0.07281506492211258,
+      "learning_rate": 5.806414591581916e-07,
+      "loss": 0.4604,
+      "step": 6811
+    },
+    {
+      "epoch": 3.3792975052749163,
+      "grad_norm": 0.07405617874881373,
+      "learning_rate": 5.797275265578034e-07,
+      "loss": 0.4454,
+      "step": 6812
+    },
+    {
+      "epoch": 3.3797939679781557,
+      "grad_norm": 0.07084226874488449,
+      "learning_rate": 5.788142695251897e-07,
+      "loss": 0.4195,
+      "step": 6813
+    },
+    {
+      "epoch": 3.380290430681395,
+      "grad_norm": 0.07461650004859605,
+      "learning_rate": 5.779016881999267e-07,
+      "loss": 0.4474,
+      "step": 6814
+    },
+    {
+      "epoch": 3.3807868933846343,
+      "grad_norm": 0.0739634619535666,
+      "learning_rate": 5.769897827214871e-07,
+      "loss": 0.4467,
+      "step": 6815
+    },
+    {
+      "epoch": 3.3812833560878737,
+      "grad_norm": 0.07162359935116724,
+      "learning_rate": 5.760785532292424e-07,
+      "loss": 0.4381,
+      "step": 6816
+    },
+    {
+      "epoch": 3.3817798187911134,
+      "grad_norm": 0.07321476360130366,
+      "learning_rate": 5.751679998624571e-07,
+      "loss": 0.4679,
+      "step": 6817
+    },
+    {
+      "epoch": 3.3822762814943528,
+      "grad_norm": 0.07121014109939985,
+      "learning_rate": 5.742581227602978e-07,
+      "loss": 0.421,
+      "step": 6818
+    },
+    {
+      "epoch": 3.382772744197592,
+      "grad_norm": 0.0713190118930726,
+      "learning_rate": 5.733489220618232e-07,
+      "loss": 0.4041,
+      "step": 6819
+    },
+    {
+      "epoch": 3.3832692069008314,
+      "grad_norm": 0.07061075340016978,
+      "learning_rate": 5.724403979059884e-07,
+      "loss": 0.4381,
+      "step": 6820
+    },
+    {
+      "epoch": 3.383765669604071,
+      "grad_norm": 0.07239685849294974,
+      "learning_rate": 5.715325504316493e-07,
+      "loss": 0.4477,
+      "step": 6821
+    },
+    {
+      "epoch": 3.3842621323073105,
+      "grad_norm": 0.07300640702874911,
+      "learning_rate": 5.70625379777554e-07,
+      "loss": 0.4121,
+      "step": 6822
+    },
+    {
+      "epoch": 3.38475859501055,
+      "grad_norm": 0.07236966548859305,
+      "learning_rate": 5.69718886082351e-07,
+      "loss": 0.4273,
+      "step": 6823
+    },
+    {
+      "epoch": 3.385255057713789,
+      "grad_norm": 0.07644248273040816,
+      "learning_rate": 5.688130694845817e-07,
+      "loss": 0.4504,
+      "step": 6824
+    },
+    {
+      "epoch": 3.3857515204170285,
+      "grad_norm": 0.07547400432157439,
+      "learning_rate": 5.679079301226853e-07,
+      "loss": 0.4404,
+      "step": 6825
+    },
+    {
+      "epoch": 3.386247983120268,
+      "grad_norm": 0.07373569405446405,
+      "learning_rate": 5.670034681349995e-07,
+      "loss": 0.4498,
+      "step": 6826
+    },
+    {
+      "epoch": 3.3867444458235076,
+      "grad_norm": 0.07244341261231016,
+      "learning_rate": 5.66099683659756e-07,
+      "loss": 0.4322,
+      "step": 6827
+    },
+    {
+      "epoch": 3.387240908526747,
+      "grad_norm": 0.07268650675591956,
+      "learning_rate": 5.651965768350836e-07,
+      "loss": 0.4398,
+      "step": 6828
+    },
+    {
+      "epoch": 3.3877373712299863,
+      "grad_norm": 0.07528774455110801,
+      "learning_rate": 5.642941477990078e-07,
+      "loss": 0.4612,
+      "step": 6829
+    },
+    {
+      "epoch": 3.3882338339332256,
+      "grad_norm": 0.07471360863771179,
+      "learning_rate": 5.633923966894495e-07,
+      "loss": 0.4464,
+      "step": 6830
+    },
+    {
+      "epoch": 3.3887302966364654,
+      "grad_norm": 0.07306230039019465,
+      "learning_rate": 5.624913236442287e-07,
+      "loss": 0.4333,
+      "step": 6831
+    },
+    {
+      "epoch": 3.3892267593397047,
+      "grad_norm": 0.07490854044047122,
+      "learning_rate": 5.615909288010579e-07,
+      "loss": 0.4461,
+      "step": 6832
+    },
+    {
+      "epoch": 3.389723222042944,
+      "grad_norm": 0.07476389460652377,
+      "learning_rate": 5.606912122975499e-07,
+      "loss": 0.4588,
+      "step": 6833
+    },
+    {
+      "epoch": 3.3902196847461834,
+      "grad_norm": 0.07136672326716272,
+      "learning_rate": 5.597921742712115e-07,
+      "loss": 0.461,
+      "step": 6834
+    },
+    {
+      "epoch": 3.3907161474494227,
+      "grad_norm": 0.07352205733020133,
+      "learning_rate": 5.588938148594452e-07,
+      "loss": 0.4571,
+      "step": 6835
+    },
+    {
+      "epoch": 3.391212610152662,
+      "grad_norm": 0.0733940024162979,
+      "learning_rate": 5.579961341995521e-07,
+      "loss": 0.4801,
+      "step": 6836
+    },
+    {
+      "epoch": 3.391709072855902,
+      "grad_norm": 0.07133932398326485,
+      "learning_rate": 5.570991324287273e-07,
+      "loss": 0.4106,
+      "step": 6837
+    },
+    {
+      "epoch": 3.392205535559141,
+      "grad_norm": 0.07094672541772178,
+      "learning_rate": 5.562028096840638e-07,
+      "loss": 0.4385,
+      "step": 6838
+    },
+    {
+      "epoch": 3.3927019982623805,
+      "grad_norm": 0.07449748299710027,
+      "learning_rate": 5.553071661025505e-07,
+      "loss": 0.4311,
+      "step": 6839
+    },
+    {
+      "epoch": 3.39319846096562,
+      "grad_norm": 0.07301207533923466,
+      "learning_rate": 5.544122018210707e-07,
+      "loss": 0.4501,
+      "step": 6840
+    },
+    {
+      "epoch": 3.3936949236688596,
+      "grad_norm": 0.07324337727308855,
+      "learning_rate": 5.535179169764071e-07,
+      "loss": 0.4628,
+      "step": 6841
+    },
+    {
+      "epoch": 3.394191386372099,
+      "grad_norm": 0.0727988496068958,
+      "learning_rate": 5.526243117052354e-07,
+      "loss": 0.4393,
+      "step": 6842
+    },
+    {
+      "epoch": 3.3946878490753383,
+      "grad_norm": 0.07377019414054901,
+      "learning_rate": 5.517313861441309e-07,
+      "loss": 0.4411,
+      "step": 6843
+    },
+    {
+      "epoch": 3.3951843117785776,
+      "grad_norm": 0.07336493283660565,
+      "learning_rate": 5.508391404295593e-07,
+      "loss": 0.4247,
+      "step": 6844
+    },
+    {
+      "epoch": 3.395680774481817,
+      "grad_norm": 0.07302798948644608,
+      "learning_rate": 5.499475746978899e-07,
+      "loss": 0.4614,
+      "step": 6845
+    },
+    {
+      "epoch": 3.3961772371850563,
+      "grad_norm": 0.07418133519525473,
+      "learning_rate": 5.490566890853822e-07,
+      "loss": 0.4425,
+      "step": 6846
+    },
+    {
+      "epoch": 3.396673699888296,
+      "grad_norm": 0.07046780669261539,
+      "learning_rate": 5.48166483728193e-07,
+      "loss": 0.4097,
+      "step": 6847
+    },
+    {
+      "epoch": 3.3971701625915354,
+      "grad_norm": 0.07472406252686546,
+      "learning_rate": 5.472769587623783e-07,
+      "loss": 0.4413,
+      "step": 6848
+    },
+    {
+      "epoch": 3.3976666252947747,
+      "grad_norm": 0.07340409078178213,
+      "learning_rate": 5.463881143238852e-07,
+      "loss": 0.425,
+      "step": 6849
+    },
+    {
+      "epoch": 3.398163087998014,
+      "grad_norm": 0.0752792541480088,
+      "learning_rate": 5.454999505485614e-07,
+      "loss": 0.4841,
+      "step": 6850
+    },
+    {
+      "epoch": 3.398659550701254,
+      "grad_norm": 0.07103666393741491,
+      "learning_rate": 5.446124675721482e-07,
+      "loss": 0.4576,
+      "step": 6851
+    },
+    {
+      "epoch": 3.399156013404493,
+      "grad_norm": 0.07172020968736212,
+      "learning_rate": 5.437256655302814e-07,
+      "loss": 0.4163,
+      "step": 6852
+    },
+    {
+      "epoch": 3.3996524761077325,
+      "grad_norm": 0.07079432917083367,
+      "learning_rate": 5.428395445584967e-07,
+      "loss": 0.4185,
+      "step": 6853
+    },
+    {
+      "epoch": 3.400148938810972,
+      "grad_norm": 0.07213578508832098,
+      "learning_rate": 5.419541047922217e-07,
+      "loss": 0.44,
+      "step": 6854
+    },
+    {
+      "epoch": 3.400645401514211,
+      "grad_norm": 0.07221119180833911,
+      "learning_rate": 5.410693463667827e-07,
+      "loss": 0.441,
+      "step": 6855
+    },
+    {
+      "epoch": 3.4011418642174505,
+      "grad_norm": 0.07210326161429813,
+      "learning_rate": 5.401852694174015e-07,
+      "loss": 0.4461,
+      "step": 6856
+    },
+    {
+      "epoch": 3.4016383269206902,
+      "grad_norm": 0.07588456212802579,
+      "learning_rate": 5.393018740791928e-07,
+      "loss": 0.4971,
+      "step": 6857
+    },
+    {
+      "epoch": 3.4021347896239296,
+      "grad_norm": 0.07226863179149022,
+      "learning_rate": 5.384191604871714e-07,
+      "loss": 0.4659,
+      "step": 6858
+    },
+    {
+      "epoch": 3.402631252327169,
+      "grad_norm": 0.07113951574127732,
+      "learning_rate": 5.37537128776246e-07,
+      "loss": 0.4367,
+      "step": 6859
+    },
+    {
+      "epoch": 3.4031277150304082,
+      "grad_norm": 0.0699606538530446,
+      "learning_rate": 5.3665577908122e-07,
+      "loss": 0.4076,
+      "step": 6860
+    },
+    {
+      "epoch": 3.403624177733648,
+      "grad_norm": 0.07404117594298328,
+      "learning_rate": 5.357751115367927e-07,
+      "loss": 0.4679,
+      "step": 6861
+    },
+    {
+      "epoch": 3.4041206404368873,
+      "grad_norm": 0.07338136610437596,
+      "learning_rate": 5.348951262775626e-07,
+      "loss": 0.4255,
+      "step": 6862
+    },
+    {
+      "epoch": 3.4046171031401267,
+      "grad_norm": 0.07467626851816644,
+      "learning_rate": 5.340158234380194e-07,
+      "loss": 0.4465,
+      "step": 6863
+    },
+    {
+      "epoch": 3.405113565843366,
+      "grad_norm": 0.07499822378987879,
+      "learning_rate": 5.331372031525506e-07,
+      "loss": 0.4405,
+      "step": 6864
+    },
+    {
+      "epoch": 3.4056100285466053,
+      "grad_norm": 0.07267811874010739,
+      "learning_rate": 5.322592655554404e-07,
+      "loss": 0.4457,
+      "step": 6865
+    },
+    {
+      "epoch": 3.4061064912498447,
+      "grad_norm": 0.07389241397863644,
+      "learning_rate": 5.313820107808665e-07,
+      "loss": 0.4707,
+      "step": 6866
+    },
+    {
+      "epoch": 3.4066029539530844,
+      "grad_norm": 0.07501551342292902,
+      "learning_rate": 5.305054389629022e-07,
+      "loss": 0.4596,
+      "step": 6867
+    },
+    {
+      "epoch": 3.4070994166563238,
+      "grad_norm": 0.07386900003627883,
+      "learning_rate": 5.296295502355203e-07,
+      "loss": 0.4547,
+      "step": 6868
+    },
+    {
+      "epoch": 3.407595879359563,
+      "grad_norm": 0.07222630259324854,
+      "learning_rate": 5.287543447325832e-07,
+      "loss": 0.4221,
+      "step": 6869
+    },
+    {
+      "epoch": 3.4080923420628024,
+      "grad_norm": 0.07411540724831181,
+      "learning_rate": 5.278798225878546e-07,
+      "loss": 0.4245,
+      "step": 6870
+    },
+    {
+      "epoch": 3.4085888047660418,
+      "grad_norm": 0.072855657027542,
+      "learning_rate": 5.2700598393499e-07,
+      "loss": 0.4279,
+      "step": 6871
+    },
+    {
+      "epoch": 3.4090852674692815,
+      "grad_norm": 0.07286142487660482,
+      "learning_rate": 5.261328289075413e-07,
+      "loss": 0.432,
+      "step": 6872
+    },
+    {
+      "epoch": 3.409581730172521,
+      "grad_norm": 0.07717464831068466,
+      "learning_rate": 5.25260357638957e-07,
+      "loss": 0.5038,
+      "step": 6873
+    },
+    {
+      "epoch": 3.41007819287576,
+      "grad_norm": 0.07193552232157858,
+      "learning_rate": 5.243885702625795e-07,
+      "loss": 0.4228,
+      "step": 6874
+    },
+    {
+      "epoch": 3.4105746555789995,
+      "grad_norm": 0.07131623721625098,
+      "learning_rate": 5.235174669116499e-07,
+      "loss": 0.4349,
+      "step": 6875
+    },
+    {
+      "epoch": 3.411071118282239,
+      "grad_norm": 0.0741642097125669,
+      "learning_rate": 5.226470477192991e-07,
+      "loss": 0.4459,
+      "step": 6876
+    },
+    {
+      "epoch": 3.4115675809854786,
+      "grad_norm": 0.07392174863668086,
+      "learning_rate": 5.217773128185582e-07,
+      "loss": 0.4392,
+      "step": 6877
+    },
+    {
+      "epoch": 3.412064043688718,
+      "grad_norm": 0.07591407769626606,
+      "learning_rate": 5.209082623423528e-07,
+      "loss": 0.4488,
+      "step": 6878
+    },
+    {
+      "epoch": 3.4125605063919573,
+      "grad_norm": 0.0728908444161917,
+      "learning_rate": 5.200398964235015e-07,
+      "loss": 0.4363,
+      "step": 6879
+    },
+    {
+      "epoch": 3.4130569690951966,
+      "grad_norm": 0.07095828506917863,
+      "learning_rate": 5.191722151947227e-07,
+      "loss": 0.4358,
+      "step": 6880
+    },
+    {
+      "epoch": 3.413553431798436,
+      "grad_norm": 0.07433082837337116,
+      "learning_rate": 5.18305218788625e-07,
+      "loss": 0.437,
+      "step": 6881
+    },
+    {
+      "epoch": 3.4140498945016757,
+      "grad_norm": 0.07254351436267927,
+      "learning_rate": 5.174389073377167e-07,
+      "loss": 0.4266,
+      "step": 6882
+    },
+    {
+      "epoch": 3.414546357204915,
+      "grad_norm": 0.07367975798141735,
+      "learning_rate": 5.165732809743995e-07,
+      "loss": 0.4083,
+      "step": 6883
+    },
+    {
+      "epoch": 3.4150428199081544,
+      "grad_norm": 0.07256116559937056,
+      "learning_rate": 5.157083398309687e-07,
+      "loss": 0.4158,
+      "step": 6884
+    },
+    {
+      "epoch": 3.4155392826113937,
+      "grad_norm": 0.0741371545114394,
+      "learning_rate": 5.148440840396191e-07,
+      "loss": 0.4409,
+      "step": 6885
+    },
+    {
+      "epoch": 3.416035745314633,
+      "grad_norm": 0.0747069253378787,
+      "learning_rate": 5.139805137324366e-07,
+      "loss": 0.4372,
+      "step": 6886
+    },
+    {
+      "epoch": 3.416532208017873,
+      "grad_norm": 0.07110773162800223,
+      "learning_rate": 5.131176290414053e-07,
+      "loss": 0.4526,
+      "step": 6887
+    },
+    {
+      "epoch": 3.417028670721112,
+      "grad_norm": 0.06964999302529894,
+      "learning_rate": 5.122554300984028e-07,
+      "loss": 0.41,
+      "step": 6888
+    },
+    {
+      "epoch": 3.4175251334243515,
+      "grad_norm": 0.07556264542454232,
+      "learning_rate": 5.113939170352012e-07,
+      "loss": 0.4554,
+      "step": 6889
+    },
+    {
+      "epoch": 3.418021596127591,
+      "grad_norm": 0.07304702051349063,
+      "learning_rate": 5.105330899834715e-07,
+      "loss": 0.4468,
+      "step": 6890
+    },
+    {
+      "epoch": 3.41851805883083,
+      "grad_norm": 0.07193806001341835,
+      "learning_rate": 5.096729490747754e-07,
+      "loss": 0.4452,
+      "step": 6891
+    },
+    {
+      "epoch": 3.41901452153407,
+      "grad_norm": 0.07089186955516436,
+      "learning_rate": 5.08813494440572e-07,
+      "loss": 0.4214,
+      "step": 6892
+    },
+    {
+      "epoch": 3.4195109842373093,
+      "grad_norm": 0.07265896324265066,
+      "learning_rate": 5.079547262122147e-07,
+      "loss": 0.4419,
+      "step": 6893
+    },
+    {
+      "epoch": 3.4200074469405486,
+      "grad_norm": 0.07104510900451609,
+      "learning_rate": 5.07096644520954e-07,
+      "loss": 0.3984,
+      "step": 6894
+    },
+    {
+      "epoch": 3.420503909643788,
+      "grad_norm": 0.07132578371343347,
+      "learning_rate": 5.062392494979329e-07,
+      "loss": 0.4506,
+      "step": 6895
+    },
+    {
+      "epoch": 3.4210003723470273,
+      "grad_norm": 0.07208462645684753,
+      "learning_rate": 5.053825412741892e-07,
+      "loss": 0.4283,
+      "step": 6896
+    },
+    {
+      "epoch": 3.421496835050267,
+      "grad_norm": 0.0736822957018377,
+      "learning_rate": 5.045265199806599e-07,
+      "loss": 0.4546,
+      "step": 6897
+    },
+    {
+      "epoch": 3.4219932977535064,
+      "grad_norm": 0.0744193833215387,
+      "learning_rate": 5.036711857481713e-07,
+      "loss": 0.4315,
+      "step": 6898
+    },
+    {
+      "epoch": 3.4224897604567457,
+      "grad_norm": 0.07275794000618309,
+      "learning_rate": 5.028165387074496e-07,
+      "loss": 0.4233,
+      "step": 6899
+    },
+    {
+      "epoch": 3.422986223159985,
+      "grad_norm": 0.07341302980067879,
+      "learning_rate": 5.019625789891136e-07,
+      "loss": 0.4438,
+      "step": 6900
+    },
+    {
+      "epoch": 3.4234826858632244,
+      "grad_norm": 0.07474191570866216,
+      "learning_rate": 5.011093067236756e-07,
+      "loss": 0.3999,
+      "step": 6901
+    },
+    {
+      "epoch": 3.423979148566464,
+      "grad_norm": 0.07628483834441871,
+      "learning_rate": 5.002567220415467e-07,
+      "loss": 0.4371,
+      "step": 6902
+    },
+    {
+      "epoch": 3.4244756112697035,
+      "grad_norm": 0.07391342730693265,
+      "learning_rate": 4.994048250730299e-07,
+      "loss": 0.4255,
+      "step": 6903
+    },
+    {
+      "epoch": 3.424972073972943,
+      "grad_norm": 0.07509989493312617,
+      "learning_rate": 4.985536159483234e-07,
+      "loss": 0.4223,
+      "step": 6904
+    },
+    {
+      "epoch": 3.425468536676182,
+      "grad_norm": 0.07190502503603871,
+      "learning_rate": 4.97703094797522e-07,
+      "loss": 0.4394,
+      "step": 6905
+    },
+    {
+      "epoch": 3.4259649993794214,
+      "grad_norm": 0.07336221763620508,
+      "learning_rate": 4.968532617506133e-07,
+      "loss": 0.4376,
+      "step": 6906
+    },
+    {
+      "epoch": 3.4264614620826612,
+      "grad_norm": 0.07212699547695878,
+      "learning_rate": 4.960041169374824e-07,
+      "loss": 0.4522,
+      "step": 6907
+    },
+    {
+      "epoch": 3.4269579247859006,
+      "grad_norm": 0.07207613536735617,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.4439,
+      "step": 6908
+    },
+    {
+      "epoch": 3.42745438748914,
+      "grad_norm": 0.07366223517554166,
+      "learning_rate": 4.943078925315553e-07,
+      "loss": 0.4529,
+      "step": 6909
+    },
+    {
+      "epoch": 3.427950850192379,
+      "grad_norm": 0.07002804566268289,
+      "learning_rate": 4.934608131980012e-07,
+      "loss": 0.4486,
+      "step": 6910
+    },
+    {
+      "epoch": 3.4284473128956185,
+      "grad_norm": 0.07405308706476917,
+      "learning_rate": 4.926144226167045e-07,
+      "loss": 0.4445,
+      "step": 6911
+    },
+    {
+      "epoch": 3.4289437755988583,
+      "grad_norm": 0.0730847232323671,
+      "learning_rate": 4.917687209170235e-07,
+      "loss": 0.4352,
+      "step": 6912
+    },
+    {
+      "epoch": 3.4294402383020977,
+      "grad_norm": 0.07075714549242612,
+      "learning_rate": 4.909237082282081e-07,
+      "loss": 0.4322,
+      "step": 6913
+    },
+    {
+      "epoch": 3.429936701005337,
+      "grad_norm": 0.07218842619876654,
+      "learning_rate": 4.900793846794077e-07,
+      "loss": 0.456,
+      "step": 6914
+    },
+    {
+      "epoch": 3.4304331637085763,
+      "grad_norm": 0.07554674078927479,
+      "learning_rate": 4.892357503996625e-07,
+      "loss": 0.454,
+      "step": 6915
+    },
+    {
+      "epoch": 3.4309296264118156,
+      "grad_norm": 0.07090126832846179,
+      "learning_rate": 4.883928055179072e-07,
+      "loss": 0.4395,
+      "step": 6916
+    },
+    {
+      "epoch": 3.4314260891150554,
+      "grad_norm": 0.07396293062611849,
+      "learning_rate": 4.87550550162974e-07,
+      "loss": 0.4567,
+      "step": 6917
+    },
+    {
+      "epoch": 3.4319225518182948,
+      "grad_norm": 0.0734029474539442,
+      "learning_rate": 4.867089844635875e-07,
+      "loss": 0.4442,
+      "step": 6918
+    },
+    {
+      "epoch": 3.432419014521534,
+      "grad_norm": 0.07125295252218236,
+      "learning_rate": 4.85868108548368e-07,
+      "loss": 0.4483,
+      "step": 6919
+    },
+    {
+      "epoch": 3.4329154772247734,
+      "grad_norm": 0.07308178603947592,
+      "learning_rate": 4.850279225458293e-07,
+      "loss": 0.4446,
+      "step": 6920
+    },
+    {
+      "epoch": 3.4334119399280127,
+      "grad_norm": 0.07120202733530115,
+      "learning_rate": 4.841884265843799e-07,
+      "loss": 0.4461,
+      "step": 6921
+    },
+    {
+      "epoch": 3.4339084026312525,
+      "grad_norm": 0.07536856568384741,
+      "learning_rate": 4.83349620792325e-07,
+      "loss": 0.4696,
+      "step": 6922
+    },
+    {
+      "epoch": 3.434404865334492,
+      "grad_norm": 0.07182558717594947,
+      "learning_rate": 4.825115052978613e-07,
+      "loss": 0.4384,
+      "step": 6923
+    },
+    {
+      "epoch": 3.434901328037731,
+      "grad_norm": 0.07693529697183651,
+      "learning_rate": 4.816740802290814e-07,
+      "loss": 0.4412,
+      "step": 6924
+    },
+    {
+      "epoch": 3.4353977907409705,
+      "grad_norm": 0.07251550677690793,
+      "learning_rate": 4.80837345713972e-07,
+      "loss": 0.4171,
+      "step": 6925
+    },
+    {
+      "epoch": 3.43589425344421,
+      "grad_norm": 0.0728053609741973,
+      "learning_rate": 4.800013018804156e-07,
+      "loss": 0.4297,
+      "step": 6926
+    },
+    {
+      "epoch": 3.436390716147449,
+      "grad_norm": 0.07209631348497708,
+      "learning_rate": 4.791659488561878e-07,
+      "loss": 0.4595,
+      "step": 6927
+    },
+    {
+      "epoch": 3.436887178850689,
+      "grad_norm": 0.07294278149924639,
+      "learning_rate": 4.783312867689577e-07,
+      "loss": 0.4695,
+      "step": 6928
+    },
+    {
+      "epoch": 3.4373836415539283,
+      "grad_norm": 0.07281794317221295,
+      "learning_rate": 4.77497315746292e-07,
+      "loss": 0.4364,
+      "step": 6929
+    },
+    {
+      "epoch": 3.4378801042571676,
+      "grad_norm": 0.07092418301279209,
+      "learning_rate": 4.766640359156477e-07,
+      "loss": 0.4218,
+      "step": 6930
+    },
+    {
+      "epoch": 3.438376566960407,
+      "grad_norm": 0.07039331739060295,
+      "learning_rate": 4.7583144740438015e-07,
+      "loss": 0.4332,
+      "step": 6931
+    },
+    {
+      "epoch": 3.4388730296636467,
+      "grad_norm": 0.07087669079862956,
+      "learning_rate": 4.74999550339737e-07,
+      "loss": 0.405,
+      "step": 6932
+    },
+    {
+      "epoch": 3.439369492366886,
+      "grad_norm": 0.0710694632443892,
+      "learning_rate": 4.741683448488582e-07,
+      "loss": 0.4212,
+      "step": 6933
+    },
+    {
+      "epoch": 3.4398659550701254,
+      "grad_norm": 0.07324992738813729,
+      "learning_rate": 4.733378310587827e-07,
+      "loss": 0.4639,
+      "step": 6934
+    },
+    {
+      "epoch": 3.4403624177733647,
+      "grad_norm": 0.07189516683464231,
+      "learning_rate": 4.72508009096439e-07,
+      "loss": 0.4173,
+      "step": 6935
+    },
+    {
+      "epoch": 3.440858880476604,
+      "grad_norm": 0.07210803887462554,
+      "learning_rate": 4.716788790886545e-07,
+      "loss": 0.442,
+      "step": 6936
+    },
+    {
+      "epoch": 3.4413553431798434,
+      "grad_norm": 0.07444855352560995,
+      "learning_rate": 4.708504411621473e-07,
+      "loss": 0.4557,
+      "step": 6937
+    },
+    {
+      "epoch": 3.441851805883083,
+      "grad_norm": 0.07405575332699776,
+      "learning_rate": 4.7002269544352996e-07,
+      "loss": 0.4949,
+      "step": 6938
+    },
+    {
+      "epoch": 3.4423482685863225,
+      "grad_norm": 0.07091669568635707,
+      "learning_rate": 4.6919564205931244e-07,
+      "loss": 0.431,
+      "step": 6939
+    },
+    {
+      "epoch": 3.442844731289562,
+      "grad_norm": 0.07403138375012017,
+      "learning_rate": 4.683692811358936e-07,
+      "loss": 0.4486,
+      "step": 6940
+    },
+    {
+      "epoch": 3.443341193992801,
+      "grad_norm": 0.07295796845958885,
+      "learning_rate": 4.6754361279957193e-07,
+      "loss": 0.4425,
+      "step": 6941
+    },
+    {
+      "epoch": 3.443837656696041,
+      "grad_norm": 0.07398747244263879,
+      "learning_rate": 4.667186371765364e-07,
+      "loss": 0.4608,
+      "step": 6942
+    },
+    {
+      "epoch": 3.4443341193992802,
+      "grad_norm": 0.07362634337316348,
+      "learning_rate": 4.658943543928707e-07,
+      "loss": 0.4205,
+      "step": 6943
+    },
+    {
+      "epoch": 3.4448305821025196,
+      "grad_norm": 0.07217553443395912,
+      "learning_rate": 4.6507076457455445e-07,
+      "loss": 0.458,
+      "step": 6944
+    },
+    {
+      "epoch": 3.445327044805759,
+      "grad_norm": 0.0729885922925346,
+      "learning_rate": 4.6424786784745936e-07,
+      "loss": 0.4691,
+      "step": 6945
+    },
+    {
+      "epoch": 3.4458235075089982,
+      "grad_norm": 0.07252264968726019,
+      "learning_rate": 4.634256643373536e-07,
+      "loss": 0.4258,
+      "step": 6946
+    },
+    {
+      "epoch": 3.4463199702122376,
+      "grad_norm": 0.07482809179360157,
+      "learning_rate": 4.6260415416989613e-07,
+      "loss": 0.4692,
+      "step": 6947
+    },
+    {
+      "epoch": 3.4468164329154773,
+      "grad_norm": 0.07322451229011788,
+      "learning_rate": 4.617833374706415e-07,
+      "loss": 0.4517,
+      "step": 6948
+    },
+    {
+      "epoch": 3.4473128956187167,
+      "grad_norm": 0.07452852578412698,
+      "learning_rate": 4.6096321436504e-07,
+      "loss": 0.4663,
+      "step": 6949
+    },
+    {
+      "epoch": 3.447809358321956,
+      "grad_norm": 0.07774629413412555,
+      "learning_rate": 4.601437849784318e-07,
+      "loss": 0.4618,
+      "step": 6950
+    },
+    {
+      "epoch": 3.4483058210251953,
+      "grad_norm": 0.07401524988974062,
+      "learning_rate": 4.593250494360563e-07,
+      "loss": 0.4552,
+      "step": 6951
+    },
+    {
+      "epoch": 3.448802283728435,
+      "grad_norm": 0.07257419698991296,
+      "learning_rate": 4.585070078630427e-07,
+      "loss": 0.4465,
+      "step": 6952
+    },
+    {
+      "epoch": 3.4492987464316744,
+      "grad_norm": 0.0749883510950849,
+      "learning_rate": 4.57689660384415e-07,
+      "loss": 0.4798,
+      "step": 6953
+    },
+    {
+      "epoch": 3.4497952091349138,
+      "grad_norm": 0.07509551841042462,
+      "learning_rate": 4.568730071250926e-07,
+      "loss": 0.4721,
+      "step": 6954
+    },
+    {
+      "epoch": 3.450291671838153,
+      "grad_norm": 0.07320749923185751,
+      "learning_rate": 4.560570482098875e-07,
+      "loss": 0.4258,
+      "step": 6955
+    },
+    {
+      "epoch": 3.4507881345413924,
+      "grad_norm": 0.0732758502467655,
+      "learning_rate": 4.5524178376350703e-07,
+      "loss": 0.4442,
+      "step": 6956
+    },
+    {
+      "epoch": 3.4512845972446318,
+      "grad_norm": 0.0738737475124385,
+      "learning_rate": 4.544272139105488e-07,
+      "loss": 0.4272,
+      "step": 6957
+    },
+    {
+      "epoch": 3.4517810599478715,
+      "grad_norm": 0.07500508451311125,
+      "learning_rate": 4.536133387755093e-07,
+      "loss": 0.4261,
+      "step": 6958
+    },
+    {
+      "epoch": 3.452277522651111,
+      "grad_norm": 0.06947963423355868,
+      "learning_rate": 4.528001584827746e-07,
+      "loss": 0.3949,
+      "step": 6959
+    },
+    {
+      "epoch": 3.45277398535435,
+      "grad_norm": 0.07202057902623601,
+      "learning_rate": 4.519876731566264e-07,
+      "loss": 0.4467,
+      "step": 6960
+    },
+    {
+      "epoch": 3.4532704480575895,
+      "grad_norm": 0.07415885781551207,
+      "learning_rate": 4.511758829212415e-07,
+      "loss": 0.4295,
+      "step": 6961
+    },
+    {
+      "epoch": 3.4537669107608293,
+      "grad_norm": 0.07532471992904242,
+      "learning_rate": 4.5036478790068673e-07,
+      "loss": 0.4598,
+      "step": 6962
+    },
+    {
+      "epoch": 3.4542633734640686,
+      "grad_norm": 0.07091409618049023,
+      "learning_rate": 4.495543882189274e-07,
+      "loss": 0.4336,
+      "step": 6963
+    },
+    {
+      "epoch": 3.454759836167308,
+      "grad_norm": 0.07498174921876814,
+      "learning_rate": 4.487446839998194e-07,
+      "loss": 0.4352,
+      "step": 6964
+    },
+    {
+      "epoch": 3.4552562988705473,
+      "grad_norm": 0.07424799522056441,
+      "learning_rate": 4.47935675367111e-07,
+      "loss": 0.4472,
+      "step": 6965
+    },
+    {
+      "epoch": 3.4557527615737866,
+      "grad_norm": 0.07135977542277129,
+      "learning_rate": 4.4712736244444943e-07,
+      "loss": 0.437,
+      "step": 6966
+    },
+    {
+      "epoch": 3.456249224277026,
+      "grad_norm": 0.07419910319307793,
+      "learning_rate": 4.4631974535536914e-07,
+      "loss": 0.4626,
+      "step": 6967
+    },
+    {
+      "epoch": 3.4567456869802657,
+      "grad_norm": 0.07283909548202254,
+      "learning_rate": 4.455128242233042e-07,
+      "loss": 0.4374,
+      "step": 6968
+    },
+    {
+      "epoch": 3.457242149683505,
+      "grad_norm": 0.0740400261684498,
+      "learning_rate": 4.4470659917157877e-07,
+      "loss": 0.4847,
+      "step": 6969
+    },
+    {
+      "epoch": 3.4577386123867444,
+      "grad_norm": 0.07116917467308426,
+      "learning_rate": 4.439010703234098e-07,
+      "loss": 0.4155,
+      "step": 6970
+    },
+    {
+      "epoch": 3.4582350750899837,
+      "grad_norm": 0.07270125442975392,
+      "learning_rate": 4.4309623780191214e-07,
+      "loss": 0.4087,
+      "step": 6971
+    },
+    {
+      "epoch": 3.4587315377932235,
+      "grad_norm": 0.07391297126366274,
+      "learning_rate": 4.4229210173008964e-07,
+      "loss": 0.4322,
+      "step": 6972
+    },
+    {
+      "epoch": 3.459228000496463,
+      "grad_norm": 0.07503149368964877,
+      "learning_rate": 4.414886622308423e-07,
+      "loss": 0.466,
+      "step": 6973
+    },
+    {
+      "epoch": 3.459724463199702,
+      "grad_norm": 0.07132078609972732,
+      "learning_rate": 4.406859194269619e-07,
+      "loss": 0.4556,
+      "step": 6974
+    },
+    {
+      "epoch": 3.4602209259029415,
+      "grad_norm": 0.07621463735331618,
+      "learning_rate": 4.3988387344113647e-07,
+      "loss": 0.446,
+      "step": 6975
+    },
+    {
+      "epoch": 3.460717388606181,
+      "grad_norm": 0.07167692072989904,
+      "learning_rate": 4.390825243959451e-07,
+      "loss": 0.4559,
+      "step": 6976
+    },
+    {
+      "epoch": 3.46121385130942,
+      "grad_norm": 0.07245327069025828,
+      "learning_rate": 4.3828187241385987e-07,
+      "loss": 0.4435,
+      "step": 6977
+    },
+    {
+      "epoch": 3.46171031401266,
+      "grad_norm": 0.07229654905654437,
+      "learning_rate": 4.374819176172501e-07,
+      "loss": 0.4376,
+      "step": 6978
+    },
+    {
+      "epoch": 3.4622067767158993,
+      "grad_norm": 0.07266118777505072,
+      "learning_rate": 4.3668266012837523e-07,
+      "loss": 0.4354,
+      "step": 6979
+    },
+    {
+      "epoch": 3.4627032394191386,
+      "grad_norm": 0.07247688735375024,
+      "learning_rate": 4.35884100069387e-07,
+      "loss": 0.4195,
+      "step": 6980
+    },
+    {
+      "epoch": 3.463199702122378,
+      "grad_norm": 0.07168303899184163,
+      "learning_rate": 4.350862375623349e-07,
+      "loss": 0.428,
+      "step": 6981
+    },
+    {
+      "epoch": 3.4636961648256177,
+      "grad_norm": 0.07207520542686019,
+      "learning_rate": 4.3428907272915823e-07,
+      "loss": 0.424,
+      "step": 6982
+    },
+    {
+      "epoch": 3.464192627528857,
+      "grad_norm": 0.07432618035966075,
+      "learning_rate": 4.334926056916916e-07,
+      "loss": 0.4574,
+      "step": 6983
+    },
+    {
+      "epoch": 3.4646890902320964,
+      "grad_norm": 0.07394453174221209,
+      "learning_rate": 4.326968365716622e-07,
+      "loss": 0.4408,
+      "step": 6984
+    },
+    {
+      "epoch": 3.4651855529353357,
+      "grad_norm": 0.07210933065781855,
+      "learning_rate": 4.319017654906887e-07,
+      "loss": 0.4131,
+      "step": 6985
+    },
+    {
+      "epoch": 3.465682015638575,
+      "grad_norm": 0.07351614551970871,
+      "learning_rate": 4.31107392570288e-07,
+      "loss": 0.4648,
+      "step": 6986
+    },
+    {
+      "epoch": 3.4661784783418144,
+      "grad_norm": 0.0727033243995911,
+      "learning_rate": 4.303137179318645e-07,
+      "loss": 0.4454,
+      "step": 6987
+    },
+    {
+      "epoch": 3.466674941045054,
+      "grad_norm": 0.0726353997298416,
+      "learning_rate": 4.2952074169672175e-07,
+      "loss": 0.4401,
+      "step": 6988
+    },
+    {
+      "epoch": 3.4671714037482935,
+      "grad_norm": 0.07393255753739562,
+      "learning_rate": 4.287284639860495e-07,
+      "loss": 0.4799,
+      "step": 6989
+    },
+    {
+      "epoch": 3.467667866451533,
+      "grad_norm": 0.07015546639264951,
+      "learning_rate": 4.279368849209381e-07,
+      "loss": 0.4463,
+      "step": 6990
+    },
+    {
+      "epoch": 3.468164329154772,
+      "grad_norm": 0.07314951227584136,
+      "learning_rate": 4.271460046223663e-07,
+      "loss": 0.4375,
+      "step": 6991
+    },
+    {
+      "epoch": 3.468660791858012,
+      "grad_norm": 0.07076454092546122,
+      "learning_rate": 4.263558232112064e-07,
+      "loss": 0.4472,
+      "step": 6992
+    },
+    {
+      "epoch": 3.4691572545612512,
+      "grad_norm": 0.0718955639453092,
+      "learning_rate": 4.255663408082272e-07,
+      "loss": 0.4359,
+      "step": 6993
+    },
+    {
+      "epoch": 3.4696537172644906,
+      "grad_norm": 0.07275376732319178,
+      "learning_rate": 4.2477755753408625e-07,
+      "loss": 0.4332,
+      "step": 6994
+    },
+    {
+      "epoch": 3.47015017996773,
+      "grad_norm": 0.07130960107957575,
+      "learning_rate": 4.239894735093386e-07,
+      "loss": 0.439,
+      "step": 6995
+    },
+    {
+      "epoch": 3.4706466426709692,
+      "grad_norm": 0.0714544722893366,
+      "learning_rate": 4.2320208885442917e-07,
+      "loss": 0.4646,
+      "step": 6996
+    },
+    {
+      "epoch": 3.4711431053742086,
+      "grad_norm": 0.0740861914603814,
+      "learning_rate": 4.2241540368969604e-07,
+      "loss": 0.4442,
+      "step": 6997
+    },
+    {
+      "epoch": 3.4716395680774483,
+      "grad_norm": 0.07174384340724754,
+      "learning_rate": 4.216294181353736e-07,
+      "loss": 0.4337,
+      "step": 6998
+    },
+    {
+      "epoch": 3.4721360307806877,
+      "grad_norm": 0.07355509238233153,
+      "learning_rate": 4.2084413231158473e-07,
+      "loss": 0.4593,
+      "step": 6999
+    },
+    {
+      "epoch": 3.472632493483927,
+      "grad_norm": 0.0712277631637039,
+      "learning_rate": 4.2005954633835055e-07,
+      "loss": 0.4212,
+      "step": 7000
+    },
+    {
+      "epoch": 3.4731289561871663,
+      "grad_norm": 0.07241974680475426,
+      "learning_rate": 4.1927566033558075e-07,
+      "loss": 0.4101,
+      "step": 7001
+    },
+    {
+      "epoch": 3.473625418890406,
+      "grad_norm": 0.07326424285200016,
+      "learning_rate": 4.184924744230784e-07,
+      "loss": 0.446,
+      "step": 7002
+    },
+    {
+      "epoch": 3.4741218815936454,
+      "grad_norm": 0.07283954065452414,
+      "learning_rate": 4.1770998872054436e-07,
+      "loss": 0.4473,
+      "step": 7003
+    },
+    {
+      "epoch": 3.4746183442968848,
+      "grad_norm": 0.07298687385051912,
+      "learning_rate": 4.169282033475663e-07,
+      "loss": 0.4355,
+      "step": 7004
+    },
+    {
+      "epoch": 3.475114807000124,
+      "grad_norm": 0.06950575035689337,
+      "learning_rate": 4.1614711842362876e-07,
+      "loss": 0.4087,
+      "step": 7005
+    },
+    {
+      "epoch": 3.4756112697033634,
+      "grad_norm": 0.07351689625797672,
+      "learning_rate": 4.153667340681067e-07,
+      "loss": 0.441,
+      "step": 7006
+    },
+    {
+      "epoch": 3.4761077324066028,
+      "grad_norm": 0.07164469420788898,
+      "learning_rate": 4.1458705040027135e-07,
+      "loss": 0.4304,
+      "step": 7007
+    },
+    {
+      "epoch": 3.4766041951098425,
+      "grad_norm": 0.07412787005202179,
+      "learning_rate": 4.138080675392836e-07,
+      "loss": 0.4427,
+      "step": 7008
+    },
+    {
+      "epoch": 3.477100657813082,
+      "grad_norm": 0.07512842436789635,
+      "learning_rate": 4.130297856041976e-07,
+      "loss": 0.418,
+      "step": 7009
+    },
+    {
+      "epoch": 3.477597120516321,
+      "grad_norm": 0.07599135302617657,
+      "learning_rate": 4.1225220471396376e-07,
+      "loss": 0.4542,
+      "step": 7010
+    },
+    {
+      "epoch": 3.4780935832195605,
+      "grad_norm": 0.07028592568982106,
+      "learning_rate": 4.1147532498742035e-07,
+      "loss": 0.4326,
+      "step": 7011
+    },
+    {
+      "epoch": 3.4785900459228,
+      "grad_norm": 0.0723766237238303,
+      "learning_rate": 4.1069914654330357e-07,
+      "loss": 0.4337,
+      "step": 7012
+    },
+    {
+      "epoch": 3.4790865086260396,
+      "grad_norm": 0.07060596015984845,
+      "learning_rate": 4.099236695002379e-07,
+      "loss": 0.4206,
+      "step": 7013
+    },
+    {
+      "epoch": 3.479582971329279,
+      "grad_norm": 0.07124772160966228,
+      "learning_rate": 4.0914889397674243e-07,
+      "loss": 0.4288,
+      "step": 7014
+    },
+    {
+      "epoch": 3.4800794340325183,
+      "grad_norm": 0.07437517655189398,
+      "learning_rate": 4.0837482009123017e-07,
+      "loss": 0.4377,
+      "step": 7015
+    },
+    {
+      "epoch": 3.4805758967357576,
+      "grad_norm": 0.07370022244994137,
+      "learning_rate": 4.0760144796200605e-07,
+      "loss": 0.4463,
+      "step": 7016
+    },
+    {
+      "epoch": 3.481072359438997,
+      "grad_norm": 0.07031536555381257,
+      "learning_rate": 4.06828777707266e-07,
+      "loss": 0.4149,
+      "step": 7017
+    },
+    {
+      "epoch": 3.4815688221422367,
+      "grad_norm": 0.07604069978172456,
+      "learning_rate": 4.060568094451023e-07,
+      "loss": 0.4738,
+      "step": 7018
+    },
+    {
+      "epoch": 3.482065284845476,
+      "grad_norm": 0.0701995822374433,
+      "learning_rate": 4.05285543293496e-07,
+      "loss": 0.4036,
+      "step": 7019
+    },
+    {
+      "epoch": 3.4825617475487154,
+      "grad_norm": 0.07424172134755505,
+      "learning_rate": 4.045149793703257e-07,
+      "loss": 0.4318,
+      "step": 7020
+    },
+    {
+      "epoch": 3.4830582102519547,
+      "grad_norm": 0.07285730183823594,
+      "learning_rate": 4.03745117793356e-07,
+      "loss": 0.4597,
+      "step": 7021
+    },
+    {
+      "epoch": 3.483554672955194,
+      "grad_norm": 0.07295991960910402,
+      "learning_rate": 4.0297595868025065e-07,
+      "loss": 0.4283,
+      "step": 7022
+    },
+    {
+      "epoch": 3.484051135658434,
+      "grad_norm": 0.0723258436974514,
+      "learning_rate": 4.022075021485622e-07,
+      "loss": 0.424,
+      "step": 7023
+    },
+    {
+      "epoch": 3.484547598361673,
+      "grad_norm": 0.07298322109584299,
+      "learning_rate": 4.014397483157362e-07,
+      "loss": 0.423,
+      "step": 7024
+    },
+    {
+      "epoch": 3.4850440610649125,
+      "grad_norm": 0.07205478016555279,
+      "learning_rate": 4.0067269729911316e-07,
+      "loss": 0.4437,
+      "step": 7025
+    },
+    {
+      "epoch": 3.485540523768152,
+      "grad_norm": 0.07307966961163839,
+      "learning_rate": 3.999063492159233e-07,
+      "loss": 0.452,
+      "step": 7026
+    },
+    {
+      "epoch": 3.486036986471391,
+      "grad_norm": 0.07439585552315807,
+      "learning_rate": 3.9914070418329123e-07,
+      "loss": 0.4404,
+      "step": 7027
+    },
+    {
+      "epoch": 3.486533449174631,
+      "grad_norm": 0.07114088102753215,
+      "learning_rate": 3.983757623182338e-07,
+      "loss": 0.4224,
+      "step": 7028
+    },
+    {
+      "epoch": 3.4870299118778703,
+      "grad_norm": 0.07336142749381283,
+      "learning_rate": 3.9761152373765875e-07,
+      "loss": 0.4412,
+      "step": 7029
+    },
+    {
+      "epoch": 3.4875263745811096,
+      "grad_norm": 0.07072524821639858,
+      "learning_rate": 3.968479885583698e-07,
+      "loss": 0.4458,
+      "step": 7030
+    },
+    {
+      "epoch": 3.488022837284349,
+      "grad_norm": 0.07262332837310295,
+      "learning_rate": 3.960851568970586e-07,
+      "loss": 0.4425,
+      "step": 7031
+    },
+    {
+      "epoch": 3.4885192999875883,
+      "grad_norm": 0.07243265698774211,
+      "learning_rate": 3.953230288703136e-07,
+      "loss": 0.4293,
+      "step": 7032
+    },
+    {
+      "epoch": 3.489015762690828,
+      "grad_norm": 0.07096323367904923,
+      "learning_rate": 3.945616045946138e-07,
+      "loss": 0.4238,
+      "step": 7033
+    },
+    {
+      "epoch": 3.4895122253940674,
+      "grad_norm": 0.07077384511779652,
+      "learning_rate": 3.938008841863289e-07,
+      "loss": 0.4137,
+      "step": 7034
+    },
+    {
+      "epoch": 3.4900086880973067,
+      "grad_norm": 0.07527318638659324,
+      "learning_rate": 3.9304086776172535e-07,
+      "loss": 0.4349,
+      "step": 7035
+    },
+    {
+      "epoch": 3.490505150800546,
+      "grad_norm": 0.06952756070762323,
+      "learning_rate": 3.9228155543695803e-07,
+      "loss": 0.4102,
+      "step": 7036
+    },
+    {
+      "epoch": 3.4910016135037854,
+      "grad_norm": 0.07224901371189203,
+      "learning_rate": 3.915229473280757e-07,
+      "loss": 0.4401,
+      "step": 7037
+    },
+    {
+      "epoch": 3.491498076207025,
+      "grad_norm": 0.07204345824938724,
+      "learning_rate": 3.907650435510185e-07,
+      "loss": 0.4125,
+      "step": 7038
+    },
+    {
+      "epoch": 3.4919945389102645,
+      "grad_norm": 0.07180263804640478,
+      "learning_rate": 3.900078442216221e-07,
+      "loss": 0.4066,
+      "step": 7039
+    },
+    {
+      "epoch": 3.492491001613504,
+      "grad_norm": 0.0722475836347384,
+      "learning_rate": 3.8925134945561107e-07,
+      "loss": 0.4146,
+      "step": 7040
+    },
+    {
+      "epoch": 3.492987464316743,
+      "grad_norm": 0.06995833245259021,
+      "learning_rate": 3.8849555936860296e-07,
+      "loss": 0.4489,
+      "step": 7041
+    },
+    {
+      "epoch": 3.4934839270199824,
+      "grad_norm": 0.07158725913239221,
+      "learning_rate": 3.877404740761093e-07,
+      "loss": 0.4211,
+      "step": 7042
+    },
+    {
+      "epoch": 3.4939803897232222,
+      "grad_norm": 0.07263199309595254,
+      "learning_rate": 3.86986093693531e-07,
+      "loss": 0.4258,
+      "step": 7043
+    },
+    {
+      "epoch": 3.4944768524264616,
+      "grad_norm": 0.07346394269935339,
+      "learning_rate": 3.8623241833616543e-07,
+      "loss": 0.4543,
+      "step": 7044
+    },
+    {
+      "epoch": 3.494973315129701,
+      "grad_norm": 0.0726445809581451,
+      "learning_rate": 3.854794481191987e-07,
+      "loss": 0.435,
+      "step": 7045
+    },
+    {
+      "epoch": 3.49546977783294,
+      "grad_norm": 0.07285345373233935,
+      "learning_rate": 3.847271831577093e-07,
+      "loss": 0.4339,
+      "step": 7046
+    },
+    {
+      "epoch": 3.4959662405361795,
+      "grad_norm": 0.07409485080802107,
+      "learning_rate": 3.8397562356667026e-07,
+      "loss": 0.446,
+      "step": 7047
+    },
+    {
+      "epoch": 3.4964627032394193,
+      "grad_norm": 0.07012737357757955,
+      "learning_rate": 3.832247694609442e-07,
+      "loss": 0.4208,
+      "step": 7048
+    },
+    {
+      "epoch": 3.4969591659426587,
+      "grad_norm": 0.07193086143076212,
+      "learning_rate": 3.824746209552882e-07,
+      "loss": 0.4454,
+      "step": 7049
+    },
+    {
+      "epoch": 3.497455628645898,
+      "grad_norm": 0.07795249327587464,
+      "learning_rate": 3.817251781643505e-07,
+      "loss": 0.431,
+      "step": 7050
+    },
+    {
+      "epoch": 3.4979520913491373,
+      "grad_norm": 0.0716902556228763,
+      "learning_rate": 3.8097644120266954e-07,
+      "loss": 0.4214,
+      "step": 7051
+    },
+    {
+      "epoch": 3.4984485540523766,
+      "grad_norm": 0.07223590607474627,
+      "learning_rate": 3.8022841018468147e-07,
+      "loss": 0.4416,
+      "step": 7052
+    },
+    {
+      "epoch": 3.4989450167556164,
+      "grad_norm": 0.07436690378938308,
+      "learning_rate": 3.794810852247066e-07,
+      "loss": 0.457,
+      "step": 7053
+    },
+    {
+      "epoch": 3.4994414794588558,
+      "grad_norm": 0.07112956963476424,
+      "learning_rate": 3.787344664369641e-07,
+      "loss": 0.4184,
+      "step": 7054
+    },
+    {
+      "epoch": 3.499937942162095,
+      "grad_norm": 0.07104327637378509,
+      "learning_rate": 3.779885539355621e-07,
+      "loss": 0.4098,
+      "step": 7055
+    },
+    {
+      "epoch": 3.5004344048653344,
+      "grad_norm": 0.07306606499748441,
+      "learning_rate": 3.7724334783450054e-07,
+      "loss": 0.4592,
+      "step": 7056
+    },
+    {
+      "epoch": 3.5004344048653344,
+      "eval_loss": 0.5162577629089355,
+      "eval_runtime": 259.1404,
+      "eval_samples_per_second": 117.13,
+      "eval_steps_per_second": 14.645,
+      "step": 7056
+    },
+    {
+      "epoch": 3.5009308675685737,
+      "grad_norm": 0.07203781559335148,
+      "learning_rate": 3.764988482476739e-07,
+      "loss": 0.4234,
+      "step": 7057
+    },
+    {
+      "epoch": 3.501427330271813,
+      "grad_norm": 0.07546147841789971,
+      "learning_rate": 3.75755055288865e-07,
+      "loss": 0.4316,
+      "step": 7058
+    },
+    {
+      "epoch": 3.501923792975053,
+      "grad_norm": 0.07488137656265763,
+      "learning_rate": 3.7501196907175297e-07,
+      "loss": 0.4289,
+      "step": 7059
+    },
+    {
+      "epoch": 3.502420255678292,
+      "grad_norm": 0.07033502879304558,
+      "learning_rate": 3.742695897099052e-07,
+      "loss": 0.403,
+      "step": 7060
+    },
+    {
+      "epoch": 3.5029167183815315,
+      "grad_norm": 0.07491132145559015,
+      "learning_rate": 3.7352791731678164e-07,
+      "loss": 0.4701,
+      "step": 7061
+    },
+    {
+      "epoch": 3.503413181084771,
+      "grad_norm": 0.07385912091891961,
+      "learning_rate": 3.7278695200573754e-07,
+      "loss": 0.4438,
+      "step": 7062
+    },
+    {
+      "epoch": 3.5039096437880106,
+      "grad_norm": 0.07340689241571162,
+      "learning_rate": 3.720466938900147e-07,
+      "loss": 0.4538,
+      "step": 7063
+    },
+    {
+      "epoch": 3.50440610649125,
+      "grad_norm": 0.07295251691291561,
+      "learning_rate": 3.7130714308275196e-07,
+      "loss": 0.4239,
+      "step": 7064
+    },
+    {
+      "epoch": 3.5049025691944893,
+      "grad_norm": 0.07080688080671606,
+      "learning_rate": 3.705682996969773e-07,
+      "loss": 0.4475,
+      "step": 7065
+    },
+    {
+      "epoch": 3.5053990318977286,
+      "grad_norm": 0.07369474252686231,
+      "learning_rate": 3.6983016384560975e-07,
+      "loss": 0.449,
+      "step": 7066
+    },
+    {
+      "epoch": 3.505895494600968,
+      "grad_norm": 0.07286840009039136,
+      "learning_rate": 3.6909273564146366e-07,
+      "loss": 0.4371,
+      "step": 7067
+    },
+    {
+      "epoch": 3.5063919573042073,
+      "grad_norm": 0.07326717587537919,
+      "learning_rate": 3.683560151972415e-07,
+      "loss": 0.4451,
+      "step": 7068
+    },
+    {
+      "epoch": 3.506888420007447,
+      "grad_norm": 0.0730059610532952,
+      "learning_rate": 3.6762000262554e-07,
+      "loss": 0.4496,
+      "step": 7069
+    },
+    {
+      "epoch": 3.5073848827106864,
+      "grad_norm": 0.0725619435516716,
+      "learning_rate": 3.668846980388452e-07,
+      "loss": 0.4546,
+      "step": 7070
+    },
+    {
+      "epoch": 3.5078813454139257,
+      "grad_norm": 0.07202927835061003,
+      "learning_rate": 3.661501015495389e-07,
+      "loss": 0.4223,
+      "step": 7071
+    },
+    {
+      "epoch": 3.508377808117165,
+      "grad_norm": 0.06993486720630124,
+      "learning_rate": 3.6541621326989183e-07,
+      "loss": 0.4114,
+      "step": 7072
+    },
+    {
+      "epoch": 3.508874270820405,
+      "grad_norm": 0.07257818917447946,
+      "learning_rate": 3.6468303331206546e-07,
+      "loss": 0.4011,
+      "step": 7073
+    },
+    {
+      "epoch": 3.509370733523644,
+      "grad_norm": 0.07288833908931067,
+      "learning_rate": 3.6395056178811725e-07,
+      "loss": 0.4481,
+      "step": 7074
+    },
+    {
+      "epoch": 3.5098671962268835,
+      "grad_norm": 0.07092586904115811,
+      "learning_rate": 3.632187988099906e-07,
+      "loss": 0.4481,
+      "step": 7075
+    },
+    {
+      "epoch": 3.510363658930123,
+      "grad_norm": 0.07214368911148672,
+      "learning_rate": 3.6248774448952695e-07,
+      "loss": 0.4237,
+      "step": 7076
+    },
+    {
+      "epoch": 3.510860121633362,
+      "grad_norm": 0.0722242659165623,
+      "learning_rate": 3.617573989384543e-07,
+      "loss": 0.4321,
+      "step": 7077
+    },
+    {
+      "epoch": 3.5113565843366015,
+      "grad_norm": 0.06952702140904304,
+      "learning_rate": 3.6102776226839386e-07,
+      "loss": 0.4119,
+      "step": 7078
+    },
+    {
+      "epoch": 3.5118530470398412,
+      "grad_norm": 0.07076164720383553,
+      "learning_rate": 3.602988345908609e-07,
+      "loss": 0.4208,
+      "step": 7079
+    },
+    {
+      "epoch": 3.5123495097430806,
+      "grad_norm": 0.07340766789593617,
+      "learning_rate": 3.5957061601725797e-07,
+      "loss": 0.4519,
+      "step": 7080
+    },
+    {
+      "epoch": 3.51284597244632,
+      "grad_norm": 0.07459512778773453,
+      "learning_rate": 3.588431066588832e-07,
+      "loss": 0.4399,
+      "step": 7081
+    },
+    {
+      "epoch": 3.5133424351495592,
+      "grad_norm": 0.0714307806848993,
+      "learning_rate": 3.58116306626925e-07,
+      "loss": 0.4329,
+      "step": 7082
+    },
+    {
+      "epoch": 3.513838897852799,
+      "grad_norm": 0.07363200094283813,
+      "learning_rate": 3.5739021603246104e-07,
+      "loss": 0.4387,
+      "step": 7083
+    },
+    {
+      "epoch": 3.5143353605560383,
+      "grad_norm": 0.07259395908343577,
+      "learning_rate": 3.56664834986466e-07,
+      "loss": 0.4516,
+      "step": 7084
+    },
+    {
+      "epoch": 3.5148318232592777,
+      "grad_norm": 0.07184213899212576,
+      "learning_rate": 3.5594016359979886e-07,
+      "loss": 0.4273,
+      "step": 7085
+    },
+    {
+      "epoch": 3.515328285962517,
+      "grad_norm": 0.07294685594344688,
+      "learning_rate": 3.552162019832167e-07,
+      "loss": 0.4387,
+      "step": 7086
+    },
+    {
+      "epoch": 3.5158247486657563,
+      "grad_norm": 0.07441906850778905,
+      "learning_rate": 3.5449295024736374e-07,
+      "loss": 0.4337,
+      "step": 7087
+    },
+    {
+      "epoch": 3.5163212113689957,
+      "grad_norm": 0.07368084403298997,
+      "learning_rate": 3.5377040850277935e-07,
+      "loss": 0.4512,
+      "step": 7088
+    },
+    {
+      "epoch": 3.5168176740722354,
+      "grad_norm": 0.0713595992730785,
+      "learning_rate": 3.5304857685989125e-07,
+      "loss": 0.4433,
+      "step": 7089
+    },
+    {
+      "epoch": 3.5173141367754748,
+      "grad_norm": 0.071877961919271,
+      "learning_rate": 3.52327455429019e-07,
+      "loss": 0.4355,
+      "step": 7090
+    },
+    {
+      "epoch": 3.517810599478714,
+      "grad_norm": 0.07338459914123707,
+      "learning_rate": 3.5160704432037616e-07,
+      "loss": 0.4804,
+      "step": 7091
+    },
+    {
+      "epoch": 3.5183070621819534,
+      "grad_norm": 0.06996501674731023,
+      "learning_rate": 3.5088734364406573e-07,
+      "loss": 0.4079,
+      "step": 7092
+    },
+    {
+      "epoch": 3.518803524885193,
+      "grad_norm": 0.07605308956893601,
+      "learning_rate": 3.5016835351008083e-07,
+      "loss": 0.4442,
+      "step": 7093
+    },
+    {
+      "epoch": 3.5192999875884325,
+      "grad_norm": 0.0721478352591386,
+      "learning_rate": 3.4945007402830964e-07,
+      "loss": 0.4262,
+      "step": 7094
+    },
+    {
+      "epoch": 3.519796450291672,
+      "grad_norm": 0.07622812677268571,
+      "learning_rate": 3.487325053085283e-07,
+      "loss": 0.465,
+      "step": 7095
+    },
+    {
+      "epoch": 3.520292912994911,
+      "grad_norm": 0.07288424413448973,
+      "learning_rate": 3.480156474604063e-07,
+      "loss": 0.4483,
+      "step": 7096
+    },
+    {
+      "epoch": 3.5207893756981505,
+      "grad_norm": 0.07327517647672152,
+      "learning_rate": 3.472995005935037e-07,
+      "loss": 0.4246,
+      "step": 7097
+    },
+    {
+      "epoch": 3.52128583840139,
+      "grad_norm": 0.07373687082587846,
+      "learning_rate": 3.465840648172719e-07,
+      "loss": 0.4583,
+      "step": 7098
+    },
+    {
+      "epoch": 3.5217823011046296,
+      "grad_norm": 0.07400232258061422,
+      "learning_rate": 3.45869340241054e-07,
+      "loss": 0.4386,
+      "step": 7099
+    },
+    {
+      "epoch": 3.522278763807869,
+      "grad_norm": 0.07429626916320163,
+      "learning_rate": 3.451553269740848e-07,
+      "loss": 0.4362,
+      "step": 7100
+    },
+    {
+      "epoch": 3.5227752265111083,
+      "grad_norm": 0.07327933612968662,
+      "learning_rate": 3.4444202512548874e-07,
+      "loss": 0.4365,
+      "step": 7101
+    },
+    {
+      "epoch": 3.5232716892143476,
+      "grad_norm": 0.07178454799022094,
+      "learning_rate": 3.437294348042819e-07,
+      "loss": 0.4208,
+      "step": 7102
+    },
+    {
+      "epoch": 3.5237681519175874,
+      "grad_norm": 0.07363781565375832,
+      "learning_rate": 3.4301755611937435e-07,
+      "loss": 0.419,
+      "step": 7103
+    },
+    {
+      "epoch": 3.5242646146208267,
+      "grad_norm": 0.07279288571151912,
+      "learning_rate": 3.423063891795647e-07,
+      "loss": 0.4364,
+      "step": 7104
+    },
+    {
+      "epoch": 3.524761077324066,
+      "grad_norm": 0.07073245127069415,
+      "learning_rate": 3.415959340935415e-07,
+      "loss": 0.4227,
+      "step": 7105
+    },
+    {
+      "epoch": 3.5252575400273054,
+      "grad_norm": 0.07027468642692776,
+      "learning_rate": 3.408861909698896e-07,
+      "loss": 0.4351,
+      "step": 7106
+    },
+    {
+      "epoch": 3.5257540027305447,
+      "grad_norm": 0.07470813023985937,
+      "learning_rate": 3.401771599170789e-07,
+      "loss": 0.4514,
+      "step": 7107
+    },
+    {
+      "epoch": 3.526250465433784,
+      "grad_norm": 0.07204009719364936,
+      "learning_rate": 3.3946884104347543e-07,
+      "loss": 0.438,
+      "step": 7108
+    },
+    {
+      "epoch": 3.526746928137024,
+      "grad_norm": 0.07185790187129391,
+      "learning_rate": 3.3876123445733376e-07,
+      "loss": 0.4442,
+      "step": 7109
+    },
+    {
+      "epoch": 3.527243390840263,
+      "grad_norm": 0.07484063183218079,
+      "learning_rate": 3.380543402667996e-07,
+      "loss": 0.4577,
+      "step": 7110
+    },
+    {
+      "epoch": 3.5277398535435025,
+      "grad_norm": 0.07228052121532155,
+      "learning_rate": 3.3734815857991155e-07,
+      "loss": 0.4441,
+      "step": 7111
+    },
+    {
+      "epoch": 3.528236316246742,
+      "grad_norm": 0.07094925816469587,
+      "learning_rate": 3.366426895045966e-07,
+      "loss": 0.4406,
+      "step": 7112
+    },
+    {
+      "epoch": 3.5287327789499816,
+      "grad_norm": 0.06883784070072635,
+      "learning_rate": 3.359379331486762e-07,
+      "loss": 0.424,
+      "step": 7113
+    },
+    {
+      "epoch": 3.529229241653221,
+      "grad_norm": 0.07220676053662756,
+      "learning_rate": 3.352338896198598e-07,
+      "loss": 0.4723,
+      "step": 7114
+    },
+    {
+      "epoch": 3.5297257043564603,
+      "grad_norm": 0.07167224348889963,
+      "learning_rate": 3.3453055902574915e-07,
+      "loss": 0.4114,
+      "step": 7115
+    },
+    {
+      "epoch": 3.5302221670596996,
+      "grad_norm": 0.07542897494207551,
+      "learning_rate": 3.3382794147383877e-07,
+      "loss": 0.4556,
+      "step": 7116
+    },
+    {
+      "epoch": 3.530718629762939,
+      "grad_norm": 0.07316883893388977,
+      "learning_rate": 3.3312603707151006e-07,
+      "loss": 0.4451,
+      "step": 7117
+    },
+    {
+      "epoch": 3.5312150924661783,
+      "grad_norm": 0.07512076327321777,
+      "learning_rate": 3.324248459260393e-07,
+      "loss": 0.4601,
+      "step": 7118
+    },
+    {
+      "epoch": 3.531711555169418,
+      "grad_norm": 0.0720588479676324,
+      "learning_rate": 3.317243681445914e-07,
+      "loss": 0.427,
+      "step": 7119
+    },
+    {
+      "epoch": 3.5322080178726574,
+      "grad_norm": 0.07030183758489095,
+      "learning_rate": 3.310246038342246e-07,
+      "loss": 0.4007,
+      "step": 7120
+    },
+    {
+      "epoch": 3.5327044805758967,
+      "grad_norm": 0.07426185203798406,
+      "learning_rate": 3.3032555310188566e-07,
+      "loss": 0.4381,
+      "step": 7121
+    },
+    {
+      "epoch": 3.533200943279136,
+      "grad_norm": 0.07266412590830498,
+      "learning_rate": 3.2962721605441227e-07,
+      "loss": 0.4422,
+      "step": 7122
+    },
+    {
+      "epoch": 3.533697405982376,
+      "grad_norm": 0.07337165886257081,
+      "learning_rate": 3.289295927985364e-07,
+      "loss": 0.4564,
+      "step": 7123
+    },
+    {
+      "epoch": 3.534193868685615,
+      "grad_norm": 0.07660405077180722,
+      "learning_rate": 3.282326834408761e-07,
+      "loss": 0.4615,
+      "step": 7124
+    },
+    {
+      "epoch": 3.5346903313888545,
+      "grad_norm": 0.07351730627486232,
+      "learning_rate": 3.2753648808794505e-07,
+      "loss": 0.4277,
+      "step": 7125
+    },
+    {
+      "epoch": 3.535186794092094,
+      "grad_norm": 0.0716057122840003,
+      "learning_rate": 3.268410068461447e-07,
+      "loss": 0.4415,
+      "step": 7126
+    },
+    {
+      "epoch": 3.535683256795333,
+      "grad_norm": 0.07415164949108556,
+      "learning_rate": 3.261462398217674e-07,
+      "loss": 0.4648,
+      "step": 7127
+    },
+    {
+      "epoch": 3.5361797194985725,
+      "grad_norm": 0.06981491738014588,
+      "learning_rate": 3.254521871209981e-07,
+      "loss": 0.4273,
+      "step": 7128
+    },
+    {
+      "epoch": 3.5366761822018122,
+      "grad_norm": 0.07090923551472,
+      "learning_rate": 3.247588488499115e-07,
+      "loss": 0.417,
+      "step": 7129
+    },
+    {
+      "epoch": 3.5371726449050516,
+      "grad_norm": 0.07266912902613439,
+      "learning_rate": 3.240662251144727e-07,
+      "loss": 0.4403,
+      "step": 7130
+    },
+    {
+      "epoch": 3.537669107608291,
+      "grad_norm": 0.07248447075293059,
+      "learning_rate": 3.233743160205388e-07,
+      "loss": 0.4516,
+      "step": 7131
+    },
+    {
+      "epoch": 3.5381655703115302,
+      "grad_norm": 0.0741972809931787,
+      "learning_rate": 3.2268312167385687e-07,
+      "loss": 0.4377,
+      "step": 7132
+    },
+    {
+      "epoch": 3.53866203301477,
+      "grad_norm": 0.07189591452664673,
+      "learning_rate": 3.2199264218006453e-07,
+      "loss": 0.4497,
+      "step": 7133
+    },
+    {
+      "epoch": 3.5391584957180093,
+      "grad_norm": 0.07147144486299516,
+      "learning_rate": 3.213028776446903e-07,
+      "loss": 0.4289,
+      "step": 7134
+    },
+    {
+      "epoch": 3.5396549584212487,
+      "grad_norm": 0.07361435471493433,
+      "learning_rate": 3.206138281731547e-07,
+      "loss": 0.4726,
+      "step": 7135
+    },
+    {
+      "epoch": 3.540151421124488,
+      "grad_norm": 0.07246314148121354,
+      "learning_rate": 3.1992549387076685e-07,
+      "loss": 0.4396,
+      "step": 7136
+    },
+    {
+      "epoch": 3.5406478838277273,
+      "grad_norm": 0.072942581284602,
+      "learning_rate": 3.1923787484272717e-07,
+      "loss": 0.4445,
+      "step": 7137
+    },
+    {
+      "epoch": 3.5411443465309667,
+      "grad_norm": 0.07465515413450105,
+      "learning_rate": 3.1855097119412924e-07,
+      "loss": 0.4613,
+      "step": 7138
+    },
+    {
+      "epoch": 3.5416408092342064,
+      "grad_norm": 0.07283341180650725,
+      "learning_rate": 3.1786478302995305e-07,
+      "loss": 0.4536,
+      "step": 7139
+    },
+    {
+      "epoch": 3.5421372719374458,
+      "grad_norm": 0.07424965634235413,
+      "learning_rate": 3.17179310455073e-07,
+      "loss": 0.4871,
+      "step": 7140
+    },
+    {
+      "epoch": 3.542633734640685,
+      "grad_norm": 0.07378328684390752,
+      "learning_rate": 3.164945535742525e-07,
+      "loss": 0.489,
+      "step": 7141
+    },
+    {
+      "epoch": 3.5431301973439244,
+      "grad_norm": 0.07472723192771431,
+      "learning_rate": 3.15810512492144e-07,
+      "loss": 0.4809,
+      "step": 7142
+    },
+    {
+      "epoch": 3.543626660047164,
+      "grad_norm": 0.07196375675496666,
+      "learning_rate": 3.151271873132944e-07,
+      "loss": 0.466,
+      "step": 7143
+    },
+    {
+      "epoch": 3.5441231227504035,
+      "grad_norm": 0.07180812934665125,
+      "learning_rate": 3.1444457814213736e-07,
+      "loss": 0.4587,
+      "step": 7144
+    },
+    {
+      "epoch": 3.544619585453643,
+      "grad_norm": 0.07119979957545244,
+      "learning_rate": 3.13762685083e-07,
+      "loss": 0.4392,
+      "step": 7145
+    },
+    {
+      "epoch": 3.545116048156882,
+      "grad_norm": 0.07433975376299118,
+      "learning_rate": 3.1308150824009785e-07,
+      "loss": 0.4415,
+      "step": 7146
+    },
+    {
+      "epoch": 3.5456125108601215,
+      "grad_norm": 0.07140407776914781,
+      "learning_rate": 3.1240104771753765e-07,
+      "loss": 0.4345,
+      "step": 7147
+    },
+    {
+      "epoch": 3.546108973563361,
+      "grad_norm": 0.07269789822773806,
+      "learning_rate": 3.1172130361931894e-07,
+      "loss": 0.4347,
+      "step": 7148
+    },
+    {
+      "epoch": 3.5466054362666006,
+      "grad_norm": 0.07110519917373652,
+      "learning_rate": 3.1104227604932644e-07,
+      "loss": 0.4149,
+      "step": 7149
+    },
+    {
+      "epoch": 3.54710189896984,
+      "grad_norm": 0.07206882812848532,
+      "learning_rate": 3.10363965111341e-07,
+      "loss": 0.4277,
+      "step": 7150
+    },
+    {
+      "epoch": 3.5475983616730793,
+      "grad_norm": 0.0732811113576226,
+      "learning_rate": 3.096863709090303e-07,
+      "loss": 0.4377,
+      "step": 7151
+    },
+    {
+      "epoch": 3.5480948243763186,
+      "grad_norm": 0.0750770669128101,
+      "learning_rate": 3.0900949354595535e-07,
+      "loss": 0.4593,
+      "step": 7152
+    },
+    {
+      "epoch": 3.5485912870795584,
+      "grad_norm": 0.07486470898140467,
+      "learning_rate": 3.0833333312556446e-07,
+      "loss": 0.4725,
+      "step": 7153
+    },
+    {
+      "epoch": 3.5490877497827977,
+      "grad_norm": 0.07116823477747528,
+      "learning_rate": 3.076578897511978e-07,
+      "loss": 0.4339,
+      "step": 7154
+    },
+    {
+      "epoch": 3.549584212486037,
+      "grad_norm": 0.07283679758350045,
+      "learning_rate": 3.069831635260878e-07,
+      "loss": 0.4172,
+      "step": 7155
+    },
+    {
+      "epoch": 3.5500806751892764,
+      "grad_norm": 0.07304141500792913,
+      "learning_rate": 3.0630915455335365e-07,
+      "loss": 0.4604,
+      "step": 7156
+    },
+    {
+      "epoch": 3.5505771378925157,
+      "grad_norm": 0.07196716548331811,
+      "learning_rate": 3.0563586293600846e-07,
+      "loss": 0.4383,
+      "step": 7157
+    },
+    {
+      "epoch": 3.551073600595755,
+      "grad_norm": 0.07490926531623762,
+      "learning_rate": 3.049632887769527e-07,
+      "loss": 0.4298,
+      "step": 7158
+    },
+    {
+      "epoch": 3.551570063298995,
+      "grad_norm": 0.07673976524598099,
+      "learning_rate": 3.0429143217897863e-07,
+      "loss": 0.484,
+      "step": 7159
+    },
+    {
+      "epoch": 3.552066526002234,
+      "grad_norm": 0.0718037382316523,
+      "learning_rate": 3.0362029324477015e-07,
+      "loss": 0.4102,
+      "step": 7160
+    },
+    {
+      "epoch": 3.5525629887054735,
+      "grad_norm": 0.07090278460304152,
+      "learning_rate": 3.0294987207689805e-07,
+      "loss": 0.443,
+      "step": 7161
+    },
+    {
+      "epoch": 3.553059451408713,
+      "grad_norm": 0.0701796762434687,
+      "learning_rate": 3.022801687778276e-07,
+      "loss": 0.4085,
+      "step": 7162
+    },
+    {
+      "epoch": 3.5535559141119526,
+      "grad_norm": 0.07463259506007328,
+      "learning_rate": 3.0161118344991083e-07,
+      "loss": 0.4718,
+      "step": 7163
+    },
+    {
+      "epoch": 3.554052376815192,
+      "grad_norm": 0.07268649175220862,
+      "learning_rate": 3.0094291619539084e-07,
+      "loss": 0.4477,
+      "step": 7164
+    },
+    {
+      "epoch": 3.5545488395184313,
+      "grad_norm": 0.07293852406475182,
+      "learning_rate": 3.0027536711640436e-07,
+      "loss": 0.4708,
+      "step": 7165
+    },
+    {
+      "epoch": 3.5550453022216706,
+      "grad_norm": 0.07040439954277569,
+      "learning_rate": 2.99608536314972e-07,
+      "loss": 0.4195,
+      "step": 7166
+    },
+    {
+      "epoch": 3.55554176492491,
+      "grad_norm": 0.07409754906539101,
+      "learning_rate": 2.9894242389301053e-07,
+      "loss": 0.4622,
+      "step": 7167
+    },
+    {
+      "epoch": 3.5560382276281493,
+      "grad_norm": 0.07423989038436342,
+      "learning_rate": 2.982770299523241e-07,
+      "loss": 0.4518,
+      "step": 7168
+    },
+    {
+      "epoch": 3.5565346903313886,
+      "grad_norm": 0.07487884963289175,
+      "learning_rate": 2.976123545946064e-07,
+      "loss": 0.4503,
+      "step": 7169
+    },
+    {
+      "epoch": 3.5570311530346284,
+      "grad_norm": 0.07367120972684456,
+      "learning_rate": 2.969483979214438e-07,
+      "loss": 0.4417,
+      "step": 7170
+    },
+    {
+      "epoch": 3.5575276157378677,
+      "grad_norm": 0.07103468441793306,
+      "learning_rate": 2.9628516003430974e-07,
+      "loss": 0.4461,
+      "step": 7171
+    },
+    {
+      "epoch": 3.558024078441107,
+      "grad_norm": 0.07187761985962852,
+      "learning_rate": 2.9562264103457196e-07,
+      "loss": 0.4685,
+      "step": 7172
+    },
+    {
+      "epoch": 3.558520541144347,
+      "grad_norm": 0.07424526932350758,
+      "learning_rate": 2.9496084102348443e-07,
+      "loss": 0.4502,
+      "step": 7173
+    },
+    {
+      "epoch": 3.559017003847586,
+      "grad_norm": 0.07072411570855557,
+      "learning_rate": 2.942997601021924e-07,
+      "loss": 0.4137,
+      "step": 7174
+    },
+    {
+      "epoch": 3.5595134665508255,
+      "grad_norm": 0.0724840565932664,
+      "learning_rate": 2.936393983717323e-07,
+      "loss": 0.4323,
+      "step": 7175
+    },
+    {
+      "epoch": 3.560009929254065,
+      "grad_norm": 0.07234915777182487,
+      "learning_rate": 2.929797559330283e-07,
+      "loss": 0.4621,
+      "step": 7176
+    },
+    {
+      "epoch": 3.560506391957304,
+      "grad_norm": 0.07094228491783852,
+      "learning_rate": 2.9232083288689814e-07,
+      "loss": 0.4311,
+      "step": 7177
+    },
+    {
+      "epoch": 3.5610028546605434,
+      "grad_norm": 0.07285222821523794,
+      "learning_rate": 2.916626293340474e-07,
+      "loss": 0.4496,
+      "step": 7178
+    },
+    {
+      "epoch": 3.561499317363783,
+      "grad_norm": 0.073098707674849,
+      "learning_rate": 2.910051453750701e-07,
+      "loss": 0.4304,
+      "step": 7179
+    },
+    {
+      "epoch": 3.5619957800670226,
+      "grad_norm": 0.07327812181997499,
+      "learning_rate": 2.9034838111045406e-07,
+      "loss": 0.4383,
+      "step": 7180
+    },
+    {
+      "epoch": 3.562492242770262,
+      "grad_norm": 0.07343319803974091,
+      "learning_rate": 2.896923366405746e-07,
+      "loss": 0.4452,
+      "step": 7181
+    },
+    {
+      "epoch": 3.562988705473501,
+      "grad_norm": 0.07205469997223532,
+      "learning_rate": 2.890370120656971e-07,
+      "loss": 0.419,
+      "step": 7182
+    },
+    {
+      "epoch": 3.563485168176741,
+      "grad_norm": 0.07112166085797217,
+      "learning_rate": 2.8838240748597757e-07,
+      "loss": 0.4192,
+      "step": 7183
+    },
+    {
+      "epoch": 3.5639816308799803,
+      "grad_norm": 0.0724456402776545,
+      "learning_rate": 2.87728523001462e-07,
+      "loss": 0.4247,
+      "step": 7184
+    },
+    {
+      "epoch": 3.5644780935832197,
+      "grad_norm": 0.07155775449072983,
+      "learning_rate": 2.8707535871208667e-07,
+      "loss": 0.4436,
+      "step": 7185
+    },
+    {
+      "epoch": 3.564974556286459,
+      "grad_norm": 0.0722910340184464,
+      "learning_rate": 2.864229147176761e-07,
+      "loss": 0.4576,
+      "step": 7186
+    },
+    {
+      "epoch": 3.5654710189896983,
+      "grad_norm": 0.07227263041241058,
+      "learning_rate": 2.8577119111794725e-07,
+      "loss": 0.4493,
+      "step": 7187
+    },
+    {
+      "epoch": 3.5659674816929376,
+      "grad_norm": 0.0742062560955642,
+      "learning_rate": 2.851201880125043e-07,
+      "loss": 0.4319,
+      "step": 7188
+    },
+    {
+      "epoch": 3.566463944396177,
+      "grad_norm": 0.07020998189715937,
+      "learning_rate": 2.8446990550084373e-07,
+      "loss": 0.4276,
+      "step": 7189
+    },
+    {
+      "epoch": 3.5669604070994168,
+      "grad_norm": 0.0719695731390821,
+      "learning_rate": 2.8382034368235003e-07,
+      "loss": 0.4354,
+      "step": 7190
+    },
+    {
+      "epoch": 3.567456869802656,
+      "grad_norm": 0.07144612364466402,
+      "learning_rate": 2.8317150265629813e-07,
+      "loss": 0.4308,
+      "step": 7191
+    },
+    {
+      "epoch": 3.5679533325058954,
+      "grad_norm": 0.07389313275934502,
+      "learning_rate": 2.825233825218543e-07,
+      "loss": 0.4377,
+      "step": 7192
+    },
+    {
+      "epoch": 3.568449795209135,
+      "grad_norm": 0.07368487689346809,
+      "learning_rate": 2.818759833780721e-07,
+      "loss": 0.456,
+      "step": 7193
+    },
+    {
+      "epoch": 3.5689462579123745,
+      "grad_norm": 0.07254773669872457,
+      "learning_rate": 2.8122930532389683e-07,
+      "loss": 0.4253,
+      "step": 7194
+    },
+    {
+      "epoch": 3.569442720615614,
+      "grad_norm": 0.07141227637028748,
+      "learning_rate": 2.8058334845816214e-07,
+      "loss": 0.4265,
+      "step": 7195
+    },
+    {
+      "epoch": 3.569939183318853,
+      "grad_norm": 0.07158291382565189,
+      "learning_rate": 2.799381128795925e-07,
+      "loss": 0.4584,
+      "step": 7196
+    },
+    {
+      "epoch": 3.5704356460220925,
+      "grad_norm": 0.07470124461409806,
+      "learning_rate": 2.7929359868680283e-07,
+      "loss": 0.464,
+      "step": 7197
+    },
+    {
+      "epoch": 3.570932108725332,
+      "grad_norm": 0.07348846639386283,
+      "learning_rate": 2.7864980597829495e-07,
+      "loss": 0.46,
+      "step": 7198
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.07182543905438588,
+      "learning_rate": 2.7800673485246343e-07,
+      "loss": 0.462,
+      "step": 7199
+    },
+    {
+      "epoch": 3.571925034131811,
+      "grad_norm": 0.07144046247551303,
+      "learning_rate": 2.7736438540759026e-07,
+      "loss": 0.44,
+      "step": 7200
+    },
+    {
+      "epoch": 3.5724214968350503,
+      "grad_norm": 0.07192453287481038,
+      "learning_rate": 2.7672275774184967e-07,
+      "loss": 0.4412,
+      "step": 7201
+    },
+    {
+      "epoch": 3.5729179595382896,
+      "grad_norm": 0.0729370214192708,
+      "learning_rate": 2.760818519533037e-07,
+      "loss": 0.4589,
+      "step": 7202
+    },
+    {
+      "epoch": 3.573414422241529,
+      "grad_norm": 0.07311855870542439,
+      "learning_rate": 2.754416681399041e-07,
+      "loss": 0.4523,
+      "step": 7203
+    },
+    {
+      "epoch": 3.5739108849447687,
+      "grad_norm": 0.07161270410308067,
+      "learning_rate": 2.748022063994932e-07,
+      "loss": 0.4334,
+      "step": 7204
+    },
+    {
+      "epoch": 3.574407347648008,
+      "grad_norm": 0.0741000290693234,
+      "learning_rate": 2.7416346682980264e-07,
+      "loss": 0.4461,
+      "step": 7205
+    },
+    {
+      "epoch": 3.5749038103512474,
+      "grad_norm": 0.07146630300410399,
+      "learning_rate": 2.7352544952845226e-07,
+      "loss": 0.4439,
+      "step": 7206
+    },
+    {
+      "epoch": 3.5754002730544867,
+      "grad_norm": 0.07101804224421061,
+      "learning_rate": 2.728881545929546e-07,
+      "loss": 0.4301,
+      "step": 7207
+    },
+    {
+      "epoch": 3.575896735757726,
+      "grad_norm": 0.07617485536695313,
+      "learning_rate": 2.722515821207078e-07,
+      "loss": 0.4331,
+      "step": 7208
+    },
+    {
+      "epoch": 3.5763931984609654,
+      "grad_norm": 0.07175749391331662,
+      "learning_rate": 2.716157322090041e-07,
+      "loss": 0.4423,
+      "step": 7209
+    },
+    {
+      "epoch": 3.576889661164205,
+      "grad_norm": 0.07385981567472476,
+      "learning_rate": 2.709806049550218e-07,
+      "loss": 0.4132,
+      "step": 7210
+    },
+    {
+      "epoch": 3.5773861238674445,
+      "grad_norm": 0.07213037434962473,
+      "learning_rate": 2.7034620045582937e-07,
+      "loss": 0.416,
+      "step": 7211
+    },
+    {
+      "epoch": 3.577882586570684,
+      "grad_norm": 0.07190591120086846,
+      "learning_rate": 2.6971251880838657e-07,
+      "loss": 0.4106,
+      "step": 7212
+    },
+    {
+      "epoch": 3.578379049273923,
+      "grad_norm": 0.07414440862138535,
+      "learning_rate": 2.6907956010954086e-07,
+      "loss": 0.4741,
+      "step": 7213
+    },
+    {
+      "epoch": 3.578875511977163,
+      "grad_norm": 0.07271101396039249,
+      "learning_rate": 2.684473244560298e-07,
+      "loss": 0.4323,
+      "step": 7214
+    },
+    {
+      "epoch": 3.5793719746804022,
+      "grad_norm": 0.07000619994876894,
+      "learning_rate": 2.678158119444796e-07,
+      "loss": 0.4145,
+      "step": 7215
+    },
+    {
+      "epoch": 3.5798684373836416,
+      "grad_norm": 0.0720654558912617,
+      "learning_rate": 2.6718502267140844e-07,
+      "loss": 0.4412,
+      "step": 7216
+    },
+    {
+      "epoch": 3.580364900086881,
+      "grad_norm": 0.07265652672638535,
+      "learning_rate": 2.6655495673322205e-07,
+      "loss": 0.4383,
+      "step": 7217
+    },
+    {
+      "epoch": 3.5808613627901202,
+      "grad_norm": 0.07569100909955953,
+      "learning_rate": 2.659256142262145e-07,
+      "loss": 0.4671,
+      "step": 7218
+    },
+    {
+      "epoch": 3.5813578254933596,
+      "grad_norm": 0.0734469350012236,
+      "learning_rate": 2.652969952465728e-07,
+      "loss": 0.411,
+      "step": 7219
+    },
+    {
+      "epoch": 3.5818542881965993,
+      "grad_norm": 0.07126251992525097,
+      "learning_rate": 2.646690998903689e-07,
+      "loss": 0.4144,
+      "step": 7220
+    },
+    {
+      "epoch": 3.5823507508998387,
+      "grad_norm": 0.07256713526228342,
+      "learning_rate": 2.640419282535689e-07,
+      "loss": 0.4074,
+      "step": 7221
+    },
+    {
+      "epoch": 3.582847213603078,
+      "grad_norm": 0.0753410981599593,
+      "learning_rate": 2.634154804320249e-07,
+      "loss": 0.4538,
+      "step": 7222
+    },
+    {
+      "epoch": 3.5833436763063173,
+      "grad_norm": 0.07215766584838977,
+      "learning_rate": 2.6278975652147875e-07,
+      "loss": 0.438,
+      "step": 7223
+    },
+    {
+      "epoch": 3.583840139009557,
+      "grad_norm": 0.07323954472644308,
+      "learning_rate": 2.6216475661756336e-07,
+      "loss": 0.4397,
+      "step": 7224
+    },
+    {
+      "epoch": 3.5843366017127964,
+      "grad_norm": 0.0764315762368446,
+      "learning_rate": 2.6154048081579897e-07,
+      "loss": 0.4896,
+      "step": 7225
+    },
+    {
+      "epoch": 3.5848330644160358,
+      "grad_norm": 0.07373114519490154,
+      "learning_rate": 2.6091692921159705e-07,
+      "loss": 0.4582,
+      "step": 7226
+    },
+    {
+      "epoch": 3.585329527119275,
+      "grad_norm": 0.07338166822291779,
+      "learning_rate": 2.602941019002575e-07,
+      "loss": 0.4599,
+      "step": 7227
+    },
+    {
+      "epoch": 3.5858259898225144,
+      "grad_norm": 0.07156011806994722,
+      "learning_rate": 2.5967199897696806e-07,
+      "loss": 0.4445,
+      "step": 7228
+    },
+    {
+      "epoch": 3.5863224525257538,
+      "grad_norm": 0.07461471442375366,
+      "learning_rate": 2.5905062053680984e-07,
+      "loss": 0.4621,
+      "step": 7229
+    },
+    {
+      "epoch": 3.5868189152289935,
+      "grad_norm": 0.07397449325671485,
+      "learning_rate": 2.584299666747475e-07,
+      "loss": 0.4622,
+      "step": 7230
+    },
+    {
+      "epoch": 3.587315377932233,
+      "grad_norm": 0.0733072439245948,
+      "learning_rate": 2.578100374856402e-07,
+      "loss": 0.4317,
+      "step": 7231
+    },
+    {
+      "epoch": 3.587811840635472,
+      "grad_norm": 0.07337499878496877,
+      "learning_rate": 2.571908330642325e-07,
+      "loss": 0.4868,
+      "step": 7232
+    },
+    {
+      "epoch": 3.5883083033387115,
+      "grad_norm": 0.0713814077159316,
+      "learning_rate": 2.5657235350516175e-07,
+      "loss": 0.4651,
+      "step": 7233
+    },
+    {
+      "epoch": 3.5888047660419513,
+      "grad_norm": 0.07316923475034787,
+      "learning_rate": 2.5595459890295106e-07,
+      "loss": 0.4265,
+      "step": 7234
+    },
+    {
+      "epoch": 3.5893012287451906,
+      "grad_norm": 0.07245152311967211,
+      "learning_rate": 2.553375693520149e-07,
+      "loss": 0.4354,
+      "step": 7235
+    },
+    {
+      "epoch": 3.58979769144843,
+      "grad_norm": 0.07337580274838068,
+      "learning_rate": 2.547212649466568e-07,
+      "loss": 0.4466,
+      "step": 7236
+    },
+    {
+      "epoch": 3.5902941541516693,
+      "grad_norm": 0.07299768932189318,
+      "learning_rate": 2.541056857810681e-07,
+      "loss": 0.4236,
+      "step": 7237
+    },
+    {
+      "epoch": 3.5907906168549086,
+      "grad_norm": 0.07410300395331985,
+      "learning_rate": 2.534908319493312e-07,
+      "loss": 0.4335,
+      "step": 7238
+    },
+    {
+      "epoch": 3.591287079558148,
+      "grad_norm": 0.07202380812272681,
+      "learning_rate": 2.528767035454166e-07,
+      "loss": 0.4323,
+      "step": 7239
+    },
+    {
+      "epoch": 3.5917835422613877,
+      "grad_norm": 0.07133969448762374,
+      "learning_rate": 2.522633006631825e-07,
+      "loss": 0.4409,
+      "step": 7240
+    },
+    {
+      "epoch": 3.592280004964627,
+      "grad_norm": 0.07511748521712497,
+      "learning_rate": 2.516506233963795e-07,
+      "loss": 0.4438,
+      "step": 7241
+    },
+    {
+      "epoch": 3.5927764676678664,
+      "grad_norm": 0.07436474096664406,
+      "learning_rate": 2.5103867183864425e-07,
+      "loss": 0.433,
+      "step": 7242
+    },
+    {
+      "epoch": 3.5932729303711057,
+      "grad_norm": 0.07322180245682137,
+      "learning_rate": 2.504274460835038e-07,
+      "loss": 0.409,
+      "step": 7243
+    },
+    {
+      "epoch": 3.5937693930743455,
+      "grad_norm": 0.07174823455141795,
+      "learning_rate": 2.4981694622437546e-07,
+      "loss": 0.4248,
+      "step": 7244
+    },
+    {
+      "epoch": 3.594265855777585,
+      "grad_norm": 0.07219829398726958,
+      "learning_rate": 2.492071723545636e-07,
+      "loss": 0.4469,
+      "step": 7245
+    },
+    {
+      "epoch": 3.594762318480824,
+      "grad_norm": 0.07416423979824675,
+      "learning_rate": 2.4859812456726195e-07,
+      "loss": 0.4398,
+      "step": 7246
+    },
+    {
+      "epoch": 3.5952587811840635,
+      "grad_norm": 0.072503482237734,
+      "learning_rate": 2.479898029555533e-07,
+      "loss": 0.4255,
+      "step": 7247
+    },
+    {
+      "epoch": 3.595755243887303,
+      "grad_norm": 0.07210104204463237,
+      "learning_rate": 2.473822076124116e-07,
+      "loss": 0.4718,
+      "step": 7248
+    },
+    {
+      "epoch": 3.596251706590542,
+      "grad_norm": 0.07646935551434451,
+      "learning_rate": 2.4677533863069705e-07,
+      "loss": 0.4709,
+      "step": 7249
+    },
+    {
+      "epoch": 3.596748169293782,
+      "grad_norm": 0.07040667756464702,
+      "learning_rate": 2.461691961031587e-07,
+      "loss": 0.4347,
+      "step": 7250
+    },
+    {
+      "epoch": 3.5972446319970213,
+      "grad_norm": 0.07550119093373847,
+      "learning_rate": 2.4556378012243807e-07,
+      "loss": 0.4966,
+      "step": 7251
+    },
+    {
+      "epoch": 3.5977410947002606,
+      "grad_norm": 0.07199950299189856,
+      "learning_rate": 2.449590907810612e-07,
+      "loss": 0.4439,
+      "step": 7252
+    },
+    {
+      "epoch": 3.5982375574035,
+      "grad_norm": 0.0702318610232438,
+      "learning_rate": 2.4435512817144625e-07,
+      "loss": 0.4417,
+      "step": 7253
+    },
+    {
+      "epoch": 3.5987340201067397,
+      "grad_norm": 0.07549606737474215,
+      "learning_rate": 2.4375189238589945e-07,
+      "loss": 0.4632,
+      "step": 7254
+    },
+    {
+      "epoch": 3.599230482809979,
+      "grad_norm": 0.07479633717728829,
+      "learning_rate": 2.4314938351661486e-07,
+      "loss": 0.4317,
+      "step": 7255
+    },
+    {
+      "epoch": 3.5997269455132184,
+      "grad_norm": 0.07362432336615526,
+      "learning_rate": 2.425476016556766e-07,
+      "loss": 0.47,
+      "step": 7256
+    },
+    {
+      "epoch": 3.6002234082164577,
+      "grad_norm": 0.07466644313513754,
+      "learning_rate": 2.4194654689505716e-07,
+      "loss": 0.4256,
+      "step": 7257
+    },
+    {
+      "epoch": 3.600719870919697,
+      "grad_norm": 0.07352553344538089,
+      "learning_rate": 2.4134621932661916e-07,
+      "loss": 0.4307,
+      "step": 7258
+    },
+    {
+      "epoch": 3.6012163336229364,
+      "grad_norm": 0.07137696369934685,
+      "learning_rate": 2.407466190421126e-07,
+      "loss": 0.4212,
+      "step": 7259
+    },
+    {
+      "epoch": 3.601712796326176,
+      "grad_norm": 0.07388793112030832,
+      "learning_rate": 2.4014774613317525e-07,
+      "loss": 0.4695,
+      "step": 7260
+    },
+    {
+      "epoch": 3.6022092590294155,
+      "grad_norm": 0.07553122083868365,
+      "learning_rate": 2.3954960069133837e-07,
+      "loss": 0.4511,
+      "step": 7261
+    },
+    {
+      "epoch": 3.602705721732655,
+      "grad_norm": 0.07526863616354296,
+      "learning_rate": 2.3895218280801547e-07,
+      "loss": 0.432,
+      "step": 7262
+    },
+    {
+      "epoch": 3.603202184435894,
+      "grad_norm": 0.07314299399999061,
+      "learning_rate": 2.3835549257451408e-07,
+      "loss": 0.4385,
+      "step": 7263
+    },
+    {
+      "epoch": 3.603698647139134,
+      "grad_norm": 0.07240540018544343,
+      "learning_rate": 2.3775953008202847e-07,
+      "loss": 0.4319,
+      "step": 7264
+    },
+    {
+      "epoch": 3.6041951098423732,
+      "grad_norm": 0.07352272917174481,
+      "learning_rate": 2.3716429542164244e-07,
+      "loss": 0.4563,
+      "step": 7265
+    },
+    {
+      "epoch": 3.6046915725456126,
+      "grad_norm": 0.07016741575846668,
+      "learning_rate": 2.365697886843271e-07,
+      "loss": 0.4203,
+      "step": 7266
+    },
+    {
+      "epoch": 3.605188035248852,
+      "grad_norm": 0.0725771299653585,
+      "learning_rate": 2.359760099609437e-07,
+      "loss": 0.4473,
+      "step": 7267
+    },
+    {
+      "epoch": 3.6056844979520912,
+      "grad_norm": 0.07475172685609481,
+      "learning_rate": 2.353829593422424e-07,
+      "loss": 0.4599,
+      "step": 7268
+    },
+    {
+      "epoch": 3.6061809606553306,
+      "grad_norm": 0.0750639449163914,
+      "learning_rate": 2.3479063691886018e-07,
+      "loss": 0.4261,
+      "step": 7269
+    },
+    {
+      "epoch": 3.6066774233585703,
+      "grad_norm": 0.07078876083450575,
+      "learning_rate": 2.3419904278132565e-07,
+      "loss": 0.4384,
+      "step": 7270
+    },
+    {
+      "epoch": 3.6071738860618097,
+      "grad_norm": 0.07311201285095503,
+      "learning_rate": 2.3360817702005323e-07,
+      "loss": 0.4862,
+      "step": 7271
+    },
+    {
+      "epoch": 3.607670348765049,
+      "grad_norm": 0.07579752198532307,
+      "learning_rate": 2.330180397253473e-07,
+      "loss": 0.457,
+      "step": 7272
+    },
+    {
+      "epoch": 3.6081668114682883,
+      "grad_norm": 0.07207424841476477,
+      "learning_rate": 2.3242863098740187e-07,
+      "loss": 0.4296,
+      "step": 7273
+    },
+    {
+      "epoch": 3.608663274171528,
+      "grad_norm": 0.07384130182890736,
+      "learning_rate": 2.3183995089629707e-07,
+      "loss": 0.4589,
+      "step": 7274
+    },
+    {
+      "epoch": 3.6091597368747674,
+      "grad_norm": 0.07191996401035634,
+      "learning_rate": 2.3125199954200482e-07,
+      "loss": 0.4498,
+      "step": 7275
+    },
+    {
+      "epoch": 3.6096561995780068,
+      "grad_norm": 0.07344068487583766,
+      "learning_rate": 2.3066477701438383e-07,
+      "loss": 0.435,
+      "step": 7276
+    },
+    {
+      "epoch": 3.610152662281246,
+      "grad_norm": 0.07526670786555902,
+      "learning_rate": 2.3007828340318117e-07,
+      "loss": 0.4632,
+      "step": 7277
+    },
+    {
+      "epoch": 3.6106491249844854,
+      "grad_norm": 0.07578434183870061,
+      "learning_rate": 2.2949251879803236e-07,
+      "loss": 0.4655,
+      "step": 7278
+    },
+    {
+      "epoch": 3.6111455876877248,
+      "grad_norm": 0.0706605039902432,
+      "learning_rate": 2.28907483288463e-07,
+      "loss": 0.4075,
+      "step": 7279
+    },
+    {
+      "epoch": 3.6116420503909645,
+      "grad_norm": 0.07303631071464331,
+      "learning_rate": 2.2832317696388607e-07,
+      "loss": 0.4445,
+      "step": 7280
+    },
+    {
+      "epoch": 3.612138513094204,
+      "grad_norm": 0.070072324686116,
+      "learning_rate": 2.2773959991360394e-07,
+      "loss": 0.439,
+      "step": 7281
+    },
+    {
+      "epoch": 3.612634975797443,
+      "grad_norm": 0.07386234745655829,
+      "learning_rate": 2.2715675222680588e-07,
+      "loss": 0.4522,
+      "step": 7282
+    },
+    {
+      "epoch": 3.6131314385006825,
+      "grad_norm": 0.0719746863855503,
+      "learning_rate": 2.265746339925723e-07,
+      "loss": 0.4513,
+      "step": 7283
+    },
+    {
+      "epoch": 3.6136279012039223,
+      "grad_norm": 0.07171417706809774,
+      "learning_rate": 2.2599324529986866e-07,
+      "loss": 0.4085,
+      "step": 7284
+    },
+    {
+      "epoch": 3.6141243639071616,
+      "grad_norm": 0.07465108871234516,
+      "learning_rate": 2.2541258623755334e-07,
+      "loss": 0.4803,
+      "step": 7285
+    },
+    {
+      "epoch": 3.614620826610401,
+      "grad_norm": 0.07537298779351806,
+      "learning_rate": 2.2483265689436929e-07,
+      "loss": 0.4569,
+      "step": 7286
+    },
+    {
+      "epoch": 3.6151172893136403,
+      "grad_norm": 0.07750838694730808,
+      "learning_rate": 2.2425345735894888e-07,
+      "loss": 0.4839,
+      "step": 7287
+    },
+    {
+      "epoch": 3.6156137520168796,
+      "grad_norm": 0.07299486180670603,
+      "learning_rate": 2.2367498771981522e-07,
+      "loss": 0.4417,
+      "step": 7288
+    },
+    {
+      "epoch": 3.616110214720119,
+      "grad_norm": 0.07301228623195481,
+      "learning_rate": 2.230972480653759e-07,
+      "loss": 0.4334,
+      "step": 7289
+    },
+    {
+      "epoch": 3.6166066774233587,
+      "grad_norm": 0.07226158965409316,
+      "learning_rate": 2.2252023848393144e-07,
+      "loss": 0.4418,
+      "step": 7290
+    },
+    {
+      "epoch": 3.617103140126598,
+      "grad_norm": 0.07130018174773109,
+      "learning_rate": 2.219439590636674e-07,
+      "loss": 0.4359,
+      "step": 7291
+    },
+    {
+      "epoch": 3.6175996028298374,
+      "grad_norm": 0.07284789034528366,
+      "learning_rate": 2.213684098926583e-07,
+      "loss": 0.4229,
+      "step": 7292
+    },
+    {
+      "epoch": 3.6180960655330767,
+      "grad_norm": 0.07327554168996199,
+      "learning_rate": 2.2079359105886989e-07,
+      "loss": 0.4299,
+      "step": 7293
+    },
+    {
+      "epoch": 3.6185925282363165,
+      "grad_norm": 0.07272115811559865,
+      "learning_rate": 2.202195026501508e-07,
+      "loss": 0.4454,
+      "step": 7294
+    },
+    {
+      "epoch": 3.619088990939556,
+      "grad_norm": 0.07328239249201607,
+      "learning_rate": 2.1964614475424306e-07,
+      "loss": 0.4301,
+      "step": 7295
+    },
+    {
+      "epoch": 3.619585453642795,
+      "grad_norm": 0.07206975572008621,
+      "learning_rate": 2.1907351745877437e-07,
+      "loss": 0.4209,
+      "step": 7296
+    },
+    {
+      "epoch": 3.6200819163460345,
+      "grad_norm": 0.07177592916609533,
+      "learning_rate": 2.1850162085126303e-07,
+      "loss": 0.4084,
+      "step": 7297
+    },
+    {
+      "epoch": 3.620578379049274,
+      "grad_norm": 0.07030122376900125,
+      "learning_rate": 2.179304550191136e-07,
+      "loss": 0.4225,
+      "step": 7298
+    },
+    {
+      "epoch": 3.621074841752513,
+      "grad_norm": 0.07258135447311445,
+      "learning_rate": 2.1736002004961898e-07,
+      "loss": 0.4344,
+      "step": 7299
+    },
+    {
+      "epoch": 3.621571304455753,
+      "grad_norm": 0.07397369089030137,
+      "learning_rate": 2.167903160299617e-07,
+      "loss": 0.4452,
+      "step": 7300
+    },
+    {
+      "epoch": 3.6220677671589923,
+      "grad_norm": 0.06982375268398766,
+      "learning_rate": 2.1622134304721098e-07,
+      "loss": 0.4241,
+      "step": 7301
+    },
+    {
+      "epoch": 3.6225642298622316,
+      "grad_norm": 0.07186374761971,
+      "learning_rate": 2.156531011883267e-07,
+      "loss": 0.4109,
+      "step": 7302
+    },
+    {
+      "epoch": 3.623060692565471,
+      "grad_norm": 0.07247529471597118,
+      "learning_rate": 2.15085590540155e-07,
+      "loss": 0.4511,
+      "step": 7303
+    },
+    {
+      "epoch": 3.6235571552687107,
+      "grad_norm": 0.07458540665068539,
+      "learning_rate": 2.1451881118942975e-07,
+      "loss": 0.4617,
+      "step": 7304
+    },
+    {
+      "epoch": 3.62405361797195,
+      "grad_norm": 0.07582139648778306,
+      "learning_rate": 2.1395276322277504e-07,
+      "loss": 0.4627,
+      "step": 7305
+    },
+    {
+      "epoch": 3.6245500806751894,
+      "grad_norm": 0.07402453362459176,
+      "learning_rate": 2.1338744672670165e-07,
+      "loss": 0.4924,
+      "step": 7306
+    },
+    {
+      "epoch": 3.6250465433784287,
+      "grad_norm": 0.07619031513083764,
+      "learning_rate": 2.1282286178761046e-07,
+      "loss": 0.4432,
+      "step": 7307
+    },
+    {
+      "epoch": 3.625543006081668,
+      "grad_norm": 0.07527533962278696,
+      "learning_rate": 2.1225900849178804e-07,
+      "loss": 0.4662,
+      "step": 7308
+    },
+    {
+      "epoch": 3.6260394687849073,
+      "grad_norm": 0.07256160181172347,
+      "learning_rate": 2.11695886925411e-07,
+      "loss": 0.4579,
+      "step": 7309
+    },
+    {
+      "epoch": 3.6265359314881467,
+      "grad_norm": 0.07459927144810276,
+      "learning_rate": 2.1113349717454267e-07,
+      "loss": 0.4597,
+      "step": 7310
+    },
+    {
+      "epoch": 3.6270323941913865,
+      "grad_norm": 0.07210064448559347,
+      "learning_rate": 2.105718393251349e-07,
+      "loss": 0.4384,
+      "step": 7311
+    },
+    {
+      "epoch": 3.627528856894626,
+      "grad_norm": 0.07283451452803122,
+      "learning_rate": 2.100109134630296e-07,
+      "loss": 0.4451,
+      "step": 7312
+    },
+    {
+      "epoch": 3.628025319597865,
+      "grad_norm": 0.0768403360547824,
+      "learning_rate": 2.094507196739537e-07,
+      "loss": 0.4738,
+      "step": 7313
+    },
+    {
+      "epoch": 3.628521782301105,
+      "grad_norm": 0.0744758874009114,
+      "learning_rate": 2.0889125804352595e-07,
+      "loss": 0.4502,
+      "step": 7314
+    },
+    {
+      "epoch": 3.6290182450043442,
+      "grad_norm": 0.0716895568860105,
+      "learning_rate": 2.0833252865724907e-07,
+      "loss": 0.4421,
+      "step": 7315
+    },
+    {
+      "epoch": 3.6295147077075836,
+      "grad_norm": 0.0726724197567022,
+      "learning_rate": 2.077745316005164e-07,
+      "loss": 0.4341,
+      "step": 7316
+    },
+    {
+      "epoch": 3.630011170410823,
+      "grad_norm": 0.07434405417937585,
+      "learning_rate": 2.0721726695860977e-07,
+      "loss": 0.4293,
+      "step": 7317
+    },
+    {
+      "epoch": 3.630507633114062,
+      "grad_norm": 0.07143412055686629,
+      "learning_rate": 2.0666073481669714e-07,
+      "loss": 0.4097,
+      "step": 7318
+    },
+    {
+      "epoch": 3.6310040958173015,
+      "grad_norm": 0.07276002545567115,
+      "learning_rate": 2.0610493525983544e-07,
+      "loss": 0.4411,
+      "step": 7319
+    },
+    {
+      "epoch": 3.631500558520541,
+      "grad_norm": 0.07075461918486434,
+      "learning_rate": 2.0554986837297064e-07,
+      "loss": 0.427,
+      "step": 7320
+    },
+    {
+      "epoch": 3.6319970212237807,
+      "grad_norm": 0.07420935677934005,
+      "learning_rate": 2.049955342409349e-07,
+      "loss": 0.426,
+      "step": 7321
+    },
+    {
+      "epoch": 3.63249348392702,
+      "grad_norm": 0.06984495924090936,
+      "learning_rate": 2.044419329484504e-07,
+      "loss": 0.4552,
+      "step": 7322
+    },
+    {
+      "epoch": 3.6329899466302593,
+      "grad_norm": 0.06939603455821021,
+      "learning_rate": 2.0388906458012503e-07,
+      "loss": 0.4255,
+      "step": 7323
+    },
+    {
+      "epoch": 3.633486409333499,
+      "grad_norm": 0.07035364818165041,
+      "learning_rate": 2.0333692922045623e-07,
+      "loss": 0.4284,
+      "step": 7324
+    },
+    {
+      "epoch": 3.6339828720367384,
+      "grad_norm": 0.07382774656321925,
+      "learning_rate": 2.0278552695383036e-07,
+      "loss": 0.4388,
+      "step": 7325
+    },
+    {
+      "epoch": 3.6344793347399778,
+      "grad_norm": 0.07532508040952224,
+      "learning_rate": 2.022348578645178e-07,
+      "loss": 0.4571,
+      "step": 7326
+    },
+    {
+      "epoch": 3.634975797443217,
+      "grad_norm": 0.07251073324334695,
+      "learning_rate": 2.0168492203668122e-07,
+      "loss": 0.4427,
+      "step": 7327
+    },
+    {
+      "epoch": 3.6354722601464564,
+      "grad_norm": 0.0706357298880477,
+      "learning_rate": 2.0113571955436895e-07,
+      "loss": 0.4272,
+      "step": 7328
+    },
+    {
+      "epoch": 3.6359687228496957,
+      "grad_norm": 0.07216767093440042,
+      "learning_rate": 2.0058725050151828e-07,
+      "loss": 0.4375,
+      "step": 7329
+    },
+    {
+      "epoch": 3.636465185552935,
+      "grad_norm": 0.07330320428541527,
+      "learning_rate": 2.0003951496195385e-07,
+      "loss": 0.4575,
+      "step": 7330
+    },
+    {
+      "epoch": 3.636961648256175,
+      "grad_norm": 0.07144970418486908,
+      "learning_rate": 1.9949251301938756e-07,
+      "loss": 0.4409,
+      "step": 7331
+    },
+    {
+      "epoch": 3.637458110959414,
+      "grad_norm": 0.07503845282729875,
+      "learning_rate": 1.9894624475742086e-07,
+      "loss": 0.4707,
+      "step": 7332
+    },
+    {
+      "epoch": 3.6379545736626535,
+      "grad_norm": 0.07117560793351785,
+      "learning_rate": 1.9840071025954089e-07,
+      "loss": 0.4067,
+      "step": 7333
+    },
+    {
+      "epoch": 3.6384510363658933,
+      "grad_norm": 0.07277799392519127,
+      "learning_rate": 1.9785590960912538e-07,
+      "loss": 0.4395,
+      "step": 7334
+    },
+    {
+      "epoch": 3.6389474990691326,
+      "grad_norm": 0.07147057359788965,
+      "learning_rate": 1.9731184288943772e-07,
+      "loss": 0.4275,
+      "step": 7335
+    },
+    {
+      "epoch": 3.639443961772372,
+      "grad_norm": 0.11171374930937528,
+      "learning_rate": 1.9676851018362865e-07,
+      "loss": 0.4633,
+      "step": 7336
+    },
+    {
+      "epoch": 3.6399404244756113,
+      "grad_norm": 0.07646171474556615,
+      "learning_rate": 1.9622591157473946e-07,
+      "loss": 0.4648,
+      "step": 7337
+    },
+    {
+      "epoch": 3.6404368871788506,
+      "grad_norm": 0.07372439939792098,
+      "learning_rate": 1.9568404714569666e-07,
+      "loss": 0.453,
+      "step": 7338
+    },
+    {
+      "epoch": 3.64093334988209,
+      "grad_norm": 0.07240599601324296,
+      "learning_rate": 1.951429169793162e-07,
+      "loss": 0.4366,
+      "step": 7339
+    },
+    {
+      "epoch": 3.6414298125853293,
+      "grad_norm": 0.072422143521238,
+      "learning_rate": 1.9460252115830137e-07,
+      "loss": 0.4423,
+      "step": 7340
+    },
+    {
+      "epoch": 3.641926275288569,
+      "grad_norm": 0.07145807098907717,
+      "learning_rate": 1.940628597652422e-07,
+      "loss": 0.4506,
+      "step": 7341
+    },
+    {
+      "epoch": 3.6424227379918084,
+      "grad_norm": 0.07421632819020778,
+      "learning_rate": 1.9352393288261717e-07,
+      "loss": 0.4659,
+      "step": 7342
+    },
+    {
+      "epoch": 3.6429192006950477,
+      "grad_norm": 0.07249759916938332,
+      "learning_rate": 1.9298574059279263e-07,
+      "loss": 0.4437,
+      "step": 7343
+    },
+    {
+      "epoch": 3.643415663398287,
+      "grad_norm": 0.07546293310432843,
+      "learning_rate": 1.9244828297802386e-07,
+      "loss": 0.4802,
+      "step": 7344
+    },
+    {
+      "epoch": 3.643912126101527,
+      "grad_norm": 0.07424234347976809,
+      "learning_rate": 1.919115601204513e-07,
+      "loss": 0.4367,
+      "step": 7345
+    },
+    {
+      "epoch": 3.644408588804766,
+      "grad_norm": 0.07359058194638433,
+      "learning_rate": 1.9137557210210544e-07,
+      "loss": 0.4675,
+      "step": 7346
+    },
+    {
+      "epoch": 3.6449050515080055,
+      "grad_norm": 0.07322009392359763,
+      "learning_rate": 1.9084031900490297e-07,
+      "loss": 0.4305,
+      "step": 7347
+    },
+    {
+      "epoch": 3.645401514211245,
+      "grad_norm": 0.07376178653192299,
+      "learning_rate": 1.9030580091064787e-07,
+      "loss": 0.4631,
+      "step": 7348
+    },
+    {
+      "epoch": 3.645897976914484,
+      "grad_norm": 0.07327292909717786,
+      "learning_rate": 1.8977201790103428e-07,
+      "loss": 0.424,
+      "step": 7349
+    },
+    {
+      "epoch": 3.6463944396177235,
+      "grad_norm": 0.07114499977321155,
+      "learning_rate": 1.892389700576408e-07,
+      "loss": 0.426,
+      "step": 7350
+    },
+    {
+      "epoch": 3.6468909023209632,
+      "grad_norm": 0.069110037267425,
+      "learning_rate": 1.8870665746193672e-07,
+      "loss": 0.4088,
+      "step": 7351
+    },
+    {
+      "epoch": 3.6473873650242026,
+      "grad_norm": 0.07252850374903859,
+      "learning_rate": 1.8817508019527696e-07,
+      "loss": 0.4345,
+      "step": 7352
+    },
+    {
+      "epoch": 3.647883827727442,
+      "grad_norm": 0.07267951521063351,
+      "learning_rate": 1.8764423833890434e-07,
+      "loss": 0.4111,
+      "step": 7353
+    },
+    {
+      "epoch": 3.6483802904306812,
+      "grad_norm": 0.07179521032898407,
+      "learning_rate": 1.8711413197394944e-07,
+      "loss": 0.4216,
+      "step": 7354
+    },
+    {
+      "epoch": 3.648876753133921,
+      "grad_norm": 0.07389459033510468,
+      "learning_rate": 1.8658476118143086e-07,
+      "loss": 0.4613,
+      "step": 7355
+    },
+    {
+      "epoch": 3.6493732158371603,
+      "grad_norm": 0.0723475247866929,
+      "learning_rate": 1.8605612604225388e-07,
+      "loss": 0.4221,
+      "step": 7356
+    },
+    {
+      "epoch": 3.6498696785403997,
+      "grad_norm": 0.07341268249466791,
+      "learning_rate": 1.8552822663721382e-07,
+      "loss": 0.4354,
+      "step": 7357
+    },
+    {
+      "epoch": 3.650366141243639,
+      "grad_norm": 0.07049114822555134,
+      "learning_rate": 1.850010630469884e-07,
+      "loss": 0.4192,
+      "step": 7358
+    },
+    {
+      "epoch": 3.6508626039468783,
+      "grad_norm": 0.07264265024902371,
+      "learning_rate": 1.8447463535214872e-07,
+      "loss": 0.4275,
+      "step": 7359
+    },
+    {
+      "epoch": 3.6513590666501177,
+      "grad_norm": 0.06929144716295033,
+      "learning_rate": 1.839489436331493e-07,
+      "loss": 0.433,
+      "step": 7360
+    },
+    {
+      "epoch": 3.6518555293533574,
+      "grad_norm": 0.07198194923441253,
+      "learning_rate": 1.8342398797033479e-07,
+      "loss": 0.4572,
+      "step": 7361
+    },
+    {
+      "epoch": 3.6523519920565968,
+      "grad_norm": 0.07200270067094541,
+      "learning_rate": 1.8289976844393599e-07,
+      "loss": 0.4322,
+      "step": 7362
+    },
+    {
+      "epoch": 3.652848454759836,
+      "grad_norm": 0.06941931192926973,
+      "learning_rate": 1.8237628513407046e-07,
+      "loss": 0.4091,
+      "step": 7363
+    },
+    {
+      "epoch": 3.6533449174630754,
+      "grad_norm": 0.07070649329985829,
+      "learning_rate": 1.818535381207459e-07,
+      "loss": 0.4351,
+      "step": 7364
+    },
+    {
+      "epoch": 3.653841380166315,
+      "grad_norm": 0.07445472823058359,
+      "learning_rate": 1.8133152748385397e-07,
+      "loss": 0.4628,
+      "step": 7365
+    },
+    {
+      "epoch": 3.6543378428695545,
+      "grad_norm": 0.07447413719165662,
+      "learning_rate": 1.8081025330317748e-07,
+      "loss": 0.4695,
+      "step": 7366
+    },
+    {
+      "epoch": 3.654834305572794,
+      "grad_norm": 0.07359381894990621,
+      "learning_rate": 1.8028971565838381e-07,
+      "loss": 0.437,
+      "step": 7367
+    },
+    {
+      "epoch": 3.655330768276033,
+      "grad_norm": 0.07046676873498374,
+      "learning_rate": 1.7976991462902827e-07,
+      "loss": 0.447,
+      "step": 7368
+    },
+    {
+      "epoch": 3.6558272309792725,
+      "grad_norm": 0.07269725462449012,
+      "learning_rate": 1.7925085029455558e-07,
+      "loss": 0.4503,
+      "step": 7369
+    },
+    {
+      "epoch": 3.656323693682512,
+      "grad_norm": 0.07852134438582638,
+      "learning_rate": 1.787325227342951e-07,
+      "loss": 0.4914,
+      "step": 7370
+    },
+    {
+      "epoch": 3.6568201563857516,
+      "grad_norm": 0.07509607274320566,
+      "learning_rate": 1.7821493202746565e-07,
+      "loss": 0.4357,
+      "step": 7371
+    },
+    {
+      "epoch": 3.657316619088991,
+      "grad_norm": 0.07142845307398488,
+      "learning_rate": 1.7769807825317232e-07,
+      "loss": 0.4394,
+      "step": 7372
+    },
+    {
+      "epoch": 3.6578130817922303,
+      "grad_norm": 0.07178759846116196,
+      "learning_rate": 1.77181961490408e-07,
+      "loss": 0.4345,
+      "step": 7373
+    },
+    {
+      "epoch": 3.6583095444954696,
+      "grad_norm": 0.07158932483763085,
+      "learning_rate": 1.7666658181805295e-07,
+      "loss": 0.434,
+      "step": 7374
+    },
+    {
+      "epoch": 3.6588060071987094,
+      "grad_norm": 0.07280743494373647,
+      "learning_rate": 1.7615193931487417e-07,
+      "loss": 0.4337,
+      "step": 7375
+    },
+    {
+      "epoch": 3.6593024699019487,
+      "grad_norm": 0.07453278405125927,
+      "learning_rate": 1.7563803405952761e-07,
+      "loss": 0.4368,
+      "step": 7376
+    },
+    {
+      "epoch": 3.659798932605188,
+      "grad_norm": 0.07010280774336348,
+      "learning_rate": 1.751248661305538e-07,
+      "loss": 0.4187,
+      "step": 7377
+    },
+    {
+      "epoch": 3.6602953953084274,
+      "grad_norm": 0.0726927043570313,
+      "learning_rate": 1.7461243560638442e-07,
+      "loss": 0.4559,
+      "step": 7378
+    },
+    {
+      "epoch": 3.6607918580116667,
+      "grad_norm": 0.07233361327364231,
+      "learning_rate": 1.741007425653346e-07,
+      "loss": 0.4505,
+      "step": 7379
+    },
+    {
+      "epoch": 3.661288320714906,
+      "grad_norm": 0.07363675822337719,
+      "learning_rate": 1.7358978708560848e-07,
+      "loss": 0.4361,
+      "step": 7380
+    },
+    {
+      "epoch": 3.661784783418146,
+      "grad_norm": 0.07199309976813066,
+      "learning_rate": 1.73079569245298e-07,
+      "loss": 0.4289,
+      "step": 7381
+    },
+    {
+      "epoch": 3.662281246121385,
+      "grad_norm": 0.07317353186147292,
+      "learning_rate": 1.7257008912238138e-07,
+      "loss": 0.4487,
+      "step": 7382
+    },
+    {
+      "epoch": 3.6627777088246245,
+      "grad_norm": 0.07372588029874809,
+      "learning_rate": 1.720613467947252e-07,
+      "loss": 0.4389,
+      "step": 7383
+    },
+    {
+      "epoch": 3.663274171527864,
+      "grad_norm": 0.07320921827857442,
+      "learning_rate": 1.715533423400817e-07,
+      "loss": 0.4193,
+      "step": 7384
+    },
+    {
+      "epoch": 3.6637706342311036,
+      "grad_norm": 0.07522855796845176,
+      "learning_rate": 1.7104607583609157e-07,
+      "loss": 0.4581,
+      "step": 7385
+    },
+    {
+      "epoch": 3.664267096934343,
+      "grad_norm": 0.0734689312284619,
+      "learning_rate": 1.7053954736028222e-07,
+      "loss": 0.4576,
+      "step": 7386
+    },
+    {
+      "epoch": 3.6647635596375823,
+      "grad_norm": 0.07075618371467093,
+      "learning_rate": 1.700337569900684e-07,
+      "loss": 0.4227,
+      "step": 7387
+    },
+    {
+      "epoch": 3.6652600223408216,
+      "grad_norm": 0.06987389893800902,
+      "learning_rate": 1.6952870480275273e-07,
+      "loss": 0.4298,
+      "step": 7388
+    },
+    {
+      "epoch": 3.665756485044061,
+      "grad_norm": 0.07197270042011136,
+      "learning_rate": 1.6902439087552402e-07,
+      "loss": 0.4571,
+      "step": 7389
+    },
+    {
+      "epoch": 3.6662529477473003,
+      "grad_norm": 0.07106276503884373,
+      "learning_rate": 1.6852081528545838e-07,
+      "loss": 0.4175,
+      "step": 7390
+    },
+    {
+      "epoch": 3.66674941045054,
+      "grad_norm": 0.07285710553611387,
+      "learning_rate": 1.680179781095187e-07,
+      "loss": 0.4527,
+      "step": 7391
+    },
+    {
+      "epoch": 3.6672458731537794,
+      "grad_norm": 0.0737065341456534,
+      "learning_rate": 1.6751587942455627e-07,
+      "loss": 0.4609,
+      "step": 7392
+    },
+    {
+      "epoch": 3.6677423358570187,
+      "grad_norm": 0.07056335676130286,
+      "learning_rate": 1.670145193073086e-07,
+      "loss": 0.4241,
+      "step": 7393
+    },
+    {
+      "epoch": 3.668238798560258,
+      "grad_norm": 0.07405405366597342,
+      "learning_rate": 1.665138978344011e-07,
+      "loss": 0.4321,
+      "step": 7394
+    },
+    {
+      "epoch": 3.668735261263498,
+      "grad_norm": 0.07170622222913631,
+      "learning_rate": 1.6601401508234417e-07,
+      "loss": 0.4267,
+      "step": 7395
+    },
+    {
+      "epoch": 3.669231723966737,
+      "grad_norm": 0.07178680400866755,
+      "learning_rate": 1.6551487112753893e-07,
+      "loss": 0.4264,
+      "step": 7396
+    },
+    {
+      "epoch": 3.6697281866699765,
+      "grad_norm": 0.0721285124574552,
+      "learning_rate": 1.6501646604626997e-07,
+      "loss": 0.4489,
+      "step": 7397
+    },
+    {
+      "epoch": 3.670224649373216,
+      "grad_norm": 0.07082806212736886,
+      "learning_rate": 1.6451879991471186e-07,
+      "loss": 0.4538,
+      "step": 7398
+    },
+    {
+      "epoch": 3.670721112076455,
+      "grad_norm": 0.0729535155041323,
+      "learning_rate": 1.640218728089238e-07,
+      "loss": 0.4467,
+      "step": 7399
+    },
+    {
+      "epoch": 3.6712175747796945,
+      "grad_norm": 0.07230176500116002,
+      "learning_rate": 1.6352568480485277e-07,
+      "loss": 0.4326,
+      "step": 7400
+    },
+    {
+      "epoch": 3.6717140374829342,
+      "grad_norm": 0.07345615326458771,
+      "learning_rate": 1.6303023597833478e-07,
+      "loss": 0.462,
+      "step": 7401
+    },
+    {
+      "epoch": 3.6722105001861736,
+      "grad_norm": 0.07151370299052442,
+      "learning_rate": 1.6253552640508985e-07,
+      "loss": 0.4369,
+      "step": 7402
+    },
+    {
+      "epoch": 3.672706962889413,
+      "grad_norm": 0.07148248573684546,
+      "learning_rate": 1.6204155616072693e-07,
+      "loss": 0.4302,
+      "step": 7403
+    },
+    {
+      "epoch": 3.6732034255926522,
+      "grad_norm": 0.07043625217762424,
+      "learning_rate": 1.615483253207417e-07,
+      "loss": 0.4506,
+      "step": 7404
+    },
+    {
+      "epoch": 3.673699888295892,
+      "grad_norm": 0.0728731177103944,
+      "learning_rate": 1.610558339605156e-07,
+      "loss": 0.4342,
+      "step": 7405
+    },
+    {
+      "epoch": 3.6741963509991313,
+      "grad_norm": 0.07369450751879054,
+      "learning_rate": 1.6056408215532005e-07,
+      "loss": 0.4324,
+      "step": 7406
+    },
+    {
+      "epoch": 3.6746928137023707,
+      "grad_norm": 0.07185327188054835,
+      "learning_rate": 1.600730699803088e-07,
+      "loss": 0.4597,
+      "step": 7407
+    },
+    {
+      "epoch": 3.67518927640561,
+      "grad_norm": 0.07372288477305014,
+      "learning_rate": 1.5958279751052686e-07,
+      "loss": 0.4277,
+      "step": 7408
+    },
+    {
+      "epoch": 3.6756857391088493,
+      "grad_norm": 0.07364912954293501,
+      "learning_rate": 1.5909326482090371e-07,
+      "loss": 0.4667,
+      "step": 7409
+    },
+    {
+      "epoch": 3.6761822018120887,
+      "grad_norm": 0.07528543823766569,
+      "learning_rate": 1.5860447198625784e-07,
+      "loss": 0.4458,
+      "step": 7410
+    },
+    {
+      "epoch": 3.6766786645153284,
+      "grad_norm": 0.07312472664500712,
+      "learning_rate": 1.5811641908129226e-07,
+      "loss": 0.4478,
+      "step": 7411
+    },
+    {
+      "epoch": 3.6771751272185678,
+      "grad_norm": 0.07443070106907346,
+      "learning_rate": 1.576291061805979e-07,
+      "loss": 0.4292,
+      "step": 7412
+    },
+    {
+      "epoch": 3.677671589921807,
+      "grad_norm": 0.07226946918899885,
+      "learning_rate": 1.571425333586535e-07,
+      "loss": 0.4423,
+      "step": 7413
+    },
+    {
+      "epoch": 3.6781680526250464,
+      "grad_norm": 0.07206464093796312,
+      "learning_rate": 1.5665670068982286e-07,
+      "loss": 0.4505,
+      "step": 7414
+    },
+    {
+      "epoch": 3.678664515328286,
+      "grad_norm": 0.07188587481501309,
+      "learning_rate": 1.5617160824835942e-07,
+      "loss": 0.4254,
+      "step": 7415
+    },
+    {
+      "epoch": 3.6791609780315255,
+      "grad_norm": 0.07265685993339371,
+      "learning_rate": 1.556872561084005e-07,
+      "loss": 0.415,
+      "step": 7416
+    },
+    {
+      "epoch": 3.679657440734765,
+      "grad_norm": 0.07162493038184804,
+      "learning_rate": 1.552036443439714e-07,
+      "loss": 0.4041,
+      "step": 7417
+    },
+    {
+      "epoch": 3.680153903438004,
+      "grad_norm": 0.07155610729032282,
+      "learning_rate": 1.5472077302898515e-07,
+      "loss": 0.4076,
+      "step": 7418
+    },
+    {
+      "epoch": 3.6806503661412435,
+      "grad_norm": 0.07355344416089545,
+      "learning_rate": 1.542386422372405e-07,
+      "loss": 0.4398,
+      "step": 7419
+    },
+    {
+      "epoch": 3.681146828844483,
+      "grad_norm": 0.07116724820478201,
+      "learning_rate": 1.5375725204242407e-07,
+      "loss": 0.4096,
+      "step": 7420
+    },
+    {
+      "epoch": 3.6816432915477226,
+      "grad_norm": 0.07241791000339823,
+      "learning_rate": 1.532766025181076e-07,
+      "loss": 0.4413,
+      "step": 7421
+    },
+    {
+      "epoch": 3.682139754250962,
+      "grad_norm": 0.07047095111897257,
+      "learning_rate": 1.5279669373775118e-07,
+      "loss": 0.4271,
+      "step": 7422
+    },
+    {
+      "epoch": 3.6826362169542013,
+      "grad_norm": 0.07313074541716791,
+      "learning_rate": 1.523175257747017e-07,
+      "loss": 0.4202,
+      "step": 7423
+    },
+    {
+      "epoch": 3.6831326796574406,
+      "grad_norm": 0.0728777056119168,
+      "learning_rate": 1.518390987021906e-07,
+      "loss": 0.4475,
+      "step": 7424
+    },
+    {
+      "epoch": 3.6836291423606804,
+      "grad_norm": 0.07442455746276211,
+      "learning_rate": 1.5136141259333992e-07,
+      "loss": 0.4543,
+      "step": 7425
+    },
+    {
+      "epoch": 3.6841256050639197,
+      "grad_norm": 0.07267058678491543,
+      "learning_rate": 1.5088446752115403e-07,
+      "loss": 0.4238,
+      "step": 7426
+    },
+    {
+      "epoch": 3.684622067767159,
+      "grad_norm": 0.07096583850524037,
+      "learning_rate": 1.504082635585291e-07,
+      "loss": 0.4389,
+      "step": 7427
+    },
+    {
+      "epoch": 3.6851185304703984,
+      "grad_norm": 0.07447238809266303,
+      "learning_rate": 1.49932800778243e-07,
+      "loss": 0.4358,
+      "step": 7428
+    },
+    {
+      "epoch": 3.6856149931736377,
+      "grad_norm": 0.0726313401883301,
+      "learning_rate": 1.494580792529632e-07,
+      "loss": 0.4654,
+      "step": 7429
+    },
+    {
+      "epoch": 3.686111455876877,
+      "grad_norm": 0.07211262857098678,
+      "learning_rate": 1.4898409905524436e-07,
+      "loss": 0.4304,
+      "step": 7430
+    },
+    {
+      "epoch": 3.686607918580117,
+      "grad_norm": 0.07363037113750746,
+      "learning_rate": 1.4851086025752525e-07,
+      "loss": 0.4148,
+      "step": 7431
+    },
+    {
+      "epoch": 3.687104381283356,
+      "grad_norm": 0.07434935712528255,
+      "learning_rate": 1.4803836293213303e-07,
+      "loss": 0.4506,
+      "step": 7432
+    },
+    {
+      "epoch": 3.6876008439865955,
+      "grad_norm": 0.07194422077130175,
+      "learning_rate": 1.4756660715128267e-07,
+      "loss": 0.4413,
+      "step": 7433
+    },
+    {
+      "epoch": 3.688097306689835,
+      "grad_norm": 0.07294248103909735,
+      "learning_rate": 1.4709559298707265e-07,
+      "loss": 0.4236,
+      "step": 7434
+    },
+    {
+      "epoch": 3.6885937693930746,
+      "grad_norm": 0.07226694533909107,
+      "learning_rate": 1.4662532051149149e-07,
+      "loss": 0.4567,
+      "step": 7435
+    },
+    {
+      "epoch": 3.689090232096314,
+      "grad_norm": 0.07333144120478627,
+      "learning_rate": 1.4615578979641164e-07,
+      "loss": 0.4091,
+      "step": 7436
+    },
+    {
+      "epoch": 3.6895866947995533,
+      "grad_norm": 0.07599661651446776,
+      "learning_rate": 1.4568700091359412e-07,
+      "loss": 0.4766,
+      "step": 7437
+    },
+    {
+      "epoch": 3.6900831575027926,
+      "grad_norm": 0.07267348396613163,
+      "learning_rate": 1.45218953934686e-07,
+      "loss": 0.4541,
+      "step": 7438
+    },
+    {
+      "epoch": 3.690579620206032,
+      "grad_norm": 0.07154268790112295,
+      "learning_rate": 1.4475164893121952e-07,
+      "loss": 0.445,
+      "step": 7439
+    },
+    {
+      "epoch": 3.6910760829092712,
+      "grad_norm": 0.07736575888556468,
+      "learning_rate": 1.4428508597461587e-07,
+      "loss": 0.4725,
+      "step": 7440
+    },
+    {
+      "epoch": 3.691572545612511,
+      "grad_norm": 0.07338375042541488,
+      "learning_rate": 1.4381926513618139e-07,
+      "loss": 0.4707,
+      "step": 7441
+    },
+    {
+      "epoch": 3.6920690083157504,
+      "grad_norm": 0.07366882489429853,
+      "learning_rate": 1.4335418648710907e-07,
+      "loss": 0.4544,
+      "step": 7442
+    },
+    {
+      "epoch": 3.6925654710189897,
+      "grad_norm": 0.07366490558844847,
+      "learning_rate": 1.4288985009847932e-07,
+      "loss": 0.4729,
+      "step": 7443
+    },
+    {
+      "epoch": 3.693061933722229,
+      "grad_norm": 0.07561083067989416,
+      "learning_rate": 1.4242625604125758e-07,
+      "loss": 0.4583,
+      "step": 7444
+    },
+    {
+      "epoch": 3.693558396425469,
+      "grad_norm": 0.07254485416644309,
+      "learning_rate": 1.4196340438629774e-07,
+      "loss": 0.4202,
+      "step": 7445
+    },
+    {
+      "epoch": 3.694054859128708,
+      "grad_norm": 0.07400616431660505,
+      "learning_rate": 1.415012952043382e-07,
+      "loss": 0.4553,
+      "step": 7446
+    },
+    {
+      "epoch": 3.6945513218319475,
+      "grad_norm": 0.07556294028601898,
+      "learning_rate": 1.4103992856600634e-07,
+      "loss": 0.4469,
+      "step": 7447
+    },
+    {
+      "epoch": 3.695047784535187,
+      "grad_norm": 0.07015152787078588,
+      "learning_rate": 1.4057930454181412e-07,
+      "loss": 0.4226,
+      "step": 7448
+    },
+    {
+      "epoch": 3.695544247238426,
+      "grad_norm": 0.07176355041550919,
+      "learning_rate": 1.4011942320215964e-07,
+      "loss": 0.4476,
+      "step": 7449
+    },
+    {
+      "epoch": 3.6960407099416654,
+      "grad_norm": 0.0723388702967807,
+      "learning_rate": 1.396602846173295e-07,
+      "loss": 0.4106,
+      "step": 7450
+    },
+    {
+      "epoch": 3.696537172644905,
+      "grad_norm": 0.07139271451498826,
+      "learning_rate": 1.3920188885749475e-07,
+      "loss": 0.4268,
+      "step": 7451
+    },
+    {
+      "epoch": 3.6970336353481446,
+      "grad_norm": 0.0730447884583585,
+      "learning_rate": 1.3874423599271435e-07,
+      "loss": 0.4533,
+      "step": 7452
+    },
+    {
+      "epoch": 3.697530098051384,
+      "grad_norm": 0.07339909768171726,
+      "learning_rate": 1.3828732609293404e-07,
+      "loss": 0.4243,
+      "step": 7453
+    },
+    {
+      "epoch": 3.698026560754623,
+      "grad_norm": 0.07116727161227962,
+      "learning_rate": 1.378311592279835e-07,
+      "loss": 0.4267,
+      "step": 7454
+    },
+    {
+      "epoch": 3.698523023457863,
+      "grad_norm": 0.07034248657623826,
+      "learning_rate": 1.3737573546758198e-07,
+      "loss": 0.4024,
+      "step": 7455
+    },
+    {
+      "epoch": 3.6990194861611023,
+      "grad_norm": 0.07127374652005736,
+      "learning_rate": 1.3692105488133211e-07,
+      "loss": 0.4129,
+      "step": 7456
+    },
+    {
+      "epoch": 3.6995159488643417,
+      "grad_norm": 0.07375826456205288,
+      "learning_rate": 1.364671175387261e-07,
+      "loss": 0.4344,
+      "step": 7457
+    },
+    {
+      "epoch": 3.700012411567581,
+      "grad_norm": 0.07357945355081973,
+      "learning_rate": 1.3601392350913957e-07,
+      "loss": 0.4299,
+      "step": 7458
+    },
+    {
+      "epoch": 3.7005088742708203,
+      "grad_norm": 0.06876318428670228,
+      "learning_rate": 1.3556147286183762e-07,
+      "loss": 0.4066,
+      "step": 7459
+    },
+    {
+      "epoch": 3.7010053369740596,
+      "grad_norm": 0.07282356150500176,
+      "learning_rate": 1.3510976566596946e-07,
+      "loss": 0.4544,
+      "step": 7460
+    },
+    {
+      "epoch": 3.701501799677299,
+      "grad_norm": 0.07378253520368942,
+      "learning_rate": 1.346588019905698e-07,
+      "loss": 0.4738,
+      "step": 7461
+    },
+    {
+      "epoch": 3.7019982623805388,
+      "grad_norm": 0.07330034917523164,
+      "learning_rate": 1.3420858190456353e-07,
+      "loss": 0.4493,
+      "step": 7462
+    },
+    {
+      "epoch": 3.702494725083778,
+      "grad_norm": 0.07280228471411862,
+      "learning_rate": 1.3375910547675785e-07,
+      "loss": 0.4559,
+      "step": 7463
+    },
+    {
+      "epoch": 3.7029911877870174,
+      "grad_norm": 0.07247121072667194,
+      "learning_rate": 1.333103727758489e-07,
+      "loss": 0.4429,
+      "step": 7464
+    },
+    {
+      "epoch": 3.703487650490257,
+      "grad_norm": 0.07188303430763324,
+      "learning_rate": 1.328623838704185e-07,
+      "loss": 0.4578,
+      "step": 7465
+    },
+    {
+      "epoch": 3.7039841131934965,
+      "grad_norm": 0.07198733038911835,
+      "learning_rate": 1.3241513882893297e-07,
+      "loss": 0.4456,
+      "step": 7466
+    },
+    {
+      "epoch": 3.704480575896736,
+      "grad_norm": 0.07444942706026461,
+      "learning_rate": 1.3196863771974877e-07,
+      "loss": 0.4482,
+      "step": 7467
+    },
+    {
+      "epoch": 3.704977038599975,
+      "grad_norm": 0.07331451044578703,
+      "learning_rate": 1.3152288061110518e-07,
+      "loss": 0.4722,
+      "step": 7468
+    },
+    {
+      "epoch": 3.7054735013032145,
+      "grad_norm": 0.0711213868596691,
+      "learning_rate": 1.3107786757112827e-07,
+      "loss": 0.4449,
+      "step": 7469
+    },
+    {
+      "epoch": 3.705969964006454,
+      "grad_norm": 0.07371066642583825,
+      "learning_rate": 1.3063359866783365e-07,
+      "loss": 0.4485,
+      "step": 7470
+    },
+    {
+      "epoch": 3.706466426709693,
+      "grad_norm": 0.07218421529119273,
+      "learning_rate": 1.3019007396911809e-07,
+      "loss": 0.4218,
+      "step": 7471
+    },
+    {
+      "epoch": 3.706962889412933,
+      "grad_norm": 0.07372364917278044,
+      "learning_rate": 1.297472935427685e-07,
+      "loss": 0.4183,
+      "step": 7472
+    },
+    {
+      "epoch": 3.7074593521161723,
+      "grad_norm": 0.0728757938451939,
+      "learning_rate": 1.2930525745645572e-07,
+      "loss": 0.4456,
+      "step": 7473
+    },
+    {
+      "epoch": 3.7079558148194116,
+      "grad_norm": 0.07611551431202836,
+      "learning_rate": 1.2886396577773963e-07,
+      "loss": 0.4598,
+      "step": 7474
+    },
+    {
+      "epoch": 3.7084522775226514,
+      "grad_norm": 0.07087427054206022,
+      "learning_rate": 1.284234185740635e-07,
+      "loss": 0.4217,
+      "step": 7475
+    },
+    {
+      "epoch": 3.7089487402258907,
+      "grad_norm": 0.07179631403204342,
+      "learning_rate": 1.2798361591275788e-07,
+      "loss": 0.4418,
+      "step": 7476
+    },
+    {
+      "epoch": 3.70944520292913,
+      "grad_norm": 0.07054594755591478,
+      "learning_rate": 1.2754455786104015e-07,
+      "loss": 0.4274,
+      "step": 7477
+    },
+    {
+      "epoch": 3.7099416656323694,
+      "grad_norm": 0.0753189956312603,
+      "learning_rate": 1.2710624448601216e-07,
+      "loss": 0.4701,
+      "step": 7478
+    },
+    {
+      "epoch": 3.7104381283356087,
+      "grad_norm": 0.074245085561885,
+      "learning_rate": 1.2666867585466426e-07,
+      "loss": 0.4503,
+      "step": 7479
+    },
+    {
+      "epoch": 3.710934591038848,
+      "grad_norm": 0.07609399666272167,
+      "learning_rate": 1.2623185203387124e-07,
+      "loss": 0.4751,
+      "step": 7480
+    },
+    {
+      "epoch": 3.7114310537420874,
+      "grad_norm": 0.0737868894777158,
+      "learning_rate": 1.2579577309039416e-07,
+      "loss": 0.4444,
+      "step": 7481
+    },
+    {
+      "epoch": 3.711927516445327,
+      "grad_norm": 0.07286497148501907,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.4249,
+      "step": 7482
+    },
+    {
+      "epoch": 3.7124239791485665,
+      "grad_norm": 0.07058913943698754,
+      "learning_rate": 1.249258501018674e-07,
+      "loss": 0.4333,
+      "step": 7483
+    },
+    {
+      "epoch": 3.712920441851806,
+      "grad_norm": 0.07179867347108665,
+      "learning_rate": 1.2449200618977087e-07,
+      "loss": 0.407,
+      "step": 7484
+    },
+    {
+      "epoch": 3.713416904555045,
+      "grad_norm": 0.07661219061733442,
+      "learning_rate": 1.2405890742089866e-07,
+      "loss": 0.4703,
+      "step": 7485
+    },
+    {
+      "epoch": 3.713913367258285,
+      "grad_norm": 0.0743605137172351,
+      "learning_rate": 1.2362655386144285e-07,
+      "loss": 0.469,
+      "step": 7486
+    },
+    {
+      "epoch": 3.7144098299615242,
+      "grad_norm": 0.07115634995620582,
+      "learning_rate": 1.2319494557748112e-07,
+      "loss": 0.3979,
+      "step": 7487
+    },
+    {
+      "epoch": 3.7149062926647636,
+      "grad_norm": 0.07471501232365835,
+      "learning_rate": 1.2276408263497796e-07,
+      "loss": 0.4321,
+      "step": 7488
+    },
+    {
+      "epoch": 3.715402755368003,
+      "grad_norm": 0.07250546434057126,
+      "learning_rate": 1.2233396509978513e-07,
+      "loss": 0.4548,
+      "step": 7489
+    },
+    {
+      "epoch": 3.7158992180712422,
+      "grad_norm": 0.07214773644289035,
+      "learning_rate": 1.2190459303763723e-07,
+      "loss": 0.4651,
+      "step": 7490
+    },
+    {
+      "epoch": 3.7163956807744816,
+      "grad_norm": 0.0718965188875489,
+      "learning_rate": 1.2147596651415906e-07,
+      "loss": 0.4139,
+      "step": 7491
+    },
+    {
+      "epoch": 3.7168921434777213,
+      "grad_norm": 0.07277477056759525,
+      "learning_rate": 1.2104808559485758e-07,
+      "loss": 0.4387,
+      "step": 7492
+    },
+    {
+      "epoch": 3.7173886061809607,
+      "grad_norm": 0.07317195671805793,
+      "learning_rate": 1.2062095034512832e-07,
+      "loss": 0.4226,
+      "step": 7493
+    },
+    {
+      "epoch": 3.7178850688842,
+      "grad_norm": 0.0733610221756854,
+      "learning_rate": 1.2019456083025184e-07,
+      "loss": 0.4299,
+      "step": 7494
+    },
+    {
+      "epoch": 3.7183815315874393,
+      "grad_norm": 0.07031093760031494,
+      "learning_rate": 1.1976891711539485e-07,
+      "loss": 0.4241,
+      "step": 7495
+    },
+    {
+      "epoch": 3.718877994290679,
+      "grad_norm": 0.0732484652938987,
+      "learning_rate": 1.193440192656109e-07,
+      "loss": 0.4521,
+      "step": 7496
+    },
+    {
+      "epoch": 3.7193744569939184,
+      "grad_norm": 0.0736922672961387,
+      "learning_rate": 1.1891986734583805e-07,
+      "loss": 0.4912,
+      "step": 7497
+    },
+    {
+      "epoch": 3.7198709196971578,
+      "grad_norm": 0.07318471954689998,
+      "learning_rate": 1.1849646142090054e-07,
+      "loss": 0.4442,
+      "step": 7498
+    },
+    {
+      "epoch": 3.720367382400397,
+      "grad_norm": 0.07406641358059599,
+      "learning_rate": 1.1807380155551107e-07,
+      "loss": 0.4461,
+      "step": 7499
+    },
+    {
+      "epoch": 3.7208638451036364,
+      "grad_norm": 0.07588040681542361,
+      "learning_rate": 1.1765188781426406e-07,
+      "loss": 0.4549,
+      "step": 7500
+    },
+    {
+      "epoch": 3.7213603078068758,
+      "grad_norm": 0.07167415013929496,
+      "learning_rate": 1.1723072026164462e-07,
+      "loss": 0.4408,
+      "step": 7501
+    },
+    {
+      "epoch": 3.7218567705101155,
+      "grad_norm": 0.07461061552940207,
+      "learning_rate": 1.1681029896202011e-07,
+      "loss": 0.4485,
+      "step": 7502
+    },
+    {
+      "epoch": 3.722353233213355,
+      "grad_norm": 0.07238913693556961,
+      "learning_rate": 1.163906239796453e-07,
+      "loss": 0.4396,
+      "step": 7503
+    },
+    {
+      "epoch": 3.722849695916594,
+      "grad_norm": 0.07380424841271917,
+      "learning_rate": 1.1597169537866104e-07,
+      "loss": 0.461,
+      "step": 7504
+    },
+    {
+      "epoch": 3.7233461586198335,
+      "grad_norm": 0.07294741460152432,
+      "learning_rate": 1.1555351322309339e-07,
+      "loss": 0.419,
+      "step": 7505
+    },
+    {
+      "epoch": 3.7238426213230733,
+      "grad_norm": 0.07201224370610604,
+      "learning_rate": 1.1513607757685508e-07,
+      "loss": 0.4185,
+      "step": 7506
+    },
+    {
+      "epoch": 3.7243390840263126,
+      "grad_norm": 0.07345882234607866,
+      "learning_rate": 1.1471938850374509e-07,
+      "loss": 0.4589,
+      "step": 7507
+    },
+    {
+      "epoch": 3.724835546729552,
+      "grad_norm": 0.07540856132370888,
+      "learning_rate": 1.1430344606744581e-07,
+      "loss": 0.4805,
+      "step": 7508
+    },
+    {
+      "epoch": 3.7253320094327913,
+      "grad_norm": 0.069747428732833,
+      "learning_rate": 1.1388825033152973e-07,
+      "loss": 0.4207,
+      "step": 7509
+    },
+    {
+      "epoch": 3.7258284721360306,
+      "grad_norm": 0.0723745653939202,
+      "learning_rate": 1.1347380135945108e-07,
+      "loss": 0.4215,
+      "step": 7510
+    },
+    {
+      "epoch": 3.72632493483927,
+      "grad_norm": 0.07205562171665118,
+      "learning_rate": 1.1306009921455307e-07,
+      "loss": 0.4283,
+      "step": 7511
+    },
+    {
+      "epoch": 3.7268213975425097,
+      "grad_norm": 0.07103586324604899,
+      "learning_rate": 1.1264714396006238e-07,
+      "loss": 0.4365,
+      "step": 7512
+    },
+    {
+      "epoch": 3.727317860245749,
+      "grad_norm": 0.07359648786780269,
+      "learning_rate": 1.1223493565909238e-07,
+      "loss": 0.426,
+      "step": 7513
+    },
+    {
+      "epoch": 3.7278143229489884,
+      "grad_norm": 0.0731775813226635,
+      "learning_rate": 1.1182347437464436e-07,
+      "loss": 0.3941,
+      "step": 7514
+    },
+    {
+      "epoch": 3.7283107856522277,
+      "grad_norm": 0.07398743238728402,
+      "learning_rate": 1.1141276016960134e-07,
+      "loss": 0.4517,
+      "step": 7515
+    },
+    {
+      "epoch": 3.7288072483554675,
+      "grad_norm": 0.07149681628651035,
+      "learning_rate": 1.110027931067359e-07,
+      "loss": 0.4327,
+      "step": 7516
+    },
+    {
+      "epoch": 3.729303711058707,
+      "grad_norm": 0.07604656685426203,
+      "learning_rate": 1.1059357324870456e-07,
+      "loss": 0.5081,
+      "step": 7517
+    },
+    {
+      "epoch": 3.729800173761946,
+      "grad_norm": 0.07339999849617872,
+      "learning_rate": 1.1018510065804954e-07,
+      "loss": 0.422,
+      "step": 7518
+    },
+    {
+      "epoch": 3.7302966364651855,
+      "grad_norm": 0.07454026734027142,
+      "learning_rate": 1.097773753971998e-07,
+      "loss": 0.4468,
+      "step": 7519
+    },
+    {
+      "epoch": 3.730793099168425,
+      "grad_norm": 0.07104200483397377,
+      "learning_rate": 1.0937039752846934e-07,
+      "loss": 0.4052,
+      "step": 7520
+    },
+    {
+      "epoch": 3.731289561871664,
+      "grad_norm": 0.07364091604071907,
+      "learning_rate": 1.0896416711405844e-07,
+      "loss": 0.476,
+      "step": 7521
+    },
+    {
+      "epoch": 3.731786024574904,
+      "grad_norm": 0.07229456530948636,
+      "learning_rate": 1.0855868421605242e-07,
+      "loss": 0.4509,
+      "step": 7522
+    },
+    {
+      "epoch": 3.7322824872781433,
+      "grad_norm": 0.07751160754378096,
+      "learning_rate": 1.0815394889642339e-07,
+      "loss": 0.4764,
+      "step": 7523
+    },
+    {
+      "epoch": 3.7327789499813826,
+      "grad_norm": 0.07244200225598892,
+      "learning_rate": 1.0774996121702907e-07,
+      "loss": 0.4451,
+      "step": 7524
+    },
+    {
+      "epoch": 3.733275412684622,
+      "grad_norm": 0.07498395699972732,
+      "learning_rate": 1.073467212396112e-07,
+      "loss": 0.4413,
+      "step": 7525
+    },
+    {
+      "epoch": 3.7337718753878617,
+      "grad_norm": 0.073491623792054,
+      "learning_rate": 1.0694422902579937e-07,
+      "loss": 0.4481,
+      "step": 7526
+    },
+    {
+      "epoch": 3.734268338091101,
+      "grad_norm": 0.07329100270406572,
+      "learning_rate": 1.0654248463710826e-07,
+      "loss": 0.4361,
+      "step": 7527
+    },
+    {
+      "epoch": 3.7347648007943404,
+      "grad_norm": 0.07118864173737534,
+      "learning_rate": 1.0614148813493764e-07,
+      "loss": 0.4608,
+      "step": 7528
+    },
+    {
+      "epoch": 3.7352612634975797,
+      "grad_norm": 0.0706601013127105,
+      "learning_rate": 1.0574123958057347e-07,
+      "loss": 0.4538,
+      "step": 7529
+    },
+    {
+      "epoch": 3.735757726200819,
+      "grad_norm": 0.07220402969613111,
+      "learning_rate": 1.053417390351874e-07,
+      "loss": 0.4513,
+      "step": 7530
+    },
+    {
+      "epoch": 3.7362541889040584,
+      "grad_norm": 0.07169252185126151,
+      "learning_rate": 1.049429865598367e-07,
+      "loss": 0.4262,
+      "step": 7531
+    },
+    {
+      "epoch": 3.736750651607298,
+      "grad_norm": 0.073324508656316,
+      "learning_rate": 1.0454498221546372e-07,
+      "loss": 0.4238,
+      "step": 7532
+    },
+    {
+      "epoch": 3.7372471143105375,
+      "grad_norm": 0.07198044107510718,
+      "learning_rate": 1.0414772606289814e-07,
+      "loss": 0.4454,
+      "step": 7533
+    },
+    {
+      "epoch": 3.737743577013777,
+      "grad_norm": 0.07118370980297188,
+      "learning_rate": 1.0375121816285361e-07,
+      "loss": 0.4057,
+      "step": 7534
+    },
+    {
+      "epoch": 3.738240039717016,
+      "grad_norm": 0.07160281716815883,
+      "learning_rate": 1.0335545857592999e-07,
+      "loss": 0.421,
+      "step": 7535
+    },
+    {
+      "epoch": 3.738736502420256,
+      "grad_norm": 0.07145401626454782,
+      "learning_rate": 1.0296044736261279e-07,
+      "loss": 0.4177,
+      "step": 7536
+    },
+    {
+      "epoch": 3.7392329651234952,
+      "grad_norm": 0.07322743586878572,
+      "learning_rate": 1.025661845832726e-07,
+      "loss": 0.4592,
+      "step": 7537
+    },
+    {
+      "epoch": 3.7397294278267346,
+      "grad_norm": 0.07544486411820084,
+      "learning_rate": 1.0217267029816736e-07,
+      "loss": 0.4377,
+      "step": 7538
+    },
+    {
+      "epoch": 3.740225890529974,
+      "grad_norm": 0.07049498098803661,
+      "learning_rate": 1.0177990456743835e-07,
+      "loss": 0.4324,
+      "step": 7539
+    },
+    {
+      "epoch": 3.7407223532332132,
+      "grad_norm": 0.07235376143354762,
+      "learning_rate": 1.0138788745111427e-07,
+      "loss": 0.4507,
+      "step": 7540
+    },
+    {
+      "epoch": 3.7412188159364526,
+      "grad_norm": 0.07288344961458237,
+      "learning_rate": 1.0099661900910829e-07,
+      "loss": 0.4552,
+      "step": 7541
+    },
+    {
+      "epoch": 3.7417152786396923,
+      "grad_norm": 0.07193444729394458,
+      "learning_rate": 1.0060609930121923e-07,
+      "loss": 0.4278,
+      "step": 7542
+    },
+    {
+      "epoch": 3.7422117413429317,
+      "grad_norm": 0.07227753500713993,
+      "learning_rate": 1.0021632838713213e-07,
+      "loss": 0.4435,
+      "step": 7543
+    },
+    {
+      "epoch": 3.742708204046171,
+      "grad_norm": 0.07170481264345718,
+      "learning_rate": 9.982730632641768e-08,
+      "loss": 0.4204,
+      "step": 7544
+    },
+    {
+      "epoch": 3.7432046667494103,
+      "grad_norm": 0.07643936687734741,
+      "learning_rate": 9.943903317853055e-08,
+      "loss": 0.457,
+      "step": 7545
+    },
+    {
+      "epoch": 3.74370112945265,
+      "grad_norm": 0.07330117834237307,
+      "learning_rate": 9.905150900281325e-08,
+      "loss": 0.4408,
+      "step": 7546
+    },
+    {
+      "epoch": 3.7441975921558894,
+      "grad_norm": 0.07001701440144002,
+      "learning_rate": 9.866473385849117e-08,
+      "loss": 0.3988,
+      "step": 7547
+    },
+    {
+      "epoch": 3.7446940548591288,
+      "grad_norm": 0.07348218035906592,
+      "learning_rate": 9.827870780467819e-08,
+      "loss": 0.4546,
+      "step": 7548
+    },
+    {
+      "epoch": 3.745190517562368,
+      "grad_norm": 0.07149853576907274,
+      "learning_rate": 9.789343090037207e-08,
+      "loss": 0.4354,
+      "step": 7549
+    },
+    {
+      "epoch": 3.7456869802656074,
+      "grad_norm": 0.07237011239351258,
+      "learning_rate": 9.750890320445517e-08,
+      "loss": 0.4282,
+      "step": 7550
+    },
+    {
+      "epoch": 3.7461834429688468,
+      "grad_norm": 0.069848514948421,
+      "learning_rate": 9.712512477569713e-08,
+      "loss": 0.4113,
+      "step": 7551
+    },
+    {
+      "epoch": 3.7466799056720865,
+      "grad_norm": 0.07450332694516261,
+      "learning_rate": 9.674209567275161e-08,
+      "loss": 0.4375,
+      "step": 7552
+    },
+    {
+      "epoch": 3.747176368375326,
+      "grad_norm": 0.07318998690584826,
+      "learning_rate": 9.635981595415955e-08,
+      "loss": 0.4656,
+      "step": 7553
+    },
+    {
+      "epoch": 3.747672831078565,
+      "grad_norm": 0.07210166747045929,
+      "learning_rate": 9.597828567834589e-08,
+      "loss": 0.4431,
+      "step": 7554
+    },
+    {
+      "epoch": 3.7481692937818045,
+      "grad_norm": 0.07136638830380382,
+      "learning_rate": 9.559750490362063e-08,
+      "loss": 0.4189,
+      "step": 7555
+    },
+    {
+      "epoch": 3.7486657564850443,
+      "grad_norm": 0.07291966530697377,
+      "learning_rate": 9.521747368818112e-08,
+      "loss": 0.4365,
+      "step": 7556
+    },
+    {
+      "epoch": 3.7491622191882836,
+      "grad_norm": 0.07022951970666523,
+      "learning_rate": 9.483819209010813e-08,
+      "loss": 0.4128,
+      "step": 7557
+    },
+    {
+      "epoch": 3.749658681891523,
+      "grad_norm": 0.07527555997003625,
+      "learning_rate": 9.445966016736974e-08,
+      "loss": 0.496,
+      "step": 7558
+    },
+    {
+      "epoch": 3.7501551445947623,
+      "grad_norm": 0.0759036245963917,
+      "learning_rate": 9.408187797781743e-08,
+      "loss": 0.4411,
+      "step": 7559
+    },
+    {
+      "epoch": 3.7506516072980016,
+      "grad_norm": 0.07328521345891655,
+      "learning_rate": 9.370484557919002e-08,
+      "loss": 0.4489,
+      "step": 7560
+    },
+    {
+      "epoch": 3.7506516072980016,
+      "eval_loss": 0.5161105394363403,
+      "eval_runtime": 258.8004,
+      "eval_samples_per_second": 117.283,
+      "eval_steps_per_second": 14.664,
+      "step": 7560
+    },
+    {
+      "epoch": 3.751148070001241,
+      "grad_norm": 0.07384410860926094,
+      "learning_rate": 9.33285630291103e-08,
+      "loss": 0.4631,
+      "step": 7561
+    },
+    {
+      "epoch": 3.7516445327044807,
+      "grad_norm": 0.07337397188656725,
+      "learning_rate": 9.295303038508729e-08,
+      "loss": 0.4382,
+      "step": 7562
+    },
+    {
+      "epoch": 3.75214099540772,
+      "grad_norm": 0.07247752832773147,
+      "learning_rate": 9.257824770451507e-08,
+      "loss": 0.4332,
+      "step": 7563
+    },
+    {
+      "epoch": 3.7526374581109594,
+      "grad_norm": 0.07263876716289892,
+      "learning_rate": 9.22042150446728e-08,
+      "loss": 0.461,
+      "step": 7564
+    },
+    {
+      "epoch": 3.7531339208141987,
+      "grad_norm": 0.07229693840018254,
+      "learning_rate": 9.183093246272645e-08,
+      "loss": 0.4416,
+      "step": 7565
+    },
+    {
+      "epoch": 3.7536303835174385,
+      "grad_norm": 0.07183061018007368,
+      "learning_rate": 9.145840001572537e-08,
+      "loss": 0.4415,
+      "step": 7566
+    },
+    {
+      "epoch": 3.754126846220678,
+      "grad_norm": 0.07461055074980504,
+      "learning_rate": 9.108661776060568e-08,
+      "loss": 0.4713,
+      "step": 7567
+    },
+    {
+      "epoch": 3.754623308923917,
+      "grad_norm": 0.07409143152983287,
+      "learning_rate": 9.071558575418749e-08,
+      "loss": 0.451,
+      "step": 7568
+    },
+    {
+      "epoch": 3.7551197716271565,
+      "grad_norm": 0.07229577276590579,
+      "learning_rate": 9.034530405317765e-08,
+      "loss": 0.4364,
+      "step": 7569
+    },
+    {
+      "epoch": 3.755616234330396,
+      "grad_norm": 0.07086020583436162,
+      "learning_rate": 8.997577271416758e-08,
+      "loss": 0.4171,
+      "step": 7570
+    },
+    {
+      "epoch": 3.756112697033635,
+      "grad_norm": 0.07011907855229332,
+      "learning_rate": 8.96069917936343e-08,
+      "loss": 0.421,
+      "step": 7571
+    },
+    {
+      "epoch": 3.756609159736875,
+      "grad_norm": 0.07276329677023631,
+      "learning_rate": 8.923896134794053e-08,
+      "loss": 0.4565,
+      "step": 7572
+    },
+    {
+      "epoch": 3.7571056224401143,
+      "grad_norm": 0.07401906530715995,
+      "learning_rate": 8.887168143333402e-08,
+      "loss": 0.43,
+      "step": 7573
+    },
+    {
+      "epoch": 3.7576020851433536,
+      "grad_norm": 0.0726548623776061,
+      "learning_rate": 8.850515210594601e-08,
+      "loss": 0.4225,
+      "step": 7574
+    },
+    {
+      "epoch": 3.758098547846593,
+      "grad_norm": 0.07209307721520765,
+      "learning_rate": 8.813937342179613e-08,
+      "loss": 0.3859,
+      "step": 7575
+    },
+    {
+      "epoch": 3.7585950105498327,
+      "grad_norm": 0.07235868664387755,
+      "learning_rate": 8.777434543678687e-08,
+      "loss": 0.4178,
+      "step": 7576
+    },
+    {
+      "epoch": 3.759091473253072,
+      "grad_norm": 0.07384631212067913,
+      "learning_rate": 8.741006820670805e-08,
+      "loss": 0.4502,
+      "step": 7577
+    },
+    {
+      "epoch": 3.7595879359563114,
+      "grad_norm": 0.07160271493652788,
+      "learning_rate": 8.704654178723293e-08,
+      "loss": 0.4224,
+      "step": 7578
+    },
+    {
+      "epoch": 3.7600843986595507,
+      "grad_norm": 0.07175008162985384,
+      "learning_rate": 8.668376623391983e-08,
+      "loss": 0.4291,
+      "step": 7579
+    },
+    {
+      "epoch": 3.76058086136279,
+      "grad_norm": 0.071081695755974,
+      "learning_rate": 8.632174160221496e-08,
+      "loss": 0.4432,
+      "step": 7580
+    },
+    {
+      "epoch": 3.7610773240660293,
+      "grad_norm": 0.07161415164284515,
+      "learning_rate": 8.596046794744683e-08,
+      "loss": 0.447,
+      "step": 7581
+    },
+    {
+      "epoch": 3.761573786769269,
+      "grad_norm": 0.07457076039250243,
+      "learning_rate": 8.559994532483074e-08,
+      "loss": 0.4739,
+      "step": 7582
+    },
+    {
+      "epoch": 3.7620702494725085,
+      "grad_norm": 0.0705624068638462,
+      "learning_rate": 8.524017378946592e-08,
+      "loss": 0.4159,
+      "step": 7583
+    },
+    {
+      "epoch": 3.762566712175748,
+      "grad_norm": 0.07177507462194573,
+      "learning_rate": 8.48811533963384e-08,
+      "loss": 0.435,
+      "step": 7584
+    },
+    {
+      "epoch": 3.763063174878987,
+      "grad_norm": 0.07268354722345835,
+      "learning_rate": 8.452288420031929e-08,
+      "loss": 0.4301,
+      "step": 7585
+    },
+    {
+      "epoch": 3.763559637582227,
+      "grad_norm": 0.07380471538302204,
+      "learning_rate": 8.416536625616312e-08,
+      "loss": 0.4226,
+      "step": 7586
+    },
+    {
+      "epoch": 3.764056100285466,
+      "grad_norm": 0.07641290940010709,
+      "learning_rate": 8.380859961851174e-08,
+      "loss": 0.444,
+      "step": 7587
+    },
+    {
+      "epoch": 3.7645525629887056,
+      "grad_norm": 0.07167215516692764,
+      "learning_rate": 8.345258434189041e-08,
+      "loss": 0.44,
+      "step": 7588
+    },
+    {
+      "epoch": 3.765049025691945,
+      "grad_norm": 0.07420714066871849,
+      "learning_rate": 8.309732048071062e-08,
+      "loss": 0.449,
+      "step": 7589
+    },
+    {
+      "epoch": 3.765545488395184,
+      "grad_norm": 0.07177597297596648,
+      "learning_rate": 8.274280808926893e-08,
+      "loss": 0.404,
+      "step": 7590
+    },
+    {
+      "epoch": 3.7660419510984235,
+      "grad_norm": 0.07439229401969116,
+      "learning_rate": 8.238904722174701e-08,
+      "loss": 0.4384,
+      "step": 7591
+    },
+    {
+      "epoch": 3.766538413801663,
+      "grad_norm": 0.07206472800094606,
+      "learning_rate": 8.203603793221104e-08,
+      "loss": 0.4363,
+      "step": 7592
+    },
+    {
+      "epoch": 3.7670348765049027,
+      "grad_norm": 0.07751943428429821,
+      "learning_rate": 8.168378027461343e-08,
+      "loss": 0.4822,
+      "step": 7593
+    },
+    {
+      "epoch": 3.767531339208142,
+      "grad_norm": 0.07116288444682155,
+      "learning_rate": 8.133227430279055e-08,
+      "loss": 0.4153,
+      "step": 7594
+    },
+    {
+      "epoch": 3.7680278019113813,
+      "grad_norm": 0.07418819356407474,
+      "learning_rate": 8.0981520070465e-08,
+      "loss": 0.4585,
+      "step": 7595
+    },
+    {
+      "epoch": 3.768524264614621,
+      "grad_norm": 0.0733105625391331,
+      "learning_rate": 8.063151763124332e-08,
+      "loss": 0.4455,
+      "step": 7596
+    },
+    {
+      "epoch": 3.7690207273178604,
+      "grad_norm": 0.07300590027724989,
+      "learning_rate": 8.02822670386183e-08,
+      "loss": 0.4423,
+      "step": 7597
+    },
+    {
+      "epoch": 3.7695171900210998,
+      "grad_norm": 0.07357361670187289,
+      "learning_rate": 7.993376834596722e-08,
+      "loss": 0.4414,
+      "step": 7598
+    },
+    {
+      "epoch": 3.770013652724339,
+      "grad_norm": 0.07297895745552392,
+      "learning_rate": 7.958602160655193e-08,
+      "loss": 0.4465,
+      "step": 7599
+    },
+    {
+      "epoch": 3.7705101154275784,
+      "grad_norm": 0.0743257455513542,
+      "learning_rate": 7.923902687352103e-08,
+      "loss": 0.4503,
+      "step": 7600
+    },
+    {
+      "epoch": 3.7710065781308177,
+      "grad_norm": 0.07258222433746735,
+      "learning_rate": 7.889278419990598e-08,
+      "loss": 0.4757,
+      "step": 7601
+    },
+    {
+      "epoch": 3.771503040834057,
+      "grad_norm": 0.07186680285906648,
+      "learning_rate": 7.854729363862502e-08,
+      "loss": 0.4384,
+      "step": 7602
+    },
+    {
+      "epoch": 3.771999503537297,
+      "grad_norm": 0.07320648320061995,
+      "learning_rate": 7.820255524248032e-08,
+      "loss": 0.4455,
+      "step": 7603
+    },
+    {
+      "epoch": 3.772495966240536,
+      "grad_norm": 0.07359622206653958,
+      "learning_rate": 7.78585690641609e-08,
+      "loss": 0.4377,
+      "step": 7604
+    },
+    {
+      "epoch": 3.7729924289437755,
+      "grad_norm": 0.07749700381908971,
+      "learning_rate": 7.7515335156238e-08,
+      "loss": 0.4363,
+      "step": 7605
+    },
+    {
+      "epoch": 3.7734888916470153,
+      "grad_norm": 0.07211493802132227,
+      "learning_rate": 7.717285357117022e-08,
+      "loss": 0.4297,
+      "step": 7606
+    },
+    {
+      "epoch": 3.7739853543502546,
+      "grad_norm": 0.07104624651291085,
+      "learning_rate": 7.68311243613007e-08,
+      "loss": 0.4409,
+      "step": 7607
+    },
+    {
+      "epoch": 3.774481817053494,
+      "grad_norm": 0.0726414983389291,
+      "learning_rate": 7.649014757885597e-08,
+      "loss": 0.4284,
+      "step": 7608
+    },
+    {
+      "epoch": 3.7749782797567333,
+      "grad_norm": 0.07053503146406671,
+      "learning_rate": 7.614992327594994e-08,
+      "loss": 0.4136,
+      "step": 7609
+    },
+    {
+      "epoch": 3.7754747424599726,
+      "grad_norm": 0.0748476111263376,
+      "learning_rate": 7.581045150458099e-08,
+      "loss": 0.4532,
+      "step": 7610
+    },
+    {
+      "epoch": 3.775971205163212,
+      "grad_norm": 0.07253655365457137,
+      "learning_rate": 7.547173231663041e-08,
+      "loss": 0.4345,
+      "step": 7611
+    },
+    {
+      "epoch": 3.7764676678664513,
+      "grad_norm": 0.0714135317327364,
+      "learning_rate": 7.513376576386678e-08,
+      "loss": 0.4414,
+      "step": 7612
+    },
+    {
+      "epoch": 3.776964130569691,
+      "grad_norm": 0.07205257695786416,
+      "learning_rate": 7.479655189794266e-08,
+      "loss": 0.4374,
+      "step": 7613
+    },
+    {
+      "epoch": 3.7774605932729304,
+      "grad_norm": 0.07257848974462944,
+      "learning_rate": 7.446009077039629e-08,
+      "loss": 0.4548,
+      "step": 7614
+    },
+    {
+      "epoch": 3.7779570559761697,
+      "grad_norm": 0.07242154741962721,
+      "learning_rate": 7.41243824326504e-08,
+      "loss": 0.4307,
+      "step": 7615
+    },
+    {
+      "epoch": 3.7784535186794095,
+      "grad_norm": 0.0730945396679734,
+      "learning_rate": 7.378942693601177e-08,
+      "loss": 0.4675,
+      "step": 7616
+    },
+    {
+      "epoch": 3.778949981382649,
+      "grad_norm": 0.07274778188236991,
+      "learning_rate": 7.345522433167385e-08,
+      "loss": 0.4497,
+      "step": 7617
+    },
+    {
+      "epoch": 3.779446444085888,
+      "grad_norm": 0.07027696943602645,
+      "learning_rate": 7.312177467071302e-08,
+      "loss": 0.4272,
+      "step": 7618
+    },
+    {
+      "epoch": 3.7799429067891275,
+      "grad_norm": 0.07173677048643133,
+      "learning_rate": 7.278907800409296e-08,
+      "loss": 0.4569,
+      "step": 7619
+    },
+    {
+      "epoch": 3.780439369492367,
+      "grad_norm": 0.0713370611452363,
+      "learning_rate": 7.245713438266022e-08,
+      "loss": 0.4181,
+      "step": 7620
+    },
+    {
+      "epoch": 3.780935832195606,
+      "grad_norm": 0.07240382585870658,
+      "learning_rate": 7.212594385714755e-08,
+      "loss": 0.4274,
+      "step": 7621
+    },
+    {
+      "epoch": 3.7814322948988455,
+      "grad_norm": 0.0729393270005386,
+      "learning_rate": 7.179550647817224e-08,
+      "loss": 0.4517,
+      "step": 7622
+    },
+    {
+      "epoch": 3.7819287576020852,
+      "grad_norm": 0.07434877364398472,
+      "learning_rate": 7.1465822296235e-08,
+      "loss": 0.4713,
+      "step": 7623
+    },
+    {
+      "epoch": 3.7824252203053246,
+      "grad_norm": 0.07110157033397635,
+      "learning_rate": 7.113689136172441e-08,
+      "loss": 0.4262,
+      "step": 7624
+    },
+    {
+      "epoch": 3.782921683008564,
+      "grad_norm": 0.07200827074812058,
+      "learning_rate": 7.080871372491193e-08,
+      "loss": 0.4307,
+      "step": 7625
+    },
+    {
+      "epoch": 3.7834181457118032,
+      "grad_norm": 0.07128597824580814,
+      "learning_rate": 7.048128943595356e-08,
+      "loss": 0.4464,
+      "step": 7626
+    },
+    {
+      "epoch": 3.783914608415043,
+      "grad_norm": 0.07378865528886532,
+      "learning_rate": 7.015461854489148e-08,
+      "loss": 0.4447,
+      "step": 7627
+    },
+    {
+      "epoch": 3.7844110711182823,
+      "grad_norm": 0.07477304509027059,
+      "learning_rate": 6.982870110165185e-08,
+      "loss": 0.4768,
+      "step": 7628
+    },
+    {
+      "epoch": 3.7849075338215217,
+      "grad_norm": 0.07292149129680923,
+      "learning_rate": 6.950353715604597e-08,
+      "loss": 0.4242,
+      "step": 7629
+    },
+    {
+      "epoch": 3.785403996524761,
+      "grad_norm": 0.07333502503957316,
+      "learning_rate": 6.917912675777016e-08,
+      "loss": 0.4362,
+      "step": 7630
+    },
+    {
+      "epoch": 3.7859004592280003,
+      "grad_norm": 0.07190268268817648,
+      "learning_rate": 6.885546995640479e-08,
+      "loss": 0.4507,
+      "step": 7631
+    },
+    {
+      "epoch": 3.7863969219312397,
+      "grad_norm": 0.07061760849169589,
+      "learning_rate": 6.85325668014164e-08,
+      "loss": 0.4109,
+      "step": 7632
+    },
+    {
+      "epoch": 3.7868933846344794,
+      "grad_norm": 0.07427625826680745,
+      "learning_rate": 6.821041734215438e-08,
+      "loss": 0.4956,
+      "step": 7633
+    },
+    {
+      "epoch": 3.7873898473377188,
+      "grad_norm": 0.07358870722147563,
+      "learning_rate": 6.788902162785549e-08,
+      "loss": 0.4449,
+      "step": 7634
+    },
+    {
+      "epoch": 3.787886310040958,
+      "grad_norm": 0.0720882277469726,
+      "learning_rate": 6.756837970763875e-08,
+      "loss": 0.4546,
+      "step": 7635
+    },
+    {
+      "epoch": 3.7883827727441974,
+      "grad_norm": 0.07358476569511135,
+      "learning_rate": 6.724849163050995e-08,
+      "loss": 0.4937,
+      "step": 7636
+    },
+    {
+      "epoch": 3.788879235447437,
+      "grad_norm": 0.07112676903545612,
+      "learning_rate": 6.692935744535889e-08,
+      "loss": 0.4274,
+      "step": 7637
+    },
+    {
+      "epoch": 3.7893756981506765,
+      "grad_norm": 0.07339375287303061,
+      "learning_rate": 6.661097720095877e-08,
+      "loss": 0.4433,
+      "step": 7638
+    },
+    {
+      "epoch": 3.789872160853916,
+      "grad_norm": 0.07241100188549718,
+      "learning_rate": 6.629335094597067e-08,
+      "loss": 0.4727,
+      "step": 7639
+    },
+    {
+      "epoch": 3.790368623557155,
+      "grad_norm": 0.07495755848213913,
+      "learning_rate": 6.597647872893798e-08,
+      "loss": 0.4613,
+      "step": 7640
+    },
+    {
+      "epoch": 3.7908650862603945,
+      "grad_norm": 0.07058571062118439,
+      "learning_rate": 6.566036059828918e-08,
+      "loss": 0.4577,
+      "step": 7641
+    },
+    {
+      "epoch": 3.791361548963634,
+      "grad_norm": 0.07198596199162723,
+      "learning_rate": 6.53449966023384e-08,
+      "loss": 0.4468,
+      "step": 7642
+    },
+    {
+      "epoch": 3.7918580116668736,
+      "grad_norm": 0.07094089700135145,
+      "learning_rate": 6.50303867892832e-08,
+      "loss": 0.4782,
+      "step": 7643
+    },
+    {
+      "epoch": 3.792354474370113,
+      "grad_norm": 0.0722320181545589,
+      "learning_rate": 6.471653120720733e-08,
+      "loss": 0.4428,
+      "step": 7644
+    },
+    {
+      "epoch": 3.7928509370733523,
+      "grad_norm": 0.07167870597974875,
+      "learning_rate": 6.440342990407856e-08,
+      "loss": 0.4252,
+      "step": 7645
+    },
+    {
+      "epoch": 3.7933473997765916,
+      "grad_norm": 0.07092657802659981,
+      "learning_rate": 6.409108292774912e-08,
+      "loss": 0.4242,
+      "step": 7646
+    },
+    {
+      "epoch": 3.7938438624798314,
+      "grad_norm": 0.0755376186724099,
+      "learning_rate": 6.377949032595699e-08,
+      "loss": 0.4451,
+      "step": 7647
+    },
+    {
+      "epoch": 3.7943403251830707,
+      "grad_norm": 0.07150643420264564,
+      "learning_rate": 6.346865214632292e-08,
+      "loss": 0.4109,
+      "step": 7648
+    },
+    {
+      "epoch": 3.79483678788631,
+      "grad_norm": 0.07421329049090035,
+      "learning_rate": 6.315856843635449e-08,
+      "loss": 0.4141,
+      "step": 7649
+    },
+    {
+      "epoch": 3.7953332505895494,
+      "grad_norm": 0.07365889529760761,
+      "learning_rate": 6.28492392434421e-08,
+      "loss": 0.4379,
+      "step": 7650
+    },
+    {
+      "epoch": 3.7958297132927887,
+      "grad_norm": 0.07265832592753993,
+      "learning_rate": 6.254066461486241e-08,
+      "loss": 0.4228,
+      "step": 7651
+    },
+    {
+      "epoch": 3.796326175996028,
+      "grad_norm": 0.07236918113007618,
+      "learning_rate": 6.223284459777601e-08,
+      "loss": 0.4388,
+      "step": 7652
+    },
+    {
+      "epoch": 3.796822638699268,
+      "grad_norm": 0.07135326402955866,
+      "learning_rate": 6.19257792392286e-08,
+      "loss": 0.4314,
+      "step": 7653
+    },
+    {
+      "epoch": 3.797319101402507,
+      "grad_norm": 0.07192211849326689,
+      "learning_rate": 6.161946858614931e-08,
+      "loss": 0.4284,
+      "step": 7654
+    },
+    {
+      "epoch": 3.7978155641057465,
+      "grad_norm": 0.07425977862800254,
+      "learning_rate": 6.131391268535347e-08,
+      "loss": 0.4783,
+      "step": 7655
+    },
+    {
+      "epoch": 3.798312026808986,
+      "grad_norm": 0.07256891885712644,
+      "learning_rate": 6.100911158354039e-08,
+      "loss": 0.4139,
+      "step": 7656
+    },
+    {
+      "epoch": 3.7988084895122256,
+      "grad_norm": 0.07132789042499738,
+      "learning_rate": 6.070506532729393e-08,
+      "loss": 0.4487,
+      "step": 7657
+    },
+    {
+      "epoch": 3.799304952215465,
+      "grad_norm": 0.071378541512784,
+      "learning_rate": 6.040177396308244e-08,
+      "loss": 0.4264,
+      "step": 7658
+    },
+    {
+      "epoch": 3.7998014149187043,
+      "grad_norm": 0.0725960156414164,
+      "learning_rate": 6.009923753725999e-08,
+      "loss": 0.4477,
+      "step": 7659
+    },
+    {
+      "epoch": 3.8002978776219436,
+      "grad_norm": 0.0736973528236666,
+      "learning_rate": 5.979745609606291e-08,
+      "loss": 0.4299,
+      "step": 7660
+    },
+    {
+      "epoch": 3.800794340325183,
+      "grad_norm": 0.07299851087260749,
+      "learning_rate": 5.949642968561542e-08,
+      "loss": 0.4331,
+      "step": 7661
+    },
+    {
+      "epoch": 3.8012908030284223,
+      "grad_norm": 0.07473707812315147,
+      "learning_rate": 5.9196158351923496e-08,
+      "loss": 0.4818,
+      "step": 7662
+    },
+    {
+      "epoch": 3.801787265731662,
+      "grad_norm": 0.07336642968051621,
+      "learning_rate": 5.889664214087876e-08,
+      "loss": 0.4387,
+      "step": 7663
+    },
+    {
+      "epoch": 3.8022837284349014,
+      "grad_norm": 0.07245571259074714,
+      "learning_rate": 5.8597881098257924e-08,
+      "loss": 0.441,
+      "step": 7664
+    },
+    {
+      "epoch": 3.8027801911381407,
+      "grad_norm": 0.07236547353401815,
+      "learning_rate": 5.829987526972114e-08,
+      "loss": 0.4435,
+      "step": 7665
+    },
+    {
+      "epoch": 3.80327665384138,
+      "grad_norm": 0.07503876585403813,
+      "learning_rate": 5.8002624700814744e-08,
+      "loss": 0.4763,
+      "step": 7666
+    },
+    {
+      "epoch": 3.80377311654462,
+      "grad_norm": 0.0732106784277791,
+      "learning_rate": 5.770612943696741e-08,
+      "loss": 0.4082,
+      "step": 7667
+    },
+    {
+      "epoch": 3.804269579247859,
+      "grad_norm": 0.07331334362720807,
+      "learning_rate": 5.741038952349565e-08,
+      "loss": 0.435,
+      "step": 7668
+    },
+    {
+      "epoch": 3.8047660419510985,
+      "grad_norm": 0.07081532219477497,
+      "learning_rate": 5.71154050055972e-08,
+      "loss": 0.4333,
+      "step": 7669
+    },
+    {
+      "epoch": 3.805262504654338,
+      "grad_norm": 0.0729540189307146,
+      "learning_rate": 5.682117592835545e-08,
+      "loss": 0.439,
+      "step": 7670
+    },
+    {
+      "epoch": 3.805758967357577,
+      "grad_norm": 0.07107821559478245,
+      "learning_rate": 5.652770233673943e-08,
+      "loss": 0.4316,
+      "step": 7671
+    },
+    {
+      "epoch": 3.8062554300608165,
+      "grad_norm": 0.0712561932838854,
+      "learning_rate": 5.623498427560159e-08,
+      "loss": 0.4298,
+      "step": 7672
+    },
+    {
+      "epoch": 3.8067518927640562,
+      "grad_norm": 0.07389039590125981,
+      "learning_rate": 5.594302178967892e-08,
+      "loss": 0.4223,
+      "step": 7673
+    },
+    {
+      "epoch": 3.8072483554672956,
+      "grad_norm": 0.07327860460759891,
+      "learning_rate": 5.565181492359406e-08,
+      "loss": 0.4317,
+      "step": 7674
+    },
+    {
+      "epoch": 3.807744818170535,
+      "grad_norm": 0.07032477492674219,
+      "learning_rate": 5.536136372185197e-08,
+      "loss": 0.418,
+      "step": 7675
+    },
+    {
+      "epoch": 3.8082412808737742,
+      "grad_norm": 0.07096786204176252,
+      "learning_rate": 5.507166822884435e-08,
+      "loss": 0.4008,
+      "step": 7676
+    },
+    {
+      "epoch": 3.808737743577014,
+      "grad_norm": 0.07408752101095577,
+      "learning_rate": 5.478272848884636e-08,
+      "loss": 0.4616,
+      "step": 7677
+    },
+    {
+      "epoch": 3.8092342062802533,
+      "grad_norm": 0.0728659361480815,
+      "learning_rate": 5.4494544546018216e-08,
+      "loss": 0.4349,
+      "step": 7678
+    },
+    {
+      "epoch": 3.8097306689834927,
+      "grad_norm": 0.07048801473333352,
+      "learning_rate": 5.420711644440357e-08,
+      "loss": 0.4175,
+      "step": 7679
+    },
+    {
+      "epoch": 3.810227131686732,
+      "grad_norm": 0.07360760755959432,
+      "learning_rate": 5.392044422793119e-08,
+      "loss": 0.4632,
+      "step": 7680
+    },
+    {
+      "epoch": 3.8107235943899713,
+      "grad_norm": 0.07173217601575797,
+      "learning_rate": 5.3634527940414903e-08,
+      "loss": 0.4181,
+      "step": 7681
+    },
+    {
+      "epoch": 3.8112200570932107,
+      "grad_norm": 0.07142680853692407,
+      "learning_rate": 5.334936762555198e-08,
+      "loss": 0.4561,
+      "step": 7682
+    },
+    {
+      "epoch": 3.8117165197964504,
+      "grad_norm": 0.07222635525660184,
+      "learning_rate": 5.3064963326924235e-08,
+      "loss": 0.4435,
+      "step": 7683
+    },
+    {
+      "epoch": 3.8122129824996898,
+      "grad_norm": 0.07122565629989039,
+      "learning_rate": 5.2781315087999106e-08,
+      "loss": 0.4317,
+      "step": 7684
+    },
+    {
+      "epoch": 3.812709445202929,
+      "grad_norm": 0.07193918182893863,
+      "learning_rate": 5.249842295212748e-08,
+      "loss": 0.4184,
+      "step": 7685
+    },
+    {
+      "epoch": 3.8132059079061684,
+      "grad_norm": 0.06929577784349823,
+      "learning_rate": 5.221628696254477e-08,
+      "loss": 0.3988,
+      "step": 7686
+    },
+    {
+      "epoch": 3.813702370609408,
+      "grad_norm": 0.07224933484530108,
+      "learning_rate": 5.1934907162370374e-08,
+      "loss": 0.441,
+      "step": 7687
+    },
+    {
+      "epoch": 3.8141988333126475,
+      "grad_norm": 0.07296837062153164,
+      "learning_rate": 5.165428359460989e-08,
+      "loss": 0.4522,
+      "step": 7688
+    },
+    {
+      "epoch": 3.814695296015887,
+      "grad_norm": 0.0737214515339532,
+      "learning_rate": 5.137441630215123e-08,
+      "loss": 0.4502,
+      "step": 7689
+    },
+    {
+      "epoch": 3.815191758719126,
+      "grad_norm": 0.07404136133329889,
+      "learning_rate": 5.1095305327767965e-08,
+      "loss": 0.469,
+      "step": 7690
+    },
+    {
+      "epoch": 3.8156882214223655,
+      "grad_norm": 0.07376348125016616,
+      "learning_rate": 5.081695071411763e-08,
+      "loss": 0.4454,
+      "step": 7691
+    },
+    {
+      "epoch": 3.816184684125605,
+      "grad_norm": 0.07453116586826475,
+      "learning_rate": 5.0539352503741756e-08,
+      "loss": 0.4473,
+      "step": 7692
+    },
+    {
+      "epoch": 3.8166811468288446,
+      "grad_norm": 0.07315513997164763,
+      "learning_rate": 5.026251073906807e-08,
+      "loss": 0.4248,
+      "step": 7693
+    },
+    {
+      "epoch": 3.817177609532084,
+      "grad_norm": 0.0753470001876728,
+      "learning_rate": 4.998642546240606e-08,
+      "loss": 0.4602,
+      "step": 7694
+    },
+    {
+      "epoch": 3.8176740722353233,
+      "grad_norm": 0.07377153988075652,
+      "learning_rate": 4.9711096715951977e-08,
+      "loss": 0.4182,
+      "step": 7695
+    },
+    {
+      "epoch": 3.8181705349385626,
+      "grad_norm": 0.07252299098795054,
+      "learning_rate": 4.9436524541784384e-08,
+      "loss": 0.4297,
+      "step": 7696
+    },
+    {
+      "epoch": 3.8186669976418024,
+      "grad_norm": 0.0750530077316559,
+      "learning_rate": 4.9162708981868034e-08,
+      "loss": 0.4418,
+      "step": 7697
+    },
+    {
+      "epoch": 3.8191634603450417,
+      "grad_norm": 0.07255075825129673,
+      "learning_rate": 4.888965007805113e-08,
+      "loss": 0.4417,
+      "step": 7698
+    },
+    {
+      "epoch": 3.819659923048281,
+      "grad_norm": 0.07321503165402111,
+      "learning_rate": 4.861734787206529e-08,
+      "loss": 0.4585,
+      "step": 7699
+    },
+    {
+      "epoch": 3.8201563857515204,
+      "grad_norm": 0.0747663686249205,
+      "learning_rate": 4.834580240552944e-08,
+      "loss": 0.4267,
+      "step": 7700
+    },
+    {
+      "epoch": 3.8206528484547597,
+      "grad_norm": 0.070900126955607,
+      "learning_rate": 4.807501371994372e-08,
+      "loss": 0.4307,
+      "step": 7701
+    },
+    {
+      "epoch": 3.821149311157999,
+      "grad_norm": 0.07040814280606325,
+      "learning_rate": 4.780498185669391e-08,
+      "loss": 0.4374,
+      "step": 7702
+    },
+    {
+      "epoch": 3.821645773861239,
+      "grad_norm": 0.06980323783933094,
+      "learning_rate": 4.753570685705033e-08,
+      "loss": 0.4215,
+      "step": 7703
+    },
+    {
+      "epoch": 3.822142236564478,
+      "grad_norm": 0.0737712431960967,
+      "learning_rate": 4.726718876216674e-08,
+      "loss": 0.4528,
+      "step": 7704
+    },
+    {
+      "epoch": 3.8226386992677175,
+      "grad_norm": 0.06996481879965229,
+      "learning_rate": 4.699942761308307e-08,
+      "loss": 0.4095,
+      "step": 7705
+    },
+    {
+      "epoch": 3.823135161970957,
+      "grad_norm": 0.07212443730436509,
+      "learning_rate": 4.67324234507216e-08,
+      "loss": 0.4208,
+      "step": 7706
+    },
+    {
+      "epoch": 3.8236316246741966,
+      "grad_norm": 0.0701011409340594,
+      "learning_rate": 4.646617631588912e-08,
+      "loss": 0.4538,
+      "step": 7707
+    },
+    {
+      "epoch": 3.824128087377436,
+      "grad_norm": 0.07139643476891598,
+      "learning_rate": 4.620068624927754e-08,
+      "loss": 0.4357,
+      "step": 7708
+    },
+    {
+      "epoch": 3.8246245500806753,
+      "grad_norm": 0.07197096113145784,
+      "learning_rate": 4.593595329146327e-08,
+      "loss": 0.4173,
+      "step": 7709
+    },
+    {
+      "epoch": 3.8251210127839146,
+      "grad_norm": 0.07144844528918787,
+      "learning_rate": 4.567197748290619e-08,
+      "loss": 0.4693,
+      "step": 7710
+    },
+    {
+      "epoch": 3.825617475487154,
+      "grad_norm": 0.07178497814452507,
+      "learning_rate": 4.5408758863950685e-08,
+      "loss": 0.4556,
+      "step": 7711
+    },
+    {
+      "epoch": 3.8261139381903932,
+      "grad_norm": 0.07187992019343,
+      "learning_rate": 4.5146297474825684e-08,
+      "loss": 0.4478,
+      "step": 7712
+    },
+    {
+      "epoch": 3.826610400893633,
+      "grad_norm": 0.0696093030235306,
+      "learning_rate": 4.48845933556441e-08,
+      "loss": 0.409,
+      "step": 7713
+    },
+    {
+      "epoch": 3.8271068635968724,
+      "grad_norm": 0.07230955744554561,
+      "learning_rate": 4.462364654640283e-08,
+      "loss": 0.4346,
+      "step": 7714
+    },
+    {
+      "epoch": 3.8276033263001117,
+      "grad_norm": 0.0731872540090233,
+      "learning_rate": 4.4363457086984416e-08,
+      "loss": 0.4461,
+      "step": 7715
+    },
+    {
+      "epoch": 3.828099789003351,
+      "grad_norm": 0.0733655897084953,
+      "learning_rate": 4.4104025017153165e-08,
+      "loss": 0.4661,
+      "step": 7716
+    },
+    {
+      "epoch": 3.828596251706591,
+      "grad_norm": 0.06899392563017782,
+      "learning_rate": 4.384535037656068e-08,
+      "loss": 0.4181,
+      "step": 7717
+    },
+    {
+      "epoch": 3.82909271440983,
+      "grad_norm": 0.07049518740890774,
+      "learning_rate": 4.3587433204739795e-08,
+      "loss": 0.4303,
+      "step": 7718
+    },
+    {
+      "epoch": 3.8295891771130695,
+      "grad_norm": 0.0746969522914595,
+      "learning_rate": 4.333027354111008e-08,
+      "loss": 0.4639,
+      "step": 7719
+    },
+    {
+      "epoch": 3.830085639816309,
+      "grad_norm": 0.07287444592387665,
+      "learning_rate": 4.307387142497399e-08,
+      "loss": 0.465,
+      "step": 7720
+    },
+    {
+      "epoch": 3.830582102519548,
+      "grad_norm": 0.07368450889542699,
+      "learning_rate": 4.281822689551795e-08,
+      "loss": 0.4309,
+      "step": 7721
+    },
+    {
+      "epoch": 3.8310785652227874,
+      "grad_norm": 0.07167812835385413,
+      "learning_rate": 4.256333999181406e-08,
+      "loss": 0.422,
+      "step": 7722
+    },
+    {
+      "epoch": 3.831575027926027,
+      "grad_norm": 0.07185303282623809,
+      "learning_rate": 4.230921075281724e-08,
+      "loss": 0.4336,
+      "step": 7723
+    },
+    {
+      "epoch": 3.8320714906292666,
+      "grad_norm": 0.07492267375306622,
+      "learning_rate": 4.205583921736644e-08,
+      "loss": 0.4489,
+      "step": 7724
+    },
+    {
+      "epoch": 3.832567953332506,
+      "grad_norm": 0.0735973347845139,
+      "learning_rate": 4.18032254241868e-08,
+      "loss": 0.4628,
+      "step": 7725
+    },
+    {
+      "epoch": 3.833064416035745,
+      "grad_norm": 0.07490391828605589,
+      "learning_rate": 4.155136941188465e-08,
+      "loss": 0.4698,
+      "step": 7726
+    },
+    {
+      "epoch": 3.833560878738985,
+      "grad_norm": 0.07323970805963602,
+      "learning_rate": 4.130027121895419e-08,
+      "loss": 0.4248,
+      "step": 7727
+    },
+    {
+      "epoch": 3.8340573414422243,
+      "grad_norm": 0.07439139948354823,
+      "learning_rate": 4.104993088376974e-08,
+      "loss": 0.4538,
+      "step": 7728
+    },
+    {
+      "epoch": 3.8345538041454637,
+      "grad_norm": 0.07164520350912759,
+      "learning_rate": 4.0800348444592354e-08,
+      "loss": 0.4484,
+      "step": 7729
+    },
+    {
+      "epoch": 3.835050266848703,
+      "grad_norm": 0.07438774813373158,
+      "learning_rate": 4.0551523939567626e-08,
+      "loss": 0.4545,
+      "step": 7730
+    },
+    {
+      "epoch": 3.8355467295519423,
+      "grad_norm": 0.07505882519637529,
+      "learning_rate": 4.030345740672348e-08,
+      "loss": 0.4427,
+      "step": 7731
+    },
+    {
+      "epoch": 3.8360431922551816,
+      "grad_norm": 0.07161856615140008,
+      "learning_rate": 4.005614888397347e-08,
+      "loss": 0.399,
+      "step": 7732
+    },
+    {
+      "epoch": 3.836539654958421,
+      "grad_norm": 0.07156914586364191,
+      "learning_rate": 3.980959840911402e-08,
+      "loss": 0.4366,
+      "step": 7733
+    },
+    {
+      "epoch": 3.8370361176616608,
+      "grad_norm": 0.07024940982721835,
+      "learning_rate": 3.956380601982668e-08,
+      "loss": 0.4231,
+      "step": 7734
+    },
+    {
+      "epoch": 3.8375325803649,
+      "grad_norm": 0.0721047312960411,
+      "learning_rate": 3.931877175367749e-08,
+      "loss": 0.4326,
+      "step": 7735
+    },
+    {
+      "epoch": 3.8380290430681394,
+      "grad_norm": 0.07392564844521746,
+      "learning_rate": 3.9074495648115384e-08,
+      "loss": 0.4199,
+      "step": 7736
+    },
+    {
+      "epoch": 3.838525505771379,
+      "grad_norm": 0.07194179417323361,
+      "learning_rate": 3.88309777404744e-08,
+      "loss": 0.4501,
+      "step": 7737
+    },
+    {
+      "epoch": 3.8390219684746185,
+      "grad_norm": 0.07293171801245724,
+      "learning_rate": 3.8588218067972526e-08,
+      "loss": 0.4632,
+      "step": 7738
+    },
+    {
+      "epoch": 3.839518431177858,
+      "grad_norm": 0.0728548840052456,
+      "learning_rate": 3.8346216667710653e-08,
+      "loss": 0.4458,
+      "step": 7739
+    },
+    {
+      "epoch": 3.840014893881097,
+      "grad_norm": 0.07218780052939112,
+      "learning_rate": 3.8104973576675863e-08,
+      "loss": 0.4098,
+      "step": 7740
+    },
+    {
+      "epoch": 3.8405113565843365,
+      "grad_norm": 0.07300164587754282,
+      "learning_rate": 3.786448883173755e-08,
+      "loss": 0.4407,
+      "step": 7741
+    },
+    {
+      "epoch": 3.841007819287576,
+      "grad_norm": 0.07402710952167288,
+      "learning_rate": 3.762476246965019e-08,
+      "loss": 0.4467,
+      "step": 7742
+    },
+    {
+      "epoch": 3.841504281990815,
+      "grad_norm": 0.0741556939673995,
+      "learning_rate": 3.738579452705282e-08,
+      "loss": 0.4528,
+      "step": 7743
+    },
+    {
+      "epoch": 3.842000744694055,
+      "grad_norm": 0.07379438521836251,
+      "learning_rate": 3.7147585040467336e-08,
+      "loss": 0.4341,
+      "step": 7744
+    },
+    {
+      "epoch": 3.8424972073972943,
+      "grad_norm": 0.07020108294788051,
+      "learning_rate": 3.691013404630017e-08,
+      "loss": 0.4323,
+      "step": 7745
+    },
+    {
+      "epoch": 3.8429936701005336,
+      "grad_norm": 0.0708905708801741,
+      "learning_rate": 3.667344158084118e-08,
+      "loss": 0.4259,
+      "step": 7746
+    },
+    {
+      "epoch": 3.8434901328037734,
+      "grad_norm": 0.07450166176521068,
+      "learning_rate": 3.643750768026644e-08,
+      "loss": 0.447,
+      "step": 7747
+    },
+    {
+      "epoch": 3.8439865955070127,
+      "grad_norm": 0.07581994277980696,
+      "learning_rate": 3.620233238063375e-08,
+      "loss": 0.4254,
+      "step": 7748
+    },
+    {
+      "epoch": 3.844483058210252,
+      "grad_norm": 0.07482993843850519,
+      "learning_rate": 3.596791571788605e-08,
+      "loss": 0.4533,
+      "step": 7749
+    },
+    {
+      "epoch": 3.8449795209134914,
+      "grad_norm": 0.0745584907129804,
+      "learning_rate": 3.573425772785077e-08,
+      "loss": 0.4275,
+      "step": 7750
+    },
+    {
+      "epoch": 3.8454759836167307,
+      "grad_norm": 0.07337402169038784,
+      "learning_rate": 3.55013584462377e-08,
+      "loss": 0.4518,
+      "step": 7751
+    },
+    {
+      "epoch": 3.84597244631997,
+      "grad_norm": 0.0741595197126386,
+      "learning_rate": 3.526921790864224e-08,
+      "loss": 0.4334,
+      "step": 7752
+    },
+    {
+      "epoch": 3.8464689090232094,
+      "grad_norm": 0.07075236082454928,
+      "learning_rate": 3.503783615054324e-08,
+      "loss": 0.4057,
+      "step": 7753
+    },
+    {
+      "epoch": 3.846965371726449,
+      "grad_norm": 0.07066552410408557,
+      "learning_rate": 3.4807213207304624e-08,
+      "loss": 0.43,
+      "step": 7754
+    },
+    {
+      "epoch": 3.8474618344296885,
+      "grad_norm": 0.07297525480330684,
+      "learning_rate": 3.45773491141721e-08,
+      "loss": 0.4413,
+      "step": 7755
+    },
+    {
+      "epoch": 3.847958297132928,
+      "grad_norm": 0.07206069774315897,
+      "learning_rate": 3.4348243906277554e-08,
+      "loss": 0.4347,
+      "step": 7756
+    },
+    {
+      "epoch": 3.8484547598361676,
+      "grad_norm": 0.07137753032462671,
+      "learning_rate": 3.411989761863577e-08,
+      "loss": 0.4156,
+      "step": 7757
+    },
+    {
+      "epoch": 3.848951222539407,
+      "grad_norm": 0.07449584893291657,
+      "learning_rate": 3.389231028614548e-08,
+      "loss": 0.4427,
+      "step": 7758
+    },
+    {
+      "epoch": 3.8494476852426462,
+      "grad_norm": 0.07267303968637863,
+      "learning_rate": 3.3665481943590536e-08,
+      "loss": 0.436,
+      "step": 7759
+    },
+    {
+      "epoch": 3.8499441479458856,
+      "grad_norm": 0.0733567394109518,
+      "learning_rate": 3.34394126256371e-08,
+      "loss": 0.4537,
+      "step": 7760
+    },
+    {
+      "epoch": 3.850440610649125,
+      "grad_norm": 0.07014161013435151,
+      "learning_rate": 3.3214102366836974e-08,
+      "loss": 0.4138,
+      "step": 7761
+    },
+    {
+      "epoch": 3.8509370733523642,
+      "grad_norm": 0.07137409206356522,
+      "learning_rate": 3.2989551201624836e-08,
+      "loss": 0.4311,
+      "step": 7762
+    },
+    {
+      "epoch": 3.8514335360556036,
+      "grad_norm": 0.07205143541112742,
+      "learning_rate": 3.27657591643199e-08,
+      "loss": 0.416,
+      "step": 7763
+    },
+    {
+      "epoch": 3.8519299987588433,
+      "grad_norm": 0.0706412399665651,
+      "learning_rate": 3.2542726289124804e-08,
+      "loss": 0.4325,
+      "step": 7764
+    },
+    {
+      "epoch": 3.8524264614620827,
+      "grad_norm": 0.07095456941872033,
+      "learning_rate": 3.232045261012728e-08,
+      "loss": 0.4188,
+      "step": 7765
+    },
+    {
+      "epoch": 3.852922924165322,
+      "grad_norm": 0.07211475153795083,
+      "learning_rate": 3.2098938161297945e-08,
+      "loss": 0.4464,
+      "step": 7766
+    },
+    {
+      "epoch": 3.8534193868685613,
+      "grad_norm": 0.0701709114389938,
+      "learning_rate": 3.1878182976491366e-08,
+      "loss": 0.4042,
+      "step": 7767
+    },
+    {
+      "epoch": 3.853915849571801,
+      "grad_norm": 0.07153435007874302,
+      "learning_rate": 3.165818708944668e-08,
+      "loss": 0.4543,
+      "step": 7768
+    },
+    {
+      "epoch": 3.8544123122750404,
+      "grad_norm": 0.0712219761262089,
+      "learning_rate": 3.143895053378698e-08,
+      "loss": 0.4205,
+      "step": 7769
+    },
+    {
+      "epoch": 3.8549087749782798,
+      "grad_norm": 0.06959699545376342,
+      "learning_rate": 3.122047334301881e-08,
+      "loss": 0.4077,
+      "step": 7770
+    },
+    {
+      "epoch": 3.855405237681519,
+      "grad_norm": 0.07165436362760237,
+      "learning_rate": 3.100275555053323e-08,
+      "loss": 0.4264,
+      "step": 7771
+    },
+    {
+      "epoch": 3.8559017003847584,
+      "grad_norm": 0.0696543457598655,
+      "learning_rate": 3.0785797189604725e-08,
+      "loss": 0.4359,
+      "step": 7772
+    },
+    {
+      "epoch": 3.8563981630879978,
+      "grad_norm": 0.07115194392658761,
+      "learning_rate": 3.0569598293391235e-08,
+      "loss": 0.4155,
+      "step": 7773
+    },
+    {
+      "epoch": 3.8568946257912375,
+      "grad_norm": 0.07347182112534703,
+      "learning_rate": 3.035415889493631e-08,
+      "loss": 0.4571,
+      "step": 7774
+    },
+    {
+      "epoch": 3.857391088494477,
+      "grad_norm": 0.07383870491369782,
+      "learning_rate": 3.0139479027165855e-08,
+      "loss": 0.4633,
+      "step": 7775
+    },
+    {
+      "epoch": 3.857887551197716,
+      "grad_norm": 0.07222560472754942,
+      "learning_rate": 2.992555872289082e-08,
+      "loss": 0.4567,
+      "step": 7776
+    },
+    {
+      "epoch": 3.8583840139009555,
+      "grad_norm": 0.07311389448186495,
+      "learning_rate": 2.971239801480452e-08,
+      "loss": 0.4393,
+      "step": 7777
+    },
+    {
+      "epoch": 3.8588804766041953,
+      "grad_norm": 0.07205166481588295,
+      "learning_rate": 2.949999693548533e-08,
+      "loss": 0.4118,
+      "step": 7778
+    },
+    {
+      "epoch": 3.8593769393074346,
+      "grad_norm": 0.07175495025624617,
+      "learning_rate": 2.9288355517396726e-08,
+      "loss": 0.4252,
+      "step": 7779
+    },
+    {
+      "epoch": 3.859873402010674,
+      "grad_norm": 0.07242985283004692,
+      "learning_rate": 2.9077473792882837e-08,
+      "loss": 0.4192,
+      "step": 7780
+    },
+    {
+      "epoch": 3.8603698647139133,
+      "grad_norm": 0.07324306870128854,
+      "learning_rate": 2.8867351794174547e-08,
+      "loss": 0.4543,
+      "step": 7781
+    },
+    {
+      "epoch": 3.8608663274171526,
+      "grad_norm": 0.07133057449898085,
+      "learning_rate": 2.8657989553385614e-08,
+      "loss": 0.4362,
+      "step": 7782
+    },
+    {
+      "epoch": 3.861362790120392,
+      "grad_norm": 0.07255116932077542,
+      "learning_rate": 2.844938710251377e-08,
+      "loss": 0.4327,
+      "step": 7783
+    },
+    {
+      "epoch": 3.8618592528236317,
+      "grad_norm": 0.07084270382910844,
+      "learning_rate": 2.8241544473440185e-08,
+      "loss": 0.4079,
+      "step": 7784
+    },
+    {
+      "epoch": 3.862355715526871,
+      "grad_norm": 0.06919189412064118,
+      "learning_rate": 2.8034461697930005e-08,
+      "loss": 0.3972,
+      "step": 7785
+    },
+    {
+      "epoch": 3.8628521782301104,
+      "grad_norm": 0.07340701474171342,
+      "learning_rate": 2.7828138807633465e-08,
+      "loss": 0.4318,
+      "step": 7786
+    },
+    {
+      "epoch": 3.8633486409333497,
+      "grad_norm": 0.06966816916473437,
+      "learning_rate": 2.762257583408312e-08,
+      "loss": 0.4434,
+      "step": 7787
+    },
+    {
+      "epoch": 3.8638451036365895,
+      "grad_norm": 0.07393027151147268,
+      "learning_rate": 2.7417772808696065e-08,
+      "loss": 0.446,
+      "step": 7788
+    },
+    {
+      "epoch": 3.864341566339829,
+      "grad_norm": 0.0725801905832486,
+      "learning_rate": 2.7213729762773366e-08,
+      "loss": 0.4501,
+      "step": 7789
+    },
+    {
+      "epoch": 3.864838029043068,
+      "grad_norm": 0.07182022009850554,
+      "learning_rate": 2.7010446727498974e-08,
+      "loss": 0.4327,
+      "step": 7790
+    },
+    {
+      "epoch": 3.8653344917463075,
+      "grad_norm": 0.07026862262470093,
+      "learning_rate": 2.6807923733942474e-08,
+      "loss": 0.4239,
+      "step": 7791
+    },
+    {
+      "epoch": 3.865830954449547,
+      "grad_norm": 0.0700408038106669,
+      "learning_rate": 2.6606160813055225e-08,
+      "loss": 0.4386,
+      "step": 7792
+    },
+    {
+      "epoch": 3.866327417152786,
+      "grad_norm": 0.07021156797431176,
+      "learning_rate": 2.640515799567478e-08,
+      "loss": 0.4227,
+      "step": 7793
+    },
+    {
+      "epoch": 3.866823879856026,
+      "grad_norm": 0.07378239177728425,
+      "learning_rate": 2.6204915312519898e-08,
+      "loss": 0.4347,
+      "step": 7794
+    },
+    {
+      "epoch": 3.8673203425592653,
+      "grad_norm": 0.0709349128055407,
+      "learning_rate": 2.6005432794194985e-08,
+      "loss": 0.404,
+      "step": 7795
+    },
+    {
+      "epoch": 3.8678168052625046,
+      "grad_norm": 0.07371643425293671,
+      "learning_rate": 2.5806710471187323e-08,
+      "loss": 0.4593,
+      "step": 7796
+    },
+    {
+      "epoch": 3.868313267965744,
+      "grad_norm": 0.07127165318042873,
+      "learning_rate": 2.5608748373869285e-08,
+      "loss": 0.4214,
+      "step": 7797
+    },
+    {
+      "epoch": 3.8688097306689837,
+      "grad_norm": 0.07447420956274946,
+      "learning_rate": 2.5411546532496113e-08,
+      "loss": 0.4593,
+      "step": 7798
+    },
+    {
+      "epoch": 3.869306193372223,
+      "grad_norm": 0.07438841127488162,
+      "learning_rate": 2.5215104977205918e-08,
+      "loss": 0.4456,
+      "step": 7799
+    },
+    {
+      "epoch": 3.8698026560754624,
+      "grad_norm": 0.07230538935773687,
+      "learning_rate": 2.5019423738022464e-08,
+      "loss": 0.4437,
+      "step": 7800
+    },
+    {
+      "epoch": 3.8702991187787017,
+      "grad_norm": 0.07139035223489404,
+      "learning_rate": 2.4824502844852938e-08,
+      "loss": 0.4292,
+      "step": 7801
+    },
+    {
+      "epoch": 3.870795581481941,
+      "grad_norm": 0.07179684932883365,
+      "learning_rate": 2.463034232748629e-08,
+      "loss": 0.4277,
+      "step": 7802
+    },
+    {
+      "epoch": 3.8712920441851804,
+      "grad_norm": 0.07397952129422569,
+      "learning_rate": 2.443694221559878e-08,
+      "loss": 0.4437,
+      "step": 7803
+    },
+    {
+      "epoch": 3.87178850688842,
+      "grad_norm": 0.07249524619609739,
+      "learning_rate": 2.4244302538746765e-08,
+      "loss": 0.4558,
+      "step": 7804
+    },
+    {
+      "epoch": 3.8722849695916595,
+      "grad_norm": 0.07270288733195596,
+      "learning_rate": 2.4052423326373364e-08,
+      "loss": 0.4578,
+      "step": 7805
+    },
+    {
+      "epoch": 3.872781432294899,
+      "grad_norm": 0.0726698326519595,
+      "learning_rate": 2.3861304607804005e-08,
+      "loss": 0.4179,
+      "step": 7806
+    },
+    {
+      "epoch": 3.873277894998138,
+      "grad_norm": 0.07307247352030227,
+      "learning_rate": 2.367094641224754e-08,
+      "loss": 0.425,
+      "step": 7807
+    },
+    {
+      "epoch": 3.873774357701378,
+      "grad_norm": 0.06997583746928546,
+      "learning_rate": 2.348134876879793e-08,
+      "loss": 0.4366,
+      "step": 7808
+    },
+    {
+      "epoch": 3.8742708204046172,
+      "grad_norm": 0.07301178424469724,
+      "learning_rate": 2.3292511706431432e-08,
+      "loss": 0.4397,
+      "step": 7809
+    },
+    {
+      "epoch": 3.8747672831078566,
+      "grad_norm": 0.07162155795689573,
+      "learning_rate": 2.3104435254008852e-08,
+      "loss": 0.4414,
+      "step": 7810
+    },
+    {
+      "epoch": 3.875263745811096,
+      "grad_norm": 0.07393752464541645,
+      "learning_rate": 2.2917119440275524e-08,
+      "loss": 0.4407,
+      "step": 7811
+    },
+    {
+      "epoch": 3.8757602085143352,
+      "grad_norm": 0.07452520362796672,
+      "learning_rate": 2.273056429385856e-08,
+      "loss": 0.4543,
+      "step": 7812
+    },
+    {
+      "epoch": 3.8762566712175746,
+      "grad_norm": 0.07272499079941327,
+      "learning_rate": 2.25447698432707e-08,
+      "loss": 0.4385,
+      "step": 7813
+    },
+    {
+      "epoch": 3.8767531339208143,
+      "grad_norm": 0.07405819111966294,
+      "learning_rate": 2.2359736116907006e-08,
+      "loss": 0.4587,
+      "step": 7814
+    },
+    {
+      "epoch": 3.8772495966240537,
+      "grad_norm": 0.07297572414669824,
+      "learning_rate": 2.2175463143047636e-08,
+      "loss": 0.4528,
+      "step": 7815
+    },
+    {
+      "epoch": 3.877746059327293,
+      "grad_norm": 0.072798895598875,
+      "learning_rate": 2.1991950949855067e-08,
+      "loss": 0.4457,
+      "step": 7816
+    },
+    {
+      "epoch": 3.8782425220305323,
+      "grad_norm": 0.07416129028729428,
+      "learning_rate": 2.1809199565376305e-08,
+      "loss": 0.4544,
+      "step": 7817
+    },
+    {
+      "epoch": 3.878738984733772,
+      "grad_norm": 0.07049086209545863,
+      "learning_rate": 2.162720901754234e-08,
+      "loss": 0.4315,
+      "step": 7818
+    },
+    {
+      "epoch": 3.8792354474370114,
+      "grad_norm": 0.07035749557978963,
+      "learning_rate": 2.144597933416759e-08,
+      "loss": 0.4339,
+      "step": 7819
+    },
+    {
+      "epoch": 3.8797319101402508,
+      "grad_norm": 0.07354958199948448,
+      "learning_rate": 2.1265510542949895e-08,
+      "loss": 0.4473,
+      "step": 7820
+    },
+    {
+      "epoch": 3.88022837284349,
+      "grad_norm": 0.07115599998557434,
+      "learning_rate": 2.1085802671470533e-08,
+      "loss": 0.4474,
+      "step": 7821
+    },
+    {
+      "epoch": 3.8807248355467294,
+      "grad_norm": 0.07502940181703499,
+      "learning_rate": 2.0906855747195864e-08,
+      "loss": 0.4541,
+      "step": 7822
+    },
+    {
+      "epoch": 3.8812212982499688,
+      "grad_norm": 0.07376974132135386,
+      "learning_rate": 2.0728669797474565e-08,
+      "loss": 0.4585,
+      "step": 7823
+    },
+    {
+      "epoch": 3.8817177609532085,
+      "grad_norm": 0.07225019533734907,
+      "learning_rate": 2.05512448495393e-08,
+      "loss": 0.4423,
+      "step": 7824
+    },
+    {
+      "epoch": 3.882214223656448,
+      "grad_norm": 0.07441920631682299,
+      "learning_rate": 2.037458093050726e-08,
+      "loss": 0.4407,
+      "step": 7825
+    },
+    {
+      "epoch": 3.882710686359687,
+      "grad_norm": 0.07540349843464134,
+      "learning_rate": 2.0198678067377965e-08,
+      "loss": 0.4355,
+      "step": 7826
+    },
+    {
+      "epoch": 3.8832071490629265,
+      "grad_norm": 0.07116267191584182,
+      "learning_rate": 2.002353628703546e-08,
+      "loss": 0.4147,
+      "step": 7827
+    },
+    {
+      "epoch": 3.8837036117661663,
+      "grad_norm": 0.07237432491317197,
+      "learning_rate": 1.984915561624834e-08,
+      "loss": 0.4307,
+      "step": 7828
+    },
+    {
+      "epoch": 3.8842000744694056,
+      "grad_norm": 0.0712437413651589,
+      "learning_rate": 1.9675536081666392e-08,
+      "loss": 0.4266,
+      "step": 7829
+    },
+    {
+      "epoch": 3.884696537172645,
+      "grad_norm": 0.07248638071409952,
+      "learning_rate": 1.9502677709825613e-08,
+      "loss": 0.4531,
+      "step": 7830
+    },
+    {
+      "epoch": 3.8851929998758843,
+      "grad_norm": 0.07092717694139716,
+      "learning_rate": 1.9330580527144315e-08,
+      "loss": 0.432,
+      "step": 7831
+    },
+    {
+      "epoch": 3.8856894625791236,
+      "grad_norm": 0.06979256151546342,
+      "learning_rate": 1.9159244559924795e-08,
+      "loss": 0.4214,
+      "step": 7832
+    },
+    {
+      "epoch": 3.886185925282363,
+      "grad_norm": 0.0717442354707785,
+      "learning_rate": 1.8988669834352767e-08,
+      "loss": 0.4483,
+      "step": 7833
+    },
+    {
+      "epoch": 3.8866823879856027,
+      "grad_norm": 0.07453108762459253,
+      "learning_rate": 1.8818856376498497e-08,
+      "loss": 0.4479,
+      "step": 7834
+    },
+    {
+      "epoch": 3.887178850688842,
+      "grad_norm": 0.07154902113522246,
+      "learning_rate": 1.8649804212315103e-08,
+      "loss": 0.4168,
+      "step": 7835
+    },
+    {
+      "epoch": 3.8876753133920814,
+      "grad_norm": 0.07514204973247401,
+      "learning_rate": 1.8481513367638593e-08,
+      "loss": 0.4346,
+      "step": 7836
+    },
+    {
+      "epoch": 3.8881717760953207,
+      "grad_norm": 0.07279588195975031,
+      "learning_rate": 1.831398386819061e-08,
+      "loss": 0.4302,
+      "step": 7837
+    },
+    {
+      "epoch": 3.8886682387985605,
+      "grad_norm": 0.07083433789230371,
+      "learning_rate": 1.8147215739575118e-08,
+      "loss": 0.4343,
+      "step": 7838
+    },
+    {
+      "epoch": 3.8891647015018,
+      "grad_norm": 0.07264082852536816,
+      "learning_rate": 1.7981209007278956e-08,
+      "loss": 0.4373,
+      "step": 7839
+    },
+    {
+      "epoch": 3.889661164205039,
+      "grad_norm": 0.07218971964839162,
+      "learning_rate": 1.7815963696675153e-08,
+      "loss": 0.4335,
+      "step": 7840
+    },
+    {
+      "epoch": 3.8901576269082785,
+      "grad_norm": 0.07279006779354186,
+      "learning_rate": 1.765147983301796e-08,
+      "loss": 0.4327,
+      "step": 7841
+    },
+    {
+      "epoch": 3.890654089611518,
+      "grad_norm": 0.07169296606291194,
+      "learning_rate": 1.7487757441446152e-08,
+      "loss": 0.457,
+      "step": 7842
+    },
+    {
+      "epoch": 3.891150552314757,
+      "grad_norm": 0.07436589540232992,
+      "learning_rate": 1.7324796546981937e-08,
+      "loss": 0.4483,
+      "step": 7843
+    },
+    {
+      "epoch": 3.891647015017997,
+      "grad_norm": 0.0746696408772963,
+      "learning_rate": 1.7162597174531503e-08,
+      "loss": 0.4514,
+      "step": 7844
+    },
+    {
+      "epoch": 3.8921434777212363,
+      "grad_norm": 0.07588571149435935,
+      "learning_rate": 1.7001159348884466e-08,
+      "loss": 0.4471,
+      "step": 7845
+    },
+    {
+      "epoch": 3.8926399404244756,
+      "grad_norm": 0.07245883306650445,
+      "learning_rate": 1.6840483094713867e-08,
+      "loss": 0.4646,
+      "step": 7846
+    },
+    {
+      "epoch": 3.893136403127715,
+      "grad_norm": 0.0705640191588085,
+      "learning_rate": 1.6680568436576726e-08,
+      "loss": 0.4345,
+      "step": 7847
+    },
+    {
+      "epoch": 3.8936328658309547,
+      "grad_norm": 0.07208993186469582,
+      "learning_rate": 1.6521415398912942e-08,
+      "loss": 0.434,
+      "step": 7848
+    },
+    {
+      "epoch": 3.894129328534194,
+      "grad_norm": 0.07197941174858101,
+      "learning_rate": 1.6363024006046945e-08,
+      "loss": 0.4348,
+      "step": 7849
+    },
+    {
+      "epoch": 3.8946257912374334,
+      "grad_norm": 0.07302725303638385,
+      "learning_rate": 1.6205394282186037e-08,
+      "loss": 0.4585,
+      "step": 7850
+    },
+    {
+      "epoch": 3.8951222539406727,
+      "grad_norm": 0.07479123755996081,
+      "learning_rate": 1.6048526251421502e-08,
+      "loss": 0.4697,
+      "step": 7851
+    },
+    {
+      "epoch": 3.895618716643912,
+      "grad_norm": 0.0714460020972371,
+      "learning_rate": 1.589241993772861e-08,
+      "loss": 0.4547,
+      "step": 7852
+    },
+    {
+      "epoch": 3.8961151793471513,
+      "grad_norm": 0.06989138523119054,
+      "learning_rate": 1.573707536496494e-08,
+      "loss": 0.4448,
+      "step": 7853
+    },
+    {
+      "epoch": 3.896611642050391,
+      "grad_norm": 0.07408706106103906,
+      "learning_rate": 1.5582492556872608e-08,
+      "loss": 0.4332,
+      "step": 7854
+    },
+    {
+      "epoch": 3.8971081047536305,
+      "grad_norm": 0.07365674527289237,
+      "learning_rate": 1.5428671537077168e-08,
+      "loss": 0.4464,
+      "step": 7855
+    },
+    {
+      "epoch": 3.89760456745687,
+      "grad_norm": 0.07171203833147921,
+      "learning_rate": 1.527561232908814e-08,
+      "loss": 0.4478,
+      "step": 7856
+    },
+    {
+      "epoch": 3.898101030160109,
+      "grad_norm": 0.07137804282473109,
+      "learning_rate": 1.5123314956297375e-08,
+      "loss": 0.4175,
+      "step": 7857
+    },
+    {
+      "epoch": 3.898597492863349,
+      "grad_norm": 0.07128942518964267,
+      "learning_rate": 1.4971779441981804e-08,
+      "loss": 0.4415,
+      "step": 7858
+    },
+    {
+      "epoch": 3.899093955566588,
+      "grad_norm": 0.07515082405610274,
+      "learning_rate": 1.4821005809300681e-08,
+      "loss": 0.4675,
+      "step": 7859
+    },
+    {
+      "epoch": 3.8995904182698276,
+      "grad_norm": 0.07434300726999438,
+      "learning_rate": 1.4670994081297796e-08,
+      "loss": 0.4691,
+      "step": 7860
+    },
+    {
+      "epoch": 3.900086880973067,
+      "grad_norm": 0.07227320230556432,
+      "learning_rate": 1.4521744280899808e-08,
+      "loss": 0.4602,
+      "step": 7861
+    },
+    {
+      "epoch": 3.900583343676306,
+      "grad_norm": 0.07159390508729566,
+      "learning_rate": 1.4373256430916805e-08,
+      "loss": 0.4324,
+      "step": 7862
+    },
+    {
+      "epoch": 3.9010798063795455,
+      "grad_norm": 0.07127322741170937,
+      "learning_rate": 1.4225530554043409e-08,
+      "loss": 0.4531,
+      "step": 7863
+    },
+    {
+      "epoch": 3.9015762690827853,
+      "grad_norm": 0.07397431432659492,
+      "learning_rate": 1.4078566672856564e-08,
+      "loss": 0.4535,
+      "step": 7864
+    },
+    {
+      "epoch": 3.9020727317860247,
+      "grad_norm": 0.07726727419027592,
+      "learning_rate": 1.3932364809817745e-08,
+      "loss": 0.4815,
+      "step": 7865
+    },
+    {
+      "epoch": 3.902569194489264,
+      "grad_norm": 0.07289550295671977,
+      "learning_rate": 1.3786924987271299e-08,
+      "loss": 0.4266,
+      "step": 7866
+    },
+    {
+      "epoch": 3.9030656571925033,
+      "grad_norm": 0.07308451132181172,
+      "learning_rate": 1.3642247227446114e-08,
+      "loss": 0.4157,
+      "step": 7867
+    },
+    {
+      "epoch": 3.903562119895743,
+      "grad_norm": 0.07590479426568846,
+      "learning_rate": 1.3498331552452837e-08,
+      "loss": 0.4762,
+      "step": 7868
+    },
+    {
+      "epoch": 3.9040585825989824,
+      "grad_norm": 0.06921959548820725,
+      "learning_rate": 1.3355177984287205e-08,
+      "loss": 0.4036,
+      "step": 7869
+    },
+    {
+      "epoch": 3.9045550453022217,
+      "grad_norm": 0.07118954201021534,
+      "learning_rate": 1.3212786544827828e-08,
+      "loss": 0.4435,
+      "step": 7870
+    },
+    {
+      "epoch": 3.905051508005461,
+      "grad_norm": 0.07403036754736678,
+      "learning_rate": 1.3071157255836742e-08,
+      "loss": 0.4469,
+      "step": 7871
+    },
+    {
+      "epoch": 3.9055479707087004,
+      "grad_norm": 0.07467206775977153,
+      "learning_rate": 1.2930290138960522e-08,
+      "loss": 0.4566,
+      "step": 7872
+    },
+    {
+      "epoch": 3.9060444334119397,
+      "grad_norm": 0.07396114789472796,
+      "learning_rate": 1.2790185215727501e-08,
+      "loss": 0.4306,
+      "step": 7873
+    },
+    {
+      "epoch": 3.906540896115179,
+      "grad_norm": 0.07240021119883902,
+      "learning_rate": 1.2650842507550554e-08,
+      "loss": 0.4314,
+      "step": 7874
+    },
+    {
+      "epoch": 3.907037358818419,
+      "grad_norm": 0.07189908371677901,
+      "learning_rate": 1.251226203572653e-08,
+      "loss": 0.4538,
+      "step": 7875
+    },
+    {
+      "epoch": 3.907533821521658,
+      "grad_norm": 0.07336991053621658,
+      "learning_rate": 1.2374443821435156e-08,
+      "loss": 0.4604,
+      "step": 7876
+    },
+    {
+      "epoch": 3.9080302842248975,
+      "grad_norm": 0.0720966340794103,
+      "learning_rate": 1.2237387885739582e-08,
+      "loss": 0.4556,
+      "step": 7877
+    },
+    {
+      "epoch": 3.9085267469281373,
+      "grad_norm": 0.07273980475231961,
+      "learning_rate": 1.2101094249585832e-08,
+      "loss": 0.4215,
+      "step": 7878
+    },
+    {
+      "epoch": 3.9090232096313766,
+      "grad_norm": 0.0743906178360643,
+      "learning_rate": 1.1965562933805575e-08,
+      "loss": 0.4663,
+      "step": 7879
+    },
+    {
+      "epoch": 3.909519672334616,
+      "grad_norm": 0.07354772952353111,
+      "learning_rate": 1.1830793959112241e-08,
+      "loss": 0.4241,
+      "step": 7880
+    },
+    {
+      "epoch": 3.9100161350378553,
+      "grad_norm": 0.07363233034652644,
+      "learning_rate": 1.1696787346102134e-08,
+      "loss": 0.4644,
+      "step": 7881
+    },
+    {
+      "epoch": 3.9105125977410946,
+      "grad_norm": 0.0752133492026765,
+      "learning_rate": 1.1563543115257203e-08,
+      "loss": 0.4649,
+      "step": 7882
+    },
+    {
+      "epoch": 3.911009060444334,
+      "grad_norm": 0.07207085299674393,
+      "learning_rate": 1.1431061286941159e-08,
+      "loss": 0.438,
+      "step": 7883
+    },
+    {
+      "epoch": 3.9115055231475733,
+      "grad_norm": 0.07235157876028925,
+      "learning_rate": 1.1299341881401693e-08,
+      "loss": 0.4298,
+      "step": 7884
+    },
+    {
+      "epoch": 3.912001985850813,
+      "grad_norm": 0.06988345938969277,
+      "learning_rate": 1.1168384918769926e-08,
+      "loss": 0.4237,
+      "step": 7885
+    },
+    {
+      "epoch": 3.9124984485540524,
+      "grad_norm": 0.07545700325415562,
+      "learning_rate": 1.1038190419060957e-08,
+      "loss": 0.4584,
+      "step": 7886
+    },
+    {
+      "epoch": 3.9129949112572917,
+      "grad_norm": 0.07367556891620859,
+      "learning_rate": 1.090875840217276e-08,
+      "loss": 0.4421,
+      "step": 7887
+    },
+    {
+      "epoch": 3.9134913739605315,
+      "grad_norm": 0.07316765299975153,
+      "learning_rate": 1.078008888788673e-08,
+      "loss": 0.4282,
+      "step": 7888
+    },
+    {
+      "epoch": 3.913987836663771,
+      "grad_norm": 0.07406156551740832,
+      "learning_rate": 1.0652181895867697e-08,
+      "loss": 0.4556,
+      "step": 7889
+    },
+    {
+      "epoch": 3.91448429936701,
+      "grad_norm": 0.07194509976912991,
+      "learning_rate": 1.052503744566502e-08,
+      "loss": 0.4311,
+      "step": 7890
+    },
+    {
+      "epoch": 3.9149807620702495,
+      "grad_norm": 0.07084174131860045,
+      "learning_rate": 1.0398655556709824e-08,
+      "loss": 0.4387,
+      "step": 7891
+    },
+    {
+      "epoch": 3.915477224773489,
+      "grad_norm": 0.07592270875366632,
+      "learning_rate": 1.0273036248318325e-08,
+      "loss": 0.4898,
+      "step": 7892
+    },
+    {
+      "epoch": 3.915973687476728,
+      "grad_norm": 0.07159994780886023,
+      "learning_rate": 1.0148179539689051e-08,
+      "loss": 0.4386,
+      "step": 7893
+    },
+    {
+      "epoch": 3.9164701501799675,
+      "grad_norm": 0.07249131497488212,
+      "learning_rate": 1.0024085449903964e-08,
+      "loss": 0.459,
+      "step": 7894
+    },
+    {
+      "epoch": 3.9169666128832072,
+      "grad_norm": 0.07215634562957392,
+      "learning_rate": 9.900753997929557e-09,
+      "loss": 0.4351,
+      "step": 7895
+    },
+    {
+      "epoch": 3.9174630755864466,
+      "grad_norm": 0.07306229934044586,
+      "learning_rate": 9.778185202614643e-09,
+      "loss": 0.4542,
+      "step": 7896
+    },
+    {
+      "epoch": 3.917959538289686,
+      "grad_norm": 0.07464935978086107,
+      "learning_rate": 9.656379082692014e-09,
+      "loss": 0.4696,
+      "step": 7897
+    },
+    {
+      "epoch": 3.9184560009929257,
+      "grad_norm": 0.06959882315202705,
+      "learning_rate": 9.535335656777333e-09,
+      "loss": 0.4045,
+      "step": 7898
+    },
+    {
+      "epoch": 3.918952463696165,
+      "grad_norm": 0.07189424265721825,
+      "learning_rate": 9.41505494337136e-09,
+      "loss": 0.444,
+      "step": 7899
+    },
+    {
+      "epoch": 3.9194489263994043,
+      "grad_norm": 0.07512861736253441,
+      "learning_rate": 9.295536960856055e-09,
+      "loss": 0.4615,
+      "step": 7900
+    },
+    {
+      "epoch": 3.9199453891026437,
+      "grad_norm": 0.07184217453560283,
+      "learning_rate": 9.176781727497919e-09,
+      "loss": 0.449,
+      "step": 7901
+    },
+    {
+      "epoch": 3.920441851805883,
+      "grad_norm": 0.0692935480523205,
+      "learning_rate": 9.058789261446876e-09,
+      "loss": 0.4199,
+      "step": 7902
+    },
+    {
+      "epoch": 3.9209383145091223,
+      "grad_norm": 0.07158569849881712,
+      "learning_rate": 8.94155958073628e-09,
+      "loss": 0.4446,
+      "step": 7903
+    },
+    {
+      "epoch": 3.9214347772123617,
+      "grad_norm": 0.07128555959690813,
+      "learning_rate": 8.825092703282912e-09,
+      "loss": 0.4417,
+      "step": 7904
+    },
+    {
+      "epoch": 3.9219312399156014,
+      "grad_norm": 0.07386013608300344,
+      "learning_rate": 8.709388646886419e-09,
+      "loss": 0.4383,
+      "step": 7905
+    },
+    {
+      "epoch": 3.9224277026188408,
+      "grad_norm": 0.07077271842701878,
+      "learning_rate": 8.59444742923099e-09,
+      "loss": 0.4483,
+      "step": 7906
+    },
+    {
+      "epoch": 3.92292416532208,
+      "grad_norm": 0.07214791377825751,
+      "learning_rate": 8.480269067882574e-09,
+      "loss": 0.4524,
+      "step": 7907
+    },
+    {
+      "epoch": 3.9234206280253194,
+      "grad_norm": 0.07498863312685188,
+      "learning_rate": 8.366853580292767e-09,
+      "loss": 0.4706,
+      "step": 7908
+    },
+    {
+      "epoch": 3.923917090728559,
+      "grad_norm": 0.07559995398531838,
+      "learning_rate": 8.254200983794369e-09,
+      "loss": 0.4325,
+      "step": 7909
+    },
+    {
+      "epoch": 3.9244135534317985,
+      "grad_norm": 0.07226155025300252,
+      "learning_rate": 8.14231129560472e-09,
+      "loss": 0.4312,
+      "step": 7910
+    },
+    {
+      "epoch": 3.924910016135038,
+      "grad_norm": 0.07127750193403352,
+      "learning_rate": 8.031184532824588e-09,
+      "loss": 0.4491,
+      "step": 7911
+    },
+    {
+      "epoch": 3.925406478838277,
+      "grad_norm": 0.07415159903823959,
+      "learning_rate": 7.920820712437604e-09,
+      "loss": 0.4306,
+      "step": 7912
+    },
+    {
+      "epoch": 3.9259029415415165,
+      "grad_norm": 0.07378377183619132,
+      "learning_rate": 7.811219851311392e-09,
+      "loss": 0.4494,
+      "step": 7913
+    },
+    {
+      "epoch": 3.926399404244756,
+      "grad_norm": 0.07275277293764515,
+      "learning_rate": 7.702381966196437e-09,
+      "loss": 0.4556,
+      "step": 7914
+    },
+    {
+      "epoch": 3.9268958669479956,
+      "grad_norm": 0.0746244877378339,
+      "learning_rate": 7.594307073727214e-09,
+      "loss": 0.434,
+      "step": 7915
+    },
+    {
+      "epoch": 3.927392329651235,
+      "grad_norm": 0.07584813299523505,
+      "learning_rate": 7.48699519042051e-09,
+      "loss": 0.4589,
+      "step": 7916
+    },
+    {
+      "epoch": 3.9278887923544743,
+      "grad_norm": 0.07282040520441282,
+      "learning_rate": 7.380446332678204e-09,
+      "loss": 0.4531,
+      "step": 7917
+    },
+    {
+      "epoch": 3.9283852550577136,
+      "grad_norm": 0.07229666000116906,
+      "learning_rate": 7.274660516783938e-09,
+      "loss": 0.4335,
+      "step": 7918
+    },
+    {
+      "epoch": 3.9288817177609534,
+      "grad_norm": 0.07213635388094723,
+      "learning_rate": 7.16963775890589e-09,
+      "loss": 0.4284,
+      "step": 7919
+    },
+    {
+      "epoch": 3.9293781804641927,
+      "grad_norm": 0.0734265619388905,
+      "learning_rate": 7.0653780750945534e-09,
+      "loss": 0.4937,
+      "step": 7920
+    },
+    {
+      "epoch": 3.929874643167432,
+      "grad_norm": 0.07242700143100422,
+      "learning_rate": 6.961881481284405e-09,
+      "loss": 0.4483,
+      "step": 7921
+    },
+    {
+      "epoch": 3.9303711058706714,
+      "grad_norm": 0.07132732254668113,
+      "learning_rate": 6.859147993293347e-09,
+      "loss": 0.4248,
+      "step": 7922
+    },
+    {
+      "epoch": 3.9308675685739107,
+      "grad_norm": 0.0735034320107039,
+      "learning_rate": 6.757177626822709e-09,
+      "loss": 0.4509,
+      "step": 7923
+    },
+    {
+      "epoch": 3.93136403127715,
+      "grad_norm": 0.07314542671884339,
+      "learning_rate": 6.655970397457245e-09,
+      "loss": 0.4424,
+      "step": 7924
+    },
+    {
+      "epoch": 3.93186049398039,
+      "grad_norm": 0.0711518739203714,
+      "learning_rate": 6.555526320664029e-09,
+      "loss": 0.4151,
+      "step": 7925
+    },
+    {
+      "epoch": 3.932356956683629,
+      "grad_norm": 0.07142566153152452,
+      "learning_rate": 6.455845411795225e-09,
+      "loss": 0.4423,
+      "step": 7926
+    },
+    {
+      "epoch": 3.9328534193868685,
+      "grad_norm": 0.07246197419030367,
+      "learning_rate": 6.356927686084757e-09,
+      "loss": 0.4286,
+      "step": 7927
+    },
+    {
+      "epoch": 3.933349882090108,
+      "grad_norm": 0.0735996074037667,
+      "learning_rate": 6.258773158650533e-09,
+      "loss": 0.4461,
+      "step": 7928
+    },
+    {
+      "epoch": 3.9338463447933476,
+      "grad_norm": 0.07195374563717467,
+      "learning_rate": 6.161381844494995e-09,
+      "loss": 0.404,
+      "step": 7929
+    },
+    {
+      "epoch": 3.934342807496587,
+      "grad_norm": 0.07236629384914703,
+      "learning_rate": 6.0647537585017956e-09,
+      "loss": 0.447,
+      "step": 7930
+    },
+    {
+      "epoch": 3.9348392701998263,
+      "grad_norm": 0.07330477178938324,
+      "learning_rate": 5.968888915439675e-09,
+      "loss": 0.4312,
+      "step": 7931
+    },
+    {
+      "epoch": 3.9353357329030656,
+      "grad_norm": 0.07297599221536687,
+      "learning_rate": 5.873787329959135e-09,
+      "loss": 0.4588,
+      "step": 7932
+    },
+    {
+      "epoch": 3.935832195606305,
+      "grad_norm": 0.0725622873374946,
+      "learning_rate": 5.779449016595773e-09,
+      "loss": 0.4248,
+      "step": 7933
+    },
+    {
+      "epoch": 3.9363286583095443,
+      "grad_norm": 0.07113219855206085,
+      "learning_rate": 5.685873989767499e-09,
+      "loss": 0.414,
+      "step": 7934
+    },
+    {
+      "epoch": 3.936825121012784,
+      "grad_norm": 0.07443343638570427,
+      "learning_rate": 5.593062263775095e-09,
+      "loss": 0.4223,
+      "step": 7935
+    },
+    {
+      "epoch": 3.9373215837160234,
+      "grad_norm": 0.07409075328129311,
+      "learning_rate": 5.501013852804438e-09,
+      "loss": 0.4621,
+      "step": 7936
+    },
+    {
+      "epoch": 3.9378180464192627,
+      "grad_norm": 0.07382741390740961,
+      "learning_rate": 5.409728770923162e-09,
+      "loss": 0.4646,
+      "step": 7937
+    },
+    {
+      "epoch": 3.938314509122502,
+      "grad_norm": 0.07174126251091134,
+      "learning_rate": 5.31920703208233e-09,
+      "loss": 0.4315,
+      "step": 7938
+    },
+    {
+      "epoch": 3.938810971825742,
+      "grad_norm": 0.07545133094403188,
+      "learning_rate": 5.2294486501175415e-09,
+      "loss": 0.4343,
+      "step": 7939
+    },
+    {
+      "epoch": 3.939307434528981,
+      "grad_norm": 0.0738239885507223,
+      "learning_rate": 5.140453638746156e-09,
+      "loss": 0.434,
+      "step": 7940
+    },
+    {
+      "epoch": 3.9398038972322205,
+      "grad_norm": 0.07164997663670053,
+      "learning_rate": 5.052222011570074e-09,
+      "loss": 0.4165,
+      "step": 7941
+    },
+    {
+      "epoch": 3.94030035993546,
+      "grad_norm": 0.07283733967363415,
+      "learning_rate": 4.964753782073506e-09,
+      "loss": 0.4388,
+      "step": 7942
+    },
+    {
+      "epoch": 3.940796822638699,
+      "grad_norm": 0.0739064888698837,
+      "learning_rate": 4.878048963625759e-09,
+      "loss": 0.4847,
+      "step": 7943
+    },
+    {
+      "epoch": 3.9412932853419385,
+      "grad_norm": 0.07381267380019754,
+      "learning_rate": 4.792107569476789e-09,
+      "loss": 0.4413,
+      "step": 7944
+    },
+    {
+      "epoch": 3.9417897480451782,
+      "grad_norm": 0.07393797317527814,
+      "learning_rate": 4.706929612762756e-09,
+      "loss": 0.4765,
+      "step": 7945
+    },
+    {
+      "epoch": 3.9422862107484176,
+      "grad_norm": 0.07426952582219104,
+      "learning_rate": 4.6225151065004695e-09,
+      "loss": 0.474,
+      "step": 7946
+    },
+    {
+      "epoch": 3.942782673451657,
+      "grad_norm": 0.07352326988482331,
+      "learning_rate": 4.5388640635923855e-09,
+      "loss": 0.4547,
+      "step": 7947
+    },
+    {
+      "epoch": 3.9432791361548962,
+      "grad_norm": 0.07224179319878705,
+      "learning_rate": 4.455976496822723e-09,
+      "loss": 0.4379,
+      "step": 7948
+    },
+    {
+      "epoch": 3.943775598858136,
+      "grad_norm": 0.07259290717504588,
+      "learning_rate": 4.373852418859681e-09,
+      "loss": 0.467,
+      "step": 7949
+    },
+    {
+      "epoch": 3.9442720615613753,
+      "grad_norm": 0.07293484053231977,
+      "learning_rate": 4.292491842254331e-09,
+      "loss": 0.4172,
+      "step": 7950
+    },
+    {
+      "epoch": 3.9447685242646147,
+      "grad_norm": 0.07146552534187618,
+      "learning_rate": 4.211894779441727e-09,
+      "loss": 0.4531,
+      "step": 7951
+    },
+    {
+      "epoch": 3.945264986967854,
+      "grad_norm": 0.07296431475149752,
+      "learning_rate": 4.1320612427397935e-09,
+      "loss": 0.429,
+      "step": 7952
+    },
+    {
+      "epoch": 3.9457614496710933,
+      "grad_norm": 0.07229619087165308,
+      "learning_rate": 4.0529912443493246e-09,
+      "loss": 0.445,
+      "step": 7953
+    },
+    {
+      "epoch": 3.9462579123743327,
+      "grad_norm": 0.07254327033190938,
+      "learning_rate": 3.974684796355099e-09,
+      "loss": 0.4445,
+      "step": 7954
+    },
+    {
+      "epoch": 3.9467543750775724,
+      "grad_norm": 0.07171593333450604,
+      "learning_rate": 3.897141910725321e-09,
+      "loss": 0.4538,
+      "step": 7955
+    },
+    {
+      "epoch": 3.9472508377808118,
+      "grad_norm": 0.07196674522189404,
+      "learning_rate": 3.820362599311067e-09,
+      "loss": 0.4271,
+      "step": 7956
+    },
+    {
+      "epoch": 3.947747300484051,
+      "grad_norm": 0.07219094545893444,
+      "learning_rate": 3.744346873846838e-09,
+      "loss": 0.4191,
+      "step": 7957
+    },
+    {
+      "epoch": 3.9482437631872904,
+      "grad_norm": 0.07204485023562479,
+      "learning_rate": 3.669094745950008e-09,
+      "loss": 0.4072,
+      "step": 7958
+    },
+    {
+      "epoch": 3.94874022589053,
+      "grad_norm": 0.07363711105033145,
+      "learning_rate": 3.594606227121933e-09,
+      "loss": 0.4216,
+      "step": 7959
+    },
+    {
+      "epoch": 3.9492366885937695,
+      "grad_norm": 0.07456848670081903,
+      "learning_rate": 3.520881328747394e-09,
+      "loss": 0.4366,
+      "step": 7960
+    },
+    {
+      "epoch": 3.949733151297009,
+      "grad_norm": 0.07378103181258867,
+      "learning_rate": 3.4479200620934904e-09,
+      "loss": 0.4531,
+      "step": 7961
+    },
+    {
+      "epoch": 3.950229614000248,
+      "grad_norm": 0.07390658868890232,
+      "learning_rate": 3.375722438311302e-09,
+      "loss": 0.4638,
+      "step": 7962
+    },
+    {
+      "epoch": 3.9507260767034875,
+      "grad_norm": 0.07300728961397632,
+      "learning_rate": 3.304288468435335e-09,
+      "loss": 0.4514,
+      "step": 7963
+    },
+    {
+      "epoch": 3.951222539406727,
+      "grad_norm": 0.07023027048085913,
+      "learning_rate": 3.2336181633829676e-09,
+      "loss": 0.4128,
+      "step": 7964
+    },
+    {
+      "epoch": 3.9517190021099666,
+      "grad_norm": 0.07505027569492374,
+      "learning_rate": 3.16371153395445e-09,
+      "loss": 0.4764,
+      "step": 7965
+    },
+    {
+      "epoch": 3.952215464813206,
+      "grad_norm": 0.073450603120353,
+      "learning_rate": 3.094568590835123e-09,
+      "loss": 0.4575,
+      "step": 7966
+    },
+    {
+      "epoch": 3.9527119275164453,
+      "grad_norm": 0.07273796979259504,
+      "learning_rate": 3.0261893445915346e-09,
+      "loss": 0.4542,
+      "step": 7967
+    },
+    {
+      "epoch": 3.9532083902196846,
+      "grad_norm": 0.07073293634698218,
+      "learning_rate": 2.958573805674214e-09,
+      "loss": 0.4398,
+      "step": 7968
+    },
+    {
+      "epoch": 3.9537048529229244,
+      "grad_norm": 0.07475951665447415,
+      "learning_rate": 2.8917219844176724e-09,
+      "loss": 0.4416,
+      "step": 7969
+    },
+    {
+      "epoch": 3.9542013156261637,
+      "grad_norm": 0.07299123908045202,
+      "learning_rate": 2.825633891039292e-09,
+      "loss": 0.466,
+      "step": 7970
+    },
+    {
+      "epoch": 3.954697778329403,
+      "grad_norm": 0.0726272322838916,
+      "learning_rate": 2.760309535638772e-09,
+      "loss": 0.4298,
+      "step": 7971
+    },
+    {
+      "epoch": 3.9551942410326424,
+      "grad_norm": 0.07061022132016685,
+      "learning_rate": 2.6957489281997927e-09,
+      "loss": 0.4039,
+      "step": 7972
+    },
+    {
+      "epoch": 3.9556907037358817,
+      "grad_norm": 0.07486006728424369,
+      "learning_rate": 2.631952078590572e-09,
+      "loss": 0.4831,
+      "step": 7973
+    },
+    {
+      "epoch": 3.956187166439121,
+      "grad_norm": 0.0732371630222974,
+      "learning_rate": 2.568918996560532e-09,
+      "loss": 0.4318,
+      "step": 7974
+    },
+    {
+      "epoch": 3.956683629142361,
+      "grad_norm": 0.07101131445407827,
+      "learning_rate": 2.5066496917436346e-09,
+      "loss": 0.4261,
+      "step": 7975
+    },
+    {
+      "epoch": 3.9571800918456,
+      "grad_norm": 0.06824274012821278,
+      "learning_rate": 2.4451441736567106e-09,
+      "loss": 0.4106,
+      "step": 7976
+    },
+    {
+      "epoch": 3.9576765545488395,
+      "grad_norm": 0.07220652300245425,
+      "learning_rate": 2.3844024516994634e-09,
+      "loss": 0.4322,
+      "step": 7977
+    },
+    {
+      "epoch": 3.958173017252079,
+      "grad_norm": 0.07315711417550587,
+      "learning_rate": 2.3244245351561334e-09,
+      "loss": 0.4572,
+      "step": 7978
+    },
+    {
+      "epoch": 3.9586694799553186,
+      "grad_norm": 0.07367506621638123,
+      "learning_rate": 2.2652104331921664e-09,
+      "loss": 0.4404,
+      "step": 7979
+    },
+    {
+      "epoch": 3.959165942658558,
+      "grad_norm": 0.0733571536697855,
+      "learning_rate": 2.206760154858656e-09,
+      "loss": 0.4415,
+      "step": 7980
+    },
+    {
+      "epoch": 3.9596624053617973,
+      "grad_norm": 0.07267523211050667,
+      "learning_rate": 2.149073709088456e-09,
+      "loss": 0.4413,
+      "step": 7981
+    },
+    {
+      "epoch": 3.9601588680650366,
+      "grad_norm": 0.07303058441794463,
+      "learning_rate": 2.0921511046978482e-09,
+      "loss": 0.4389,
+      "step": 7982
+    },
+    {
+      "epoch": 3.960655330768276,
+      "grad_norm": 0.07206286293004117,
+      "learning_rate": 2.0359923503859844e-09,
+      "loss": 0.4429,
+      "step": 7983
+    },
+    {
+      "epoch": 3.9611517934715152,
+      "grad_norm": 0.07229480870759866,
+      "learning_rate": 1.98059745473711e-09,
+      "loss": 0.4496,
+      "step": 7984
+    },
+    {
+      "epoch": 3.961648256174755,
+      "grad_norm": 0.07330560265167406,
+      "learning_rate": 1.9259664262166745e-09,
+      "loss": 0.4469,
+      "step": 7985
+    },
+    {
+      "epoch": 3.9621447188779944,
+      "grad_norm": 0.07347536565439004,
+      "learning_rate": 1.8720992731741104e-09,
+      "loss": 0.4503,
+      "step": 7986
+    },
+    {
+      "epoch": 3.9626411815812337,
+      "grad_norm": 0.06805756827226955,
+      "learning_rate": 1.818996003842277e-09,
+      "loss": 0.4025,
+      "step": 7987
+    },
+    {
+      "epoch": 3.963137644284473,
+      "grad_norm": 0.07707538276879311,
+      "learning_rate": 1.7666566263374596e-09,
+      "loss": 0.5152,
+      "step": 7988
+    },
+    {
+      "epoch": 3.963634106987713,
+      "grad_norm": 0.07205588748513372,
+      "learning_rate": 1.7150811486582597e-09,
+      "loss": 0.4058,
+      "step": 7989
+    },
+    {
+      "epoch": 3.964130569690952,
+      "grad_norm": 0.07140882741949466,
+      "learning_rate": 1.664269578687261e-09,
+      "loss": 0.4337,
+      "step": 7990
+    },
+    {
+      "epoch": 3.9646270323941915,
+      "grad_norm": 0.07212114622615089,
+      "learning_rate": 1.6142219241910284e-09,
+      "loss": 0.4594,
+      "step": 7991
+    },
+    {
+      "epoch": 3.965123495097431,
+      "grad_norm": 0.07371435390405398,
+      "learning_rate": 1.5649381928173335e-09,
+      "loss": 0.4353,
+      "step": 7992
+    },
+    {
+      "epoch": 3.96561995780067,
+      "grad_norm": 0.07281281216816414,
+      "learning_rate": 1.5164183920995946e-09,
+      "loss": 0.4208,
+      "step": 7993
+    },
+    {
+      "epoch": 3.9661164205039094,
+      "grad_norm": 0.07163297655742279,
+      "learning_rate": 1.4686625294524359e-09,
+      "loss": 0.4264,
+      "step": 7994
+    },
+    {
+      "epoch": 3.966612883207149,
+      "grad_norm": 0.07175362399757045,
+      "learning_rate": 1.421670612175019e-09,
+      "loss": 0.4193,
+      "step": 7995
+    },
+    {
+      "epoch": 3.9671093459103886,
+      "grad_norm": 0.07426390382442767,
+      "learning_rate": 1.3754426474488213e-09,
+      "loss": 0.4553,
+      "step": 7996
+    },
+    {
+      "epoch": 3.967605808613628,
+      "grad_norm": 0.07549104779425102,
+      "learning_rate": 1.3299786423393025e-09,
+      "loss": 0.434,
+      "step": 7997
+    },
+    {
+      "epoch": 3.968102271316867,
+      "grad_norm": 0.07037254959583114,
+      "learning_rate": 1.285278603795348e-09,
+      "loss": 0.4229,
+      "step": 7998
+    },
+    {
+      "epoch": 3.968598734020107,
+      "grad_norm": 0.07276741874477172,
+      "learning_rate": 1.2413425386481604e-09,
+      "loss": 0.4719,
+      "step": 7999
+    },
+    {
+      "epoch": 3.9690951967233463,
+      "grad_norm": 0.07322622193035082,
+      "learning_rate": 1.1981704536129234e-09,
+      "loss": 0.4335,
+      "step": 8000
+    },
+    {
+      "epoch": 3.9695916594265857,
+      "grad_norm": 0.07225963682092588,
+      "learning_rate": 1.1557623552871378e-09,
+      "loss": 0.4515,
+      "step": 8001
+    },
+    {
+      "epoch": 3.970088122129825,
+      "grad_norm": 0.06965915089959167,
+      "learning_rate": 1.1141182501533954e-09,
+      "loss": 0.4038,
+      "step": 8002
+    },
+    {
+      "epoch": 3.9705845848330643,
+      "grad_norm": 0.0702457996988929,
+      "learning_rate": 1.0732381445749396e-09,
+      "loss": 0.4679,
+      "step": 8003
+    },
+    {
+      "epoch": 3.9710810475363036,
+      "grad_norm": 0.07355844745120076,
+      "learning_rate": 1.0331220448006606e-09,
+      "loss": 0.4619,
+      "step": 8004
+    },
+    {
+      "epoch": 3.9715775102395434,
+      "grad_norm": 0.07306605760434766,
+      "learning_rate": 9.937699569617654e-10,
+      "loss": 0.4395,
+      "step": 8005
+    },
+    {
+      "epoch": 3.9720739729427827,
+      "grad_norm": 0.07274946485326253,
+      "learning_rate": 9.551818870712214e-10,
+      "loss": 0.4733,
+      "step": 8006
+    },
+    {
+      "epoch": 3.972570435646022,
+      "grad_norm": 0.07309474485727571,
+      "learning_rate": 9.173578410281992e-10,
+      "loss": 0.4631,
+      "step": 8007
+    },
+    {
+      "epoch": 3.9730668983492614,
+      "grad_norm": 0.0725914055312676,
+      "learning_rate": 8.802978246130745e-10,
+      "loss": 0.4182,
+      "step": 8008
+    },
+    {
+      "epoch": 3.973563361052501,
+      "grad_norm": 0.0739704011459376,
+      "learning_rate": 8.440018434890951e-10,
+      "loss": 0.4589,
+      "step": 8009
+    },
+    {
+      "epoch": 3.9740598237557405,
+      "grad_norm": 0.07236480524788812,
+      "learning_rate": 8.084699032040455e-10,
+      "loss": 0.4535,
+      "step": 8010
+    },
+    {
+      "epoch": 3.97455628645898,
+      "grad_norm": 0.07415395830044395,
+      "learning_rate": 7.737020091885816e-10,
+      "loss": 0.4642,
+      "step": 8011
+    },
+    {
+      "epoch": 3.975052749162219,
+      "grad_norm": 0.07422221753321762,
+      "learning_rate": 7.396981667562308e-10,
+      "loss": 0.47,
+      "step": 8012
+    },
+    {
+      "epoch": 3.9755492118654585,
+      "grad_norm": 0.07579519395700349,
+      "learning_rate": 7.064583811039472e-10,
+      "loss": 0.4164,
+      "step": 8013
+    },
+    {
+      "epoch": 3.976045674568698,
+      "grad_norm": 0.07189369806693677,
+      "learning_rate": 6.739826573121111e-10,
+      "loss": 0.4366,
+      "step": 8014
+    },
+    {
+      "epoch": 3.976542137271937,
+      "grad_norm": 0.07405554121755314,
+      "learning_rate": 6.422710003439747e-10,
+      "loss": 0.4445,
+      "step": 8015
+    },
+    {
+      "epoch": 3.977038599975177,
+      "grad_norm": 0.07085365042841145,
+      "learning_rate": 6.113234150462166e-10,
+      "loss": 0.4333,
+      "step": 8016
+    },
+    {
+      "epoch": 3.9775350626784163,
+      "grad_norm": 0.07432592377837327,
+      "learning_rate": 5.811399061478318e-10,
+      "loss": 0.4393,
+      "step": 8017
+    },
+    {
+      "epoch": 3.9780315253816556,
+      "grad_norm": 0.07604342053211292,
+      "learning_rate": 5.517204782634622e-10,
+      "loss": 0.428,
+      "step": 8018
+    },
+    {
+      "epoch": 3.9785279880848954,
+      "grad_norm": 0.07578805782078543,
+      "learning_rate": 5.230651358878458e-10,
+      "loss": 0.4992,
+      "step": 8019
+    },
+    {
+      "epoch": 3.9790244507881347,
+      "grad_norm": 0.07353741636356836,
+      "learning_rate": 4.951738834019226e-10,
+      "loss": 0.4536,
+      "step": 8020
+    },
+    {
+      "epoch": 3.979520913491374,
+      "grad_norm": 0.07402907843843444,
+      "learning_rate": 4.680467250672837e-10,
+      "loss": 0.4799,
+      "step": 8021
+    },
+    {
+      "epoch": 3.9800173761946134,
+      "grad_norm": 0.07183089412737431,
+      "learning_rate": 4.416836650300571e-10,
+      "loss": 0.44,
+      "step": 8022
+    },
+    {
+      "epoch": 3.9805138388978527,
+      "grad_norm": 0.07147423691021713,
+      "learning_rate": 4.160847073203522e-10,
+      "loss": 0.4499,
+      "step": 8023
+    },
+    {
+      "epoch": 3.981010301601092,
+      "grad_norm": 0.07213831950971986,
+      "learning_rate": 3.912498558494848e-10,
+      "loss": 0.4279,
+      "step": 8024
+    },
+    {
+      "epoch": 3.9815067643043314,
+      "grad_norm": 0.0728362024466595,
+      "learning_rate": 3.671791144133074e-10,
+      "loss": 0.4278,
+      "step": 8025
+    },
+    {
+      "epoch": 3.982003227007571,
+      "grad_norm": 0.07470742291399306,
+      "learning_rate": 3.438724866910992e-10,
+      "loss": 0.4686,
+      "step": 8026
+    },
+    {
+      "epoch": 3.9824996897108105,
+      "grad_norm": 0.07293936423797455,
+      "learning_rate": 3.213299762444555e-10,
+      "loss": 0.4369,
+      "step": 8027
+    },
+    {
+      "epoch": 3.98299615241405,
+      "grad_norm": 0.07378429528311958,
+      "learning_rate": 2.995515865183984e-10,
+      "loss": 0.426,
+      "step": 8028
+    },
+    {
+      "epoch": 3.9834926151172896,
+      "grad_norm": 0.07632103507293397,
+      "learning_rate": 2.7853732084248687e-10,
+      "loss": 0.4461,
+      "step": 8029
+    },
+    {
+      "epoch": 3.983989077820529,
+      "grad_norm": 0.0726281969225929,
+      "learning_rate": 2.5828718242693064e-10,
+      "loss": 0.4328,
+      "step": 8030
+    },
+    {
+      "epoch": 3.9844855405237682,
+      "grad_norm": 0.07534194717846308,
+      "learning_rate": 2.3880117436814176e-10,
+      "loss": 0.4417,
+      "step": 8031
+    },
+    {
+      "epoch": 3.9849820032270076,
+      "grad_norm": 0.0765137544246794,
+      "learning_rate": 2.200792996431833e-10,
+      "loss": 0.4883,
+      "step": 8032
+    },
+    {
+      "epoch": 3.985478465930247,
+      "grad_norm": 0.07125139382665711,
+      "learning_rate": 2.0212156111365512e-10,
+      "loss": 0.4274,
+      "step": 8033
+    },
+    {
+      "epoch": 3.9859749286334862,
+      "grad_norm": 0.07279854809314834,
+      "learning_rate": 1.849279615240285e-10,
+      "loss": 0.4246,
+      "step": 8034
+    },
+    {
+      "epoch": 3.9864713913367256,
+      "grad_norm": 0.07217756295907013,
+      "learning_rate": 1.6849850350275643e-10,
+      "loss": 0.4515,
+      "step": 8035
+    },
+    {
+      "epoch": 3.9869678540399653,
+      "grad_norm": 0.07128111812935736,
+      "learning_rate": 1.528331895600532e-10,
+      "loss": 0.4098,
+      "step": 8036
+    },
+    {
+      "epoch": 3.9874643167432047,
+      "grad_norm": 0.07351939432494661,
+      "learning_rate": 1.3793202209011481e-10,
+      "loss": 0.4502,
+      "step": 8037
+    },
+    {
+      "epoch": 3.987960779446444,
+      "grad_norm": 0.07180638250150356,
+      "learning_rate": 1.2379500337056372e-10,
+      "loss": 0.4431,
+      "step": 8038
+    },
+    {
+      "epoch": 3.988457242149684,
+      "grad_norm": 0.07389007991886898,
+      "learning_rate": 1.1042213556244907e-10,
+      "loss": 0.4535,
+      "step": 8039
+    },
+    {
+      "epoch": 3.988953704852923,
+      "grad_norm": 0.07322600834668609,
+      "learning_rate": 9.781342070913635e-11,
+      "loss": 0.4267,
+      "step": 8040
+    },
+    {
+      "epoch": 3.9894501675561624,
+      "grad_norm": 0.07292913398722783,
+      "learning_rate": 8.596886073741761e-11,
+      "loss": 0.4277,
+      "step": 8041
+    },
+    {
+      "epoch": 3.9899466302594018,
+      "grad_norm": 0.07133143471334574,
+      "learning_rate": 7.48884574575115e-11,
+      "loss": 0.4277,
+      "step": 8042
+    },
+    {
+      "epoch": 3.990443092962641,
+      "grad_norm": 0.07206100534353532,
+      "learning_rate": 6.457221256361834e-11,
+      "loss": 0.4427,
+      "step": 8043
+    },
+    {
+      "epoch": 3.9909395556658804,
+      "grad_norm": 0.06961673178845708,
+      "learning_rate": 5.502012763225484e-11,
+      "loss": 0.4377,
+      "step": 8044
+    },
+    {
+      "epoch": 3.9914360183691198,
+      "grad_norm": 0.07234539655461314,
+      "learning_rate": 4.623220412280916e-11,
+      "loss": 0.4542,
+      "step": 8045
+    },
+    {
+      "epoch": 3.9919324810723595,
+      "grad_norm": 0.07365297569379811,
+      "learning_rate": 3.820844337865115e-11,
+      "loss": 0.4656,
+      "step": 8046
+    },
+    {
+      "epoch": 3.992428943775599,
+      "grad_norm": 0.07182593616667256,
+      "learning_rate": 3.094884662602215e-11,
+      "loss": 0.4482,
+      "step": 8047
+    },
+    {
+      "epoch": 3.992925406478838,
+      "grad_norm": 0.07176683987772409,
+      "learning_rate": 2.4453414974034972e-11,
+      "loss": 0.428,
+      "step": 8048
+    },
+    {
+      "epoch": 3.9934218691820775,
+      "grad_norm": 0.07117401113949938,
+      "learning_rate": 1.872214941633921e-11,
+      "loss": 0.4418,
+      "step": 8049
+    },
+    {
+      "epoch": 3.9939183318853173,
+      "grad_norm": 0.07501244871652403,
+      "learning_rate": 1.3755050828345717e-11,
+      "loss": 0.4443,
+      "step": 8050
+    },
+    {
+      "epoch": 3.9944147945885566,
+      "grad_norm": 0.071852960560573,
+      "learning_rate": 9.552119968891937e-12,
+      "loss": 0.445,
+      "step": 8051
+    },
+    {
+      "epoch": 3.994911257291796,
+      "grad_norm": 0.070757041234484,
+      "learning_rate": 6.1133574807969995e-12,
+      "loss": 0.4219,
+      "step": 8052
+    },
+    {
+      "epoch": 3.9954077199950353,
+      "grad_norm": 0.07282539751959509,
+      "learning_rate": 3.4387638897515065e-12,
+      "loss": 0.4521,
+      "step": 8053
+    },
+    {
+      "epoch": 3.9959041826982746,
+      "grad_norm": 0.07154126903717001,
+      "learning_rate": 1.5283396037624188e-12,
+      "loss": 0.4225,
+      "step": 8054
+    },
+    {
+      "epoch": 3.996400645401514,
+      "grad_norm": 0.07092968323664142,
+      "learning_rate": 3.82084915373504e-13,
+      "loss": 0.4317,
+      "step": 8055
+    },
+    {
+      "epoch": 3.9968971081047537,
+      "grad_norm": 0.07410447458205609,
+      "learning_rate": 0.0,
+      "loss": 0.4481,
+      "step": 8056
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8056,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 2014,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6115693053345792.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}