diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.2076465855114595,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.160980224609375,
+      "learning_rate": 5e-06,
+      "loss": 9.2944,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.949506759643555,
+      "learning_rate": 1e-05,
+      "loss": 9.0982,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.4793701171875,
+      "learning_rate": 1.5e-05,
+      "loss": 8.7635,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.469444274902344,
+      "learning_rate": 2e-05,
+      "loss": 8.4563,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.251720428466797,
+      "learning_rate": 2.5e-05,
+      "loss": 8.2254,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.546104431152344,
+      "learning_rate": 3e-05,
+      "loss": 7.8594,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.340007781982422,
+      "learning_rate": 3.5e-05,
+      "loss": 7.5332,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.783193588256836,
+      "learning_rate": 4e-05,
+      "loss": 7.1389,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.952728271484375,
+      "learning_rate": 4.5e-05,
+      "loss": 6.6343,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.553371906280518,
+      "learning_rate": 5e-05,
+      "loss": 5.9556,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.343291997909546,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 5.5788,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6398086547851562,
+      "learning_rate": 6e-05,
+      "loss": 5.2846,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.117621660232544,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 4.9595,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6999884843826294,
+      "learning_rate": 7e-05,
+      "loss": 4.7634,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7466254234313965,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 4.5928,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4681142568588257,
+      "learning_rate": 8e-05,
+      "loss": 4.4051,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4110980033874512,
+      "learning_rate": 8.5e-05,
+      "loss": 4.3196,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2879364490509033,
+      "learning_rate": 9e-05,
+      "loss": 4.2043,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2290191650390625,
+      "learning_rate": 9.5e-05,
+      "loss": 4.0071,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0562479496002197,
+      "learning_rate": 0.0001,
+      "loss": 3.901,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0664085149765015,
+      "learning_rate": 9.999999582259295e-05,
+      "loss": 3.8521,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0484060049057007,
+      "learning_rate": 9.999998329037248e-05,
+      "loss": 3.8531,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0460189580917358,
+      "learning_rate": 9.999996240334068e-05,
+      "loss": 3.7799,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9307366013526917,
+      "learning_rate": 9.999993316150106e-05,
+      "loss": 3.7412,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0051379203796387,
+      "learning_rate": 9.999989556485852e-05,
+      "loss": 3.6291,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.063181757926941,
+      "learning_rate": 9.999984961341928e-05,
+      "loss": 3.5635,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9018902778625488,
+      "learning_rate": 9.999979530719108e-05,
+      "loss": 3.5891,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8319661021232605,
+      "learning_rate": 9.999973264618297e-05,
+      "loss": 3.5904,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.875512421131134,
+      "learning_rate": 9.999966163040541e-05,
+      "loss": 3.4533,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8168103098869324,
+      "learning_rate": 9.99995822598703e-05,
+      "loss": 3.5466,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7946365475654602,
+      "learning_rate": 9.999949453459088e-05,
+      "loss": 3.5335,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7695865631103516,
+      "learning_rate": 9.999939845458179e-05,
+      "loss": 3.3519,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8073988556861877,
+      "learning_rate": 9.999929401985911e-05,
+      "loss": 3.4354,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7681397199630737,
+      "learning_rate": 9.99991812304403e-05,
+      "loss": 3.4518,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7409121990203857,
+      "learning_rate": 9.999906008634416e-05,
+      "loss": 3.4055,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8215389251708984,
+      "learning_rate": 9.9998930587591e-05,
+      "loss": 3.403,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7428345680236816,
+      "learning_rate": 9.99987927342024e-05,
+      "loss": 3.4346,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7144994139671326,
+      "learning_rate": 9.999864652620143e-05,
+      "loss": 3.3442,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7305485606193542,
+      "learning_rate": 9.999849196361251e-05,
+      "loss": 3.4202,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7057986855506897,
+      "learning_rate": 9.999832904646147e-05,
+      "loss": 3.299,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7311503291130066,
+      "learning_rate": 9.99981577747755e-05,
+      "loss": 3.332,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6890970468521118,
+      "learning_rate": 9.999797814858327e-05,
+      "loss": 3.2917,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7627695798873901,
+      "learning_rate": 9.999779016791477e-05,
+      "loss": 3.3367,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4635143280029297,
+      "learning_rate": 9.999759383280141e-05,
+      "loss": 3.2634,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8400528430938721,
+      "learning_rate": 9.9997389143276e-05,
+      "loss": 3.2858,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.750618577003479,
+      "learning_rate": 9.999717609937275e-05,
+      "loss": 3.316,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7571924924850464,
+      "learning_rate": 9.999695470112725e-05,
+      "loss": 3.3644,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7450360655784607,
+      "learning_rate": 9.99967249485765e-05,
+      "loss": 3.2841,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6675312519073486,
+      "learning_rate": 9.999648684175888e-05,
+      "loss": 3.2448,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7598224878311157,
+      "learning_rate": 9.999624038071418e-05,
+      "loss": 3.1478,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7733460068702698,
+      "learning_rate": 9.99959855654836e-05,
+      "loss": 3.2424,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7883107662200928,
+      "learning_rate": 9.999572239610968e-05,
+      "loss": 3.2829,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.639010488986969,
+      "learning_rate": 9.999545087263645e-05,
+      "loss": 3.2392,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6680151224136353,
+      "learning_rate": 9.999517099510923e-05,
+      "loss": 3.2375,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6659584641456604,
+      "learning_rate": 9.999488276357481e-05,
+      "loss": 3.2633,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6495057940483093,
+      "learning_rate": 9.999458617808137e-05,
+      "loss": 3.1589,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7257037162780762,
+      "learning_rate": 9.999428123867843e-05,
+      "loss": 3.2007,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6328142881393433,
+      "learning_rate": 9.999396794541695e-05,
+      "loss": 3.2497,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6253700256347656,
+      "learning_rate": 9.999364629834931e-05,
+      "loss": 3.1886,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6815829873085022,
+      "learning_rate": 9.999331629752924e-05,
+      "loss": 3.2416,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6481777429580688,
+      "learning_rate": 9.999297794301188e-05,
+      "loss": 3.2043,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6525105237960815,
+      "learning_rate": 9.999263123485376e-05,
+      "loss": 3.126,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6912250518798828,
+      "learning_rate": 9.999227617311282e-05,
+      "loss": 3.1982,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6551243662834167,
+      "learning_rate": 9.99919127578484e-05,
+      "loss": 3.0945,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6423667073249817,
+      "learning_rate": 9.999154098912122e-05,
+      "loss": 3.1507,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6301848888397217,
+      "learning_rate": 9.999116086699338e-05,
+      "loss": 3.0636,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6644454598426819,
+      "learning_rate": 9.999077239152843e-05,
+      "loss": 3.1549,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.649381697177887,
+      "learning_rate": 9.999037556279127e-05,
+      "loss": 3.14,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6491669416427612,
+      "learning_rate": 9.998997038084818e-05,
+      "loss": 3.1106,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6729897856712341,
+      "learning_rate": 9.998955684576692e-05,
+      "loss": 3.1246,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6555104851722717,
+      "learning_rate": 9.998913495761654e-05,
+      "loss": 3.0931,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5893439650535583,
+      "learning_rate": 9.998870471646757e-05,
+      "loss": 3.1359,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.620822548866272,
+      "learning_rate": 9.99882661223919e-05,
+      "loss": 3.1319,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5989420413970947,
+      "learning_rate": 9.998781917546278e-05,
+      "loss": 3.1262,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6204272508621216,
+      "learning_rate": 9.998736387575493e-05,
+      "loss": 3.1542,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5934432148933411,
+      "learning_rate": 9.998690022334442e-05,
+      "loss": 3.094,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6479682326316833,
+      "learning_rate": 9.998642821830873e-05,
+      "loss": 3.0546,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6053611636161804,
+      "learning_rate": 9.99859478607267e-05,
+      "loss": 3.0638,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5799993872642517,
+      "learning_rate": 9.998545915067864e-05,
+      "loss": 3.0802,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6771678924560547,
+      "learning_rate": 9.998496208824618e-05,
+      "loss": 3.0518,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6179514527320862,
+      "learning_rate": 9.99844566735124e-05,
+      "loss": 3.1331,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6352304816246033,
+      "learning_rate": 9.998394290656172e-05,
+      "loss": 3.0533,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6237308979034424,
+      "learning_rate": 9.998342078748002e-05,
+      "loss": 3.0024,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6074531674385071,
+      "learning_rate": 9.998289031635453e-05,
+      "loss": 3.1082,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6042516827583313,
+      "learning_rate": 9.998235149327389e-05,
+      "loss": 3.0204,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6022238731384277,
+      "learning_rate": 9.998180431832816e-05,
+      "loss": 3.0845,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6218652725219727,
+      "learning_rate": 9.998124879160872e-05,
+      "loss": 3.1369,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6044962406158447,
+      "learning_rate": 9.998068491320844e-05,
+      "loss": 2.9992,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7303677201271057,
+      "learning_rate": 9.998011268322152e-05,
+      "loss": 3.0496,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6162992715835571,
+      "learning_rate": 9.997953210174361e-05,
+      "loss": 3.0851,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58905029296875,
+      "learning_rate": 9.997894316887167e-05,
+      "loss": 3.1067,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5957860350608826,
+      "learning_rate": 9.997834588470414e-05,
+      "loss": 3.107,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5831341743469238,
+      "learning_rate": 9.997774024934082e-05,
+      "loss": 2.9424,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6075407266616821,
+      "learning_rate": 9.997712626288292e-05,
+      "loss": 3.0407,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5903592705726624,
+      "learning_rate": 9.997650392543303e-05,
+      "loss": 2.9692,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5757583379745483,
+      "learning_rate": 9.997587323709512e-05,
+      "loss": 2.9844,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6342194080352783,
+      "learning_rate": 9.997523419797459e-05,
+      "loss": 2.96,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.557708203792572,
+      "learning_rate": 9.997458680817824e-05,
+      "loss": 2.9914,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6286576986312866,
+      "learning_rate": 9.997393106781421e-05,
+      "loss": 2.9658,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5779178738594055,
+      "learning_rate": 9.99732669769921e-05,
+      "loss": 2.9706,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.583254873752594,
+      "learning_rate": 9.997259453582286e-05,
+      "loss": 2.9969,
+      "step": 505
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5486466288566589,
+      "learning_rate": 9.997191374441887e-05,
+      "loss": 3.04,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5430684089660645,
+      "learning_rate": 9.997122460289387e-05,
+      "loss": 2.9766,
+      "step": 515
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6035254001617432,
+      "learning_rate": 9.997052711136301e-05,
+      "loss": 3.0756,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5668070316314697,
+      "learning_rate": 9.996982126994287e-05,
+      "loss": 3.0363,
+      "step": 525
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.532256007194519,
+      "learning_rate": 9.996910707875136e-05,
+      "loss": 3.0188,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6169216632843018,
+      "learning_rate": 9.996838453790784e-05,
+      "loss": 2.9957,
+      "step": 535
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.543561577796936,
+      "learning_rate": 9.996765364753302e-05,
+      "loss": 2.9775,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5566686391830444,
+      "learning_rate": 9.996691440774906e-05,
+      "loss": 2.9691,
+      "step": 545
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5432292222976685,
+      "learning_rate": 9.996616681867945e-05,
+      "loss": 2.9525,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5466824173927307,
+      "learning_rate": 9.996541088044913e-05,
+      "loss": 2.9389,
+      "step": 555
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58321213722229,
+      "learning_rate": 9.996464659318442e-05,
+      "loss": 3.0056,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5711489915847778,
+      "learning_rate": 9.996387395701302e-05,
+      "loss": 2.9395,
+      "step": 565
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5725634098052979,
+      "learning_rate": 9.996309297206403e-05,
+      "loss": 3.0319,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5800571441650391,
+      "learning_rate": 9.996230363846797e-05,
+      "loss": 2.947,
+      "step": 575
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5644673109054565,
+      "learning_rate": 9.996150595635671e-05,
+      "loss": 2.975,
+      "step": 580
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5683979392051697,
+      "learning_rate": 9.996069992586355e-05,
+      "loss": 2.9174,
+      "step": 585
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5549993515014648,
+      "learning_rate": 9.995988554712318e-05,
+      "loss": 2.9191,
+      "step": 590
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.560452938079834,
+      "learning_rate": 9.995906282027166e-05,
+      "loss": 2.9623,
+      "step": 595
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5923082828521729,
+      "learning_rate": 9.995823174544649e-05,
+      "loss": 2.9652,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6109171509742737,
+      "learning_rate": 9.995739232278654e-05,
+      "loss": 2.9117,
+      "step": 605
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5340412855148315,
+      "learning_rate": 9.995654455243204e-05,
+      "loss": 2.9302,
+      "step": 610
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5512436032295227,
+      "learning_rate": 9.995568843452468e-05,
+      "loss": 2.9294,
+      "step": 615
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5419607162475586,
+      "learning_rate": 9.99548239692075e-05,
+      "loss": 2.9984,
+      "step": 620
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5700150728225708,
+      "learning_rate": 9.995395115662498e-05,
+      "loss": 2.931,
+      "step": 625
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5535702705383301,
+      "learning_rate": 9.995306999692293e-05,
+      "loss": 2.8863,
+      "step": 630
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5817483067512512,
+      "learning_rate": 9.995218049024858e-05,
+      "loss": 2.9504,
+      "step": 635
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5571666955947876,
+      "learning_rate": 9.995128263675058e-05,
+      "loss": 2.9315,
+      "step": 640
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5520240068435669,
+      "learning_rate": 9.995037643657899e-05,
+      "loss": 2.8846,
+      "step": 645
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5881261229515076,
+      "learning_rate": 9.994946188988516e-05,
+      "loss": 2.9243,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.556154727935791,
+      "learning_rate": 9.994853899682197e-05,
+      "loss": 2.9253,
+      "step": 655
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5392178893089294,
+      "learning_rate": 9.994760775754362e-05,
+      "loss": 2.955,
+      "step": 660
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5867845416069031,
+      "learning_rate": 9.994666817220571e-05,
+      "loss": 2.9416,
+      "step": 665
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5594280362129211,
+      "learning_rate": 9.994572024096521e-05,
+      "loss": 2.9694,
+      "step": 670
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.514578640460968,
+      "learning_rate": 9.994476396398055e-05,
+      "loss": 2.8679,
+      "step": 675
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5166343450546265,
+      "learning_rate": 9.994379934141153e-05,
+      "loss": 2.8414,
+      "step": 680
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5623914003372192,
+      "learning_rate": 9.994282637341932e-05,
+      "loss": 2.9364,
+      "step": 685
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5347405076026917,
+      "learning_rate": 9.994184506016651e-05,
+      "loss": 2.8684,
+      "step": 690
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.52596116065979,
+      "learning_rate": 9.994085540181703e-05,
+      "loss": 2.8634,
+      "step": 695
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5294895172119141,
+      "learning_rate": 9.99398573985363e-05,
+      "loss": 2.8834,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8429256081581116,
+      "learning_rate": 9.993885105049107e-05,
+      "loss": 2.904,
+      "step": 705
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5511649250984192,
+      "learning_rate": 9.993783635784947e-05,
+      "loss": 2.94,
+      "step": 710
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.551913857460022,
+      "learning_rate": 9.993681332078108e-05,
+      "loss": 2.9614,
+      "step": 715
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5377745032310486,
+      "learning_rate": 9.993578193945685e-05,
+      "loss": 2.8614,
+      "step": 720
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5551316738128662,
+      "learning_rate": 9.99347422140491e-05,
+      "loss": 2.9005,
+      "step": 725
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5425111055374146,
+      "learning_rate": 9.993369414473157e-05,
+      "loss": 2.8611,
+      "step": 730
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5574254989624023,
+      "learning_rate": 9.993263773167939e-05,
+      "loss": 2.952,
+      "step": 735
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5425921678543091,
+      "learning_rate": 9.993157297506907e-05,
+      "loss": 2.9328,
+      "step": 740
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5733354687690735,
+      "learning_rate": 9.993049987507856e-05,
+      "loss": 2.8503,
+      "step": 745
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5580013990402222,
+      "learning_rate": 9.992941843188715e-05,
+      "loss": 2.9358,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5442509651184082,
+      "learning_rate": 9.992832864567554e-05,
+      "loss": 2.9079,
+      "step": 755
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5204207301139832,
+      "learning_rate": 9.992723051662584e-05,
+      "loss": 2.8611,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5190364122390747,
+      "learning_rate": 9.992612404492153e-05,
+      "loss": 2.9282,
+      "step": 765
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5260244607925415,
+      "learning_rate": 9.992500923074752e-05,
+      "loss": 2.8352,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5538665652275085,
+      "learning_rate": 9.992388607429006e-05,
+      "loss": 2.9037,
+      "step": 775
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5304633378982544,
+      "learning_rate": 9.992275457573685e-05,
+      "loss": 2.939,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5454950928688049,
+      "learning_rate": 9.992161473527695e-05,
+      "loss": 2.8539,
+      "step": 785
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5111319422721863,
+      "learning_rate": 9.992046655310083e-05,
+      "loss": 2.8266,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5481424927711487,
+      "learning_rate": 9.991931002940034e-05,
+      "loss": 2.9467,
+      "step": 795
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5543172955513,
+      "learning_rate": 9.991814516436873e-05,
+      "loss": 2.9136,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.530357837677002,
+      "learning_rate": 9.991697195820065e-05,
+      "loss": 2.8215,
+      "step": 805
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5149588584899902,
+      "learning_rate": 9.991579041109212e-05,
+      "loss": 2.8568,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5527454018592834,
+      "learning_rate": 9.991460052324061e-05,
+      "loss": 2.911,
+      "step": 815
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5437872409820557,
+      "learning_rate": 9.991340229484491e-05,
+      "loss": 2.795,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5620787143707275,
+      "learning_rate": 9.991219572610526e-05,
+      "loss": 2.8531,
+      "step": 825
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5435457229614258,
+      "learning_rate": 9.991098081722325e-05,
+      "loss": 2.862,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5630514621734619,
+      "learning_rate": 9.990975756840193e-05,
+      "loss": 2.8975,
+      "step": 835
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5202612280845642,
+      "learning_rate": 9.990852597984566e-05,
+      "loss": 2.8694,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5396909713745117,
+      "learning_rate": 9.990728605176025e-05,
+      "loss": 2.7987,
+      "step": 845
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.522987961769104,
+      "learning_rate": 9.990603778435288e-05,
+      "loss": 2.8292,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5228624939918518,
+      "learning_rate": 9.990478117783214e-05,
+      "loss": 2.8059,
+      "step": 855
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5820401906967163,
+      "learning_rate": 9.990351623240799e-05,
+      "loss": 2.8827,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4799802303314209,
+      "learning_rate": 9.990224294829181e-05,
+      "loss": 2.882,
+      "step": 865
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5819080471992493,
+      "learning_rate": 9.990096132569636e-05,
+      "loss": 2.9988,
+      "step": 870
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6007246375083923,
+      "learning_rate": 9.98996713648358e-05,
+      "loss": 2.7952,
+      "step": 875
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5428991913795471,
+      "learning_rate": 9.989837306592567e-05,
+      "loss": 2.9317,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5621552467346191,
+      "learning_rate": 9.989706642918291e-05,
+      "loss": 2.8335,
+      "step": 885
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5007776021957397,
+      "learning_rate": 9.989575145482583e-05,
+      "loss": 2.7971,
+      "step": 890
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5121883153915405,
+      "learning_rate": 9.989442814307422e-05,
+      "loss": 2.9152,
+      "step": 895
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5302690863609314,
+      "learning_rate": 9.989309649414914e-05,
+      "loss": 2.8574,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5046696662902832,
+      "learning_rate": 9.989175650827314e-05,
+      "loss": 2.8988,
+      "step": 905
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5041911005973816,
+      "learning_rate": 9.98904081856701e-05,
+      "loss": 2.9083,
+      "step": 910
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5146932005882263,
+      "learning_rate": 9.988905152656534e-05,
+      "loss": 2.8666,
+      "step": 915
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5520676970481873,
+      "learning_rate": 9.988768653118554e-05,
+      "loss": 2.9537,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5169224143028259,
+      "learning_rate": 9.988631319975881e-05,
+      "loss": 2.8081,
+      "step": 925
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5197150111198425,
+      "learning_rate": 9.988493153251459e-05,
+      "loss": 2.8594,
+      "step": 930
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5101742148399353,
+      "learning_rate": 9.988354152968379e-05,
+      "loss": 2.7817,
+      "step": 935
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49797070026397705,
+      "learning_rate": 9.988214319149865e-05,
+      "loss": 2.8427,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5219241976737976,
+      "learning_rate": 9.988073651819281e-05,
+      "loss": 2.8797,
+      "step": 945
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5001295208930969,
+      "learning_rate": 9.987932151000137e-05,
+      "loss": 2.835,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5077839493751526,
+      "learning_rate": 9.987789816716073e-05,
+      "loss": 2.8477,
+      "step": 955
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5334168076515198,
+      "learning_rate": 9.987646648990876e-05,
+      "loss": 2.8305,
+      "step": 960
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6205976009368896,
+      "learning_rate": 9.987502647848466e-05,
+      "loss": 2.8197,
+      "step": 965
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5302903056144714,
+      "learning_rate": 9.987357813312905e-05,
+      "loss": 2.8581,
+      "step": 970
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4849225878715515,
+      "learning_rate": 9.987212145408395e-05,
+      "loss": 2.8229,
+      "step": 975
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5107173919677734,
+      "learning_rate": 9.987065644159278e-05,
+      "loss": 2.797,
+      "step": 980
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5267950892448425,
+      "learning_rate": 9.986918309590031e-05,
+      "loss": 2.8294,
+      "step": 985
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5336316227912903,
+      "learning_rate": 9.986770141725277e-05,
+      "loss": 2.7954,
+      "step": 990
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5624336004257202,
+      "learning_rate": 9.986621140589772e-05,
+      "loss": 2.8395,
+      "step": 995
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.551811933517456,
+      "learning_rate": 9.986471306208412e-05,
+      "loss": 2.7963,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5234202146530151,
+      "learning_rate": 9.986320638606235e-05,
+      "loss": 2.8197,
+      "step": 1005
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5127961039543152,
+      "learning_rate": 9.986169137808419e-05,
+      "loss": 2.8468,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47998014092445374,
+      "learning_rate": 9.986016803840275e-05,
+      "loss": 2.7798,
+      "step": 1015
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5014148950576782,
+      "learning_rate": 9.985863636727262e-05,
+      "loss": 2.8805,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.497785359621048,
+      "learning_rate": 9.985709636494971e-05,
+      "loss": 2.8157,
+      "step": 1025
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5217312574386597,
+      "learning_rate": 9.985554803169134e-05,
+      "loss": 2.7743,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2653827667236328,
+      "learning_rate": 9.985399136775627e-05,
+      "loss": 2.7561,
+      "step": 1035
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5163415670394897,
+      "learning_rate": 9.985242637340458e-05,
+      "loss": 2.7324,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48676174879074097,
+      "learning_rate": 9.985085304889778e-05,
+      "loss": 2.8276,
+      "step": 1045
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5088127851486206,
+      "learning_rate": 9.984927139449877e-05,
+      "loss": 2.7711,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49691009521484375,
+      "learning_rate": 9.984768141047184e-05,
+      "loss": 2.901,
+      "step": 1055
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5436579585075378,
+      "learning_rate": 9.984608309708266e-05,
+      "loss": 2.8264,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4915987253189087,
+      "learning_rate": 9.984447645459831e-05,
+      "loss": 2.8711,
+      "step": 1065
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5676747560501099,
+      "learning_rate": 9.984286148328728e-05,
+      "loss": 2.8161,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5167434215545654,
+      "learning_rate": 9.984123818341937e-05,
+      "loss": 2.8174,
+      "step": 1075
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5058668255805969,
+      "learning_rate": 9.983960655526587e-05,
+      "loss": 2.8798,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4952792227268219,
+      "learning_rate": 9.983796659909941e-05,
+      "loss": 2.8375,
+      "step": 1085
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5071784257888794,
+      "learning_rate": 9.983631831519401e-05,
+      "loss": 2.8344,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4810205101966858,
+      "learning_rate": 9.983466170382511e-05,
+      "loss": 2.745,
+      "step": 1095
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5021117925643921,
+      "learning_rate": 9.983299676526949e-05,
+      "loss": 2.8164,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5290012359619141,
+      "learning_rate": 9.98313234998054e-05,
+      "loss": 2.7549,
+      "step": 1105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48396316170692444,
+      "learning_rate": 9.982964190771242e-05,
+      "loss": 2.8327,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5137316584587097,
+      "learning_rate": 9.98279519892715e-05,
+      "loss": 2.7648,
+      "step": 1115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49960365891456604,
+      "learning_rate": 9.982625374476508e-05,
+      "loss": 2.6926,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48757389187812805,
+      "learning_rate": 9.98245471744769e-05,
+      "loss": 2.7819,
+      "step": 1125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.482384592294693,
+      "learning_rate": 9.982283227869211e-05,
+      "loss": 2.7444,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5238838791847229,
+      "learning_rate": 9.982110905769729e-05,
+      "loss": 2.7466,
+      "step": 1135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5288039445877075,
+      "learning_rate": 9.981937751178037e-05,
+      "loss": 2.8421,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5548570156097412,
+      "learning_rate": 9.981763764123067e-05,
+      "loss": 2.8427,
+      "step": 1145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4949299991130829,
+      "learning_rate": 9.981588944633893e-05,
+      "loss": 2.8078,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5175497531890869,
+      "learning_rate": 9.981413292739727e-05,
+      "loss": 2.8117,
+      "step": 1155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48642298579216003,
+      "learning_rate": 9.981236808469922e-05,
+      "loss": 2.7792,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49451473355293274,
+      "learning_rate": 9.981059491853964e-05,
+      "loss": 2.7458,
+      "step": 1165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5019437074661255,
+      "learning_rate": 9.980881342921482e-05,
+      "loss": 2.7379,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5109878182411194,
+      "learning_rate": 9.980702361702246e-05,
+      "loss": 2.6942,
+      "step": 1175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4959583580493927,
+      "learning_rate": 9.980522548226162e-05,
+      "loss": 2.777,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49855005741119385,
+      "learning_rate": 9.980341902523279e-05,
+      "loss": 2.8059,
+      "step": 1185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5099169611930847,
+      "learning_rate": 9.980160424623776e-05,
+      "loss": 2.7694,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5042619109153748,
+      "learning_rate": 9.979978114557984e-05,
+      "loss": 2.7244,
+      "step": 1195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4988323152065277,
+      "learning_rate": 9.979794972356362e-05,
+      "loss": 2.7859,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5333328247070312,
+      "learning_rate": 9.979610998049516e-05,
+      "loss": 2.7653,
+      "step": 1205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5010849237442017,
+      "learning_rate": 9.979426191668182e-05,
+      "loss": 2.8015,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5304666757583618,
+      "learning_rate": 9.979240553243246e-05,
+      "loss": 2.6927,
+      "step": 1215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5194066166877747,
+      "learning_rate": 9.979054082805726e-05,
+      "loss": 2.7521,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49911245703697205,
+      "learning_rate": 9.978866780386777e-05,
+      "loss": 2.7792,
+      "step": 1225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5120240449905396,
+      "learning_rate": 9.978678646017701e-05,
+      "loss": 2.7592,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5028381943702698,
+      "learning_rate": 9.978489679729934e-05,
+      "loss": 2.8505,
+      "step": 1235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4768087863922119,
+      "learning_rate": 9.978299881555048e-05,
+      "loss": 2.8349,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5030097365379333,
+      "learning_rate": 9.97810925152476e-05,
+      "loss": 2.7769,
+      "step": 1245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5423885583877563,
+      "learning_rate": 9.977917789670925e-05,
+      "loss": 2.7514,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5045027732849121,
+      "learning_rate": 9.977725496025535e-05,
+      "loss": 2.8112,
+      "step": 1255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49109435081481934,
+      "learning_rate": 9.977532370620718e-05,
+      "loss": 2.807,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48550117015838623,
+      "learning_rate": 9.977338413488748e-05,
+      "loss": 2.7612,
+      "step": 1265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4963570237159729,
+      "learning_rate": 9.977143624662035e-05,
+      "loss": 2.7751,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48572275042533875,
+      "learning_rate": 9.976948004173125e-05,
+      "loss": 2.7192,
+      "step": 1275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4798545241355896,
+      "learning_rate": 9.976751552054706e-05,
+      "loss": 2.8371,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48611730337142944,
+      "learning_rate": 9.976554268339607e-05,
+      "loss": 2.8112,
+      "step": 1285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47927358746528625,
+      "learning_rate": 9.97635615306079e-05,
+      "loss": 2.721,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5214134454727173,
+      "learning_rate": 9.976157206251361e-05,
+      "loss": 2.7476,
+      "step": 1295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4856017529964447,
+      "learning_rate": 9.975957427944563e-05,
+      "loss": 2.7951,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.495883047580719,
+      "learning_rate": 9.975756818173778e-05,
+      "loss": 2.7558,
+      "step": 1305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4841471016407013,
+      "learning_rate": 9.975555376972529e-05,
+      "loss": 2.708,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5039588809013367,
+      "learning_rate": 9.975353104374474e-05,
+      "loss": 2.6686,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5142002105712891,
+      "learning_rate": 9.975150000413411e-05,
+      "loss": 2.7949,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5143137574195862,
+      "learning_rate": 9.97494606512328e-05,
+      "loss": 2.7997,
+      "step": 1325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4714558720588684,
+      "learning_rate": 9.974741298538158e-05,
+      "loss": 2.7144,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49831250309944153,
+      "learning_rate": 9.974535700692257e-05,
+      "loss": 2.6826,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48340100049972534,
+      "learning_rate": 9.974329271619938e-05,
+      "loss": 2.7369,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4722755551338196,
+      "learning_rate": 9.97412201135569e-05,
+      "loss": 2.7252,
+      "step": 1345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4970778822898865,
+      "learning_rate": 9.973913919934147e-05,
+      "loss": 2.6869,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.464353084564209,
+      "learning_rate": 9.973704997390079e-05,
+      "loss": 2.73,
+      "step": 1355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4834035634994507,
+      "learning_rate": 9.973495243758398e-05,
+      "loss": 2.6838,
+      "step": 1360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47598889470100403,
+      "learning_rate": 9.973284659074151e-05,
+      "loss": 2.7753,
+      "step": 1365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46175307035446167,
+      "learning_rate": 9.973073243372528e-05,
+      "loss": 2.7773,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5173929929733276,
+      "learning_rate": 9.972860996688853e-05,
+      "loss": 2.6713,
+      "step": 1375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46748581528663635,
+      "learning_rate": 9.972647919058595e-05,
+      "loss": 2.7839,
+      "step": 1380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49214664101600647,
+      "learning_rate": 9.972434010517358e-05,
+      "loss": 2.6958,
+      "step": 1385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4839484393596649,
+      "learning_rate": 9.972219271100882e-05,
+      "loss": 2.7914,
+      "step": 1390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.501247227191925,
+      "learning_rate": 9.972003700845052e-05,
+      "loss": 2.8315,
+      "step": 1395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4902189075946808,
+      "learning_rate": 9.971787299785888e-05,
+      "loss": 2.8166,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4697709381580353,
+      "learning_rate": 9.97157006795955e-05,
+      "loss": 2.797,
+      "step": 1405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4714774191379547,
+      "learning_rate": 9.971352005402338e-05,
+      "loss": 2.7549,
+      "step": 1410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46433117985725403,
+      "learning_rate": 9.971133112150688e-05,
+      "loss": 2.7234,
+      "step": 1415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4823758900165558,
+      "learning_rate": 9.970913388241177e-05,
+      "loss": 2.6864,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4849538803100586,
+      "learning_rate": 9.970692833710517e-05,
+      "loss": 2.7507,
+      "step": 1425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5179776549339294,
+      "learning_rate": 9.970471448595566e-05,
+      "loss": 2.7641,
+      "step": 1430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4658927321434021,
+      "learning_rate": 9.970249232933315e-05,
+      "loss": 2.8148,
+      "step": 1435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5238528251647949,
+      "learning_rate": 9.970026186760896e-05,
+      "loss": 2.6929,
+      "step": 1440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49536195397377014,
+      "learning_rate": 9.969802310115577e-05,
+      "loss": 2.6731,
+      "step": 1445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4937048852443695,
+      "learning_rate": 9.96957760303477e-05,
+      "loss": 2.7598,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47988590598106384,
+      "learning_rate": 9.969352065556022e-05,
+      "loss": 2.7501,
+      "step": 1455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5316676497459412,
+      "learning_rate": 9.969125697717017e-05,
+      "loss": 2.7495,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4968559145927429,
+      "learning_rate": 9.968898499555582e-05,
+      "loss": 2.7515,
+      "step": 1465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4648337662220001,
+      "learning_rate": 9.968670471109682e-05,
+      "loss": 2.7506,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4796505868434906,
+      "learning_rate": 9.968441612417418e-05,
+      "loss": 2.7654,
+      "step": 1475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.510708212852478,
+      "learning_rate": 9.968211923517032e-05,
+      "loss": 2.725,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4823015332221985,
+      "learning_rate": 9.967981404446905e-05,
+      "loss": 2.7221,
+      "step": 1485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.120663642883301,
+      "learning_rate": 9.967750055245555e-05,
+      "loss": 2.8237,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47696441411972046,
+      "learning_rate": 9.967517875951638e-05,
+      "loss": 2.7311,
+      "step": 1495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4683150053024292,
+      "learning_rate": 9.967284866603955e-05,
+      "loss": 2.8298,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5164295434951782,
+      "learning_rate": 9.967051027241436e-05,
+      "loss": 2.6982,
+      "step": 1505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4870600998401642,
+      "learning_rate": 9.966816357903155e-05,
+      "loss": 2.6942,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49420836567878723,
+      "learning_rate": 9.96658085862833e-05,
+      "loss": 2.7344,
+      "step": 1515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5017637014389038,
+      "learning_rate": 9.966344529456305e-05,
+      "loss": 2.7207,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4823729395866394,
+      "learning_rate": 9.966107370426573e-05,
+      "loss": 2.7539,
+      "step": 1525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4994487464427948,
+      "learning_rate": 9.965869381578765e-05,
+      "loss": 2.7598,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5090956091880798,
+      "learning_rate": 9.965630562952643e-05,
+      "loss": 2.7031,
+      "step": 1535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5098832845687866,
+      "learning_rate": 9.965390914588115e-05,
+      "loss": 2.7677,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48966941237449646,
+      "learning_rate": 9.965150436525225e-05,
+      "loss": 2.7003,
+      "step": 1545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5065357685089111,
+      "learning_rate": 9.964909128804159e-05,
+      "loss": 2.7188,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49941563606262207,
+      "learning_rate": 9.964666991465234e-05,
+      "loss": 2.706,
+      "step": 1555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.511374294757843,
+      "learning_rate": 9.964424024548914e-05,
+      "loss": 2.7565,
+      "step": 1560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4580732583999634,
+      "learning_rate": 9.964180228095793e-05,
+      "loss": 2.6838,
+      "step": 1565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48295167088508606,
+      "learning_rate": 9.963935602146612e-05,
+      "loss": 2.7807,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49552053213119507,
+      "learning_rate": 9.963690146742248e-05,
+      "loss": 2.729,
+      "step": 1575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.449606329202652,
+      "learning_rate": 9.963443861923715e-05,
+      "loss": 2.715,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5116463303565979,
+      "learning_rate": 9.963196747732164e-05,
+      "loss": 2.7219,
+      "step": 1585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4788565933704376,
+      "learning_rate": 9.96294880420889e-05,
+      "loss": 2.6921,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46824705600738525,
+      "learning_rate": 9.96270003139532e-05,
+      "loss": 2.7907,
+      "step": 1595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45418503880500793,
+      "learning_rate": 9.962450429333026e-05,
+      "loss": 2.7184,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47127294540405273,
+      "learning_rate": 9.962199998063715e-05,
+      "loss": 2.6906,
+      "step": 1605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46570026874542236,
+      "learning_rate": 9.961948737629231e-05,
+      "loss": 2.6565,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49575963616371155,
+      "learning_rate": 9.961696648071561e-05,
+      "loss": 2.6851,
+      "step": 1615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4820267856121063,
+      "learning_rate": 9.961443729432828e-05,
+      "loss": 2.6751,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4520756006240845,
+      "learning_rate": 9.961189981755294e-05,
+      "loss": 2.6879,
+      "step": 1625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4633789658546448,
+      "learning_rate": 9.960935405081357e-05,
+      "loss": 2.7423,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47979599237442017,
+      "learning_rate": 9.960679999453559e-05,
+      "loss": 2.7661,
+      "step": 1635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5591489672660828,
+      "learning_rate": 9.960423764914575e-05,
+      "loss": 2.6786,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4861874580383301,
+      "learning_rate": 9.960166701507221e-05,
+      "loss": 2.6721,
+      "step": 1645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.506767213344574,
+      "learning_rate": 9.959908809274453e-05,
+      "loss": 2.7089,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4622741937637329,
+      "learning_rate": 9.959650088259362e-05,
+      "loss": 2.647,
+      "step": 1655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4801636040210724,
+      "learning_rate": 9.959390538505181e-05,
+      "loss": 2.7563,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44462332129478455,
+      "learning_rate": 9.959130160055279e-05,
+      "loss": 2.6154,
+      "step": 1665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5009552836418152,
+      "learning_rate": 9.958868952953163e-05,
+      "loss": 2.734,
+      "step": 1670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4875224530696869,
+      "learning_rate": 9.95860691724248e-05,
+      "loss": 2.7211,
+      "step": 1675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45669278502464294,
+      "learning_rate": 9.958344052967018e-05,
+      "loss": 2.6983,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4848799705505371,
+      "learning_rate": 9.958080360170698e-05,
+      "loss": 2.7129,
+      "step": 1685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46313369274139404,
+      "learning_rate": 9.957815838897585e-05,
+      "loss": 2.5847,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.477213054895401,
+      "learning_rate": 9.957550489191874e-05,
+      "loss": 2.7088,
+      "step": 1695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.479230135679245,
+      "learning_rate": 9.957284311097908e-05,
+      "loss": 2.7188,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45532065629959106,
+      "learning_rate": 9.957017304660164e-05,
+      "loss": 2.6807,
+      "step": 1705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4766122102737427,
+      "learning_rate": 9.956749469923258e-05,
+      "loss": 2.6594,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4664555490016937,
+      "learning_rate": 9.956480806931942e-05,
+      "loss": 2.7222,
+      "step": 1715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.521416962146759,
+      "learning_rate": 9.956211315731112e-05,
+      "loss": 2.7158,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4583393633365631,
+      "learning_rate": 9.955940996365796e-05,
+      "loss": 2.6784,
+      "step": 1725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5257468819618225,
+      "learning_rate": 9.955669848881166e-05,
+      "loss": 2.6276,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45799389481544495,
+      "learning_rate": 9.955397873322527e-05,
+      "loss": 2.6276,
+      "step": 1735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47336339950561523,
+      "learning_rate": 9.955125069735327e-05,
+      "loss": 2.5799,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47599855065345764,
+      "learning_rate": 9.95485143816515e-05,
+      "loss": 2.6564,
+      "step": 1745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45576903223991394,
+      "learning_rate": 9.95457697865772e-05,
+      "loss": 2.6286,
+      "step": 1750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45217999815940857,
+      "learning_rate": 9.954301691258894e-05,
+      "loss": 2.6821,
+      "step": 1755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4789089560508728,
+      "learning_rate": 9.954025576014674e-05,
+      "loss": 2.7074,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46593227982521057,
+      "learning_rate": 9.953748632971201e-05,
+      "loss": 2.5947,
+      "step": 1765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5342175960540771,
+      "learning_rate": 9.953470862174748e-05,
+      "loss": 2.7184,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45817553997039795,
+      "learning_rate": 9.953192263671728e-05,
+      "loss": 2.7264,
+      "step": 1775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45365849137306213,
+      "learning_rate": 9.952912837508697e-05,
+      "loss": 2.6409,
+      "step": 1780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45007672905921936,
+      "learning_rate": 9.952632583732343e-05,
+      "loss": 2.5394,
+      "step": 1785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4655129015445709,
+      "learning_rate": 9.952351502389498e-05,
+      "loss": 2.6724,
+      "step": 1790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4790281653404236,
+      "learning_rate": 9.95206959352713e-05,
+      "loss": 2.6857,
+      "step": 1795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4348864257335663,
+      "learning_rate": 9.951786857192341e-05,
+      "loss": 2.6034,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47006234526634216,
+      "learning_rate": 9.951503293432378e-05,
+      "loss": 2.6805,
+      "step": 1805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6105749607086182,
+      "learning_rate": 9.951218902294625e-05,
+      "loss": 2.7566,
+      "step": 1810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4893050193786621,
+      "learning_rate": 9.9509336838266e-05,
+      "loss": 2.5982,
+      "step": 1815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4411056339740753,
+      "learning_rate": 9.950647638075963e-05,
+      "loss": 2.6809,
+      "step": 1820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45695966482162476,
+      "learning_rate": 9.950360765090511e-05,
+      "loss": 2.6528,
+      "step": 1825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4605148434638977,
+      "learning_rate": 9.950073064918179e-05,
+      "loss": 2.7023,
+      "step": 1830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4560076892375946,
+      "learning_rate": 9.949784537607042e-05,
+      "loss": 2.6536,
+      "step": 1835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46272727847099304,
+      "learning_rate": 9.949495183205311e-05,
+      "loss": 2.6191,
+      "step": 1840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4522585868835449,
+      "learning_rate": 9.949205001761334e-05,
+      "loss": 2.7353,
+      "step": 1845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46910274028778076,
+      "learning_rate": 9.948913993323603e-05,
+      "loss": 2.6895,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4879136085510254,
+      "learning_rate": 9.948622157940744e-05,
+      "loss": 2.6111,
+      "step": 1855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4619349539279938,
+      "learning_rate": 9.948329495661517e-05,
+      "loss": 2.7491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4735822081565857,
+      "learning_rate": 9.94803600653483e-05,
+      "loss": 2.6891,
+      "step": 1865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4559914469718933,
+      "learning_rate": 9.947741690609722e-05,
+      "loss": 2.6498,
+      "step": 1870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44435858726501465,
+      "learning_rate": 9.94744654793537e-05,
+      "loss": 2.767,
+      "step": 1875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47376713156700134,
+      "learning_rate": 9.947150578561095e-05,
+      "loss": 2.6386,
+      "step": 1880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46026942133903503,
+      "learning_rate": 9.946853782536351e-05,
+      "loss": 2.6239,
+      "step": 1885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46101972460746765,
+      "learning_rate": 9.946556159910732e-05,
+      "loss": 2.7182,
+      "step": 1890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45950689911842346,
+      "learning_rate": 9.946257710733966e-05,
+      "loss": 2.6276,
+      "step": 1895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46267133951187134,
+      "learning_rate": 9.945958435055927e-05,
+      "loss": 2.6397,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45391443371772766,
+      "learning_rate": 9.945658332926622e-05,
+      "loss": 2.6417,
+      "step": 1905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4590759575366974,
+      "learning_rate": 9.945357404396197e-05,
+      "loss": 2.6439,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4561498165130615,
+      "learning_rate": 9.945055649514934e-05,
+      "loss": 2.6086,
+      "step": 1915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4349052608013153,
+      "learning_rate": 9.944753068333256e-05,
+      "loss": 2.6948,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4439074695110321,
+      "learning_rate": 9.944449660901725e-05,
+      "loss": 2.6571,
+      "step": 1925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4369959235191345,
+      "learning_rate": 9.944145427271036e-05,
+      "loss": 2.7004,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.49056270718574524,
+      "learning_rate": 9.94384036749203e-05,
+      "loss": 2.6308,
+      "step": 1935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4483295679092407,
+      "learning_rate": 9.943534481615677e-05,
+      "loss": 2.713,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4711354970932007,
+      "learning_rate": 9.943227769693091e-05,
+      "loss": 2.7482,
+      "step": 1945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44868361949920654,
+      "learning_rate": 9.942920231775524e-05,
+      "loss": 2.6423,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4746469259262085,
+      "learning_rate": 9.942611867914363e-05,
+      "loss": 2.6525,
+      "step": 1955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4656049907207489,
+      "learning_rate": 9.942302678161133e-05,
+      "loss": 2.636,
+      "step": 1960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44452813267707825,
+      "learning_rate": 9.9419926625675e-05,
+      "loss": 2.6663,
+      "step": 1965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4340547025203705,
+      "learning_rate": 9.941681821185265e-05,
+      "loss": 2.6831,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45119673013687134,
+      "learning_rate": 9.941370154066371e-05,
+      "loss": 2.6358,
+      "step": 1975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46128201484680176,
+      "learning_rate": 9.941057661262895e-05,
+      "loss": 2.6727,
+      "step": 1980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4847337305545807,
+      "learning_rate": 9.940744342827055e-05,
+      "loss": 2.6188,
+      "step": 1985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4718155562877655,
+      "learning_rate": 9.940430198811202e-05,
+      "loss": 2.6652,
+      "step": 1990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.474626362323761,
+      "learning_rate": 9.940115229267831e-05,
+      "loss": 2.6504,
+      "step": 1995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44636523723602295,
+      "learning_rate": 9.939799434249571e-05,
+      "loss": 2.6456,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48442310094833374,
+      "learning_rate": 9.93948281380919e-05,
+      "loss": 2.6624,
+      "step": 2005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45707738399505615,
+      "learning_rate": 9.939165367999597e-05,
+      "loss": 2.6557,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4694918990135193,
+      "learning_rate": 9.938847096873831e-05,
+      "loss": 2.6816,
+      "step": 2015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4502624273300171,
+      "learning_rate": 9.938528000485078e-05,
+      "loss": 2.6529,
+      "step": 2020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4520168602466583,
+      "learning_rate": 9.938208078886655e-05,
+      "loss": 2.6498,
+      "step": 2025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46525925397872925,
+      "learning_rate": 9.937887332132023e-05,
+      "loss": 2.6794,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4674282371997833,
+      "learning_rate": 9.937565760274776e-05,
+      "loss": 2.5771,
+      "step": 2035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4519534111022949,
+      "learning_rate": 9.937243363368645e-05,
+      "loss": 2.5551,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4409465789794922,
+      "learning_rate": 9.936920141467505e-05,
+      "loss": 2.615,
+      "step": 2045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44015103578567505,
+      "learning_rate": 9.936596094625364e-05,
+      "loss": 2.5819,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44725537300109863,
+      "learning_rate": 9.936271222896367e-05,
+      "loss": 2.6466,
+      "step": 2055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5157557129859924,
+      "learning_rate": 9.935945526334803e-05,
+      "loss": 2.6184,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45462557673454285,
+      "learning_rate": 9.935619004995089e-05,
+      "loss": 2.6506,
+      "step": 2065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4460848271846771,
+      "learning_rate": 9.935291658931791e-05,
+      "loss": 2.7038,
+      "step": 2070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.463811457157135,
+      "learning_rate": 9.934963488199604e-05,
+      "loss": 2.6344,
+      "step": 2075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46156156063079834,
+      "learning_rate": 9.934634492853366e-05,
+      "loss": 2.7354,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4508422911167145,
+      "learning_rate": 9.93430467294805e-05,
+      "loss": 2.6276,
+      "step": 2085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43860721588134766,
+      "learning_rate": 9.933974028538768e-05,
+      "loss": 2.6465,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4292071461677551,
+      "learning_rate": 9.933642559680769e-05,
+      "loss": 2.6169,
+      "step": 2095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47586333751678467,
+      "learning_rate": 9.93331026642944e-05,
+      "loss": 2.5673,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44137105345726013,
+      "learning_rate": 9.932977148840307e-05,
+      "loss": 2.6607,
+      "step": 2105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4636947512626648,
+      "learning_rate": 9.93264320696903e-05,
+      "loss": 2.6544,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4485258460044861,
+      "learning_rate": 9.932308440871413e-05,
+      "loss": 2.6694,
+      "step": 2115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.444821834564209,
+      "learning_rate": 9.931972850603394e-05,
+      "loss": 2.6146,
+      "step": 2120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45808660984039307,
+      "learning_rate": 9.931636436221048e-05,
+      "loss": 2.6775,
+      "step": 2125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4476850926876068,
+      "learning_rate": 9.931299197780586e-05,
+      "loss": 2.637,
+      "step": 2130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48579394817352295,
+      "learning_rate": 9.930961135338361e-05,
+      "loss": 2.6024,
+      "step": 2135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4691111743450165,
+      "learning_rate": 9.930622248950864e-05,
+      "loss": 2.6788,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45097053050994873,
+      "learning_rate": 9.93028253867472e-05,
+      "loss": 2.6411,
+      "step": 2145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45346006751060486,
+      "learning_rate": 9.929942004566695e-05,
+      "loss": 2.643,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44862473011016846,
+      "learning_rate": 9.929600646683686e-05,
+      "loss": 2.6106,
+      "step": 2155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4611954092979431,
+      "learning_rate": 9.929258465082738e-05,
+      "loss": 2.6088,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44688645005226135,
+      "learning_rate": 9.928915459821027e-05,
+      "loss": 2.5968,
+      "step": 2165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4423767030239105,
+      "learning_rate": 9.928571630955865e-05,
+      "loss": 2.5848,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4694935381412506,
+      "learning_rate": 9.928226978544708e-05,
+      "loss": 2.6088,
+      "step": 2175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4587995409965515,
+      "learning_rate": 9.927881502645146e-05,
+      "loss": 2.6382,
+      "step": 2180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45115649700164795,
+      "learning_rate": 9.927535203314904e-05,
+      "loss": 2.5999,
+      "step": 2185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44753918051719666,
+      "learning_rate": 9.927188080611849e-05,
+      "loss": 2.6653,
+      "step": 2190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4874328076839447,
+      "learning_rate": 9.926840134593984e-05,
+      "loss": 2.5932,
+      "step": 2195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46999648213386536,
+      "learning_rate": 9.92649136531945e-05,
+      "loss": 2.4948,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45272403955459595,
+      "learning_rate": 9.926141772846525e-05,
+      "loss": 2.6178,
+      "step": 2205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4452803134918213,
+      "learning_rate": 9.925791357233623e-05,
+      "loss": 2.6415,
+      "step": 2210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4459554851055145,
+      "learning_rate": 9.925440118539298e-05,
+      "loss": 2.637,
+      "step": 2215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45208442211151123,
+      "learning_rate": 9.925088056822241e-05,
+      "loss": 2.5838,
+      "step": 2220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4446738064289093,
+      "learning_rate": 9.924735172141281e-05,
+      "loss": 2.6553,
+      "step": 2225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4695885181427002,
+      "learning_rate": 9.924381464555381e-05,
+      "loss": 2.6842,
+      "step": 2230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43879666924476624,
+      "learning_rate": 9.924026934123647e-05,
+      "loss": 2.6533,
+      "step": 2235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45654571056365967,
+      "learning_rate": 9.92367158090532e-05,
+      "loss": 2.6431,
+      "step": 2240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4459069073200226,
+      "learning_rate": 9.923315404959775e-05,
+      "loss": 2.6049,
+      "step": 2245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4485475718975067,
+      "learning_rate": 9.92295840634653e-05,
+      "loss": 2.6174,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44199803471565247,
+      "learning_rate": 9.922600585125237e-05,
+      "loss": 2.669,
+      "step": 2255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46444666385650635,
+      "learning_rate": 9.922241941355688e-05,
+      "loss": 2.693,
+      "step": 2260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45787563920021057,
+      "learning_rate": 9.921882475097811e-05,
+      "loss": 2.6221,
+      "step": 2265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45151183009147644,
+      "learning_rate": 9.921522186411671e-05,
+      "loss": 2.6358,
+      "step": 2270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44339460134506226,
+      "learning_rate": 9.92116107535747e-05,
+      "loss": 2.6411,
+      "step": 2275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44401928782463074,
+      "learning_rate": 9.920799141995549e-05,
+      "loss": 2.6219,
+      "step": 2280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45282503962516785,
+      "learning_rate": 9.920436386386387e-05,
+      "loss": 2.6531,
+      "step": 2285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4505927860736847,
+      "learning_rate": 9.920072808590596e-05,
+      "loss": 2.6389,
+      "step": 2290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4532349109649658,
+      "learning_rate": 9.91970840866893e-05,
+      "loss": 2.5722,
+      "step": 2295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42472127079963684,
+      "learning_rate": 9.919343186682282e-05,
+      "loss": 2.6057,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4235581159591675,
+      "learning_rate": 9.918977142691675e-05,
+      "loss": 2.5853,
+      "step": 2305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4401482045650482,
+      "learning_rate": 9.918610276758273e-05,
+      "loss": 2.6211,
+      "step": 2310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43106457591056824,
+      "learning_rate": 9.918242588943383e-05,
+      "loss": 2.6216,
+      "step": 2315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4415753483772278,
+      "learning_rate": 9.91787407930844e-05,
+      "loss": 2.5615,
+      "step": 2320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4242173731327057,
+      "learning_rate": 9.917504747915023e-05,
+      "loss": 2.5681,
+      "step": 2325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4465724229812622,
+      "learning_rate": 9.917134594824843e-05,
+      "loss": 2.5681,
+      "step": 2330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4571617841720581,
+      "learning_rate": 9.916763620099754e-05,
+      "loss": 2.5286,
+      "step": 2335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4609034061431885,
+      "learning_rate": 9.916391823801743e-05,
+      "loss": 2.6776,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46449869871139526,
+      "learning_rate": 9.916019205992937e-05,
+      "loss": 2.6434,
+      "step": 2345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4459689259529114,
+      "learning_rate": 9.915645766735597e-05,
+      "loss": 2.6883,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4326188564300537,
+      "learning_rate": 9.915271506092126e-05,
+      "loss": 2.6747,
+      "step": 2355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43627214431762695,
+      "learning_rate": 9.91489642412506e-05,
+      "loss": 2.6762,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44020766019821167,
+      "learning_rate": 9.914520520897072e-05,
+      "loss": 2.6291,
+      "step": 2365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.433481901884079,
+      "learning_rate": 9.914143796470978e-05,
+      "loss": 2.6389,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47297364473342896,
+      "learning_rate": 9.913766250909726e-05,
+      "loss": 2.6942,
+      "step": 2375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4542173445224762,
+      "learning_rate": 9.9133878842764e-05,
+      "loss": 2.5598,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.464533269405365,
+      "learning_rate": 9.913008696634226e-05,
+      "loss": 2.5735,
+      "step": 2385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43234264850616455,
+      "learning_rate": 9.912628688046564e-05,
+      "loss": 2.5362,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4512435495853424,
+      "learning_rate": 9.912247858576914e-05,
+      "loss": 2.6877,
+      "step": 2395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44196251034736633,
+      "learning_rate": 9.911866208288906e-05,
+      "loss": 2.6208,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4485480785369873,
+      "learning_rate": 9.911483737246319e-05,
+      "loss": 2.5892,
+      "step": 2405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44690027832984924,
+      "learning_rate": 9.911100445513058e-05,
+      "loss": 2.6271,
+      "step": 2410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45327118039131165,
+      "learning_rate": 9.910716333153169e-05,
+      "loss": 2.6059,
+      "step": 2415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4754297435283661,
+      "learning_rate": 9.910331400230841e-05,
+      "loss": 2.6068,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46986591815948486,
+      "learning_rate": 9.90994564681039e-05,
+      "loss": 2.5867,
+      "step": 2425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4482470452785492,
+      "learning_rate": 9.909559072956275e-05,
+      "loss": 2.6696,
+      "step": 2430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4418042302131653,
+      "learning_rate": 9.909171678733091e-05,
+      "loss": 2.6438,
+      "step": 2435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42180028557777405,
+      "learning_rate": 9.908783464205573e-05,
+      "loss": 2.656,
+      "step": 2440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4549803137779236,
+      "learning_rate": 9.908394429438585e-05,
+      "loss": 2.5381,
+      "step": 2445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5138429999351501,
+      "learning_rate": 9.908004574497139e-05,
+      "loss": 2.5694,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4442288875579834,
+      "learning_rate": 9.907613899446372e-05,
+      "loss": 2.609,
+      "step": 2455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4641910493373871,
+      "learning_rate": 9.90722240435157e-05,
+      "loss": 2.6068,
+      "step": 2460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4786643385887146,
+      "learning_rate": 9.906830089278148e-05,
+      "loss": 2.5208,
+      "step": 2465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47780269384384155,
+      "learning_rate": 9.906436954291661e-05,
+      "loss": 2.5779,
+      "step": 2470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4562578499317169,
+      "learning_rate": 9.906042999457798e-05,
+      "loss": 2.5898,
+      "step": 2475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4474030137062073,
+      "learning_rate": 9.905648224842391e-05,
+      "loss": 2.6801,
+      "step": 2480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44603195786476135,
+      "learning_rate": 9.905252630511405e-05,
+      "loss": 2.563,
+      "step": 2485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4402001202106476,
+      "learning_rate": 9.904856216530938e-05,
+      "loss": 2.6412,
+      "step": 2490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43054819107055664,
+      "learning_rate": 9.904458982967235e-05,
+      "loss": 2.5002,
+      "step": 2495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44241154193878174,
+      "learning_rate": 9.904060929886668e-05,
+      "loss": 2.5913,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42241841554641724,
+      "learning_rate": 9.903662057355751e-05,
+      "loss": 2.5745,
+      "step": 2505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4497540593147278,
+      "learning_rate": 9.903262365441137e-05,
+      "loss": 2.6443,
+      "step": 2510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45087742805480957,
+      "learning_rate": 9.90286185420961e-05,
+      "loss": 2.5638,
+      "step": 2515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43662863969802856,
+      "learning_rate": 9.902460523728094e-05,
+      "loss": 2.6155,
+      "step": 2520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44917747378349304,
+      "learning_rate": 9.902058374063651e-05,
+      "loss": 2.6738,
+      "step": 2525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4536852538585663,
+      "learning_rate": 9.90165540528348e-05,
+      "loss": 2.6245,
+      "step": 2530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9800692200660706,
+      "learning_rate": 9.901251617454913e-05,
+      "loss": 2.5946,
+      "step": 2535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43859627842903137,
+      "learning_rate": 9.900847010645422e-05,
+      "loss": 2.5426,
+      "step": 2540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4306222200393677,
+      "learning_rate": 9.900441584922616e-05,
+      "loss": 2.5121,
+      "step": 2545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4464203417301178,
+      "learning_rate": 9.90003534035424e-05,
+      "loss": 2.5507,
+      "step": 2550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42890608310699463,
+      "learning_rate": 9.899628277008175e-05,
+      "loss": 2.6102,
+      "step": 2555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43994754552841187,
+      "learning_rate": 9.899220394952441e-05,
+      "loss": 2.6125,
+      "step": 2560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42872101068496704,
+      "learning_rate": 9.898811694255195e-05,
+      "loss": 2.6061,
+      "step": 2565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4538576304912567,
+      "learning_rate": 9.898402174984726e-05,
+      "loss": 2.572,
+      "step": 2570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4641314148902893,
+      "learning_rate": 9.897991837209464e-05,
+      "loss": 2.6129,
+      "step": 2575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4662421941757202,
+      "learning_rate": 9.897580680997977e-05,
+      "loss": 2.6215,
+      "step": 2580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46297523379325867,
+      "learning_rate": 9.897168706418965e-05,
+      "loss": 2.6466,
+      "step": 2585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45740410685539246,
+      "learning_rate": 9.89675591354127e-05,
+      "loss": 2.5388,
+      "step": 2590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42665278911590576,
+      "learning_rate": 9.896342302433866e-05,
+      "loss": 2.6016,
+      "step": 2595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4522157311439514,
+      "learning_rate": 9.895927873165867e-05,
+      "loss": 2.548,
+      "step": 2600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5546777844429016,
+      "learning_rate": 9.895512625806522e-05,
+      "loss": 2.5977,
+      "step": 2605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4505189061164856,
+      "learning_rate": 9.895096560425218e-05,
+      "loss": 2.6907,
+      "step": 2610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4391687512397766,
+      "learning_rate": 9.894679677091477e-05,
+      "loss": 2.6069,
+      "step": 2615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42620375752449036,
+      "learning_rate": 9.89426197587496e-05,
+      "loss": 2.5544,
+      "step": 2620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4493272602558136,
+      "learning_rate": 9.893843456845462e-05,
+      "loss": 2.5362,
+      "step": 2625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4310363829135895,
+      "learning_rate": 9.893424120072916e-05,
+      "loss": 2.5762,
+      "step": 2630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4640781581401825,
+      "learning_rate": 9.893003965627394e-05,
+      "loss": 2.5165,
+      "step": 2635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45891815423965454,
+      "learning_rate": 9.8925829935791e-05,
+      "loss": 2.6072,
+      "step": 2640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.444256454706192,
+      "learning_rate": 9.892161203998375e-05,
+      "loss": 2.6378,
+      "step": 2645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45140671730041504,
+      "learning_rate": 9.891738596955703e-05,
+      "loss": 2.638,
+      "step": 2650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4445360004901886,
+      "learning_rate": 9.891315172521696e-05,
+      "loss": 2.5793,
+      "step": 2655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45720961689949036,
+      "learning_rate": 9.890890930767111e-05,
+      "loss": 2.5904,
+      "step": 2660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40399590134620667,
+      "learning_rate": 9.890465871762832e-05,
+      "loss": 2.5499,
+      "step": 2665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4426371157169342,
+      "learning_rate": 9.89003999557989e-05,
+      "loss": 2.5871,
+      "step": 2670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4417699873447418,
+      "learning_rate": 9.889613302289443e-05,
+      "loss": 2.5698,
+      "step": 2675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43354278802871704,
+      "learning_rate": 9.889185791962792e-05,
+      "loss": 2.599,
+      "step": 2680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42457714676856995,
+      "learning_rate": 9.888757464671372e-05,
+      "loss": 2.6147,
+      "step": 2685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4489341378211975,
+      "learning_rate": 9.888328320486755e-05,
+      "loss": 2.5846,
+      "step": 2690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42269960045814514,
+      "learning_rate": 9.88789835948065e-05,
+      "loss": 2.5899,
+      "step": 2695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42992448806762695,
+      "learning_rate": 9.887467581724901e-05,
+      "loss": 2.6055,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4255000948905945,
+      "learning_rate": 9.887035987291488e-05,
+      "loss": 2.6163,
+      "step": 2705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4316226840019226,
+      "learning_rate": 9.88660357625253e-05,
+      "loss": 2.5898,
+      "step": 2710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4635344445705414,
+      "learning_rate": 9.886170348680284e-05,
+      "loss": 2.5275,
+      "step": 2715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4186500012874603,
+      "learning_rate": 9.885736304647139e-05,
+      "loss": 2.5521,
+      "step": 2720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41054481267929077,
+      "learning_rate": 9.885301444225619e-05,
+      "loss": 2.613,
+      "step": 2725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45451268553733826,
+      "learning_rate": 9.884865767488391e-05,
+      "loss": 2.5652,
+      "step": 2730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4316863715648651,
+      "learning_rate": 9.884429274508256e-05,
+      "loss": 2.6285,
+      "step": 2735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.445164293050766,
+      "learning_rate": 9.883991965358146e-05,
+      "loss": 2.487,
+      "step": 2740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4452510178089142,
+      "learning_rate": 9.883553840111136e-05,
+      "loss": 2.6354,
+      "step": 2745
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40874549746513367,
+      "learning_rate": 9.883114898840437e-05,
+      "loss": 2.651,
+      "step": 2750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42711570858955383,
+      "learning_rate": 9.882675141619391e-05,
+      "loss": 2.5483,
+      "step": 2755
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4276675581932068,
+      "learning_rate": 9.882234568521482e-05,
+      "loss": 2.6237,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.434202641248703,
+      "learning_rate": 9.881793179620328e-05,
+      "loss": 2.6127,
+      "step": 2765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45742687582969666,
+      "learning_rate": 9.881350974989682e-05,
+      "loss": 2.5726,
+      "step": 2770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42267629504203796,
+      "learning_rate": 9.880907954703437e-05,
+      "loss": 2.5951,
+      "step": 2775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42969274520874023,
+      "learning_rate": 9.880464118835619e-05,
+      "loss": 2.5837,
+      "step": 2780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42939257621765137,
+      "learning_rate": 9.880019467460391e-05,
+      "loss": 2.6079,
+      "step": 2785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43971049785614014,
+      "learning_rate": 9.879574000652052e-05,
+      "loss": 2.592,
+      "step": 2790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4274647831916809,
+      "learning_rate": 9.87912771848504e-05,
+      "loss": 2.5851,
+      "step": 2795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47127288579940796,
+      "learning_rate": 9.878680621033925e-05,
+      "loss": 2.5651,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45299553871154785,
+      "learning_rate": 9.878232708373415e-05,
+      "loss": 2.5159,
+      "step": 2805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4647405445575714,
+      "learning_rate": 9.877783980578357e-05,
+      "loss": 2.5693,
+      "step": 2810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44908735156059265,
+      "learning_rate": 9.877334437723728e-05,
+      "loss": 2.6016,
+      "step": 2815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4499860405921936,
+      "learning_rate": 9.87688407988465e-05,
+      "loss": 2.5228,
+      "step": 2820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4227310121059418,
+      "learning_rate": 9.876432907136372e-05,
+      "loss": 2.6048,
+      "step": 2825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43216729164123535,
+      "learning_rate": 9.875980919554287e-05,
+      "loss": 2.6192,
+      "step": 2830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4405534863471985,
+      "learning_rate": 9.875528117213915e-05,
+      "loss": 2.5635,
+      "step": 2835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42040860652923584,
+      "learning_rate": 9.875074500190922e-05,
+      "loss": 2.5739,
+      "step": 2840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42552614212036133,
+      "learning_rate": 9.874620068561104e-05,
+      "loss": 2.5751,
+      "step": 2845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4174400866031647,
+      "learning_rate": 9.874164822400396e-05,
+      "loss": 2.5636,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43836450576782227,
+      "learning_rate": 9.873708761784866e-05,
+      "loss": 2.624,
+      "step": 2855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4259657859802246,
+      "learning_rate": 9.873251886790722e-05,
+      "loss": 2.5635,
+      "step": 2860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4272574782371521,
+      "learning_rate": 9.872794197494307e-05,
+      "loss": 2.6556,
+      "step": 2865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40173599123954773,
+      "learning_rate": 9.872335693972094e-05,
+      "loss": 2.5263,
+      "step": 2870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46419671177864075,
+      "learning_rate": 9.871876376300702e-05,
+      "loss": 2.5036,
+      "step": 2875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42429250478744507,
+      "learning_rate": 9.871416244556881e-05,
+      "loss": 2.5468,
+      "step": 2880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4398246109485626,
+      "learning_rate": 9.870955298817517e-05,
+      "loss": 2.5657,
+      "step": 2885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42618173360824585,
+      "learning_rate": 9.87049353915963e-05,
+      "loss": 2.5785,
+      "step": 2890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44166162610054016,
+      "learning_rate": 9.87003096566038e-05,
+      "loss": 2.5598,
+      "step": 2895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4448752701282501,
+      "learning_rate": 9.869567578397062e-05,
+      "loss": 2.6235,
+      "step": 2900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4359762966632843,
+      "learning_rate": 9.869103377447105e-05,
+      "loss": 2.5113,
+      "step": 2905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43620043992996216,
+      "learning_rate": 9.868638362888076e-05,
+      "loss": 2.6072,
+      "step": 2910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43776997923851013,
+      "learning_rate": 9.868172534797678e-05,
+      "loss": 2.6018,
+      "step": 2915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4316084384918213,
+      "learning_rate": 9.867705893253747e-05,
+      "loss": 2.5778,
+      "step": 2920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45051875710487366,
+      "learning_rate": 9.867238438334259e-05,
+      "loss": 2.5123,
+      "step": 2925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42391207814216614,
+      "learning_rate": 9.866770170117324e-05,
+      "loss": 2.5328,
+      "step": 2930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44185560941696167,
+      "learning_rate": 9.866301088681185e-05,
+      "loss": 2.6471,
+      "step": 2935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4389095902442932,
+      "learning_rate": 9.865831194104227e-05,
+      "loss": 2.4701,
+      "step": 2940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49202775955200195,
+      "learning_rate": 9.865360486464967e-05,
+      "loss": 2.4295,
+      "step": 2945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4340350329875946,
+      "learning_rate": 9.864888965842056e-05,
+      "loss": 2.5621,
+      "step": 2950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44590312242507935,
+      "learning_rate": 9.864416632314287e-05,
+      "loss": 2.5293,
+      "step": 2955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42789819836616516,
+      "learning_rate": 9.863943485960583e-05,
+      "loss": 2.4969,
+      "step": 2960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40523675084114075,
+      "learning_rate": 9.863469526860005e-05,
+      "loss": 2.5759,
+      "step": 2965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.427165687084198,
+      "learning_rate": 9.862994755091752e-05,
+      "loss": 2.5282,
+      "step": 2970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4273890256881714,
+      "learning_rate": 9.862519170735154e-05,
+      "loss": 2.5377,
+      "step": 2975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4300043284893036,
+      "learning_rate": 9.86204277386968e-05,
+      "loss": 2.5037,
+      "step": 2980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43027305603027344,
+      "learning_rate": 9.861565564574935e-05,
+      "loss": 2.599,
+      "step": 2985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4259815216064453,
+      "learning_rate": 9.861087542930658e-05,
+      "loss": 2.5155,
+      "step": 2990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4154326915740967,
+      "learning_rate": 9.860608709016726e-05,
+      "loss": 2.5004,
+      "step": 2995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42423099279403687,
+      "learning_rate": 9.860129062913147e-05,
+      "loss": 2.5769,
+      "step": 3000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4301806390285492,
+      "learning_rate": 9.859648604700072e-05,
+      "loss": 2.6132,
+      "step": 3005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44970008730888367,
+      "learning_rate": 9.859167334457781e-05,
+      "loss": 2.6191,
+      "step": 3010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.447582870721817,
+      "learning_rate": 9.858685252266696e-05,
+      "loss": 2.5098,
+      "step": 3015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42455363273620605,
+      "learning_rate": 9.858202358207367e-05,
+      "loss": 2.542,
+      "step": 3020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4264921247959137,
+      "learning_rate": 9.857718652360486e-05,
+      "loss": 2.5063,
+      "step": 3025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43237215280532837,
+      "learning_rate": 9.857234134806878e-05,
+      "loss": 2.5246,
+      "step": 3030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43131309747695923,
+      "learning_rate": 9.856748805627503e-05,
+      "loss": 2.5649,
+      "step": 3035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4574224352836609,
+      "learning_rate": 9.856262664903462e-05,
+      "loss": 2.5547,
+      "step": 3040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.431661456823349,
+      "learning_rate": 9.855775712715981e-05,
+      "loss": 2.552,
+      "step": 3045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4171062111854553,
+      "learning_rate": 9.855287949146432e-05,
+      "loss": 2.5005,
+      "step": 3050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44946810603141785,
+      "learning_rate": 9.854799374276318e-05,
+      "loss": 2.4453,
+      "step": 3055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4324861764907837,
+      "learning_rate": 9.854309988187277e-05,
+      "loss": 2.5579,
+      "step": 3060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44608044624328613,
+      "learning_rate": 9.853819790961086e-05,
+      "loss": 2.5451,
+      "step": 3065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41277486085891724,
+      "learning_rate": 9.853328782679652e-05,
+      "loss": 2.4952,
+      "step": 3070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4486488699913025,
+      "learning_rate": 9.852836963425021e-05,
+      "loss": 2.5336,
+      "step": 3075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43473899364471436,
+      "learning_rate": 9.852344333279375e-05,
+      "loss": 2.5323,
+      "step": 3080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.410753458738327,
+      "learning_rate": 9.851850892325032e-05,
+      "loss": 2.5515,
+      "step": 3085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42531758546829224,
+      "learning_rate": 9.851356640644443e-05,
+      "loss": 2.5166,
+      "step": 3090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42476287484169006,
+      "learning_rate": 9.850861578320196e-05,
+      "loss": 2.5988,
+      "step": 3095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4397708773612976,
+      "learning_rate": 9.850365705435013e-05,
+      "loss": 2.5748,
+      "step": 3100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44073066115379333,
+      "learning_rate": 9.849869022071753e-05,
+      "loss": 2.5481,
+      "step": 3105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4238365590572357,
+      "learning_rate": 9.849371528313411e-05,
+      "loss": 2.574,
+      "step": 3110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4107390344142914,
+      "learning_rate": 9.848873224243114e-05,
+      "loss": 2.5393,
+      "step": 3115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43788957595825195,
+      "learning_rate": 9.848374109944128e-05,
+      "loss": 2.5168,
+      "step": 3120
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43675801157951355,
+      "learning_rate": 9.847874185499855e-05,
+      "loss": 2.5778,
+      "step": 3125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4615284502506256,
+      "learning_rate": 9.847373450993828e-05,
+      "loss": 2.5608,
+      "step": 3130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46616578102111816,
+      "learning_rate": 9.846871906509719e-05,
+      "loss": 2.5761,
+      "step": 3135
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44282686710357666,
+      "learning_rate": 9.846369552131334e-05,
+      "loss": 2.6427,
+      "step": 3140
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41826507449150085,
+      "learning_rate": 9.845866387942614e-05,
+      "loss": 2.6266,
+      "step": 3145
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4180801808834076,
+      "learning_rate": 9.845362414027637e-05,
+      "loss": 2.4762,
+      "step": 3150
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41745540499687195,
+      "learning_rate": 9.844857630470614e-05,
+      "loss": 2.5788,
+      "step": 3155
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44452258944511414,
+      "learning_rate": 9.844352037355893e-05,
+      "loss": 2.496,
+      "step": 3160
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.430630087852478,
+      "learning_rate": 9.843845634767957e-05,
+      "loss": 2.4841,
+      "step": 3165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42415493726730347,
+      "learning_rate": 9.843338422791424e-05,
+      "loss": 2.5263,
+      "step": 3170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44700491428375244,
+      "learning_rate": 9.842830401511048e-05,
+      "loss": 2.5042,
+      "step": 3175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4231417179107666,
+      "learning_rate": 9.842321571011714e-05,
+      "loss": 2.4946,
+      "step": 3180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42834043502807617,
+      "learning_rate": 9.841811931378448e-05,
+      "loss": 2.5392,
+      "step": 3185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4569108486175537,
+      "learning_rate": 9.841301482696413e-05,
+      "loss": 2.4699,
+      "step": 3190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4273216724395752,
+      "learning_rate": 9.840790225050896e-05,
+      "loss": 2.534,
+      "step": 3195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4343474507331848,
+      "learning_rate": 9.840278158527328e-05,
+      "loss": 2.5098,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41953644156455994,
+      "learning_rate": 9.839765283211278e-05,
+      "loss": 2.5145,
+      "step": 3205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44883257150650024,
+      "learning_rate": 9.83925159918844e-05,
+      "loss": 2.5063,
+      "step": 3210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45871350169181824,
+      "learning_rate": 9.838737106544651e-05,
+      "loss": 2.5102,
+      "step": 3215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4468609690666199,
+      "learning_rate": 9.838221805365881e-05,
+      "loss": 2.4405,
+      "step": 3220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42680245637893677,
+      "learning_rate": 9.837705695738236e-05,
+      "loss": 2.4868,
+      "step": 3225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.420076847076416,
+      "learning_rate": 9.837188777747954e-05,
+      "loss": 2.5414,
+      "step": 3230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44864481687545776,
+      "learning_rate": 9.836671051481411e-05,
+      "loss": 2.4801,
+      "step": 3235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41252875328063965,
+      "learning_rate": 9.836152517025116e-05,
+      "loss": 2.4651,
+      "step": 3240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4185810387134552,
+      "learning_rate": 9.835633174465715e-05,
+      "loss": 2.4788,
+      "step": 3245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4506584405899048,
+      "learning_rate": 9.83511302388999e-05,
+      "loss": 2.4897,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43840909004211426,
+      "learning_rate": 9.834592065384853e-05,
+      "loss": 2.5417,
+      "step": 3255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4125819206237793,
+      "learning_rate": 9.834070299037357e-05,
+      "loss": 2.5157,
+      "step": 3260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9324262142181396,
+      "learning_rate": 9.833547724934687e-05,
+      "loss": 2.4584,
+      "step": 3265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41954514384269714,
+      "learning_rate": 9.833024343164159e-05,
+      "loss": 2.5581,
+      "step": 3270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44926130771636963,
+      "learning_rate": 9.832500153813234e-05,
+      "loss": 2.5359,
+      "step": 3275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4236092269420624,
+      "learning_rate": 9.831975156969498e-05,
+      "loss": 2.5315,
+      "step": 3280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4339849650859833,
+      "learning_rate": 9.831449352720677e-05,
+      "loss": 2.5645,
+      "step": 3285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40436044335365295,
+      "learning_rate": 9.830922741154631e-05,
+      "loss": 2.6055,
+      "step": 3290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41621994972229004,
+      "learning_rate": 9.830395322359357e-05,
+      "loss": 2.5243,
+      "step": 3295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41897112131118774,
+      "learning_rate": 9.829867096422983e-05,
+      "loss": 2.5246,
+      "step": 3300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41987916827201843,
+      "learning_rate": 9.829338063433773e-05,
+      "loss": 2.5516,
+      "step": 3305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4368478059768677,
+      "learning_rate": 9.828808223480126e-05,
+      "loss": 2.5637,
+      "step": 3310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.413682222366333,
+      "learning_rate": 9.828277576650578e-05,
+      "loss": 2.5733,
+      "step": 3315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4406430721282959,
+      "learning_rate": 9.827746123033796e-05,
+      "loss": 2.5109,
+      "step": 3320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42099863290786743,
+      "learning_rate": 9.827213862718587e-05,
+      "loss": 2.4509,
+      "step": 3325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44899240136146545,
+      "learning_rate": 9.826680795793887e-05,
+      "loss": 2.574,
+      "step": 3330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4200420081615448,
+      "learning_rate": 9.826146922348773e-05,
+      "loss": 2.5599,
+      "step": 3335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44432705640792847,
+      "learning_rate": 9.825612242472448e-05,
+      "loss": 2.4997,
+      "step": 3340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4281298518180847,
+      "learning_rate": 9.82507675625426e-05,
+      "loss": 2.5209,
+      "step": 3345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4019240140914917,
+      "learning_rate": 9.824540463783682e-05,
+      "loss": 2.5855,
+      "step": 3350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43513137102127075,
+      "learning_rate": 9.824003365150329e-05,
+      "loss": 2.5552,
+      "step": 3355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41877481341362,
+      "learning_rate": 9.82346546044395e-05,
+      "loss": 2.3943,
+      "step": 3360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41558438539505005,
+      "learning_rate": 9.822926749754424e-05,
+      "loss": 2.5527,
+      "step": 3365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4194242060184479,
+      "learning_rate": 9.822387233171771e-05,
+      "loss": 2.5297,
+      "step": 3370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4012485146522522,
+      "learning_rate": 9.821846910786136e-05,
+      "loss": 2.5276,
+      "step": 3375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41538599133491516,
+      "learning_rate": 9.821305782687812e-05,
+      "loss": 2.4905,
+      "step": 3380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4214002192020416,
+      "learning_rate": 9.820763848967215e-05,
+      "loss": 2.4216,
+      "step": 3385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.438284695148468,
+      "learning_rate": 9.820221109714902e-05,
+      "loss": 2.4626,
+      "step": 3390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42304956912994385,
+      "learning_rate": 9.81967756502156e-05,
+      "loss": 2.4886,
+      "step": 3395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42296892404556274,
+      "learning_rate": 9.819133214978017e-05,
+      "loss": 2.5711,
+      "step": 3400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42140474915504456,
+      "learning_rate": 9.818588059675229e-05,
+      "loss": 2.5106,
+      "step": 3405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42303505539894104,
+      "learning_rate": 9.818042099204292e-05,
+      "loss": 2.5621,
+      "step": 3410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3970809280872345,
+      "learning_rate": 9.817495333656433e-05,
+      "loss": 2.5711,
+      "step": 3415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42582881450653076,
+      "learning_rate": 9.816947763123013e-05,
+      "loss": 2.5472,
+      "step": 3420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41324761509895325,
+      "learning_rate": 9.816399387695529e-05,
+      "loss": 2.5344,
+      "step": 3425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41110068559646606,
+      "learning_rate": 9.815850207465616e-05,
+      "loss": 2.5079,
+      "step": 3430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43681082129478455,
+      "learning_rate": 9.815300222525035e-05,
+      "loss": 2.5419,
+      "step": 3435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39850080013275146,
+      "learning_rate": 9.814749432965691e-05,
+      "loss": 2.5236,
+      "step": 3440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4373910427093506,
+      "learning_rate": 9.814197838879615e-05,
+      "loss": 2.566,
+      "step": 3445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42678138613700867,
+      "learning_rate": 9.813645440358981e-05,
+      "loss": 2.493,
+      "step": 3450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43437913060188293,
+      "learning_rate": 9.813092237496087e-05,
+      "loss": 2.5331,
+      "step": 3455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40660277009010315,
+      "learning_rate": 9.812538230383376e-05,
+      "loss": 2.4906,
+      "step": 3460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41363874077796936,
+      "learning_rate": 9.811983419113418e-05,
+      "loss": 2.5329,
+      "step": 3465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.434317946434021,
+      "learning_rate": 9.81142780377892e-05,
+      "loss": 2.5264,
+      "step": 3470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43473848700523376,
+      "learning_rate": 9.810871384472724e-05,
+      "loss": 2.5079,
+      "step": 3475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4020531475543976,
+      "learning_rate": 9.810314161287806e-05,
+      "loss": 2.6195,
+      "step": 3480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4274156391620636,
+      "learning_rate": 9.809756134317276e-05,
+      "loss": 2.5581,
+      "step": 3485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44298937916755676,
+      "learning_rate": 9.809197303654376e-05,
+      "loss": 2.4829,
+      "step": 3490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42849424481391907,
+      "learning_rate": 9.808637669392487e-05,
+      "loss": 2.5925,
+      "step": 3495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4287078082561493,
+      "learning_rate": 9.808077231625119e-05,
+      "loss": 2.5345,
+      "step": 3500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4214578866958618,
+      "learning_rate": 9.807515990445922e-05,
+      "loss": 2.4857,
+      "step": 3505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4156367778778076,
+      "learning_rate": 9.806953945948676e-05,
+      "loss": 2.6164,
+      "step": 3510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43715494871139526,
+      "learning_rate": 9.806391098227297e-05,
+      "loss": 2.5622,
+      "step": 3515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43101805448532104,
+      "learning_rate": 9.805827447375835e-05,
+      "loss": 2.542,
+      "step": 3520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.398406058549881,
+      "learning_rate": 9.805262993488473e-05,
+      "loss": 2.5503,
+      "step": 3525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41427746415138245,
+      "learning_rate": 9.804697736659532e-05,
+      "loss": 2.6348,
+      "step": 3530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4119768738746643,
+      "learning_rate": 9.804131676983458e-05,
+      "loss": 2.488,
+      "step": 3535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41676729917526245,
+      "learning_rate": 9.803564814554844e-05,
+      "loss": 2.5037,
+      "step": 3540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4316596984863281,
+      "learning_rate": 9.802997149468408e-05,
+      "loss": 2.4355,
+      "step": 3545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4278164207935333,
+      "learning_rate": 9.802428681819003e-05,
+      "loss": 2.4749,
+      "step": 3550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4602886438369751,
+      "learning_rate": 9.801859411701621e-05,
+      "loss": 2.5181,
+      "step": 3555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42494112253189087,
+      "learning_rate": 9.801289339211384e-05,
+      "loss": 2.5328,
+      "step": 3560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4145127236843109,
+      "learning_rate": 9.800718464443547e-05,
+      "loss": 2.501,
+      "step": 3565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42412570118904114,
+      "learning_rate": 9.800146787493503e-05,
+      "loss": 2.5671,
+      "step": 3570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40215352177619934,
+      "learning_rate": 9.799574308456775e-05,
+      "loss": 2.4407,
+      "step": 3575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4359852969646454,
+      "learning_rate": 9.799001027429026e-05,
+      "loss": 2.3716,
+      "step": 3580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43408599495887756,
+      "learning_rate": 9.798426944506048e-05,
+      "loss": 2.5697,
+      "step": 3585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42817550897598267,
+      "learning_rate": 9.797852059783766e-05,
+      "loss": 2.6235,
+      "step": 3590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4207446575164795,
+      "learning_rate": 9.79727637335824e-05,
+      "loss": 2.52,
+      "step": 3595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5072458386421204,
+      "learning_rate": 9.79669988532567e-05,
+      "loss": 2.479,
+      "step": 3600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4309207797050476,
+      "learning_rate": 9.79612259578238e-05,
+      "loss": 2.4712,
+      "step": 3605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4545679986476898,
+      "learning_rate": 9.795544504824836e-05,
+      "loss": 2.4557,
+      "step": 3610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3990359902381897,
+      "learning_rate": 9.794965612549635e-05,
+      "loss": 2.4645,
+      "step": 3615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42837396264076233,
+      "learning_rate": 9.794385919053506e-05,
+      "loss": 2.5074,
+      "step": 3620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43564605712890625,
+      "learning_rate": 9.793805424433313e-05,
+      "loss": 2.4692,
+      "step": 3625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44082972407341003,
+      "learning_rate": 9.793224128786058e-05,
+      "loss": 2.4441,
+      "step": 3630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4154863655567169,
+      "learning_rate": 9.79264203220887e-05,
+      "loss": 2.5793,
+      "step": 3635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43978703022003174,
+      "learning_rate": 9.792059134799016e-05,
+      "loss": 2.528,
+      "step": 3640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40013542771339417,
+      "learning_rate": 9.791475436653896e-05,
+      "loss": 2.5361,
+      "step": 3645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4278464913368225,
+      "learning_rate": 9.790890937871046e-05,
+      "loss": 2.4812,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4060390293598175,
+      "learning_rate": 9.79030563854813e-05,
+      "loss": 2.5556,
+      "step": 3655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3993265628814697,
+      "learning_rate": 9.789719538782952e-05,
+      "loss": 2.4935,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40739354491233826,
+      "learning_rate": 9.789132638673445e-05,
+      "loss": 2.5943,
+      "step": 3665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42229223251342773,
+      "learning_rate": 9.78854493831768e-05,
+      "loss": 2.5869,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42648905515670776,
+      "learning_rate": 9.787956437813859e-05,
+      "loss": 2.5571,
+      "step": 3675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42417335510253906,
+      "learning_rate": 9.787367137260317e-05,
+      "loss": 2.5379,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43959033489227295,
+      "learning_rate": 9.786777036755525e-05,
+      "loss": 2.5532,
+      "step": 3685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4540771245956421,
+      "learning_rate": 9.786186136398085e-05,
+      "loss": 2.5188,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3966066539287567,
+      "learning_rate": 9.785594436286738e-05,
+      "loss": 2.5652,
+      "step": 3695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4079882502555847,
+      "learning_rate": 9.78500193652035e-05,
+      "loss": 2.5161,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4097125828266144,
+      "learning_rate": 9.784408637197929e-05,
+      "loss": 2.4174,
+      "step": 3705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43045660853385925,
+      "learning_rate": 9.783814538418612e-05,
+      "loss": 2.5513,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4513573944568634,
+      "learning_rate": 9.783219640281671e-05,
+      "loss": 2.5754,
+      "step": 3715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41416653990745544,
+      "learning_rate": 9.78262394288651e-05,
+      "loss": 2.5326,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43572360277175903,
+      "learning_rate": 9.782027446332671e-05,
+      "loss": 2.5481,
+      "step": 3725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45901769399642944,
+      "learning_rate": 9.781430150719822e-05,
+      "loss": 2.5228,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4297926425933838,
+      "learning_rate": 9.780832056147772e-05,
+      "loss": 2.6206,
+      "step": 3735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41366881132125854,
+      "learning_rate": 9.78023316271646e-05,
+      "loss": 2.5885,
+      "step": 3740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41044357419013977,
+      "learning_rate": 9.779633470525957e-05,
+      "loss": 2.5025,
+      "step": 3745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42426109313964844,
+      "learning_rate": 9.77903297967647e-05,
+      "loss": 2.5398,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41779428720474243,
+      "learning_rate": 9.778431690268341e-05,
+      "loss": 2.5833,
+      "step": 3755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43967461585998535,
+      "learning_rate": 9.777829602402043e-05,
+      "loss": 2.5285,
+      "step": 3760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42884010076522827,
+      "learning_rate": 9.777226716178178e-05,
+      "loss": 2.5437,
+      "step": 3765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42278072237968445,
+      "learning_rate": 9.776623031697492e-05,
+      "loss": 2.463,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4245304465293884,
+      "learning_rate": 9.776018549060855e-05,
+      "loss": 2.5149,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4190027415752411,
+      "learning_rate": 9.775413268369274e-05,
+      "loss": 2.4762,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4035365879535675,
+      "learning_rate": 9.77480718972389e-05,
+      "loss": 2.4996,
+      "step": 3785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4182158410549164,
+      "learning_rate": 9.774200313225976e-05,
+      "loss": 2.5208,
+      "step": 3790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4341569244861603,
+      "learning_rate": 9.773592638976939e-05,
+      "loss": 2.518,
+      "step": 3795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9493876695632935,
+      "learning_rate": 9.77298416707832e-05,
+      "loss": 2.5636,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4347634017467499,
+      "learning_rate": 9.772374897631791e-05,
+      "loss": 2.5022,
+      "step": 3805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3983246982097626,
+      "learning_rate": 9.771764830739159e-05,
+      "loss": 2.4537,
+      "step": 3810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.510066568851471,
+      "learning_rate": 9.771153966502365e-05,
+      "loss": 2.4749,
+      "step": 3815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.419828861951828,
+      "learning_rate": 9.77054230502348e-05,
+      "loss": 2.4922,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41545891761779785,
+      "learning_rate": 9.769929846404711e-05,
+      "loss": 2.5053,
+      "step": 3825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40990546345710754,
+      "learning_rate": 9.769316590748401e-05,
+      "loss": 2.4729,
+      "step": 3830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3951980173587799,
+      "learning_rate": 9.768702538157018e-05,
+      "loss": 2.4706,
+      "step": 3835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41250839829444885,
+      "learning_rate": 9.76808768873317e-05,
+      "loss": 2.553,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40834102034568787,
+      "learning_rate": 9.767472042579595e-05,
+      "loss": 2.475,
+      "step": 3845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40338069200515747,
+      "learning_rate": 9.766855599799168e-05,
+      "loss": 2.446,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43315252661705017,
+      "learning_rate": 9.76623836049489e-05,
+      "loss": 2.4572,
+      "step": 3855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41872233152389526,
+      "learning_rate": 9.765620324769903e-05,
+      "loss": 2.548,
+      "step": 3860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43121737241744995,
+      "learning_rate": 9.765001492727476e-05,
+      "loss": 2.4989,
+      "step": 3865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40181416273117065,
+      "learning_rate": 9.764381864471017e-05,
+      "loss": 2.498,
+      "step": 3870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4011335074901581,
+      "learning_rate": 9.76376144010406e-05,
+      "loss": 2.5563,
+      "step": 3875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4199112057685852,
+      "learning_rate": 9.763140219730278e-05,
+      "loss": 2.4891,
+      "step": 3880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4199535548686981,
+      "learning_rate": 9.762518203453472e-05,
+      "loss": 2.4249,
+      "step": 3885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39125925302505493,
+      "learning_rate": 9.761895391377581e-05,
+      "loss": 2.4555,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41029882431030273,
+      "learning_rate": 9.761271783606674e-05,
+      "loss": 2.5041,
+      "step": 3895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42846694588661194,
+      "learning_rate": 9.760647380244952e-05,
+      "loss": 2.5406,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40966662764549255,
+      "learning_rate": 9.760022181396753e-05,
+      "loss": 2.586,
+      "step": 3905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40147584676742554,
+      "learning_rate": 9.759396187166543e-05,
+      "loss": 2.6054,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43804046511650085,
+      "learning_rate": 9.758769397658926e-05,
+      "loss": 2.4896,
+      "step": 3915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40173789858818054,
+      "learning_rate": 9.758141812978634e-05,
+      "loss": 2.4098,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44153958559036255,
+      "learning_rate": 9.757513433230534e-05,
+      "loss": 2.4614,
+      "step": 3925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4390711784362793,
+      "learning_rate": 9.756884258519627e-05,
+      "loss": 2.4976,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3993587791919708,
+      "learning_rate": 9.756254288951046e-05,
+      "loss": 2.4904,
+      "step": 3935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4028399586677551,
+      "learning_rate": 9.755623524630055e-05,
+      "loss": 2.4293,
+      "step": 3940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40059199929237366,
+      "learning_rate": 9.754991965662055e-05,
+      "loss": 2.5171,
+      "step": 3945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7557455897331238,
+      "learning_rate": 9.754359612152574e-05,
+      "loss": 2.5198,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40708720684051514,
+      "learning_rate": 9.753726464207278e-05,
+      "loss": 2.4357,
+      "step": 3955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42117127776145935,
+      "learning_rate": 9.753092521931963e-05,
+      "loss": 2.4911,
+      "step": 3960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40833884477615356,
+      "learning_rate": 9.752457785432558e-05,
+      "loss": 2.4357,
+      "step": 3965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4177716374397278,
+      "learning_rate": 9.751822254815126e-05,
+      "loss": 2.4625,
+      "step": 3970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4074014127254486,
+      "learning_rate": 9.751185930185861e-05,
+      "loss": 2.461,
+      "step": 3975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42372092604637146,
+      "learning_rate": 9.750548811651091e-05,
+      "loss": 2.5232,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.410813570022583,
+      "learning_rate": 9.749910899317276e-05,
+      "loss": 2.495,
+      "step": 3985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40567073225975037,
+      "learning_rate": 9.74927219329101e-05,
+      "loss": 2.4573,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41241776943206787,
+      "learning_rate": 9.748632693679016e-05,
+      "loss": 2.466,
+      "step": 3995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.39371126890182495,
+      "learning_rate": 9.747992400588153e-05,
+      "loss": 2.4949,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4394828677177429,
+      "learning_rate": 9.747351314125413e-05,
+      "loss": 2.468,
+      "step": 4005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4056907892227173,
+      "learning_rate": 9.746709434397916e-05,
+      "loss": 2.5545,
+      "step": 4010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41539931297302246,
+      "learning_rate": 9.746066761512921e-05,
+      "loss": 2.5471,
+      "step": 4015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42427510023117065,
+      "learning_rate": 9.745423295577815e-05,
+      "loss": 2.5144,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4156424403190613,
+      "learning_rate": 9.744779036700118e-05,
+      "loss": 2.5966,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43414202332496643,
+      "learning_rate": 9.744133984987485e-05,
+      "loss": 2.5132,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44093331694602966,
+      "learning_rate": 9.743488140547701e-05,
+      "loss": 2.5677,
+      "step": 4035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4202071726322174,
+      "learning_rate": 9.742841503488682e-05,
+      "loss": 2.4504,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4083489179611206,
+      "learning_rate": 9.742194073918482e-05,
+      "loss": 2.4738,
+      "step": 4045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4300188720226288,
+      "learning_rate": 9.741545851945282e-05,
+      "loss": 2.4479,
+      "step": 4050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4043237566947937,
+      "learning_rate": 9.740896837677398e-05,
+      "loss": 2.5155,
+      "step": 4055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40921083092689514,
+      "learning_rate": 9.740247031223279e-05,
+      "loss": 2.5693,
+      "step": 4060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41888317465782166,
+      "learning_rate": 9.739596432691502e-05,
+      "loss": 2.4741,
+      "step": 4065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41925787925720215,
+      "learning_rate": 9.738945042190785e-05,
+      "loss": 2.4781,
+      "step": 4070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4528825581073761,
+      "learning_rate": 9.738292859829968e-05,
+      "loss": 2.5273,
+      "step": 4075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4149487316608429,
+      "learning_rate": 9.73763988571803e-05,
+      "loss": 2.4776,
+      "step": 4080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4186002016067505,
+      "learning_rate": 9.73698611996408e-05,
+      "loss": 2.5515,
+      "step": 4085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42035502195358276,
+      "learning_rate": 9.736331562677361e-05,
+      "loss": 2.5778,
+      "step": 4090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40942299365997314,
+      "learning_rate": 9.735676213967247e-05,
+      "loss": 2.4888,
+      "step": 4095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41778773069381714,
+      "learning_rate": 9.735020073943244e-05,
+      "loss": 2.5651,
+      "step": 4100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40230756998062134,
+      "learning_rate": 9.734363142714988e-05,
+      "loss": 2.4308,
+      "step": 4105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40558311343193054,
+      "learning_rate": 9.733705420392254e-05,
+      "loss": 2.5208,
+      "step": 4110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3955559730529785,
+      "learning_rate": 9.733046907084944e-05,
+      "loss": 2.5429,
+      "step": 4115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39566823840141296,
+      "learning_rate": 9.732387602903091e-05,
+      "loss": 2.45,
+      "step": 4120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39663997292518616,
+      "learning_rate": 9.731727507956862e-05,
+      "loss": 2.403,
+      "step": 4125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4241093099117279,
+      "learning_rate": 9.73106662235656e-05,
+      "loss": 2.5536,
+      "step": 4130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41281378269195557,
+      "learning_rate": 9.730404946212614e-05,
+      "loss": 2.4726,
+      "step": 4135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39781469106674194,
+      "learning_rate": 9.729742479635589e-05,
+      "loss": 2.5006,
+      "step": 4140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.437139630317688,
+      "learning_rate": 9.729079222736177e-05,
+      "loss": 2.5021,
+      "step": 4145
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4049411714076996,
+      "learning_rate": 9.728415175625211e-05,
+      "loss": 2.5342,
+      "step": 4150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3936813175678253,
+      "learning_rate": 9.727750338413647e-05,
+      "loss": 2.5415,
+      "step": 4155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38825124502182007,
+      "learning_rate": 9.727084711212579e-05,
+      "loss": 2.4308,
+      "step": 4160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4265642464160919,
+      "learning_rate": 9.726418294133229e-05,
+      "loss": 2.4843,
+      "step": 4165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4016586244106293,
+      "learning_rate": 9.725751087286954e-05,
+      "loss": 2.4498,
+      "step": 4170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40944746136665344,
+      "learning_rate": 9.725083090785242e-05,
+      "loss": 2.4426,
+      "step": 4175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40711575746536255,
+      "learning_rate": 9.724414304739712e-05,
+      "loss": 2.5904,
+      "step": 4180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3982873857021332,
+      "learning_rate": 9.723744729262116e-05,
+      "loss": 2.3971,
+      "step": 4185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4062134027481079,
+      "learning_rate": 9.723074364464337e-05,
+      "loss": 2.4902,
+      "step": 4190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4072529077529907,
+      "learning_rate": 9.722403210458393e-05,
+      "loss": 2.5323,
+      "step": 4195
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3999941945075989,
+      "learning_rate": 9.721731267356426e-05,
+      "loss": 2.531,
+      "step": 4200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4010932147502899,
+      "learning_rate": 9.721058535270722e-05,
+      "loss": 2.5003,
+      "step": 4205
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4328639507293701,
+      "learning_rate": 9.720385014313686e-05,
+      "loss": 2.4673,
+      "step": 4210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4059986472129822,
+      "learning_rate": 9.719710704597865e-05,
+      "loss": 2.4759,
+      "step": 4215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39341679215431213,
+      "learning_rate": 9.719035606235931e-05,
+      "loss": 2.5241,
+      "step": 4220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4048922657966614,
+      "learning_rate": 9.718359719340693e-05,
+      "loss": 2.419,
+      "step": 4225
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4021875262260437,
+      "learning_rate": 9.717683044025086e-05,
+      "loss": 2.5174,
+      "step": 4230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3886715769767761,
+      "learning_rate": 9.717005580402184e-05,
+      "loss": 2.4935,
+      "step": 4235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4156700670719147,
+      "learning_rate": 9.716327328585183e-05,
+      "loss": 2.5138,
+      "step": 4240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4113134741783142,
+      "learning_rate": 9.715648288687421e-05,
+      "loss": 2.6068,
+      "step": 4245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.404624879360199,
+      "learning_rate": 9.714968460822363e-05,
+      "loss": 2.4663,
+      "step": 4250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.416970431804657,
+      "learning_rate": 9.714287845103602e-05,
+      "loss": 2.4827,
+      "step": 4255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4054573178291321,
+      "learning_rate": 9.71360644164487e-05,
+      "loss": 2.5973,
+      "step": 4260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4178522825241089,
+      "learning_rate": 9.712924250560025e-05,
+      "loss": 2.4807,
+      "step": 4265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41729670763015747,
+      "learning_rate": 9.712241271963059e-05,
+      "loss": 2.5212,
+      "step": 4270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41174185276031494,
+      "learning_rate": 9.711557505968097e-05,
+      "loss": 2.3974,
+      "step": 4275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.381398469209671,
+      "learning_rate": 9.710872952689391e-05,
+      "loss": 2.5191,
+      "step": 4280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41683611273765564,
+      "learning_rate": 9.71018761224133e-05,
+      "loss": 2.5079,
+      "step": 4285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40159493684768677,
+      "learning_rate": 9.709501484738429e-05,
+      "loss": 2.4326,
+      "step": 4290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39853864908218384,
+      "learning_rate": 9.70881457029534e-05,
+      "loss": 2.4732,
+      "step": 4295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4155634939670563,
+      "learning_rate": 9.708126869026841e-05,
+      "loss": 2.5241,
+      "step": 4300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.38230100274086,
+      "learning_rate": 9.707438381047846e-05,
+      "loss": 2.3583,
+      "step": 4305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41651979088783264,
+      "learning_rate": 9.7067491064734e-05,
+      "loss": 2.502,
+      "step": 4310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41017279028892517,
+      "learning_rate": 9.706059045418676e-05,
+      "loss": 2.5243,
+      "step": 4315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.403313547372818,
+      "learning_rate": 9.705368197998982e-05,
+      "loss": 2.5066,
+      "step": 4320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.436592161655426,
+      "learning_rate": 9.704676564329756e-05,
+      "loss": 2.5078,
+      "step": 4325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.410942405462265,
+      "learning_rate": 9.703984144526565e-05,
+      "loss": 2.4587,
+      "step": 4330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41615357995033264,
+      "learning_rate": 9.703290938705114e-05,
+      "loss": 2.429,
+      "step": 4335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4069333076477051,
+      "learning_rate": 9.702596946981232e-05,
+      "loss": 2.4348,
+      "step": 4340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39312222599983215,
+      "learning_rate": 9.701902169470883e-05,
+      "loss": 2.4533,
+      "step": 4345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39762037992477417,
+      "learning_rate": 9.701206606290164e-05,
+      "loss": 2.4986,
+      "step": 4350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4414960741996765,
+      "learning_rate": 9.700510257555295e-05,
+      "loss": 2.4563,
+      "step": 4355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4010964334011078,
+      "learning_rate": 9.69981312338264e-05,
+      "loss": 2.4263,
+      "step": 4360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39740073680877686,
+      "learning_rate": 9.699115203888686e-05,
+      "loss": 2.4403,
+      "step": 4365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40610843896865845,
+      "learning_rate": 9.69841649919005e-05,
+      "loss": 2.4537,
+      "step": 4370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4123898446559906,
+      "learning_rate": 9.697717009403485e-05,
+      "loss": 2.4695,
+      "step": 4375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4198593199253082,
+      "learning_rate": 9.697016734645873e-05,
+      "loss": 2.5147,
+      "step": 4380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41294312477111816,
+      "learning_rate": 9.696315675034227e-05,
+      "loss": 2.3829,
+      "step": 4385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4083223342895508,
+      "learning_rate": 9.69561383068569e-05,
+      "loss": 2.5098,
+      "step": 4390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4225003123283386,
+      "learning_rate": 9.694911201717541e-05,
+      "loss": 2.4537,
+      "step": 4395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4027034640312195,
+      "learning_rate": 9.694207788247186e-05,
+      "loss": 2.5143,
+      "step": 4400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.41473209857940674,
+      "learning_rate": 9.693503590392159e-05,
+      "loss": 2.4866,
+      "step": 4405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4090598523616791,
+      "learning_rate": 9.692798608270132e-05,
+      "loss": 2.4102,
+      "step": 4410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4264538288116455,
+      "learning_rate": 9.692092841998904e-05,
+      "loss": 2.4915,
+      "step": 4415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.394085556268692,
+      "learning_rate": 9.691386291696407e-05,
+      "loss": 2.4817,
+      "step": 4420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4355419874191284,
+      "learning_rate": 9.690678957480702e-05,
+      "loss": 2.5428,
+      "step": 4425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42167991399765015,
+      "learning_rate": 9.689970839469981e-05,
+      "loss": 2.5219,
+      "step": 4430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41902291774749756,
+      "learning_rate": 9.689261937782571e-05,
+      "loss": 2.4345,
+      "step": 4435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39265498518943787,
+      "learning_rate": 9.688552252536924e-05,
+      "loss": 2.4548,
+      "step": 4440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41109657287597656,
+      "learning_rate": 9.687841783851626e-05,
+      "loss": 2.496,
+      "step": 4445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4115400016307831,
+      "learning_rate": 9.687130531845394e-05,
+      "loss": 2.5212,
+      "step": 4450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.410237580537796,
+      "learning_rate": 9.686418496637076e-05,
+      "loss": 2.5207,
+      "step": 4455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4169434905052185,
+      "learning_rate": 9.685705678345652e-05,
+      "loss": 2.4483,
+      "step": 4460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3927488923072815,
+      "learning_rate": 9.684992077090227e-05,
+      "loss": 2.4732,
+      "step": 4465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40053266286849976,
+      "learning_rate": 9.684277692990045e-05,
+      "loss": 2.4081,
+      "step": 4470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40560659766197205,
+      "learning_rate": 9.683562526164475e-05,
+      "loss": 2.437,
+      "step": 4475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41578593850135803,
+      "learning_rate": 9.68284657673302e-05,
+      "loss": 2.4911,
+      "step": 4480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40104132890701294,
+      "learning_rate": 9.682129844815311e-05,
+      "loss": 2.4657,
+      "step": 4485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39522239565849304,
+      "learning_rate": 9.681412330531112e-05,
+      "loss": 2.535,
+      "step": 4490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40305182337760925,
+      "learning_rate": 9.680694034000318e-05,
+      "loss": 2.4414,
+      "step": 4495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40569767355918884,
+      "learning_rate": 9.67997495534295e-05,
+      "loss": 2.526,
+      "step": 4500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3997107446193695,
+      "learning_rate": 9.679255094679169e-05,
+      "loss": 2.4905,
+      "step": 4505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41955456137657166,
+      "learning_rate": 9.678534452129256e-05,
+      "loss": 2.4691,
+      "step": 4510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4139942228794098,
+      "learning_rate": 9.677813027813632e-05,
+      "loss": 2.4667,
+      "step": 4515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42155954241752625,
+      "learning_rate": 9.677090821852842e-05,
+      "loss": 2.4479,
+      "step": 4520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41461116075515747,
+      "learning_rate": 9.676367834367562e-05,
+      "loss": 2.5793,
+      "step": 4525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41037657856941223,
+      "learning_rate": 9.675644065478602e-05,
+      "loss": 2.4791,
+      "step": 4530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4126727283000946,
+      "learning_rate": 9.674919515306904e-05,
+      "loss": 2.5096,
+      "step": 4535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39963558316230774,
+      "learning_rate": 9.674194183973533e-05,
+      "loss": 2.39,
+      "step": 4540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4051569104194641,
+      "learning_rate": 9.673468071599694e-05,
+      "loss": 2.5443,
+      "step": 4545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42032814025878906,
+      "learning_rate": 9.672741178306713e-05,
+      "loss": 2.4756,
+      "step": 4550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39982062578201294,
+      "learning_rate": 9.672013504216054e-05,
+      "loss": 2.4975,
+      "step": 4555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4073619544506073,
+      "learning_rate": 9.671285049449307e-05,
+      "loss": 2.4437,
+      "step": 4560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.430736780166626,
+      "learning_rate": 9.670555814128197e-05,
+      "loss": 2.468,
+      "step": 4565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4162490665912628,
+      "learning_rate": 9.669825798374574e-05,
+      "loss": 2.4479,
+      "step": 4570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40805068612098694,
+      "learning_rate": 9.66909500231042e-05,
+      "loss": 2.4613,
+      "step": 4575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4133608937263489,
+      "learning_rate": 9.668363426057852e-05,
+      "loss": 2.4906,
+      "step": 4580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38110628724098206,
+      "learning_rate": 9.66763106973911e-05,
+      "loss": 2.3995,
+      "step": 4585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4035351574420929,
+      "learning_rate": 9.666897933476568e-05,
+      "loss": 2.3792,
+      "step": 4590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4113156199455261,
+      "learning_rate": 9.666164017392734e-05,
+      "loss": 2.5416,
+      "step": 4595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4139128625392914,
+      "learning_rate": 9.665429321610239e-05,
+      "loss": 2.4044,
+      "step": 4600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4196441173553467,
+      "learning_rate": 9.66469384625185e-05,
+      "loss": 2.4943,
+      "step": 4605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4043656289577484,
+      "learning_rate": 9.66395759144046e-05,
+      "loss": 2.4743,
+      "step": 4610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4052789807319641,
+      "learning_rate": 9.6632205572991e-05,
+      "loss": 2.4485,
+      "step": 4615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40383073687553406,
+      "learning_rate": 9.662482743950918e-05,
+      "loss": 2.5287,
+      "step": 4620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40504327416419983,
+      "learning_rate": 9.661744151519206e-05,
+      "loss": 2.5068,
+      "step": 4625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4095880091190338,
+      "learning_rate": 9.661004780127373e-05,
+      "loss": 2.494,
+      "step": 4630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3840869069099426,
+      "learning_rate": 9.660264629898974e-05,
+      "loss": 2.508,
+      "step": 4635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3871769905090332,
+      "learning_rate": 9.65952370095768e-05,
+      "loss": 2.4531,
+      "step": 4640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44232887029647827,
+      "learning_rate": 9.6587819934273e-05,
+      "loss": 2.5047,
+      "step": 4645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4164472222328186,
+      "learning_rate": 9.658039507431768e-05,
+      "loss": 2.4038,
+      "step": 4650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3993256092071533,
+      "learning_rate": 9.657296243095151e-05,
+      "loss": 2.4376,
+      "step": 4655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4035753011703491,
+      "learning_rate": 9.656552200541648e-05,
+      "loss": 2.4529,
+      "step": 4660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4034801423549652,
+      "learning_rate": 9.655807379895583e-05,
+      "loss": 2.4739,
+      "step": 4665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40048956871032715,
+      "learning_rate": 9.655061781281413e-05,
+      "loss": 2.4635,
+      "step": 4670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3884316086769104,
+      "learning_rate": 9.654315404823729e-05,
+      "loss": 2.4749,
+      "step": 4675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4250464141368866,
+      "learning_rate": 9.653568250647242e-05,
+      "loss": 2.4126,
+      "step": 4680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43094536662101746,
+      "learning_rate": 9.652820318876802e-05,
+      "loss": 2.4485,
+      "step": 4685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4177592098712921,
+      "learning_rate": 9.652071609637385e-05,
+      "loss": 2.4697,
+      "step": 4690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41387543082237244,
+      "learning_rate": 9.651322123054097e-05,
+      "loss": 2.4799,
+      "step": 4695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.386737197637558,
+      "learning_rate": 9.650571859252175e-05,
+      "loss": 2.4297,
+      "step": 4700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.38170427083969116,
+      "learning_rate": 9.649820818356985e-05,
+      "loss": 2.5144,
+      "step": 4705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40773892402648926,
+      "learning_rate": 9.649069000494025e-05,
+      "loss": 2.5384,
+      "step": 4710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4061359465122223,
+      "learning_rate": 9.648316405788916e-05,
+      "loss": 2.4258,
+      "step": 4715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4290323257446289,
+      "learning_rate": 9.647563034367419e-05,
+      "loss": 2.4773,
+      "step": 4720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40184494853019714,
+      "learning_rate": 9.646808886355417e-05,
+      "loss": 2.4048,
+      "step": 4725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40181830525398254,
+      "learning_rate": 9.646053961878925e-05,
+      "loss": 2.5029,
+      "step": 4730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41655445098876953,
+      "learning_rate": 9.64529826106409e-05,
+      "loss": 2.4163,
+      "step": 4735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4431319236755371,
+      "learning_rate": 9.644541784037186e-05,
+      "loss": 2.4221,
+      "step": 4740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4178268313407898,
+      "learning_rate": 9.643784530924616e-05,
+      "loss": 2.4491,
+      "step": 4745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4163326025009155,
+      "learning_rate": 9.643026501852916e-05,
+      "loss": 2.4934,
+      "step": 4750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4302927851676941,
+      "learning_rate": 9.64226769694875e-05,
+      "loss": 2.4371,
+      "step": 4755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4373737871646881,
+      "learning_rate": 9.641508116338908e-05,
+      "loss": 2.384,
+      "step": 4760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4135447144508362,
+      "learning_rate": 9.640747760150319e-05,
+      "loss": 2.5036,
+      "step": 4765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42020687460899353,
+      "learning_rate": 9.639986628510032e-05,
+      "loss": 2.4444,
+      "step": 4770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40116384625434875,
+      "learning_rate": 9.639224721545229e-05,
+      "loss": 2.3508,
+      "step": 4775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40288034081459045,
+      "learning_rate": 9.638462039383223e-05,
+      "loss": 2.3954,
+      "step": 4780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3996966779232025,
+      "learning_rate": 9.637698582151454e-05,
+      "loss": 2.4574,
+      "step": 4785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42408308386802673,
+      "learning_rate": 9.636934349977493e-05,
+      "loss": 2.4063,
+      "step": 4790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39666783809661865,
+      "learning_rate": 9.636169342989043e-05,
+      "loss": 2.4849,
+      "step": 4795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3939882814884186,
+      "learning_rate": 9.635403561313932e-05,
+      "loss": 2.4394,
+      "step": 4800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4091871976852417,
+      "learning_rate": 9.634637005080119e-05,
+      "loss": 2.4773,
+      "step": 4805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40521231293678284,
+      "learning_rate": 9.633869674415693e-05,
+      "loss": 2.4294,
+      "step": 4810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3937463164329529,
+      "learning_rate": 9.633101569448873e-05,
+      "loss": 2.4716,
+      "step": 4815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4142480492591858,
+      "learning_rate": 9.632332690308004e-05,
+      "loss": 2.4419,
+      "step": 4820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40592092275619507,
+      "learning_rate": 9.631563037121565e-05,
+      "loss": 2.4973,
+      "step": 4825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39700067043304443,
+      "learning_rate": 9.63079261001816e-05,
+      "loss": 2.4923,
+      "step": 4830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4007636606693268,
+      "learning_rate": 9.630021409126529e-05,
+      "loss": 2.4348,
+      "step": 4835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38609179854393005,
+      "learning_rate": 9.629249434575531e-05,
+      "loss": 2.4617,
+      "step": 4840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4212386906147003,
+      "learning_rate": 9.628476686494164e-05,
+      "loss": 2.4658,
+      "step": 4845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4285244941711426,
+      "learning_rate": 9.627703165011551e-05,
+      "loss": 2.4602,
+      "step": 4850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40462741255760193,
+      "learning_rate": 9.626928870256944e-05,
+      "loss": 2.5474,
+      "step": 4855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4404885470867157,
+      "learning_rate": 9.626153802359724e-05,
+      "loss": 2.483,
+      "step": 4860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40638217329978943,
+      "learning_rate": 9.625377961449402e-05,
+      "loss": 2.4489,
+      "step": 4865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4204055070877075,
+      "learning_rate": 9.62460134765562e-05,
+      "loss": 2.4416,
+      "step": 4870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4029890298843384,
+      "learning_rate": 9.623823961108145e-05,
+      "loss": 2.4467,
+      "step": 4875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3889310657978058,
+      "learning_rate": 9.623045801936876e-05,
+      "loss": 2.4983,
+      "step": 4880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42707693576812744,
+      "learning_rate": 9.622266870271842e-05,
+      "loss": 2.4849,
+      "step": 4885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4149658977985382,
+      "learning_rate": 9.6214871662432e-05,
+      "loss": 2.46,
+      "step": 4890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41261357069015503,
+      "learning_rate": 9.620706689981232e-05,
+      "loss": 2.3546,
+      "step": 4895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40974244475364685,
+      "learning_rate": 9.619925441616354e-05,
+      "loss": 2.4291,
+      "step": 4900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39375630021095276,
+      "learning_rate": 9.619143421279112e-05,
+      "loss": 2.5077,
+      "step": 4905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4208444058895111,
+      "learning_rate": 9.618360629100176e-05,
+      "loss": 2.5134,
+      "step": 4910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3980614244937897,
+      "learning_rate": 9.61757706521035e-05,
+      "loss": 2.4028,
+      "step": 4915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38089823722839355,
+      "learning_rate": 9.616792729740564e-05,
+      "loss": 2.4856,
+      "step": 4920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4199288785457611,
+      "learning_rate": 9.616007622821877e-05,
+      "loss": 2.487,
+      "step": 4925
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40086662769317627,
+      "learning_rate": 9.615221744585476e-05,
+      "loss": 2.4267,
+      "step": 4930
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39533543586730957,
+      "learning_rate": 9.614435095162679e-05,
+      "loss": 2.4578,
+      "step": 4935
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3918769955635071,
+      "learning_rate": 9.613647674684936e-05,
+      "loss": 2.4703,
+      "step": 4940
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4021318256855011,
+      "learning_rate": 9.612859483283818e-05,
+      "loss": 2.4738,
+      "step": 4945
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39459228515625,
+      "learning_rate": 9.612070521091029e-05,
+      "loss": 2.4253,
+      "step": 4950
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38013097643852234,
+      "learning_rate": 9.611280788238403e-05,
+      "loss": 2.4573,
+      "step": 4955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3951113820075989,
+      "learning_rate": 9.610490284857901e-05,
+      "loss": 2.4251,
+      "step": 4960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3925844728946686,
+      "learning_rate": 9.609699011081613e-05,
+      "loss": 2.4596,
+      "step": 4965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38971617817878723,
+      "learning_rate": 9.608906967041757e-05,
+      "loss": 2.4814,
+      "step": 4970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41488340497016907,
+      "learning_rate": 9.608114152870684e-05,
+      "loss": 2.3808,
+      "step": 4975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4069625437259674,
+      "learning_rate": 9.607320568700866e-05,
+      "loss": 2.3428,
+      "step": 4980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40587151050567627,
+      "learning_rate": 9.606526214664911e-05,
+      "loss": 2.5119,
+      "step": 4985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4078567624092102,
+      "learning_rate": 9.605731090895551e-05,
+      "loss": 2.4886,
+      "step": 4990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4103011190891266,
+      "learning_rate": 9.604935197525649e-05,
+      "loss": 2.5059,
+      "step": 4995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3972037136554718,
+      "learning_rate": 9.604138534688196e-05,
+      "loss": 2.4198,
+      "step": 5000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43941161036491394,
+      "learning_rate": 9.60334110251631e-05,
+      "loss": 2.3465,
+      "step": 5005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4046022891998291,
+      "learning_rate": 9.602542901143241e-05,
+      "loss": 2.4767,
+      "step": 5010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4021283686161041,
+      "learning_rate": 9.601743930702363e-05,
+      "loss": 2.4854,
+      "step": 5015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42118024826049805,
+      "learning_rate": 9.600944191327184e-05,
+      "loss": 2.3643,
+      "step": 5020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39496299624443054,
+      "learning_rate": 9.600143683151335e-05,
+      "loss": 2.4641,
+      "step": 5025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40399137139320374,
+      "learning_rate": 9.59934240630858e-05,
+      "loss": 2.4594,
+      "step": 5030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4074048697948456,
+      "learning_rate": 9.598540360932806e-05,
+      "loss": 2.4522,
+      "step": 5035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38558343052864075,
+      "learning_rate": 9.597737547158035e-05,
+      "loss": 2.4426,
+      "step": 5040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3814246356487274,
+      "learning_rate": 9.596933965118413e-05,
+      "loss": 2.3511,
+      "step": 5045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40011048316955566,
+      "learning_rate": 9.596129614948217e-05,
+      "loss": 2.4408,
+      "step": 5050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39631766080856323,
+      "learning_rate": 9.595324496781849e-05,
+      "loss": 2.4206,
+      "step": 5055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39802655577659607,
+      "learning_rate": 9.594518610753841e-05,
+      "loss": 2.4386,
+      "step": 5060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37002652883529663,
+      "learning_rate": 9.593711956998855e-05,
+      "loss": 2.4547,
+      "step": 5065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41581979393959045,
+      "learning_rate": 9.59290453565168e-05,
+      "loss": 2.4528,
+      "step": 5070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39202460646629333,
+      "learning_rate": 9.592096346847232e-05,
+      "loss": 2.4427,
+      "step": 5075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39303281903266907,
+      "learning_rate": 9.591287390720555e-05,
+      "loss": 2.4354,
+      "step": 5080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39366966485977173,
+      "learning_rate": 9.590477667406828e-05,
+      "loss": 2.5116,
+      "step": 5085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4268970191478729,
+      "learning_rate": 9.589667177041347e-05,
+      "loss": 2.3854,
+      "step": 5090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39613068103790283,
+      "learning_rate": 9.588855919759544e-05,
+      "loss": 2.3764,
+      "step": 5095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3998851180076599,
+      "learning_rate": 9.588043895696978e-05,
+      "loss": 2.4832,
+      "step": 5100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39427807927131653,
+      "learning_rate": 9.587231104989334e-05,
+      "loss": 2.3869,
+      "step": 5105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41599225997924805,
+      "learning_rate": 9.586417547772426e-05,
+      "loss": 2.4732,
+      "step": 5110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38469091057777405,
+      "learning_rate": 9.585603224182196e-05,
+      "loss": 2.4784,
+      "step": 5115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40092623233795166,
+      "learning_rate": 9.584788134354718e-05,
+      "loss": 2.4946,
+      "step": 5120
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41128620505332947,
+      "learning_rate": 9.583972278426186e-05,
+      "loss": 2.4858,
+      "step": 5125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4057711660861969,
+      "learning_rate": 9.583155656532929e-05,
+      "loss": 2.4557,
+      "step": 5130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4270320534706116,
+      "learning_rate": 9.582338268811402e-05,
+      "loss": 2.44,
+      "step": 5135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3847527801990509,
+      "learning_rate": 9.581520115398185e-05,
+      "loss": 2.4209,
+      "step": 5140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4241020381450653,
+      "learning_rate": 9.580701196429991e-05,
+      "loss": 2.4704,
+      "step": 5145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40340834856033325,
+      "learning_rate": 9.579881512043655e-05,
+      "loss": 2.4625,
+      "step": 5150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4153643548488617,
+      "learning_rate": 9.579061062376146e-05,
+      "loss": 2.4528,
+      "step": 5155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39770615100860596,
+      "learning_rate": 9.578239847564559e-05,
+      "loss": 2.4097,
+      "step": 5160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3793421983718872,
+      "learning_rate": 9.577417867746113e-05,
+      "loss": 2.4882,
+      "step": 5165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39867520332336426,
+      "learning_rate": 9.576595123058158e-05,
+      "loss": 2.4704,
+      "step": 5170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4178451597690582,
+      "learning_rate": 9.575771613638175e-05,
+      "loss": 2.3811,
+      "step": 5175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39590251445770264,
+      "learning_rate": 9.574947339623765e-05,
+      "loss": 2.4877,
+      "step": 5180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4338364005088806,
+      "learning_rate": 9.574122301152665e-05,
+      "loss": 2.4229,
+      "step": 5185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.379643976688385,
+      "learning_rate": 9.573296498362733e-05,
+      "loss": 2.4613,
+      "step": 5190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38619285821914673,
+      "learning_rate": 9.572469931391959e-05,
+      "loss": 2.3996,
+      "step": 5195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40440914034843445,
+      "learning_rate": 9.571642600378459e-05,
+      "loss": 2.4049,
+      "step": 5200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41194647550582886,
+      "learning_rate": 9.570814505460475e-05,
+      "loss": 2.4301,
+      "step": 5205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4201595187187195,
+      "learning_rate": 9.569985646776382e-05,
+      "loss": 2.4759,
+      "step": 5210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39703169465065,
+      "learning_rate": 9.569156024464677e-05,
+      "loss": 2.5305,
+      "step": 5215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39247146248817444,
+      "learning_rate": 9.568325638663989e-05,
+      "loss": 2.4129,
+      "step": 5220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3878908157348633,
+      "learning_rate": 9.567494489513068e-05,
+      "loss": 2.4125,
+      "step": 5225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4074651300907135,
+      "learning_rate": 9.5666625771508e-05,
+      "loss": 2.3479,
+      "step": 5230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39090603590011597,
+      "learning_rate": 9.565829901716194e-05,
+      "loss": 2.4624,
+      "step": 5235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.401667982339859,
+      "learning_rate": 9.564996463348384e-05,
+      "loss": 2.3351,
+      "step": 5240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4153296947479248,
+      "learning_rate": 9.564162262186638e-05,
+      "loss": 2.4917,
+      "step": 5245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4176900088787079,
+      "learning_rate": 9.563327298370347e-05,
+      "loss": 2.4223,
+      "step": 5250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37582558393478394,
+      "learning_rate": 9.56249157203903e-05,
+      "loss": 2.4407,
+      "step": 5255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4079509675502777,
+      "learning_rate": 9.561655083332333e-05,
+      "loss": 2.393,
+      "step": 5260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40055549144744873,
+      "learning_rate": 9.56081783239003e-05,
+      "loss": 2.4334,
+      "step": 5265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39478209614753723,
+      "learning_rate": 9.559979819352024e-05,
+      "loss": 2.4013,
+      "step": 5270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4110153317451477,
+      "learning_rate": 9.559141044358342e-05,
+      "loss": 2.4603,
+      "step": 5275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3973243832588196,
+      "learning_rate": 9.558301507549143e-05,
+      "loss": 2.4569,
+      "step": 5280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42099717259407043,
+      "learning_rate": 9.557461209064708e-05,
+      "loss": 2.4163,
+      "step": 5285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40605753660202026,
+      "learning_rate": 9.556620149045447e-05,
+      "loss": 2.4169,
+      "step": 5290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3915252089500427,
+      "learning_rate": 9.555778327631903e-05,
+      "loss": 2.439,
+      "step": 5295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4309784770011902,
+      "learning_rate": 9.554935744964734e-05,
+      "loss": 2.5046,
+      "step": 5300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3887786865234375,
+      "learning_rate": 9.55409240118474e-05,
+      "loss": 2.363,
+      "step": 5305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39914610981941223,
+      "learning_rate": 9.553248296432835e-05,
+      "loss": 2.4983,
+      "step": 5310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39786607027053833,
+      "learning_rate": 9.552403430850068e-05,
+      "loss": 2.4001,
+      "step": 5315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3891544044017792,
+      "learning_rate": 9.551557804577612e-05,
+      "loss": 2.5082,
+      "step": 5320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4040165841579437,
+      "learning_rate": 9.550711417756767e-05,
+      "loss": 2.4373,
+      "step": 5325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3960551917552948,
+      "learning_rate": 9.549864270528965e-05,
+      "loss": 2.4134,
+      "step": 5330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4053976535797119,
+      "learning_rate": 9.549016363035758e-05,
+      "loss": 2.4591,
+      "step": 5335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4130423665046692,
+      "learning_rate": 9.548167695418828e-05,
+      "loss": 2.3855,
+      "step": 5340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.397362619638443,
+      "learning_rate": 9.547318267819986e-05,
+      "loss": 2.4776,
+      "step": 5345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37679511308670044,
+      "learning_rate": 9.546468080381167e-05,
+      "loss": 2.4228,
+      "step": 5350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40481284260749817,
+      "learning_rate": 9.545617133244436e-05,
+      "loss": 2.4866,
+      "step": 5355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3954336643218994,
+      "learning_rate": 9.544765426551981e-05,
+      "loss": 2.3357,
+      "step": 5360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38092270493507385,
+      "learning_rate": 9.543912960446117e-05,
+      "loss": 2.4064,
+      "step": 5365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3876906931400299,
+      "learning_rate": 9.543059735069293e-05,
+      "loss": 2.3808,
+      "step": 5370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3989468812942505,
+      "learning_rate": 9.542205750564077e-05,
+      "loss": 2.4977,
+      "step": 5375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.43544018268585205,
+      "learning_rate": 9.541351007073167e-05,
+      "loss": 2.5574,
+      "step": 5380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39048171043395996,
+      "learning_rate": 9.540495504739387e-05,
+      "loss": 2.3934,
+      "step": 5385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3827662467956543,
+      "learning_rate": 9.53963924370569e-05,
+      "loss": 2.4266,
+      "step": 5390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3935962915420532,
+      "learning_rate": 9.538782224115151e-05,
+      "loss": 2.3951,
+      "step": 5395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38034117221832275,
+      "learning_rate": 9.537924446110976e-05,
+      "loss": 2.4441,
+      "step": 5400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3968839943408966,
+      "learning_rate": 9.537065909836497e-05,
+      "loss": 2.4804,
+      "step": 5405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3775872588157654,
+      "learning_rate": 9.536206615435174e-05,
+      "loss": 2.4965,
+      "step": 5410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38062402606010437,
+      "learning_rate": 9.535346563050591e-05,
+      "loss": 2.388,
+      "step": 5415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40295079350471497,
+      "learning_rate": 9.534485752826456e-05,
+      "loss": 2.5121,
+      "step": 5420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38589558005332947,
+      "learning_rate": 9.53362418490661e-05,
+      "loss": 2.4828,
+      "step": 5425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38293901085853577,
+      "learning_rate": 9.53276185943502e-05,
+      "loss": 2.4172,
+      "step": 5430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38192740082740784,
+      "learning_rate": 9.531898776555773e-05,
+      "loss": 2.4655,
+      "step": 5435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42262181639671326,
+      "learning_rate": 9.531034936413091e-05,
+      "loss": 2.5089,
+      "step": 5440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4176526665687561,
+      "learning_rate": 9.530170339151314e-05,
+      "loss": 2.4139,
+      "step": 5445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4093663990497589,
+      "learning_rate": 9.529304984914917e-05,
+      "loss": 2.4872,
+      "step": 5450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39079737663269043,
+      "learning_rate": 9.528438873848496e-05,
+      "loss": 2.3985,
+      "step": 5455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4138225317001343,
+      "learning_rate": 9.527572006096773e-05,
+      "loss": 2.3646,
+      "step": 5460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3815244734287262,
+      "learning_rate": 9.526704381804601e-05,
+      "loss": 2.3994,
+      "step": 5465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39213261008262634,
+      "learning_rate": 9.525836001116957e-05,
+      "loss": 2.4579,
+      "step": 5470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38995885848999023,
+      "learning_rate": 9.524966864178943e-05,
+      "loss": 2.3882,
+      "step": 5475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40092089772224426,
+      "learning_rate": 9.524096971135789e-05,
+      "loss": 2.4613,
+      "step": 5480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3962084650993347,
+      "learning_rate": 9.52322632213285e-05,
+      "loss": 2.4514,
+      "step": 5485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.397453635931015,
+      "learning_rate": 9.52235491731561e-05,
+      "loss": 2.4904,
+      "step": 5490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.408712774515152,
+      "learning_rate": 9.521482756829675e-05,
+      "loss": 2.472,
+      "step": 5495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3814677596092224,
+      "learning_rate": 9.520609840820782e-05,
+      "loss": 2.4547,
+      "step": 5500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3791644871234894,
+      "learning_rate": 9.51973616943479e-05,
+      "loss": 2.4531,
+      "step": 5505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3921375274658203,
+      "learning_rate": 9.518861742817689e-05,
+      "loss": 2.5012,
+      "step": 5510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40556326508522034,
+      "learning_rate": 9.51798656111559e-05,
+      "loss": 2.4992,
+      "step": 5515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41510578989982605,
+      "learning_rate": 9.517110624474733e-05,
+      "loss": 2.4534,
+      "step": 5520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38701075315475464,
+      "learning_rate": 9.516233933041485e-05,
+      "loss": 2.4327,
+      "step": 5525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4303721487522125,
+      "learning_rate": 9.515356486962337e-05,
+      "loss": 2.4574,
+      "step": 5530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.415055513381958,
+      "learning_rate": 9.514478286383907e-05,
+      "loss": 2.3824,
+      "step": 5535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38568180799484253,
+      "learning_rate": 9.513599331452939e-05,
+      "loss": 2.4145,
+      "step": 5540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37910565733909607,
+      "learning_rate": 9.512719622316304e-05,
+      "loss": 2.3781,
+      "step": 5545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39074745774269104,
+      "learning_rate": 9.511839159120997e-05,
+      "loss": 2.4549,
+      "step": 5550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3895575702190399,
+      "learning_rate": 9.51095794201414e-05,
+      "loss": 2.3949,
+      "step": 5555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3753109574317932,
+      "learning_rate": 9.510075971142983e-05,
+      "loss": 2.3692,
+      "step": 5560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40365585684776306,
+      "learning_rate": 9.509193246654896e-05,
+      "loss": 2.3999,
+      "step": 5565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40175122022628784,
+      "learning_rate": 9.508309768697384e-05,
+      "loss": 2.3752,
+      "step": 5570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39689481258392334,
+      "learning_rate": 9.507425537418068e-05,
+      "loss": 2.4099,
+      "step": 5575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40409594774246216,
+      "learning_rate": 9.506540552964704e-05,
+      "loss": 2.3648,
+      "step": 5580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40519729256629944,
+      "learning_rate": 9.505654815485168e-05,
+      "loss": 2.4591,
+      "step": 5585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40129414200782776,
+      "learning_rate": 9.504768325127464e-05,
+      "loss": 2.4656,
+      "step": 5590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3933297097682953,
+      "learning_rate": 9.503881082039719e-05,
+      "loss": 2.3779,
+      "step": 5595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42640215158462524,
+      "learning_rate": 9.50299308637019e-05,
+      "loss": 2.5053,
+      "step": 5600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4122491776943207,
+      "learning_rate": 9.502104338267257e-05,
+      "loss": 2.4981,
+      "step": 5605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37676215171813965,
+      "learning_rate": 9.501214837879428e-05,
+      "loss": 2.4098,
+      "step": 5610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38315221667289734,
+      "learning_rate": 9.500324585355333e-05,
+      "loss": 2.459,
+      "step": 5615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3952159285545349,
+      "learning_rate": 9.499433580843732e-05,
+      "loss": 2.3639,
+      "step": 5620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3769629895687103,
+      "learning_rate": 9.498541824493506e-05,
+      "loss": 2.3755,
+      "step": 5625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3893836438655853,
+      "learning_rate": 9.497649316453667e-05,
+      "loss": 2.3779,
+      "step": 5630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4384431838989258,
+      "learning_rate": 9.496756056873348e-05,
+      "loss": 2.5225,
+      "step": 5635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3812290132045746,
+      "learning_rate": 9.49586204590181e-05,
+      "loss": 2.4341,
+      "step": 5640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3952697515487671,
+      "learning_rate": 9.494967283688438e-05,
+      "loss": 2.3608,
+      "step": 5645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39532315731048584,
+      "learning_rate": 9.494071770382745e-05,
+      "loss": 2.4671,
+      "step": 5650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3892088234424591,
+      "learning_rate": 9.493175506134367e-05,
+      "loss": 2.4154,
+      "step": 5655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3913530111312866,
+      "learning_rate": 9.492278491093067e-05,
+      "loss": 2.4106,
+      "step": 5660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40939587354660034,
+      "learning_rate": 9.491380725408731e-05,
+      "loss": 2.4123,
+      "step": 5665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3894113004207611,
+      "learning_rate": 9.490482209231374e-05,
+      "loss": 2.4491,
+      "step": 5670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45230886340141296,
+      "learning_rate": 9.489582942711135e-05,
+      "loss": 2.4819,
+      "step": 5675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4150765538215637,
+      "learning_rate": 9.488682925998277e-05,
+      "loss": 2.3844,
+      "step": 5680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3884260952472687,
+      "learning_rate": 9.48778215924319e-05,
+      "loss": 2.4679,
+      "step": 5685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40213847160339355,
+      "learning_rate": 9.48688064259639e-05,
+      "loss": 2.5037,
+      "step": 5690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3800782263278961,
+      "learning_rate": 9.485978376208513e-05,
+      "loss": 2.3961,
+      "step": 5695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39880743622779846,
+      "learning_rate": 9.485075360230328e-05,
+      "loss": 2.4583,
+      "step": 5700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39435383677482605,
+      "learning_rate": 9.484171594812728e-05,
+      "loss": 2.4626,
+      "step": 5705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3846075236797333,
+      "learning_rate": 9.483267080106723e-05,
+      "loss": 2.4126,
+      "step": 5710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39223694801330566,
+      "learning_rate": 9.482361816263456e-05,
+      "loss": 2.41,
+      "step": 5715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39790600538253784,
+      "learning_rate": 9.481455803434195e-05,
+      "loss": 2.3927,
+      "step": 5720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3857821822166443,
+      "learning_rate": 9.48054904177033e-05,
+      "loss": 2.4473,
+      "step": 5725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40189844369888306,
+      "learning_rate": 9.479641531423377e-05,
+      "loss": 2.4702,
+      "step": 5730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5108299851417542,
+      "learning_rate": 9.47873327254498e-05,
+      "loss": 2.324,
+      "step": 5735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40055859088897705,
+      "learning_rate": 9.477824265286904e-05,
+      "loss": 2.4045,
+      "step": 5740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3885335624217987,
+      "learning_rate": 9.476914509801039e-05,
+      "loss": 2.4126,
+      "step": 5745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3765617609024048,
+      "learning_rate": 9.476004006239406e-05,
+      "loss": 2.4271,
+      "step": 5750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38724055886268616,
+      "learning_rate": 9.475092754754142e-05,
+      "loss": 2.4676,
+      "step": 5755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4004720449447632,
+      "learning_rate": 9.474180755497517e-05,
+      "loss": 2.4736,
+      "step": 5760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4086596369743347,
+      "learning_rate": 9.473268008621922e-05,
+      "loss": 2.4402,
+      "step": 5765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37814053893089294,
+      "learning_rate": 9.472354514279874e-05,
+      "loss": 2.401,
+      "step": 5770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38661399483680725,
+      "learning_rate": 9.471440272624012e-05,
+      "loss": 2.4925,
+      "step": 5775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38773465156555176,
+      "learning_rate": 9.470525283807105e-05,
+      "loss": 2.438,
+      "step": 5780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3897664248943329,
+      "learning_rate": 9.469609547982044e-05,
+      "loss": 2.4454,
+      "step": 5785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38839736580848694,
+      "learning_rate": 9.468693065301843e-05,
+      "loss": 2.4442,
+      "step": 5790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3962596654891968,
+      "learning_rate": 9.467775835919646e-05,
+      "loss": 2.4548,
+      "step": 5795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.420775830745697,
+      "learning_rate": 9.466857859988715e-05,
+      "loss": 2.5167,
+      "step": 5800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4196512699127197,
+      "learning_rate": 9.465939137662444e-05,
+      "loss": 2.4244,
+      "step": 5805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3760147988796234,
+      "learning_rate": 9.465019669094346e-05,
+      "loss": 2.4172,
+      "step": 5810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3944096565246582,
+      "learning_rate": 9.464099454438062e-05,
+      "loss": 2.4217,
+      "step": 5815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3874826729297638,
+      "learning_rate": 9.463178493847355e-05,
+      "loss": 2.4286,
+      "step": 5820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.371317595243454,
+      "learning_rate": 9.462256787476113e-05,
+      "loss": 2.4435,
+      "step": 5825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.392147034406662,
+      "learning_rate": 9.461334335478354e-05,
+      "loss": 2.4258,
+      "step": 5830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4052322208881378,
+      "learning_rate": 9.460411138008214e-05,
+      "loss": 2.4804,
+      "step": 5835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4012030065059662,
+      "learning_rate": 9.459487195219954e-05,
+      "loss": 2.3621,
+      "step": 5840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37887150049209595,
+      "learning_rate": 9.458562507267963e-05,
+      "loss": 2.3396,
+      "step": 5845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3839370310306549,
+      "learning_rate": 9.457637074306754e-05,
+      "loss": 2.4381,
+      "step": 5850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3762620985507965,
+      "learning_rate": 9.456710896490961e-05,
+      "loss": 2.3873,
+      "step": 5855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39563941955566406,
+      "learning_rate": 9.455783973975348e-05,
+      "loss": 2.4681,
+      "step": 5860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3977762460708618,
+      "learning_rate": 9.454856306914797e-05,
+      "loss": 2.3781,
+      "step": 5865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38467150926589966,
+      "learning_rate": 9.453927895464321e-05,
+      "loss": 2.4272,
+      "step": 5870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38536497950553894,
+      "learning_rate": 9.452998739779051e-05,
+      "loss": 2.4255,
+      "step": 5875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4174638092517853,
+      "learning_rate": 9.452068840014246e-05,
+      "loss": 2.3326,
+      "step": 5880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37771713733673096,
+      "learning_rate": 9.451138196325291e-05,
+      "loss": 2.4433,
+      "step": 5885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4092395305633545,
+      "learning_rate": 9.45020680886769e-05,
+      "loss": 2.4069,
+      "step": 5890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40871527791023254,
+      "learning_rate": 9.449274677797078e-05,
+      "loss": 2.4577,
+      "step": 5895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38769423961639404,
+      "learning_rate": 9.448341803269208e-05,
+      "loss": 2.4234,
+      "step": 5900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40050870180130005,
+      "learning_rate": 9.44740818543996e-05,
+      "loss": 2.4633,
+      "step": 5905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4020121693611145,
+      "learning_rate": 9.44647382446534e-05,
+      "loss": 2.3396,
+      "step": 5910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3885359764099121,
+      "learning_rate": 9.445538720501474e-05,
+      "loss": 2.4243,
+      "step": 5915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4251491129398346,
+      "learning_rate": 9.444602873704616e-05,
+      "loss": 2.4437,
+      "step": 5920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38989201188087463,
+      "learning_rate": 9.443666284231141e-05,
+      "loss": 2.4232,
+      "step": 5925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3798200190067291,
+      "learning_rate": 9.44272895223755e-05,
+      "loss": 2.4086,
+      "step": 5930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39673200249671936,
+      "learning_rate": 9.441790877880469e-05,
+      "loss": 2.4662,
+      "step": 5935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37593281269073486,
+      "learning_rate": 9.440852061316645e-05,
+      "loss": 2.5326,
+      "step": 5940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3969704508781433,
+      "learning_rate": 9.439912502702952e-05,
+      "loss": 2.4703,
+      "step": 5945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4016268849372864,
+      "learning_rate": 9.438972202196387e-05,
+      "loss": 2.4129,
+      "step": 5950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3959368169307709,
+      "learning_rate": 9.43803115995407e-05,
+      "loss": 2.3916,
+      "step": 5955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3908936381340027,
+      "learning_rate": 9.437089376133246e-05,
+      "loss": 2.5037,
+      "step": 5960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48665153980255127,
+      "learning_rate": 9.436146850891283e-05,
+      "loss": 2.4688,
+      "step": 5965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3895121216773987,
+      "learning_rate": 9.435203584385673e-05,
+      "loss": 2.3468,
+      "step": 5970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40200427174568176,
+      "learning_rate": 9.434259576774034e-05,
+      "loss": 2.3399,
+      "step": 5975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40870627760887146,
+      "learning_rate": 9.433314828214103e-05,
+      "loss": 2.4714,
+      "step": 5980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39395463466644287,
+      "learning_rate": 9.43236933886375e-05,
+      "loss": 2.3708,
+      "step": 5985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3657365143299103,
+      "learning_rate": 9.431423108880957e-05,
+      "loss": 2.4139,
+      "step": 5990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3777257800102234,
+      "learning_rate": 9.430476138423838e-05,
+      "loss": 2.4375,
+      "step": 5995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40575021505355835,
+      "learning_rate": 9.429528427650627e-05,
+      "loss": 2.3679,
+      "step": 6000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3914792239665985,
+      "learning_rate": 9.428579976719684e-05,
+      "loss": 2.3786,
+      "step": 6005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3925821781158447,
+      "learning_rate": 9.427630785789491e-05,
+      "loss": 2.4876,
+      "step": 6010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37913426756858826,
+      "learning_rate": 9.426680855018654e-05,
+      "loss": 2.3947,
+      "step": 6015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3748929798603058,
+      "learning_rate": 9.425730184565907e-05,
+      "loss": 2.3483,
+      "step": 6020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.383169561624527,
+      "learning_rate": 9.424778774590096e-05,
+      "loss": 2.3929,
+      "step": 6025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.419164776802063,
+      "learning_rate": 9.423826625250204e-05,
+      "loss": 2.3447,
+      "step": 6030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39967161417007446,
+      "learning_rate": 9.422873736705329e-05,
+      "loss": 2.4077,
+      "step": 6035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39952757954597473,
+      "learning_rate": 9.421920109114696e-05,
+      "loss": 2.4031,
+      "step": 6040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4203836917877197,
+      "learning_rate": 9.420965742637653e-05,
+      "loss": 2.4356,
+      "step": 6045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3993808329105377,
+      "learning_rate": 9.420010637433669e-05,
+      "loss": 2.497,
+      "step": 6050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38815444707870483,
+      "learning_rate": 9.419054793662342e-05,
+      "loss": 2.4432,
+      "step": 6055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4046403765678406,
+      "learning_rate": 9.418098211483387e-05,
+      "loss": 2.3448,
+      "step": 6060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3980211317539215,
+      "learning_rate": 9.417140891056646e-05,
+      "loss": 2.4336,
+      "step": 6065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37483617663383484,
+      "learning_rate": 9.416182832542085e-05,
+      "loss": 2.4201,
+      "step": 6070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3780595064163208,
+      "learning_rate": 9.41522403609979e-05,
+      "loss": 2.4001,
+      "step": 6075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40683647990226746,
+      "learning_rate": 9.414264501889975e-05,
+      "loss": 2.4874,
+      "step": 6080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4124225676059723,
+      "learning_rate": 9.41330423007297e-05,
+      "loss": 2.4717,
+      "step": 6085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3934960663318634,
+      "learning_rate": 9.412343220809239e-05,
+      "loss": 2.4173,
+      "step": 6090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37683990597724915,
+      "learning_rate": 9.41138147425936e-05,
+      "loss": 2.3285,
+      "step": 6095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37970080971717834,
+      "learning_rate": 9.410418990584034e-05,
+      "loss": 2.3397,
+      "step": 6100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3704308271408081,
+      "learning_rate": 9.409455769944093e-05,
+      "loss": 2.4172,
+      "step": 6105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3822551965713501,
+      "learning_rate": 9.408491812500486e-05,
+      "loss": 2.4202,
+      "step": 6110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38785162568092346,
+      "learning_rate": 9.407527118414287e-05,
+      "loss": 2.3783,
+      "step": 6115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.396711140871048,
+      "learning_rate": 9.406561687846693e-05,
+      "loss": 2.4089,
+      "step": 6120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40308666229248047,
+      "learning_rate": 9.405595520959025e-05,
+      "loss": 2.3443,
+      "step": 6125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38664910197257996,
+      "learning_rate": 9.404628617912723e-05,
+      "loss": 2.4525,
+      "step": 6130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3996780812740326,
+      "learning_rate": 9.403660978869353e-05,
+      "loss": 2.3944,
+      "step": 6135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39979174733161926,
+      "learning_rate": 9.402692603990608e-05,
+      "loss": 2.5282,
+      "step": 6140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39089006185531616,
+      "learning_rate": 9.401723493438295e-05,
+      "loss": 2.4322,
+      "step": 6145
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3940250873565674,
+      "learning_rate": 9.400753647374352e-05,
+      "loss": 2.3821,
+      "step": 6150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4002339243888855,
+      "learning_rate": 9.399783065960835e-05,
+      "loss": 2.4235,
+      "step": 6155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3870840072631836,
+      "learning_rate": 9.398811749359924e-05,
+      "loss": 2.4365,
+      "step": 6160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4181632101535797,
+      "learning_rate": 9.397839697733924e-05,
+      "loss": 2.4216,
+      "step": 6165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3910233676433563,
+      "learning_rate": 9.396866911245261e-05,
+      "loss": 2.4368,
+      "step": 6170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3747069537639618,
+      "learning_rate": 9.395893390056484e-05,
+      "loss": 2.3396,
+      "step": 6175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3902587294578552,
+      "learning_rate": 9.394919134330264e-05,
+      "loss": 2.392,
+      "step": 6180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4008750319480896,
+      "learning_rate": 9.393944144229396e-05,
+      "loss": 2.3188,
+      "step": 6185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3749091923236847,
+      "learning_rate": 9.392968419916797e-05,
+      "loss": 2.3062,
+      "step": 6190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3987392485141754,
+      "learning_rate": 9.391991961555507e-05,
+      "loss": 2.3832,
+      "step": 6195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.393453449010849,
+      "learning_rate": 9.39101476930869e-05,
+      "loss": 2.3764,
+      "step": 6200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3956431746482849,
+      "learning_rate": 9.390036843339629e-05,
+      "loss": 2.3156,
+      "step": 6205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39189404249191284,
+      "learning_rate": 9.389058183811732e-05,
+      "loss": 2.3806,
+      "step": 6210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3952174782752991,
+      "learning_rate": 9.38807879088853e-05,
+      "loss": 2.4435,
+      "step": 6215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.381058007478714,
+      "learning_rate": 9.387098664733676e-05,
+      "loss": 2.3024,
+      "step": 6220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38341376185417175,
+      "learning_rate": 9.386117805510945e-05,
+      "loss": 2.495,
+      "step": 6225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38799741864204407,
+      "learning_rate": 9.385136213384237e-05,
+      "loss": 2.4104,
+      "step": 6230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4043133556842804,
+      "learning_rate": 9.38415388851757e-05,
+      "loss": 2.4016,
+      "step": 6235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.399460107088089,
+      "learning_rate": 9.383170831075086e-05,
+      "loss": 2.3916,
+      "step": 6240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3926413953304291,
+      "learning_rate": 9.382187041221054e-05,
+      "loss": 2.3741,
+      "step": 6245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38559192419052124,
+      "learning_rate": 9.381202519119858e-05,
+      "loss": 2.4605,
+      "step": 6250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4008232057094574,
+      "learning_rate": 9.38021726493601e-05,
+      "loss": 2.3779,
+      "step": 6255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38708487153053284,
+      "learning_rate": 9.37923127883414e-05,
+      "loss": 2.4033,
+      "step": 6260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42726290225982666,
+      "learning_rate": 9.378244560979007e-05,
+      "loss": 2.4309,
+      "step": 6265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.401485800743103,
+      "learning_rate": 9.377257111535482e-05,
+      "loss": 2.4537,
+      "step": 6270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43589451909065247,
+      "learning_rate": 9.376268930668569e-05,
+      "loss": 2.3739,
+      "step": 6275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39421597123146057,
+      "learning_rate": 9.375280018543387e-05,
+      "loss": 2.3996,
+      "step": 6280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43092796206474304,
+      "learning_rate": 9.374290375325181e-05,
+      "loss": 2.4209,
+      "step": 6285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4012669622898102,
+      "learning_rate": 9.373300001179316e-05,
+      "loss": 2.3706,
+      "step": 6290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3664630949497223,
+      "learning_rate": 9.37230889627128e-05,
+      "loss": 2.3824,
+      "step": 6295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3839011490345001,
+      "learning_rate": 9.371317060766682e-05,
+      "loss": 2.3156,
+      "step": 6300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3944704532623291,
+      "learning_rate": 9.370324494831254e-05,
+      "loss": 2.488,
+      "step": 6305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39250221848487854,
+      "learning_rate": 9.369331198630852e-05,
+      "loss": 2.3703,
+      "step": 6310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39754363894462585,
+      "learning_rate": 9.36833717233145e-05,
+      "loss": 2.3823,
+      "step": 6315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3803838789463043,
+      "learning_rate": 9.367342416099147e-05,
+      "loss": 2.4152,
+      "step": 6320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3671237826347351,
+      "learning_rate": 9.366346930100163e-05,
+      "loss": 2.3829,
+      "step": 6325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4007478356361389,
+      "learning_rate": 9.36535071450084e-05,
+      "loss": 2.4236,
+      "step": 6330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37629905343055725,
+      "learning_rate": 9.364353769467643e-05,
+      "loss": 2.4095,
+      "step": 6335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39403778314590454,
+      "learning_rate": 9.363356095167156e-05,
+      "loss": 2.3801,
+      "step": 6340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39557212591171265,
+      "learning_rate": 9.362357691766088e-05,
+      "loss": 2.4434,
+      "step": 6345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.40919315814971924,
+      "learning_rate": 9.361358559431266e-05,
+      "loss": 2.4564,
+      "step": 6350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36079251766204834,
+      "learning_rate": 9.360358698329646e-05,
+      "loss": 2.4224,
+      "step": 6355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39171838760375977,
+      "learning_rate": 9.359358108628298e-05,
+      "loss": 2.4422,
+      "step": 6360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3822607696056366,
+      "learning_rate": 9.358356790494416e-05,
+      "loss": 2.4013,
+      "step": 6365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4140118956565857,
+      "learning_rate": 9.357354744095317e-05,
+      "loss": 2.4445,
+      "step": 6370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38803187012672424,
+      "learning_rate": 9.35635196959844e-05,
+      "loss": 2.443,
+      "step": 6375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4136311113834381,
+      "learning_rate": 9.355348467171347e-05,
+      "loss": 2.3055,
+      "step": 6380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38874727487564087,
+      "learning_rate": 9.354344236981716e-05,
+      "loss": 2.2899,
+      "step": 6385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3795917332172394,
+      "learning_rate": 9.353339279197351e-05,
+      "loss": 2.3876,
+      "step": 6390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39464622735977173,
+      "learning_rate": 9.352333593986176e-05,
+      "loss": 2.4079,
+      "step": 6395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3705817461013794,
+      "learning_rate": 9.35132718151624e-05,
+      "loss": 2.4248,
+      "step": 6400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3777334988117218,
+      "learning_rate": 9.35032004195571e-05,
+      "loss": 2.4728,
+      "step": 6405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37659400701522827,
+      "learning_rate": 9.349312175472873e-05,
+      "loss": 2.4948,
+      "step": 6410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3972844183444977,
+      "learning_rate": 9.348303582236142e-05,
+      "loss": 2.3748,
+      "step": 6415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3764602839946747,
+      "learning_rate": 9.347294262414047e-05,
+      "loss": 2.4377,
+      "step": 6420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4085213541984558,
+      "learning_rate": 9.346284216175245e-05,
+      "loss": 2.3518,
+      "step": 6425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3861781656742096,
+      "learning_rate": 9.345273443688509e-05,
+      "loss": 2.2892,
+      "step": 6430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39590325951576233,
+      "learning_rate": 9.344261945122734e-05,
+      "loss": 2.5099,
+      "step": 6435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38799160718917847,
+      "learning_rate": 9.343249720646939e-05,
+      "loss": 2.3532,
+      "step": 6440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40191248059272766,
+      "learning_rate": 9.342236770430264e-05,
+      "loss": 2.4279,
+      "step": 6445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37825435400009155,
+      "learning_rate": 9.341223094641967e-05,
+      "loss": 2.3651,
+      "step": 6450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39434614777565,
+      "learning_rate": 9.340208693451431e-05,
+      "loss": 2.3741,
+      "step": 6455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37207460403442383,
+      "learning_rate": 9.339193567028157e-05,
+      "loss": 2.3596,
+      "step": 6460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3963320553302765,
+      "learning_rate": 9.338177715541772e-05,
+      "loss": 2.3998,
+      "step": 6465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36525920033454895,
+      "learning_rate": 9.337161139162019e-05,
+      "loss": 2.3769,
+      "step": 6470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37490904331207275,
+      "learning_rate": 9.336143838058761e-05,
+      "loss": 2.456,
+      "step": 6475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3879038095474243,
+      "learning_rate": 9.335125812401993e-05,
+      "loss": 2.3809,
+      "step": 6480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3937845230102539,
+      "learning_rate": 9.334107062361816e-05,
+      "loss": 2.3845,
+      "step": 6485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3957381248474121,
+      "learning_rate": 9.333087588108462e-05,
+      "loss": 2.4641,
+      "step": 6490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38195934891700745,
+      "learning_rate": 9.332067389812282e-05,
+      "loss": 2.338,
+      "step": 6495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3919680118560791,
+      "learning_rate": 9.331046467643746e-05,
+      "loss": 2.3498,
+      "step": 6500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38248682022094727,
+      "learning_rate": 9.330024821773448e-05,
+      "loss": 2.3887,
+      "step": 6505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4023313820362091,
+      "learning_rate": 9.329002452372102e-05,
+      "loss": 2.442,
+      "step": 6510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39164501428604126,
+      "learning_rate": 9.327979359610537e-05,
+      "loss": 2.4106,
+      "step": 6515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4116343855857849,
+      "learning_rate": 9.326955543659714e-05,
+      "loss": 2.4123,
+      "step": 6520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38497623801231384,
+      "learning_rate": 9.325931004690706e-05,
+      "loss": 2.4136,
+      "step": 6525
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39222922921180725,
+      "learning_rate": 9.324905742874709e-05,
+      "loss": 2.3942,
+      "step": 6530
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38775062561035156,
+      "learning_rate": 9.323879758383042e-05,
+      "loss": 2.4002,
+      "step": 6535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3795422613620758,
+      "learning_rate": 9.322853051387141e-05,
+      "loss": 2.3682,
+      "step": 6540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0328624248504639,
+      "learning_rate": 9.321825622058568e-05,
+      "loss": 2.4439,
+      "step": 6545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3691748082637787,
+      "learning_rate": 9.320797470569001e-05,
+      "loss": 2.3791,
+      "step": 6550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39962154626846313,
+      "learning_rate": 9.319768597090242e-05,
+      "loss": 2.4168,
+      "step": 6555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3808695673942566,
+      "learning_rate": 9.318739001794208e-05,
+      "loss": 2.4082,
+      "step": 6560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3802781105041504,
+      "learning_rate": 9.317708684852942e-05,
+      "loss": 2.4627,
+      "step": 6565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3847859799861908,
+      "learning_rate": 9.31667764643861e-05,
+      "loss": 2.3788,
+      "step": 6570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4047774076461792,
+      "learning_rate": 9.31564588672349e-05,
+      "loss": 2.296,
+      "step": 6575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38137468695640564,
+      "learning_rate": 9.314613405879986e-05,
+      "loss": 2.3322,
+      "step": 6580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39926302433013916,
+      "learning_rate": 9.313580204080623e-05,
+      "loss": 2.4401,
+      "step": 6585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3853113055229187,
+      "learning_rate": 9.312546281498047e-05,
+      "loss": 2.4124,
+      "step": 6590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37341800332069397,
+      "learning_rate": 9.311511638305018e-05,
+      "loss": 2.3842,
+      "step": 6595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38309094309806824,
+      "learning_rate": 9.310476274674425e-05,
+      "loss": 2.371,
+      "step": 6600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39156097173690796,
+      "learning_rate": 9.30944019077927e-05,
+      "loss": 2.4805,
+      "step": 6605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38827601075172424,
+      "learning_rate": 9.30840338679268e-05,
+      "loss": 2.445,
+      "step": 6610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38264748454093933,
+      "learning_rate": 9.307365862887905e-05,
+      "loss": 2.3485,
+      "step": 6615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3985002934932709,
+      "learning_rate": 9.306327619238306e-05,
+      "loss": 2.3971,
+      "step": 6620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3684330880641937,
+      "learning_rate": 9.305288656017372e-05,
+      "loss": 2.369,
+      "step": 6625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.396390438079834,
+      "learning_rate": 9.304248973398711e-05,
+      "loss": 2.3929,
+      "step": 6630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39240720868110657,
+      "learning_rate": 9.303208571556048e-05,
+      "loss": 2.324,
+      "step": 6635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5668953061103821,
+      "learning_rate": 9.30216745066323e-05,
+      "loss": 2.3826,
+      "step": 6640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38838455080986023,
+      "learning_rate": 9.301125610894226e-05,
+      "loss": 2.3891,
+      "step": 6645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39860889315605164,
+      "learning_rate": 9.300083052423124e-05,
+      "loss": 2.3569,
+      "step": 6650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38251256942749023,
+      "learning_rate": 9.29903977542413e-05,
+      "loss": 2.5157,
+      "step": 6655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4268370568752289,
+      "learning_rate": 9.297995780071571e-05,
+      "loss": 2.4678,
+      "step": 6660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37537509202957153,
+      "learning_rate": 9.296951066539899e-05,
+      "loss": 2.3955,
+      "step": 6665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40345779061317444,
+      "learning_rate": 9.295905635003677e-05,
+      "loss": 2.4013,
+      "step": 6670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38834914565086365,
+      "learning_rate": 9.294859485637595e-05,
+      "loss": 2.3756,
+      "step": 6675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40468353033065796,
+      "learning_rate": 9.29381261861646e-05,
+      "loss": 2.409,
+      "step": 6680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38255539536476135,
+      "learning_rate": 9.292765034115198e-05,
+      "loss": 2.3784,
+      "step": 6685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3684251010417938,
+      "learning_rate": 9.291716732308862e-05,
+      "loss": 2.3769,
+      "step": 6690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3653717339038849,
+      "learning_rate": 9.290667713372613e-05,
+      "loss": 2.3494,
+      "step": 6695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3789008557796478,
+      "learning_rate": 9.289617977481741e-05,
+      "loss": 2.4134,
+      "step": 6700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4086689352989197,
+      "learning_rate": 9.288567524811654e-05,
+      "loss": 2.3273,
+      "step": 6705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38342830538749695,
+      "learning_rate": 9.287516355537876e-05,
+      "loss": 2.3883,
+      "step": 6710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3849562406539917,
+      "learning_rate": 9.286464469836056e-05,
+      "loss": 2.4048,
+      "step": 6715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41361159086227417,
+      "learning_rate": 9.28541186788196e-05,
+      "loss": 2.424,
+      "step": 6720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38070642948150635,
+      "learning_rate": 9.284358549851471e-05,
+      "loss": 2.459,
+      "step": 6725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37537115812301636,
+      "learning_rate": 9.283304515920598e-05,
+      "loss": 2.3576,
+      "step": 6730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4081230163574219,
+      "learning_rate": 9.282249766265464e-05,
+      "loss": 2.3705,
+      "step": 6735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37754902243614197,
+      "learning_rate": 9.281194301062315e-05,
+      "loss": 2.3637,
+      "step": 6740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3944016098976135,
+      "learning_rate": 9.280138120487514e-05,
+      "loss": 2.3114,
+      "step": 6745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3707534372806549,
+      "learning_rate": 9.279081224717547e-05,
+      "loss": 2.4297,
+      "step": 6750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39825552701950073,
+      "learning_rate": 9.278023613929014e-05,
+      "loss": 2.4556,
+      "step": 6755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3916734457015991,
+      "learning_rate": 9.276965288298642e-05,
+      "loss": 2.4062,
+      "step": 6760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42393219470977783,
+      "learning_rate": 9.27590624800327e-05,
+      "loss": 2.3912,
+      "step": 6765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41341716051101685,
+      "learning_rate": 9.27484649321986e-05,
+      "loss": 2.3485,
+      "step": 6770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3912891149520874,
+      "learning_rate": 9.273786024125496e-05,
+      "loss": 2.4043,
+      "step": 6775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38186100125312805,
+      "learning_rate": 9.272724840897375e-05,
+      "loss": 2.4503,
+      "step": 6780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3757629096508026,
+      "learning_rate": 9.271662943712818e-05,
+      "loss": 2.372,
+      "step": 6785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3694356679916382,
+      "learning_rate": 9.270600332749264e-05,
+      "loss": 2.4195,
+      "step": 6790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41043514013290405,
+      "learning_rate": 9.269537008184273e-05,
+      "loss": 2.4258,
+      "step": 6795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44519883394241333,
+      "learning_rate": 9.268472970195519e-05,
+      "loss": 2.3653,
+      "step": 6800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38669469952583313,
+      "learning_rate": 9.267408218960803e-05,
+      "loss": 2.2797,
+      "step": 6805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37843766808509827,
+      "learning_rate": 9.266342754658036e-05,
+      "loss": 2.4337,
+      "step": 6810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3688774108886719,
+      "learning_rate": 9.265276577465259e-05,
+      "loss": 2.4109,
+      "step": 6815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3883999288082123,
+      "learning_rate": 9.264209687560621e-05,
+      "loss": 2.3808,
+      "step": 6820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37799063324928284,
+      "learning_rate": 9.263142085122399e-05,
+      "loss": 2.3046,
+      "step": 6825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3867829442024231,
+      "learning_rate": 9.262073770328983e-05,
+      "loss": 2.3697,
+      "step": 6830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37894579768180847,
+      "learning_rate": 9.261004743358886e-05,
+      "loss": 2.376,
+      "step": 6835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3978067934513092,
+      "learning_rate": 9.259935004390737e-05,
+      "loss": 2.4565,
+      "step": 6840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37866368889808655,
+      "learning_rate": 9.258864553603286e-05,
+      "loss": 2.3883,
+      "step": 6845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39498814940452576,
+      "learning_rate": 9.257793391175402e-05,
+      "loss": 2.3269,
+      "step": 6850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3729490041732788,
+      "learning_rate": 9.256721517286073e-05,
+      "loss": 2.4256,
+      "step": 6855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37837979197502136,
+      "learning_rate": 9.255648932114403e-05,
+      "loss": 2.3757,
+      "step": 6860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.383259654045105,
+      "learning_rate": 9.254575635839618e-05,
+      "loss": 2.2894,
+      "step": 6865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4073196351528168,
+      "learning_rate": 9.253501628641061e-05,
+      "loss": 2.3384,
+      "step": 6870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3790021240711212,
+      "learning_rate": 9.252426910698198e-05,
+      "loss": 2.3198,
+      "step": 6875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41724133491516113,
+      "learning_rate": 9.251351482190604e-05,
+      "loss": 2.3207,
+      "step": 6880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39029765129089355,
+      "learning_rate": 9.250275343297985e-05,
+      "loss": 2.4101,
+      "step": 6885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40110868215560913,
+      "learning_rate": 9.249198494200158e-05,
+      "loss": 2.4675,
+      "step": 6890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37624987959861755,
+      "learning_rate": 9.24812093507706e-05,
+      "loss": 2.3833,
+      "step": 6895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3729467988014221,
+      "learning_rate": 9.247042666108747e-05,
+      "loss": 2.3511,
+      "step": 6900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37379905581474304,
+      "learning_rate": 9.245963687475395e-05,
+      "loss": 2.3275,
+      "step": 6905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41176095604896545,
+      "learning_rate": 9.244883999357293e-05,
+      "loss": 2.3726,
+      "step": 6910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3875267803668976,
+      "learning_rate": 9.24380360193486e-05,
+      "loss": 2.3593,
+      "step": 6915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3936648368835449,
+      "learning_rate": 9.242722495388621e-05,
+      "loss": 2.3666,
+      "step": 6920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40441811084747314,
+      "learning_rate": 9.241640679899226e-05,
+      "loss": 2.4289,
+      "step": 6925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40688925981521606,
+      "learning_rate": 9.240558155647443e-05,
+      "loss": 2.3847,
+      "step": 6930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4118880331516266,
+      "learning_rate": 9.239474922814157e-05,
+      "loss": 2.4386,
+      "step": 6935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39184510707855225,
+      "learning_rate": 9.238390981580375e-05,
+      "loss": 2.4162,
+      "step": 6940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39151880145072937,
+      "learning_rate": 9.237306332127215e-05,
+      "loss": 2.462,
+      "step": 6945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3826306164264679,
+      "learning_rate": 9.236220974635923e-05,
+      "loss": 2.3723,
+      "step": 6950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.390306293964386,
+      "learning_rate": 9.235134909287853e-05,
+      "loss": 2.336,
+      "step": 6955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3989279568195343,
+      "learning_rate": 9.234048136264485e-05,
+      "loss": 2.376,
+      "step": 6960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37056881189346313,
+      "learning_rate": 9.232960655747416e-05,
+      "loss": 2.3802,
+      "step": 6965
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41524413228034973,
+      "learning_rate": 9.231872467918358e-05,
+      "loss": 2.3503,
+      "step": 6970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4049029052257538,
+      "learning_rate": 9.230783572959143e-05,
+      "loss": 2.3997,
+      "step": 6975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38071203231811523,
+      "learning_rate": 9.229693971051723e-05,
+      "loss": 2.321,
+      "step": 6980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3961276113986969,
+      "learning_rate": 9.228603662378166e-05,
+      "loss": 2.4306,
+      "step": 6985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3823302388191223,
+      "learning_rate": 9.227512647120656e-05,
+      "loss": 2.3923,
+      "step": 6990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3967801630496979,
+      "learning_rate": 9.2264209254615e-05,
+      "loss": 2.4169,
+      "step": 6995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3876556158065796,
+      "learning_rate": 9.225328497583122e-05,
+      "loss": 2.3606,
+      "step": 7000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4217390716075897,
+      "learning_rate": 9.224235363668058e-05,
+      "loss": 2.2739,
+      "step": 7005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41772782802581787,
+      "learning_rate": 9.223141523898972e-05,
+      "loss": 2.378,
+      "step": 7010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3794209659099579,
+      "learning_rate": 9.222046978458639e-05,
+      "loss": 2.3878,
+      "step": 7015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3847315311431885,
+      "learning_rate": 9.22095172752995e-05,
+      "loss": 2.3823,
+      "step": 7020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3783474862575531,
+      "learning_rate": 9.21985577129592e-05,
+      "loss": 2.4179,
+      "step": 7025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3810170888900757,
+      "learning_rate": 9.218759109939681e-05,
+      "loss": 2.4027,
+      "step": 7030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3804612159729004,
+      "learning_rate": 9.217661743644477e-05,
+      "loss": 2.366,
+      "step": 7035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4066917300224304,
+      "learning_rate": 9.216563672593677e-05,
+      "loss": 2.3829,
+      "step": 7040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37400779128074646,
+      "learning_rate": 9.215464896970763e-05,
+      "loss": 2.3257,
+      "step": 7045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3796880841255188,
+      "learning_rate": 9.214365416959336e-05,
+      "loss": 2.3125,
+      "step": 7050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38020390272140503,
+      "learning_rate": 9.213265232743119e-05,
+      "loss": 2.3959,
+      "step": 7055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38993194699287415,
+      "learning_rate": 9.212164344505942e-05,
+      "loss": 2.4311,
+      "step": 7060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37751904129981995,
+      "learning_rate": 9.211062752431765e-05,
+      "loss": 2.3832,
+      "step": 7065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39901599287986755,
+      "learning_rate": 9.209960456704656e-05,
+      "loss": 2.4727,
+      "step": 7070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3747711777687073,
+      "learning_rate": 9.208857457508809e-05,
+      "loss": 2.4089,
+      "step": 7075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.419518381357193,
+      "learning_rate": 9.207753755028526e-05,
+      "loss": 2.3393,
+      "step": 7080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36917147040367126,
+      "learning_rate": 9.206649349448235e-05,
+      "loss": 2.4157,
+      "step": 7085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3834037184715271,
+      "learning_rate": 9.205544240952477e-05,
+      "loss": 2.386,
+      "step": 7090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3780665695667267,
+      "learning_rate": 9.204438429725911e-05,
+      "loss": 2.4698,
+      "step": 7095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3808090388774872,
+      "learning_rate": 9.203331915953314e-05,
+      "loss": 2.3706,
+      "step": 7100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38215988874435425,
+      "learning_rate": 9.202224699819584e-05,
+      "loss": 2.3501,
+      "step": 7105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37141138315200806,
+      "learning_rate": 9.201116781509726e-05,
+      "loss": 2.4176,
+      "step": 7110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.405146062374115,
+      "learning_rate": 9.200008161208875e-05,
+      "loss": 2.3103,
+      "step": 7115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3942197561264038,
+      "learning_rate": 9.198898839102273e-05,
+      "loss": 2.402,
+      "step": 7120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.372393935918808,
+      "learning_rate": 9.197788815375287e-05,
+      "loss": 2.3152,
+      "step": 7125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3773491382598877,
+      "learning_rate": 9.196678090213395e-05,
+      "loss": 2.4189,
+      "step": 7130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3947962820529938,
+      "learning_rate": 9.195566663802198e-05,
+      "loss": 2.4057,
+      "step": 7135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.371697336435318,
+      "learning_rate": 9.194454536327408e-05,
+      "loss": 2.3023,
+      "step": 7140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3776797652244568,
+      "learning_rate": 9.19334170797486e-05,
+      "loss": 2.3371,
+      "step": 7145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38953882455825806,
+      "learning_rate": 9.192228178930503e-05,
+      "loss": 2.4074,
+      "step": 7150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38433557748794556,
+      "learning_rate": 9.191113949380402e-05,
+      "loss": 2.3481,
+      "step": 7155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37438058853149414,
+      "learning_rate": 9.189999019510742e-05,
+      "loss": 2.3468,
+      "step": 7160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40474867820739746,
+      "learning_rate": 9.188883389507823e-05,
+      "loss": 2.3475,
+      "step": 7165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3682725727558136,
+      "learning_rate": 9.187767059558064e-05,
+      "loss": 2.3811,
+      "step": 7170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5979679822921753,
+      "learning_rate": 9.186650029847998e-05,
+      "loss": 2.351,
+      "step": 7175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4136204719543457,
+      "learning_rate": 9.185532300564276e-05,
+      "loss": 2.4076,
+      "step": 7180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38111093640327454,
+      "learning_rate": 9.184413871893668e-05,
+      "loss": 2.3769,
+      "step": 7185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38824647665023804,
+      "learning_rate": 9.183294744023058e-05,
+      "loss": 2.449,
+      "step": 7190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3786308765411377,
+      "learning_rate": 9.18217491713945e-05,
+      "loss": 2.363,
+      "step": 7195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3908441960811615,
+      "learning_rate": 9.181054391429962e-05,
+      "loss": 2.3325,
+      "step": 7200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3615347146987915,
+      "learning_rate": 9.179933167081828e-05,
+      "loss": 2.3556,
+      "step": 7205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37007030844688416,
+      "learning_rate": 9.178811244282404e-05,
+      "loss": 2.3446,
+      "step": 7210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3700975477695465,
+      "learning_rate": 9.177688623219154e-05,
+      "loss": 2.4459,
+      "step": 7215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3784796893596649,
+      "learning_rate": 9.176565304079669e-05,
+      "loss": 2.3231,
+      "step": 7220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39437800645828247,
+      "learning_rate": 9.175441287051648e-05,
+      "loss": 2.3456,
+      "step": 7225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43581029772758484,
+      "learning_rate": 9.174316572322912e-05,
+      "loss": 2.3869,
+      "step": 7230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.399662584066391,
+      "learning_rate": 9.173191160081395e-05,
+      "loss": 2.2628,
+      "step": 7235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3843769431114197,
+      "learning_rate": 9.17206505051515e-05,
+      "loss": 2.3747,
+      "step": 7240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4066675901412964,
+      "learning_rate": 9.170938243812349e-05,
+      "loss": 2.3821,
+      "step": 7245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38300344347953796,
+      "learning_rate": 9.16981074016127e-05,
+      "loss": 2.4028,
+      "step": 7250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37688159942626953,
+      "learning_rate": 9.16868253975032e-05,
+      "loss": 2.3676,
+      "step": 7255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.405192106962204,
+      "learning_rate": 9.167553642768017e-05,
+      "loss": 2.4095,
+      "step": 7260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3879852890968323,
+      "learning_rate": 9.166424049402993e-05,
+      "loss": 2.3287,
+      "step": 7265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3803970217704773,
+      "learning_rate": 9.165293759844e-05,
+      "loss": 2.3763,
+      "step": 7270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3930979073047638,
+      "learning_rate": 9.164162774279905e-05,
+      "loss": 2.3598,
+      "step": 7275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4087534248828888,
+      "learning_rate": 9.163031092899693e-05,
+      "loss": 2.484,
+      "step": 7280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3704996407032013,
+      "learning_rate": 9.161898715892463e-05,
+      "loss": 2.3571,
+      "step": 7285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3666727542877197,
+      "learning_rate": 9.160765643447429e-05,
+      "loss": 2.4284,
+      "step": 7290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3848545253276825,
+      "learning_rate": 9.159631875753925e-05,
+      "loss": 2.4555,
+      "step": 7295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38789573311805725,
+      "learning_rate": 9.1584974130014e-05,
+      "loss": 2.3403,
+      "step": 7300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4025851786136627,
+      "learning_rate": 9.157362255379419e-05,
+      "loss": 2.3946,
+      "step": 7305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37643733620643616,
+      "learning_rate": 9.156226403077657e-05,
+      "loss": 2.3521,
+      "step": 7310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3807030916213989,
+      "learning_rate": 9.155089856285917e-05,
+      "loss": 2.3396,
+      "step": 7315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3814091980457306,
+      "learning_rate": 9.153952615194111e-05,
+      "loss": 2.4389,
+      "step": 7320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3711268901824951,
+      "learning_rate": 9.152814679992265e-05,
+      "loss": 2.3985,
+      "step": 7325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37775880098342896,
+      "learning_rate": 9.151676050870526e-05,
+      "loss": 2.3456,
+      "step": 7330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38292786478996277,
+      "learning_rate": 9.150536728019154e-05,
+      "loss": 2.459,
+      "step": 7335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3671378493309021,
+      "learning_rate": 9.149396711628523e-05,
+      "loss": 2.4025,
+      "step": 7340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4122884273529053,
+      "learning_rate": 9.14825600188913e-05,
+      "loss": 2.2827,
+      "step": 7345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3597871959209442,
+      "learning_rate": 9.147114598991583e-05,
+      "loss": 2.3434,
+      "step": 7350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3855368494987488,
+      "learning_rate": 9.145972503126603e-05,
+      "loss": 2.4095,
+      "step": 7355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3860791325569153,
+      "learning_rate": 9.14482971448503e-05,
+      "loss": 2.3006,
+      "step": 7360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3844926953315735,
+      "learning_rate": 9.143686233257824e-05,
+      "loss": 2.3494,
+      "step": 7365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3981086015701294,
+      "learning_rate": 9.142542059636053e-05,
+      "loss": 2.3484,
+      "step": 7370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38726404309272766,
+      "learning_rate": 9.141397193810903e-05,
+      "loss": 2.3237,
+      "step": 7375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3799365162849426,
+      "learning_rate": 9.14025163597368e-05,
+      "loss": 2.3584,
+      "step": 7380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3762057423591614,
+      "learning_rate": 9.1391053863158e-05,
+      "loss": 2.4078,
+      "step": 7385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3738326132297516,
+      "learning_rate": 9.137958445028799e-05,
+      "loss": 2.4267,
+      "step": 7390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4022557735443115,
+      "learning_rate": 9.136810812304326e-05,
+      "loss": 2.3933,
+      "step": 7395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3797825276851654,
+      "learning_rate": 9.135662488334147e-05,
+      "loss": 2.3838,
+      "step": 7400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38085460662841797,
+      "learning_rate": 9.13451347331014e-05,
+      "loss": 2.2891,
+      "step": 7405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38730528950691223,
+      "learning_rate": 9.133363767424303e-05,
+      "loss": 2.3946,
+      "step": 7410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38261866569519043,
+      "learning_rate": 9.132213370868748e-05,
+      "loss": 2.371,
+      "step": 7415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42055457830429077,
+      "learning_rate": 9.131062283835701e-05,
+      "loss": 2.4203,
+      "step": 7420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38609468936920166,
+      "learning_rate": 9.129910506517503e-05,
+      "loss": 2.4278,
+      "step": 7425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3772518038749695,
+      "learning_rate": 9.128758039106616e-05,
+      "loss": 2.4,
+      "step": 7430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.385397344827652,
+      "learning_rate": 9.127604881795609e-05,
+      "loss": 2.3922,
+      "step": 7435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37878987193107605,
+      "learning_rate": 9.126451034777173e-05,
+      "loss": 2.3527,
+      "step": 7440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39951297640800476,
+      "learning_rate": 9.12529649824411e-05,
+      "loss": 2.336,
+      "step": 7445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4087267220020294,
+      "learning_rate": 9.124141272389337e-05,
+      "loss": 2.3559,
+      "step": 7450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38972747325897217,
+      "learning_rate": 9.122985357405892e-05,
+      "loss": 2.399,
+      "step": 7455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37987199425697327,
+      "learning_rate": 9.121828753486921e-05,
+      "loss": 2.4099,
+      "step": 7460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44305419921875,
+      "learning_rate": 9.120671460825691e-05,
+      "loss": 2.3934,
+      "step": 7465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39322808384895325,
+      "learning_rate": 9.119513479615577e-05,
+      "loss": 2.3941,
+      "step": 7470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38651329278945923,
+      "learning_rate": 9.118354810050079e-05,
+      "loss": 2.3759,
+      "step": 7475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40869197249412537,
+      "learning_rate": 9.117195452322801e-05,
+      "loss": 2.3564,
+      "step": 7480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37995222210884094,
+      "learning_rate": 9.11603540662747e-05,
+      "loss": 2.3702,
+      "step": 7485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38796737790107727,
+      "learning_rate": 9.114874673157927e-05,
+      "loss": 2.3316,
+      "step": 7490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3650217354297638,
+      "learning_rate": 9.113713252108123e-05,
+      "loss": 2.3441,
+      "step": 7495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37982380390167236,
+      "learning_rate": 9.112551143672127e-05,
+      "loss": 2.2421,
+      "step": 7500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3857923150062561,
+      "learning_rate": 9.111388348044125e-05,
+      "loss": 2.3858,
+      "step": 7505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3823206424713135,
+      "learning_rate": 9.110224865418417e-05,
+      "loss": 2.4058,
+      "step": 7510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38922521471977234,
+      "learning_rate": 9.109060695989413e-05,
+      "loss": 2.3641,
+      "step": 7515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3994320034980774,
+      "learning_rate": 9.107895839951643e-05,
+      "loss": 2.3895,
+      "step": 7520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38006478548049927,
+      "learning_rate": 9.106730297499752e-05,
+      "loss": 2.3997,
+      "step": 7525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3826669454574585,
+      "learning_rate": 9.105564068828497e-05,
+      "loss": 2.3525,
+      "step": 7530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38146936893463135,
+      "learning_rate": 9.104397154132747e-05,
+      "loss": 2.4408,
+      "step": 7535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3917499780654907,
+      "learning_rate": 9.103229553607493e-05,
+      "loss": 2.4484,
+      "step": 7540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4011600911617279,
+      "learning_rate": 9.102061267447835e-05,
+      "loss": 2.3566,
+      "step": 7545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3795270621776581,
+      "learning_rate": 9.100892295848989e-05,
+      "loss": 2.3059,
+      "step": 7550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3861050307750702,
+      "learning_rate": 9.099722639006287e-05,
+      "loss": 2.3613,
+      "step": 7555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.366303414106369,
+      "learning_rate": 9.098552297115175e-05,
+      "loss": 2.3768,
+      "step": 7560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39337867498397827,
+      "learning_rate": 9.097381270371212e-05,
+      "loss": 2.3235,
+      "step": 7565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37198737263679504,
+      "learning_rate": 9.09620955897007e-05,
+      "loss": 2.3738,
+      "step": 7570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37840840220451355,
+      "learning_rate": 9.09503716310754e-05,
+      "loss": 2.3139,
+      "step": 7575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4021204113960266,
+      "learning_rate": 9.093864082979525e-05,
+      "loss": 2.3262,
+      "step": 7580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3721044361591339,
+      "learning_rate": 9.092690318782042e-05,
+      "loss": 2.36,
+      "step": 7585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37759456038475037,
+      "learning_rate": 9.091515870711221e-05,
+      "loss": 2.3842,
+      "step": 7590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36816081404685974,
+      "learning_rate": 9.09034073896331e-05,
+      "loss": 2.4037,
+      "step": 7595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3743402063846588,
+      "learning_rate": 9.08916492373467e-05,
+      "loss": 2.3444,
+      "step": 7600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3930553197860718,
+      "learning_rate": 9.087988425221772e-05,
+      "loss": 2.3207,
+      "step": 7605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38041192293167114,
+      "learning_rate": 9.086811243621206e-05,
+      "loss": 2.2864,
+      "step": 7610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3887983560562134,
+      "learning_rate": 9.085633379129677e-05,
+      "loss": 2.3253,
+      "step": 7615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4027004837989807,
+      "learning_rate": 9.084454831943998e-05,
+      "loss": 2.3277,
+      "step": 7620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3974758982658386,
+      "learning_rate": 9.083275602261101e-05,
+      "loss": 2.3536,
+      "step": 7625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39558038115501404,
+      "learning_rate": 9.082095690278034e-05,
+      "loss": 2.3537,
+      "step": 7630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36775341629981995,
+      "learning_rate": 9.080915096191951e-05,
+      "loss": 2.4182,
+      "step": 7635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4096049666404724,
+      "learning_rate": 9.07973382020013e-05,
+      "loss": 2.3812,
+      "step": 7640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3859267234802246,
+      "learning_rate": 9.078551862499953e-05,
+      "loss": 2.411,
+      "step": 7645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3925534188747406,
+      "learning_rate": 9.077369223288923e-05,
+      "loss": 2.3249,
+      "step": 7650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38505223393440247,
+      "learning_rate": 9.076185902764656e-05,
+      "loss": 2.3677,
+      "step": 7655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39727380871772766,
+      "learning_rate": 9.075001901124878e-05,
+      "loss": 2.3515,
+      "step": 7660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.374025821685791,
+      "learning_rate": 9.073817218567431e-05,
+      "loss": 2.3236,
+      "step": 7665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3736701011657715,
+      "learning_rate": 9.072631855290275e-05,
+      "loss": 2.2798,
+      "step": 7670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3726747930049896,
+      "learning_rate": 9.071445811491475e-05,
+      "loss": 2.3168,
+      "step": 7675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3843863010406494,
+      "learning_rate": 9.070259087369218e-05,
+      "loss": 2.3395,
+      "step": 7680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.42971405386924744,
+      "learning_rate": 9.069071683121799e-05,
+      "loss": 2.4447,
+      "step": 7685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37763288617134094,
+      "learning_rate": 9.067883598947628e-05,
+      "loss": 2.3498,
+      "step": 7690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37887459993362427,
+      "learning_rate": 9.066694835045233e-05,
+      "loss": 2.4201,
+      "step": 7695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3693830072879791,
+      "learning_rate": 9.06550539161325e-05,
+      "loss": 2.4084,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39116841554641724,
+      "learning_rate": 9.06431526885043e-05,
+      "loss": 2.2941,
+      "step": 7705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38062620162963867,
+      "learning_rate": 9.063124466955639e-05,
+      "loss": 2.3548,
+      "step": 7710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.360323429107666,
+      "learning_rate": 9.061932986127857e-05,
+      "loss": 2.3898,
+      "step": 7715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4148135185241699,
+      "learning_rate": 9.060740826566172e-05,
+      "loss": 2.3463,
+      "step": 7720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.42293551564216614,
+      "learning_rate": 9.059547988469792e-05,
+      "loss": 2.3496,
+      "step": 7725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37806662917137146,
+      "learning_rate": 9.058354472038037e-05,
+      "loss": 2.3759,
+      "step": 7730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37789738178253174,
+      "learning_rate": 9.057160277470337e-05,
+      "loss": 2.4303,
+      "step": 7735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38087910413742065,
+      "learning_rate": 9.055965404966238e-05,
+      "loss": 2.4412,
+      "step": 7740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3670419454574585,
+      "learning_rate": 9.054769854725399e-05,
+      "loss": 2.2852,
+      "step": 7745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3937246799468994,
+      "learning_rate": 9.053573626947593e-05,
+      "loss": 2.4295,
+      "step": 7750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3743956387042999,
+      "learning_rate": 9.052376721832702e-05,
+      "loss": 2.3669,
+      "step": 7755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36857497692108154,
+      "learning_rate": 9.051179139580727e-05,
+      "loss": 2.3225,
+      "step": 7760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37284955382347107,
+      "learning_rate": 9.049980880391782e-05,
+      "loss": 2.4075,
+      "step": 7765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4266446530818939,
+      "learning_rate": 9.048781944466087e-05,
+      "loss": 2.3076,
+      "step": 7770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3883683681488037,
+      "learning_rate": 9.04758233200398e-05,
+      "loss": 2.3558,
+      "step": 7775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38533520698547363,
+      "learning_rate": 9.046382043205914e-05,
+      "loss": 2.4108,
+      "step": 7780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3737938106060028,
+      "learning_rate": 9.045181078272453e-05,
+      "loss": 2.2961,
+      "step": 7785
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39717981219291687,
+      "learning_rate": 9.043979437404273e-05,
+      "loss": 2.4455,
+      "step": 7790
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3694390058517456,
+      "learning_rate": 9.042777120802164e-05,
+      "loss": 2.4223,
+      "step": 7795
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39764177799224854,
+      "learning_rate": 9.041574128667027e-05,
+      "loss": 2.3735,
+      "step": 7800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3743802011013031,
+      "learning_rate": 9.040370461199879e-05,
+      "loss": 2.3242,
+      "step": 7805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37303426861763,
+      "learning_rate": 9.039166118601849e-05,
+      "loss": 2.3624,
+      "step": 7810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3795291781425476,
+      "learning_rate": 9.037961101074175e-05,
+      "loss": 2.3948,
+      "step": 7815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3655077815055847,
+      "learning_rate": 9.036755408818217e-05,
+      "loss": 2.4246,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3666858971118927,
+      "learning_rate": 9.035549042035434e-05,
+      "loss": 2.3954,
+      "step": 7825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38421571254730225,
+      "learning_rate": 9.03434200092741e-05,
+      "loss": 2.3892,
+      "step": 7830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36218738555908203,
+      "learning_rate": 9.033134285695837e-05,
+      "loss": 2.3104,
+      "step": 7835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4026510715484619,
+      "learning_rate": 9.031925896542521e-05,
+      "loss": 2.3246,
+      "step": 7840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45338383316993713,
+      "learning_rate": 9.030716833669375e-05,
+      "loss": 2.3687,
+      "step": 7845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35163572430610657,
+      "learning_rate": 9.029507097278433e-05,
+      "loss": 2.3903,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38969340920448303,
+      "learning_rate": 9.028296687571834e-05,
+      "loss": 2.4251,
+      "step": 7855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3626038134098053,
+      "learning_rate": 9.027085604751835e-05,
+      "loss": 2.3941,
+      "step": 7860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3883587419986725,
+      "learning_rate": 9.025873849020805e-05,
+      "loss": 2.3712,
+      "step": 7865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.382510781288147,
+      "learning_rate": 9.024661420581222e-05,
+      "loss": 2.3955,
+      "step": 7870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39293593168258667,
+      "learning_rate": 9.023448319635676e-05,
+      "loss": 2.406,
+      "step": 7875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36774471402168274,
+      "learning_rate": 9.022234546386877e-05,
+      "loss": 2.336,
+      "step": 7880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3796512484550476,
+      "learning_rate": 9.021020101037637e-05,
+      "loss": 2.4263,
+      "step": 7885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39686405658721924,
+      "learning_rate": 9.019804983790888e-05,
+      "loss": 2.362,
+      "step": 7890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4062720537185669,
+      "learning_rate": 9.018589194849672e-05,
+      "loss": 2.3904,
+      "step": 7895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39817386865615845,
+      "learning_rate": 9.01737273441714e-05,
+      "loss": 2.3621,
+      "step": 7900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3820001482963562,
+      "learning_rate": 9.016155602696559e-05,
+      "loss": 2.4052,
+      "step": 7905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.390395849943161,
+      "learning_rate": 9.01493779989131e-05,
+      "loss": 2.3226,
+      "step": 7910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38826292753219604,
+      "learning_rate": 9.013719326204881e-05,
+      "loss": 2.3019,
+      "step": 7915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38358476758003235,
+      "learning_rate": 9.012500181840874e-05,
+      "loss": 2.4235,
+      "step": 7920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36424243450164795,
+      "learning_rate": 9.011280367003003e-05,
+      "loss": 2.3882,
+      "step": 7925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3870227336883545,
+      "learning_rate": 9.010059881895099e-05,
+      "loss": 2.3982,
+      "step": 7930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37909358739852905,
+      "learning_rate": 9.008838726721094e-05,
+      "loss": 2.4136,
+      "step": 7935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.404336154460907,
+      "learning_rate": 9.007616901685042e-05,
+      "loss": 2.326,
+      "step": 7940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40206748247146606,
+      "learning_rate": 9.006394406991108e-05,
+      "loss": 2.353,
+      "step": 7945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39102137088775635,
+      "learning_rate": 9.00517124284356e-05,
+      "loss": 2.3948,
+      "step": 7950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3557710349559784,
+      "learning_rate": 9.003947409446789e-05,
+      "loss": 2.3664,
+      "step": 7955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36864960193634033,
+      "learning_rate": 9.002722907005292e-05,
+      "loss": 2.3953,
+      "step": 7960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3789011240005493,
+      "learning_rate": 9.001497735723679e-05,
+      "loss": 2.4224,
+      "step": 7965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37259918451309204,
+      "learning_rate": 9.000271895806669e-05,
+      "loss": 2.4117,
+      "step": 7970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4033454358577728,
+      "learning_rate": 8.9990453874591e-05,
+      "loss": 2.4163,
+      "step": 7975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3902803659439087,
+      "learning_rate": 8.997818210885912e-05,
+      "loss": 2.3752,
+      "step": 7980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3967965841293335,
+      "learning_rate": 8.996590366292164e-05,
+      "loss": 2.3203,
+      "step": 7985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38087499141693115,
+      "learning_rate": 8.995361853883026e-05,
+      "loss": 2.3007,
+      "step": 7990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3624536991119385,
+      "learning_rate": 8.994132673863774e-05,
+      "loss": 2.3704,
+      "step": 7995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3757088780403137,
+      "learning_rate": 8.992902826439802e-05,
+      "loss": 2.3685,
+      "step": 8000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 38527,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 2.126010100010189e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}