{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.99947367395076,
  "eval_steps": 500,
  "global_step": 28854,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1.4185574054718018,
      "learning_rate": 1.9999998518172358e-05,
      "loss": 0.3131,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.2674533128738403,
      "learning_rate": 1.9999994072689862e-05,
      "loss": 0.1822,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.4997848272323608,
      "learning_rate": 1.999998666355383e-05,
      "loss": 0.2147,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.265012502670288,
      "learning_rate": 1.9999976290766464e-05,
      "loss": 0.2305,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.29109525680542,
      "learning_rate": 1.9999962954330833e-05,
      "loss": 0.2123,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.475464940071106,
      "learning_rate": 1.9999946654250892e-05,
      "loss": 0.1856,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.054665207862854,
      "learning_rate": 1.9999927390531468e-05,
      "loss": 0.1843,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.248856544494629,
      "learning_rate": 1.9999905163178273e-05,
      "loss": 0.1667,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.092337965965271,
      "learning_rate": 1.9999879972197895e-05,
      "loss": 0.1828,
      "step": 45
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.166669249534607,
      "learning_rate": 1.99998518175978e-05,
      "loss": 0.1984,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5460426807403564,
      "learning_rate": 1.999982069938633e-05,
      "loss": 0.1694,
      "step": 55
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0065215826034546,
      "learning_rate": 1.999978661757271e-05,
      "loss": 0.1549,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2282688617706299,
      "learning_rate": 1.9999749572167037e-05,
      "loss": 0.1621,
      "step": 65
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9245144724845886,
      "learning_rate": 1.9999709563180293e-05,
      "loss": 0.1693,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2537037134170532,
      "learning_rate": 1.9999666590624332e-05,
      "loss": 0.2242,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1063697338104248,
      "learning_rate": 1.9999620654511895e-05,
      "loss": 0.1641,
      "step": 80
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0977532863616943,
      "learning_rate": 1.9999571754856594e-05,
      "loss": 0.2038,
      "step": 85
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.041774034500122,
      "learning_rate": 1.9999519891672918e-05,
      "loss": 0.2041,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9609395861625671,
      "learning_rate": 1.9999465064976234e-05,
      "loss": 0.1759,
      "step": 95
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7936742305755615,
      "learning_rate": 1.9999407274782804e-05,
      "loss": 0.1415,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1698766946792603,
      "learning_rate": 1.9999346521109742e-05,
      "loss": 0.1801,
      "step": 105
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1854974031448364,
      "learning_rate": 1.9999282803975058e-05,
      "loss": 0.1927,
      "step": 110
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0320738554000854,
      "learning_rate": 1.999921612339764e-05,
      "loss": 0.1966,
      "step": 115
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8534027338027954,
      "learning_rate": 1.9999146479397245e-05,
      "loss": 0.1739,
      "step": 120
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8952992558479309,
      "learning_rate": 1.9999073871994513e-05,
      "loss": 0.151,
      "step": 125
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.941814661026001,
      "learning_rate": 1.9998998301210965e-05,
      "loss": 0.1678,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2861160039901733,
      "learning_rate": 1.9998919767068993e-05,
      "loss": 0.2027,
      "step": 135
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0701113939285278,
      "learning_rate": 1.9998838269591877e-05,
      "loss": 0.2,
      "step": 140
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.2391180992126465,
      "learning_rate": 1.9998753808803766e-05,
      "loss": 0.1855,
      "step": 145
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7739980220794678,
      "learning_rate": 1.9998666384729692e-05,
      "loss": 0.1495,
      "step": 150
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0521777868270874,
      "learning_rate": 1.9998575997395566e-05,
      "loss": 0.1829,
      "step": 155
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0798540115356445,
      "learning_rate": 1.9998482646828178e-05,
      "loss": 0.1714,
      "step": 160
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8348418474197388,
      "learning_rate": 1.9998386333055188e-05,
      "loss": 0.135,
      "step": 165
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9550169706344604,
      "learning_rate": 1.999828705610514e-05,
      "loss": 0.1329,
      "step": 170
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1403224468231201,
      "learning_rate": 1.9998184816007466e-05,
      "loss": 0.1536,
      "step": 175
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7705615758895874,
      "learning_rate": 1.9998079612792457e-05,
      "loss": 0.152,
      "step": 180
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1348049640655518,
      "learning_rate": 1.9997971446491295e-05,
      "loss": 0.183,
      "step": 185
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1489228010177612,
      "learning_rate": 1.9997860317136035e-05,
      "loss": 0.2018,
      "step": 190
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8085460662841797,
      "learning_rate": 1.9997746224759612e-05,
      "loss": 0.1793,
      "step": 195
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7113868594169617,
      "learning_rate": 1.999762916939584e-05,
      "loss": 0.1468,
      "step": 200
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.2538015842437744,
      "learning_rate": 1.9997509151079413e-05,
      "loss": 0.1668,
      "step": 205
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0342391729354858,
      "learning_rate": 1.9997386169845896e-05,
      "loss": 0.1892,
      "step": 210
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0485116243362427,
      "learning_rate": 1.999726022573174e-05,
      "loss": 0.1544,
      "step": 215
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0436828136444092,
      "learning_rate": 1.9997131318774267e-05,
      "loss": 0.2145,
      "step": 220
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9429926872253418,
      "learning_rate": 1.999699944901168e-05,
      "loss": 0.1212,
      "step": 225
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.61878901720047,
      "learning_rate": 1.9996864616483064e-05,
      "loss": 0.1253,
      "step": 230
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9540243148803711,
      "learning_rate": 1.999672682122838e-05,
      "loss": 0.1656,
      "step": 235
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9568746089935303,
      "learning_rate": 1.9996586063288458e-05,
      "loss": 0.1676,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7145164012908936,
      "learning_rate": 1.9996442342705023e-05,
      "loss": 0.1354,
      "step": 245
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8123164176940918,
      "learning_rate": 1.9996295659520663e-05,
      "loss": 0.1439,
      "step": 250
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9031246900558472,
      "learning_rate": 1.9996146013778853e-05,
      "loss": 0.1539,
      "step": 255
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7666079998016357,
      "learning_rate": 1.9995993405523943e-05,
      "loss": 0.137,
      "step": 260
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.2642663717269897,
      "learning_rate": 1.9995837834801155e-05,
      "loss": 0.1893,
      "step": 265
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8628743886947632,
      "learning_rate": 1.99956793016566e-05,
      "loss": 0.1766,
      "step": 270
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9469863772392273,
      "learning_rate": 1.9995517806137263e-05,
      "loss": 0.1306,
      "step": 275
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8292585611343384,
      "learning_rate": 1.9995353348291003e-05,
      "loss": 0.1885,
      "step": 280
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8164292573928833,
      "learning_rate": 1.999518592816656e-05,
      "loss": 0.1484,
      "step": 285
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9202820658683777,
      "learning_rate": 1.9995015545813552e-05,
      "loss": 0.1488,
      "step": 290
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8424696326255798,
      "learning_rate": 1.999484220128247e-05,
      "loss": 0.1545,
      "step": 295
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9172386527061462,
      "learning_rate": 1.9994665894624697e-05,
      "loss": 0.1279,
      "step": 300
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8169159889221191,
      "learning_rate": 1.9994486625892477e-05,
      "loss": 0.1475,
      "step": 305
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.902919352054596,
      "learning_rate": 1.9994304395138942e-05,
      "loss": 0.1662,
      "step": 310
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8908593654632568,
      "learning_rate": 1.99941192024181e-05,
      "loss": 0.1976,
      "step": 315
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8870375752449036,
      "learning_rate": 1.999393104778483e-05,
      "loss": 0.1145,
      "step": 320
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7934074401855469,
      "learning_rate": 1.9993739931294895e-05,
      "loss": 0.1481,
      "step": 325
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6324104070663452,
      "learning_rate": 1.9993545853004944e-05,
      "loss": 0.1619,
      "step": 330
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1757112741470337,
      "learning_rate": 1.9993348812972484e-05,
      "loss": 0.1661,
      "step": 335
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7598108649253845,
      "learning_rate": 1.999314881125592e-05,
      "loss": 0.1254,
      "step": 340
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7880079746246338,
      "learning_rate": 1.999294584791452e-05,
      "loss": 0.1378,
      "step": 345
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.9844062328338623,
      "learning_rate": 1.999273992300844e-05,
      "loss": 0.1837,
      "step": 350
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.672408938407898,
      "learning_rate": 1.9992531036598705e-05,
      "loss": 0.1289,
      "step": 355
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6497504711151123,
      "learning_rate": 1.9992319188747225e-05,
      "loss": 0.1444,
      "step": 360
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8867689371109009,
      "learning_rate": 1.999210437951678e-05,
      "loss": 0.1401,
      "step": 365
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.939816415309906,
      "learning_rate": 1.9991886608971037e-05,
      "loss": 0.1773,
      "step": 370
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9581828117370605,
      "learning_rate": 1.9991665877174528e-05,
      "loss": 0.164,
      "step": 375
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9168484210968018,
      "learning_rate": 1.999144218419268e-05,
      "loss": 0.1992,
      "step": 380
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9601644277572632,
      "learning_rate": 1.9991215530091782e-05,
      "loss": 0.1664,
      "step": 385
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1929875612258911,
      "learning_rate": 1.999098591493901e-05,
      "loss": 0.154,
      "step": 390
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0054327249526978,
      "learning_rate": 1.9990753338802408e-05,
      "loss": 0.139,
      "step": 395
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8742963075637817,
      "learning_rate": 1.999051780175091e-05,
      "loss": 0.1881,
      "step": 400
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6974807977676392,
      "learning_rate": 1.999027930385432e-05,
      "loss": 0.148,
      "step": 405
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7910093665122986,
      "learning_rate": 1.9990037845183313e-05,
      "loss": 0.136,
      "step": 410
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.8378175497055054,
      "learning_rate": 1.9989793425809463e-05,
      "loss": 0.1005,
      "step": 415
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7058895230293274,
      "learning_rate": 1.9989546045805195e-05,
      "loss": 0.1394,
      "step": 420
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6101320385932922,
      "learning_rate": 1.998929570524383e-05,
      "loss": 0.1416,
      "step": 425
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.8287069201469421,
      "learning_rate": 1.9989042404199558e-05,
      "loss": 0.1251,
      "step": 430
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8602848052978516,
      "learning_rate": 1.998878614274745e-05,
      "loss": 0.1405,
      "step": 435
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8005295991897583,
      "learning_rate": 1.998852692096345e-05,
      "loss": 0.1436,
      "step": 440
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1535545587539673,
      "learning_rate": 1.998826473892439e-05,
      "loss": 0.1675,
      "step": 445
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8452690839767456,
      "learning_rate": 1.998799959670796e-05,
      "loss": 0.1592,
      "step": 450
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7784842252731323,
      "learning_rate": 1.998773149439275e-05,
      "loss": 0.1079,
      "step": 455
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6018524765968323,
      "learning_rate": 1.9987460432058213e-05,
      "loss": 0.1295,
      "step": 460
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.880659282207489,
      "learning_rate": 1.998718640978468e-05,
      "loss": 0.1467,
      "step": 465
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8606876134872437,
      "learning_rate": 1.998690942765336e-05,
      "loss": 0.1585,
      "step": 470
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.73162841796875,
      "learning_rate": 1.998662948574635e-05,
      "loss": 0.1589,
      "step": 475
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6358553171157837,
      "learning_rate": 1.9986346584146607e-05,
      "loss": 0.1754,
      "step": 480
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.804429829120636,
      "learning_rate": 1.9986060722937974e-05,
      "loss": 0.1178,
      "step": 485
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.834313690662384,
      "learning_rate": 1.9985771902205172e-05,
      "loss": 0.1394,
      "step": 490
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.9056774973869324,
      "learning_rate": 1.99854801220338e-05,
      "loss": 0.1521,
      "step": 495
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8872849345207214,
      "learning_rate": 1.9985185382510327e-05,
      "loss": 0.1537,
      "step": 500
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8164851069450378,
      "learning_rate": 1.998488768372211e-05,
      "loss": 0.1508,
      "step": 505
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8496866226196289,
      "learning_rate": 1.9984587025757367e-05,
      "loss": 0.1372,
      "step": 510
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8128467798233032,
      "learning_rate": 1.998428340870521e-05,
      "loss": 0.1409,
      "step": 515
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8530411720275879,
      "learning_rate": 1.9983976832655618e-05,
      "loss": 0.1313,
      "step": 520
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7878673076629639,
      "learning_rate": 1.9983667297699453e-05,
      "loss": 0.1747,
      "step": 525
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6734103560447693,
      "learning_rate": 1.9983354803928443e-05,
      "loss": 0.1228,
      "step": 530
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.9592450261116028,
      "learning_rate": 1.9983039351435208e-05,
      "loss": 0.1349,
      "step": 535
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7146002054214478,
      "learning_rate": 1.9982720940313232e-05,
      "loss": 0.1335,
      "step": 540
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.8208376169204712,
      "learning_rate": 1.998239957065689e-05,
      "loss": 0.1627,
      "step": 545
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7010489702224731,
      "learning_rate": 1.998207524256141e-05,
      "loss": 0.1125,
      "step": 550
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6489346027374268,
      "learning_rate": 1.9981747956122928e-05,
      "loss": 0.1302,
      "step": 555
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.8398034572601318,
      "learning_rate": 1.9981417711438427e-05,
      "loss": 0.1863,
      "step": 560
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.7627323269844055,
      "learning_rate": 1.998108450860579e-05,
      "loss": 0.1263,
      "step": 565
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8145926594734192,
      "learning_rate": 1.9980748347723757e-05,
      "loss": 0.1588,
      "step": 570
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.75948166847229,
      "learning_rate": 1.9980409228891962e-05,
      "loss": 0.1425,
      "step": 575
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1241673231124878,
      "learning_rate": 1.9980067152210908e-05,
      "loss": 0.1381,
      "step": 580
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.7214015126228333,
      "learning_rate": 1.9979722117781972e-05,
      "loss": 0.1329,
      "step": 585
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6891055703163147,
      "learning_rate": 1.997937412570741e-05,
      "loss": 0.1164,
      "step": 590
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.8613958954811096,
      "learning_rate": 1.9979023176090356e-05,
      "loss": 0.1379,
      "step": 595
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7776545286178589,
      "learning_rate": 1.997866926903482e-05,
      "loss": 0.1524,
      "step": 600
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.943812906742096,
      "learning_rate": 1.9978312404645685e-05,
      "loss": 0.1449,
      "step": 605
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.75004643201828,
      "learning_rate": 1.9977952583028718e-05,
      "loss": 0.1315,
      "step": 610
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6595169305801392,
      "learning_rate": 1.9977589804290557e-05,
      "loss": 0.109,
      "step": 615
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7484739422798157,
      "learning_rate": 1.9977224068538713e-05,
      "loss": 0.1217,
      "step": 620
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.9818029999732971,
      "learning_rate": 1.997685537588158e-05,
      "loss": 0.1321,
      "step": 625
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8451599478721619,
      "learning_rate": 1.9976483726428423e-05,
      "loss": 0.1375,
      "step": 630
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7618489861488342,
      "learning_rate": 1.997610912028939e-05,
      "loss": 0.1446,
      "step": 635
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6438541412353516,
      "learning_rate": 1.99757315575755e-05,
      "loss": 0.1501,
      "step": 640
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8832077383995056,
      "learning_rate": 1.9975351038398654e-05,
      "loss": 0.122,
      "step": 645
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.9104491472244263,
      "learning_rate": 1.9974967562871616e-05,
      "loss": 0.147,
      "step": 650
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.9494624733924866,
      "learning_rate": 1.997458113110804e-05,
      "loss": 0.1319,
      "step": 655
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.7751930952072144,
      "learning_rate": 1.997419174322245e-05,
      "loss": 0.1488,
      "step": 660
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.80446857213974,
      "learning_rate": 1.997379939933025e-05,
      "loss": 0.1691,
      "step": 665
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.8367449641227722,
      "learning_rate": 1.997340409954771e-05,
      "loss": 0.1514,
      "step": 670
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.8535647392272949,
      "learning_rate": 1.997300584399199e-05,
      "loss": 0.1436,
      "step": 675
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.7626618146896362,
      "learning_rate": 1.997260463278112e-05,
      "loss": 0.1323,
      "step": 680
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.9637759923934937,
      "learning_rate": 1.9972200466034e-05,
      "loss": 0.1571,
      "step": 685
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8359416723251343,
      "learning_rate": 1.9971793343870414e-05,
      "loss": 0.1406,
      "step": 690
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7154871821403503,
      "learning_rate": 1.9971383266411015e-05,
      "loss": 0.1425,
      "step": 695
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6426644325256348,
      "learning_rate": 1.9970970233777343e-05,
      "loss": 0.1401,
      "step": 700
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8010983467102051,
      "learning_rate": 1.99705542460918e-05,
      "loss": 0.1159,
      "step": 705
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8396361470222473,
      "learning_rate": 1.9970135303477674e-05,
      "loss": 0.1577,
      "step": 710
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8315424919128418,
      "learning_rate": 1.9969713406059126e-05,
      "loss": 0.1614,
      "step": 715
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8752270340919495,
      "learning_rate": 1.996928855396119e-05,
      "loss": 0.1679,
      "step": 720
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.558314859867096,
      "learning_rate": 1.9968860747309778e-05,
      "loss": 0.1395,
      "step": 725
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7923253178596497,
      "learning_rate": 1.9968429986231678e-05,
      "loss": 0.1326,
      "step": 730
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.8511756062507629,
      "learning_rate": 1.996799627085455e-05,
      "loss": 0.1192,
      "step": 735
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7928483486175537,
      "learning_rate": 1.996755960130693e-05,
      "loss": 0.1247,
      "step": 740
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.922551691532135,
      "learning_rate": 1.9967119977718245e-05,
      "loss": 0.18,
      "step": 745
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4644975960254669,
      "learning_rate": 1.9966677400218768e-05,
      "loss": 0.1349,
      "step": 750
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6948322653770447,
      "learning_rate": 1.9966231868939672e-05,
      "loss": 0.1393,
      "step": 755
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6136431097984314,
      "learning_rate": 1.9965783384012996e-05,
      "loss": 0.1236,
      "step": 760
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.776211678981781,
      "learning_rate": 1.9965331945571656e-05,
      "loss": 0.1378,
      "step": 765
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.8530936241149902,
      "learning_rate": 1.996487755374944e-05,
      "loss": 0.1007,
      "step": 770
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.8924842476844788,
      "learning_rate": 1.9964420208681016e-05,
      "loss": 0.126,
      "step": 775
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0084704160690308,
      "learning_rate": 1.9963959910501927e-05,
      "loss": 0.1524,
      "step": 780
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7858508825302124,
      "learning_rate": 1.9963496659348588e-05,
      "loss": 0.12,
      "step": 785
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0238721370697021,
      "learning_rate": 1.9963030455358293e-05,
      "loss": 0.1261,
      "step": 790
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6840950846672058,
      "learning_rate": 1.9962561298669202e-05,
      "loss": 0.178,
      "step": 795
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.777533233165741,
      "learning_rate": 1.9962089189420365e-05,
      "loss": 0.1601,
      "step": 800
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1069340705871582,
      "learning_rate": 1.9961614127751695e-05,
      "loss": 0.131,
      "step": 805
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6931054592132568,
      "learning_rate": 1.9961136113803982e-05,
      "loss": 0.1377,
      "step": 810
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7552673816680908,
      "learning_rate": 1.9960655147718894e-05,
      "loss": 0.1438,
      "step": 815
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.6302088499069214,
      "learning_rate": 1.9960171229638977e-05,
      "loss": 0.1128,
      "step": 820
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.6855728626251221,
      "learning_rate": 1.9959684359707643e-05,
      "loss": 0.1254,
      "step": 825
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.8626918196678162,
      "learning_rate": 1.9959194538069186e-05,
      "loss": 0.1077,
      "step": 830
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5936461687088013,
      "learning_rate": 1.9958701764868768e-05,
      "loss": 0.1824,
      "step": 835
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.9230837821960449,
      "learning_rate": 1.9958206040252437e-05,
      "loss": 0.0893,
      "step": 840
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7443073391914368,
      "learning_rate": 1.9957707364367102e-05,
      "loss": 0.1643,
      "step": 845
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.9582584500312805,
      "learning_rate": 1.9957205737360557e-05,
      "loss": 0.1704,
      "step": 850
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6180398464202881,
      "learning_rate": 1.9956701159381468e-05,
      "loss": 0.1821,
      "step": 855
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7253844141960144,
      "learning_rate": 1.9956193630579368e-05,
      "loss": 0.136,
      "step": 860
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7296341061592102,
      "learning_rate": 1.995568315110468e-05,
      "loss": 0.1712,
      "step": 865
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7956708073616028,
      "learning_rate": 1.9955169721108686e-05,
      "loss": 0.136,
      "step": 870
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.6485627293586731,
      "learning_rate": 1.9954653340743552e-05,
      "loss": 0.132,
      "step": 875
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5205156207084656,
      "learning_rate": 1.9954134010162317e-05,
      "loss": 0.1179,
      "step": 880
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5457340478897095,
      "learning_rate": 1.9953611729518884e-05,
      "loss": 0.1473,
      "step": 885
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.93240886926651,
      "learning_rate": 1.995308649896805e-05,
      "loss": 0.1205,
      "step": 890
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6138172149658203,
      "learning_rate": 1.9952558318665468e-05,
      "loss": 0.1397,
      "step": 895
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7555159330368042,
      "learning_rate": 1.9952027188767674e-05,
      "loss": 0.1041,
      "step": 900
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7118566036224365,
      "learning_rate": 1.995149310943208e-05,
      "loss": 0.1189,
      "step": 905
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5578028559684753,
      "learning_rate": 1.9950956080816964e-05,
      "loss": 0.1069,
      "step": 910
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8313916325569153,
      "learning_rate": 1.9950416103081486e-05,
      "loss": 0.1292,
      "step": 915
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6720024347305298,
      "learning_rate": 1.9949873176385676e-05,
      "loss": 0.1212,
      "step": 920
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.831915020942688,
      "learning_rate": 1.994932730089044e-05,
      "loss": 0.1134,
      "step": 925
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8834342956542969,
      "learning_rate": 1.9948778476757554e-05,
      "loss": 0.1652,
      "step": 930
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7775338888168335,
      "learning_rate": 1.9948226704149668e-05,
      "loss": 0.1344,
      "step": 935
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7776153683662415,
      "learning_rate": 1.9947671983230315e-05,
      "loss": 0.1366,
      "step": 940
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.7737155556678772,
      "learning_rate": 1.9947114314163892e-05,
      "loss": 0.1685,
      "step": 945
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.6648966073989868,
      "learning_rate": 1.9946553697115672e-05,
      "loss": 0.1584,
      "step": 950
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.8067174553871155,
      "learning_rate": 1.9945990132251805e-05,
      "loss": 0.147,
      "step": 955
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.740641713142395,
      "learning_rate": 1.994542361973931e-05,
      "loss": 0.1128,
      "step": 960
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.716967761516571,
      "learning_rate": 1.9944854159746084e-05,
      "loss": 0.1438,
      "step": 965
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.7191652059555054,
      "learning_rate": 1.994428175244089e-05,
      "loss": 0.1037,
      "step": 970
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.8301136493682861,
      "learning_rate": 1.9943706397993382e-05,
      "loss": 0.1434,
      "step": 975
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7556055784225464,
      "learning_rate": 1.994312809657406e-05,
      "loss": 0.1145,
      "step": 980
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7464662194252014,
      "learning_rate": 1.9942546848354322e-05,
      "loss": 0.1387,
      "step": 985
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8151674866676331,
      "learning_rate": 1.9941962653506426e-05,
      "loss": 0.1253,
      "step": 990
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.7346523404121399,
      "learning_rate": 1.994137551220351e-05,
      "loss": 0.1459,
      "step": 995
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6668190360069275,
      "learning_rate": 1.9940785424619578e-05,
      "loss": 0.1105,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6598778963088989,
      "learning_rate": 1.9940192390929518e-05,
      "loss": 0.133,
      "step": 1005
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5903392434120178,
      "learning_rate": 1.9939596411309083e-05,
      "loss": 0.1344,
      "step": 1010
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7818740010261536,
      "learning_rate": 1.9938997485934896e-05,
      "loss": 0.1178,
      "step": 1015
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5331541895866394,
      "learning_rate": 1.993839561498446e-05,
      "loss": 0.1117,
      "step": 1020
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5732979774475098,
      "learning_rate": 1.9937790798636155e-05,
      "loss": 0.0965,
      "step": 1025
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6993948817253113,
      "learning_rate": 1.9937183037069217e-05,
      "loss": 0.1001,
      "step": 1030
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7085026502609253,
      "learning_rate": 1.9936572330463777e-05,
      "loss": 0.1209,
      "step": 1035
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5701314806938171,
      "learning_rate": 1.9935958679000816e-05,
      "loss": 0.1394,
      "step": 1040
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8212972283363342,
      "learning_rate": 1.9935342082862208e-05,
      "loss": 0.1148,
      "step": 1045
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.6771918535232544,
      "learning_rate": 1.9934722542230687e-05,
      "loss": 0.1649,
      "step": 1050
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.712510883808136,
      "learning_rate": 1.9934100057289866e-05,
      "loss": 0.1188,
      "step": 1055
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.7547973990440369,
      "learning_rate": 1.9933474628224222e-05,
      "loss": 0.1336,
      "step": 1060
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.6179800629615784,
      "learning_rate": 1.9932846255219117e-05,
      "loss": 0.1277,
      "step": 1065
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9826509356498718,
      "learning_rate": 1.993221493846078e-05,
      "loss": 0.1138,
      "step": 1070
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0901100635528564,
      "learning_rate": 1.9931580678136305e-05,
      "loss": 0.1467,
      "step": 1075
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.727664589881897,
      "learning_rate": 1.9930943474433667e-05,
      "loss": 0.0986,
      "step": 1080
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.5122860670089722,
      "learning_rate": 1.9930303327541714e-05,
      "loss": 0.1246,
      "step": 1085
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.5919774174690247,
      "learning_rate": 1.9929660237650164e-05,
      "loss": 0.1011,
      "step": 1090
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0322712659835815,
      "learning_rate": 1.9929014204949603e-05,
      "loss": 0.1939,
      "step": 1095
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.8237559199333191,
      "learning_rate": 1.9928365229631494e-05,
      "loss": 0.1441,
      "step": 1100
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.7380067706108093,
      "learning_rate": 1.992771331188817e-05,
      "loss": 0.1517,
      "step": 1105
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.5965713858604431,
      "learning_rate": 1.9927058451912843e-05,
      "loss": 0.0884,
      "step": 1110
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7433302402496338,
      "learning_rate": 1.9926400649899584e-05,
      "loss": 0.1443,
      "step": 1115
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.6493307948112488,
      "learning_rate": 1.9925739906043345e-05,
      "loss": 0.1296,
      "step": 1120
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8302630186080933,
      "learning_rate": 1.9925076220539947e-05,
      "loss": 0.1202,
      "step": 1125
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7189318537712097,
      "learning_rate": 1.9924409593586086e-05,
      "loss": 0.152,
      "step": 1130
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7712987065315247,
      "learning_rate": 1.9923740025379326e-05,
      "loss": 0.1146,
      "step": 1135
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9559170603752136,
      "learning_rate": 1.99230675161181e-05,
      "loss": 0.1279,
      "step": 1140
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9394153356552124,
      "learning_rate": 1.9922392066001724e-05,
      "loss": 0.0898,
      "step": 1145
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8282424211502075,
      "learning_rate": 1.992171367523037e-05,
      "loss": 0.1355,
      "step": 1150
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8344165682792664,
      "learning_rate": 1.9921032344005097e-05,
      "loss": 0.158,
      "step": 1155
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5974935293197632,
      "learning_rate": 1.9920348072527824e-05,
      "loss": 0.1349,
      "step": 1160
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8082500696182251,
      "learning_rate": 1.9919660861001346e-05,
      "loss": 0.1336,
      "step": 1165
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.6113331913948059,
      "learning_rate": 1.991897070962933e-05,
      "loss": 0.1098,
      "step": 1170
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.8488194942474365,
      "learning_rate": 1.9918277618616312e-05,
      "loss": 0.1632,
      "step": 1175
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.8501783609390259,
      "learning_rate": 1.99175815881677e-05,
      "loss": 0.1106,
      "step": 1180
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7749114036560059,
      "learning_rate": 1.9916882618489777e-05,
      "loss": 0.1325,
      "step": 1185
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5864267349243164,
      "learning_rate": 1.991618070978969e-05,
      "loss": 0.1081,
      "step": 1190
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7307422757148743,
      "learning_rate": 1.991547586227546e-05,
      "loss": 0.1206,
      "step": 1195
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5497795939445496,
      "learning_rate": 1.9914768076155977e-05,
      "loss": 0.1195,
      "step": 1200
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.7037328481674194,
      "learning_rate": 1.991405735164101e-05,
      "loss": 0.1272,
      "step": 1205
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6973326206207275,
      "learning_rate": 1.9913343688941192e-05,
      "loss": 0.1299,
      "step": 1210
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.7845016717910767,
      "learning_rate": 1.991262708826803e-05,
      "loss": 0.1212,
      "step": 1215
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6262115836143494,
      "learning_rate": 1.9911907549833895e-05,
      "loss": 0.1388,
      "step": 1220
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.7345215082168579,
      "learning_rate": 1.9911185073852036e-05,
      "loss": 0.1271,
      "step": 1225
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.48333102464675903,
      "learning_rate": 1.9910459660536567e-05,
      "loss": 0.1041,
      "step": 1230
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6141420006752014,
      "learning_rate": 1.9909731310102477e-05,
      "loss": 0.1335,
      "step": 1235
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8758596777915955,
      "learning_rate": 1.990900002276563e-05,
      "loss": 0.1266,
      "step": 1240
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6014870405197144,
      "learning_rate": 1.9908265798742747e-05,
      "loss": 0.1311,
      "step": 1245
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6260560154914856,
      "learning_rate": 1.9907528638251427e-05,
      "loss": 0.116,
      "step": 1250
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8104715943336487,
      "learning_rate": 1.990678854151014e-05,
      "loss": 0.1237,
      "step": 1255
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6168515682220459,
      "learning_rate": 1.990604550873823e-05,
      "loss": 0.1125,
      "step": 1260
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6618785262107849,
      "learning_rate": 1.9905299540155897e-05,
      "loss": 0.0979,
      "step": 1265
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7389912605285645,
      "learning_rate": 1.9904550635984228e-05,
      "loss": 0.1644,
      "step": 1270
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7781631350517273,
      "learning_rate": 1.9903798796445174e-05,
      "loss": 0.1228,
      "step": 1275
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7687239646911621,
      "learning_rate": 1.990304402176155e-05,
      "loss": 0.1313,
      "step": 1280
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.49261704087257385,
      "learning_rate": 1.990228631215704e-05,
      "loss": 0.0842,
      "step": 1285
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8993775844573975,
      "learning_rate": 1.9901525667856213e-05,
      "loss": 0.123,
      "step": 1290
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.5044394731521606,
      "learning_rate": 1.990076208908449e-05,
      "loss": 0.1053,
      "step": 1295
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6558696627616882,
      "learning_rate": 1.9899995576068177e-05,
      "loss": 0.1264,
      "step": 1300
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6986610889434814,
      "learning_rate": 1.9899226129034433e-05,
      "loss": 0.1344,
      "step": 1305
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.8104435205459595,
      "learning_rate": 1.9898453748211305e-05,
      "loss": 0.1053,
      "step": 1310
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6717681884765625,
      "learning_rate": 1.989767843382769e-05,
      "loss": 0.1071,
      "step": 1315
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.0109965801239014,
      "learning_rate": 1.9896900186113375e-05,
      "loss": 0.1361,
      "step": 1320
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.5133586525917053,
      "learning_rate": 1.9896119005299e-05,
      "loss": 0.0943,
      "step": 1325
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6515102386474609,
      "learning_rate": 1.989533489161608e-05,
      "loss": 0.1263,
      "step": 1330
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6868942975997925,
      "learning_rate": 1.9894547845297e-05,
      "loss": 0.0937,
      "step": 1335
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8736807107925415,
      "learning_rate": 1.9893757866575012e-05,
      "loss": 0.1256,
      "step": 1340
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.5859454274177551,
      "learning_rate": 1.989296495568424e-05,
      "loss": 0.1203,
      "step": 1345
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8191548585891724,
      "learning_rate": 1.9892169112859677e-05,
      "loss": 0.1654,
      "step": 1350
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.5680137276649475,
      "learning_rate": 1.9891370338337182e-05,
      "loss": 0.1169,
      "step": 1355
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8363646864891052,
      "learning_rate": 1.9890568632353485e-05,
      "loss": 0.1673,
      "step": 1360
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7161871194839478,
      "learning_rate": 1.9889763995146178e-05,
      "loss": 0.1491,
      "step": 1365
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.5407841205596924,
      "learning_rate": 1.9888956426953735e-05,
      "loss": 0.0926,
      "step": 1370
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9694133996963501,
      "learning_rate": 1.988814592801549e-05,
      "loss": 0.0969,
      "step": 1375
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7110317349433899,
      "learning_rate": 1.988733249857165e-05,
      "loss": 0.1375,
      "step": 1380
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8600018620491028,
      "learning_rate": 1.9886516138863276e-05,
      "loss": 0.1333,
      "step": 1385
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.6293290853500366,
      "learning_rate": 1.9885696849132323e-05,
      "loss": 0.1053,
      "step": 1390
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7569682598114014,
      "learning_rate": 1.9884874629621588e-05,
      "loss": 0.1677,
      "step": 1395
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.6882773041725159,
      "learning_rate": 1.9884049480574753e-05,
      "loss": 0.1546,
      "step": 1400
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8036536574363708,
      "learning_rate": 1.988322140223637e-05,
      "loss": 0.1645,
      "step": 1405
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8944361805915833,
      "learning_rate": 1.9882390394851846e-05,
      "loss": 0.1588,
      "step": 1410
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.6535063982009888,
      "learning_rate": 1.9881556458667463e-05,
      "loss": 0.1357,
      "step": 1415
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7615301609039307,
      "learning_rate": 1.9880719593930375e-05,
      "loss": 0.1442,
      "step": 1420
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0170069932937622,
      "learning_rate": 1.9879879800888596e-05,
      "loss": 0.1655,
      "step": 1425
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.739402174949646,
      "learning_rate": 1.9879037079791014e-05,
      "loss": 0.1644,
      "step": 1430
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.48603636026382446,
      "learning_rate": 1.9878191430887382e-05,
      "loss": 0.1349,
      "step": 1435
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.512487530708313,
      "learning_rate": 1.987734285442832e-05,
      "loss": 0.1102,
      "step": 1440
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.6387705206871033,
      "learning_rate": 1.9876491350665315e-05,
      "loss": 0.1031,
      "step": 1445
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.597423255443573,
      "learning_rate": 1.987563691985073e-05,
      "loss": 0.0989,
      "step": 1450
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5606899857521057,
      "learning_rate": 1.9874779562237784e-05,
      "loss": 0.0946,
      "step": 1455
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8332621455192566,
      "learning_rate": 1.987391927808057e-05,
      "loss": 0.147,
      "step": 1460
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6586269736289978,
      "learning_rate": 1.9873056067634043e-05,
      "loss": 0.1221,
      "step": 1465
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6761701703071594,
      "learning_rate": 1.9872189931154035e-05,
      "loss": 0.1293,
      "step": 1470
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6314623951911926,
      "learning_rate": 1.987132086889723e-05,
      "loss": 0.107,
      "step": 1475
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9015997052192688,
      "learning_rate": 1.98704488811212e-05,
      "loss": 0.1291,
      "step": 1480
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7586043477058411,
      "learning_rate": 1.986957396808436e-05,
      "loss": 0.0968,
      "step": 1485
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.6687948107719421,
      "learning_rate": 1.9868696130046016e-05,
      "loss": 0.085,
      "step": 1490
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7385773062705994,
      "learning_rate": 1.9867815367266317e-05,
      "loss": 0.1183,
      "step": 1495
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8993258476257324,
      "learning_rate": 1.98669316800063e-05,
      "loss": 0.1374,
      "step": 1500
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.6431350708007812,
      "learning_rate": 1.986604506852785e-05,
      "loss": 0.1281,
      "step": 1505
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7766661047935486,
      "learning_rate": 1.986515553309374e-05,
      "loss": 0.1193,
      "step": 1510
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9873250126838684,
      "learning_rate": 1.986426307396759e-05,
      "loss": 0.1195,
      "step": 1515
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.7213776111602783,
      "learning_rate": 1.9863367691413896e-05,
      "loss": 0.1338,
      "step": 1520
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8704515099525452,
      "learning_rate": 1.9862469385698015e-05,
      "loss": 0.1237,
      "step": 1525
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8648638725280762,
      "learning_rate": 1.9861568157086182e-05,
      "loss": 0.1406,
      "step": 1530
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7704949378967285,
      "learning_rate": 1.9860664005845482e-05,
      "loss": 0.1264,
      "step": 1535
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.5187184810638428,
      "learning_rate": 1.9859756932243878e-05,
      "loss": 0.1391,
      "step": 1540
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0362634658813477,
      "learning_rate": 1.9858846936550197e-05,
      "loss": 0.1341,
      "step": 1545
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.693230152130127,
      "learning_rate": 1.9857934019034126e-05,
      "loss": 0.1053,
      "step": 1550
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8774945139884949,
      "learning_rate": 1.9857018179966223e-05,
      "loss": 0.1829,
      "step": 1555
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8253774642944336,
      "learning_rate": 1.9856099419617915e-05,
      "loss": 0.0974,
      "step": 1560
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5385066270828247,
      "learning_rate": 1.985517773826149e-05,
      "loss": 0.1159,
      "step": 1565
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5954375267028809,
      "learning_rate": 1.9854253136170096e-05,
      "loss": 0.1122,
      "step": 1570
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5960990786552429,
      "learning_rate": 1.985332561361776e-05,
      "loss": 0.1026,
      "step": 1575
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8077450394630432,
      "learning_rate": 1.9852395170879366e-05,
      "loss": 0.1304,
      "step": 1580
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.686133086681366,
      "learning_rate": 1.9851461808230667e-05,
      "loss": 0.1017,
      "step": 1585
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9486279487609863,
      "learning_rate": 1.9850525525948277e-05,
      "loss": 0.1321,
      "step": 1590
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7093173265457153,
      "learning_rate": 1.9849586324309677e-05,
      "loss": 0.1195,
      "step": 1595
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7591297626495361,
      "learning_rate": 1.9848644203593217e-05,
      "loss": 0.135,
      "step": 1600
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6641188263893127,
      "learning_rate": 1.9847699164078105e-05,
      "loss": 0.1013,
      "step": 1605
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7843316793441772,
      "learning_rate": 1.9846751206044424e-05,
      "loss": 0.1333,
      "step": 1610
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6371971368789673,
      "learning_rate": 1.984580032977311e-05,
      "loss": 0.1338,
      "step": 1615
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.5326147079467773,
      "learning_rate": 1.9844846535545975e-05,
      "loss": 0.0938,
      "step": 1620
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7384750843048096,
      "learning_rate": 1.9843889823645685e-05,
      "loss": 0.0977,
      "step": 1625
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6112232804298401,
      "learning_rate": 1.9842930194355784e-05,
      "loss": 0.1374,
      "step": 1630
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7577245831489563,
      "learning_rate": 1.9841967647960666e-05,
      "loss": 0.1432,
      "step": 1635
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7354856133460999,
      "learning_rate": 1.9841002184745598e-05,
      "loss": 0.1481,
      "step": 1640
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6963324546813965,
      "learning_rate": 1.9840033804996715e-05,
      "loss": 0.0993,
      "step": 1645
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8006656169891357,
      "learning_rate": 1.9839062509001004e-05,
      "loss": 0.1478,
      "step": 1650
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7103011012077332,
      "learning_rate": 1.983808829704633e-05,
      "loss": 0.1289,
      "step": 1655
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9841951727867126,
      "learning_rate": 1.9837111169421416e-05,
      "loss": 0.1455,
      "step": 1660
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8196072578430176,
      "learning_rate": 1.983613112641584e-05,
      "loss": 0.1225,
      "step": 1665
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7359209656715393,
      "learning_rate": 1.9835148168320064e-05,
      "loss": 0.1356,
      "step": 1670
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.6371826529502869,
      "learning_rate": 1.9834162295425393e-05,
      "loss": 0.1227,
      "step": 1675
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5500133633613586,
      "learning_rate": 1.9833173508024017e-05,
      "loss": 0.1214,
      "step": 1680
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.7776175737380981,
      "learning_rate": 1.983218180640897e-05,
      "loss": 0.0967,
      "step": 1685
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5735242366790771,
      "learning_rate": 1.983118719087416e-05,
      "loss": 0.0962,
      "step": 1690
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4979003071784973,
      "learning_rate": 1.9830189661714358e-05,
      "loss": 0.1208,
      "step": 1695
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6308743953704834,
      "learning_rate": 1.9829189219225194e-05,
      "loss": 0.1219,
      "step": 1700
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.7372656464576721,
      "learning_rate": 1.9828185863703172e-05,
      "loss": 0.1446,
      "step": 1705
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.6539852619171143,
      "learning_rate": 1.9827179595445644e-05,
      "loss": 0.1302,
      "step": 1710
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5244619846343994,
      "learning_rate": 1.9826170414750836e-05,
      "loss": 0.1014,
      "step": 1715
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5680856108665466,
      "learning_rate": 1.9825158321917836e-05,
      "loss": 0.0799,
      "step": 1720
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8987188935279846,
      "learning_rate": 1.9824143317246595e-05,
      "loss": 0.156,
      "step": 1725
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6499376893043518,
      "learning_rate": 1.982312540103792e-05,
      "loss": 0.12,
      "step": 1730
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6216092705726624,
      "learning_rate": 1.9822104573593485e-05,
      "loss": 0.1271,
      "step": 1735
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6452203989028931,
      "learning_rate": 1.9821080835215836e-05,
      "loss": 0.0962,
      "step": 1740
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6163891553878784,
      "learning_rate": 1.982005418620837e-05,
      "loss": 0.0765,
      "step": 1745
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6932926774024963,
      "learning_rate": 1.981902462687535e-05,
      "loss": 0.1292,
      "step": 1750
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7456857562065125,
      "learning_rate": 1.9817992157521898e-05,
      "loss": 0.1224,
      "step": 1755
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.7458961009979248,
      "learning_rate": 1.9816956778454012e-05,
      "loss": 0.1101,
      "step": 1760
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.1397936344146729,
      "learning_rate": 1.9815918489978535e-05,
      "loss": 0.1421,
      "step": 1765
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.6820269227027893,
      "learning_rate": 1.981487729240318e-05,
      "loss": 0.0916,
      "step": 1770
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.1029489040374756,
      "learning_rate": 1.9813833186036526e-05,
      "loss": 0.1663,
      "step": 1775
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.127271056175232,
      "learning_rate": 1.9812786171188008e-05,
      "loss": 0.1151,
      "step": 1780
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8425104022026062,
      "learning_rate": 1.981173624816792e-05,
      "loss": 0.1362,
      "step": 1785
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.65611332654953,
      "learning_rate": 1.9810683417287437e-05,
      "loss": 0.1048,
      "step": 1790
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7015208601951599,
      "learning_rate": 1.980962767885857e-05,
      "loss": 0.1278,
      "step": 1795
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7320536375045776,
      "learning_rate": 1.9808569033194205e-05,
      "loss": 0.1443,
      "step": 1800
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.56165611743927,
      "learning_rate": 1.980750748060809e-05,
      "loss": 0.1023,
      "step": 1805
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6682567596435547,
      "learning_rate": 1.980644302141483e-05,
      "loss": 0.1398,
      "step": 1810
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6880092024803162,
      "learning_rate": 1.98053756559299e-05,
      "loss": 0.1097,
      "step": 1815
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7480456829071045,
      "learning_rate": 1.9804305384469623e-05,
      "loss": 0.1643,
      "step": 1820
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7221274375915527,
      "learning_rate": 1.9803232207351197e-05,
      "loss": 0.1072,
      "step": 1825
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6828133463859558,
      "learning_rate": 1.980215612489267e-05,
      "loss": 0.1442,
      "step": 1830
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8638070821762085,
      "learning_rate": 1.980107713741296e-05,
      "loss": 0.1219,
      "step": 1835
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.6649238467216492,
      "learning_rate": 1.9799995245231837e-05,
      "loss": 0.1035,
      "step": 1840
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8936854600906372,
      "learning_rate": 1.9798910448669935e-05,
      "loss": 0.1316,
      "step": 1845
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4825485050678253,
      "learning_rate": 1.979782274804876e-05,
      "loss": 0.0937,
      "step": 1850
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7550978064537048,
      "learning_rate": 1.9796732143690663e-05,
      "loss": 0.1424,
      "step": 1855
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.7773075699806213,
      "learning_rate": 1.979563863591886e-05,
      "loss": 0.1314,
      "step": 1860
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9723725914955139,
      "learning_rate": 1.979454222505743e-05,
      "loss": 0.1628,
      "step": 1865
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4671371877193451,
      "learning_rate": 1.9793442911431315e-05,
      "loss": 0.0999,
      "step": 1870
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.5132455825805664,
      "learning_rate": 1.9792340695366304e-05,
      "loss": 0.1186,
      "step": 1875
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.47127968072891235,
      "learning_rate": 1.9791235577189067e-05,
      "loss": 0.1121,
      "step": 1880
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.725286066532135,
      "learning_rate": 1.9790127557227115e-05,
      "loss": 0.1027,
      "step": 1885
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.6264200210571289,
      "learning_rate": 1.9789016635808836e-05,
      "loss": 0.101,
      "step": 1890
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.6502572894096375,
      "learning_rate": 1.978790281326346e-05,
      "loss": 0.089,
      "step": 1895
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8432275056838989,
      "learning_rate": 1.978678608992109e-05,
      "loss": 0.1644,
      "step": 1900
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.7738016843795776,
      "learning_rate": 1.978566646611268e-05,
      "loss": 0.0792,
      "step": 1905
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.881592869758606,
      "learning_rate": 1.9784543942170056e-05,
      "loss": 0.1287,
      "step": 1910
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7415223717689514,
      "learning_rate": 1.9783418518425885e-05,
      "loss": 0.0938,
      "step": 1915
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.86686110496521,
      "learning_rate": 1.9782290195213713e-05,
      "loss": 0.1249,
      "step": 1920
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6496177315711975,
      "learning_rate": 1.978115897286793e-05,
      "loss": 0.1146,
      "step": 1925
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8109525442123413,
      "learning_rate": 1.9780024851723795e-05,
      "loss": 0.1019,
      "step": 1930
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9621292352676392,
      "learning_rate": 1.9778887832117418e-05,
      "loss": 0.092,
      "step": 1935
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7076414227485657,
      "learning_rate": 1.9777747914385782e-05,
      "loss": 0.1485,
      "step": 1940
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.6960060596466064,
      "learning_rate": 1.9776605098866705e-05,
      "loss": 0.104,
      "step": 1945
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9698060154914856,
      "learning_rate": 1.9775459385898892e-05,
      "loss": 0.09,
      "step": 1950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.5515087842941284,
      "learning_rate": 1.9774310775821883e-05,
      "loss": 0.0888,
      "step": 1955
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.644008457660675,
      "learning_rate": 1.977315926897609e-05,
      "loss": 0.1185,
      "step": 1960
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7313825488090515,
      "learning_rate": 1.9772004865702784e-05,
      "loss": 0.1103,
      "step": 1965
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7930110096931458,
      "learning_rate": 1.9770847566344083e-05,
      "loss": 0.1038,
      "step": 1970
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6186053156852722,
      "learning_rate": 1.9769687371242973e-05,
      "loss": 0.1131,
      "step": 1975
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5100052952766418,
      "learning_rate": 1.9768524280743298e-05,
      "loss": 0.0998,
      "step": 1980
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6867700815200806,
      "learning_rate": 1.976735829518976e-05,
      "loss": 0.1045,
      "step": 1985
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7107873558998108,
      "learning_rate": 1.976618941492791e-05,
      "loss": 0.1067,
      "step": 1990
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6350167393684387,
      "learning_rate": 1.9765017640304166e-05,
      "loss": 0.1032,
      "step": 1995
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7512570023536682,
      "learning_rate": 1.976384297166581e-05,
      "loss": 0.1477,
      "step": 2000
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5293204188346863,
      "learning_rate": 1.9762665409360958e-05,
      "loss": 0.1131,
      "step": 2005
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6789517402648926,
      "learning_rate": 1.9761484953738614e-05,
      "loss": 0.1101,
      "step": 2010
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.7041693329811096,
      "learning_rate": 1.9760301605148615e-05,
      "loss": 0.1155,
      "step": 2015
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5716454982757568,
      "learning_rate": 1.9759115363941668e-05,
      "loss": 0.116,
      "step": 2020
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5885083079338074,
      "learning_rate": 1.9757926230469334e-05,
      "loss": 0.0912,
      "step": 2025
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5873716473579407,
      "learning_rate": 1.975673420508403e-05,
      "loss": 0.0879,
      "step": 2030
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8442503809928894,
      "learning_rate": 1.975553928813903e-05,
      "loss": 0.1104,
      "step": 2035
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5353924036026001,
      "learning_rate": 1.975434147998847e-05,
      "loss": 0.1202,
      "step": 2040
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7577025294303894,
      "learning_rate": 1.9753140780987334e-05,
      "loss": 0.113,
      "step": 2045
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7398855686187744,
      "learning_rate": 1.9751937191491474e-05,
      "loss": 0.1304,
      "step": 2050
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.6806811690330505,
      "learning_rate": 1.9750730711857588e-05,
      "loss": 0.0784,
      "step": 2055
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7182321548461914,
      "learning_rate": 1.974952134244324e-05,
      "loss": 0.1381,
      "step": 2060
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.6555871367454529,
      "learning_rate": 1.9748309083606834e-05,
      "loss": 0.1076,
      "step": 2065
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7293125987052917,
      "learning_rate": 1.9747093935707658e-05,
      "loss": 0.1219,
      "step": 2070
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8153331875801086,
      "learning_rate": 1.9745875899105824e-05,
      "loss": 0.0967,
      "step": 2075
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6450050473213196,
      "learning_rate": 1.974465497416233e-05,
      "loss": 0.1347,
      "step": 2080
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8893457055091858,
      "learning_rate": 1.9743431161239003e-05,
      "loss": 0.1311,
      "step": 2085
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5727622509002686,
      "learning_rate": 1.974220446069855e-05,
      "loss": 0.0853,
      "step": 2090
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.623758852481842,
      "learning_rate": 1.9740974872904517e-05,
      "loss": 0.1195,
      "step": 2095
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6649274230003357,
      "learning_rate": 1.9739742398221314e-05,
      "loss": 0.1091,
      "step": 2100
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.7457634806632996,
      "learning_rate": 1.9738507037014198e-05,
      "loss": 0.1476,
      "step": 2105
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.6186316013336182,
      "learning_rate": 1.9737268789649295e-05,
      "loss": 0.1125,
      "step": 2110
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.6898834705352783,
      "learning_rate": 1.9736027656493576e-05,
      "loss": 0.1433,
      "step": 2115
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5200203061103821,
      "learning_rate": 1.9734783637914868e-05,
      "loss": 0.1071,
      "step": 2120
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.7021803855895996,
      "learning_rate": 1.9733536734281863e-05,
      "loss": 0.0774,
      "step": 2125
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.7065944671630859,
      "learning_rate": 1.973228694596409e-05,
      "loss": 0.1142,
      "step": 2130
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5829887390136719,
      "learning_rate": 1.973103427333195e-05,
      "loss": 0.0856,
      "step": 2135
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5470687747001648,
      "learning_rate": 1.9729778716756687e-05,
      "loss": 0.1114,
      "step": 2140
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.6725425720214844,
      "learning_rate": 1.9728520276610407e-05,
      "loss": 0.1097,
      "step": 2145
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8056449890136719,
      "learning_rate": 1.972725895326607e-05,
      "loss": 0.1607,
      "step": 2150
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7054765224456787,
      "learning_rate": 1.9725994747097487e-05,
      "loss": 0.0967,
      "step": 2155
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.6087821125984192,
      "learning_rate": 1.972472765847933e-05,
      "loss": 0.1179,
      "step": 2160
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.6627722978591919,
      "learning_rate": 1.972345768778711e-05,
      "loss": 0.114,
      "step": 2165
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8059279918670654,
      "learning_rate": 1.972218483539721e-05,
      "loss": 0.1533,
      "step": 2170
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9217575788497925,
      "learning_rate": 1.972090910168686e-05,
      "loss": 0.1613,
      "step": 2175
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8698978424072266,
      "learning_rate": 1.9719630487034138e-05,
      "loss": 0.122,
      "step": 2180
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4954151213169098,
      "learning_rate": 1.9718348991817988e-05,
      "loss": 0.1374,
      "step": 2185
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7494037747383118,
      "learning_rate": 1.9717064616418195e-05,
      "loss": 0.1298,
      "step": 2190
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5662988424301147,
      "learning_rate": 1.971577736121541e-05,
      "loss": 0.0793,
      "step": 2195
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.7248169779777527,
      "learning_rate": 1.9714487226591122e-05,
      "loss": 0.1084,
      "step": 2200
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5932888984680176,
      "learning_rate": 1.9713194212927694e-05,
      "loss": 0.1141,
      "step": 2205
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5105285048484802,
      "learning_rate": 1.971189832060832e-05,
      "loss": 0.0972,
      "step": 2210
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.667962372303009,
      "learning_rate": 1.9710599550017068e-05,
      "loss": 0.1347,
      "step": 2215
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6687511205673218,
      "learning_rate": 1.9709297901538836e-05,
      "loss": 0.1006,
      "step": 2220
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.7091338038444519,
      "learning_rate": 1.9707993375559402e-05,
      "loss": 0.0885,
      "step": 2225
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6677641272544861,
      "learning_rate": 1.9706685972465372e-05,
      "loss": 0.1057,
      "step": 2230
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7264307141304016,
      "learning_rate": 1.970537569264422e-05,
      "loss": 0.136,
      "step": 2235
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5116482973098755,
      "learning_rate": 1.9704062536484264e-05,
      "loss": 0.0875,
      "step": 2240
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8132953643798828,
      "learning_rate": 1.9702746504374682e-05,
      "loss": 0.1185,
      "step": 2245
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6584734916687012,
      "learning_rate": 1.9701427596705504e-05,
      "loss": 0.1155,
      "step": 2250
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6060574650764465,
      "learning_rate": 1.9700105813867598e-05,
      "loss": 0.1025,
      "step": 2255
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8468656539916992,
      "learning_rate": 1.9698781156252703e-05,
      "loss": 0.1275,
      "step": 2260
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7912387847900391,
      "learning_rate": 1.9697453624253396e-05,
      "loss": 0.1226,
      "step": 2265
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.41651102900505066,
      "learning_rate": 1.969612321826312e-05,
      "loss": 0.1076,
      "step": 2270
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7137461304664612,
      "learning_rate": 1.9694789938676156e-05,
      "loss": 0.1335,
      "step": 2275
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7056092023849487,
      "learning_rate": 1.9693453785887643e-05,
      "loss": 0.1138,
      "step": 2280
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7696053385734558,
      "learning_rate": 1.969211476029357e-05,
      "loss": 0.1369,
      "step": 2285
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8914565443992615,
      "learning_rate": 1.969077286229078e-05,
      "loss": 0.1409,
      "step": 2290
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8970075845718384,
      "learning_rate": 1.9689428092276963e-05,
      "loss": 0.1603,
      "step": 2295
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8135008215904236,
      "learning_rate": 1.9688080450650667e-05,
      "loss": 0.1301,
      "step": 2300
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7445924878120422,
      "learning_rate": 1.968672993781128e-05,
      "loss": 0.1204,
      "step": 2305
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.631936252117157,
      "learning_rate": 1.9685376554159048e-05,
      "loss": 0.0874,
      "step": 2310
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7985293865203857,
      "learning_rate": 1.9684020300095074e-05,
      "loss": 0.1071,
      "step": 2315
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6265071630477905,
      "learning_rate": 1.9682661176021297e-05,
      "loss": 0.1446,
      "step": 2320
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7634604573249817,
      "learning_rate": 1.9681299182340523e-05,
      "loss": 0.134,
      "step": 2325
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.6241299510002136,
      "learning_rate": 1.9679934319456393e-05,
      "loss": 0.1438,
      "step": 2330
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7445600628852844,
      "learning_rate": 1.9678566587773406e-05,
      "loss": 0.1006,
      "step": 2335
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.737260103225708,
      "learning_rate": 1.9677195987696912e-05,
      "loss": 0.0995,
      "step": 2340
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8832901120185852,
      "learning_rate": 1.9675822519633113e-05,
      "loss": 0.1441,
      "step": 2345
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.1757749319076538,
      "learning_rate": 1.967444618398905e-05,
      "loss": 0.1051,
      "step": 2350
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.6915484070777893,
      "learning_rate": 1.967306698117263e-05,
      "loss": 0.1169,
      "step": 2355
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8908982276916504,
      "learning_rate": 1.9671684911592595e-05,
      "loss": 0.1209,
      "step": 2360
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.6741725206375122,
      "learning_rate": 1.9670299975658544e-05,
      "loss": 0.1251,
      "step": 2365
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8597294688224792,
      "learning_rate": 1.9668912173780925e-05,
      "loss": 0.1387,
      "step": 2370
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.6731261610984802,
      "learning_rate": 1.9667521506371036e-05,
      "loss": 0.0912,
      "step": 2375
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.6857419013977051,
      "learning_rate": 1.9666127973841022e-05,
      "loss": 0.093,
      "step": 2380
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5474216341972351,
      "learning_rate": 1.9664731576603872e-05,
      "loss": 0.0884,
      "step": 2385
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5955496430397034,
      "learning_rate": 1.966333231507344e-05,
      "loss": 0.0883,
      "step": 2390
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9464544653892517,
      "learning_rate": 1.9661930189664417e-05,
      "loss": 0.1198,
      "step": 2395
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.7434287667274475,
      "learning_rate": 1.9660525200792336e-05,
      "loss": 0.1051,
      "step": 2400
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9932650327682495,
      "learning_rate": 1.9659117348873595e-05,
      "loss": 0.1326,
      "step": 2405
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6093739867210388,
      "learning_rate": 1.9657706634325433e-05,
      "loss": 0.1051,
      "step": 2410
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.47722938656806946,
      "learning_rate": 1.9656293057565932e-05,
      "loss": 0.1202,
      "step": 2415
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.6692053079605103,
      "learning_rate": 1.9654876619014034e-05,
      "loss": 0.1106,
      "step": 2420
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8569662570953369,
      "learning_rate": 1.965345731908952e-05,
      "loss": 0.1131,
      "step": 2425
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.7757257223129272,
      "learning_rate": 1.9652035158213015e-05,
      "loss": 0.1076,
      "step": 2430
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.7739999890327454,
      "learning_rate": 1.965061013680601e-05,
      "loss": 0.1501,
      "step": 2435
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.6754980087280273,
      "learning_rate": 1.9649182255290823e-05,
      "loss": 0.1372,
      "step": 2440
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.571395993232727,
      "learning_rate": 1.9647751514090636e-05,
      "loss": 0.1411,
      "step": 2445
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8778666853904724,
      "learning_rate": 1.9646317913629464e-05,
      "loss": 0.1307,
      "step": 2450
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.6127117276191711,
      "learning_rate": 1.9644881454332183e-05,
      "loss": 0.099,
      "step": 2455
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.852408230304718,
      "learning_rate": 1.9643442136624507e-05,
      "loss": 0.0993,
      "step": 2460
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8724494576454163,
      "learning_rate": 1.9641999960932997e-05,
      "loss": 0.1092,
      "step": 2465
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0813883543014526,
      "learning_rate": 1.964055492768507e-05,
      "loss": 0.1073,
      "step": 2470
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7854318618774414,
      "learning_rate": 1.9639107037308985e-05,
      "loss": 0.112,
      "step": 2475
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7752684950828552,
      "learning_rate": 1.963765629023384e-05,
      "loss": 0.1129,
      "step": 2480
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5344607830047607,
      "learning_rate": 1.963620268688959e-05,
      "loss": 0.124,
      "step": 2485
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5747987627983093,
      "learning_rate": 1.9634746227707037e-05,
      "loss": 0.1107,
      "step": 2490
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8008213639259338,
      "learning_rate": 1.9633286913117816e-05,
      "loss": 0.135,
      "step": 2495
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5648581385612488,
      "learning_rate": 1.9631824743554425e-05,
      "loss": 0.0976,
      "step": 2500
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.7081100344657898,
      "learning_rate": 1.9630359719450197e-05,
      "loss": 0.1372,
      "step": 2505
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.71477872133255,
      "learning_rate": 1.962889184123932e-05,
      "loss": 0.1119,
      "step": 2510
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.613400936126709,
      "learning_rate": 1.9627421109356817e-05,
      "loss": 0.104,
      "step": 2515
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5026811361312866,
      "learning_rate": 1.9625947524238564e-05,
      "loss": 0.0951,
      "step": 2520
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6399575471878052,
      "learning_rate": 1.9624471086321276e-05,
      "loss": 0.0751,
      "step": 2525
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.7490134835243225,
      "learning_rate": 1.9622991796042526e-05,
      "loss": 0.1211,
      "step": 2530
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6886721253395081,
      "learning_rate": 1.9621509653840724e-05,
      "loss": 0.1058,
      "step": 2535
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.689605712890625,
      "learning_rate": 1.962002466015512e-05,
      "loss": 0.0789,
      "step": 2540
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4565759003162384,
      "learning_rate": 1.9618536815425822e-05,
      "loss": 0.1147,
      "step": 2545
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5247865319252014,
      "learning_rate": 1.9617046120093772e-05,
      "loss": 0.1168,
      "step": 2550
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.6111215353012085,
      "learning_rate": 1.961555257460076e-05,
      "loss": 0.1375,
      "step": 2555
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9100821614265442,
      "learning_rate": 1.961405617938942e-05,
      "loss": 0.1192,
      "step": 2560
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7861168384552002,
      "learning_rate": 1.9612556934903236e-05,
      "loss": 0.1096,
      "step": 2565
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7265927791595459,
      "learning_rate": 1.9611054841586532e-05,
      "loss": 0.0869,
      "step": 2570
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8864724636077881,
      "learning_rate": 1.9609549899884476e-05,
      "loss": 0.1129,
      "step": 2575
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7909520864486694,
      "learning_rate": 1.9608042110243077e-05,
      "loss": 0.1203,
      "step": 2580
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5725612640380859,
      "learning_rate": 1.96065314731092e-05,
      "loss": 0.1356,
      "step": 2585
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4555843770503998,
      "learning_rate": 1.9605017988930535e-05,
      "loss": 0.099,
      "step": 2590
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.6792924404144287,
      "learning_rate": 1.9603501658155634e-05,
      "loss": 0.1184,
      "step": 2595
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.4023492336273193,
      "learning_rate": 1.9601982481233885e-05,
      "loss": 0.0903,
      "step": 2600
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8893625140190125,
      "learning_rate": 1.9600460458615517e-05,
      "loss": 0.1313,
      "step": 2605
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.840968132019043,
      "learning_rate": 1.959893559075161e-05,
      "loss": 0.1068,
      "step": 2610
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8245032429695129,
      "learning_rate": 1.9597713648187354e-05,
      "loss": 0.1491,
      "step": 2615
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8265867829322815,
      "learning_rate": 1.9596183660020867e-05,
      "loss": 0.1156,
      "step": 2620
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9567066431045532,
      "learning_rate": 1.9594650827876332e-05,
      "loss": 0.1285,
      "step": 2625
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.0007941722869873,
      "learning_rate": 1.9593115152208034e-05,
      "loss": 0.1139,
      "step": 2630
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6162289381027222,
      "learning_rate": 1.9591576633471084e-05,
      "loss": 0.1314,
      "step": 2635
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9753004312515259,
      "learning_rate": 1.9590035272121453e-05,
      "loss": 0.1472,
      "step": 2640
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.2655069828033447,
      "learning_rate": 1.9588491068615946e-05,
      "loss": 0.114,
      "step": 2645
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.1793577671051025,
      "learning_rate": 1.9586944023412212e-05,
      "loss": 0.123,
      "step": 2650
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.7227995991706848,
      "learning_rate": 1.958539413696874e-05,
      "loss": 0.1043,
      "step": 2655
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.6717692017555237,
      "learning_rate": 1.9583841409744862e-05,
      "loss": 0.0765,
      "step": 2660
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5767033100128174,
      "learning_rate": 1.9582285842200753e-05,
      "loss": 0.1184,
      "step": 2665
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9931422472000122,
      "learning_rate": 1.9580727434797432e-05,
      "loss": 0.0844,
      "step": 2670
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.6345029473304749,
      "learning_rate": 1.957916618799676e-05,
      "loss": 0.1105,
      "step": 2675
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7955521941184998,
      "learning_rate": 1.957760210226143e-05,
      "loss": 0.1172,
      "step": 2680
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7941994071006775,
      "learning_rate": 1.9576035178054988e-05,
      "loss": 0.0976,
      "step": 2685
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.0277880430221558,
      "learning_rate": 1.957446541584181e-05,
      "loss": 0.1201,
      "step": 2690
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7474991083145142,
      "learning_rate": 1.9572892816087132e-05,
      "loss": 0.1154,
      "step": 2695
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.788794219493866,
      "learning_rate": 1.9571317379257006e-05,
      "loss": 0.1399,
      "step": 2700
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7396997809410095,
      "learning_rate": 1.9569739105818342e-05,
      "loss": 0.0995,
      "step": 2705
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9025750756263733,
      "learning_rate": 1.9568157996238884e-05,
      "loss": 0.0943,
      "step": 2710
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7392351627349854,
      "learning_rate": 1.956657405098722e-05,
      "loss": 0.1283,
      "step": 2715
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.934230625629425,
      "learning_rate": 1.9564987270532777e-05,
      "loss": 0.1131,
      "step": 2720
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.8671979308128357,
      "learning_rate": 1.9563397655345822e-05,
      "loss": 0.13,
      "step": 2725
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7905526161193848,
      "learning_rate": 1.956180520589746e-05,
      "loss": 0.0957,
      "step": 2730
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.6696571111679077,
      "learning_rate": 1.956020992265964e-05,
      "loss": 0.1187,
      "step": 2735
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.528008759021759,
      "learning_rate": 1.955861180610515e-05,
      "loss": 0.0787,
      "step": 2740
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.014591932296753,
      "learning_rate": 1.9557010856707615e-05,
      "loss": 0.1075,
      "step": 2745
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8025286197662354,
      "learning_rate": 1.9555407074941503e-05,
      "loss": 0.1274,
      "step": 2750
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.1211886405944824,
      "learning_rate": 1.9553800461282114e-05,
      "loss": 0.1212,
      "step": 2755
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.1589736938476562,
      "learning_rate": 1.95521910162056e-05,
      "loss": 0.1171,
      "step": 2760
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.9360532164573669,
      "learning_rate": 1.9550578740188945e-05,
      "loss": 0.1468,
      "step": 2765
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8516525626182556,
      "learning_rate": 1.9548963633709967e-05,
      "loss": 0.1472,
      "step": 2770
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.8036392331123352,
      "learning_rate": 1.954734569724733e-05,
      "loss": 0.0956,
      "step": 2775
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.0293813943862915,
      "learning_rate": 1.9545724931280535e-05,
      "loss": 0.1474,
      "step": 2780
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.9267975687980652,
      "learning_rate": 1.954410133628992e-05,
      "loss": 0.1416,
      "step": 2785
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.7071394920349121,
      "learning_rate": 1.9542474912756663e-05,
      "loss": 0.0968,
      "step": 2790
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.9008345007896423,
      "learning_rate": 1.954084566116278e-05,
      "loss": 0.143,
      "step": 2795
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.6870996356010437,
      "learning_rate": 1.9539213581991127e-05,
      "loss": 0.133,
      "step": 2800
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.8136075735092163,
      "learning_rate": 1.9537578675725393e-05,
      "loss": 0.1283,
      "step": 2805
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7019197344779968,
      "learning_rate": 1.953594094285011e-05,
      "loss": 0.1398,
      "step": 2810
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.6268460750579834,
      "learning_rate": 1.9534300383850643e-05,
      "loss": 0.0928,
      "step": 2815
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5861523747444153,
      "learning_rate": 1.95326569992132e-05,
      "loss": 0.1075,
      "step": 2820
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7399869561195374,
      "learning_rate": 1.953101078942482e-05,
      "loss": 0.122,
      "step": 2825
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.8029427528381348,
      "learning_rate": 1.952936175497339e-05,
      "loss": 0.109,
      "step": 2830
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7381945848464966,
      "learning_rate": 1.9527709896347623e-05,
      "loss": 0.1099,
      "step": 2835
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.680610716342926,
      "learning_rate": 1.9526055214037067e-05,
      "loss": 0.1067,
      "step": 2840
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5349668264389038,
      "learning_rate": 1.952439770853212e-05,
      "loss": 0.0878,
      "step": 2845
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.8351783156394958,
      "learning_rate": 1.952273738032401e-05,
      "loss": 0.0794,
      "step": 2850
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.887694239616394,
      "learning_rate": 1.9521074229904797e-05,
      "loss": 0.118,
      "step": 2855
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.675528883934021,
      "learning_rate": 1.9519408257767383e-05,
      "loss": 0.1485,
      "step": 2860
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.2613089084625244,
      "learning_rate": 1.9517739464405503e-05,
      "loss": 0.0914,
      "step": 2865
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6458963751792908,
      "learning_rate": 1.9516067850313734e-05,
      "loss": 0.0831,
      "step": 2870
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6229132413864136,
      "learning_rate": 1.9514393415987484e-05,
      "loss": 0.1205,
      "step": 2875
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6624443531036377,
      "learning_rate": 1.951271616192299e-05,
      "loss": 0.1246,
      "step": 2880
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.7545179724693298,
      "learning_rate": 1.9511036088617344e-05,
      "loss": 0.1181,
      "step": 2885
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.689713716506958,
      "learning_rate": 1.9509353196568454e-05,
      "loss": 0.117,
      "step": 2890
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8974180221557617,
      "learning_rate": 1.9507667486275072e-05,
      "loss": 0.1255,
      "step": 2895
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8327198028564453,
      "learning_rate": 1.9505978958236785e-05,
      "loss": 0.1371,
      "step": 2900
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.7198882102966309,
      "learning_rate": 1.9504287612954017e-05,
      "loss": 0.1352,
      "step": 2905
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5859416127204895,
      "learning_rate": 1.9502593450928025e-05,
      "loss": 0.0857,
      "step": 2910
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.7368265986442566,
      "learning_rate": 1.9500896472660896e-05,
      "loss": 0.0864,
      "step": 2915
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5910233855247498,
      "learning_rate": 1.9499196678655554e-05,
      "loss": 0.1122,
      "step": 2920
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.6884198784828186,
      "learning_rate": 1.9497494069415767e-05,
      "loss": 0.1076,
      "step": 2925
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.8379870057106018,
      "learning_rate": 1.9495788645446124e-05,
      "loss": 0.1118,
      "step": 2930
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7852750420570374,
      "learning_rate": 1.9494080407252058e-05,
      "loss": 0.1028,
      "step": 2935
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5889078378677368,
      "learning_rate": 1.9492369355339827e-05,
      "loss": 0.1474,
      "step": 2940
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.8883277177810669,
      "learning_rate": 1.9490655490216534e-05,
      "loss": 0.1115,
      "step": 2945
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.0939563512802124,
      "learning_rate": 1.9488938812390102e-05,
      "loss": 0.112,
      "step": 2950
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6624184250831604,
      "learning_rate": 1.94872193223693e-05,
      "loss": 0.1048,
      "step": 2955
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.8896359801292419,
      "learning_rate": 1.9485497020663725e-05,
      "loss": 0.1351,
      "step": 2960
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.6318917870521545,
      "learning_rate": 1.9483771907783806e-05,
      "loss": 0.1173,
      "step": 2965
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.7852448225021362,
      "learning_rate": 1.948204398424081e-05,
      "loss": 0.1057,
      "step": 2970
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.6205310225486755,
      "learning_rate": 1.9480313250546833e-05,
      "loss": 0.1229,
      "step": 2975
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.885155439376831,
      "learning_rate": 1.94785797072148e-05,
      "loss": 0.0921,
      "step": 2980
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.7539743185043335,
      "learning_rate": 1.9476843354758483e-05,
      "loss": 0.1534,
      "step": 2985
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.8679502010345459,
      "learning_rate": 1.947510419369247e-05,
      "loss": 0.0976,
      "step": 2990
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5822016000747681,
      "learning_rate": 1.947336222453219e-05,
      "loss": 0.0882,
      "step": 2995
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7897557020187378,
      "learning_rate": 1.9471617447793903e-05,
      "loss": 0.1242,
      "step": 3000
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.6063729524612427,
      "learning_rate": 1.94698698639947e-05,
      "loss": 0.1531,
      "step": 3005
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5307276844978333,
      "learning_rate": 1.9468119473652506e-05,
      "loss": 0.1106,
      "step": 3010
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.6027793884277344,
      "learning_rate": 1.9466366277286075e-05,
      "loss": 0.1413,
      "step": 3015
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.4923043251037598,
      "learning_rate": 1.9464610275414996e-05,
      "loss": 0.1398,
      "step": 3020
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7162624001502991,
      "learning_rate": 1.946285146855968e-05,
      "loss": 0.0702,
      "step": 3025
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.689167857170105,
      "learning_rate": 1.9461089857241387e-05,
      "loss": 0.1331,
      "step": 3030
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.7341986298561096,
      "learning_rate": 1.9459325441982192e-05,
      "loss": 0.062,
      "step": 3035
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5668254494667053,
      "learning_rate": 1.945755822330501e-05,
      "loss": 0.1277,
      "step": 3040
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6565370559692383,
      "learning_rate": 1.945578820173358e-05,
      "loss": 0.0937,
      "step": 3045
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6408843398094177,
      "learning_rate": 1.9454015377792478e-05,
      "loss": 0.1204,
      "step": 3050
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.683785617351532,
      "learning_rate": 1.945223975200711e-05,
      "loss": 0.0826,
      "step": 3055
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6759133338928223,
      "learning_rate": 1.94504613249037e-05,
      "loss": 0.1165,
      "step": 3060
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.9206417202949524,
      "learning_rate": 1.9448680097009326e-05,
      "loss": 0.1219,
      "step": 3065
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5556135773658752,
      "learning_rate": 1.9446896068851877e-05,
      "loss": 0.0858,
      "step": 3070
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.709461510181427,
      "learning_rate": 1.9445109240960076e-05,
      "loss": 0.0787,
      "step": 3075
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.8782566785812378,
      "learning_rate": 1.9443319613863475e-05,
      "loss": 0.1239,
      "step": 3080
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.7591779828071594,
      "learning_rate": 1.9441527188092462e-05,
      "loss": 0.1403,
      "step": 3085
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.8730403780937195,
      "learning_rate": 1.9439731964178252e-05,
      "loss": 0.1159,
      "step": 3090
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.6206468939781189,
      "learning_rate": 1.9437933942652883e-05,
      "loss": 0.1126,
      "step": 3095
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.569102942943573,
      "learning_rate": 1.9436133124049227e-05,
      "loss": 0.0954,
      "step": 3100
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8233943581581116,
      "learning_rate": 1.943432950890099e-05,
      "loss": 0.0972,
      "step": 3105
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8483831286430359,
      "learning_rate": 1.9432523097742693e-05,
      "loss": 0.1138,
      "step": 3110
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.47807222604751587,
      "learning_rate": 1.9430713891109698e-05,
      "loss": 0.1364,
      "step": 3115
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.7278242111206055,
      "learning_rate": 1.942890188953819e-05,
      "loss": 0.1297,
      "step": 3120
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.6293314695358276,
      "learning_rate": 1.942708709356519e-05,
      "loss": 0.1143,
      "step": 3125
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.9456065893173218,
      "learning_rate": 1.9425269503728536e-05,
      "loss": 0.1052,
      "step": 3130
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.8846820592880249,
      "learning_rate": 1.94234491205669e-05,
      "loss": 0.1414,
      "step": 3135
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.6343728303909302,
      "learning_rate": 1.9421625944619778e-05,
      "loss": 0.0812,
      "step": 3140
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5727220177650452,
      "learning_rate": 1.9419799976427497e-05,
      "loss": 0.1436,
      "step": 3145
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.7205094695091248,
      "learning_rate": 1.9417971216531217e-05,
      "loss": 0.1175,
      "step": 3150
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.8147839307785034,
      "learning_rate": 1.941613966547291e-05,
      "loss": 0.1143,
      "step": 3155
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.0907286405563354,
      "learning_rate": 1.941430532379539e-05,
      "loss": 0.1352,
      "step": 3160
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.6234989762306213,
      "learning_rate": 1.9412468192042298e-05,
      "loss": 0.0985,
      "step": 3165
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.9922168850898743,
      "learning_rate": 1.941062827075809e-05,
      "loss": 0.1644,
      "step": 3170
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5943562984466553,
      "learning_rate": 1.9408785560488052e-05,
      "loss": 0.1064,
      "step": 3175
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.7550861239433289,
      "learning_rate": 1.9406940061778306e-05,
      "loss": 0.0945,
      "step": 3180
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.4942474663257599,
      "learning_rate": 1.9405091775175792e-05,
      "loss": 0.1299,
      "step": 3185
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.6698021292686462,
      "learning_rate": 1.9403240701228277e-05,
      "loss": 0.1189,
      "step": 3190
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7570865154266357,
      "learning_rate": 1.940138684048436e-05,
      "loss": 0.1232,
      "step": 3195
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.40966135263442993,
      "learning_rate": 1.939953019349346e-05,
      "loss": 0.1092,
      "step": 3200
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6771470904350281,
      "learning_rate": 1.9397670760805817e-05,
      "loss": 0.1223,
      "step": 3205
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6577381491661072,
      "learning_rate": 1.9395808542972507e-05,
      "loss": 0.076,
      "step": 3210
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6543042063713074,
      "learning_rate": 1.9393943540545434e-05,
      "loss": 0.0738,
      "step": 3215
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8199294209480286,
      "learning_rate": 1.9392075754077307e-05,
      "loss": 0.0757,
      "step": 3220
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.0831599235534668,
      "learning_rate": 1.9390205184121683e-05,
      "loss": 0.063,
      "step": 3225
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.5358847975730896,
      "learning_rate": 1.9388331831232934e-05,
      "loss": 0.075,
      "step": 3230
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.7586596608161926,
      "learning_rate": 1.9386455695966253e-05,
      "loss": 0.0643,
      "step": 3235
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.8688328862190247,
      "learning_rate": 1.9384576778877663e-05,
      "loss": 0.0563,
      "step": 3240
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.305880069732666,
      "learning_rate": 1.9382695080524013e-05,
      "loss": 0.0794,
      "step": 3245
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.7951090931892395,
      "learning_rate": 1.9380810601462974e-05,
      "loss": 0.0795,
      "step": 3250
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.5801917314529419,
      "learning_rate": 1.9378923342253035e-05,
      "loss": 0.0567,
      "step": 3255
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.7360835075378418,
      "learning_rate": 1.9377033303453522e-05,
      "loss": 0.0528,
      "step": 3260
    },
    {
      "epoch": 1.02,
      "grad_norm": 13.658980369567871,
      "learning_rate": 1.9375140485624568e-05,
      "loss": 0.066,
      "step": 3265
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.7239563465118408,
      "learning_rate": 1.937324488932715e-05,
      "loss": 0.0806,
      "step": 3270
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.6245361566543579,
      "learning_rate": 1.9371346515123047e-05,
      "loss": 0.0443,
      "step": 3275
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.6881046295166016,
      "learning_rate": 1.9369445363574877e-05,
      "loss": 0.0639,
      "step": 3280
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.8322771191596985,
      "learning_rate": 1.9367541435246077e-05,
      "loss": 0.0843,
      "step": 3285
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.9991777539253235,
      "learning_rate": 1.9365634730700903e-05,
      "loss": 0.0498,
      "step": 3290
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.6667386293411255,
      "learning_rate": 1.9363725250504438e-05,
      "loss": 0.0881,
      "step": 3295
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.7544474005699158,
      "learning_rate": 1.9361812995222586e-05,
      "loss": 0.0717,
      "step": 3300
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.7047283053398132,
      "learning_rate": 1.935989796542207e-05,
      "loss": 0.0863,
      "step": 3305
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.7891948819160461,
      "learning_rate": 1.9357980161670443e-05,
      "loss": 0.0479,
      "step": 3310
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.8909279704093933,
      "learning_rate": 1.935605958453608e-05,
      "loss": 0.0662,
      "step": 3315
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.6332457065582275,
      "learning_rate": 1.9354136234588163e-05,
      "loss": 0.058,
      "step": 3320
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.6614800691604614,
      "learning_rate": 1.9352210112396713e-05,
      "loss": 0.0642,
      "step": 3325
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.7803705334663391,
      "learning_rate": 1.935028121853257e-05,
      "loss": 0.063,
      "step": 3330
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.6813333034515381,
      "learning_rate": 1.934834955356738e-05,
      "loss": 0.0656,
      "step": 3335
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.1551873683929443,
      "learning_rate": 1.9346415118073634e-05,
      "loss": 0.087,
      "step": 3340
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.6966256499290466,
      "learning_rate": 1.934447791262463e-05,
      "loss": 0.0621,
      "step": 3345
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.8793096542358398,
      "learning_rate": 1.934253793779448e-05,
      "loss": 0.0504,
      "step": 3350
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.9607953429222107,
      "learning_rate": 1.934059519415813e-05,
      "loss": 0.0789,
      "step": 3355
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.8557941317558289,
      "learning_rate": 1.933864968229135e-05,
      "loss": 0.106,
      "step": 3360
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.7457416653633118,
      "learning_rate": 1.9336701402770716e-05,
      "loss": 0.0633,
      "step": 3365
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.5383043885231018,
      "learning_rate": 1.9334750356173627e-05,
      "loss": 0.0534,
      "step": 3370
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.9611634612083435,
      "learning_rate": 1.9332796543078313e-05,
      "loss": 0.057,
      "step": 3375
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.6954233646392822,
      "learning_rate": 1.9330839964063815e-05,
      "loss": 0.0562,
      "step": 3380
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.7774192094802856,
      "learning_rate": 1.9328880619709995e-05,
      "loss": 0.068,
      "step": 3385
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.7439762949943542,
      "learning_rate": 1.9326918510597534e-05,
      "loss": 0.0469,
      "step": 3390
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.5823460221290588,
      "learning_rate": 1.9324953637307935e-05,
      "loss": 0.0599,
      "step": 3395
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.6018666625022888,
      "learning_rate": 1.932298600042352e-05,
      "loss": 0.048,
      "step": 3400
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.6874041557312012,
      "learning_rate": 1.9321015600527425e-05,
      "loss": 0.0947,
      "step": 3405
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.7596636414527893,
      "learning_rate": 1.931904243820361e-05,
      "loss": 0.0719,
      "step": 3410
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.7127168774604797,
      "learning_rate": 1.931706651403685e-05,
      "loss": 0.0699,
      "step": 3415
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.7314068675041199,
      "learning_rate": 1.9315087828612747e-05,
      "loss": 0.0626,
      "step": 3420
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.0029963254928589,
      "learning_rate": 1.9313106382517714e-05,
      "loss": 0.0673,
      "step": 3425
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.8213211297988892,
      "learning_rate": 1.931112217633898e-05,
      "loss": 0.0641,
      "step": 3430
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.6091497540473938,
      "learning_rate": 1.9309135210664592e-05,
      "loss": 0.0706,
      "step": 3435
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.019183874130249,
      "learning_rate": 1.9307145486083426e-05,
      "loss": 0.0753,
      "step": 3440
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.7182345986366272,
      "learning_rate": 1.9305153003185167e-05,
      "loss": 0.07,
      "step": 3445
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.9115707874298096,
      "learning_rate": 1.930315776256031e-05,
      "loss": 0.0686,
      "step": 3450
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.6539191007614136,
      "learning_rate": 1.9301159764800185e-05,
      "loss": 0.068,
      "step": 3455
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.5159173607826233,
      "learning_rate": 1.9299159010496925e-05,
      "loss": 0.0555,
      "step": 3460
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.8276963233947754,
      "learning_rate": 1.9297155500243485e-05,
      "loss": 0.0509,
      "step": 3465
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.9333387613296509,
      "learning_rate": 1.9295149234633637e-05,
      "loss": 0.058,
      "step": 3470
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.083387851715088,
      "learning_rate": 1.9293140214261967e-05,
      "loss": 0.0626,
      "step": 3475
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.7512198090553284,
      "learning_rate": 1.9291128439723886e-05,
      "loss": 0.0548,
      "step": 3480
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.9256929755210876,
      "learning_rate": 1.9289113911615605e-05,
      "loss": 0.0795,
      "step": 3485
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.6555706858634949,
      "learning_rate": 1.9287096630534168e-05,
      "loss": 0.0551,
      "step": 3490
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.8473435044288635,
      "learning_rate": 1.9285076597077425e-05,
      "loss": 0.0676,
      "step": 3495
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.7201313972473145,
      "learning_rate": 1.928305381184404e-05,
      "loss": 0.0758,
      "step": 3500
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.5640273094177246,
      "learning_rate": 1.9281028275433505e-05,
      "loss": 0.0706,
      "step": 3505
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.6947470903396606,
      "learning_rate": 1.9278999988446118e-05,
      "loss": 0.0799,
      "step": 3510
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.9326536059379578,
      "learning_rate": 1.9276968951482987e-05,
      "loss": 0.0681,
      "step": 3515
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.6781297922134399,
      "learning_rate": 1.9274935165146044e-05,
      "loss": 0.0565,
      "step": 3520
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.8024562001228333,
      "learning_rate": 1.927289863003804e-05,
      "loss": 0.0762,
      "step": 3525
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.5281566977500916,
      "learning_rate": 1.9270859346762522e-05,
      "loss": 0.0701,
      "step": 3530
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.5051602721214294,
      "learning_rate": 1.926881731592387e-05,
      "loss": 0.0612,
      "step": 3535
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.5649287104606628,
      "learning_rate": 1.926677253812727e-05,
      "loss": 0.076,
      "step": 3540
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.5975441932678223,
      "learning_rate": 1.9264725013978727e-05,
      "loss": 0.0699,
      "step": 3545
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.5763807892799377,
      "learning_rate": 1.9262674744085054e-05,
      "loss": 0.0537,
      "step": 3550
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.8457044363021851,
      "learning_rate": 1.926062172905388e-05,
      "loss": 0.0728,
      "step": 3555
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.7114633321762085,
      "learning_rate": 1.9258565969493647e-05,
      "loss": 0.0566,
      "step": 3560
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.7445443868637085,
      "learning_rate": 1.925650746601361e-05,
      "loss": 0.0562,
      "step": 3565
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.5508908033370972,
      "learning_rate": 1.9254446219223845e-05,
      "loss": 0.0617,
      "step": 3570
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.8591492176055908,
      "learning_rate": 1.925238222973523e-05,
      "loss": 0.071,
      "step": 3575
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.834518551826477,
      "learning_rate": 1.925031549815946e-05,
      "loss": 0.0787,
      "step": 3580
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.718428909778595,
      "learning_rate": 1.9248246025109044e-05,
      "loss": 0.0849,
      "step": 3585
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.7240105271339417,
      "learning_rate": 1.9246173811197302e-05,
      "loss": 0.0561,
      "step": 3590
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.5630722045898438,
      "learning_rate": 1.924409885703837e-05,
      "loss": 0.0515,
      "step": 3595
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.63336580991745,
      "learning_rate": 1.9242021163247187e-05,
      "loss": 0.0576,
      "step": 3600
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.6805610060691833,
      "learning_rate": 1.923994073043951e-05,
      "loss": 0.0727,
      "step": 3605
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.3520512580871582,
      "learning_rate": 1.9237857559231914e-05,
      "loss": 0.0758,
      "step": 3610
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.8155093789100647,
      "learning_rate": 1.9235771650241776e-05,
      "loss": 0.079,
      "step": 3615
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.9271829128265381,
      "learning_rate": 1.923368300408729e-05,
      "loss": 0.0808,
      "step": 3620
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.8426121473312378,
      "learning_rate": 1.923159162138745e-05,
      "loss": 0.0671,
      "step": 3625
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.6503888368606567,
      "learning_rate": 1.9229497502762075e-05,
      "loss": 0.0624,
      "step": 3630
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.5869102478027344,
      "learning_rate": 1.922740064883179e-05,
      "loss": 0.0661,
      "step": 3635
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.7934098839759827,
      "learning_rate": 1.9225301060218032e-05,
      "loss": 0.073,
      "step": 3640
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.608038067817688,
      "learning_rate": 1.9223198737543046e-05,
      "loss": 0.0725,
      "step": 3645
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.5883069038391113,
      "learning_rate": 1.9221093681429887e-05,
      "loss": 0.0647,
      "step": 3650
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.8434496521949768,
      "learning_rate": 1.921898589250242e-05,
      "loss": 0.0582,
      "step": 3655
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.7228140830993652,
      "learning_rate": 1.9216875371385323e-05,
      "loss": 0.0604,
      "step": 3660
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.7881928086280823,
      "learning_rate": 1.921476211870408e-05,
      "loss": 0.0674,
      "step": 3665
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.7950953841209412,
      "learning_rate": 1.9212646135084986e-05,
      "loss": 0.0647,
      "step": 3670
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.4857760667800903,
      "learning_rate": 1.9210527421155145e-05,
      "loss": 0.0701,
      "step": 3675
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.7841891050338745,
      "learning_rate": 1.9208405977542474e-05,
      "loss": 0.0633,
      "step": 3680
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.9583932757377625,
      "learning_rate": 1.9206281804875695e-05,
      "loss": 0.0905,
      "step": 3685
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.8628225922584534,
      "learning_rate": 1.9204154903784337e-05,
      "loss": 0.0739,
      "step": 3690
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.9285526275634766,
      "learning_rate": 1.9202025274898744e-05,
      "loss": 0.0659,
      "step": 3695
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.1224682331085205,
      "learning_rate": 1.919989291885006e-05,
      "loss": 0.062,
      "step": 3700
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.8893318772315979,
      "learning_rate": 1.9197757836270245e-05,
      "loss": 0.0844,
      "step": 3705
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.6806356906890869,
      "learning_rate": 1.9195620027792063e-05,
      "loss": 0.0459,
      "step": 3710
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.7901091575622559,
      "learning_rate": 1.9193479494049088e-05,
      "loss": 0.0829,
      "step": 3715
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.6605781316757202,
      "learning_rate": 1.9191336235675698e-05,
      "loss": 0.0566,
      "step": 3720
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.8274343609809875,
      "learning_rate": 1.9189190253307082e-05,
      "loss": 0.0723,
      "step": 3725
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.6665601134300232,
      "learning_rate": 1.9187041547579234e-05,
      "loss": 0.0623,
      "step": 3730
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.6877213716506958,
      "learning_rate": 1.9184890119128963e-05,
      "loss": 0.0582,
      "step": 3735
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.7861253619194031,
      "learning_rate": 1.9182735968593865e-05,
      "loss": 0.0684,
      "step": 3740
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.7444209456443787,
      "learning_rate": 1.918057909661237e-05,
      "loss": 0.0677,
      "step": 3745
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.5835312604904175,
      "learning_rate": 1.9178419503823692e-05,
      "loss": 0.0677,
      "step": 3750
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.8612051606178284,
      "learning_rate": 1.9176257190867864e-05,
      "loss": 0.0868,
      "step": 3755
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.8530820608139038,
      "learning_rate": 1.9174092158385716e-05,
      "loss": 0.0629,
      "step": 3760
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.6976132392883301,
      "learning_rate": 1.9171924407018895e-05,
      "loss": 0.0742,
      "step": 3765
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.8631473183631897,
      "learning_rate": 1.9169753937409844e-05,
      "loss": 0.0783,
      "step": 3770
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.6785290241241455,
      "learning_rate": 1.9167580750201816e-05,
      "loss": 0.062,
      "step": 3775
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.9032337665557861,
      "learning_rate": 1.9165404846038868e-05,
      "loss": 0.071,
      "step": 3780
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.2370872497558594,
      "learning_rate": 1.9163226225565867e-05,
      "loss": 0.0724,
      "step": 3785
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.7235238552093506,
      "learning_rate": 1.9161044889428476e-05,
      "loss": 0.0482,
      "step": 3790
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.558603048324585,
      "learning_rate": 1.915886083827317e-05,
      "loss": 0.0522,
      "step": 3795
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.9623020887374878,
      "learning_rate": 1.915667407274723e-05,
      "loss": 0.0667,
      "step": 3800
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.6551677584648132,
      "learning_rate": 1.915448459349873e-05,
      "loss": 0.0478,
      "step": 3805
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.7484034895896912,
      "learning_rate": 1.9152292401176563e-05,
      "loss": 0.0731,
      "step": 3810
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.8798559308052063,
      "learning_rate": 1.9150097496430415e-05,
      "loss": 0.0743,
      "step": 3815
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.0757192373275757,
      "learning_rate": 1.9147899879910783e-05,
      "loss": 0.0715,
      "step": 3820
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.9112750887870789,
      "learning_rate": 1.914569955226896e-05,
      "loss": 0.0818,
      "step": 3825
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.5911403298377991,
      "learning_rate": 1.9143496514157056e-05,
      "loss": 0.0709,
      "step": 3830
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.9122008085250854,
      "learning_rate": 1.914129076622797e-05,
      "loss": 0.0885,
      "step": 3835
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.8840839862823486,
      "learning_rate": 1.913908230913541e-05,
      "loss": 0.0642,
      "step": 3840
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.6602410078048706,
      "learning_rate": 1.9136871143533884e-05,
      "loss": 0.0495,
      "step": 3845
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.5362725853919983,
      "learning_rate": 1.913465727007871e-05,
      "loss": 0.0692,
      "step": 3850
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.8106977343559265,
      "learning_rate": 1.9132440689426e-05,
      "loss": 0.0856,
      "step": 3855
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.6339511871337891,
      "learning_rate": 1.9130221402232676e-05,
      "loss": 0.0601,
      "step": 3860
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.7605702877044678,
      "learning_rate": 1.9127999409156454e-05,
      "loss": 0.0627,
      "step": 3865
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.622286856174469,
      "learning_rate": 1.912577471085586e-05,
      "loss": 0.0624,
      "step": 3870
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.5903645157814026,
      "learning_rate": 1.9123547307990215e-05,
      "loss": 0.048,
      "step": 3875
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.6491131782531738,
      "learning_rate": 1.9121317201219645e-05,
      "loss": 0.0706,
      "step": 3880
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.7359797358512878,
      "learning_rate": 1.911908439120508e-05,
      "loss": 0.0626,
      "step": 3885
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.683800458908081,
      "learning_rate": 1.9116848878608243e-05,
      "loss": 0.0506,
      "step": 3890
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.6683903336524963,
      "learning_rate": 1.9114610664091665e-05,
      "loss": 0.064,
      "step": 3895
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.6654495596885681,
      "learning_rate": 1.911236974831868e-05,
      "loss": 0.0674,
      "step": 3900
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.8582611680030823,
      "learning_rate": 1.9110126131953408e-05,
      "loss": 0.0738,
      "step": 3905
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.9522505402565002,
      "learning_rate": 1.9107879815660788e-05,
      "loss": 0.0609,
      "step": 3910
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.6455296277999878,
      "learning_rate": 1.910563080010655e-05,
      "loss": 0.0682,
      "step": 3915
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.6582437753677368,
      "learning_rate": 1.910337908595722e-05,
      "loss": 0.0585,
      "step": 3920
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.7527994513511658,
      "learning_rate": 1.9101124673880132e-05,
      "loss": 0.0541,
      "step": 3925
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.9205194115638733,
      "learning_rate": 1.909886756454341e-05,
      "loss": 0.0913,
      "step": 3930
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.8828095197677612,
      "learning_rate": 1.9096607758615998e-05,
      "loss": 0.0674,
      "step": 3935
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.6131769418716431,
      "learning_rate": 1.9094345256767614e-05,
      "loss": 0.0587,
      "step": 3940
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.0256649255752563,
      "learning_rate": 1.9092080059668784e-05,
      "loss": 0.0627,
      "step": 3945
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.8035500645637512,
      "learning_rate": 1.9089812167990836e-05,
      "loss": 0.0805,
      "step": 3950
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.6150037050247192,
      "learning_rate": 1.9087541582405897e-05,
      "loss": 0.055,
      "step": 3955
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.6392942667007446,
      "learning_rate": 1.9085268303586892e-05,
      "loss": 0.0633,
      "step": 3960
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.6936980485916138,
      "learning_rate": 1.908299233220754e-05,
      "loss": 0.0857,
      "step": 3965
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.823484480381012,
      "learning_rate": 1.9080713668942356e-05,
      "loss": 0.0834,
      "step": 3970
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.6093165278434753,
      "learning_rate": 1.9078432314466665e-05,
      "loss": 0.0637,
      "step": 3975
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.6953601241111755,
      "learning_rate": 1.9076148269456576e-05,
      "loss": 0.0637,
      "step": 3980
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.8668028116226196,
      "learning_rate": 1.9073861534589006e-05,
      "loss": 0.0522,
      "step": 3985
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.9046851396560669,
      "learning_rate": 1.907157211054166e-05,
      "loss": 0.0541,
      "step": 3990
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.8799270391464233,
      "learning_rate": 1.906927999799305e-05,
      "loss": 0.0907,
      "step": 3995
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.908135175704956,
      "learning_rate": 1.906698519762247e-05,
      "loss": 0.0756,
      "step": 4000
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.7531886100769043,
      "learning_rate": 1.906468771011003e-05,
      "loss": 0.0706,
      "step": 4005
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.8982405662536621,
      "learning_rate": 1.9062387536136623e-05,
      "loss": 0.0897,
      "step": 4010
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.8098554611206055,
      "learning_rate": 1.9060084676383934e-05,
      "loss": 0.0654,
      "step": 4015
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.5767571926116943,
      "learning_rate": 1.905777913153446e-05,
      "loss": 0.0658,
      "step": 4020
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.9261816143989563,
      "learning_rate": 1.905547090227148e-05,
      "loss": 0.0552,
      "step": 4025
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.8063369989395142,
      "learning_rate": 1.905315998927908e-05,
      "loss": 0.0567,
      "step": 4030
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.8452303409576416,
      "learning_rate": 1.9050846393242127e-05,
      "loss": 0.0626,
      "step": 4035
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.695500373840332,
      "learning_rate": 1.9048530114846295e-05,
      "loss": 0.0756,
      "step": 4040
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.6767045855522156,
      "learning_rate": 1.904621115477805e-05,
      "loss": 0.0648,
      "step": 4045
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.8446992635726929,
      "learning_rate": 1.9043889513724648e-05,
      "loss": 0.0806,
      "step": 4050
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.8408441543579102,
      "learning_rate": 1.904156519237415e-05,
      "loss": 0.0677,
      "step": 4055
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.7592993974685669,
      "learning_rate": 1.9039238191415395e-05,
      "loss": 0.0756,
      "step": 4060
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.6402657628059387,
      "learning_rate": 1.9036908511538036e-05,
      "loss": 0.0531,
      "step": 4065
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.7251437902450562,
      "learning_rate": 1.9034576153432504e-05,
      "loss": 0.079,
      "step": 4070
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.8086432814598083,
      "learning_rate": 1.903224111779003e-05,
      "loss": 0.0588,
      "step": 4075
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.5242979526519775,
      "learning_rate": 1.902990340530264e-05,
      "loss": 0.0706,
      "step": 4080
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.258223295211792,
      "learning_rate": 1.902756301666315e-05,
      "loss": 0.0768,
      "step": 4085
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.7568871378898621,
      "learning_rate": 1.902521995256517e-05,
      "loss": 0.0829,
      "step": 4090
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.726394534111023,
      "learning_rate": 1.90228742137031e-05,
      "loss": 0.0832,
      "step": 4095
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.7445221543312073,
      "learning_rate": 1.9020525800772148e-05,
      "loss": 0.0553,
      "step": 4100
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.864100992679596,
      "learning_rate": 1.9018174714468292e-05,
      "loss": 0.0836,
      "step": 4105
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.6084670424461365,
      "learning_rate": 1.9015820955488316e-05,
      "loss": 0.0665,
      "step": 4110
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.6526731252670288,
      "learning_rate": 1.9013464524529793e-05,
      "loss": 0.0595,
      "step": 4115
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.8703616261482239,
      "learning_rate": 1.9011105422291087e-05,
      "loss": 0.095,
      "step": 4120
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.6150163412094116,
      "learning_rate": 1.9008743649471356e-05,
      "loss": 0.0676,
      "step": 4125
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.7024520635604858,
      "learning_rate": 1.900637920677055e-05,
      "loss": 0.0615,
      "step": 4130
    },
    {
      "epoch": 1.29,
      "grad_norm": 3.2384016513824463,
      "learning_rate": 1.90040120948894e-05,
      "loss": 0.0737,
      "step": 4135
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.7834177017211914,
      "learning_rate": 1.900164231452945e-05,
      "loss": 0.0814,
      "step": 4140
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.411509245634079,
      "learning_rate": 1.8999269866393005e-05,
      "loss": 0.0487,
      "step": 4145
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.9382750988006592,
      "learning_rate": 1.899689475118319e-05,
      "loss": 0.0619,
      "step": 4150
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.5898329615592957,
      "learning_rate": 1.8994516969603903e-05,
      "loss": 0.0496,
      "step": 4155
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.8763757348060608,
      "learning_rate": 1.8992136522359837e-05,
      "loss": 0.0761,
      "step": 4160
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.7113845348358154,
      "learning_rate": 1.898975341015647e-05,
      "loss": 0.0684,
      "step": 4165
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.5834293365478516,
      "learning_rate": 1.8987367633700076e-05,
      "loss": 0.0569,
      "step": 4170
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.7183889150619507,
      "learning_rate": 1.8984979193697727e-05,
      "loss": 0.059,
      "step": 4175
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.8346315622329712,
      "learning_rate": 1.8982588090857257e-05,
      "loss": 0.0603,
      "step": 4180
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.5111814141273499,
      "learning_rate": 1.8980194325887323e-05,
      "loss": 0.0601,
      "step": 4185
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.2758036851882935,
      "learning_rate": 1.8977797899497343e-05,
      "loss": 0.0487,
      "step": 4190
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.698638916015625,
      "learning_rate": 1.897539881239754e-05,
      "loss": 0.064,
      "step": 4195
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.6919531226158142,
      "learning_rate": 1.897299706529892e-05,
      "loss": 0.0556,
      "step": 4200
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.2566674947738647,
      "learning_rate": 1.8970592658913278e-05,
      "loss": 0.0821,
      "step": 4205
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.5100722312927246,
      "learning_rate": 1.8968185593953195e-05,
      "loss": 0.0447,
      "step": 4210
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.5391504168510437,
      "learning_rate": 1.8965775871132047e-05,
      "loss": 0.0802,
      "step": 4215
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.5908461213111877,
      "learning_rate": 1.8963363491163985e-05,
      "loss": 0.0927,
      "step": 4220
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.5374833345413208,
      "learning_rate": 1.8960948454763964e-05,
      "loss": 0.0561,
      "step": 4225
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.8269191384315491,
      "learning_rate": 1.8958530762647714e-05,
      "loss": 0.0835,
      "step": 4230
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.6975058317184448,
      "learning_rate": 1.8956110415531755e-05,
      "loss": 0.0584,
      "step": 4235
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.0302953720092773,
      "learning_rate": 1.8953687414133394e-05,
      "loss": 0.0698,
      "step": 4240
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.6055700778961182,
      "learning_rate": 1.8951261759170724e-05,
      "loss": 0.0631,
      "step": 4245
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.6735210418701172,
      "learning_rate": 1.8948833451362632e-05,
      "loss": 0.0633,
      "step": 4250
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.7031700611114502,
      "learning_rate": 1.8946402491428778e-05,
      "loss": 0.0619,
      "step": 4255
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.5998573899269104,
      "learning_rate": 1.8943968880089617e-05,
      "loss": 0.0671,
      "step": 4260
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.5515586137771606,
      "learning_rate": 1.8941532618066385e-05,
      "loss": 0.0569,
      "step": 4265
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.8202019333839417,
      "learning_rate": 1.893909370608111e-05,
      "loss": 0.0752,
      "step": 4270
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.8019958734512329,
      "learning_rate": 1.89366521448566e-05,
      "loss": 0.0595,
      "step": 4275
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.5395796298980713,
      "learning_rate": 1.8934207935116447e-05,
      "loss": 0.0788,
      "step": 4280
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.8099183440208435,
      "learning_rate": 1.8931761077585037e-05,
      "loss": 0.0634,
      "step": 4285
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.6752781867980957,
      "learning_rate": 1.8929311572987528e-05,
      "loss": 0.0872,
      "step": 4290
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.7430433034896851,
      "learning_rate": 1.892685942204987e-05,
      "loss": 0.0779,
      "step": 4295
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.45912253856658936,
      "learning_rate": 1.89244046254988e-05,
      "loss": 0.0703,
      "step": 4300
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.6058170199394226,
      "learning_rate": 1.892194718406183e-05,
      "loss": 0.0602,
      "step": 4305
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.5597983002662659,
      "learning_rate": 1.8919487098467263e-05,
      "loss": 0.0879,
      "step": 4310
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.7414512038230896,
      "learning_rate": 1.891702436944418e-05,
      "loss": 0.0754,
      "step": 4315
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.7263529300689697,
      "learning_rate": 1.891455899772246e-05,
      "loss": 0.0581,
      "step": 4320
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.49107396602630615,
      "learning_rate": 1.891209098403274e-05,
      "loss": 0.0576,
      "step": 4325
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.8759298324584961,
      "learning_rate": 1.8909620329106463e-05,
      "loss": 0.0685,
      "step": 4330
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.5367811918258667,
      "learning_rate": 1.8907147033675847e-05,
      "loss": 0.0616,
      "step": 4335
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.6136344075202942,
      "learning_rate": 1.8904671098473885e-05,
      "loss": 0.059,
      "step": 4340
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.7203885912895203,
      "learning_rate": 1.890219252423436e-05,
      "loss": 0.0824,
      "step": 4345
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.7837348580360413,
      "learning_rate": 1.8899711311691843e-05,
      "loss": 0.0721,
      "step": 4350
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.9603199362754822,
      "learning_rate": 1.8897227461581673e-05,
      "loss": 0.0773,
      "step": 4355
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.6584271788597107,
      "learning_rate": 1.889474097463998e-05,
      "loss": 0.0582,
      "step": 4360
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.7230022549629211,
      "learning_rate": 1.889225185160367e-05,
      "loss": 0.0792,
      "step": 4365
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.5705015063285828,
      "learning_rate": 1.888976009321044e-05,
      "loss": 0.0603,
      "step": 4370
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.7777138352394104,
      "learning_rate": 1.8887265700198755e-05,
      "loss": 0.0595,
      "step": 4375
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.6518700122833252,
      "learning_rate": 1.8884768673307867e-05,
      "loss": 0.0489,
      "step": 4380
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.7976190447807312,
      "learning_rate": 1.8882269013277813e-05,
      "loss": 0.0698,
      "step": 4385
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.5520892143249512,
      "learning_rate": 1.8879766720849404e-05,
      "loss": 0.0872,
      "step": 4390
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.8835591673851013,
      "learning_rate": 1.8877261796764235e-05,
      "loss": 0.0721,
      "step": 4395
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.8427669405937195,
      "learning_rate": 1.8874754241764676e-05,
      "loss": 0.0889,
      "step": 4400
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.7144712805747986,
      "learning_rate": 1.887224405659388e-05,
      "loss": 0.0631,
      "step": 4405
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.4469461739063263,
      "learning_rate": 1.886973124199578e-05,
      "loss": 0.0715,
      "step": 4410
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.7447325587272644,
      "learning_rate": 1.886721579871509e-05,
      "loss": 0.0872,
      "step": 4415
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.6704967021942139,
      "learning_rate": 1.8864697727497297e-05,
      "loss": 0.0783,
      "step": 4420
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.5041611194610596,
      "learning_rate": 1.8862177029088674e-05,
      "loss": 0.0629,
      "step": 4425
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.5847105383872986,
      "learning_rate": 1.885965370423627e-05,
      "loss": 0.0529,
      "step": 4430
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.5569111108779907,
      "learning_rate": 1.8857127753687904e-05,
      "loss": 0.0537,
      "step": 4435
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.5260555744171143,
      "learning_rate": 1.885459917819219e-05,
      "loss": 0.0511,
      "step": 4440
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.490197092294693,
      "learning_rate": 1.8852067978498503e-05,
      "loss": 0.069,
      "step": 4445
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.7321019172668457,
      "learning_rate": 1.884953415535701e-05,
      "loss": 0.0814,
      "step": 4450
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.7469559907913208,
      "learning_rate": 1.8846997709518642e-05,
      "loss": 0.078,
      "step": 4455
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.76277095079422,
      "learning_rate": 1.884445864173512e-05,
      "loss": 0.0596,
      "step": 4460
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.583543062210083,
      "learning_rate": 1.8841916952758933e-05,
      "loss": 0.0636,
      "step": 4465
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.7911220788955688,
      "learning_rate": 1.883937264334335e-05,
      "loss": 0.0634,
      "step": 4470
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.6241563558578491,
      "learning_rate": 1.8836825714242417e-05,
      "loss": 0.0612,
      "step": 4475
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.1762306690216064,
      "learning_rate": 1.883427616621096e-05,
      "loss": 0.0703,
      "step": 4480
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.086243987083435,
      "learning_rate": 1.883172400000457e-05,
      "loss": 0.0843,
      "step": 4485
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.8462454676628113,
      "learning_rate": 1.8829169216379623e-05,
      "loss": 0.0776,
      "step": 4490
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.5264838933944702,
      "learning_rate": 1.8826611816093274e-05,
      "loss": 0.0611,
      "step": 4495
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.6226249933242798,
      "learning_rate": 1.8824051799903442e-05,
      "loss": 0.0637,
      "step": 4500
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.830549418926239,
      "learning_rate": 1.8821489168568832e-05,
      "loss": 0.0577,
      "step": 4505
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.9017519354820251,
      "learning_rate": 1.8818923922848918e-05,
      "loss": 0.0698,
      "step": 4510
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.8163900971412659,
      "learning_rate": 1.881635606350395e-05,
      "loss": 0.0724,
      "step": 4515
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.715849757194519,
      "learning_rate": 1.8813785591294952e-05,
      "loss": 0.056,
      "step": 4520
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.6897075176239014,
      "learning_rate": 1.8811212506983724e-05,
      "loss": 0.0697,
      "step": 4525
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.8066615462303162,
      "learning_rate": 1.8808636811332843e-05,
      "loss": 0.076,
      "step": 4530
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.8620597720146179,
      "learning_rate": 1.880605850510565e-05,
      "loss": 0.0802,
      "step": 4535
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.921575665473938,
      "learning_rate": 1.8803477589066272e-05,
      "loss": 0.0776,
      "step": 4540
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.7957607507705688,
      "learning_rate": 1.8800894063979602e-05,
      "loss": 0.0997,
      "step": 4545
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.5893496870994568,
      "learning_rate": 1.8798307930611303e-05,
      "loss": 0.0613,
      "step": 4550
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.9337874054908752,
      "learning_rate": 1.8795719189727824e-05,
      "loss": 0.07,
      "step": 4555
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.7042413353919983,
      "learning_rate": 1.8793127842096373e-05,
      "loss": 0.0813,
      "step": 4560
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.7177154421806335,
      "learning_rate": 1.8790533888484937e-05,
      "loss": 0.0447,
      "step": 4565
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.9277735948562622,
      "learning_rate": 1.8787937329662273e-05,
      "loss": 0.0787,
      "step": 4570
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.850438117980957,
      "learning_rate": 1.8785338166397917e-05,
      "loss": 0.0487,
      "step": 4575
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.6746247410774231,
      "learning_rate": 1.878273639946216e-05,
      "loss": 0.0717,
      "step": 4580
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.7712878584861755,
      "learning_rate": 1.8780132029626093e-05,
      "loss": 0.0586,
      "step": 4585
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.8856165409088135,
      "learning_rate": 1.8777525057661547e-05,
      "loss": 0.0745,
      "step": 4590
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.8087002038955688,
      "learning_rate": 1.8774915484341147e-05,
      "loss": 0.082,
      "step": 4595
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.7779855728149414,
      "learning_rate": 1.8772303310438275e-05,
      "loss": 0.0775,
      "step": 4600
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.7559125423431396,
      "learning_rate": 1.8769688536727094e-05,
      "loss": 0.0714,
      "step": 4605
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.6574954986572266,
      "learning_rate": 1.8767071163982525e-05,
      "loss": 0.0695,
      "step": 4610
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.8335013389587402,
      "learning_rate": 1.8764451192980278e-05,
      "loss": 0.0564,
      "step": 4615
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.6353960633277893,
      "learning_rate": 1.8761828624496816e-05,
      "loss": 0.0765,
      "step": 4620
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.7106569409370422,
      "learning_rate": 1.8759203459309375e-05,
      "loss": 0.0714,
      "step": 4625
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.7167492508888245,
      "learning_rate": 1.8756575698195968e-05,
      "loss": 0.0664,
      "step": 4630
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.6576385498046875,
      "learning_rate": 1.8753945341935376e-05,
      "loss": 0.0796,
      "step": 4635
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.501663088798523,
      "learning_rate": 1.875131239130714e-05,
      "loss": 0.0755,
      "step": 4640
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.831375777721405,
      "learning_rate": 1.8748676847091575e-05,
      "loss": 0.0801,
      "step": 4645
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.9495288133621216,
      "learning_rate": 1.874603871006977e-05,
      "loss": 0.0841,
      "step": 4650
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.5677212476730347,
      "learning_rate": 1.8743397981023574e-05,
      "loss": 0.0757,
      "step": 4655
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.6868023872375488,
      "learning_rate": 1.8740754660735612e-05,
      "loss": 0.0695,
      "step": 4660
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.635217010974884,
      "learning_rate": 1.873810874998927e-05,
      "loss": 0.0661,
      "step": 4665
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.582550048828125,
      "learning_rate": 1.8735460249568708e-05,
      "loss": 0.047,
      "step": 4670
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.8971719741821289,
      "learning_rate": 1.8732809160258846e-05,
      "loss": 0.0812,
      "step": 4675
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.7387098073959351,
      "learning_rate": 1.8730155482845374e-05,
      "loss": 0.0844,
      "step": 4680
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.7902190685272217,
      "learning_rate": 1.872749921811476e-05,
      "loss": 0.0813,
      "step": 4685
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.6167246103286743,
      "learning_rate": 1.8724840366854218e-05,
      "loss": 0.0868,
      "step": 4690
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.6898187398910522,
      "learning_rate": 1.8722178929851748e-05,
      "loss": 0.0692,
      "step": 4695
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.6871895790100098,
      "learning_rate": 1.8719514907896104e-05,
      "loss": 0.0718,
      "step": 4700
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.7227907776832581,
      "learning_rate": 1.871684830177681e-05,
      "loss": 0.0521,
      "step": 4705
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.124375820159912,
      "learning_rate": 1.8714179112284162e-05,
      "loss": 0.091,
      "step": 4710
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.9651860594749451,
      "learning_rate": 1.871150734020921e-05,
      "loss": 0.0891,
      "step": 4715
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.6683818101882935,
      "learning_rate": 1.8708832986343775e-05,
      "loss": 0.049,
      "step": 4720
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.743218719959259,
      "learning_rate": 1.8706156051480447e-05,
      "loss": 0.0792,
      "step": 4725
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.8588913083076477,
      "learning_rate": 1.870347653641257e-05,
      "loss": 0.0704,
      "step": 4730
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.6899648308753967,
      "learning_rate": 1.8700794441934272e-05,
      "loss": 0.073,
      "step": 4735
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.5917913913726807,
      "learning_rate": 1.869810976884042e-05,
      "loss": 0.0985,
      "step": 4740
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.697803795337677,
      "learning_rate": 1.869542251792667e-05,
      "loss": 0.0535,
      "step": 4745
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.6701207160949707,
      "learning_rate": 1.8692732689989423e-05,
      "loss": 0.0529,
      "step": 4750
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.58310467004776,
      "learning_rate": 1.8690040285825852e-05,
      "loss": 0.0826,
      "step": 4755
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.6153090000152588,
      "learning_rate": 1.8687345306233896e-05,
      "loss": 0.0907,
      "step": 4760
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.6057466864585876,
      "learning_rate": 1.8684647752012255e-05,
      "loss": 0.0619,
      "step": 4765
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.7822586894035339,
      "learning_rate": 1.8681947623960387e-05,
      "loss": 0.0758,
      "step": 4770
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.0661760568618774,
      "learning_rate": 1.8679244922878516e-05,
      "loss": 0.0641,
      "step": 4775
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.7023093104362488,
      "learning_rate": 1.8676539649567636e-05,
      "loss": 0.0697,
      "step": 4780
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.6697647571563721,
      "learning_rate": 1.867383180482949e-05,
      "loss": 0.0672,
      "step": 4785
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.7119118571281433,
      "learning_rate": 1.8671121389466592e-05,
      "loss": 0.0619,
      "step": 4790
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.7365993857383728,
      "learning_rate": 1.866840840428222e-05,
      "loss": 0.0655,
      "step": 4795
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.8320115208625793,
      "learning_rate": 1.86656928500804e-05,
      "loss": 0.0563,
      "step": 4800
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.1454790830612183,
      "learning_rate": 1.866297472766594e-05,
      "loss": 0.0685,
      "step": 4805
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.6646517515182495,
      "learning_rate": 1.866025403784439e-05,
      "loss": 0.0621,
      "step": 4810
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.848318874835968,
      "learning_rate": 1.8657530781422067e-05,
      "loss": 0.0708,
      "step": 4815
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.9369855523109436,
      "learning_rate": 1.865480495920606e-05,
      "loss": 0.1123,
      "step": 4820
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.175233006477356,
      "learning_rate": 1.8652076572004197e-05,
      "loss": 0.0645,
      "step": 4825
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.8523390293121338,
      "learning_rate": 1.8649345620625087e-05,
      "loss": 0.0617,
      "step": 4830
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.7075837254524231,
      "learning_rate": 1.8646612105878087e-05,
      "loss": 0.0706,
      "step": 4835
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.7256972193717957,
      "learning_rate": 1.8643876028573312e-05,
      "loss": 0.0432,
      "step": 4840
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.8678048253059387,
      "learning_rate": 1.8641137389521648e-05,
      "loss": 0.0854,
      "step": 4845
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.7625493407249451,
      "learning_rate": 1.8638396189534728e-05,
      "loss": 0.0645,
      "step": 4850
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.0163854360580444,
      "learning_rate": 1.8635652429424956e-05,
      "loss": 0.0796,
      "step": 4855
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.7053332328796387,
      "learning_rate": 1.863290611000548e-05,
      "loss": 0.0609,
      "step": 4860
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.7304905652999878,
      "learning_rate": 1.8630157232090215e-05,
      "loss": 0.0733,
      "step": 4865
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.7250766754150391,
      "learning_rate": 1.8627405796493843e-05,
      "loss": 0.0482,
      "step": 4870
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.680020272731781,
      "learning_rate": 1.8624651804031782e-05,
      "loss": 0.0683,
      "step": 4875
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.7620866894721985,
      "learning_rate": 1.862189525552023e-05,
      "loss": 0.0884,
      "step": 4880
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.6260718107223511,
      "learning_rate": 1.861913615177613e-05,
      "loss": 0.0577,
      "step": 4885
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.621261477470398,
      "learning_rate": 1.8616374493617184e-05,
      "loss": 0.0598,
      "step": 4890
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.8694585561752319,
      "learning_rate": 1.8613610281861853e-05,
      "loss": 0.0826,
      "step": 4895
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.922036349773407,
      "learning_rate": 1.8610843517329353e-05,
      "loss": 0.0644,
      "step": 4900
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.8628506064414978,
      "learning_rate": 1.8608074200839658e-05,
      "loss": 0.0769,
      "step": 4905
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.8055910468101501,
      "learning_rate": 1.8605302333213497e-05,
      "loss": 0.0643,
      "step": 4910
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.6417886018753052,
      "learning_rate": 1.860252791527236e-05,
      "loss": 0.0435,
      "step": 4915
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.8504335880279541,
      "learning_rate": 1.859975094783849e-05,
      "loss": 0.0696,
      "step": 4920
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.3809876441955566,
      "learning_rate": 1.8596971431734878e-05,
      "loss": 0.0743,
      "step": 4925
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.6176561713218689,
      "learning_rate": 1.8594189367785276e-05,
      "loss": 0.0555,
      "step": 4930
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.7021178603172302,
      "learning_rate": 1.85914047568142e-05,
      "loss": 0.0572,
      "step": 4935
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.7485069632530212,
      "learning_rate": 1.8588617599646906e-05,
      "loss": 0.0726,
      "step": 4940
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.8081046342849731,
      "learning_rate": 1.8585827897109415e-05,
      "loss": 0.0476,
      "step": 4945
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.890154242515564,
      "learning_rate": 1.8583035650028497e-05,
      "loss": 0.0693,
      "step": 4950
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.524368166923523,
      "learning_rate": 1.8580240859231677e-05,
      "loss": 0.0601,
      "step": 4955
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.6123557686805725,
      "learning_rate": 1.8577443525547237e-05,
      "loss": 0.0771,
      "step": 4960
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.5457882285118103,
      "learning_rate": 1.857464364980421e-05,
      "loss": 0.0533,
      "step": 4965
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.6672859787940979,
      "learning_rate": 1.857184123283238e-05,
      "loss": 0.0605,
      "step": 4970
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.5480169653892517,
      "learning_rate": 1.8569036275462287e-05,
      "loss": 0.0588,
      "step": 4975
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.2137045860290527,
      "learning_rate": 1.8566228778525226e-05,
      "loss": 0.084,
      "step": 4980
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.751510500907898,
      "learning_rate": 1.856341874285324e-05,
      "loss": 0.0779,
      "step": 4985
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.9870890378952026,
      "learning_rate": 1.856060616927913e-05,
      "loss": 0.0772,
      "step": 4990
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.8300113677978516,
      "learning_rate": 1.8557791058636444e-05,
      "loss": 0.0739,
      "step": 4995
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.845734715461731,
      "learning_rate": 1.8554973411759484e-05,
      "loss": 0.0825,
      "step": 5000
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.8794209361076355,
      "learning_rate": 1.8552153229483302e-05,
      "loss": 0.0462,
      "step": 5005
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.5634300708770752,
      "learning_rate": 1.8549330512643707e-05,
      "loss": 0.0727,
      "step": 5010
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.6634950041770935,
      "learning_rate": 1.8546505262077245e-05,
      "loss": 0.0674,
      "step": 5015
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.9048967957496643,
      "learning_rate": 1.8543677478621236e-05,
      "loss": 0.0814,
      "step": 5020
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.0099159479141235,
      "learning_rate": 1.854084716311373e-05,
      "loss": 0.0674,
      "step": 5025
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.7578509449958801,
      "learning_rate": 1.8538014316393534e-05,
      "loss": 0.0722,
      "step": 5030
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.8153797388076782,
      "learning_rate": 1.853517893930021e-05,
      "loss": 0.0811,
      "step": 5035
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.6564578413963318,
      "learning_rate": 1.8532341032674063e-05,
      "loss": 0.0535,
      "step": 5040
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.7237277030944824,
      "learning_rate": 1.8529500597356155e-05,
      "loss": 0.0774,
      "step": 5045
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.7941708564758301,
      "learning_rate": 1.8526657634188288e-05,
      "loss": 0.0844,
      "step": 5050
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.2459404468536377,
      "learning_rate": 1.852381214401302e-05,
      "loss": 0.1087,
      "step": 5055
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.005991816520691,
      "learning_rate": 1.8520964127673658e-05,
      "loss": 0.087,
      "step": 5060
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.7758017182350159,
      "learning_rate": 1.8518113586014253e-05,
      "loss": 0.095,
      "step": 5065
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.7437559366226196,
      "learning_rate": 1.851526051987961e-05,
      "loss": 0.0649,
      "step": 5070
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.8242641687393188,
      "learning_rate": 1.851240493011528e-05,
      "loss": 0.0678,
      "step": 5075
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.0162962675094604,
      "learning_rate": 1.8509546817567553e-05,
      "loss": 0.0714,
      "step": 5080
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.7507639527320862,
      "learning_rate": 1.850668618308349e-05,
      "loss": 0.0783,
      "step": 5085
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.0634666681289673,
      "learning_rate": 1.850382302751087e-05,
      "loss": 0.0815,
      "step": 5090
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.8045139312744141,
      "learning_rate": 1.8500957351698244e-05,
      "loss": 0.0585,
      "step": 5095
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.7873166799545288,
      "learning_rate": 1.8498089156494895e-05,
      "loss": 0.0607,
      "step": 5100
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.6100918650627136,
      "learning_rate": 1.849521844275085e-05,
      "loss": 0.0473,
      "step": 5105
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.6852316856384277,
      "learning_rate": 1.8492345211316908e-05,
      "loss": 0.066,
      "step": 5110
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.80754554271698,
      "learning_rate": 1.848946946304458e-05,
      "loss": 0.0838,
      "step": 5115
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.8571280241012573,
      "learning_rate": 1.8486591198786143e-05,
      "loss": 0.084,
      "step": 5120
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.5786952972412109,
      "learning_rate": 1.8483710419394616e-05,
      "loss": 0.0704,
      "step": 5125
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.5984448194503784,
      "learning_rate": 1.8480827125723764e-05,
      "loss": 0.0946,
      "step": 5130
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.7940416932106018,
      "learning_rate": 1.8477941318628093e-05,
      "loss": 0.0779,
      "step": 5135
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.4904640913009644,
      "learning_rate": 1.8475052998962856e-05,
      "loss": 0.0994,
      "step": 5140
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.8121379017829895,
      "learning_rate": 1.8472162167584058e-05,
      "loss": 0.0734,
      "step": 5145
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.5815505981445312,
      "learning_rate": 1.8469268825348435e-05,
      "loss": 0.0556,
      "step": 5150
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.8653185963630676,
      "learning_rate": 1.8466372973113476e-05,
      "loss": 0.0561,
      "step": 5155
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.5899277329444885,
      "learning_rate": 1.8463474611737412e-05,
      "loss": 0.0752,
      "step": 5160
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.6971949338912964,
      "learning_rate": 1.8460573742079216e-05,
      "loss": 0.061,
      "step": 5165
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.7547398805618286,
      "learning_rate": 1.845767036499861e-05,
      "loss": 0.1052,
      "step": 5170
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.8991611003875732,
      "learning_rate": 1.845476448135605e-05,
      "loss": 0.0631,
      "step": 5175
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.0087960958480835,
      "learning_rate": 1.8451856092012743e-05,
      "loss": 0.0643,
      "step": 5180
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.7402775287628174,
      "learning_rate": 1.8448945197830635e-05,
      "loss": 0.0755,
      "step": 5185
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.6131215691566467,
      "learning_rate": 1.8446031799672412e-05,
      "loss": 0.0821,
      "step": 5190
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.7232556939125061,
      "learning_rate": 1.8443115898401505e-05,
      "loss": 0.0816,
      "step": 5195
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.876368522644043,
      "learning_rate": 1.844019749488209e-05,
      "loss": 0.0711,
      "step": 5200
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.8438897728919983,
      "learning_rate": 1.843727658997908e-05,
      "loss": 0.0626,
      "step": 5205
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.7738611102104187,
      "learning_rate": 1.843435318455813e-05,
      "loss": 0.0632,
      "step": 5210
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.8397727012634277,
      "learning_rate": 1.8431427279485638e-05,
      "loss": 0.07,
      "step": 5215
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.9025486707687378,
      "learning_rate": 1.842849887562873e-05,
      "loss": 0.0765,
      "step": 5220
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.8906568884849548,
      "learning_rate": 1.8425567973855306e-05,
      "loss": 0.083,
      "step": 5225
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.8898297548294067,
      "learning_rate": 1.842263457503397e-05,
      "loss": 0.0582,
      "step": 5230
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.7294970750808716,
      "learning_rate": 1.8419698680034076e-05,
      "loss": 0.0804,
      "step": 5235
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.635528028011322,
      "learning_rate": 1.8416760289725736e-05,
      "loss": 0.0662,
      "step": 5240
    },
    {
      "epoch": 1.64,
      "grad_norm": 2.0224263668060303,
      "learning_rate": 1.8413819404979776e-05,
      "loss": 0.0635,
      "step": 5245
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.8897799253463745,
      "learning_rate": 1.841087602666778e-05,
      "loss": 0.077,
      "step": 5250
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.9975351095199585,
      "learning_rate": 1.8407930155662056e-05,
      "loss": 0.0896,
      "step": 5255
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.5616067051887512,
      "learning_rate": 1.8404981792835665e-05,
      "loss": 0.0556,
      "step": 5260
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.4890785217285156,
      "learning_rate": 1.840262130905084e-05,
      "loss": 0.0705,
      "step": 5265
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.7033029198646545,
      "learning_rate": 1.83996684631497e-05,
      "loss": 0.0816,
      "step": 5270
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.6858252882957458,
      "learning_rate": 1.8396713127876373e-05,
      "loss": 0.0723,
      "step": 5275
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.7298487424850464,
      "learning_rate": 1.839375530410672e-05,
      "loss": 0.0748,
      "step": 5280
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.9825119376182556,
      "learning_rate": 1.8390794992717333e-05,
      "loss": 0.055,
      "step": 5285
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.856184720993042,
      "learning_rate": 1.838783219458555e-05,
      "loss": 0.0636,
      "step": 5290
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.7891317009925842,
      "learning_rate": 1.8384866910589442e-05,
      "loss": 0.0405,
      "step": 5295
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.9762969017028809,
      "learning_rate": 1.8381899141607817e-05,
      "loss": 0.0707,
      "step": 5300
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.5405159592628479,
      "learning_rate": 1.8378928888520216e-05,
      "loss": 0.0639,
      "step": 5305
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.8812373876571655,
      "learning_rate": 1.8375956152206926e-05,
      "loss": 0.0606,
      "step": 5310
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.6692848205566406,
      "learning_rate": 1.8372980933548957e-05,
      "loss": 0.0772,
      "step": 5315
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.2220298051834106,
      "learning_rate": 1.8370003233428067e-05,
      "loss": 0.0653,
      "step": 5320
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.8167205452919006,
      "learning_rate": 1.836702305272674e-05,
      "loss": 0.0746,
      "step": 5325
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.952875018119812,
      "learning_rate": 1.8364040392328197e-05,
      "loss": 0.072,
      "step": 5330
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.9271166324615479,
      "learning_rate": 1.83610552531164e-05,
      "loss": 0.0542,
      "step": 5335
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.6192010641098022,
      "learning_rate": 1.835806763597604e-05,
      "loss": 0.073,
      "step": 5340
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.7979380488395691,
      "learning_rate": 1.8355077541792543e-05,
      "loss": 0.0774,
      "step": 5345
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.6753906011581421,
      "learning_rate": 1.835208497145207e-05,
      "loss": 0.0733,
      "step": 5350
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.069570779800415,
      "learning_rate": 1.8349089925841516e-05,
      "loss": 0.0671,
      "step": 5355
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.5559250712394714,
      "learning_rate": 1.8346092405848513e-05,
      "loss": 0.0641,
      "step": 5360
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.2465052604675293,
      "learning_rate": 1.8343092412361414e-05,
      "loss": 0.0676,
      "step": 5365
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.8193307518959045,
      "learning_rate": 1.834008994626932e-05,
      "loss": 0.0482,
      "step": 5370
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.9308450818061829,
      "learning_rate": 1.8337085008462056e-05,
      "loss": 0.0897,
      "step": 5375
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.6304412484169006,
      "learning_rate": 1.8334077599830188e-05,
      "loss": 0.0767,
      "step": 5380
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.5548502206802368,
      "learning_rate": 1.8331067721265e-05,
      "loss": 0.0664,
      "step": 5385
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.9302047491073608,
      "learning_rate": 1.8328055373658517e-05,
      "loss": 0.0734,
      "step": 5390
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.5483214259147644,
      "learning_rate": 1.83250405579035e-05,
      "loss": 0.1012,
      "step": 5395
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.8517399430274963,
      "learning_rate": 1.8322023274893435e-05,
      "loss": 0.0836,
      "step": 5400
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.6601105332374573,
      "learning_rate": 1.8319003525522537e-05,
      "loss": 0.0459,
      "step": 5405
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.583982527256012,
      "learning_rate": 1.8315981310685756e-05,
      "loss": 0.0694,
      "step": 5410
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.3079158067703247,
      "learning_rate": 1.831295663127878e-05,
      "loss": 0.0832,
      "step": 5415
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.6640372276306152,
      "learning_rate": 1.8309929488198012e-05,
      "loss": 0.0543,
      "step": 5420
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.2028993368148804,
      "learning_rate": 1.8306899882340597e-05,
      "loss": 0.0812,
      "step": 5425
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.8884703516960144,
      "learning_rate": 1.8303867814604402e-05,
      "loss": 0.0768,
      "step": 5430
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.8457720875740051,
      "learning_rate": 1.830083328588803e-05,
      "loss": 0.0662,
      "step": 5435
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.8695236444473267,
      "learning_rate": 1.8297796297090808e-05,
      "loss": 0.0609,
      "step": 5440
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.763415515422821,
      "learning_rate": 1.82947568491128e-05,
      "loss": 0.0654,
      "step": 5445
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.3075530529022217,
      "learning_rate": 1.8291714942854787e-05,
      "loss": 0.0789,
      "step": 5450
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.7080858945846558,
      "learning_rate": 1.828867057921829e-05,
      "loss": 0.0663,
      "step": 5455
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.6674808263778687,
      "learning_rate": 1.828562375910555e-05,
      "loss": 0.1035,
      "step": 5460
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.0941526889801025,
      "learning_rate": 1.8282574483419544e-05,
      "loss": 0.0645,
      "step": 5465
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.6409357190132141,
      "learning_rate": 1.8279522753063966e-05,
      "loss": 0.0806,
      "step": 5470
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.7128939032554626,
      "learning_rate": 1.827646856894325e-05,
      "loss": 0.0675,
      "step": 5475
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.8401421308517456,
      "learning_rate": 1.827341193196255e-05,
      "loss": 0.0792,
      "step": 5480
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.0783287286758423,
      "learning_rate": 1.8270352843027744e-05,
      "loss": 0.0746,
      "step": 5485
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.5906087160110474,
      "learning_rate": 1.8267291303045443e-05,
      "loss": 0.0572,
      "step": 5490
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.6586350798606873,
      "learning_rate": 1.826422731292298e-05,
      "loss": 0.0708,
      "step": 5495
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.787606418132782,
      "learning_rate": 1.826116087356842e-05,
      "loss": 0.0666,
      "step": 5500
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.7802187204360962,
      "learning_rate": 1.8258091985890547e-05,
      "loss": 0.0648,
      "step": 5505
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.8204883337020874,
      "learning_rate": 1.825502065079887e-05,
      "loss": 0.0664,
      "step": 5510
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.5513288378715515,
      "learning_rate": 1.8251946869203635e-05,
      "loss": 0.0661,
      "step": 5515
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.000407338142395,
      "learning_rate": 1.82488706420158e-05,
      "loss": 0.0769,
      "step": 5520
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.827264666557312,
      "learning_rate": 1.824579197014705e-05,
      "loss": 0.0863,
      "step": 5525
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.0817997455596924,
      "learning_rate": 1.8242710854509807e-05,
      "loss": 0.083,
      "step": 5530
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.7575589418411255,
      "learning_rate": 1.8239627296017196e-05,
      "loss": 0.0823,
      "step": 5535
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.9375240206718445,
      "learning_rate": 1.8236541295583083e-05,
      "loss": 0.0782,
      "step": 5540
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.5978908538818359,
      "learning_rate": 1.8233452854122054e-05,
      "loss": 0.0817,
      "step": 5545
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.9602009654045105,
      "learning_rate": 1.823036197254941e-05,
      "loss": 0.0771,
      "step": 5550
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.5891567468643188,
      "learning_rate": 1.822726865178119e-05,
      "loss": 0.0677,
      "step": 5555
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.8885820508003235,
      "learning_rate": 1.822417289273414e-05,
      "loss": 0.0575,
      "step": 5560
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.7974045276641846,
      "learning_rate": 1.822107469632574e-05,
      "loss": 0.0884,
      "step": 5565
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.5491849184036255,
      "learning_rate": 1.8217974063474196e-05,
      "loss": 0.0665,
      "step": 5570
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.8097928166389465,
      "learning_rate": 1.821487099509841e-05,
      "loss": 0.0685,
      "step": 5575
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.8077875971794128,
      "learning_rate": 1.8211765492118046e-05,
      "loss": 0.0668,
      "step": 5580
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.768814206123352,
      "learning_rate": 1.820865755545345e-05,
      "loss": 0.0713,
      "step": 5585
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.9624421000480652,
      "learning_rate": 1.8205547186025724e-05,
      "loss": 0.0679,
      "step": 5590
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.6452382206916809,
      "learning_rate": 1.8202434384756656e-05,
      "loss": 0.0722,
      "step": 5595
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.7202498912811279,
      "learning_rate": 1.8199319152568786e-05,
      "loss": 0.0656,
      "step": 5600
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.7296916842460632,
      "learning_rate": 1.819620149038536e-05,
      "loss": 0.0849,
      "step": 5605
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.9542629718780518,
      "learning_rate": 1.819308139913034e-05,
      "loss": 0.0948,
      "step": 5610
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.7899646759033203,
      "learning_rate": 1.8189958879728418e-05,
      "loss": 0.0643,
      "step": 5615
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.6902501583099365,
      "learning_rate": 1.8186833933105e-05,
      "loss": 0.0612,
      "step": 5620
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.8489237427711487,
      "learning_rate": 1.8183706560186212e-05,
      "loss": 0.0723,
      "step": 5625
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.46526649594306946,
      "learning_rate": 1.81805767618989e-05,
      "loss": 0.0549,
      "step": 5630
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.7680943608283997,
      "learning_rate": 1.8177444539170627e-05,
      "loss": 0.0777,
      "step": 5635
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.6338277459144592,
      "learning_rate": 1.817430989292968e-05,
      "loss": 0.0612,
      "step": 5640
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.8167020082473755,
      "learning_rate": 1.8171172824105056e-05,
      "loss": 0.0579,
      "step": 5645
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.9066957235336304,
      "learning_rate": 1.8168033333626474e-05,
      "loss": 0.0729,
      "step": 5650
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.146350622177124,
      "learning_rate": 1.8164891422424372e-05,
      "loss": 0.1003,
      "step": 5655
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.7212271690368652,
      "learning_rate": 1.8161747091429903e-05,
      "loss": 0.0733,
      "step": 5660
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.6951694488525391,
      "learning_rate": 1.8158600341574946e-05,
      "loss": 0.0605,
      "step": 5665
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.7327526211738586,
      "learning_rate": 1.8155451173792077e-05,
      "loss": 0.0617,
      "step": 5670
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.4549367427825928,
      "learning_rate": 1.815229958901461e-05,
      "loss": 0.0796,
      "step": 5675
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.9679293036460876,
      "learning_rate": 1.8149145588176557e-05,
      "loss": 0.0842,
      "step": 5680
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.6489354372024536,
      "learning_rate": 1.8145989172212666e-05,
      "loss": 0.0769,
      "step": 5685
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.8843717575073242,
      "learning_rate": 1.8142830342058383e-05,
      "loss": 0.0787,
      "step": 5690
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.8279662132263184,
      "learning_rate": 1.8139669098649876e-05,
      "loss": 0.0563,
      "step": 5695
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.659649670124054,
      "learning_rate": 1.8136505442924033e-05,
      "loss": 0.0679,
      "step": 5700
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.9309002161026001,
      "learning_rate": 1.813333937581845e-05,
      "loss": 0.0748,
      "step": 5705
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.8285350203514099,
      "learning_rate": 1.813017089827144e-05,
      "loss": 0.0567,
      "step": 5710
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.6510990858078003,
      "learning_rate": 1.812700001122203e-05,
      "loss": 0.0616,
      "step": 5715
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.7334113717079163,
      "learning_rate": 1.8123826715609963e-05,
      "loss": 0.0577,
      "step": 5720
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.475523978471756,
      "learning_rate": 1.8120651012375694e-05,
      "loss": 0.0654,
      "step": 5725
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.5141718983650208,
      "learning_rate": 1.811747290246039e-05,
      "loss": 0.0633,
      "step": 5730
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.7106728553771973,
      "learning_rate": 1.8114292386805935e-05,
      "loss": 0.0583,
      "step": 5735
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.6758169531822205,
      "learning_rate": 1.8111109466354926e-05,
      "loss": 0.0673,
      "step": 5740
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.7288673520088196,
      "learning_rate": 1.810792414205067e-05,
      "loss": 0.065,
      "step": 5745
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.0398705005645752,
      "learning_rate": 1.8104736414837183e-05,
      "loss": 0.0903,
      "step": 5750
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.5896938443183899,
      "learning_rate": 1.81015462856592e-05,
      "loss": 0.0636,
      "step": 5755
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.6056891679763794,
      "learning_rate": 1.809835375546217e-05,
      "loss": 0.0599,
      "step": 5760
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.9472219347953796,
      "learning_rate": 1.8095158825192248e-05,
      "loss": 0.0695,
      "step": 5765
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.7288747429847717,
      "learning_rate": 1.8091961495796294e-05,
      "loss": 0.0733,
      "step": 5770
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5857644081115723,
      "learning_rate": 1.8088761768221887e-05,
      "loss": 0.0613,
      "step": 5775
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.1260770559310913,
      "learning_rate": 1.8085559643417326e-05,
      "loss": 0.082,
      "step": 5780
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5587033033370972,
      "learning_rate": 1.80823551223316e-05,
      "loss": 0.06,
      "step": 5785
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.8423185348510742,
      "learning_rate": 1.8079148205914423e-05,
      "loss": 0.0546,
      "step": 5790
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.0525543689727783,
      "learning_rate": 1.807593889511621e-05,
      "loss": 0.0681,
      "step": 5795
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.8642443418502808,
      "learning_rate": 1.8072727190888102e-05,
      "loss": 0.0574,
      "step": 5800
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.6415408253669739,
      "learning_rate": 1.8069513094181925e-05,
      "loss": 0.0632,
      "step": 5805
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.572545051574707,
      "learning_rate": 1.8066296605950232e-05,
      "loss": 0.0705,
      "step": 5810
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.7826864123344421,
      "learning_rate": 1.806307772714628e-05,
      "loss": 0.0686,
      "step": 5815
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.5731449723243713,
      "learning_rate": 1.8059856458724028e-05,
      "loss": 0.0607,
      "step": 5820
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.8855318427085876,
      "learning_rate": 1.8056632801638155e-05,
      "loss": 0.0749,
      "step": 5825
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.1506929397583008,
      "learning_rate": 1.805340675684404e-05,
      "loss": 0.0887,
      "step": 5830
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.8381348252296448,
      "learning_rate": 1.8050178325297767e-05,
      "loss": 0.0708,
      "step": 5835
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.8033125996589661,
      "learning_rate": 1.8046947507956138e-05,
      "loss": 0.0649,
      "step": 5840
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.6807813048362732,
      "learning_rate": 1.8043714305776657e-05,
      "loss": 0.0773,
      "step": 5845
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.0260368585586548,
      "learning_rate": 1.804047871971753e-05,
      "loss": 0.0921,
      "step": 5850
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.8171194195747375,
      "learning_rate": 1.803724075073767e-05,
      "loss": 0.0797,
      "step": 5855
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.7568709254264832,
      "learning_rate": 1.8034000399796706e-05,
      "loss": 0.0578,
      "step": 5860
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.6419884562492371,
      "learning_rate": 1.803075766785496e-05,
      "loss": 0.0658,
      "step": 5865
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.9156310558319092,
      "learning_rate": 1.8027512555873472e-05,
      "loss": 0.0793,
      "step": 5870
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.7987212538719177,
      "learning_rate": 1.802426506481398e-05,
      "loss": 0.0695,
      "step": 5875
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.49292904138565063,
      "learning_rate": 1.8021015195638926e-05,
      "loss": 0.0472,
      "step": 5880
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.8221922516822815,
      "learning_rate": 1.8017762949311456e-05,
      "loss": 0.0679,
      "step": 5885
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.5154405832290649,
      "learning_rate": 1.8014508326795435e-05,
      "loss": 0.0732,
      "step": 5890
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.601098895072937,
      "learning_rate": 1.801125132905541e-05,
      "loss": 0.0595,
      "step": 5895
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.9453170895576477,
      "learning_rate": 1.800799195705665e-05,
      "loss": 0.0777,
      "step": 5900
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.7187973260879517,
      "learning_rate": 1.800473021176511e-05,
      "loss": 0.0679,
      "step": 5905
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.686946451663971,
      "learning_rate": 1.8001466094147472e-05,
      "loss": 0.058,
      "step": 5910
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.6171877384185791,
      "learning_rate": 1.7998199605171102e-05,
      "loss": 0.0728,
      "step": 5915
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.7061054110527039,
      "learning_rate": 1.7994930745804072e-05,
      "loss": 0.0567,
      "step": 5920
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.7402878999710083,
      "learning_rate": 1.7991659517015166e-05,
      "loss": 0.0578,
      "step": 5925
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.6647749543190002,
      "learning_rate": 1.7988385919773858e-05,
      "loss": 0.077,
      "step": 5930
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.6972184181213379,
      "learning_rate": 1.7985109955050332e-05,
      "loss": 0.0601,
      "step": 5935
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.0716334581375122,
      "learning_rate": 1.7981831623815465e-05,
      "loss": 0.0675,
      "step": 5940
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.6735240817070007,
      "learning_rate": 1.7978550927040848e-05,
      "loss": 0.071,
      "step": 5945
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.908909022808075,
      "learning_rate": 1.7975267865698765e-05,
      "loss": 0.073,
      "step": 5950
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.6877546906471252,
      "learning_rate": 1.7971982440762204e-05,
      "loss": 0.0844,
      "step": 5955
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.1860584020614624,
      "learning_rate": 1.796869465320485e-05,
      "loss": 0.0761,
      "step": 5960
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.8039729595184326,
      "learning_rate": 1.7965404504001087e-05,
      "loss": 0.0551,
      "step": 5965
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.7674239873886108,
      "learning_rate": 1.7962111994126004e-05,
      "loss": 0.0795,
      "step": 5970
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.7315002679824829,
      "learning_rate": 1.795881712455539e-05,
      "loss": 0.0534,
      "step": 5975
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.46990859508514404,
      "learning_rate": 1.7955519896265727e-05,
      "loss": 0.0599,
      "step": 5980
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.890020489692688,
      "learning_rate": 1.7952220310234205e-05,
      "loss": 0.0724,
      "step": 5985
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.7626070976257324,
      "learning_rate": 1.7948918367438697e-05,
      "loss": 0.0459,
      "step": 5990
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.978101909160614,
      "learning_rate": 1.7945614068857797e-05,
      "loss": 0.0572,
      "step": 5995
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.6801090836524963,
      "learning_rate": 1.7942307415470777e-05,
      "loss": 0.076,
      "step": 6000
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.9605746269226074,
      "learning_rate": 1.7938998408257622e-05,
      "loss": 0.053,
      "step": 6005
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.6000919938087463,
      "learning_rate": 1.7935687048199e-05,
      "loss": 0.0595,
      "step": 6010
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.5320618152618408,
      "learning_rate": 1.7932373336276292e-05,
      "loss": 0.058,
      "step": 6015
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.695774495601654,
      "learning_rate": 1.792905727347156e-05,
      "loss": 0.0906,
      "step": 6020
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.6470620632171631,
      "learning_rate": 1.7925738860767574e-05,
      "loss": 0.0557,
      "step": 6025
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.1034142971038818,
      "learning_rate": 1.79224180991478e-05,
      "loss": 0.0903,
      "step": 6030
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.8370234966278076,
      "learning_rate": 1.7919094989596395e-05,
      "loss": 0.0655,
      "step": 6035
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.8852375149726868,
      "learning_rate": 1.7915769533098214e-05,
      "loss": 0.0971,
      "step": 6040
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.7123667597770691,
      "learning_rate": 1.7912441730638806e-05,
      "loss": 0.0835,
      "step": 6045
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.6630333065986633,
      "learning_rate": 1.790911158320442e-05,
      "loss": 0.0695,
      "step": 6050
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.7879079580307007,
      "learning_rate": 1.7905779091781996e-05,
      "loss": 0.0747,
      "step": 6055
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.8282826542854309,
      "learning_rate": 1.7902444257359167e-05,
      "loss": 0.044,
      "step": 6060
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.0165438652038574,
      "learning_rate": 1.789910708092427e-05,
      "loss": 0.0762,
      "step": 6065
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.9525620937347412,
      "learning_rate": 1.789576756346632e-05,
      "loss": 0.0752,
      "step": 6070
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.36143264174461365,
      "learning_rate": 1.789242570597504e-05,
      "loss": 0.051,
      "step": 6075
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.7918498516082764,
      "learning_rate": 1.7889081509440843e-05,
      "loss": 0.0705,
      "step": 6080
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.9866046905517578,
      "learning_rate": 1.7885734974854824e-05,
      "loss": 0.0664,
      "step": 6085
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.7423363327980042,
      "learning_rate": 1.7882386103208795e-05,
      "loss": 0.0749,
      "step": 6090
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.7900855541229248,
      "learning_rate": 1.787903489549523e-05,
      "loss": 0.1023,
      "step": 6095
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.7296101450920105,
      "learning_rate": 1.787568135270733e-05,
      "loss": 0.0583,
      "step": 6100
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.6891100406646729,
      "learning_rate": 1.7872325475838956e-05,
      "loss": 0.068,
      "step": 6105
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.1211246252059937,
      "learning_rate": 1.7868967265884673e-05,
      "loss": 0.1065,
      "step": 6110
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.8416553735733032,
      "learning_rate": 1.786560672383975e-05,
      "loss": 0.0785,
      "step": 6115
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.5475817918777466,
      "learning_rate": 1.7862243850700126e-05,
      "loss": 0.0632,
      "step": 6120
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.7562156319618225,
      "learning_rate": 1.7858878647462443e-05,
      "loss": 0.0705,
      "step": 6125
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.5753487348556519,
      "learning_rate": 1.7855511115124035e-05,
      "loss": 0.0759,
      "step": 6130
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.6446362137794495,
      "learning_rate": 1.785214125468292e-05,
      "loss": 0.0651,
      "step": 6135
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.6154198050498962,
      "learning_rate": 1.7848769067137803e-05,
      "loss": 0.0901,
      "step": 6140
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.6295109987258911,
      "learning_rate": 1.7845394553488093e-05,
      "loss": 0.0578,
      "step": 6145
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.8939473032951355,
      "learning_rate": 1.7842017714733875e-05,
      "loss": 0.0557,
      "step": 6150
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.9612230658531189,
      "learning_rate": 1.7838638551875927e-05,
      "loss": 0.071,
      "step": 6155
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.8438981771469116,
      "learning_rate": 1.7835257065915722e-05,
      "loss": 0.0795,
      "step": 6160
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.8041678071022034,
      "learning_rate": 1.7831873257855407e-05,
      "loss": 0.0652,
      "step": 6165
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.49747854471206665,
      "learning_rate": 1.7828487128697832e-05,
      "loss": 0.0744,
      "step": 6170
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.5716861486434937,
      "learning_rate": 1.7825098679446528e-05,
      "loss": 0.0697,
      "step": 6175
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.8185078501701355,
      "learning_rate": 1.7821707911105716e-05,
      "loss": 0.0753,
      "step": 6180
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.1662544012069702,
      "learning_rate": 1.78183148246803e-05,
      "loss": 0.0953,
      "step": 6185
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.6486769914627075,
      "learning_rate": 1.7814919421175877e-05,
      "loss": 0.0829,
      "step": 6190
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.0896178483963013,
      "learning_rate": 1.7811521701598723e-05,
      "loss": 0.0706,
      "step": 6195
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.5178818106651306,
      "learning_rate": 1.7808121666955807e-05,
      "loss": 0.0658,
      "step": 6200
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.69623202085495,
      "learning_rate": 1.7804719318254782e-05,
      "loss": 0.0651,
      "step": 6205
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.9311226010322571,
      "learning_rate": 1.780131465650399e-05,
      "loss": 0.0608,
      "step": 6210
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.5501450896263123,
      "learning_rate": 1.779790768271245e-05,
      "loss": 0.0662,
      "step": 6215
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.9106300473213196,
      "learning_rate": 1.7794498397889873e-05,
      "loss": 0.0617,
      "step": 6220
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.7548519372940063,
      "learning_rate": 1.779108680304666e-05,
      "loss": 0.0817,
      "step": 6225
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.5478365421295166,
      "learning_rate": 1.7787672899193882e-05,
      "loss": 0.0485,
      "step": 6230
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.9088108539581299,
      "learning_rate": 1.7784256687343308e-05,
      "loss": 0.0629,
      "step": 6235
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.8210974931716919,
      "learning_rate": 1.778083816850738e-05,
      "loss": 0.0484,
      "step": 6240
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.8573375940322876,
      "learning_rate": 1.777741734369923e-05,
      "loss": 0.0619,
      "step": 6245
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.6718164682388306,
      "learning_rate": 1.777399421393268e-05,
      "loss": 0.0698,
      "step": 6250
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.9652662873268127,
      "learning_rate": 1.777056878022222e-05,
      "loss": 0.0901,
      "step": 6255
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.97198885679245,
      "learning_rate": 1.776714104358303e-05,
      "loss": 0.0844,
      "step": 6260
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.513791561126709,
      "learning_rate": 1.7763711005030977e-05,
      "loss": 0.0638,
      "step": 6265
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.7584071755409241,
      "learning_rate": 1.7760278665582606e-05,
      "loss": 0.0612,
      "step": 6270
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.8927552103996277,
      "learning_rate": 1.7756844026255144e-05,
      "loss": 0.0713,
      "step": 6275
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.7739115953445435,
      "learning_rate": 1.7753407088066494e-05,
      "loss": 0.072,
      "step": 6280
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.0160481929779053,
      "learning_rate": 1.7749967852035257e-05,
      "loss": 0.1012,
      "step": 6285
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.8235144019126892,
      "learning_rate": 1.774652631918069e-05,
      "loss": 0.0784,
      "step": 6290
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.7568656802177429,
      "learning_rate": 1.774308249052276e-05,
      "loss": 0.0891,
      "step": 6295
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.6839597821235657,
      "learning_rate": 1.7739636367082087e-05,
      "loss": 0.0881,
      "step": 6300
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.680732786655426,
      "learning_rate": 1.7736187949879992e-05,
      "loss": 0.0871,
      "step": 6305
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.8405312895774841,
      "learning_rate": 1.7732737239938455e-05,
      "loss": 0.0988,
      "step": 6310
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.7258217334747314,
      "learning_rate": 1.7729284238280163e-05,
      "loss": 0.0622,
      "step": 6315
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.7236860394477844,
      "learning_rate": 1.7725828945928457e-05,
      "loss": 0.0849,
      "step": 6320
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.48804813623428345,
      "learning_rate": 1.7722371363907366e-05,
      "loss": 0.0567,
      "step": 6325
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.7919111847877502,
      "learning_rate": 1.77189114932416e-05,
      "loss": 0.044,
      "step": 6330
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.82615727186203,
      "learning_rate": 1.7715449334956547e-05,
      "loss": 0.0798,
      "step": 6335
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.9560028314590454,
      "learning_rate": 1.771198489007827e-05,
      "loss": 0.0804,
      "step": 6340
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.0600428581237793,
      "learning_rate": 1.770851815963351e-05,
      "loss": 0.0926,
      "step": 6345
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.0330913066864014,
      "learning_rate": 1.770504914464969e-05,
      "loss": 0.0753,
      "step": 6350
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.5192542672157288,
      "learning_rate": 1.7701577846154903e-05,
      "loss": 0.0773,
      "step": 6355
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.9962630867958069,
      "learning_rate": 1.7698104265177925e-05,
      "loss": 0.1456,
      "step": 6360
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.7099242210388184,
      "learning_rate": 1.7694628402748203e-05,
      "loss": 0.0738,
      "step": 6365
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.23063325881958,
      "learning_rate": 1.7691150259895865e-05,
      "loss": 0.0753,
      "step": 6370
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.5459660887718201,
      "learning_rate": 1.7687669837651707e-05,
      "loss": 0.0606,
      "step": 6375
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.8204454183578491,
      "learning_rate": 1.7684187137047214e-05,
      "loss": 0.0663,
      "step": 6380
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.3700021505355835,
      "learning_rate": 1.7680702159114537e-05,
      "loss": 0.0893,
      "step": 6385
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.9551741480827332,
      "learning_rate": 1.7677214904886496e-05,
      "loss": 0.068,
      "step": 6390
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.195141077041626,
      "learning_rate": 1.76737253753966e-05,
      "loss": 0.0781,
      "step": 6395
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2067859172821045,
      "learning_rate": 1.7670233571679025e-05,
      "loss": 0.097,
      "step": 6400
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8323922753334045,
      "learning_rate": 1.7666739494768622e-05,
      "loss": 0.0658,
      "step": 6405
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7578905820846558,
      "learning_rate": 1.766324314570091e-05,
      "loss": 0.0901,
      "step": 6410
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9802036285400391,
      "learning_rate": 1.765974452551209e-05,
      "loss": 0.0394,
      "step": 6415
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6872645616531372,
      "learning_rate": 1.765624363523903e-05,
      "loss": 0.0387,
      "step": 6420
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9163408875465393,
      "learning_rate": 1.7652740475919275e-05,
      "loss": 0.0289,
      "step": 6425
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.6171188950538635,
      "learning_rate": 1.7649235048591044e-05,
      "loss": 0.0283,
      "step": 6430
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.5377746820449829,
      "learning_rate": 1.7645727354293218e-05,
      "loss": 0.0304,
      "step": 6435
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.9587904214859009,
      "learning_rate": 1.764221739406536e-05,
      "loss": 0.0331,
      "step": 6440
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.0330588817596436,
      "learning_rate": 1.7638705168947702e-05,
      "loss": 0.0486,
      "step": 6445
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.749311089515686,
      "learning_rate": 1.7635190679981146e-05,
      "loss": 0.035,
      "step": 6450
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.9288537502288818,
      "learning_rate": 1.763167392820726e-05,
      "loss": 0.0366,
      "step": 6455
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.7338812351226807,
      "learning_rate": 1.7628154914668297e-05,
      "loss": 0.0421,
      "step": 6460
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.4167785048484802,
      "learning_rate": 1.762463364040717e-05,
      "loss": 0.026,
      "step": 6465
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.8623115420341492,
      "learning_rate": 1.7621110106467458e-05,
      "loss": 0.039,
      "step": 6470
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.7953075766563416,
      "learning_rate": 1.7617584313893417e-05,
      "loss": 0.0418,
      "step": 6475
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.6112560629844666,
      "learning_rate": 1.761405626372997e-05,
      "loss": 0.0305,
      "step": 6480
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.6794976592063904,
      "learning_rate": 1.7610525957022712e-05,
      "loss": 0.0351,
      "step": 6485
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.6737608313560486,
      "learning_rate": 1.7606993394817903e-05,
      "loss": 0.0337,
      "step": 6490
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.1872364282608032,
      "learning_rate": 1.760345857816247e-05,
      "loss": 0.0461,
      "step": 6495
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.8426690697669983,
      "learning_rate": 1.7599921508104018e-05,
      "loss": 0.0306,
      "step": 6500
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.825894832611084,
      "learning_rate": 1.7596382185690802e-05,
      "loss": 0.0419,
      "step": 6505
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.8509790897369385,
      "learning_rate": 1.7592840611971765e-05,
      "loss": 0.0366,
      "step": 6510
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.7118868231773376,
      "learning_rate": 1.7589296787996503e-05,
      "loss": 0.0491,
      "step": 6515
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.9368664026260376,
      "learning_rate": 1.758575071481528e-05,
      "loss": 0.0305,
      "step": 6520
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.8669412732124329,
      "learning_rate": 1.7582202393479037e-05,
      "loss": 0.0427,
      "step": 6525
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.7687955498695374,
      "learning_rate": 1.757865182503937e-05,
      "loss": 0.0378,
      "step": 6530
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.9307061433792114,
      "learning_rate": 1.7575099010548543e-05,
      "loss": 0.0396,
      "step": 6535
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.670947790145874,
      "learning_rate": 1.7571543951059493e-05,
      "loss": 0.0451,
      "step": 6540
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.6400846838951111,
      "learning_rate": 1.7567986647625813e-05,
      "loss": 0.0339,
      "step": 6545
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.0342981815338135,
      "learning_rate": 1.7564427101301767e-05,
      "loss": 0.0347,
      "step": 6550
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.1251486539840698,
      "learning_rate": 1.756086531314228e-05,
      "loss": 0.028,
      "step": 6555
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.1043672561645508,
      "learning_rate": 1.7557301284202947e-05,
      "loss": 0.0357,
      "step": 6560
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.6798744797706604,
      "learning_rate": 1.7553735015540022e-05,
      "loss": 0.0299,
      "step": 6565
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.8796838521957397,
      "learning_rate": 1.755016650821042e-05,
      "loss": 0.0404,
      "step": 6570
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.696263313293457,
      "learning_rate": 1.754659576327173e-05,
      "loss": 0.0434,
      "step": 6575
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.7451015114784241,
      "learning_rate": 1.754302278178219e-05,
      "loss": 0.0362,
      "step": 6580
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.7645043730735779,
      "learning_rate": 1.7539447564800714e-05,
      "loss": 0.037,
      "step": 6585
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.9263461828231812,
      "learning_rate": 1.7535870113386873e-05,
      "loss": 0.028,
      "step": 6590
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.7381824254989624,
      "learning_rate": 1.75322904286009e-05,
      "loss": 0.0425,
      "step": 6595
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.8345758318901062,
      "learning_rate": 1.752870851150369e-05,
      "loss": 0.0413,
      "step": 6600
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.4364377558231354,
      "learning_rate": 1.7525124363156796e-05,
      "loss": 0.0384,
      "step": 6605
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.7056708931922913,
      "learning_rate": 1.7521537984622444e-05,
      "loss": 0.0344,
      "step": 6610
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.7152959704399109,
      "learning_rate": 1.7517949376963503e-05,
      "loss": 0.0348,
      "step": 6615
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.824553370475769,
      "learning_rate": 1.7514358541243518e-05,
      "loss": 0.0351,
      "step": 6620
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.7878523468971252,
      "learning_rate": 1.751076547852669e-05,
      "loss": 0.0337,
      "step": 6625
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.6791169047355652,
      "learning_rate": 1.750717018987788e-05,
      "loss": 0.03,
      "step": 6630
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.6202698945999146,
      "learning_rate": 1.75035726763626e-05,
      "loss": 0.0367,
      "step": 6635
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.7299143075942993,
      "learning_rate": 1.749997293904704e-05,
      "loss": 0.0423,
      "step": 6640
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.9602674841880798,
      "learning_rate": 1.749637097899803e-05,
      "loss": 0.0367,
      "step": 6645
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.9754753112792969,
      "learning_rate": 1.749276679728307e-05,
      "loss": 0.0367,
      "step": 6650
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.645286500453949,
      "learning_rate": 1.7489160394970314e-05,
      "loss": 0.0391,
      "step": 6655
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.5990299582481384,
      "learning_rate": 1.7485551773128574e-05,
      "loss": 0.0327,
      "step": 6660
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.062504529953003,
      "learning_rate": 1.7481940932827324e-05,
      "loss": 0.0433,
      "step": 6665
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.6965630650520325,
      "learning_rate": 1.747832787513669e-05,
      "loss": 0.0411,
      "step": 6670
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.509995698928833,
      "learning_rate": 1.747471260112746e-05,
      "loss": 0.0338,
      "step": 6675
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.990881621837616,
      "learning_rate": 1.7471095111871076e-05,
      "loss": 0.0375,
      "step": 6680
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.8508270382881165,
      "learning_rate": 1.7467475408439636e-05,
      "loss": 0.0384,
      "step": 6685
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.7476538419723511,
      "learning_rate": 1.74638534919059e-05,
      "loss": 0.0408,
      "step": 6690
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.6179473400115967,
      "learning_rate": 1.746022936334327e-05,
      "loss": 0.0395,
      "step": 6695
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.7255807518959045,
      "learning_rate": 1.745660302382582e-05,
      "loss": 0.033,
      "step": 6700
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.7685034871101379,
      "learning_rate": 1.745297447442827e-05,
      "loss": 0.0269,
      "step": 6705
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.8111138343811035,
      "learning_rate": 1.7449343716225998e-05,
      "loss": 0.0358,
      "step": 6710
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.9519994854927063,
      "learning_rate": 1.7445710750295034e-05,
      "loss": 0.0482,
      "step": 6715
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.8502340316772461,
      "learning_rate": 1.7442075577712064e-05,
      "loss": 0.0441,
      "step": 6720
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.5589675903320312,
      "learning_rate": 1.743843819955443e-05,
      "loss": 0.0267,
      "step": 6725
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.6813309788703918,
      "learning_rate": 1.7434798616900122e-05,
      "loss": 0.0298,
      "step": 6730
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.6360641717910767,
      "learning_rate": 1.743115683082779e-05,
      "loss": 0.0391,
      "step": 6735
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.5595855712890625,
      "learning_rate": 1.742751284241673e-05,
      "loss": 0.0285,
      "step": 6740
    },
    {
      "epoch": 2.1,
      "grad_norm": Infinity,
      "learning_rate": 1.7424596066729708e-05,
      "loss": 0.1078,
      "step": 6745
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.6349712014198303,
      "learning_rate": 1.7420948116830864e-05,
      "loss": 0.0482,
      "step": 6750
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.8908722400665283,
      "learning_rate": 1.7417297967618806e-05,
      "loss": 0.0535,
      "step": 6755
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.2324929237365723,
      "learning_rate": 1.7413645620175313e-05,
      "loss": 0.0334,
      "step": 6760
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.7902947664260864,
      "learning_rate": 1.740999107558281e-05,
      "loss": 0.0339,
      "step": 6765
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.8780156970024109,
      "learning_rate": 1.7406334334924386e-05,
      "loss": 0.0568,
      "step": 6770
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.7588701248168945,
      "learning_rate": 1.7402675399283762e-05,
      "loss": 0.0384,
      "step": 6775
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.9184271693229675,
      "learning_rate": 1.7399014269745332e-05,
      "loss": 0.0348,
      "step": 6780
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.895255982875824,
      "learning_rate": 1.7395350947394127e-05,
      "loss": 0.0226,
      "step": 6785
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.495691180229187,
      "learning_rate": 1.739168543331582e-05,
      "loss": 0.0316,
      "step": 6790
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.8194505572319031,
      "learning_rate": 1.7388017728596752e-05,
      "loss": 0.0396,
      "step": 6795
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.6622453927993774,
      "learning_rate": 1.73843478343239e-05,
      "loss": 0.0344,
      "step": 6800
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.5545173287391663,
      "learning_rate": 1.73806757515849e-05,
      "loss": 0.0454,
      "step": 6805
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.9523460268974304,
      "learning_rate": 1.7377001481468017e-05,
      "loss": 0.0372,
      "step": 6810
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.9859544634819031,
      "learning_rate": 1.7373325025062197e-05,
      "loss": 0.0368,
      "step": 6815
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.64875727891922,
      "learning_rate": 1.7369646383457e-05,
      "loss": 0.0375,
      "step": 6820
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.5710052847862244,
      "learning_rate": 1.7365965557742658e-05,
      "loss": 0.0306,
      "step": 6825
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.690978467464447,
      "learning_rate": 1.736228254901003e-05,
      "loss": 0.0387,
      "step": 6830
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.9192633628845215,
      "learning_rate": 1.7358597358350646e-05,
      "loss": 0.035,
      "step": 6835
    },
    {
      "epoch": 2.13,
      "grad_norm": 15.089350700378418,
      "learning_rate": 1.7354909986856664e-05,
      "loss": 0.059,
      "step": 6840
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.888706088066101,
      "learning_rate": 1.735122043562089e-05,
      "loss": 0.0421,
      "step": 6845
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.481156587600708,
      "learning_rate": 1.734752870573678e-05,
      "loss": 0.0639,
      "step": 6850
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.3059687614440918,
      "learning_rate": 1.734383479829844e-05,
      "loss": 0.0459,
      "step": 6855
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.741889238357544,
      "learning_rate": 1.7340138714400616e-05,
      "loss": 0.0677,
      "step": 6860
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.8907132744789124,
      "learning_rate": 1.7336440455138695e-05,
      "loss": 0.0521,
      "step": 6865
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.2927290201187134,
      "learning_rate": 1.7332740021608722e-05,
      "loss": 0.0434,
      "step": 6870
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.7092378735542297,
      "learning_rate": 1.7329037414907374e-05,
      "loss": 0.0414,
      "step": 6875
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.9778251647949219,
      "learning_rate": 1.732533263613197e-05,
      "loss": 0.0402,
      "step": 6880
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.6234639883041382,
      "learning_rate": 1.732162568638049e-05,
      "loss": 0.0401,
      "step": 6885
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.4143224954605103,
      "learning_rate": 1.7317916566751537e-05,
      "loss": 0.0458,
      "step": 6890
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.670375645160675,
      "learning_rate": 1.7314205278344372e-05,
      "loss": 0.0365,
      "step": 6895
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.0456308126449585,
      "learning_rate": 1.731049182225889e-05,
      "loss": 0.0475,
      "step": 6900
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.8300198316574097,
      "learning_rate": 1.730677619959563e-05,
      "loss": 0.0385,
      "step": 6905
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.7571542263031006,
      "learning_rate": 1.7303058411455774e-05,
      "loss": 0.0291,
      "step": 6910
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.3860149383544922,
      "learning_rate": 1.729933845894115e-05,
      "loss": 0.0425,
      "step": 6915
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4162564277648926,
      "learning_rate": 1.7295616343154224e-05,
      "loss": 0.0488,
      "step": 6920
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.6601019501686096,
      "learning_rate": 1.72918920651981e-05,
      "loss": 0.0461,
      "step": 6925
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.9042936563491821,
      "learning_rate": 1.728816562617653e-05,
      "loss": 0.0331,
      "step": 6930
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.8178083896636963,
      "learning_rate": 1.728443702719389e-05,
      "loss": 0.0459,
      "step": 6935
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.077895164489746,
      "learning_rate": 1.7280706269355222e-05,
      "loss": 0.0413,
      "step": 6940
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.7887989282608032,
      "learning_rate": 1.7276973353766187e-05,
      "loss": 0.0478,
      "step": 6945
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.1501586437225342,
      "learning_rate": 1.7273238281533097e-05,
      "loss": 0.0378,
      "step": 6950
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.1501015424728394,
      "learning_rate": 1.7269501053762896e-05,
      "loss": 0.0338,
      "step": 6955
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.111640453338623,
      "learning_rate": 1.7265761671563167e-05,
      "loss": 0.039,
      "step": 6960
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.6609378457069397,
      "learning_rate": 1.7262020136042138e-05,
      "loss": 0.0403,
      "step": 6965
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.8092169165611267,
      "learning_rate": 1.7258276448308672e-05,
      "loss": 0.0411,
      "step": 6970
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.6540818810462952,
      "learning_rate": 1.725453060947226e-05,
      "loss": 0.0359,
      "step": 6975
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.087636947631836,
      "learning_rate": 1.7250782620643055e-05,
      "loss": 0.0426,
      "step": 6980
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.8799102902412415,
      "learning_rate": 1.7247032482931817e-05,
      "loss": 0.0377,
      "step": 6985
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.72962486743927,
      "learning_rate": 1.7243280197449963e-05,
      "loss": 0.041,
      "step": 6990
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.7479212284088135,
      "learning_rate": 1.7239525765309543e-05,
      "loss": 0.0276,
      "step": 6995
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.8470936417579651,
      "learning_rate": 1.723576918762324e-05,
      "loss": 0.0428,
      "step": 7000
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.8011202216148376,
      "learning_rate": 1.7232010465504376e-05,
      "loss": 0.0392,
      "step": 7005
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.085710048675537,
      "learning_rate": 1.72282496000669e-05,
      "loss": 0.0585,
      "step": 7010
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.6330513954162598,
      "learning_rate": 1.7224486592425407e-05,
      "loss": 0.033,
      "step": 7015
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.7976099252700806,
      "learning_rate": 1.7220721443695126e-05,
      "loss": 0.0331,
      "step": 7020
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.0006788969039917,
      "learning_rate": 1.7216954154991917e-05,
      "loss": 0.0399,
      "step": 7025
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.693641185760498,
      "learning_rate": 1.7213184727432267e-05,
      "loss": 0.0336,
      "step": 7030
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.958171010017395,
      "learning_rate": 1.720941316213331e-05,
      "loss": 0.0476,
      "step": 7035
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.7659122347831726,
      "learning_rate": 1.7205639460212807e-05,
      "loss": 0.05,
      "step": 7040
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.8464353680610657,
      "learning_rate": 1.7201863622789158e-05,
      "loss": 0.047,
      "step": 7045
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.8198861479759216,
      "learning_rate": 1.719808565098138e-05,
      "loss": 0.041,
      "step": 7050
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.4542665183544159,
      "learning_rate": 1.7194305545909146e-05,
      "loss": 0.0388,
      "step": 7055
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.7713673710823059,
      "learning_rate": 1.719052330869274e-05,
      "loss": 0.0338,
      "step": 7060
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.8043854236602783,
      "learning_rate": 1.7186738940453092e-05,
      "loss": 0.026,
      "step": 7065
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.8994637131690979,
      "learning_rate": 1.7182952442311752e-05,
      "loss": 0.0435,
      "step": 7070
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.6519581079483032,
      "learning_rate": 1.7179163815390912e-05,
      "loss": 0.031,
      "step": 7075
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.0966442823410034,
      "learning_rate": 1.7175373060813396e-05,
      "loss": 0.0342,
      "step": 7080
    },
    {
      "epoch": 2.21,
      "grad_norm": 6.207887172698975,
      "learning_rate": 1.717158017970264e-05,
      "loss": 0.0642,
      "step": 7085
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.5999218821525574,
      "learning_rate": 1.7167785173182734e-05,
      "loss": 0.0343,
      "step": 7090
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.7367837429046631,
      "learning_rate": 1.716398804237838e-05,
      "loss": 0.0468,
      "step": 7095
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.121308445930481,
      "learning_rate": 1.7160188788414923e-05,
      "loss": 0.0411,
      "step": 7100
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.8983088135719299,
      "learning_rate": 1.7156387412418328e-05,
      "loss": 0.0345,
      "step": 7105
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.9664926528930664,
      "learning_rate": 1.7152583915515188e-05,
      "loss": 0.0378,
      "step": 7110
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.8459317088127136,
      "learning_rate": 1.7148778298832737e-05,
      "loss": 0.0403,
      "step": 7115
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.0068672895431519,
      "learning_rate": 1.714497056349882e-05,
      "loss": 0.0497,
      "step": 7120
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.114620327949524,
      "learning_rate": 1.714116071064192e-05,
      "loss": 0.042,
      "step": 7125
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.9400085806846619,
      "learning_rate": 1.7137348741391155e-05,
      "loss": 0.0432,
      "step": 7130
    },
    {
      "epoch": 2.23,
      "grad_norm": 7.140718936920166,
      "learning_rate": 1.713353465687625e-05,
      "loss": 0.0664,
      "step": 7135
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.289583206176758,
      "learning_rate": 1.7129718458227574e-05,
      "loss": 0.0388,
      "step": 7140
    },
    {
      "epoch": 2.23,
      "grad_norm": 25.73358917236328,
      "learning_rate": 1.7125900146576112e-05,
      "loss": 0.0408,
      "step": 7145
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.9193993210792542,
      "learning_rate": 1.7122079723053486e-05,
      "loss": 0.0376,
      "step": 7150
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.834202289581299,
      "learning_rate": 1.711825718879194e-05,
      "loss": 0.0444,
      "step": 7155
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.4898897409439087,
      "learning_rate": 1.711443254492433e-05,
      "loss": 0.0416,
      "step": 7160
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.7508093118667603,
      "learning_rate": 1.7110605792584157e-05,
      "loss": 0.055,
      "step": 7165
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.8517021536827087,
      "learning_rate": 1.7106776932905535e-05,
      "loss": 0.0457,
      "step": 7170
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.6700179576873779,
      "learning_rate": 1.710294596702321e-05,
      "loss": 0.034,
      "step": 7175
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.8382918238639832,
      "learning_rate": 1.7099112896072546e-05,
      "loss": 0.0483,
      "step": 7180
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.9156063795089722,
      "learning_rate": 1.709527772118953e-05,
      "loss": 0.0423,
      "step": 7185
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.8718563318252563,
      "learning_rate": 1.7091440443510784e-05,
      "loss": 0.0397,
      "step": 7190
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.9186751246452332,
      "learning_rate": 1.7087601064173535e-05,
      "loss": 0.0402,
      "step": 7195
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.8474892377853394,
      "learning_rate": 1.7083759584315648e-05,
      "loss": 0.0404,
      "step": 7200
    },
    {
      "epoch": 2.25,
      "grad_norm": 6.393630027770996,
      "learning_rate": 1.7079916005075604e-05,
      "loss": 0.0432,
      "step": 7205
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.7541725635528564,
      "learning_rate": 1.707607032759251e-05,
      "loss": 0.0417,
      "step": 7210
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.794830858707428,
      "learning_rate": 1.7072222553006086e-05,
      "loss": 0.0446,
      "step": 7215
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.8188607692718506,
      "learning_rate": 1.7068372682456686e-05,
      "loss": 0.0395,
      "step": 7220
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.7906588912010193,
      "learning_rate": 1.7064520717085276e-05,
      "loss": 0.0305,
      "step": 7225
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.7647044062614441,
      "learning_rate": 1.706066665803345e-05,
      "loss": 0.0422,
      "step": 7230
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.5535922050476074,
      "learning_rate": 1.705681050644341e-05,
      "loss": 0.0426,
      "step": 7235
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.8452866077423096,
      "learning_rate": 1.7052952263457993e-05,
      "loss": 0.0404,
      "step": 7240
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.6865875720977783,
      "learning_rate": 1.7049091930220647e-05,
      "loss": 0.0505,
      "step": 7245
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.904468834400177,
      "learning_rate": 1.7045229507875443e-05,
      "loss": 0.0345,
      "step": 7250
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.7027440071105957,
      "learning_rate": 1.7041364997567064e-05,
      "loss": 0.0424,
      "step": 7255
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.8030436038970947,
      "learning_rate": 1.703749840044083e-05,
      "loss": 0.0385,
      "step": 7260
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.8274464011192322,
      "learning_rate": 1.7033629717642656e-05,
      "loss": 0.0431,
      "step": 7265
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.9646808505058289,
      "learning_rate": 1.702975895031909e-05,
      "loss": 0.0376,
      "step": 7270
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.9568483233451843,
      "learning_rate": 1.7025886099617293e-05,
      "loss": 0.0328,
      "step": 7275
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.6486111879348755,
      "learning_rate": 1.7022011166685046e-05,
      "loss": 0.035,
      "step": 7280
    },
    {
      "epoch": 2.27,
      "grad_norm": 5.381010055541992,
      "learning_rate": 1.7018134152670745e-05,
      "loss": 0.0321,
      "step": 7285
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.302991509437561,
      "learning_rate": 1.7014255058723403e-05,
      "loss": 0.0423,
      "step": 7290
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.7025021910667419,
      "learning_rate": 1.701037388599265e-05,
      "loss": 0.0482,
      "step": 7295
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.8713043928146362,
      "learning_rate": 1.7006490635628733e-05,
      "loss": 0.0457,
      "step": 7300
    },
    {
      "epoch": 2.28,
      "grad_norm": 21.033382415771484,
      "learning_rate": 1.700260530878251e-05,
      "loss": 0.0937,
      "step": 7305
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.8523858189582825,
      "learning_rate": 1.699871790660546e-05,
      "loss": 0.0375,
      "step": 7310
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.4806874990463257,
      "learning_rate": 1.6994828430249676e-05,
      "loss": 0.0477,
      "step": 7315
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.6420323252677917,
      "learning_rate": 1.6990936880867863e-05,
      "loss": 0.0303,
      "step": 7320
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.7890008687973022,
      "learning_rate": 1.698704325961334e-05,
      "loss": 0.0437,
      "step": 7325
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.2356423139572144,
      "learning_rate": 1.698314756764005e-05,
      "loss": 0.0398,
      "step": 7330
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.7098721265792847,
      "learning_rate": 1.6979249806102535e-05,
      "loss": 0.0411,
      "step": 7335
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.9438479542732239,
      "learning_rate": 1.697534997615596e-05,
      "loss": 0.0373,
      "step": 7340
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.758943498134613,
      "learning_rate": 1.6971448078956098e-05,
      "loss": 0.0446,
      "step": 7345
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.7948986291885376,
      "learning_rate": 1.6967544115659344e-05,
      "loss": 0.0359,
      "step": 7350
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.1407995223999023,
      "learning_rate": 1.6963638087422686e-05,
      "loss": 0.0414,
      "step": 7355
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.8077024221420288,
      "learning_rate": 1.6959729995403743e-05,
      "loss": 0.0407,
      "step": 7360
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.3187479972839355,
      "learning_rate": 1.6955819840760736e-05,
      "loss": 0.0369,
      "step": 7365
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.6189730763435364,
      "learning_rate": 1.6951907624652506e-05,
      "loss": 0.0364,
      "step": 7370
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.8181635141372681,
      "learning_rate": 1.6947993348238492e-05,
      "loss": 0.038,
      "step": 7375
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.5456126928329468,
      "learning_rate": 1.6944077012678758e-05,
      "loss": 0.0479,
      "step": 7380
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.996551036834717,
      "learning_rate": 1.6940158619133965e-05,
      "loss": 0.0402,
      "step": 7385
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.6797996759414673,
      "learning_rate": 1.693623816876539e-05,
      "loss": 0.0552,
      "step": 7390
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.6198251843452454,
      "learning_rate": 1.693231566273492e-05,
      "loss": 0.0368,
      "step": 7395
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.764711320400238,
      "learning_rate": 1.692839110220505e-05,
      "loss": 0.0427,
      "step": 7400
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.841666042804718,
      "learning_rate": 1.692446448833889e-05,
      "loss": 0.0401,
      "step": 7405
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.7802782654762268,
      "learning_rate": 1.6920535822300145e-05,
      "loss": 0.0464,
      "step": 7410
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.9961202144622803,
      "learning_rate": 1.6916605105253144e-05,
      "loss": 0.0443,
      "step": 7415
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.8209372758865356,
      "learning_rate": 1.691267233836281e-05,
      "loss": 0.0442,
      "step": 7420
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.8145427703857422,
      "learning_rate": 1.6908737522794682e-05,
      "loss": 0.037,
      "step": 7425
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.7817115187644958,
      "learning_rate": 1.6904800659714902e-05,
      "loss": 0.0333,
      "step": 7430
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.8446786403656006,
      "learning_rate": 1.690086175029022e-05,
      "loss": 0.0337,
      "step": 7435
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.0260515213012695,
      "learning_rate": 1.6896920795687998e-05,
      "loss": 0.0423,
      "step": 7440
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.1411818265914917,
      "learning_rate": 1.6892977797076197e-05,
      "loss": 0.0281,
      "step": 7445
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.8356356024742126,
      "learning_rate": 1.6889032755623382e-05,
      "loss": 0.0392,
      "step": 7450
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.90280681848526,
      "learning_rate": 1.6885085672498726e-05,
      "loss": 0.0372,
      "step": 7455
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.8793005347251892,
      "learning_rate": 1.688113654887202e-05,
      "loss": 0.0345,
      "step": 7460
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.5199700593948364,
      "learning_rate": 1.6877185385913636e-05,
      "loss": 0.0291,
      "step": 7465
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.762761652469635,
      "learning_rate": 1.6873232184794566e-05,
      "loss": 0.0432,
      "step": 7470
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.3521060943603516,
      "learning_rate": 1.686927694668641e-05,
      "loss": 0.0395,
      "step": 7475
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.9418859481811523,
      "learning_rate": 1.686531967276135e-05,
      "loss": 0.0475,
      "step": 7480
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.8807092905044556,
      "learning_rate": 1.6861360364192195e-05,
      "loss": 0.0473,
      "step": 7485
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.8254885673522949,
      "learning_rate": 1.685739902215235e-05,
      "loss": 0.0348,
      "step": 7490
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.0653108358383179,
      "learning_rate": 1.6853435647815812e-05,
      "loss": 0.0506,
      "step": 7495
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.9384851455688477,
      "learning_rate": 1.6849470242357197e-05,
      "loss": 0.0391,
      "step": 7500
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.7083196640014648,
      "learning_rate": 1.6845502806951706e-05,
      "loss": 0.0252,
      "step": 7505
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.5872582793235779,
      "learning_rate": 1.6841533342775153e-05,
      "loss": 0.0424,
      "step": 7510
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.6886513829231262,
      "learning_rate": 1.6837561851003956e-05,
      "loss": 0.0551,
      "step": 7515
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.934777557849884,
      "learning_rate": 1.683358833281512e-05,
      "loss": 0.0367,
      "step": 7520
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.0215473175048828,
      "learning_rate": 1.6829612789386265e-05,
      "loss": 0.0424,
      "step": 7525
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.6930414438247681,
      "learning_rate": 1.68256352218956e-05,
      "loss": 0.0299,
      "step": 7530
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.0204532146453857,
      "learning_rate": 1.6821655631521947e-05,
      "loss": 0.0628,
      "step": 7535
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.1268620491027832,
      "learning_rate": 1.681767401944471e-05,
      "loss": 0.0364,
      "step": 7540
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.9842634201049805,
      "learning_rate": 1.6813690386843905e-05,
      "loss": 0.0375,
      "step": 7545
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.8676565885543823,
      "learning_rate": 1.6809704734900148e-05,
      "loss": 0.0338,
      "step": 7550
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.8421193361282349,
      "learning_rate": 1.680571706479464e-05,
      "loss": 0.0419,
      "step": 7555
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.695347547531128,
      "learning_rate": 1.6801727377709195e-05,
      "loss": 0.0341,
      "step": 7560
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.8723491430282593,
      "learning_rate": 1.679773567482622e-05,
      "loss": 0.0373,
      "step": 7565
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.7483975291252136,
      "learning_rate": 1.6793741957328713e-05,
      "loss": 0.0489,
      "step": 7570
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.1286433935165405,
      "learning_rate": 1.6789746226400277e-05,
      "loss": 0.0392,
      "step": 7575
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.7481610774993896,
      "learning_rate": 1.678574848322511e-05,
      "loss": 0.0394,
      "step": 7580
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.645449697971344,
      "learning_rate": 1.6781748728988003e-05,
      "loss": 0.0386,
      "step": 7585
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.9019328951835632,
      "learning_rate": 1.6777746964874346e-05,
      "loss": 0.0327,
      "step": 7590
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.5962875485420227,
      "learning_rate": 1.6773743192070124e-05,
      "loss": 0.0369,
      "step": 7595
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.7119027376174927,
      "learning_rate": 1.6769737411761917e-05,
      "loss": 0.0344,
      "step": 7600
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.7260536551475525,
      "learning_rate": 1.6765729625136906e-05,
      "loss": 0.0356,
      "step": 7605
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.9319799542427063,
      "learning_rate": 1.6761719833382846e-05,
      "loss": 0.0405,
      "step": 7610
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.6937612891197205,
      "learning_rate": 1.6757708037688117e-05,
      "loss": 0.0273,
      "step": 7615
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.7246429324150085,
      "learning_rate": 1.675369423924167e-05,
      "loss": 0.0343,
      "step": 7620
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.7110769748687744,
      "learning_rate": 1.6749678439233058e-05,
      "loss": 0.0354,
      "step": 7625
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.9897186160087585,
      "learning_rate": 1.674566063885242e-05,
      "loss": 0.0541,
      "step": 7630
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.7339702248573303,
      "learning_rate": 1.67416408392905e-05,
      "loss": 0.0364,
      "step": 7635
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.0443443059921265,
      "learning_rate": 1.673761904173863e-05,
      "loss": 0.0384,
      "step": 7640
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.766381561756134,
      "learning_rate": 1.6733595247388723e-05,
      "loss": 0.0261,
      "step": 7645
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.9241385459899902,
      "learning_rate": 1.67295694574333e-05,
      "loss": 0.0387,
      "step": 7650
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.869219958782196,
      "learning_rate": 1.6725541673065465e-05,
      "loss": 0.0335,
      "step": 7655
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.7011229991912842,
      "learning_rate": 1.6721511895478915e-05,
      "loss": 0.0434,
      "step": 7660
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.0889408588409424,
      "learning_rate": 1.6717480125867936e-05,
      "loss": 0.0436,
      "step": 7665
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.7818471789360046,
      "learning_rate": 1.6713446365427402e-05,
      "loss": 0.0361,
      "step": 7670
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.7618464231491089,
      "learning_rate": 1.6709410615352794e-05,
      "loss": 0.0352,
      "step": 7675
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.0346121788024902,
      "learning_rate": 1.670537287684015e-05,
      "loss": 0.0339,
      "step": 7680
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.872925877571106,
      "learning_rate": 1.670133315108613e-05,
      "loss": 0.039,
      "step": 7685
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.7551829814910889,
      "learning_rate": 1.6697291439287965e-05,
      "loss": 0.0359,
      "step": 7690
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.0588226318359375,
      "learning_rate": 1.669324774264348e-05,
      "loss": 0.0284,
      "step": 7695
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.8457269668579102,
      "learning_rate": 1.6689202062351086e-05,
      "loss": 0.0409,
      "step": 7700
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.787465512752533,
      "learning_rate": 1.6685154399609787e-05,
      "loss": 0.0293,
      "step": 7705
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.9415322542190552,
      "learning_rate": 1.6681104755619166e-05,
      "loss": 0.0391,
      "step": 7710
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.9193680286407471,
      "learning_rate": 1.6677053131579397e-05,
      "loss": 0.0399,
      "step": 7715
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.6151878833770752,
      "learning_rate": 1.6672999528691245e-05,
      "loss": 0.0372,
      "step": 7720
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.5558480620384216,
      "learning_rate": 1.666894394815606e-05,
      "loss": 0.0279,
      "step": 7725
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.7361923456192017,
      "learning_rate": 1.6664886391175775e-05,
      "loss": 0.0357,
      "step": 7730
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.6120460033416748,
      "learning_rate": 1.6660826858952902e-05,
      "loss": 0.0471,
      "step": 7735
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.5619087815284729,
      "learning_rate": 1.665676535269056e-05,
      "loss": 0.0265,
      "step": 7740
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.0988489389419556,
      "learning_rate": 1.665270187359243e-05,
      "loss": 0.0401,
      "step": 7745
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.712315559387207,
      "learning_rate": 1.664863642286279e-05,
      "loss": 0.0414,
      "step": 7750
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.0724822282791138,
      "learning_rate": 1.6644569001706502e-05,
      "loss": 0.0332,
      "step": 7755
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.9803611040115356,
      "learning_rate": 1.6640499611329003e-05,
      "loss": 0.0334,
      "step": 7760
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.6442325115203857,
      "learning_rate": 1.6636428252936326e-05,
      "loss": 0.0294,
      "step": 7765
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.1340879201889038,
      "learning_rate": 1.6632354927735076e-05,
      "loss": 0.0272,
      "step": 7770
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.017544150352478,
      "learning_rate": 1.6628279636932453e-05,
      "loss": 0.0426,
      "step": 7775
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.0248128175735474,
      "learning_rate": 1.6624202381736226e-05,
      "loss": 0.0414,
      "step": 7780
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.7086641788482666,
      "learning_rate": 1.662012316335476e-05,
      "loss": 0.0396,
      "step": 7785
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.7434113621711731,
      "learning_rate": 1.6616041982996986e-05,
      "loss": 0.0437,
      "step": 7790
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.7166509628295898,
      "learning_rate": 1.6611958841872427e-05,
      "loss": 0.0305,
      "step": 7795
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.7706720232963562,
      "learning_rate": 1.6607873741191197e-05,
      "loss": 0.0342,
      "step": 7800
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.6073039174079895,
      "learning_rate": 1.6603786682163964e-05,
      "loss": 0.0418,
      "step": 7805
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.849389910697937,
      "learning_rate": 1.6599697666001998e-05,
      "loss": 0.0389,
      "step": 7810
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.6125224232673645,
      "learning_rate": 1.659560669391714e-05,
      "loss": 0.0318,
      "step": 7815
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.7484531402587891,
      "learning_rate": 1.659151376712182e-05,
      "loss": 0.0497,
      "step": 7820
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.7708289623260498,
      "learning_rate": 1.6587418886829034e-05,
      "loss": 0.0401,
      "step": 7825
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.958131730556488,
      "learning_rate": 1.6583322054252363e-05,
      "loss": 0.0336,
      "step": 7830
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.7831741571426392,
      "learning_rate": 1.6579223270605968e-05,
      "loss": 0.0298,
      "step": 7835
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.0524176359176636,
      "learning_rate": 1.657512253710459e-05,
      "loss": 0.0572,
      "step": 7840
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.0298731327056885,
      "learning_rate": 1.657101985496354e-05,
      "loss": 0.0452,
      "step": 7845
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.9892203211784363,
      "learning_rate": 1.656691522539872e-05,
      "loss": 0.0332,
      "step": 7850
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.8172314763069153,
      "learning_rate": 1.656280864962659e-05,
      "loss": 0.0485,
      "step": 7855
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.6813853979110718,
      "learning_rate": 1.6558700128864206e-05,
      "loss": 0.0351,
      "step": 7860
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.6233561635017395,
      "learning_rate": 1.6554589664329186e-05,
      "loss": 0.0336,
      "step": 7865
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.1553077697753906,
      "learning_rate": 1.6550477257239733e-05,
      "loss": 0.034,
      "step": 7870
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.5354089140892029,
      "learning_rate": 1.6546362908814623e-05,
      "loss": 0.0292,
      "step": 7875
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.7513230443000793,
      "learning_rate": 1.654224662027321e-05,
      "loss": 0.0286,
      "step": 7880
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.7038884162902832,
      "learning_rate": 1.653812839283541e-05,
      "loss": 0.0404,
      "step": 7885
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.9296578764915466,
      "learning_rate": 1.653400822772173e-05,
      "loss": 0.0377,
      "step": 7890
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.7138383388519287,
      "learning_rate": 1.652988612615325e-05,
      "loss": 0.0371,
      "step": 7895
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.8308192491531372,
      "learning_rate": 1.652576208935161e-05,
      "loss": 0.0422,
      "step": 7900
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.6145361661911011,
      "learning_rate": 1.652163611853904e-05,
      "loss": 0.0247,
      "step": 7905
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.931335985660553,
      "learning_rate": 1.6517508214938328e-05,
      "loss": 0.0348,
      "step": 7910
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.8879193067550659,
      "learning_rate": 1.651337837977285e-05,
      "loss": 0.0389,
      "step": 7915
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.9264506101608276,
      "learning_rate": 1.6509246614266535e-05,
      "loss": 0.0344,
      "step": 7920
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.5814990997314453,
      "learning_rate": 1.650511291964391e-05,
      "loss": 0.0334,
      "step": 7925
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.6779055595397949,
      "learning_rate": 1.6500977297130054e-05,
      "loss": 0.0335,
      "step": 7930
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.857691764831543,
      "learning_rate": 1.649683974795062e-05,
      "loss": 0.0377,
      "step": 7935
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.7544268369674683,
      "learning_rate": 1.6492700273331835e-05,
      "loss": 0.036,
      "step": 7940
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.754554808139801,
      "learning_rate": 1.6488558874500503e-05,
      "loss": 0.0423,
      "step": 7945
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.6650204658508301,
      "learning_rate": 1.6484415552683986e-05,
      "loss": 0.0466,
      "step": 7950
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.9802659153938293,
      "learning_rate": 1.6480270309110224e-05,
      "loss": 0.0414,
      "step": 7955
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.8128763437271118,
      "learning_rate": 1.647612314500772e-05,
      "loss": 0.0415,
      "step": 7960
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.7706061005592346,
      "learning_rate": 1.6471974061605556e-05,
      "loss": 0.0298,
      "step": 7965
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.544342577457428,
      "learning_rate": 1.6467823060133378e-05,
      "loss": 0.0357,
      "step": 7970
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.7460711598396301,
      "learning_rate": 1.6463670141821394e-05,
      "loss": 0.0255,
      "step": 7975
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.8606274724006653,
      "learning_rate": 1.645951530790039e-05,
      "loss": 0.0526,
      "step": 7980
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.019179105758667,
      "learning_rate": 1.6455358559601715e-05,
      "loss": 0.0355,
      "step": 7985
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.9450494050979614,
      "learning_rate": 1.6451199898157288e-05,
      "loss": 0.0398,
      "step": 7990
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.6935722231864929,
      "learning_rate": 1.6447039324799585e-05,
      "loss": 0.0446,
      "step": 7995
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.8222050666809082,
      "learning_rate": 1.6442876840761667e-05,
      "loss": 0.0281,
      "step": 8000
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.6318336129188538,
      "learning_rate": 1.6438712447277142e-05,
      "loss": 0.0219,
      "step": 8005
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.866169810295105,
      "learning_rate": 1.64345461455802e-05,
      "loss": 0.0468,
      "step": 8010
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.8264434337615967,
      "learning_rate": 1.643037793690558e-05,
      "loss": 0.0434,
      "step": 8015
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.024613857269287,
      "learning_rate": 1.6426207822488603e-05,
      "loss": 0.0386,
      "step": 8020
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.7298322319984436,
      "learning_rate": 1.642203580356515e-05,
      "loss": 0.0321,
      "step": 8025
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.083125352859497,
      "learning_rate": 1.6417861881371654e-05,
      "loss": 0.0382,
      "step": 8030
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.6637572050094604,
      "learning_rate": 1.641368605714513e-05,
      "loss": 0.0504,
      "step": 8035
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.6951263546943665,
      "learning_rate": 1.6409508332123138e-05,
      "loss": 0.0429,
      "step": 8040
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.8588437438011169,
      "learning_rate": 1.6405328707543823e-05,
      "loss": 0.0212,
      "step": 8045
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.7385882139205933,
      "learning_rate": 1.6401147184645877e-05,
      "loss": 0.0315,
      "step": 8050
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.523358941078186,
      "learning_rate": 1.639696376466856e-05,
      "loss": 0.0245,
      "step": 8055
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.707894504070282,
      "learning_rate": 1.639277844885169e-05,
      "loss": 0.0351,
      "step": 8060
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.8531205654144287,
      "learning_rate": 1.6388591238435654e-05,
      "loss": 0.0412,
      "step": 8065
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.8357705473899841,
      "learning_rate": 1.6384402134661397e-05,
      "loss": 0.0429,
      "step": 8070
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.7158061861991882,
      "learning_rate": 1.6380211138770424e-05,
      "loss": 0.0326,
      "step": 8075
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.0381453037261963,
      "learning_rate": 1.63760182520048e-05,
      "loss": 0.031,
      "step": 8080
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.9476194381713867,
      "learning_rate": 1.6371823475607154e-05,
      "loss": 0.0345,
      "step": 8085
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.6989636421203613,
      "learning_rate": 1.6367626810820676e-05,
      "loss": 0.0266,
      "step": 8090
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.7869669795036316,
      "learning_rate": 1.6363428258889108e-05,
      "loss": 0.04,
      "step": 8095
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.0312131643295288,
      "learning_rate": 1.6359227821056752e-05,
      "loss": 0.0448,
      "step": 8100
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.9612884521484375,
      "learning_rate": 1.6355025498568485e-05,
      "loss": 0.0416,
      "step": 8105
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.8276956677436829,
      "learning_rate": 1.6350821292669724e-05,
      "loss": 0.0361,
      "step": 8110
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.6667450666427612,
      "learning_rate": 1.6346615204606448e-05,
      "loss": 0.0316,
      "step": 8115
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.8874323964118958,
      "learning_rate": 1.6342407235625203e-05,
      "loss": 0.0581,
      "step": 8120
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.7738590240478516,
      "learning_rate": 1.633819738697308e-05,
      "loss": 0.0431,
      "step": 8125
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.710171639919281,
      "learning_rate": 1.6333985659897737e-05,
      "loss": 0.0315,
      "step": 8130
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.6763156056404114,
      "learning_rate": 1.632977205564738e-05,
      "loss": 0.0395,
      "step": 8135
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.6096729636192322,
      "learning_rate": 1.632555657547078e-05,
      "loss": 0.0346,
      "step": 8140
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.097630500793457,
      "learning_rate": 1.6321339220617262e-05,
      "loss": 0.0467,
      "step": 8145
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.7507611513137817,
      "learning_rate": 1.6317119992336698e-05,
      "loss": 0.031,
      "step": 8150
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.8424447178840637,
      "learning_rate": 1.631289889187953e-05,
      "loss": 0.0444,
      "step": 8155
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.0239499807357788,
      "learning_rate": 1.630867592049674e-05,
      "loss": 0.031,
      "step": 8160
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.6097200512886047,
      "learning_rate": 1.630445107943987e-05,
      "loss": 0.0436,
      "step": 8165
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.662420928478241,
      "learning_rate": 1.6300224369961022e-05,
      "loss": 0.0239,
      "step": 8170
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.47109031677246094,
      "learning_rate": 1.629599579331285e-05,
      "loss": 0.0251,
      "step": 8175
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.7203395366668701,
      "learning_rate": 1.6291765350748554e-05,
      "loss": 0.0403,
      "step": 8180
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.7731189727783203,
      "learning_rate": 1.6287533043521888e-05,
      "loss": 0.046,
      "step": 8185
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.7585968971252441,
      "learning_rate": 1.6283298872887166e-05,
      "loss": 0.0396,
      "step": 8190
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.827131450176239,
      "learning_rate": 1.6279062840099248e-05,
      "loss": 0.0289,
      "step": 8195
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.6474844813346863,
      "learning_rate": 1.6274824946413552e-05,
      "loss": 0.0239,
      "step": 8200
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.7048923969268799,
      "learning_rate": 1.627058519308604e-05,
      "loss": 0.0396,
      "step": 8205
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.568031370639801,
      "learning_rate": 1.626634358137323e-05,
      "loss": 0.0452,
      "step": 8210
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.8100217580795288,
      "learning_rate": 1.6262100112532194e-05,
      "loss": 0.039,
      "step": 8215
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.9056006669998169,
      "learning_rate": 1.625785478782054e-05,
      "loss": 0.038,
      "step": 8220
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.7772448658943176,
      "learning_rate": 1.6253607608496444e-05,
      "loss": 0.0369,
      "step": 8225
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.9542365670204163,
      "learning_rate": 1.6249358575818618e-05,
      "loss": 0.0297,
      "step": 8230
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.48814138770103455,
      "learning_rate": 1.6245107691046335e-05,
      "loss": 0.0372,
      "step": 8235
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.1438723802566528,
      "learning_rate": 1.6240854955439402e-05,
      "loss": 0.0459,
      "step": 8240
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.5761513113975525,
      "learning_rate": 1.6236600370258193e-05,
      "loss": 0.0291,
      "step": 8245
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.7332009077072144,
      "learning_rate": 1.6232343936763614e-05,
      "loss": 0.0346,
      "step": 8250
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.6568591594696045,
      "learning_rate": 1.6228085656217133e-05,
      "loss": 0.036,
      "step": 8255
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.6697298884391785,
      "learning_rate": 1.6223825529880742e-05,
      "loss": 0.0268,
      "step": 8260
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.789516031742096,
      "learning_rate": 1.621956355901701e-05,
      "loss": 0.0442,
      "step": 8265
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.6526065468788147,
      "learning_rate": 1.621529974488904e-05,
      "loss": 0.0284,
      "step": 8270
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.229318857192993,
      "learning_rate": 1.621103408876046e-05,
      "loss": 0.0346,
      "step": 8275
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.6409262418746948,
      "learning_rate": 1.6206766591895486e-05,
      "loss": 0.0274,
      "step": 8280
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.7699871063232422,
      "learning_rate": 1.6202497255558844e-05,
      "loss": 0.023,
      "step": 8285
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.9219619631767273,
      "learning_rate": 1.6198226081015823e-05,
      "loss": 0.0511,
      "step": 8290
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.7852801084518433,
      "learning_rate": 1.619395306953225e-05,
      "loss": 0.0361,
      "step": 8295
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.8048036098480225,
      "learning_rate": 1.61896782223745e-05,
      "loss": 0.0308,
      "step": 8300
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.6462746262550354,
      "learning_rate": 1.618540154080949e-05,
      "loss": 0.0314,
      "step": 8305
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.5217623114585876,
      "learning_rate": 1.6181123026104674e-05,
      "loss": 0.0378,
      "step": 8310
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.0315415859222412,
      "learning_rate": 1.6176842679528068e-05,
      "loss": 0.0479,
      "step": 8315
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.7805516719818115,
      "learning_rate": 1.6172560502348213e-05,
      "loss": 0.0318,
      "step": 8320
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.886910617351532,
      "learning_rate": 1.61682764958342e-05,
      "loss": 0.0359,
      "step": 8325
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.6613612771034241,
      "learning_rate": 1.616399066125566e-05,
      "loss": 0.0381,
      "step": 8330
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.9460324645042419,
      "learning_rate": 1.6159702999882763e-05,
      "loss": 0.036,
      "step": 8335
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.9295498132705688,
      "learning_rate": 1.615541351298623e-05,
      "loss": 0.0335,
      "step": 8340
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.7159280776977539,
      "learning_rate": 1.6151122201837318e-05,
      "loss": 0.0375,
      "step": 8345
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.9947106242179871,
      "learning_rate": 1.6146829067707818e-05,
      "loss": 0.0509,
      "step": 8350
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.8763145208358765,
      "learning_rate": 1.6142534111870067e-05,
      "loss": 0.0387,
      "step": 8355
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.8634383678436279,
      "learning_rate": 1.613823733559695e-05,
      "loss": 0.0398,
      "step": 8360
    },
    {
      "epoch": 2.61,
      "grad_norm": 11.380885124206543,
      "learning_rate": 1.613393874016187e-05,
      "loss": 0.035,
      "step": 8365
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.8703393936157227,
      "learning_rate": 1.6129638326838796e-05,
      "loss": 0.038,
      "step": 8370
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.8195749521255493,
      "learning_rate": 1.6125336096902214e-05,
      "loss": 0.0459,
      "step": 8375
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.143809199333191,
      "learning_rate": 1.6121032051627164e-05,
      "loss": 0.0419,
      "step": 8380
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.9704199433326721,
      "learning_rate": 1.6116726192289207e-05,
      "loss": 0.0438,
      "step": 8385
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.7254340648651123,
      "learning_rate": 1.611241852016446e-05,
      "loss": 0.0351,
      "step": 8390
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.8501814603805542,
      "learning_rate": 1.6108109036529563e-05,
      "loss": 0.0456,
      "step": 8395
    },
    {
      "epoch": 2.62,
      "grad_norm": 3.679652452468872,
      "learning_rate": 1.6103797742661698e-05,
      "loss": 0.0321,
      "step": 8400
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.1169507503509521,
      "learning_rate": 1.609948463983859e-05,
      "loss": 0.041,
      "step": 8405
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.7015743851661682,
      "learning_rate": 1.6095169729338485e-05,
      "loss": 0.0355,
      "step": 8410
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.7521771788597107,
      "learning_rate": 1.6090853012440183e-05,
      "loss": 0.032,
      "step": 8415
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.8876381516456604,
      "learning_rate": 1.6086534490423002e-05,
      "loss": 0.0406,
      "step": 8420
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.3464659452438354,
      "learning_rate": 1.6082214164566807e-05,
      "loss": 0.0404,
      "step": 8425
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.1670219898223877,
      "learning_rate": 1.6077892036151995e-05,
      "loss": 0.039,
      "step": 8430
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.5168001651763916,
      "learning_rate": 1.60735681064595e-05,
      "loss": 0.0525,
      "step": 8435
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.0316020250320435,
      "learning_rate": 1.6069242376770772e-05,
      "loss": 0.0376,
      "step": 8440
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.7442283034324646,
      "learning_rate": 1.6064914848367818e-05,
      "loss": 0.0313,
      "step": 8445
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.9239617586135864,
      "learning_rate": 1.606058552253317e-05,
      "loss": 0.0366,
      "step": 8450
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.3136714994907379,
      "learning_rate": 1.6056254400549885e-05,
      "loss": 0.0343,
      "step": 8455
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.7645418047904968,
      "learning_rate": 1.605192148370156e-05,
      "loss": 0.0302,
      "step": 8460
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.49683353304862976,
      "learning_rate": 1.6047586773272322e-05,
      "loss": 0.0404,
      "step": 8465
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.589810311794281,
      "learning_rate": 1.6043250270546832e-05,
      "loss": 0.0346,
      "step": 8470
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.5914163589477539,
      "learning_rate": 1.603891197681028e-05,
      "loss": 0.023,
      "step": 8475
    },
    {
      "epoch": 2.64,
      "grad_norm": 16.47276496887207,
      "learning_rate": 1.603457189334838e-05,
      "loss": 0.0321,
      "step": 8480
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.7461623549461365,
      "learning_rate": 1.6030230021447396e-05,
      "loss": 0.0394,
      "step": 8485
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.5548524856567383,
      "learning_rate": 1.60258863623941e-05,
      "loss": 0.0351,
      "step": 8490
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.5185383558273315,
      "learning_rate": 1.60215409174758e-05,
      "loss": 0.0349,
      "step": 8495
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.8972347378730774,
      "learning_rate": 1.601719368798034e-05,
      "loss": 0.0356,
      "step": 8500
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.979608952999115,
      "learning_rate": 1.60128446751961e-05,
      "loss": 0.0493,
      "step": 8505
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.9665581583976746,
      "learning_rate": 1.6008493880411956e-05,
      "loss": 0.0342,
      "step": 8510
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.772299587726593,
      "learning_rate": 1.6004141304917347e-05,
      "loss": 0.041,
      "step": 8515
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.6392544507980347,
      "learning_rate": 1.5999786950002227e-05,
      "loss": 0.039,
      "step": 8520
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.9064784049987793,
      "learning_rate": 1.599543081695707e-05,
      "loss": 0.0574,
      "step": 8525
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.8364819288253784,
      "learning_rate": 1.5991072907072886e-05,
      "loss": 0.0358,
      "step": 8530
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.5332923531532288,
      "learning_rate": 1.5986713221641213e-05,
      "loss": 0.0312,
      "step": 8535
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.8230621218681335,
      "learning_rate": 1.5982351761954113e-05,
      "loss": 0.0318,
      "step": 8540
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.9251091480255127,
      "learning_rate": 1.5977988529304162e-05,
      "loss": 0.0429,
      "step": 8545
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.81324702501297,
      "learning_rate": 1.5973623524984477e-05,
      "loss": 0.0442,
      "step": 8550
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5638898611068726,
      "learning_rate": 1.5969256750288702e-05,
      "loss": 0.0434,
      "step": 8555
    },
    {
      "epoch": 2.67,
      "grad_norm": 2.012404441833496,
      "learning_rate": 1.596488820651099e-05,
      "loss": 0.0361,
      "step": 8560
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.6309984922409058,
      "learning_rate": 1.5960517894946027e-05,
      "loss": 0.0313,
      "step": 8565
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.140384554862976,
      "learning_rate": 1.5956145816889028e-05,
      "loss": 0.0411,
      "step": 8570
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.0145649909973145,
      "learning_rate": 1.5951771973635722e-05,
      "loss": 0.0314,
      "step": 8575
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.9595271944999695,
      "learning_rate": 1.5947396366482365e-05,
      "loss": 0.0427,
      "step": 8580
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.4232027530670166,
      "learning_rate": 1.5943018996725738e-05,
      "loss": 0.0325,
      "step": 8585
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.8087257146835327,
      "learning_rate": 1.5938639865663142e-05,
      "loss": 0.0344,
      "step": 8590
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.615680456161499,
      "learning_rate": 1.5934258974592402e-05,
      "loss": 0.0327,
      "step": 8595
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.7677449584007263,
      "learning_rate": 1.5929876324811862e-05,
      "loss": 0.0409,
      "step": 8600
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.9178354144096375,
      "learning_rate": 1.5925491917620384e-05,
      "loss": 0.0303,
      "step": 8605
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.8850105404853821,
      "learning_rate": 1.5921105754317364e-05,
      "loss": 0.0299,
      "step": 8610
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.832343101501465,
      "learning_rate": 1.5916717836202705e-05,
      "loss": 0.053,
      "step": 8615
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.7948322892189026,
      "learning_rate": 1.591232816457683e-05,
      "loss": 0.0353,
      "step": 8620
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.8507640957832336,
      "learning_rate": 1.5907936740740693e-05,
      "loss": 0.041,
      "step": 8625
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.45911988615989685,
      "learning_rate": 1.590354356599576e-05,
      "loss": 0.043,
      "step": 8630
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.5260618925094604,
      "learning_rate": 1.5899148641644015e-05,
      "loss": 0.0317,
      "step": 8635
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.6386368274688721,
      "learning_rate": 1.5894751968987956e-05,
      "loss": 0.0435,
      "step": 8640
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.847206711769104,
      "learning_rate": 1.589035354933062e-05,
      "loss": 0.0453,
      "step": 8645
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.6548765897750854,
      "learning_rate": 1.588595338397553e-05,
      "loss": 0.0366,
      "step": 8650
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.4999925196170807,
      "learning_rate": 1.5881551474226754e-05,
      "loss": 0.0373,
      "step": 8655
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.6970962882041931,
      "learning_rate": 1.587714782138887e-05,
      "loss": 0.0367,
      "step": 8660
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.4393469095230103,
      "learning_rate": 1.5872742426766955e-05,
      "loss": 0.0408,
      "step": 8665
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.5475620627403259,
      "learning_rate": 1.5868335291666628e-05,
      "loss": 0.0215,
      "step": 8670
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.145965814590454,
      "learning_rate": 1.5863926417394004e-05,
      "loss": 0.0453,
      "step": 8675
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.7778739929199219,
      "learning_rate": 1.5859515805255728e-05,
      "loss": 0.0309,
      "step": 8680
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.993541955947876,
      "learning_rate": 1.5855103456558953e-05,
      "loss": 0.0418,
      "step": 8685
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.7978174686431885,
      "learning_rate": 1.5850689372611342e-05,
      "loss": 0.046,
      "step": 8690
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.6382090449333191,
      "learning_rate": 1.5846273554721078e-05,
      "loss": 0.037,
      "step": 8695
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.6067830324172974,
      "learning_rate": 1.5841856004196863e-05,
      "loss": 0.0505,
      "step": 8700
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.8111315965652466,
      "learning_rate": 1.5837436722347902e-05,
      "loss": 0.0422,
      "step": 8705
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.6436445116996765,
      "learning_rate": 1.5833015710483918e-05,
      "loss": 0.0327,
      "step": 8710
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.6014158725738525,
      "learning_rate": 1.5828592969915145e-05,
      "loss": 0.0358,
      "step": 8715
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.9966750741004944,
      "learning_rate": 1.5824168501952337e-05,
      "loss": 0.0407,
      "step": 8720
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.0554747581481934,
      "learning_rate": 1.5819742307906744e-05,
      "loss": 0.0501,
      "step": 8725
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.0621410608291626,
      "learning_rate": 1.5815314389090147e-05,
      "loss": 0.0267,
      "step": 8730
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.9338129758834839,
      "learning_rate": 1.5810884746814825e-05,
      "loss": 0.0459,
      "step": 8735
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.8834397196769714,
      "learning_rate": 1.5806453382393573e-05,
      "loss": 0.0505,
      "step": 8740
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.958518922328949,
      "learning_rate": 1.580202029713969e-05,
      "loss": 0.0535,
      "step": 8745
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.6857681274414062,
      "learning_rate": 1.5797585492366993e-05,
      "loss": 0.0501,
      "step": 8750
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.6894921660423279,
      "learning_rate": 1.5793148969389802e-05,
      "loss": 0.0288,
      "step": 8755
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.6848697662353516,
      "learning_rate": 1.5788710729522953e-05,
      "loss": 0.0331,
      "step": 8760
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.7238781452178955,
      "learning_rate": 1.578427077408179e-05,
      "loss": 0.0358,
      "step": 8765
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.8410157561302185,
      "learning_rate": 1.5779829104382154e-05,
      "loss": 0.0368,
      "step": 8770
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.7383372187614441,
      "learning_rate": 1.5775385721740413e-05,
      "loss": 0.0316,
      "step": 8775
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.5548520088195801,
      "learning_rate": 1.5770940627473423e-05,
      "loss": 0.0395,
      "step": 8780
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.9778690934181213,
      "learning_rate": 1.576649382289856e-05,
      "loss": 0.0483,
      "step": 8785
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.892069399356842,
      "learning_rate": 1.5762045309333707e-05,
      "loss": 0.0348,
      "step": 8790
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.5767723321914673,
      "learning_rate": 1.575759508809725e-05,
      "loss": 0.0346,
      "step": 8795
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.6027498245239258,
      "learning_rate": 1.5753143160508076e-05,
      "loss": 0.0408,
      "step": 8800
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.7998393177986145,
      "learning_rate": 1.5748689527885587e-05,
      "loss": 0.0283,
      "step": 8805
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.7058614492416382,
      "learning_rate": 1.5744234191549685e-05,
      "loss": 0.0349,
      "step": 8810
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.7166592478752136,
      "learning_rate": 1.5739777152820775e-05,
      "loss": 0.0539,
      "step": 8815
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.6599798202514648,
      "learning_rate": 1.5735318413019775e-05,
      "loss": 0.0337,
      "step": 8820
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5327255129814148,
      "learning_rate": 1.57308579734681e-05,
      "loss": 0.03,
      "step": 8825
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.7015475630760193,
      "learning_rate": 1.572639583548767e-05,
      "loss": 0.0438,
      "step": 8830
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.1382325887680054,
      "learning_rate": 1.5721932000400908e-05,
      "loss": 0.0299,
      "step": 8835
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.5569730401039124,
      "learning_rate": 1.571746646953074e-05,
      "loss": 0.0309,
      "step": 8840
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.7418985962867737,
      "learning_rate": 1.57129992442006e-05,
      "loss": 0.0415,
      "step": 8845
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.6939743161201477,
      "learning_rate": 1.5708530325734413e-05,
      "loss": 0.0315,
      "step": 8850
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.8243491053581238,
      "learning_rate": 1.570405971545662e-05,
      "loss": 0.0397,
      "step": 8855
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.8243434429168701,
      "learning_rate": 1.569958741469215e-05,
      "loss": 0.0512,
      "step": 8860
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.8044533729553223,
      "learning_rate": 1.569511342476644e-05,
      "loss": 0.0292,
      "step": 8865
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.7083048224449158,
      "learning_rate": 1.5690637747005427e-05,
      "loss": 0.0424,
      "step": 8870
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.6713657379150391,
      "learning_rate": 1.568616038273555e-05,
      "loss": 0.0304,
      "step": 8875
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.6863850355148315,
      "learning_rate": 1.5681681333283742e-05,
      "loss": 0.0329,
      "step": 8880
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.7204381227493286,
      "learning_rate": 1.567720059997744e-05,
      "loss": 0.0344,
      "step": 8885
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.0264719724655151,
      "learning_rate": 1.567271818414458e-05,
      "loss": 0.045,
      "step": 8890
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.983029305934906,
      "learning_rate": 1.566823408711359e-05,
      "loss": 0.0392,
      "step": 8895
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.7892402410507202,
      "learning_rate": 1.566374831021341e-05,
      "loss": 0.0343,
      "step": 8900
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.6004326343536377,
      "learning_rate": 1.5659260854773465e-05,
      "loss": 0.0382,
      "step": 8905
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.7688624262809753,
      "learning_rate": 1.5654771722123684e-05,
      "loss": 0.0387,
      "step": 8910
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.6235817074775696,
      "learning_rate": 1.565028091359449e-05,
      "loss": 0.0385,
      "step": 8915
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.8388769030570984,
      "learning_rate": 1.56457884305168e-05,
      "loss": 0.0333,
      "step": 8920
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.8253457546234131,
      "learning_rate": 1.564129427422204e-05,
      "loss": 0.031,
      "step": 8925
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.765497088432312,
      "learning_rate": 1.5636798446042116e-05,
      "loss": 0.0305,
      "step": 8930
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.0711227655410767,
      "learning_rate": 1.5632300947309437e-05,
      "loss": 0.0349,
      "step": 8935
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.7266090512275696,
      "learning_rate": 1.562780177935691e-05,
      "loss": 0.033,
      "step": 8940
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.9013699293136597,
      "learning_rate": 1.562330094351793e-05,
      "loss": 0.0403,
      "step": 8945
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.77614825963974,
      "learning_rate": 1.5618798441126392e-05,
      "loss": 0.031,
      "step": 8950
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.7244963049888611,
      "learning_rate": 1.5614294273516684e-05,
      "loss": 0.0459,
      "step": 8955
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.6892554759979248,
      "learning_rate": 1.560978844202368e-05,
      "loss": 0.0402,
      "step": 8960
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.7315613627433777,
      "learning_rate": 1.560528094798276e-05,
      "loss": 0.0422,
      "step": 8965
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.277847170829773,
      "learning_rate": 1.5600771792729782e-05,
      "loss": 0.045,
      "step": 8970
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.834357500076294,
      "learning_rate": 1.5596260977601113e-05,
      "loss": 0.0397,
      "step": 8975
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.7074677348136902,
      "learning_rate": 1.55917485039336e-05,
      "loss": 0.0314,
      "step": 8980
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.9965816736221313,
      "learning_rate": 1.5587234373064577e-05,
      "loss": 0.0384,
      "step": 8985
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.8888028264045715,
      "learning_rate": 1.5582718586331888e-05,
      "loss": 0.0366,
      "step": 8990
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.8134787082672119,
      "learning_rate": 1.557820114507385e-05,
      "loss": 0.0442,
      "step": 8995
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.8666115403175354,
      "learning_rate": 1.557368205062928e-05,
      "loss": 0.0277,
      "step": 9000
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.7883780598640442,
      "learning_rate": 1.556916130433748e-05,
      "loss": 0.0452,
      "step": 9005
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.1791362762451172,
      "learning_rate": 1.5564638907538246e-05,
      "loss": 0.0391,
      "step": 9010
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.7088319063186646,
      "learning_rate": 1.5560114861571853e-05,
      "loss": 0.0289,
      "step": 9015
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.5044339895248413,
      "learning_rate": 1.555558916777908e-05,
      "loss": 0.0339,
      "step": 9020
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.5306674838066101,
      "learning_rate": 1.5551061827501185e-05,
      "loss": 0.0318,
      "step": 9025
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.4773883819580078,
      "learning_rate": 1.5546532842079917e-05,
      "loss": 0.0394,
      "step": 9030
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.0266894102096558,
      "learning_rate": 1.5542002212857506e-05,
      "loss": 0.0397,
      "step": 9035
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.6027147173881531,
      "learning_rate": 1.553746994117668e-05,
      "loss": 0.0353,
      "step": 9040
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.8161146640777588,
      "learning_rate": 1.5532936028380642e-05,
      "loss": 0.0288,
      "step": 9045
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.6301555633544922,
      "learning_rate": 1.5528400475813093e-05,
      "loss": 0.039,
      "step": 9050
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.7754611968994141,
      "learning_rate": 1.552386328481821e-05,
      "loss": 0.0396,
      "step": 9055
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.8263349533081055,
      "learning_rate": 1.5519324456740665e-05,
      "loss": 0.036,
      "step": 9060
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.5884903073310852,
      "learning_rate": 1.5514783992925607e-05,
      "loss": 0.0289,
      "step": 9065
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.5677090883255005,
      "learning_rate": 1.551024189471867e-05,
      "loss": 0.0342,
      "step": 9070
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.6854045391082764,
      "learning_rate": 1.5505698163465986e-05,
      "loss": 0.0365,
      "step": 9075
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.9792048931121826,
      "learning_rate": 1.5501152800514147e-05,
      "loss": 0.033,
      "step": 9080
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.9109053015708923,
      "learning_rate": 1.5496605807210253e-05,
      "loss": 0.0429,
      "step": 9085
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.7893013954162598,
      "learning_rate": 1.5492057184901867e-05,
      "loss": 0.0392,
      "step": 9090
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.5471526384353638,
      "learning_rate": 1.5487506934937047e-05,
      "loss": 0.0398,
      "step": 9095
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.0387884378433228,
      "learning_rate": 1.5482955058664337e-05,
      "loss": 0.0432,
      "step": 9100
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.7424880266189575,
      "learning_rate": 1.547840155743275e-05,
      "loss": 0.023,
      "step": 9105
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.5438539981842041,
      "learning_rate": 1.547384643259178e-05,
      "loss": 0.0294,
      "step": 9110
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.5898480415344238,
      "learning_rate": 1.5469289685491422e-05,
      "loss": 0.0261,
      "step": 9115
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.6443936228752136,
      "learning_rate": 1.546473131748213e-05,
      "loss": 0.0413,
      "step": 9120
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.6669382452964783,
      "learning_rate": 1.5460171329914856e-05,
      "loss": 0.049,
      "step": 9125
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5739983320236206,
      "learning_rate": 1.545560972414101e-05,
      "loss": 0.0321,
      "step": 9130
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.9604806303977966,
      "learning_rate": 1.5451046501512505e-05,
      "loss": 0.0473,
      "step": 9135
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5344032049179077,
      "learning_rate": 1.544648166338172e-05,
      "loss": 0.0317,
      "step": 9140
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.9302431344985962,
      "learning_rate": 1.5441915211101518e-05,
      "loss": 0.039,
      "step": 9145
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.7158989906311035,
      "learning_rate": 1.543734714602523e-05,
      "loss": 0.031,
      "step": 9150
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.6064074635505676,
      "learning_rate": 1.5432777469506683e-05,
      "loss": 0.0263,
      "step": 9155
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.8086766004562378,
      "learning_rate": 1.5428206182900165e-05,
      "loss": 0.0398,
      "step": 9160
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.8926895260810852,
      "learning_rate": 1.542363328756045e-05,
      "loss": 0.0323,
      "step": 9165
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.9440407752990723,
      "learning_rate": 1.541905878484279e-05,
      "loss": 0.0348,
      "step": 9170
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.7965365648269653,
      "learning_rate": 1.5414482676102905e-05,
      "loss": 0.0382,
      "step": 9175
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.6618263125419617,
      "learning_rate": 1.5409904962696994e-05,
      "loss": 0.0309,
      "step": 9180
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.8952043652534485,
      "learning_rate": 1.540532564598174e-05,
      "loss": 0.0383,
      "step": 9185
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.7503764629364014,
      "learning_rate": 1.540074472731429e-05,
      "loss": 0.0354,
      "step": 9190
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.796486496925354,
      "learning_rate": 1.539616220805227e-05,
      "loss": 0.0404,
      "step": 9195
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.6163582801818848,
      "learning_rate": 1.5391578089553785e-05,
      "loss": 0.0379,
      "step": 9200
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.635602593421936,
      "learning_rate": 1.5386992373177406e-05,
      "loss": 0.0462,
      "step": 9205
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.812752366065979,
      "learning_rate": 1.5382405060282187e-05,
      "loss": 0.0385,
      "step": 9210
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.7380738854408264,
      "learning_rate": 1.537781615222764e-05,
      "loss": 0.0322,
      "step": 9215
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.7037160992622375,
      "learning_rate": 1.5373225650373763e-05,
      "loss": 0.0338,
      "step": 9220
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.8600029349327087,
      "learning_rate": 1.536863355608103e-05,
      "loss": 0.0361,
      "step": 9225
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.7166641354560852,
      "learning_rate": 1.5364039870710368e-05,
      "loss": 0.0361,
      "step": 9230
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.3040114641189575,
      "learning_rate": 1.5359444595623192e-05,
      "loss": 0.0556,
      "step": 9235
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.7275735139846802,
      "learning_rate": 1.5354847732181384e-05,
      "loss": 0.0467,
      "step": 9240
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.890828549861908,
      "learning_rate": 1.5350249281747297e-05,
      "loss": 0.05,
      "step": 9245
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.6557976007461548,
      "learning_rate": 1.5345649245683748e-05,
      "loss": 0.0439,
      "step": 9250
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.8745369911193848,
      "learning_rate": 1.5341047625354033e-05,
      "loss": 0.0373,
      "step": 9255
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.7743590474128723,
      "learning_rate": 1.5336444422121913e-05,
      "loss": 0.0293,
      "step": 9260
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.1002082824707031,
      "learning_rate": 1.5331839637351623e-05,
      "loss": 0.0362,
      "step": 9265
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.8608964681625366,
      "learning_rate": 1.532723327240785e-05,
      "loss": 0.0396,
      "step": 9270
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.35328683257102966,
      "learning_rate": 1.5322625328655776e-05,
      "loss": 0.0317,
      "step": 9275
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.7710239887237549,
      "learning_rate": 1.531801580746103e-05,
      "loss": 0.0373,
      "step": 9280
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.8235945105552673,
      "learning_rate": 1.5313404710189715e-05,
      "loss": 0.0372,
      "step": 9285
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.777231752872467,
      "learning_rate": 1.5308792038208406e-05,
      "loss": 0.0353,
      "step": 9290
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.1153934001922607,
      "learning_rate": 1.530417779288413e-05,
      "loss": 0.0429,
      "step": 9295
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.0209358930587769,
      "learning_rate": 1.52995619755844e-05,
      "loss": 0.0405,
      "step": 9300
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.0184146165847778,
      "learning_rate": 1.529494458767718e-05,
      "loss": 0.0435,
      "step": 9305
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.7970442771911621,
      "learning_rate": 1.5290325630530906e-05,
      "loss": 0.0537,
      "step": 9310
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.7759125828742981,
      "learning_rate": 1.528570510551448e-05,
      "loss": 0.0372,
      "step": 9315
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.9187195301055908,
      "learning_rate": 1.5281083013997262e-05,
      "loss": 0.0328,
      "step": 9320
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.1653680801391602,
      "learning_rate": 1.5276459357349082e-05,
      "loss": 0.0424,
      "step": 9325
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.6130503416061401,
      "learning_rate": 1.5271834136940235e-05,
      "loss": 0.036,
      "step": 9330
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.6654280424118042,
      "learning_rate": 1.5267207354141474e-05,
      "loss": 0.0251,
      "step": 9335
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.9446947574615479,
      "learning_rate": 1.5262579010324025e-05,
      "loss": 0.0378,
      "step": 9340
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.7537127733230591,
      "learning_rate": 1.5257949106859555e-05,
      "loss": 0.0372,
      "step": 9345
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.6928290128707886,
      "learning_rate": 1.5253317645120218e-05,
      "loss": 0.0382,
      "step": 9350
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.0862854719161987,
      "learning_rate": 1.5248684626478622e-05,
      "loss": 0.045,
      "step": 9355
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.0388673543930054,
      "learning_rate": 1.5244050052307829e-05,
      "loss": 0.0293,
      "step": 9360
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.8689137101173401,
      "learning_rate": 1.5239413923981365e-05,
      "loss": 0.0383,
      "step": 9365
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.059084177017212,
      "learning_rate": 1.5234776242873226e-05,
      "loss": 0.045,
      "step": 9370
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.9204691052436829,
      "learning_rate": 1.5230137010357852e-05,
      "loss": 0.0314,
      "step": 9375
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.969298779964447,
      "learning_rate": 1.522549622781016e-05,
      "loss": 0.0368,
      "step": 9380
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.2767212390899658,
      "learning_rate": 1.5220853896605508e-05,
      "loss": 0.0552,
      "step": 9385
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.731755256652832,
      "learning_rate": 1.5216210018119735e-05,
      "loss": 0.0419,
      "step": 9390
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.8874984383583069,
      "learning_rate": 1.5211564593729118e-05,
      "loss": 0.0392,
      "step": 9395
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.087302803993225,
      "learning_rate": 1.5206917624810404e-05,
      "loss": 0.0398,
      "step": 9400
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.6841315031051636,
      "learning_rate": 1.5202269112740795e-05,
      "loss": 0.0283,
      "step": 9405
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.8801833987236023,
      "learning_rate": 1.5197619058897945e-05,
      "loss": 0.041,
      "step": 9410
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.006748080253601,
      "learning_rate": 1.5192967464659976e-05,
      "loss": 0.048,
      "step": 9415
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.7825761437416077,
      "learning_rate": 1.5188314331405454e-05,
      "loss": 0.0514,
      "step": 9420
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.8253602981567383,
      "learning_rate": 1.5183659660513415e-05,
      "loss": 0.0369,
      "step": 9425
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.8670963048934937,
      "learning_rate": 1.5179003453363334e-05,
      "loss": 0.0303,
      "step": 9430
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.697066605091095,
      "learning_rate": 1.5174345711335157e-05,
      "loss": 0.0349,
      "step": 9435
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.9415462613105774,
      "learning_rate": 1.5169686435809275e-05,
      "loss": 0.039,
      "step": 9440
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.189955472946167,
      "learning_rate": 1.5165025628166538e-05,
      "loss": 0.0465,
      "step": 9445
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.801723837852478,
      "learning_rate": 1.5160363289788249e-05,
      "loss": 0.0294,
      "step": 9450
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.7355833649635315,
      "learning_rate": 1.515569942205616e-05,
      "loss": 0.0416,
      "step": 9455
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.7931221723556519,
      "learning_rate": 1.5151034026352492e-05,
      "loss": 0.0344,
      "step": 9460
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.8118154406547546,
      "learning_rate": 1.514636710405989e-05,
      "loss": 0.0364,
      "step": 9465
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.6183436512947083,
      "learning_rate": 1.5141698656561486e-05,
      "loss": 0.0297,
      "step": 9470
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.0786751508712769,
      "learning_rate": 1.5137028685240838e-05,
      "loss": 0.0464,
      "step": 9475
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.1322855949401855,
      "learning_rate": 1.5132357191481966e-05,
      "loss": 0.0668,
      "step": 9480
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.7116437554359436,
      "learning_rate": 1.512768417666934e-05,
      "loss": 0.0533,
      "step": 9485
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.8672776818275452,
      "learning_rate": 1.5123009642187879e-05,
      "loss": 0.0419,
      "step": 9490
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.6347663402557373,
      "learning_rate": 1.5118333589422953e-05,
      "loss": 0.0448,
      "step": 9495
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.9350044131278992,
      "learning_rate": 1.511365601976039e-05,
      "loss": 0.0416,
      "step": 9500
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.8266491889953613,
      "learning_rate": 1.5108976934586452e-05,
      "loss": 0.0377,
      "step": 9505
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.4777359962463379,
      "learning_rate": 1.5104296335287863e-05,
      "loss": 0.0273,
      "step": 9510
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.7290952801704407,
      "learning_rate": 1.5099614223251793e-05,
      "loss": 0.0391,
      "step": 9515
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.606194019317627,
      "learning_rate": 1.5094930599865852e-05,
      "loss": 0.0306,
      "step": 9520
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.8185514211654663,
      "learning_rate": 1.5090245466518112e-05,
      "loss": 0.0402,
      "step": 9525
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.31226646900177,
      "learning_rate": 1.5085558824597076e-05,
      "loss": 0.0432,
      "step": 9530
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.9394533634185791,
      "learning_rate": 1.5080870675491714e-05,
      "loss": 0.0419,
      "step": 9535
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.8299448490142822,
      "learning_rate": 1.5076181020591425e-05,
      "loss": 0.0301,
      "step": 9540
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.7116577625274658,
      "learning_rate": 1.5071489861286062e-05,
      "loss": 0.0246,
      "step": 9545
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.9290896058082581,
      "learning_rate": 1.5066797198965923e-05,
      "loss": 0.0386,
      "step": 9550
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.9579975008964539,
      "learning_rate": 1.5062103035021754e-05,
      "loss": 0.0434,
      "step": 9555
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.7418085336685181,
      "learning_rate": 1.5057407370844736e-05,
      "loss": 0.0359,
      "step": 9560
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.9520347118377686,
      "learning_rate": 1.5052710207826511e-05,
      "loss": 0.0497,
      "step": 9565
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.5138440132141113,
      "learning_rate": 1.504801154735915e-05,
      "loss": 0.027,
      "step": 9570
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.5888009071350098,
      "learning_rate": 1.5043311390835178e-05,
      "loss": 0.0337,
      "step": 9575
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.5701594352722168,
      "learning_rate": 1.5038609739647559e-05,
      "loss": 0.0351,
      "step": 9580
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.0097880363464355,
      "learning_rate": 1.5033906595189698e-05,
      "loss": 0.0332,
      "step": 9585
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.3016626834869385,
      "learning_rate": 1.5029201958855445e-05,
      "loss": 0.0408,
      "step": 9590
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.9115636348724365,
      "learning_rate": 1.5024495832039091e-05,
      "loss": 0.0517,
      "step": 9595
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.7627093195915222,
      "learning_rate": 1.5019788216135376e-05,
      "loss": 0.0424,
      "step": 9600
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6963844895362854,
      "learning_rate": 1.5015079112539468e-05,
      "loss": 0.0339,
      "step": 9605
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.7055118083953857,
      "learning_rate": 1.5010368522646986e-05,
      "loss": 0.0383,
      "step": 9610
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6209856271743774,
      "learning_rate": 1.5005656447853984e-05,
      "loss": 0.0436,
      "step": 9615
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5300926566123962,
      "learning_rate": 1.5000942889556965e-05,
      "loss": 0.0292,
      "step": 9620
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6936347484588623,
      "learning_rate": 1.499622784915286e-05,
      "loss": 0.0238,
      "step": 9625
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4455910325050354,
      "learning_rate": 1.4991511328039046e-05,
      "loss": 0.0218,
      "step": 9630
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.6577441096305847,
      "learning_rate": 1.4986793327613334e-05,
      "loss": 0.0235,
      "step": 9635
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.698058009147644,
      "learning_rate": 1.498207384927398e-05,
      "loss": 0.0157,
      "step": 9640
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.3731990158557892,
      "learning_rate": 1.4977352894419674e-05,
      "loss": 0.0199,
      "step": 9645
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.8323873281478882,
      "learning_rate": 1.4972630464449547e-05,
      "loss": 0.0175,
      "step": 9650
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.7002973556518555,
      "learning_rate": 1.4967906560763158e-05,
      "loss": 0.024,
      "step": 9655
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.751404345035553,
      "learning_rate": 1.4963181184760516e-05,
      "loss": 0.025,
      "step": 9660
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.7667729258537292,
      "learning_rate": 1.4958454337842054e-05,
      "loss": 0.0234,
      "step": 9665
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.39397770166397095,
      "learning_rate": 1.495372602140865e-05,
      "loss": 0.0135,
      "step": 9670
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5528745651245117,
      "learning_rate": 1.4948996236861613e-05,
      "loss": 0.0168,
      "step": 9675
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.8898313045501709,
      "learning_rate": 1.494426498560269e-05,
      "loss": 0.019,
      "step": 9680
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.6876255869865417,
      "learning_rate": 1.4939532269034058e-05,
      "loss": 0.025,
      "step": 9685
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.4512489438056946,
      "learning_rate": 1.493479808855833e-05,
      "loss": 0.0208,
      "step": 9690
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.8061213493347168,
      "learning_rate": 1.4930062445578557e-05,
      "loss": 0.0211,
      "step": 9695
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.3963974118232727,
      "learning_rate": 1.492532534149822e-05,
      "loss": 0.0144,
      "step": 9700
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.4634009301662445,
      "learning_rate": 1.4920586777721231e-05,
      "loss": 0.0178,
      "step": 9705
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.7113550901412964,
      "learning_rate": 1.491584675565194e-05,
      "loss": 0.0161,
      "step": 9710
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.6108189821243286,
      "learning_rate": 1.4911105276695126e-05,
      "loss": 0.0185,
      "step": 9715
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.8891080021858215,
      "learning_rate": 1.4906362342255996e-05,
      "loss": 0.0229,
      "step": 9720
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.5516975522041321,
      "learning_rate": 1.4901617953740196e-05,
      "loss": 0.0235,
      "step": 9725
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.7460342049598694,
      "learning_rate": 1.48968721125538e-05,
      "loss": 0.0264,
      "step": 9730
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.5245157480239868,
      "learning_rate": 1.4892124820103305e-05,
      "loss": 0.0217,
      "step": 9735
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.48179706931114197,
      "learning_rate": 1.4887376077795651e-05,
      "loss": 0.0146,
      "step": 9740
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.0343332290649414,
      "learning_rate": 1.4882625887038202e-05,
      "loss": 0.0228,
      "step": 9745
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.7103240489959717,
      "learning_rate": 1.4877874249238751e-05,
      "loss": 0.0198,
      "step": 9750
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.5932292938232422,
      "learning_rate": 1.4873121165805516e-05,
      "loss": 0.023,
      "step": 9755
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.48035189509391785,
      "learning_rate": 1.4868366638147145e-05,
      "loss": 0.0192,
      "step": 9760
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.2648720145225525,
      "learning_rate": 1.4863610667672723e-05,
      "loss": 0.0146,
      "step": 9765
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.3342938721179962,
      "learning_rate": 1.4858853255791754e-05,
      "loss": 0.0151,
      "step": 9770
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.8612216711044312,
      "learning_rate": 1.4854094403914167e-05,
      "loss": 0.0222,
      "step": 9775
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.5093660950660706,
      "learning_rate": 1.4849334113450322e-05,
      "loss": 0.0125,
      "step": 9780
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.5830073952674866,
      "learning_rate": 1.484457238581101e-05,
      "loss": 0.0205,
      "step": 9785
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.3554871678352356,
      "learning_rate": 1.4839809222407443e-05,
      "loss": 0.0193,
      "step": 9790
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.6692444086074829,
      "learning_rate": 1.4835044624651252e-05,
      "loss": 0.0181,
      "step": 9795
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.7528447508811951,
      "learning_rate": 1.4830278593954501e-05,
      "loss": 0.0182,
      "step": 9800
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.5975103974342346,
      "learning_rate": 1.4825511131729684e-05,
      "loss": 0.0216,
      "step": 9805
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.7612311840057373,
      "learning_rate": 1.4820742239389705e-05,
      "loss": 0.0207,
      "step": 9810
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.3679894208908081,
      "learning_rate": 1.48159719183479e-05,
      "loss": 0.016,
      "step": 9815
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.4748607575893402,
      "learning_rate": 1.4811200170018032e-05,
      "loss": 0.0219,
      "step": 9820
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.6697176098823547,
      "learning_rate": 1.480642699581428e-05,
      "loss": 0.0271,
      "step": 9825
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.5894784331321716,
      "learning_rate": 1.480165239715125e-05,
      "loss": 0.0217,
      "step": 9830
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.6295979619026184,
      "learning_rate": 1.4796876375443966e-05,
      "loss": 0.0236,
      "step": 9835
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.8625422716140747,
      "learning_rate": 1.4792098932107877e-05,
      "loss": 0.0214,
      "step": 9840
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.5024229288101196,
      "learning_rate": 1.4787320068558853e-05,
      "loss": 0.0191,
      "step": 9845
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.44266295433044434,
      "learning_rate": 1.4782539786213184e-05,
      "loss": 0.0175,
      "step": 9850
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.623786985874176,
      "learning_rate": 1.4777758086487581e-05,
      "loss": 0.0204,
      "step": 9855
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.5398198366165161,
      "learning_rate": 1.4772974970799177e-05,
      "loss": 0.0212,
      "step": 9860
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.4434782564640045,
      "learning_rate": 1.4768190440565517e-05,
      "loss": 0.0144,
      "step": 9865
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.8131916522979736,
      "learning_rate": 1.4763404497204578e-05,
      "loss": 0.0177,
      "step": 9870
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.6429072618484497,
      "learning_rate": 1.4758617142134749e-05,
      "loss": 0.022,
      "step": 9875
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.4914596974849701,
      "learning_rate": 1.4753828376774823e-05,
      "loss": 0.0151,
      "step": 9880
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.4990901052951813,
      "learning_rate": 1.4749038202544044e-05,
      "loss": 0.0187,
      "step": 9885
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.7438767552375793,
      "learning_rate": 1.4744246620862041e-05,
      "loss": 0.0179,
      "step": 9890
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.7084353566169739,
      "learning_rate": 1.4739453633148881e-05,
      "loss": 0.0201,
      "step": 9895
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.22215887904167175,
      "learning_rate": 1.4734659240825037e-05,
      "loss": 0.0173,
      "step": 9900
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.5357792377471924,
      "learning_rate": 1.47298634453114e-05,
      "loss": 0.0163,
      "step": 9905
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.48687347769737244,
      "learning_rate": 1.4725066248029283e-05,
      "loss": 0.0225,
      "step": 9910
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.421177476644516,
      "learning_rate": 1.472026765040041e-05,
      "loss": 0.0136,
      "step": 9915
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.3397599160671234,
      "learning_rate": 1.4715467653846917e-05,
      "loss": 0.019,
      "step": 9920
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.616377592086792,
      "learning_rate": 1.4710666259791353e-05,
      "loss": 0.0237,
      "step": 9925
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.5014125108718872,
      "learning_rate": 1.4705863469656697e-05,
      "loss": 0.0187,
      "step": 9930
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.8148841261863708,
      "learning_rate": 1.4701059284866323e-05,
      "loss": 0.0195,
      "step": 9935
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.9623835682868958,
      "learning_rate": 1.4696253706844026e-05,
      "loss": 0.0259,
      "step": 9940
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.41041064262390137,
      "learning_rate": 1.4691446737014012e-05,
      "loss": 0.0216,
      "step": 9945
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.5692448019981384,
      "learning_rate": 1.4686638376800907e-05,
      "loss": 0.0166,
      "step": 9950
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.5151691436767578,
      "learning_rate": 1.4681828627629742e-05,
      "loss": 0.0151,
      "step": 9955
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.6894605755805969,
      "learning_rate": 1.4677017490925956e-05,
      "loss": 0.0165,
      "step": 9960
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.5428750514984131,
      "learning_rate": 1.4672204968115407e-05,
      "loss": 0.0209,
      "step": 9965
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.6880739331245422,
      "learning_rate": 1.4667391060624364e-05,
      "loss": 0.0249,
      "step": 9970
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.49203822016716003,
      "learning_rate": 1.4662575769879497e-05,
      "loss": 0.0216,
      "step": 9975
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.5015491247177124,
      "learning_rate": 1.4657759097307895e-05,
      "loss": 0.015,
      "step": 9980
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.303508996963501,
      "learning_rate": 1.4652941044337057e-05,
      "loss": 0.0198,
      "step": 9985
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.5856837034225464,
      "learning_rate": 1.4648121612394884e-05,
      "loss": 0.0166,
      "step": 9990
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.7183772921562195,
      "learning_rate": 1.4643300802909687e-05,
      "loss": 0.0258,
      "step": 9995
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.8566582202911377,
      "learning_rate": 1.4638478617310194e-05,
      "loss": 0.027,
      "step": 10000
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.7371096014976501,
      "learning_rate": 1.4633655057025533e-05,
      "loss": 0.0208,
      "step": 10005
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.5916391015052795,
      "learning_rate": 1.4628830123485239e-05,
      "loss": 0.0172,
      "step": 10010
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.5476992726325989,
      "learning_rate": 1.4624003818119255e-05,
      "loss": 0.0142,
      "step": 10015
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.7170031070709229,
      "learning_rate": 1.4619176142357936e-05,
      "loss": 0.0203,
      "step": 10020
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.463932603597641,
      "learning_rate": 1.4614347097632031e-05,
      "loss": 0.0157,
      "step": 10025
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.2391690015792847,
      "learning_rate": 1.4609516685372711e-05,
      "loss": 0.025,
      "step": 10030
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.7538567781448364,
      "learning_rate": 1.4604684907011539e-05,
      "loss": 0.0164,
      "step": 10035
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.7815136909484863,
      "learning_rate": 1.4599851763980493e-05,
      "loss": 0.0223,
      "step": 10040
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.5280672907829285,
      "learning_rate": 1.459501725771194e-05,
      "loss": 0.0199,
      "step": 10045
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.6533296704292297,
      "learning_rate": 1.4590181389638667e-05,
      "loss": 0.0218,
      "step": 10050
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.8878480792045593,
      "learning_rate": 1.4585344161193862e-05,
      "loss": 0.0255,
      "step": 10055
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.6210035085678101,
      "learning_rate": 1.4580505573811105e-05,
      "loss": 0.0202,
      "step": 10060
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.6659073233604431,
      "learning_rate": 1.4575665628924392e-05,
      "loss": 0.0173,
      "step": 10065
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.6288996338844299,
      "learning_rate": 1.4570824327968111e-05,
      "loss": 0.0178,
      "step": 10070
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.5575762987136841,
      "learning_rate": 1.4565981672377065e-05,
      "loss": 0.0207,
      "step": 10075
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.6056348085403442,
      "learning_rate": 1.4561137663586444e-05,
      "loss": 0.0223,
      "step": 10080
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.4476945102214813,
      "learning_rate": 1.4556292303031847e-05,
      "loss": 0.0164,
      "step": 10085
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.841842532157898,
      "learning_rate": 1.4551445592149267e-05,
      "loss": 0.0203,
      "step": 10090
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.7189118266105652,
      "learning_rate": 1.454659753237511e-05,
      "loss": 0.0236,
      "step": 10095
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.5335758924484253,
      "learning_rate": 1.4541748125146172e-05,
      "loss": 0.0185,
      "step": 10100
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.668975293636322,
      "learning_rate": 1.4536897371899643e-05,
      "loss": 0.0192,
      "step": 10105
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.6786960363388062,
      "learning_rate": 1.4532045274073125e-05,
      "loss": 0.0167,
      "step": 10110
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.5628133416175842,
      "learning_rate": 1.4527191833104614e-05,
      "loss": 0.0158,
      "step": 10115
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.44337987899780273,
      "learning_rate": 1.4522337050432502e-05,
      "loss": 0.0126,
      "step": 10120
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.6237801313400269,
      "learning_rate": 1.4517480927495575e-05,
      "loss": 0.0164,
      "step": 10125
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.5822309255599976,
      "learning_rate": 1.4512623465733025e-05,
      "loss": 0.016,
      "step": 10130
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.6812189817428589,
      "learning_rate": 1.4507764666584431e-05,
      "loss": 0.0272,
      "step": 10135
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.6809319257736206,
      "learning_rate": 1.4502904531489778e-05,
      "loss": 0.018,
      "step": 10140
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.6676437854766846,
      "learning_rate": 1.4498043061889445e-05,
      "loss": 0.0127,
      "step": 10145
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.7798243761062622,
      "learning_rate": 1.4493180259224195e-05,
      "loss": 0.0167,
      "step": 10150
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.7573854327201843,
      "learning_rate": 1.4488316124935205e-05,
      "loss": 0.0202,
      "step": 10155
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.5917245745658875,
      "learning_rate": 1.448345066046403e-05,
      "loss": 0.0192,
      "step": 10160
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.6839116215705872,
      "learning_rate": 1.4478583867252629e-05,
      "loss": 0.0223,
      "step": 10165
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.5750868916511536,
      "learning_rate": 1.447371574674335e-05,
      "loss": 0.0231,
      "step": 10170
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.6704385876655579,
      "learning_rate": 1.4468846300378936e-05,
      "loss": 0.0201,
      "step": 10175
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.6824901103973389,
      "learning_rate": 1.4463975529602525e-05,
      "loss": 0.0225,
      "step": 10180
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.6210573315620422,
      "learning_rate": 1.4459103435857649e-05,
      "loss": 0.0223,
      "step": 10185
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.6658128499984741,
      "learning_rate": 1.445423002058822e-05,
      "loss": 0.0224,
      "step": 10190
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.5486401319503784,
      "learning_rate": 1.4449355285238553e-05,
      "loss": 0.0221,
      "step": 10195
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.792415201663971,
      "learning_rate": 1.4444479231253357e-05,
      "loss": 0.0218,
      "step": 10200
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.5030585527420044,
      "learning_rate": 1.4439601860077721e-05,
      "loss": 0.026,
      "step": 10205
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.9226459860801697,
      "learning_rate": 1.443472317315713e-05,
      "loss": 0.02,
      "step": 10210
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.5767743587493896,
      "learning_rate": 1.4429843171937459e-05,
      "loss": 0.0206,
      "step": 10215
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.5681598782539368,
      "learning_rate": 1.4424961857864976e-05,
      "loss": 0.0191,
      "step": 10220
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.6394040584564209,
      "learning_rate": 1.4420079232386329e-05,
      "loss": 0.0214,
      "step": 10225
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.8907806277275085,
      "learning_rate": 1.441519529694856e-05,
      "loss": 0.0291,
      "step": 10230
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.43610310554504395,
      "learning_rate": 1.4410310052999103e-05,
      "loss": 0.018,
      "step": 10235
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.4930350184440613,
      "learning_rate": 1.4405423501985776e-05,
      "loss": 0.0232,
      "step": 10240
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.5421867370605469,
      "learning_rate": 1.440053564535678e-05,
      "loss": 0.0184,
      "step": 10245
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.38000333309173584,
      "learning_rate": 1.4395646484560712e-05,
      "loss": 0.0177,
      "step": 10250
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.7972351312637329,
      "learning_rate": 1.4390756021046545e-05,
      "loss": 0.0219,
      "step": 10255
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.5814357995986938,
      "learning_rate": 1.4385864256263649e-05,
      "loss": 0.0309,
      "step": 10260
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.068788766860962,
      "learning_rate": 1.4380971191661773e-05,
      "loss": 0.0264,
      "step": 10265
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.6000214219093323,
      "learning_rate": 1.437607682869105e-05,
      "loss": 0.0211,
      "step": 10270
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.7504013180732727,
      "learning_rate": 1.4371181168802005e-05,
      "loss": 0.02,
      "step": 10275
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.37451502680778503,
      "learning_rate": 1.4366284213445538e-05,
      "loss": 0.0161,
      "step": 10280
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.5724549293518066,
      "learning_rate": 1.4361385964072943e-05,
      "loss": 0.0224,
      "step": 10285
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.44703173637390137,
      "learning_rate": 1.4356486422135888e-05,
      "loss": 0.0191,
      "step": 10290
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.3945390284061432,
      "learning_rate": 1.435158558908643e-05,
      "loss": 0.0219,
      "step": 10295
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.42554935812950134,
      "learning_rate": 1.4346683466377004e-05,
      "loss": 0.0225,
      "step": 10300
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.40748491883277893,
      "learning_rate": 1.4341780055460437e-05,
      "loss": 0.0205,
      "step": 10305
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.3901636004447937,
      "learning_rate": 1.4336875357789927e-05,
      "loss": 0.0184,
      "step": 10310
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.4704354405403137,
      "learning_rate": 1.4331969374819057e-05,
      "loss": 0.0153,
      "step": 10315
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.8930145502090454,
      "learning_rate": 1.4327062108001788e-05,
      "loss": 0.0204,
      "step": 10320
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.7317019701004028,
      "learning_rate": 1.4322153558792468e-05,
      "loss": 0.0142,
      "step": 10325
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.9982587695121765,
      "learning_rate": 1.4317243728645829e-05,
      "loss": 0.019,
      "step": 10330
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.6566906571388245,
      "learning_rate": 1.4312332619016964e-05,
      "loss": 0.0222,
      "step": 10335
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.5765242576599121,
      "learning_rate": 1.4307420231361357e-05,
      "loss": 0.0181,
      "step": 10340
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.8253374695777893,
      "learning_rate": 1.4302506567134881e-05,
      "loss": 0.0237,
      "step": 10345
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.43849310278892517,
      "learning_rate": 1.429759162779377e-05,
      "loss": 0.0182,
      "step": 10350
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.5556694269180298,
      "learning_rate": 1.429267541479464e-05,
      "loss": 0.0256,
      "step": 10355
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.6590076088905334,
      "learning_rate": 1.4287757929594488e-05,
      "loss": 0.018,
      "step": 10360
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.4504948854446411,
      "learning_rate": 1.4282839173650694e-05,
      "loss": 0.0216,
      "step": 10365
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.5615300536155701,
      "learning_rate": 1.4277919148421e-05,
      "loss": 0.0257,
      "step": 10370
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.49208933115005493,
      "learning_rate": 1.4272997855363532e-05,
      "loss": 0.0184,
      "step": 10375
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.47556278109550476,
      "learning_rate": 1.4268075295936796e-05,
      "loss": 0.0207,
      "step": 10380
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.0073806047439575,
      "learning_rate": 1.4263151471599668e-05,
      "loss": 0.0224,
      "step": 10385
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.6340844035148621,
      "learning_rate": 1.4258226383811397e-05,
      "loss": 0.0251,
      "step": 10390
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.6471311450004578,
      "learning_rate": 1.4253300034031613e-05,
      "loss": 0.0195,
      "step": 10395
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.81852787733078,
      "learning_rate": 1.4248372423720313e-05,
      "loss": 0.0253,
      "step": 10400
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.1031111478805542,
      "learning_rate": 1.4243443554337874e-05,
      "loss": 0.0176,
      "step": 10405
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.5694208145141602,
      "learning_rate": 1.423851342734504e-05,
      "loss": 0.0178,
      "step": 10410
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.4479517936706543,
      "learning_rate": 1.4233582044202932e-05,
      "loss": 0.0224,
      "step": 10415
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.5453998446464539,
      "learning_rate": 1.422864940637304e-05,
      "loss": 0.0133,
      "step": 10420
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.6944233179092407,
      "learning_rate": 1.4223715515317232e-05,
      "loss": 0.0171,
      "step": 10425
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.7661752104759216,
      "learning_rate": 1.421878037249774e-05,
      "loss": 0.0246,
      "step": 10430
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.1437700986862183,
      "learning_rate": 1.4213843979377173e-05,
      "loss": 0.0221,
      "step": 10435
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.5264135003089905,
      "learning_rate": 1.4208906337418502e-05,
      "loss": 0.0238,
      "step": 10440
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.4517305791378021,
      "learning_rate": 1.420396744808508e-05,
      "loss": 0.0239,
      "step": 10445
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.5580541491508484,
      "learning_rate": 1.419902731284062e-05,
      "loss": 0.0234,
      "step": 10450
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.43453770875930786,
      "learning_rate": 1.4194085933149212e-05,
      "loss": 0.0176,
      "step": 10455
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.7332136034965515,
      "learning_rate": 1.4189143310475306e-05,
      "loss": 0.0222,
      "step": 10460
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.6036754846572876,
      "learning_rate": 1.4184199446283726e-05,
      "loss": 0.0195,
      "step": 10465
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.4978627860546112,
      "learning_rate": 1.4179254342039665e-05,
      "loss": 0.0211,
      "step": 10470
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.6187092661857605,
      "learning_rate": 1.4174307999208679e-05,
      "loss": 0.022,
      "step": 10475
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.9858737587928772,
      "learning_rate": 1.4169360419256695e-05,
      "loss": 0.0197,
      "step": 10480
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.7682695984840393,
      "learning_rate": 1.4164411603650001e-05,
      "loss": 0.0292,
      "step": 10485
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.8246850371360779,
      "learning_rate": 1.4159461553855265e-05,
      "loss": 0.023,
      "step": 10490
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.6413214802742004,
      "learning_rate": 1.4154510271339504e-05,
      "loss": 0.0181,
      "step": 10495
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.546469509601593,
      "learning_rate": 1.4149557757570105e-05,
      "loss": 0.0214,
      "step": 10500
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.6325284242630005,
      "learning_rate": 1.4144604014014827e-05,
      "loss": 0.0296,
      "step": 10505
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.5793845057487488,
      "learning_rate": 1.413964904214179e-05,
      "loss": 0.022,
      "step": 10510
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.5916551351547241,
      "learning_rate": 1.4134692843419473e-05,
      "loss": 0.0218,
      "step": 10515
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.645812451839447,
      "learning_rate": 1.4129735419316725e-05,
      "loss": 0.0206,
      "step": 10520
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.526564359664917,
      "learning_rate": 1.4124776771302755e-05,
      "loss": 0.0277,
      "step": 10525
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.6979032158851624,
      "learning_rate": 1.4119816900847134e-05,
      "loss": 0.0227,
      "step": 10530
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.588740348815918,
      "learning_rate": 1.4114855809419798e-05,
      "loss": 0.0251,
      "step": 10535
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.74185711145401,
      "learning_rate": 1.4109893498491044e-05,
      "loss": 0.021,
      "step": 10540
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.7735092639923096,
      "learning_rate": 1.4104929969531527e-05,
      "loss": 0.0233,
      "step": 10545
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.47949472069740295,
      "learning_rate": 1.4099965224012269e-05,
      "loss": 0.0193,
      "step": 10550
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.35745835304260254,
      "learning_rate": 1.409499926340465e-05,
      "loss": 0.0228,
      "step": 10555
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.6792040467262268,
      "learning_rate": 1.4090032089180407e-05,
      "loss": 0.0196,
      "step": 10560
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.35544055700302124,
      "learning_rate": 1.408506370281164e-05,
      "loss": 0.0208,
      "step": 10565
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.7614099383354187,
      "learning_rate": 1.4080094105770808e-05,
      "loss": 0.0234,
      "step": 10570
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.5480696558952332,
      "learning_rate": 1.4075123299530726e-05,
      "loss": 0.026,
      "step": 10575
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.4874107241630554,
      "learning_rate": 1.4070151285564575e-05,
      "loss": 0.0192,
      "step": 10580
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.5505118370056152,
      "learning_rate": 1.4065178065345882e-05,
      "loss": 0.0185,
      "step": 10585
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.7177337408065796,
      "learning_rate": 1.4060203640348542e-05,
      "loss": 0.0243,
      "step": 10590
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.6744440197944641,
      "learning_rate": 1.40552280120468e-05,
      "loss": 0.0229,
      "step": 10595
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.6274732351303101,
      "learning_rate": 1.4050251181915268e-05,
      "loss": 0.0161,
      "step": 10600
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.5714767575263977,
      "learning_rate": 1.40452731514289e-05,
      "loss": 0.0198,
      "step": 10605
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.6027191281318665,
      "learning_rate": 1.4040293922063011e-05,
      "loss": 0.0272,
      "step": 10610
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.6635540127754211,
      "learning_rate": 1.4035313495293281e-05,
      "loss": 0.0274,
      "step": 10615
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.6043592691421509,
      "learning_rate": 1.403033187259573e-05,
      "loss": 0.0175,
      "step": 10620
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.4285273551940918,
      "learning_rate": 1.4025349055446741e-05,
      "loss": 0.0166,
      "step": 10625
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.7969894409179688,
      "learning_rate": 1.4020365045323047e-05,
      "loss": 0.0254,
      "step": 10630
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.6607228517532349,
      "learning_rate": 1.4015379843701747e-05,
      "loss": 0.0252,
      "step": 10635
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.6720756888389587,
      "learning_rate": 1.401039345206027e-05,
      "loss": 0.0248,
      "step": 10640
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.39906564354896545,
      "learning_rate": 1.4005405871876419e-05,
      "loss": 0.0195,
      "step": 10645
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.6192402839660645,
      "learning_rate": 1.4000417104628332e-05,
      "loss": 0.0281,
      "step": 10650
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.416329562664032,
      "learning_rate": 1.399542715179452e-05,
      "loss": 0.0259,
      "step": 10655
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.7425665855407715,
      "learning_rate": 1.3990436014853823e-05,
      "loss": 0.0219,
      "step": 10660
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.3826334774494171,
      "learning_rate": 1.3985443695285447e-05,
      "loss": 0.0241,
      "step": 10665
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.5703093409538269,
      "learning_rate": 1.3980450194568942e-05,
      "loss": 0.0154,
      "step": 10670
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.6470087170600891,
      "learning_rate": 1.3975455514184209e-05,
      "loss": 0.0162,
      "step": 10675
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.671140730381012,
      "learning_rate": 1.39704596556115e-05,
      "loss": 0.0222,
      "step": 10680
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.618836522102356,
      "learning_rate": 1.3965462620331414e-05,
      "loss": 0.0189,
      "step": 10685
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.49931779503822327,
      "learning_rate": 1.3960464409824899e-05,
      "loss": 0.0192,
      "step": 10690
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.7505419850349426,
      "learning_rate": 1.3955465025573256e-05,
      "loss": 0.0179,
      "step": 10695
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.0673131942749023,
      "learning_rate": 1.3950464469058126e-05,
      "loss": 0.0206,
      "step": 10700
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.4440382719039917,
      "learning_rate": 1.3945462741761509e-05,
      "loss": 0.0117,
      "step": 10705
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.48062238097190857,
      "learning_rate": 1.3940459845165735e-05,
      "loss": 0.0193,
      "step": 10710
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.620808482170105,
      "learning_rate": 1.3935455780753496e-05,
      "loss": 0.0225,
      "step": 10715
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.6680368185043335,
      "learning_rate": 1.3930450550007818e-05,
      "loss": 0.0148,
      "step": 10720
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.43228110671043396,
      "learning_rate": 1.3925444154412091e-05,
      "loss": 0.0226,
      "step": 10725
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.604020893573761,
      "learning_rate": 1.3920436595450025e-05,
      "loss": 0.0212,
      "step": 10730
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.8461281061172485,
      "learning_rate": 1.3915427874605695e-05,
      "loss": 0.018,
      "step": 10735
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.6334606409072876,
      "learning_rate": 1.3910417993363511e-05,
      "loss": 0.0216,
      "step": 10740
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.6531664133071899,
      "learning_rate": 1.3905406953208233e-05,
      "loss": 0.0271,
      "step": 10745
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.703217625617981,
      "learning_rate": 1.3900394755624955e-05,
      "loss": 0.0295,
      "step": 10750
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.6552525162696838,
      "learning_rate": 1.3895381402099119e-05,
      "loss": 0.0252,
      "step": 10755
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.9942788481712341,
      "learning_rate": 1.3890366894116517e-05,
      "loss": 0.0257,
      "step": 10760
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.4330316185951233,
      "learning_rate": 1.3885351233163271e-05,
      "loss": 0.0206,
      "step": 10765
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.6584552526473999,
      "learning_rate": 1.3880334420725854e-05,
      "loss": 0.0221,
      "step": 10770
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.38093966245651245,
      "learning_rate": 1.3875316458291069e-05,
      "loss": 0.0251,
      "step": 10775
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.5486773252487183,
      "learning_rate": 1.3870297347346075e-05,
      "loss": 0.0199,
      "step": 10780
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.45985671877861023,
      "learning_rate": 1.3865277089378357e-05,
      "loss": 0.0175,
      "step": 10785
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.54729825258255,
      "learning_rate": 1.3860255685875754e-05,
      "loss": 0.0193,
      "step": 10790
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.0338304042816162,
      "learning_rate": 1.385523313832643e-05,
      "loss": 0.0292,
      "step": 10795
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.6128621697425842,
      "learning_rate": 1.3850209448218897e-05,
      "loss": 0.0222,
      "step": 10800
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.6406121253967285,
      "learning_rate": 1.3845184617042005e-05,
      "loss": 0.0213,
      "step": 10805
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.40638235211372375,
      "learning_rate": 1.3840158646284938e-05,
      "loss": 0.0195,
      "step": 10810
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.47747060656547546,
      "learning_rate": 1.3835131537437223e-05,
      "loss": 0.0239,
      "step": 10815
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.5485616326332092,
      "learning_rate": 1.3830103291988723e-05,
      "loss": 0.0263,
      "step": 10820
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.7525022625923157,
      "learning_rate": 1.3825073911429634e-05,
      "loss": 0.0344,
      "step": 10825
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.6187776923179626,
      "learning_rate": 1.382004339725049e-05,
      "loss": 0.0183,
      "step": 10830
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.6603714823722839,
      "learning_rate": 1.3815011750942164e-05,
      "loss": 0.0219,
      "step": 10835
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.7578282356262207,
      "learning_rate": 1.3809978973995864e-05,
      "loss": 0.0167,
      "step": 10840
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.7237257361412048,
      "learning_rate": 1.3804945067903128e-05,
      "loss": 0.0253,
      "step": 10845
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.5169981122016907,
      "learning_rate": 1.3799910034155837e-05,
      "loss": 0.0209,
      "step": 10850
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.5577664971351624,
      "learning_rate": 1.3794873874246193e-05,
      "loss": 0.031,
      "step": 10855
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.9182189702987671,
      "learning_rate": 1.378983658966675e-05,
      "loss": 0.0224,
      "step": 10860
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.5072724223136902,
      "learning_rate": 1.3784798181910381e-05,
      "loss": 0.0176,
      "step": 10865
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.4871101975440979,
      "learning_rate": 1.3779758652470296e-05,
      "loss": 0.0249,
      "step": 10870
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.5638007521629333,
      "learning_rate": 1.3774718002840037e-05,
      "loss": 0.0246,
      "step": 10875
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.759212076663971,
      "learning_rate": 1.3769676234513481e-05,
      "loss": 0.0288,
      "step": 10880
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.6488536596298218,
      "learning_rate": 1.3764633348984837e-05,
      "loss": 0.0195,
      "step": 10885
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.522614061832428,
      "learning_rate": 1.3759589347748636e-05,
      "loss": 0.0252,
      "step": 10890
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.8051714301109314,
      "learning_rate": 1.3754544232299749e-05,
      "loss": 0.0217,
      "step": 10895
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.7797815799713135,
      "learning_rate": 1.374949800413337e-05,
      "loss": 0.023,
      "step": 10900
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.8559707999229431,
      "learning_rate": 1.374445066474504e-05,
      "loss": 0.0263,
      "step": 10905
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.8661820292472839,
      "learning_rate": 1.3739402215630602e-05,
      "loss": 0.0232,
      "step": 10910
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.6741891503334045,
      "learning_rate": 1.3734352658286252e-05,
      "loss": 0.0159,
      "step": 10915
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.5936999320983887,
      "learning_rate": 1.3729301994208495e-05,
      "loss": 0.026,
      "step": 10920
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.0814099311828613,
      "learning_rate": 1.3724250224894186e-05,
      "loss": 0.0246,
      "step": 10925
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.8289362192153931,
      "learning_rate": 1.3719197351840485e-05,
      "loss": 0.0342,
      "step": 10930
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.6794367432594299,
      "learning_rate": 1.3714143376544896e-05,
      "loss": 0.0182,
      "step": 10935
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.724094569683075,
      "learning_rate": 1.3709088300505237e-05,
      "loss": 0.0205,
      "step": 10940
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.6084389090538025,
      "learning_rate": 1.3704032125219664e-05,
      "loss": 0.021,
      "step": 10945
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.587928056716919,
      "learning_rate": 1.369897485218665e-05,
      "loss": 0.0161,
      "step": 10950
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.5043631792068481,
      "learning_rate": 1.3693916482905e-05,
      "loss": 0.0175,
      "step": 10955
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.6000282764434814,
      "learning_rate": 1.3688857018873831e-05,
      "loss": 0.0156,
      "step": 10960
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.5633286833763123,
      "learning_rate": 1.3683796461592604e-05,
      "loss": 0.0237,
      "step": 10965
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.647972583770752,
      "learning_rate": 1.367873481256109e-05,
      "loss": 0.0233,
      "step": 10970
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.755373477935791,
      "learning_rate": 1.367367207327939e-05,
      "loss": 0.0273,
      "step": 10975
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.5081478357315063,
      "learning_rate": 1.3668608245247916e-05,
      "loss": 0.0167,
      "step": 10980
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.5032436847686768,
      "learning_rate": 1.366354332996742e-05,
      "loss": 0.0195,
      "step": 10985
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.699909508228302,
      "learning_rate": 1.365847732893897e-05,
      "loss": 0.0277,
      "step": 10990
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.7266407012939453,
      "learning_rate": 1.3653410243663953e-05,
      "loss": 0.0218,
      "step": 10995
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.7415618896484375,
      "learning_rate": 1.3648342075644072e-05,
      "loss": 0.0206,
      "step": 11000
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.4602750539779663,
      "learning_rate": 1.3643272826381361e-05,
      "loss": 0.0212,
      "step": 11005
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.7643329501152039,
      "learning_rate": 1.3638202497378175e-05,
      "loss": 0.0271,
      "step": 11010
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.6913391947746277,
      "learning_rate": 1.3633131090137182e-05,
      "loss": 0.0175,
      "step": 11015
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.28211793303489685,
      "learning_rate": 1.3628058606161369e-05,
      "loss": 0.0187,
      "step": 11020
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.5162671208381653,
      "learning_rate": 1.3622985046954045e-05,
      "loss": 0.0208,
      "step": 11025
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.5287748575210571,
      "learning_rate": 1.3617910414018847e-05,
      "loss": 0.0233,
      "step": 11030
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.7692965865135193,
      "learning_rate": 1.3612834708859711e-05,
      "loss": 0.0233,
      "step": 11035
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.4626641273498535,
      "learning_rate": 1.3607757932980907e-05,
      "loss": 0.0157,
      "step": 11040
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.5515432357788086,
      "learning_rate": 1.360268008788701e-05,
      "loss": 0.0205,
      "step": 11045
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.39440417289733887,
      "learning_rate": 1.3597601175082931e-05,
      "loss": 0.0223,
      "step": 11050
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.8312334418296814,
      "learning_rate": 1.3592521196073872e-05,
      "loss": 0.0184,
      "step": 11055
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.3813522756099701,
      "learning_rate": 1.3587440152365368e-05,
      "loss": 0.0276,
      "step": 11060
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.4386773705482483,
      "learning_rate": 1.3582358045463264e-05,
      "loss": 0.0151,
      "step": 11065
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.6900048851966858,
      "learning_rate": 1.3577274876873727e-05,
      "loss": 0.0282,
      "step": 11070
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.7808547019958496,
      "learning_rate": 1.3572190648103227e-05,
      "loss": 0.0252,
      "step": 11075
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.9441418051719666,
      "learning_rate": 1.3567105360658555e-05,
      "loss": 0.0261,
      "step": 11080
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.8128235340118408,
      "learning_rate": 1.3562019016046814e-05,
      "loss": 0.0251,
      "step": 11085
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.0327625274658203,
      "learning_rate": 1.3556931615775426e-05,
      "loss": 0.0309,
      "step": 11090
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.5404015183448792,
      "learning_rate": 1.3551843161352116e-05,
      "loss": 0.02,
      "step": 11095
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.812623143196106,
      "learning_rate": 1.354675365428493e-05,
      "loss": 0.0291,
      "step": 11100
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.6458951830863953,
      "learning_rate": 1.354166309608222e-05,
      "loss": 0.0182,
      "step": 11105
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.45248669385910034,
      "learning_rate": 1.3536571488252655e-05,
      "loss": 0.029,
      "step": 11110
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.6244685649871826,
      "learning_rate": 1.3531478832305207e-05,
      "loss": 0.0227,
      "step": 11115
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.36896705627441406,
      "learning_rate": 1.3526385129749171e-05,
      "loss": 0.0174,
      "step": 11120
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.5725215077400208,
      "learning_rate": 1.3521290382094135e-05,
      "loss": 0.0224,
      "step": 11125
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.524311363697052,
      "learning_rate": 1.3516194590850014e-05,
      "loss": 0.0214,
      "step": 11130
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.6719222664833069,
      "learning_rate": 1.3511097757527023e-05,
      "loss": 0.0258,
      "step": 11135
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.7116886377334595,
      "learning_rate": 1.350599988363569e-05,
      "loss": 0.0254,
      "step": 11140
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.6066772937774658,
      "learning_rate": 1.3500900970686842e-05,
      "loss": 0.0203,
      "step": 11145
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.8623872995376587,
      "learning_rate": 1.349580102019163e-05,
      "loss": 0.0305,
      "step": 11150
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.6644154787063599,
      "learning_rate": 1.3490700033661497e-05,
      "loss": 0.0183,
      "step": 11155
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.7615149021148682,
      "learning_rate": 1.3485598012608204e-05,
      "loss": 0.021,
      "step": 11160
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.641488790512085,
      "learning_rate": 1.3480494958543811e-05,
      "loss": 0.0143,
      "step": 11165
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.6052054762840271,
      "learning_rate": 1.3475390872980688e-05,
      "loss": 0.0269,
      "step": 11170
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.8062578439712524,
      "learning_rate": 1.347028575743151e-05,
      "loss": 0.0278,
      "step": 11175
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.5274462103843689,
      "learning_rate": 1.3465179613409258e-05,
      "loss": 0.0219,
      "step": 11180
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.3634663224220276,
      "learning_rate": 1.3460072442427216e-05,
      "loss": 0.0174,
      "step": 11185
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.7998952269554138,
      "learning_rate": 1.3454964245998973e-05,
      "loss": 0.0263,
      "step": 11190
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.7923819422721863,
      "learning_rate": 1.344985502563843e-05,
      "loss": 0.0238,
      "step": 11195
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.5489522814750671,
      "learning_rate": 1.3444744782859772e-05,
      "loss": 0.017,
      "step": 11200
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.0061874389648438,
      "learning_rate": 1.3439633519177504e-05,
      "loss": 0.0281,
      "step": 11205
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.8746218681335449,
      "learning_rate": 1.343452123610643e-05,
      "loss": 0.0235,
      "step": 11210
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.9540253281593323,
      "learning_rate": 1.3429407935161651e-05,
      "loss": 0.0254,
      "step": 11215
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.6697771549224854,
      "learning_rate": 1.3424293617858577e-05,
      "loss": 0.0195,
      "step": 11220
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.3734327554702759,
      "learning_rate": 1.3419178285712912e-05,
      "loss": 0.0232,
      "step": 11225
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.6305715441703796,
      "learning_rate": 1.3414061940240669e-05,
      "loss": 0.0248,
      "step": 11230
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.7490862011909485,
      "learning_rate": 1.3408944582958147e-05,
      "loss": 0.0155,
      "step": 11235
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.45252835750579834,
      "learning_rate": 1.3403826215381965e-05,
      "loss": 0.0145,
      "step": 11240
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.6468937993049622,
      "learning_rate": 1.3398706839029024e-05,
      "loss": 0.0205,
      "step": 11245
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.5805210471153259,
      "learning_rate": 1.3393586455416534e-05,
      "loss": 0.023,
      "step": 11250
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.74920654296875,
      "learning_rate": 1.3388465066061997e-05,
      "loss": 0.0269,
      "step": 11255
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.5356035828590393,
      "learning_rate": 1.3383342672483219e-05,
      "loss": 0.0199,
      "step": 11260
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.4009453058242798,
      "learning_rate": 1.3378219276198303e-05,
      "loss": 0.0176,
      "step": 11265
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.7870226502418518,
      "learning_rate": 1.3373094878725638e-05,
      "loss": 0.0199,
      "step": 11270
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.760246217250824,
      "learning_rate": 1.336796948158393e-05,
      "loss": 0.024,
      "step": 11275
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.4638262689113617,
      "learning_rate": 1.336284308629216e-05,
      "loss": 0.0279,
      "step": 11280
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.5972824692726135,
      "learning_rate": 1.3357715694369626e-05,
      "loss": 0.0196,
      "step": 11285
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.6205165386199951,
      "learning_rate": 1.33525873073359e-05,
      "loss": 0.018,
      "step": 11290
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.7853833436965942,
      "learning_rate": 1.3347457926710863e-05,
      "loss": 0.0129,
      "step": 11295
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.7756471037864685,
      "learning_rate": 1.334232755401469e-05,
      "loss": 0.0228,
      "step": 11300
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.516747772693634,
      "learning_rate": 1.3337196190767841e-05,
      "loss": 0.0215,
      "step": 11305
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.693603515625,
      "learning_rate": 1.3332063838491077e-05,
      "loss": 0.0294,
      "step": 11310
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.0093333721160889,
      "learning_rate": 1.3326930498705449e-05,
      "loss": 0.0257,
      "step": 11315
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.6264593005180359,
      "learning_rate": 1.3321796172932306e-05,
      "loss": 0.0263,
      "step": 11320
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.7026219367980957,
      "learning_rate": 1.3316660862693285e-05,
      "loss": 0.0198,
      "step": 11325
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.6751310229301453,
      "learning_rate": 1.3311524569510309e-05,
      "loss": 0.0261,
      "step": 11330
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.5312600135803223,
      "learning_rate": 1.3306387294905601e-05,
      "loss": 0.0269,
      "step": 11335
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.8188959956169128,
      "learning_rate": 1.3301249040401677e-05,
      "loss": 0.0333,
      "step": 11340
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.0527750253677368,
      "learning_rate": 1.3296109807521336e-05,
      "loss": 0.0257,
      "step": 11345
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.3274320960044861,
      "learning_rate": 1.3290969597787664e-05,
      "loss": 0.024,
      "step": 11350
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.42005568742752075,
      "learning_rate": 1.3285828412724046e-05,
      "loss": 0.0259,
      "step": 11355
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.5386154055595398,
      "learning_rate": 1.3280686253854156e-05,
      "loss": 0.0206,
      "step": 11360
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.5777267813682556,
      "learning_rate": 1.3275543122701946e-05,
      "loss": 0.0247,
      "step": 11365
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.5420946478843689,
      "learning_rate": 1.3270399020791667e-05,
      "loss": 0.022,
      "step": 11370
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.4722997844219208,
      "learning_rate": 1.3265253949647851e-05,
      "loss": 0.0253,
      "step": 11375
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.6599250435829163,
      "learning_rate": 1.3260107910795324e-05,
      "loss": 0.021,
      "step": 11380
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.6490116715431213,
      "learning_rate": 1.3254960905759186e-05,
      "loss": 0.0289,
      "step": 11385
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.6252589225769043,
      "learning_rate": 1.3249812936064844e-05,
      "loss": 0.0246,
      "step": 11390
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.7048884034156799,
      "learning_rate": 1.3244664003237967e-05,
      "loss": 0.0204,
      "step": 11395
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.6130014657974243,
      "learning_rate": 1.3239514108804525e-05,
      "loss": 0.0217,
      "step": 11400
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.7588436007499695,
      "learning_rate": 1.3234363254290772e-05,
      "loss": 0.027,
      "step": 11405
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.7372493743896484,
      "learning_rate": 1.3229211441223244e-05,
      "loss": 0.0277,
      "step": 11410
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.7405737638473511,
      "learning_rate": 1.3224058671128757e-05,
      "loss": 0.0206,
      "step": 11415
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.7434006929397583,
      "learning_rate": 1.3218904945534416e-05,
      "loss": 0.0238,
      "step": 11420
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.5010198354721069,
      "learning_rate": 1.3213750265967609e-05,
      "loss": 0.0275,
      "step": 11425
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.5099480152130127,
      "learning_rate": 1.3208594633956007e-05,
      "loss": 0.0273,
      "step": 11430
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.738741934299469,
      "learning_rate": 1.3203438051027552e-05,
      "loss": 0.0281,
      "step": 11435
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.6779162883758545,
      "learning_rate": 1.3198280518710489e-05,
      "loss": 0.0223,
      "step": 11440
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.335631787776947,
      "learning_rate": 1.319312203853333e-05,
      "loss": 0.0188,
      "step": 11445
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.3947046995162964,
      "learning_rate": 1.3187962612024866e-05,
      "loss": 0.0217,
      "step": 11450
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.5501173138618469,
      "learning_rate": 1.3182802240714176e-05,
      "loss": 0.0268,
      "step": 11455
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.38181018829345703,
      "learning_rate": 1.3177640926130615e-05,
      "loss": 0.0246,
      "step": 11460
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.8614093065261841,
      "learning_rate": 1.3172478669803823e-05,
      "loss": 0.0203,
      "step": 11465
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.6074875593185425,
      "learning_rate": 1.3167315473263709e-05,
      "loss": 0.0197,
      "step": 11470
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.606174886226654,
      "learning_rate": 1.3162151338040471e-05,
      "loss": 0.0206,
      "step": 11475
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.5583638548851013,
      "learning_rate": 1.3156986265664575e-05,
      "loss": 0.0267,
      "step": 11480
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.4318353235721588,
      "learning_rate": 1.315182025766678e-05,
      "loss": 0.0256,
      "step": 11485
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.6225631237030029,
      "learning_rate": 1.3146653315578103e-05,
      "loss": 0.0246,
      "step": 11490
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.5729676485061646,
      "learning_rate": 1.3141485440929853e-05,
      "loss": 0.0265,
      "step": 11495
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.40861061215400696,
      "learning_rate": 1.313631663525361e-05,
      "loss": 0.0272,
      "step": 11500
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.6979518532752991,
      "learning_rate": 1.3131146900081226e-05,
      "loss": 0.0152,
      "step": 11505
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.5879753828048706,
      "learning_rate": 1.3125976236944835e-05,
      "loss": 0.0196,
      "step": 11510
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.6799219250679016,
      "learning_rate": 1.3120804647376844e-05,
      "loss": 0.0233,
      "step": 11515
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.5397365093231201,
      "learning_rate": 1.311563213290993e-05,
      "loss": 0.0217,
      "step": 11520
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.48731374740600586,
      "learning_rate": 1.3110458695077054e-05,
      "loss": 0.0202,
      "step": 11525
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.9724308252334595,
      "learning_rate": 1.3105284335411436e-05,
      "loss": 0.0248,
      "step": 11530
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.7429783940315247,
      "learning_rate": 1.310010905544659e-05,
      "loss": 0.0217,
      "step": 11535
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.6432991623878479,
      "learning_rate": 1.3094932856716279e-05,
      "loss": 0.0301,
      "step": 11540
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.8815121650695801,
      "learning_rate": 1.3089755740754554e-05,
      "loss": 0.0181,
      "step": 11545
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.39547908306121826,
      "learning_rate": 1.3084577709095738e-05,
      "loss": 0.0183,
      "step": 11550
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.6698602437973022,
      "learning_rate": 1.3079398763274418e-05,
      "loss": 0.0302,
      "step": 11555
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.48663145303726196,
      "learning_rate": 1.3074218904825451e-05,
      "loss": 0.0168,
      "step": 11560
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.5853646397590637,
      "learning_rate": 1.3069038135283974e-05,
      "loss": 0.0218,
      "step": 11565
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.675857424736023,
      "learning_rate": 1.3063856456185384e-05,
      "loss": 0.0193,
      "step": 11570
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.6646441221237183,
      "learning_rate": 1.3058673869065359e-05,
      "loss": 0.0239,
      "step": 11575
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.5286111831665039,
      "learning_rate": 1.3053490375459831e-05,
      "loss": 0.0188,
      "step": 11580
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.5661985874176025,
      "learning_rate": 1.3048305976905011e-05,
      "loss": 0.0175,
      "step": 11585
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.5855173468589783,
      "learning_rate": 1.3043120674937384e-05,
      "loss": 0.0195,
      "step": 11590
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.9668726325035095,
      "learning_rate": 1.3037934471093683e-05,
      "loss": 0.0266,
      "step": 11595
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.8840664625167847,
      "learning_rate": 1.3032747366910927e-05,
      "loss": 0.0219,
      "step": 11600
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.40016576647758484,
      "learning_rate": 1.302755936392639e-05,
      "loss": 0.0203,
      "step": 11605
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.678231418132782,
      "learning_rate": 1.3022370463677627e-05,
      "loss": 0.0301,
      "step": 11610
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.6121237874031067,
      "learning_rate": 1.3017180667702438e-05,
      "loss": 0.0251,
      "step": 11615
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.9894143342971802,
      "learning_rate": 1.3011989977538907e-05,
      "loss": 0.0212,
      "step": 11620
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.5731955170631409,
      "learning_rate": 1.300679839472537e-05,
      "loss": 0.0144,
      "step": 11625
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.540483295917511,
      "learning_rate": 1.3001605920800438e-05,
      "loss": 0.0267,
      "step": 11630
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.6893200874328613,
      "learning_rate": 1.2996412557302977e-05,
      "loss": 0.0182,
      "step": 11635
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.7962963581085205,
      "learning_rate": 1.2991218305772126e-05,
      "loss": 0.0231,
      "step": 11640
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.860883355140686,
      "learning_rate": 1.2986023167747278e-05,
      "loss": 0.0264,
      "step": 11645
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.4079477787017822,
      "learning_rate": 1.2980827144768093e-05,
      "loss": 0.0306,
      "step": 11650
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.602584183216095,
      "learning_rate": 1.2975630238374498e-05,
      "loss": 0.0236,
      "step": 11655
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.80098956823349,
      "learning_rate": 1.2970432450106669e-05,
      "loss": 0.0284,
      "step": 11660
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.7290306091308594,
      "learning_rate": 1.2965233781505058e-05,
      "loss": 0.0208,
      "step": 11665
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.8001556396484375,
      "learning_rate": 1.2960034234110368e-05,
      "loss": 0.0281,
      "step": 11670
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.8522571921348572,
      "learning_rate": 1.2954833809463567e-05,
      "loss": 0.0239,
      "step": 11675
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.5874483585357666,
      "learning_rate": 1.2949632509105883e-05,
      "loss": 0.0288,
      "step": 11680
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.6494091749191284,
      "learning_rate": 1.2944430334578795e-05,
      "loss": 0.0289,
      "step": 11685
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.5915369987487793,
      "learning_rate": 1.2939227287424056e-05,
      "loss": 0.0224,
      "step": 11690
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.832787275314331,
      "learning_rate": 1.2934023369183669e-05,
      "loss": 0.0267,
      "step": 11695
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.6795758605003357,
      "learning_rate": 1.2928818581399893e-05,
      "loss": 0.0281,
      "step": 11700
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.8091376423835754,
      "learning_rate": 1.2923612925615245e-05,
      "loss": 0.0228,
      "step": 11705
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.6956303715705872,
      "learning_rate": 1.2918406403372508e-05,
      "loss": 0.0222,
      "step": 11710
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.7285389304161072,
      "learning_rate": 1.2913199016214715e-05,
      "loss": 0.0288,
      "step": 11715
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.6634201407432556,
      "learning_rate": 1.2907990765685155e-05,
      "loss": 0.0229,
      "step": 11720
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.8717896342277527,
      "learning_rate": 1.290278165332737e-05,
      "loss": 0.0217,
      "step": 11725
    },
    {
      "epoch": 3.66,
      "grad_norm": 2.627668857574463,
      "learning_rate": 1.2897571680685164e-05,
      "loss": 0.0248,
      "step": 11730
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.5117378830909729,
      "learning_rate": 1.2892360849302598e-05,
      "loss": 0.022,
      "step": 11735
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.5402873754501343,
      "learning_rate": 1.2887149160723978e-05,
      "loss": 0.0245,
      "step": 11740
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.8802453875541687,
      "learning_rate": 1.2881936616493867e-05,
      "loss": 0.0259,
      "step": 11745
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.6248879432678223,
      "learning_rate": 1.2876723218157086e-05,
      "loss": 0.0229,
      "step": 11750
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.6852085590362549,
      "learning_rate": 1.287150896725871e-05,
      "loss": 0.0322,
      "step": 11755
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.573294460773468,
      "learning_rate": 1.2866293865344058e-05,
      "loss": 0.0221,
      "step": 11760
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.847869873046875,
      "learning_rate": 1.2861077913958708e-05,
      "loss": 0.024,
      "step": 11765
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.7071881294250488,
      "learning_rate": 1.285586111464849e-05,
      "loss": 0.0202,
      "step": 11770
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.7383145093917847,
      "learning_rate": 1.2850643468959479e-05,
      "loss": 0.0211,
      "step": 11775
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.8499407768249512,
      "learning_rate": 1.284542497843801e-05,
      "loss": 0.0278,
      "step": 11780
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.7504806518554688,
      "learning_rate": 1.2840205644630662e-05,
      "loss": 0.0174,
      "step": 11785
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.0881952047348022,
      "learning_rate": 1.2834985469084265e-05,
      "loss": 0.0276,
      "step": 11790
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.6055206656455994,
      "learning_rate": 1.28297644533459e-05,
      "loss": 0.0235,
      "step": 11795
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.7683250308036804,
      "learning_rate": 1.2824542598962895e-05,
      "loss": 0.025,
      "step": 11800
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.2123945951461792,
      "learning_rate": 1.2819319907482832e-05,
      "loss": 0.0241,
      "step": 11805
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.546367347240448,
      "learning_rate": 1.281409638045353e-05,
      "loss": 0.019,
      "step": 11810
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.5594978332519531,
      "learning_rate": 1.2808872019423064e-05,
      "loss": 0.0212,
      "step": 11815
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.5757129192352295,
      "learning_rate": 1.2803646825939758e-05,
      "loss": 0.0239,
      "step": 11820
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.0514737367630005,
      "learning_rate": 1.2798420801552181e-05,
      "loss": 0.0242,
      "step": 11825
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.6186000108718872,
      "learning_rate": 1.2793193947809135e-05,
      "loss": 0.0234,
      "step": 11830
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.3759470283985138,
      "learning_rate": 1.2787966266259693e-05,
      "loss": 0.0176,
      "step": 11835
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.6968141198158264,
      "learning_rate": 1.278273775845315e-05,
      "loss": 0.0255,
      "step": 11840
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.48914167284965515,
      "learning_rate": 1.2777508425939062e-05,
      "loss": 0.0258,
      "step": 11845
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.7071937918663025,
      "learning_rate": 1.2772278270267218e-05,
      "loss": 0.0297,
      "step": 11850
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.3311907649040222,
      "learning_rate": 1.2767047292987658e-05,
      "loss": 0.0299,
      "step": 11855
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.714726984500885,
      "learning_rate": 1.2761815495650666e-05,
      "loss": 0.0228,
      "step": 11860
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.4907616674900055,
      "learning_rate": 1.2756582879806761e-05,
      "loss": 0.0243,
      "step": 11865
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.594176709651947,
      "learning_rate": 1.2751349447006713e-05,
      "loss": 0.0275,
      "step": 11870
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.5221083760261536,
      "learning_rate": 1.2746115198801527e-05,
      "loss": 0.018,
      "step": 11875
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.5824384689331055,
      "learning_rate": 1.2740880136742464e-05,
      "loss": 0.0207,
      "step": 11880
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.6119704246520996,
      "learning_rate": 1.2735644262381006e-05,
      "loss": 0.0223,
      "step": 11885
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.7727152109146118,
      "learning_rate": 1.2730407577268888e-05,
      "loss": 0.0163,
      "step": 11890
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.5603602528572083,
      "learning_rate": 1.2725170082958088e-05,
      "loss": 0.0286,
      "step": 11895
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.6091859340667725,
      "learning_rate": 1.271993178100081e-05,
      "loss": 0.0294,
      "step": 11900
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.9491273164749146,
      "learning_rate": 1.2714692672949513e-05,
      "loss": 0.0308,
      "step": 11905
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.4891149401664734,
      "learning_rate": 1.2709452760356884e-05,
      "loss": 0.0243,
      "step": 11910
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.5642223358154297,
      "learning_rate": 1.2704212044775856e-05,
      "loss": 0.0282,
      "step": 11915
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.42882299423217773,
      "learning_rate": 1.2698970527759595e-05,
      "loss": 0.027,
      "step": 11920
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.4645093083381653,
      "learning_rate": 1.2693728210861505e-05,
      "loss": 0.0252,
      "step": 11925
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.7219882607460022,
      "learning_rate": 1.2688485095635229e-05,
      "loss": 0.017,
      "step": 11930
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.4665991961956024,
      "learning_rate": 1.2683241183634645e-05,
      "loss": 0.0254,
      "step": 11935
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.8865329623222351,
      "learning_rate": 1.2677996476413868e-05,
      "loss": 0.0272,
      "step": 11940
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.7679188847541809,
      "learning_rate": 1.2672750975527249e-05,
      "loss": 0.0269,
      "step": 11945
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.4800529479980469,
      "learning_rate": 1.2667504682529375e-05,
      "loss": 0.0183,
      "step": 11950
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.8061618804931641,
      "learning_rate": 1.266225759897506e-05,
      "loss": 0.0288,
      "step": 11955
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.6751968860626221,
      "learning_rate": 1.2657009726419368e-05,
      "loss": 0.0217,
      "step": 11960
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.6072093844413757,
      "learning_rate": 1.2651761066417581e-05,
      "loss": 0.024,
      "step": 11965
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.9960513114929199,
      "learning_rate": 1.2646511620525224e-05,
      "loss": 0.025,
      "step": 11970
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.3995935022830963,
      "learning_rate": 1.2641261390298048e-05,
      "loss": 0.0228,
      "step": 11975
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.6832537055015564,
      "learning_rate": 1.2636010377292042e-05,
      "loss": 0.0258,
      "step": 11980
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.5022556781768799,
      "learning_rate": 1.2630758583063428e-05,
      "loss": 0.0293,
      "step": 11985
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.8268438577651978,
      "learning_rate": 1.2625506009168658e-05,
      "loss": 0.0304,
      "step": 11990
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.5869343876838684,
      "learning_rate": 1.2620252657164404e-05,
      "loss": 0.0274,
      "step": 11995
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.0159419775009155,
      "learning_rate": 1.2614998528607589e-05,
      "loss": 0.0223,
      "step": 12000
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.9052306413650513,
      "learning_rate": 1.2609743625055353e-05,
      "loss": 0.0303,
      "step": 12005
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.7687567472457886,
      "learning_rate": 1.2604487948065066e-05,
      "loss": 0.0248,
      "step": 12010
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.9072149991989136,
      "learning_rate": 1.2599231499194328e-05,
      "loss": 0.0217,
      "step": 12015
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.5874031186103821,
      "learning_rate": 1.2593974280000971e-05,
      "loss": 0.025,
      "step": 12020
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.3685324490070343,
      "learning_rate": 1.258871629204306e-05,
      "loss": 0.019,
      "step": 12025
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.5925565361976624,
      "learning_rate": 1.2583457536878872e-05,
      "loss": 0.0215,
      "step": 12030
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.7138723731040955,
      "learning_rate": 1.2578198016066925e-05,
      "loss": 0.0311,
      "step": 12035
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.7673307657241821,
      "learning_rate": 1.2572937731165958e-05,
      "loss": 0.0156,
      "step": 12040
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.9401269555091858,
      "learning_rate": 1.256767668373494e-05,
      "loss": 0.0338,
      "step": 12045
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.7034611701965332,
      "learning_rate": 1.2562414875333062e-05,
      "loss": 0.0218,
      "step": 12050
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.6674792766571045,
      "learning_rate": 1.2557152307519746e-05,
      "loss": 0.0235,
      "step": 12055
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.6051164865493774,
      "learning_rate": 1.2551888981854632e-05,
      "loss": 0.0246,
      "step": 12060
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.6730068325996399,
      "learning_rate": 1.254662489989759e-05,
      "loss": 0.0204,
      "step": 12065
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.6211580038070679,
      "learning_rate": 1.2541360063208714e-05,
      "loss": 0.0263,
      "step": 12070
    },
    {
      "epoch": 3.77,
      "grad_norm": 2.0400476455688477,
      "learning_rate": 1.2536094473348317e-05,
      "loss": 0.0257,
      "step": 12075
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.3689804673194885,
      "learning_rate": 1.253082813187694e-05,
      "loss": 0.0282,
      "step": 12080
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.5824944376945496,
      "learning_rate": 1.2525561040355342e-05,
      "loss": 0.0259,
      "step": 12085
    },
    {
      "epoch": 3.77,
      "grad_norm": 1.1334221363067627,
      "learning_rate": 1.2520293200344516e-05,
      "loss": 0.0241,
      "step": 12090
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.6835818886756897,
      "learning_rate": 1.2515024613405662e-05,
      "loss": 0.0199,
      "step": 12095
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.6704448461532593,
      "learning_rate": 1.2509755281100201e-05,
      "loss": 0.0245,
      "step": 12100
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.8395538926124573,
      "learning_rate": 1.2504485204989794e-05,
      "loss": 0.0192,
      "step": 12105
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.6028382182121277,
      "learning_rate": 1.2499214386636302e-05,
      "loss": 0.0314,
      "step": 12110
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.4773300290107727,
      "learning_rate": 1.2493942827601819e-05,
      "loss": 0.0155,
      "step": 12115
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.4813371002674103,
      "learning_rate": 1.2488670529448648e-05,
      "loss": 0.0191,
      "step": 12120
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.48700809478759766,
      "learning_rate": 1.248339749373932e-05,
      "loss": 0.032,
      "step": 12125
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.5969123244285583,
      "learning_rate": 1.2478123722036581e-05,
      "loss": 0.0228,
      "step": 12130
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.587505578994751,
      "learning_rate": 1.2472849215903393e-05,
      "loss": 0.0185,
      "step": 12135
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.7423061728477478,
      "learning_rate": 1.2467573976902936e-05,
      "loss": 0.0303,
      "step": 12140
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.6694118976593018,
      "learning_rate": 1.2462298006598616e-05,
      "loss": 0.0214,
      "step": 12145
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.8038167357444763,
      "learning_rate": 1.2457021306554045e-05,
      "loss": 0.0264,
      "step": 12150
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.6632444858551025,
      "learning_rate": 1.2451743878333053e-05,
      "loss": 0.0287,
      "step": 12155
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.6460713148117065,
      "learning_rate": 1.244646572349969e-05,
      "loss": 0.0291,
      "step": 12160
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.5237529277801514,
      "learning_rate": 1.2441186843618216e-05,
      "loss": 0.0225,
      "step": 12165
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.7596482634544373,
      "learning_rate": 1.2435907240253116e-05,
      "loss": 0.0281,
      "step": 12170
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.7577078938484192,
      "learning_rate": 1.2430626914969074e-05,
      "loss": 0.0269,
      "step": 12175
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.4984164834022522,
      "learning_rate": 1.2425345869331004e-05,
      "loss": 0.0251,
      "step": 12180
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.62168949842453,
      "learning_rate": 1.2420064104904022e-05,
      "loss": 0.0178,
      "step": 12185
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.8107472658157349,
      "learning_rate": 1.241478162325346e-05,
      "loss": 0.0236,
      "step": 12190
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.5770743489265442,
      "learning_rate": 1.2409498425944864e-05,
      "loss": 0.0222,
      "step": 12195
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.6336897015571594,
      "learning_rate": 1.2404214514543993e-05,
      "loss": 0.0204,
      "step": 12200
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.8589664697647095,
      "learning_rate": 1.2398929890616816e-05,
      "loss": 0.0263,
      "step": 12205
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.4037541449069977,
      "learning_rate": 1.239364455572951e-05,
      "loss": 0.0233,
      "step": 12210
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.8225604295730591,
      "learning_rate": 1.2388358511448475e-05,
      "loss": 0.0187,
      "step": 12215
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.7085739970207214,
      "learning_rate": 1.2383071759340305e-05,
      "loss": 0.0329,
      "step": 12220
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.735085666179657,
      "learning_rate": 1.2377784300971807e-05,
      "loss": 0.0202,
      "step": 12225
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.9221487045288086,
      "learning_rate": 1.2372496137910011e-05,
      "loss": 0.0263,
      "step": 12230
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.550031840801239,
      "learning_rate": 1.2367207271722139e-05,
      "loss": 0.0212,
      "step": 12235
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.744546115398407,
      "learning_rate": 1.2361917703975636e-05,
      "loss": 0.0256,
      "step": 12240
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.4377955496311188,
      "learning_rate": 1.235662743623814e-05,
      "loss": 0.0167,
      "step": 12245
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.7871254086494446,
      "learning_rate": 1.2351336470077507e-05,
      "loss": 0.0223,
      "step": 12250
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.8724322319030762,
      "learning_rate": 1.2346044807061798e-05,
      "loss": 0.0266,
      "step": 12255
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.7433586120605469,
      "learning_rate": 1.2340752448759281e-05,
      "loss": 0.022,
      "step": 12260
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.7378526926040649,
      "learning_rate": 1.233545939673842e-05,
      "loss": 0.0291,
      "step": 12265
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.7587401270866394,
      "learning_rate": 1.2330165652567905e-05,
      "loss": 0.0273,
      "step": 12270
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.0976639986038208,
      "learning_rate": 1.2324871217816615e-05,
      "loss": 0.0233,
      "step": 12275
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.8238142728805542,
      "learning_rate": 1.2319576094053631e-05,
      "loss": 0.0265,
      "step": 12280
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.5005680918693542,
      "learning_rate": 1.2314280282848254e-05,
      "loss": 0.0308,
      "step": 12285
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.631725013256073,
      "learning_rate": 1.2308983785769975e-05,
      "loss": 0.0207,
      "step": 12290
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.770847499370575,
      "learning_rate": 1.2303686604388498e-05,
      "loss": 0.0311,
      "step": 12295
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.6609355211257935,
      "learning_rate": 1.2298388740273719e-05,
      "loss": 0.0316,
      "step": 12300
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.7061938643455505,
      "learning_rate": 1.2293090194995746e-05,
      "loss": 0.019,
      "step": 12305
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.6347512602806091,
      "learning_rate": 1.2287790970124884e-05,
      "loss": 0.0243,
      "step": 12310
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.8019315600395203,
      "learning_rate": 1.2282491067231641e-05,
      "loss": 0.016,
      "step": 12315
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.5562499165534973,
      "learning_rate": 1.2277190487886725e-05,
      "loss": 0.0199,
      "step": 12320
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.8764368295669556,
      "learning_rate": 1.2271889233661045e-05,
      "loss": 0.03,
      "step": 12325
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.6134083867073059,
      "learning_rate": 1.2266587306125713e-05,
      "loss": 0.0287,
      "step": 12330
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.7679027318954468,
      "learning_rate": 1.2261284706852032e-05,
      "loss": 0.0327,
      "step": 12335
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.46010440587997437,
      "learning_rate": 1.2255981437411513e-05,
      "loss": 0.0202,
      "step": 12340
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.033030390739441,
      "learning_rate": 1.2250677499375863e-05,
      "loss": 0.0283,
      "step": 12345
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.5734465718269348,
      "learning_rate": 1.2245372894316984e-05,
      "loss": 0.0214,
      "step": 12350
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.6062762141227722,
      "learning_rate": 1.224006762380698e-05,
      "loss": 0.03,
      "step": 12355
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.2957155704498291,
      "learning_rate": 1.223476168941815e-05,
      "loss": 0.0162,
      "step": 12360
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.6891098618507385,
      "learning_rate": 1.222945509272299e-05,
      "loss": 0.0276,
      "step": 12365
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.838630199432373,
      "learning_rate": 1.222414783529419e-05,
      "loss": 0.0305,
      "step": 12370
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.6958526968955994,
      "learning_rate": 1.221883991870464e-05,
      "loss": 0.0276,
      "step": 12375
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.42833009362220764,
      "learning_rate": 1.2213531344527426e-05,
      "loss": 0.0253,
      "step": 12380
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.3405981659889221,
      "learning_rate": 1.2208222114335826e-05,
      "loss": 0.0227,
      "step": 12385
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.5990917682647705,
      "learning_rate": 1.2202912229703305e-05,
      "loss": 0.0214,
      "step": 12390
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.7133599519729614,
      "learning_rate": 1.219760169220354e-05,
      "loss": 0.0183,
      "step": 12395
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.5767427086830139,
      "learning_rate": 1.2192290503410384e-05,
      "loss": 0.0208,
      "step": 12400
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.7734942436218262,
      "learning_rate": 1.2186978664897899e-05,
      "loss": 0.0291,
      "step": 12405
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.1354448795318604,
      "learning_rate": 1.2181666178240319e-05,
      "loss": 0.0275,
      "step": 12410
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.5956282019615173,
      "learning_rate": 1.2176353045012089e-05,
      "loss": 0.0251,
      "step": 12415
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.8632811307907104,
      "learning_rate": 1.217103926678784e-05,
      "loss": 0.0277,
      "step": 12420
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7880505323410034,
      "learning_rate": 1.2165724845142388e-05,
      "loss": 0.0241,
      "step": 12425
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.6448076367378235,
      "learning_rate": 1.2160409781650748e-05,
      "loss": 0.023,
      "step": 12430
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.4887710511684418,
      "learning_rate": 1.2155094077888117e-05,
      "loss": 0.0216,
      "step": 12435
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7783735394477844,
      "learning_rate": 1.2149777735429896e-05,
      "loss": 0.0285,
      "step": 12440
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.0767704248428345,
      "learning_rate": 1.2144460755851652e-05,
      "loss": 0.0262,
      "step": 12445
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.1533854007720947,
      "learning_rate": 1.2139143140729166e-05,
      "loss": 0.0224,
      "step": 12450
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.41049784421920776,
      "learning_rate": 1.2133824891638391e-05,
      "loss": 0.025,
      "step": 12455
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.6735257506370544,
      "learning_rate": 1.2128506010155472e-05,
      "loss": 0.0275,
      "step": 12460
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.9369131326675415,
      "learning_rate": 1.2123186497856743e-05,
      "loss": 0.0196,
      "step": 12465
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.5147635340690613,
      "learning_rate": 1.2117866356318724e-05,
      "loss": 0.0249,
      "step": 12470
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.5361283421516418,
      "learning_rate": 1.2112545587118124e-05,
      "loss": 0.0247,
      "step": 12475
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.621346652507782,
      "learning_rate": 1.210722419183183e-05,
      "loss": 0.0259,
      "step": 12480
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.4300675094127655,
      "learning_rate": 1.2101902172036927e-05,
      "loss": 0.0241,
      "step": 12485
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.5782265663146973,
      "learning_rate": 1.2096579529310673e-05,
      "loss": 0.0226,
      "step": 12490
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.7544804215431213,
      "learning_rate": 1.2091256265230517e-05,
      "loss": 0.0299,
      "step": 12495
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.9750189185142517,
      "learning_rate": 1.2085932381374094e-05,
      "loss": 0.0206,
      "step": 12500
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.5985661745071411,
      "learning_rate": 1.2080607879319217e-05,
      "loss": 0.0394,
      "step": 12505
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.9388284683227539,
      "learning_rate": 1.2075282760643888e-05,
      "loss": 0.0183,
      "step": 12510
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.8321487903594971,
      "learning_rate": 1.2069957026926278e-05,
      "loss": 0.0328,
      "step": 12515
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.9432512521743774,
      "learning_rate": 1.2064630679744762e-05,
      "loss": 0.0238,
      "step": 12520
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.7519678473472595,
      "learning_rate": 1.2059303720677885e-05,
      "loss": 0.0197,
      "step": 12525
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.7325347065925598,
      "learning_rate": 1.2053976151304372e-05,
      "loss": 0.0239,
      "step": 12530
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.36889663338661194,
      "learning_rate": 1.2048647973203125e-05,
      "loss": 0.023,
      "step": 12535
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.5902948379516602,
      "learning_rate": 1.2043319187953242e-05,
      "loss": 0.0196,
      "step": 12540
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.5314132571220398,
      "learning_rate": 1.2037989797133986e-05,
      "loss": 0.0241,
      "step": 12545
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.5154591202735901,
      "learning_rate": 1.2032659802324807e-05,
      "loss": 0.0218,
      "step": 12550
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.6044285893440247,
      "learning_rate": 1.2027329205105325e-05,
      "loss": 0.0205,
      "step": 12555
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.6315939426422119,
      "learning_rate": 1.2021998007055353e-05,
      "loss": 0.025,
      "step": 12560
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.7652075886726379,
      "learning_rate": 1.2016666209754877e-05,
      "loss": 0.0241,
      "step": 12565
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.668079674243927,
      "learning_rate": 1.2011333814784048e-05,
      "loss": 0.03,
      "step": 12570
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.8443313241004944,
      "learning_rate": 1.200600082372321e-05,
      "loss": 0.0348,
      "step": 12575
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.4662708640098572,
      "learning_rate": 1.2000667238152874e-05,
      "loss": 0.0171,
      "step": 12580
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.922974705696106,
      "learning_rate": 1.1995333059653739e-05,
      "loss": 0.0302,
      "step": 12585
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.6022462248802185,
      "learning_rate": 1.1989998289806663e-05,
      "loss": 0.0268,
      "step": 12590
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.7470941543579102,
      "learning_rate": 1.1984662930192692e-05,
      "loss": 0.0258,
      "step": 12595
    },
    {
      "epoch": 3.93,
      "grad_norm": 1.0102777481079102,
      "learning_rate": 1.197932698239304e-05,
      "loss": 0.032,
      "step": 12600
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.827872097492218,
      "learning_rate": 1.19739904479891e-05,
      "loss": 0.033,
      "step": 12605
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.8313772082328796,
      "learning_rate": 1.1968653328562438e-05,
      "loss": 0.0294,
      "step": 12610
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.434220552444458,
      "learning_rate": 1.1963315625694792e-05,
      "loss": 0.0183,
      "step": 12615
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.640626847743988,
      "learning_rate": 1.195797734096807e-05,
      "loss": 0.0281,
      "step": 12620
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.6942445039749146,
      "learning_rate": 1.1952638475964358e-05,
      "loss": 0.036,
      "step": 12625
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.4551185369491577,
      "learning_rate": 1.1947299032265911e-05,
      "loss": 0.0214,
      "step": 12630
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.7700601816177368,
      "learning_rate": 1.1941959011455158e-05,
      "loss": 0.0243,
      "step": 12635
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.6045627593994141,
      "learning_rate": 1.193661841511469e-05,
      "loss": 0.0211,
      "step": 12640
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.8159734606742859,
      "learning_rate": 1.1931277244827287e-05,
      "loss": 0.0223,
      "step": 12645
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.6006589531898499,
      "learning_rate": 1.1925935502175878e-05,
      "loss": 0.0209,
      "step": 12650
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.5091835260391235,
      "learning_rate": 1.1920593188743579e-05,
      "loss": 0.0276,
      "step": 12655
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.5942313075065613,
      "learning_rate": 1.1915250306113657e-05,
      "loss": 0.0269,
      "step": 12660
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.7808033227920532,
      "learning_rate": 1.1909906855869568e-05,
      "loss": 0.0321,
      "step": 12665
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.2433404922485352,
      "learning_rate": 1.1904562839594924e-05,
      "loss": 0.0388,
      "step": 12670
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.7753468751907349,
      "learning_rate": 1.1899218258873508e-05,
      "loss": 0.0189,
      "step": 12675
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.551559329032898,
      "learning_rate": 1.1893873115289266e-05,
      "loss": 0.0162,
      "step": 12680
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.5967293977737427,
      "learning_rate": 1.1888527410426317e-05,
      "loss": 0.0273,
      "step": 12685
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.51964271068573,
      "learning_rate": 1.1883181145868943e-05,
      "loss": 0.0327,
      "step": 12690
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.27552059292793274,
      "learning_rate": 1.187783432320159e-05,
      "loss": 0.0268,
      "step": 12695
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.7243056297302246,
      "learning_rate": 1.1872486944008876e-05,
      "loss": 0.025,
      "step": 12700
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.8522168397903442,
      "learning_rate": 1.186713900987558e-05,
      "loss": 0.0219,
      "step": 12705
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.6728376746177673,
      "learning_rate": 1.1861790522386643e-05,
      "loss": 0.025,
      "step": 12710
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.602962851524353,
      "learning_rate": 1.1856441483127172e-05,
      "loss": 0.0228,
      "step": 12715
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.5500351786613464,
      "learning_rate": 1.1851091893682438e-05,
      "loss": 0.028,
      "step": 12720
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.6493998765945435,
      "learning_rate": 1.1845741755637877e-05,
      "loss": 0.0242,
      "step": 12725
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.914799153804779,
      "learning_rate": 1.1840391070579083e-05,
      "loss": 0.0203,
      "step": 12730
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.7835777401924133,
      "learning_rate": 1.1835039840091818e-05,
      "loss": 0.0227,
      "step": 12735
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.7413834929466248,
      "learning_rate": 1.1829688065761997e-05,
      "loss": 0.031,
      "step": 12740
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.7562592625617981,
      "learning_rate": 1.1824335749175706e-05,
      "loss": 0.0322,
      "step": 12745
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.4942041039466858,
      "learning_rate": 1.1818982891919184e-05,
      "loss": 0.0238,
      "step": 12750
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.5303024649620056,
      "learning_rate": 1.1813629495578838e-05,
      "loss": 0.0252,
      "step": 12755
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.6929956078529358,
      "learning_rate": 1.1808275561741224e-05,
      "loss": 0.0239,
      "step": 12760
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.7698571085929871,
      "learning_rate": 1.180292109199307e-05,
      "loss": 0.0299,
      "step": 12765
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.7231923937797546,
      "learning_rate": 1.179756608792125e-05,
      "loss": 0.0234,
      "step": 12770
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.5542680621147156,
      "learning_rate": 1.1792210551112806e-05,
      "loss": 0.0201,
      "step": 12775
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.4305645823478699,
      "learning_rate": 1.1786854483154935e-05,
      "loss": 0.0255,
      "step": 12780
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.5332358479499817,
      "learning_rate": 1.1781497885634987e-05,
      "loss": 0.019,
      "step": 12785
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.7843573093414307,
      "learning_rate": 1.1776140760140477e-05,
      "loss": 0.0304,
      "step": 12790
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.668351411819458,
      "learning_rate": 1.1770783108259072e-05,
      "loss": 0.0269,
      "step": 12795
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.7774969339370728,
      "learning_rate": 1.1765424931578597e-05,
      "loss": 0.0209,
      "step": 12800
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.6049669981002808,
      "learning_rate": 1.1760066231687021e-05,
      "loss": 0.0188,
      "step": 12805
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4718351662158966,
      "learning_rate": 1.175470701017249e-05,
      "loss": 0.0156,
      "step": 12810
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7984088063240051,
      "learning_rate": 1.1749347268623289e-05,
      "loss": 0.0323,
      "step": 12815
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7956653237342834,
      "learning_rate": 1.1743987008627858e-05,
      "loss": 0.0322,
      "step": 12820
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5670791268348694,
      "learning_rate": 1.1738626231774792e-05,
      "loss": 0.0219,
      "step": 12825
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.44218578934669495,
      "learning_rate": 1.1733264939652844e-05,
      "loss": 0.0141,
      "step": 12830
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4226168692111969,
      "learning_rate": 1.172790313385092e-05,
      "loss": 0.0118,
      "step": 12835
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.47754889726638794,
      "learning_rate": 1.1722540815958066e-05,
      "loss": 0.0159,
      "step": 12840
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.3430834710597992,
      "learning_rate": 1.1717177987563488e-05,
      "loss": 0.0118,
      "step": 12845
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.49128466844558716,
      "learning_rate": 1.171181465025655e-05,
      "loss": 0.0178,
      "step": 12850
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.5054926872253418,
      "learning_rate": 1.1706450805626762e-05,
      "loss": 0.0131,
      "step": 12855
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.4532546401023865,
      "learning_rate": 1.1701086455263776e-05,
      "loss": 0.0143,
      "step": 12860
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.4453551173210144,
      "learning_rate": 1.16957216007574e-05,
      "loss": 0.0139,
      "step": 12865
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.7018569707870483,
      "learning_rate": 1.1690356243697597e-05,
      "loss": 0.0168,
      "step": 12870
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.4247054159641266,
      "learning_rate": 1.1684990385674469e-05,
      "loss": 0.0123,
      "step": 12875
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.3397986590862274,
      "learning_rate": 1.1679624028278276e-05,
      "loss": 0.0136,
      "step": 12880
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.2713475227355957,
      "learning_rate": 1.167425717309942e-05,
      "loss": 0.0184,
      "step": 12885
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.6043540239334106,
      "learning_rate": 1.166888982172845e-05,
      "loss": 0.0131,
      "step": 12890
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.438080757856369,
      "learning_rate": 1.1663521975756065e-05,
      "loss": 0.0135,
      "step": 12895
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.40145036578178406,
      "learning_rate": 1.1658153636773108e-05,
      "loss": 0.0142,
      "step": 12900
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.7116342782974243,
      "learning_rate": 1.1652784806370576e-05,
      "loss": 0.0129,
      "step": 12905
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.5809972286224365,
      "learning_rate": 1.1647415486139597e-05,
      "loss": 0.022,
      "step": 12910
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.7480118870735168,
      "learning_rate": 1.1642045677671457e-05,
      "loss": 0.0122,
      "step": 12915
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.6057562828063965,
      "learning_rate": 1.163667538255758e-05,
      "loss": 0.0132,
      "step": 12920
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.5527793169021606,
      "learning_rate": 1.163130460238954e-05,
      "loss": 0.0098,
      "step": 12925
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.6269908547401428,
      "learning_rate": 1.1625933338759045e-05,
      "loss": 0.013,
      "step": 12930
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.8735302090644836,
      "learning_rate": 1.1620561593257958e-05,
      "loss": 0.0156,
      "step": 12935
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.58750981092453,
      "learning_rate": 1.1615189367478275e-05,
      "loss": 0.0126,
      "step": 12940
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.6866432428359985,
      "learning_rate": 1.1609816663012145e-05,
      "loss": 0.0158,
      "step": 12945
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.6975976228713989,
      "learning_rate": 1.1604443481451843e-05,
      "loss": 0.0131,
      "step": 12950
    },
    {
      "epoch": 4.04,
      "grad_norm": 3.5784661769866943,
      "learning_rate": 1.1599069824389804e-05,
      "loss": 0.0129,
      "step": 12955
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.5637338757514954,
      "learning_rate": 1.159369569341859e-05,
      "loss": 0.0125,
      "step": 12960
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.9260528683662415,
      "learning_rate": 1.1588321090130907e-05,
      "loss": 0.0169,
      "step": 12965
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.36479848623275757,
      "learning_rate": 1.1582946016119602e-05,
      "loss": 0.0127,
      "step": 12970
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.7243846654891968,
      "learning_rate": 1.1577570472977665e-05,
      "loss": 0.0159,
      "step": 12975
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.4703471064567566,
      "learning_rate": 1.1572194462298225e-05,
      "loss": 0.011,
      "step": 12980
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.31925779581069946,
      "learning_rate": 1.1566817985674537e-05,
      "loss": 0.0099,
      "step": 12985
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.4469374418258667,
      "learning_rate": 1.1561441044700009e-05,
      "loss": 0.0135,
      "step": 12990
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.6372437477111816,
      "learning_rate": 1.1556063640968176e-05,
      "loss": 0.0186,
      "step": 12995
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.533165693283081,
      "learning_rate": 1.1550685776072723e-05,
      "loss": 0.0098,
      "step": 13000
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.24450081586837769,
      "learning_rate": 1.154530745160746e-05,
      "loss": 0.0091,
      "step": 13005
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.4475289583206177,
      "learning_rate": 1.1539928669166334e-05,
      "loss": 0.015,
      "step": 13010
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.44072267413139343,
      "learning_rate": 1.1534549430343436e-05,
      "loss": 0.0106,
      "step": 13015
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.7471389770507812,
      "learning_rate": 1.152916973673298e-05,
      "loss": 0.0176,
      "step": 13020
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.45366644859313965,
      "learning_rate": 1.1523789589929331e-05,
      "loss": 0.0118,
      "step": 13025
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.5312186479568481,
      "learning_rate": 1.151840899152697e-05,
      "loss": 0.0132,
      "step": 13030
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.39465898275375366,
      "learning_rate": 1.1513027943120528e-05,
      "loss": 0.0103,
      "step": 13035
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.4964543581008911,
      "learning_rate": 1.1507646446304759e-05,
      "loss": 0.0183,
      "step": 13040
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.36636462807655334,
      "learning_rate": 1.1502264502674552e-05,
      "loss": 0.0137,
      "step": 13045
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.5650526881217957,
      "learning_rate": 1.149688211382493e-05,
      "loss": 0.0116,
      "step": 13050
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.7195597290992737,
      "learning_rate": 1.149149928135105e-05,
      "loss": 0.0153,
      "step": 13055
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.6196027398109436,
      "learning_rate": 1.1486116006848196e-05,
      "loss": 0.0142,
      "step": 13060
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.45367714762687683,
      "learning_rate": 1.1480732291911786e-05,
      "loss": 0.012,
      "step": 13065
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.5096328854560852,
      "learning_rate": 1.1475348138137368e-05,
      "loss": 0.0142,
      "step": 13070
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.3048243522644043,
      "learning_rate": 1.1469963547120614e-05,
      "loss": 0.0138,
      "step": 13075
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.5990984439849854,
      "learning_rate": 1.1464578520457341e-05,
      "loss": 0.0145,
      "step": 13080
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.5580481290817261,
      "learning_rate": 1.1459193059743476e-05,
      "loss": 0.0139,
      "step": 13085
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.6747692823410034,
      "learning_rate": 1.1453807166575093e-05,
      "loss": 0.0134,
      "step": 13090
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.6296996474266052,
      "learning_rate": 1.1448420842548375e-05,
      "loss": 0.0131,
      "step": 13095
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.5080475211143494,
      "learning_rate": 1.1443034089259651e-05,
      "loss": 0.0142,
      "step": 13100
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.6452639698982239,
      "learning_rate": 1.1437646908305364e-05,
      "loss": 0.0197,
      "step": 13105
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.7623153328895569,
      "learning_rate": 1.1432259301282091e-05,
      "loss": 0.0137,
      "step": 13110
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.539873480796814,
      "learning_rate": 1.142687126978653e-05,
      "loss": 0.018,
      "step": 13115
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.4487948417663574,
      "learning_rate": 1.1421482815415513e-05,
      "loss": 0.0114,
      "step": 13120
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.3985424041748047,
      "learning_rate": 1.141609393976599e-05,
      "loss": 0.0127,
      "step": 13125
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.5115072131156921,
      "learning_rate": 1.1410704644435038e-05,
      "loss": 0.0096,
      "step": 13130
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.7904471158981323,
      "learning_rate": 1.1405314931019854e-05,
      "loss": 0.0156,
      "step": 13135
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.512873649597168,
      "learning_rate": 1.139992480111777e-05,
      "loss": 0.0128,
      "step": 13140
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.7409130930900574,
      "learning_rate": 1.1394534256326231e-05,
      "loss": 0.0126,
      "step": 13145
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.7204809188842773,
      "learning_rate": 1.1389143298242811e-05,
      "loss": 0.0148,
      "step": 13150
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.37489232420921326,
      "learning_rate": 1.13837519284652e-05,
      "loss": 0.0104,
      "step": 13155
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.38769617676734924,
      "learning_rate": 1.1378360148591217e-05,
      "loss": 0.0151,
      "step": 13160
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.6176652312278748,
      "learning_rate": 1.13729679602188e-05,
      "loss": 0.0134,
      "step": 13165
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.4696696400642395,
      "learning_rate": 1.1367575364946006e-05,
      "loss": 0.0129,
      "step": 13170
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.39097660779953003,
      "learning_rate": 1.1362182364371018e-05,
      "loss": 0.0196,
      "step": 13175
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.7201048135757446,
      "learning_rate": 1.1356788960092132e-05,
      "loss": 0.0129,
      "step": 13180
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.581547737121582,
      "learning_rate": 1.1351395153707765e-05,
      "loss": 0.0151,
      "step": 13185
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.35174039006233215,
      "learning_rate": 1.1346000946816462e-05,
      "loss": 0.0108,
      "step": 13190
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.48897290229797363,
      "learning_rate": 1.1340606341016875e-05,
      "loss": 0.018,
      "step": 13195
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.6607118844985962,
      "learning_rate": 1.1335211337907782e-05,
      "loss": 0.014,
      "step": 13200
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.3014785051345825,
      "learning_rate": 1.1329815939088072e-05,
      "loss": 0.0099,
      "step": 13205
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.34429848194122314,
      "learning_rate": 1.1324420146156757e-05,
      "loss": 0.0102,
      "step": 13210
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.3376854956150055,
      "learning_rate": 1.1319023960712972e-05,
      "loss": 0.0165,
      "step": 13215
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.4123086631298065,
      "learning_rate": 1.1313627384355943e-05,
      "loss": 0.0094,
      "step": 13220
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.47521141171455383,
      "learning_rate": 1.1308230418685045e-05,
      "loss": 0.0166,
      "step": 13225
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.629194438457489,
      "learning_rate": 1.1302833065299745e-05,
      "loss": 0.0131,
      "step": 13230
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.9634842872619629,
      "learning_rate": 1.129743532579964e-05,
      "loss": 0.0122,
      "step": 13235
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.7814756631851196,
      "learning_rate": 1.129203720178442e-05,
      "loss": 0.019,
      "step": 13240
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.4899265170097351,
      "learning_rate": 1.1286638694853916e-05,
      "loss": 0.0153,
      "step": 13245
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.5780858993530273,
      "learning_rate": 1.1281239806608056e-05,
      "loss": 0.0157,
      "step": 13250
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.7560383081436157,
      "learning_rate": 1.1275840538646878e-05,
      "loss": 0.021,
      "step": 13255
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.6205662488937378,
      "learning_rate": 1.1270440892570544e-05,
      "loss": 0.0188,
      "step": 13260
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.3816826641559601,
      "learning_rate": 1.1265040869979325e-05,
      "loss": 0.0187,
      "step": 13265
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.6150768995285034,
      "learning_rate": 1.1259640472473602e-05,
      "loss": 0.0193,
      "step": 13270
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.7604739665985107,
      "learning_rate": 1.1254239701653862e-05,
      "loss": 0.0159,
      "step": 13275
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.6447388529777527,
      "learning_rate": 1.1248838559120708e-05,
      "loss": 0.0128,
      "step": 13280
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.6438114643096924,
      "learning_rate": 1.1243437046474854e-05,
      "loss": 0.0149,
      "step": 13285
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.7647647857666016,
      "learning_rate": 1.1238035165317123e-05,
      "loss": 0.016,
      "step": 13290
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.5507582426071167,
      "learning_rate": 1.1232632917248442e-05,
      "loss": 0.0195,
      "step": 13295
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.5412368178367615,
      "learning_rate": 1.1227230303869858e-05,
      "loss": 0.016,
      "step": 13300
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.6669492125511169,
      "learning_rate": 1.1221827326782515e-05,
      "loss": 0.0126,
      "step": 13305
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.48139962553977966,
      "learning_rate": 1.121642398758767e-05,
      "loss": 0.0124,
      "step": 13310
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.6500893235206604,
      "learning_rate": 1.1211020287886685e-05,
      "loss": 0.0135,
      "step": 13315
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.6527283787727356,
      "learning_rate": 1.1205616229281033e-05,
      "loss": 0.0133,
      "step": 13320
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.4571009576320648,
      "learning_rate": 1.1200211813372288e-05,
      "loss": 0.0107,
      "step": 13325
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.3937748372554779,
      "learning_rate": 1.1194807041762137e-05,
      "loss": 0.0148,
      "step": 13330
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.4135548770427704,
      "learning_rate": 1.1189401916052362e-05,
      "loss": 0.0205,
      "step": 13335
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.6549053192138672,
      "learning_rate": 1.118399643784486e-05,
      "loss": 0.0181,
      "step": 13340
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.7284772396087646,
      "learning_rate": 1.1178590608741628e-05,
      "loss": 0.0113,
      "step": 13345
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.8119202256202698,
      "learning_rate": 1.1173184430344765e-05,
      "loss": 0.0144,
      "step": 13350
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.8524966239929199,
      "learning_rate": 1.1167777904256481e-05,
      "loss": 0.0138,
      "step": 13355
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.6740317940711975,
      "learning_rate": 1.116237103207908e-05,
      "loss": 0.0165,
      "step": 13360
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.43391916155815125,
      "learning_rate": 1.115696381541497e-05,
      "loss": 0.0118,
      "step": 13365
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.8504462838172913,
      "learning_rate": 1.1151556255866668e-05,
      "loss": 0.0148,
      "step": 13370
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.6841346025466919,
      "learning_rate": 1.1146148355036788e-05,
      "loss": 0.0118,
      "step": 13375
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.5312472581863403,
      "learning_rate": 1.1140740114528044e-05,
      "loss": 0.0162,
      "step": 13380
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.7032490372657776,
      "learning_rate": 1.113533153594325e-05,
      "loss": 0.0128,
      "step": 13385
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.5678761005401611,
      "learning_rate": 1.1129922620885327e-05,
      "loss": 0.0132,
      "step": 13390
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.5700942873954773,
      "learning_rate": 1.112451337095729e-05,
      "loss": 0.0122,
      "step": 13395
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.6323286294937134,
      "learning_rate": 1.1119103787762254e-05,
      "loss": 0.0158,
      "step": 13400
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.38179242610931396,
      "learning_rate": 1.1113693872903426e-05,
      "loss": 0.0135,
      "step": 13405
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.8654710054397583,
      "learning_rate": 1.1108283627984129e-05,
      "loss": 0.0148,
      "step": 13410
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.5700730681419373,
      "learning_rate": 1.1102873054607768e-05,
      "loss": 0.0167,
      "step": 13415
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.5876476168632507,
      "learning_rate": 1.109746215437785e-05,
      "loss": 0.0163,
      "step": 13420
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.3003236651420593,
      "learning_rate": 1.109205092889798e-05,
      "loss": 0.0151,
      "step": 13425
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.5799517035484314,
      "learning_rate": 1.1086639379771857e-05,
      "loss": 0.0159,
      "step": 13430
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.8325760364532471,
      "learning_rate": 1.1081227508603283e-05,
      "loss": 0.0162,
      "step": 13435
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.4690932631492615,
      "learning_rate": 1.1075815316996144e-05,
      "loss": 0.0145,
      "step": 13440
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.8196485638618469,
      "learning_rate": 1.107040280655443e-05,
      "loss": 0.0141,
      "step": 13445
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.7324041724205017,
      "learning_rate": 1.1064989978882221e-05,
      "loss": 0.0168,
      "step": 13450
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.47851309180259705,
      "learning_rate": 1.1059576835583694e-05,
      "loss": 0.0168,
      "step": 13455
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.32001355290412903,
      "learning_rate": 1.1054163378263119e-05,
      "loss": 0.0119,
      "step": 13460
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.42558789253234863,
      "learning_rate": 1.1048749608524856e-05,
      "loss": 0.0116,
      "step": 13465
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.6945168375968933,
      "learning_rate": 1.104333552797336e-05,
      "loss": 0.0147,
      "step": 13470
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.47573333978652954,
      "learning_rate": 1.1037921138213177e-05,
      "loss": 0.0148,
      "step": 13475
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.3896462917327881,
      "learning_rate": 1.103250644084895e-05,
      "loss": 0.0114,
      "step": 13480
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.5291655659675598,
      "learning_rate": 1.1027091437485404e-05,
      "loss": 0.0127,
      "step": 13485
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.4930930733680725,
      "learning_rate": 1.1021676129727359e-05,
      "loss": 0.0135,
      "step": 13490
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.6311109066009521,
      "learning_rate": 1.101626051917973e-05,
      "loss": 0.0177,
      "step": 13495
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.5875706076622009,
      "learning_rate": 1.101084460744751e-05,
      "loss": 0.0112,
      "step": 13500
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.8075762391090393,
      "learning_rate": 1.10054283961358e-05,
      "loss": 0.0144,
      "step": 13505
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.7476543188095093,
      "learning_rate": 1.1000011886849767e-05,
      "loss": 0.0133,
      "step": 13510
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.5966615676879883,
      "learning_rate": 1.099459508119468e-05,
      "loss": 0.0139,
      "step": 13515
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.6924692392349243,
      "learning_rate": 1.0989177980775902e-05,
      "loss": 0.0135,
      "step": 13520
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.5964117050170898,
      "learning_rate": 1.0983760587198866e-05,
      "loss": 0.0137,
      "step": 13525
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.5716801285743713,
      "learning_rate": 1.0978342902069099e-05,
      "loss": 0.0155,
      "step": 13530
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.6075406074523926,
      "learning_rate": 1.0972924926992224e-05,
      "loss": 0.0159,
      "step": 13535
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.31922054290771484,
      "learning_rate": 1.0967506663573942e-05,
      "loss": 0.0162,
      "step": 13540
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.548858642578125,
      "learning_rate": 1.0962088113420031e-05,
      "loss": 0.0173,
      "step": 13545
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.6774040460586548,
      "learning_rate": 1.0956669278136367e-05,
      "loss": 0.0124,
      "step": 13550
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.6590894460678101,
      "learning_rate": 1.0951250159328908e-05,
      "loss": 0.0136,
      "step": 13555
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.5625125169754028,
      "learning_rate": 1.094583075860369e-05,
      "loss": 0.016,
      "step": 13560
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.8643558621406555,
      "learning_rate": 1.0940411077566842e-05,
      "loss": 0.013,
      "step": 13565
    },
    {
      "epoch": 4.23,
      "grad_norm": 1.2786364555358887,
      "learning_rate": 1.0934991117824565e-05,
      "loss": 0.0156,
      "step": 13570
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.21733850240707397,
      "learning_rate": 1.0929570880983152e-05,
      "loss": 0.0128,
      "step": 13575
    },
    {
      "epoch": 4.24,
      "grad_norm": 1.16891348361969,
      "learning_rate": 1.0924150368648974e-05,
      "loss": 0.0142,
      "step": 13580
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.6256799697875977,
      "learning_rate": 1.091872958242848e-05,
      "loss": 0.0112,
      "step": 13585
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.5066383481025696,
      "learning_rate": 1.091330852392821e-05,
      "loss": 0.0123,
      "step": 13590
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.4950759708881378,
      "learning_rate": 1.0907887194754774e-05,
      "loss": 0.0151,
      "step": 13595
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.5494158267974854,
      "learning_rate": 1.0902465596514871e-05,
      "loss": 0.0103,
      "step": 13600
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.5297831296920776,
      "learning_rate": 1.0897043730815273e-05,
      "loss": 0.0098,
      "step": 13605
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.5016347169876099,
      "learning_rate": 1.0891621599262835e-05,
      "loss": 0.0134,
      "step": 13610
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.45652124285697937,
      "learning_rate": 1.0886199203464492e-05,
      "loss": 0.0138,
      "step": 13615
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.4522571265697479,
      "learning_rate": 1.0880776545027252e-05,
      "loss": 0.0125,
      "step": 13620
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.388315349817276,
      "learning_rate": 1.0875353625558205e-05,
      "loss": 0.0113,
      "step": 13625
    },
    {
      "epoch": 4.25,
      "grad_norm": 1.0517215728759766,
      "learning_rate": 1.086993044666452e-05,
      "loss": 0.0172,
      "step": 13630
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.38936805725097656,
      "learning_rate": 1.0864507009953435e-05,
      "loss": 0.0146,
      "step": 13635
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.6670991778373718,
      "learning_rate": 1.0859083317032273e-05,
      "loss": 0.0147,
      "step": 13640
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.5981236696243286,
      "learning_rate": 1.0853659369508432e-05,
      "loss": 0.0192,
      "step": 13645
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.7371012568473816,
      "learning_rate": 1.0848235168989378e-05,
      "loss": 0.0134,
      "step": 13650
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.5573756098747253,
      "learning_rate": 1.084281071708266e-05,
      "loss": 0.0146,
      "step": 13655
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.6925496459007263,
      "learning_rate": 1.0837386015395897e-05,
      "loss": 0.0106,
      "step": 13660
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.5630655288696289,
      "learning_rate": 1.0831961065536786e-05,
      "loss": 0.0098,
      "step": 13665
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.5372442007064819,
      "learning_rate": 1.0826535869113093e-05,
      "loss": 0.0172,
      "step": 13670
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.4406470060348511,
      "learning_rate": 1.0821110427732658e-05,
      "loss": 0.0138,
      "step": 13675
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.8311110138893127,
      "learning_rate": 1.0815684743003398e-05,
      "loss": 0.0141,
      "step": 13680
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.5476308465003967,
      "learning_rate": 1.08102588165333e-05,
      "loss": 0.0139,
      "step": 13685
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.3883194327354431,
      "learning_rate": 1.0804832649930416e-05,
      "loss": 0.0092,
      "step": 13690
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.8831512928009033,
      "learning_rate": 1.0799406244802881e-05,
      "loss": 0.0198,
      "step": 13695
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.6092522740364075,
      "learning_rate": 1.0793979602758889e-05,
      "loss": 0.016,
      "step": 13700
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.6196035146713257,
      "learning_rate": 1.0788552725406713e-05,
      "loss": 0.0121,
      "step": 13705
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.5230036973953247,
      "learning_rate": 1.0783125614354691e-05,
      "loss": 0.0124,
      "step": 13710
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.5164169669151306,
      "learning_rate": 1.0777698271211232e-05,
      "loss": 0.0107,
      "step": 13715
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.43008527159690857,
      "learning_rate": 1.0772270697584812e-05,
      "loss": 0.0122,
      "step": 13720
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.4724281132221222,
      "learning_rate": 1.076684289508398e-05,
      "loss": 0.0137,
      "step": 13725
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.6964073777198792,
      "learning_rate": 1.0761414865317348e-05,
      "loss": 0.0128,
      "step": 13730
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.47468698024749756,
      "learning_rate": 1.0755986609893597e-05,
      "loss": 0.0122,
      "step": 13735
    },
    {
      "epoch": 4.29,
      "grad_norm": 1.2172765731811523,
      "learning_rate": 1.0750558130421473e-05,
      "loss": 0.0129,
      "step": 13740
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.6301424503326416,
      "learning_rate": 1.0745129428509791e-05,
      "loss": 0.015,
      "step": 13745
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.38925108313560486,
      "learning_rate": 1.0739700505767433e-05,
      "loss": 0.0127,
      "step": 13750
    },
    {
      "epoch": 4.29,
      "grad_norm": 1.6824076175689697,
      "learning_rate": 1.0734271363803343e-05,
      "loss": 0.0157,
      "step": 13755
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.594070315361023,
      "learning_rate": 1.0728842004226531e-05,
      "loss": 0.0157,
      "step": 13760
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.8799555897712708,
      "learning_rate": 1.0723412428646074e-05,
      "loss": 0.0179,
      "step": 13765
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.5111485719680786,
      "learning_rate": 1.071798263867111e-05,
      "loss": 0.0168,
      "step": 13770
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.5702099800109863,
      "learning_rate": 1.0712552635910844e-05,
      "loss": 0.0144,
      "step": 13775
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.32097887992858887,
      "learning_rate": 1.0707122421974535e-05,
      "loss": 0.0115,
      "step": 13780
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.6477245092391968,
      "learning_rate": 1.0701691998471515e-05,
      "loss": 0.0166,
      "step": 13785
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.536425769329071,
      "learning_rate": 1.0696261367011178e-05,
      "loss": 0.016,
      "step": 13790
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.668903648853302,
      "learning_rate": 1.0690830529202971e-05,
      "loss": 0.0128,
      "step": 13795
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.4450022578239441,
      "learning_rate": 1.0685399486656407e-05,
      "loss": 0.0107,
      "step": 13800
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.8847901225090027,
      "learning_rate": 1.0679968240981062e-05,
      "loss": 0.0127,
      "step": 13805
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.5488216280937195,
      "learning_rate": 1.0674536793786572e-05,
      "loss": 0.0153,
      "step": 13810
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.4895569086074829,
      "learning_rate": 1.0669105146682628e-05,
      "loss": 0.0131,
      "step": 13815
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.5281688570976257,
      "learning_rate": 1.0663673301278981e-05,
      "loss": 0.0117,
      "step": 13820
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.5016342401504517,
      "learning_rate": 1.0658241259185444e-05,
      "loss": 0.0091,
      "step": 13825
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.6420480608940125,
      "learning_rate": 1.0652809022011891e-05,
      "loss": 0.0165,
      "step": 13830
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.4913181960582733,
      "learning_rate": 1.0647376591368248e-05,
      "loss": 0.014,
      "step": 13835
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.5184372663497925,
      "learning_rate": 1.0641943968864496e-05,
      "loss": 0.0116,
      "step": 13840
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.8069402575492859,
      "learning_rate": 1.063651115611068e-05,
      "loss": 0.0176,
      "step": 13845
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.593238890171051,
      "learning_rate": 1.06310781547169e-05,
      "loss": 0.0178,
      "step": 13850
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.6287248730659485,
      "learning_rate": 1.062564496629331e-05,
      "loss": 0.015,
      "step": 13855
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.38136905431747437,
      "learning_rate": 1.0620211592450116e-05,
      "loss": 0.0136,
      "step": 13860
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.45957428216934204,
      "learning_rate": 1.0614778034797586e-05,
      "loss": 0.0154,
      "step": 13865
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.38002118468284607,
      "learning_rate": 1.060934429494604e-05,
      "loss": 0.0133,
      "step": 13870
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.5099014639854431,
      "learning_rate": 1.0603910374505847e-05,
      "loss": 0.014,
      "step": 13875
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.5811852812767029,
      "learning_rate": 1.0598476275087437e-05,
      "loss": 0.0161,
      "step": 13880
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.4693862795829773,
      "learning_rate": 1.059304199830129e-05,
      "loss": 0.0161,
      "step": 13885
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.44405293464660645,
      "learning_rate": 1.0587607545757938e-05,
      "loss": 0.0147,
      "step": 13890
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.43415597081184387,
      "learning_rate": 1.0582172919067963e-05,
      "loss": 0.015,
      "step": 13895
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.6808017492294312,
      "learning_rate": 1.0576738119842005e-05,
      "loss": 0.015,
      "step": 13900
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.978279709815979,
      "learning_rate": 1.0571303149690749e-05,
      "loss": 0.0192,
      "step": 13905
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.44453221559524536,
      "learning_rate": 1.0565868010224932e-05,
      "loss": 0.0147,
      "step": 13910
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.8853317499160767,
      "learning_rate": 1.0560432703055341e-05,
      "loss": 0.0186,
      "step": 13915
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.7187196612358093,
      "learning_rate": 1.0556084337655437e-05,
      "loss": 0.0177,
      "step": 13920
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.6485170722007751,
      "learning_rate": 1.0550648732678402e-05,
      "loss": 0.0173,
      "step": 13925
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.7718275785446167,
      "learning_rate": 1.0545212964508063e-05,
      "loss": 0.0183,
      "step": 13930
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.6682250499725342,
      "learning_rate": 1.0539777034755396e-05,
      "loss": 0.0137,
      "step": 13935
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.4407562017440796,
      "learning_rate": 1.0534340945031425e-05,
      "loss": 0.013,
      "step": 13940
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.5514277219772339,
      "learning_rate": 1.0528904696947211e-05,
      "loss": 0.0171,
      "step": 13945
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.33027181029319763,
      "learning_rate": 1.052346829211388e-05,
      "loss": 0.0109,
      "step": 13950
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.758332371711731,
      "learning_rate": 1.0518031732142592e-05,
      "loss": 0.0168,
      "step": 13955
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.29376351833343506,
      "learning_rate": 1.051259501864456e-05,
      "loss": 0.011,
      "step": 13960
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.47187724709510803,
      "learning_rate": 1.0507158153231027e-05,
      "loss": 0.0143,
      "step": 13965
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.6413419246673584,
      "learning_rate": 1.0501721137513302e-05,
      "loss": 0.019,
      "step": 13970
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.6035584211349487,
      "learning_rate": 1.049628397310273e-05,
      "loss": 0.0158,
      "step": 13975
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.5032995343208313,
      "learning_rate": 1.0490846661610695e-05,
      "loss": 0.0157,
      "step": 13980
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.43947482109069824,
      "learning_rate": 1.0485409204648624e-05,
      "loss": 0.014,
      "step": 13985
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.5266494750976562,
      "learning_rate": 1.0479971603828001e-05,
      "loss": 0.0144,
      "step": 13990
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.6087942719459534,
      "learning_rate": 1.047453386076034e-05,
      "loss": 0.0161,
      "step": 13995
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.41034409403800964,
      "learning_rate": 1.04690959770572e-05,
      "loss": 0.0155,
      "step": 14000
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.48467421531677246,
      "learning_rate": 1.0463657954330182e-05,
      "loss": 0.0137,
      "step": 14005
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.4704853296279907,
      "learning_rate": 1.0458219794190927e-05,
      "loss": 0.0141,
      "step": 14010
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.41661834716796875,
      "learning_rate": 1.0452781498251126e-05,
      "loss": 0.0159,
      "step": 14015
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.5433653593063354,
      "learning_rate": 1.0447343068122493e-05,
      "loss": 0.0133,
      "step": 14020
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.6942476034164429,
      "learning_rate": 1.0441904505416796e-05,
      "loss": 0.019,
      "step": 14025
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.5101887583732605,
      "learning_rate": 1.0436465811745834e-05,
      "loss": 0.0124,
      "step": 14030
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.659727156162262,
      "learning_rate": 1.0431026988721452e-05,
      "loss": 0.0211,
      "step": 14035
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.4992978274822235,
      "learning_rate": 1.0425588037955532e-05,
      "loss": 0.0115,
      "step": 14040
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.8023609519004822,
      "learning_rate": 1.0420148961059984e-05,
      "loss": 0.0143,
      "step": 14045
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.7880727648735046,
      "learning_rate": 1.0414709759646769e-05,
      "loss": 0.0128,
      "step": 14050
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.6856178641319275,
      "learning_rate": 1.0409270435327875e-05,
      "loss": 0.0138,
      "step": 14055
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.2927437126636505,
      "learning_rate": 1.0403830989715333e-05,
      "loss": 0.0117,
      "step": 14060
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.840043306350708,
      "learning_rate": 1.0398391424421209e-05,
      "loss": 0.0112,
      "step": 14065
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.3406813144683838,
      "learning_rate": 1.0392951741057594e-05,
      "loss": 0.0127,
      "step": 14070
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.5625513195991516,
      "learning_rate": 1.038751194123663e-05,
      "loss": 0.0171,
      "step": 14075
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.6220264434814453,
      "learning_rate": 1.0382072026570485e-05,
      "loss": 0.0177,
      "step": 14080
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.3575296998023987,
      "learning_rate": 1.0376631998671365e-05,
      "loss": 0.012,
      "step": 14085
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.6170724630355835,
      "learning_rate": 1.0371191859151497e-05,
      "loss": 0.0193,
      "step": 14090
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.5731143355369568,
      "learning_rate": 1.0365751609623162e-05,
      "loss": 0.0126,
      "step": 14095
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.6981022357940674,
      "learning_rate": 1.0360311251698656e-05,
      "loss": 0.0192,
      "step": 14100
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.34567970037460327,
      "learning_rate": 1.0354870786990318e-05,
      "loss": 0.0139,
      "step": 14105
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.5584869980812073,
      "learning_rate": 1.0349430217110508e-05,
      "loss": 0.0158,
      "step": 14110
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.4890233874320984,
      "learning_rate": 1.0343989543671629e-05,
      "loss": 0.013,
      "step": 14115
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.40981024503707886,
      "learning_rate": 1.0338548768286107e-05,
      "loss": 0.0167,
      "step": 14120
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.3282528221607208,
      "learning_rate": 1.0333107892566401e-05,
      "loss": 0.0157,
      "step": 14125
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.4963136613368988,
      "learning_rate": 1.0327666918124995e-05,
      "loss": 0.0136,
      "step": 14130
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.6778050661087036,
      "learning_rate": 1.032222584657441e-05,
      "loss": 0.0117,
      "step": 14135
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.9065635800361633,
      "learning_rate": 1.0316784679527193e-05,
      "loss": 0.0217,
      "step": 14140
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.5900851488113403,
      "learning_rate": 1.0311343418595915e-05,
      "loss": 0.0126,
      "step": 14145
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.6750664114952087,
      "learning_rate": 1.0305902065393183e-05,
      "loss": 0.0181,
      "step": 14150
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.5409610271453857,
      "learning_rate": 1.0300460621531623e-05,
      "loss": 0.0104,
      "step": 14155
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.5124744772911072,
      "learning_rate": 1.029501908862389e-05,
      "loss": 0.0098,
      "step": 14160
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.3976758122444153,
      "learning_rate": 1.028957746828267e-05,
      "loss": 0.012,
      "step": 14165
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.5759036540985107,
      "learning_rate": 1.028413576212067e-05,
      "loss": 0.0143,
      "step": 14170
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.7016305923461914,
      "learning_rate": 1.0278693971750623e-05,
      "loss": 0.0141,
      "step": 14175
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.5577017068862915,
      "learning_rate": 1.0273252098785293e-05,
      "loss": 0.0163,
      "step": 14180
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.6542133688926697,
      "learning_rate": 1.0267810144837458e-05,
      "loss": 0.0141,
      "step": 14185
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.45770275592803955,
      "learning_rate": 1.0262368111519929e-05,
      "loss": 0.0158,
      "step": 14190
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.5199791193008423,
      "learning_rate": 1.0256926000445535e-05,
      "loss": 0.0108,
      "step": 14195
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.6465505361557007,
      "learning_rate": 1.025148381322713e-05,
      "loss": 0.0103,
      "step": 14200
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.5390932559967041,
      "learning_rate": 1.0246041551477592e-05,
      "loss": 0.016,
      "step": 14205
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.5630213022232056,
      "learning_rate": 1.0240599216809823e-05,
      "loss": 0.0172,
      "step": 14210
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.7540024518966675,
      "learning_rate": 1.0235156810836735e-05,
      "loss": 0.011,
      "step": 14215
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.7141256332397461,
      "learning_rate": 1.0229714335171277e-05,
      "loss": 0.0194,
      "step": 14220
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.35217756032943726,
      "learning_rate": 1.0224271791426409e-05,
      "loss": 0.0086,
      "step": 14225
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.6676583290100098,
      "learning_rate": 1.0218829181215114e-05,
      "loss": 0.012,
      "step": 14230
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.7634176015853882,
      "learning_rate": 1.0213386506150387e-05,
      "loss": 0.0141,
      "step": 14235
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.49505695700645447,
      "learning_rate": 1.0207943767845259e-05,
      "loss": 0.0159,
      "step": 14240
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.6663855910301208,
      "learning_rate": 1.0202500967912764e-05,
      "loss": 0.0117,
      "step": 14245
    },
    {
      "epoch": 4.44,
      "grad_norm": 1.1110700368881226,
      "learning_rate": 1.019705810796597e-05,
      "loss": 0.0156,
      "step": 14250
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.7089707851409912,
      "learning_rate": 1.0191615189617938e-05,
      "loss": 0.0131,
      "step": 14255
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.6198787689208984,
      "learning_rate": 1.0186172214481769e-05,
      "loss": 0.0107,
      "step": 14260
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.7208309769630432,
      "learning_rate": 1.0180729184170577e-05,
      "loss": 0.012,
      "step": 14265
    },
    {
      "epoch": 4.45,
      "grad_norm": 1.0942414999008179,
      "learning_rate": 1.0175286100297484e-05,
      "loss": 0.0121,
      "step": 14270
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.7647571563720703,
      "learning_rate": 1.0169842964475631e-05,
      "loss": 0.0169,
      "step": 14275
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.3617548942565918,
      "learning_rate": 1.0164399778318176e-05,
      "loss": 0.0155,
      "step": 14280
    },
    {
      "epoch": 4.46,
      "grad_norm": 1.1076143980026245,
      "learning_rate": 1.01589565434383e-05,
      "loss": 0.0177,
      "step": 14285
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.7034644484519958,
      "learning_rate": 1.0153513261449182e-05,
      "loss": 0.0169,
      "step": 14290
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.7932576537132263,
      "learning_rate": 1.0148069933964024e-05,
      "loss": 0.0197,
      "step": 14295
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.36392897367477417,
      "learning_rate": 1.014262656259604e-05,
      "loss": 0.0136,
      "step": 14300
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.516571044921875,
      "learning_rate": 1.0137183148958462e-05,
      "loss": 0.0134,
      "step": 14305
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.7711630463600159,
      "learning_rate": 1.0131739694664527e-05,
      "loss": 0.0132,
      "step": 14310
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.4761156439781189,
      "learning_rate": 1.0126296201327486e-05,
      "loss": 0.0192,
      "step": 14315
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.6719192266464233,
      "learning_rate": 1.0120852670560606e-05,
      "loss": 0.0126,
      "step": 14320
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.38668563961982727,
      "learning_rate": 1.0115409103977161e-05,
      "loss": 0.0107,
      "step": 14325
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.7272295951843262,
      "learning_rate": 1.0109965503190433e-05,
      "loss": 0.0144,
      "step": 14330
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.7221820950508118,
      "learning_rate": 1.0104521869813724e-05,
      "loss": 0.0207,
      "step": 14335
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.5080924034118652,
      "learning_rate": 1.0099078205460333e-05,
      "loss": 0.0109,
      "step": 14340
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.3039231598377228,
      "learning_rate": 1.0093634511743577e-05,
      "loss": 0.0097,
      "step": 14345
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.5701034665107727,
      "learning_rate": 1.0088190790276782e-05,
      "loss": 0.0147,
      "step": 14350
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.5223418474197388,
      "learning_rate": 1.0082747042673277e-05,
      "loss": 0.0131,
      "step": 14355
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6130552291870117,
      "learning_rate": 1.0077303270546397e-05,
      "loss": 0.0182,
      "step": 14360
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.5565687417984009,
      "learning_rate": 1.0071859475509496e-05,
      "loss": 0.02,
      "step": 14365
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.32723382115364075,
      "learning_rate": 1.006641565917592e-05,
      "loss": 0.0187,
      "step": 14370
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.6693105697631836,
      "learning_rate": 1.0060971823159039e-05,
      "loss": 0.0183,
      "step": 14375
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.3853606581687927,
      "learning_rate": 1.0055527969072203e-05,
      "loss": 0.0196,
      "step": 14380
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.5426690578460693,
      "learning_rate": 1.0050084098528794e-05,
      "loss": 0.0124,
      "step": 14385
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.6572005748748779,
      "learning_rate": 1.0044640213142187e-05,
      "loss": 0.0155,
      "step": 14390
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.6691354513168335,
      "learning_rate": 1.0039196314525759e-05,
      "loss": 0.0126,
      "step": 14395
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.4074169099330902,
      "learning_rate": 1.0033752404292894e-05,
      "loss": 0.0144,
      "step": 14400
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.5884296298027039,
      "learning_rate": 1.0028308484056977e-05,
      "loss": 0.0156,
      "step": 14405
    },
    {
      "epoch": 4.49,
      "grad_norm": 1.086810827255249,
      "learning_rate": 1.0022864555431405e-05,
      "loss": 0.0124,
      "step": 14410
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.624291181564331,
      "learning_rate": 1.0017420620029566e-05,
      "loss": 0.0129,
      "step": 14415
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6476021409034729,
      "learning_rate": 1.0011976679464854e-05,
      "loss": 0.0176,
      "step": 14420
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.7050710916519165,
      "learning_rate": 1.0006532735350666e-05,
      "loss": 0.0139,
      "step": 14425
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6201428174972534,
      "learning_rate": 1.0001088789300403e-05,
      "loss": 0.0142,
      "step": 14430
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6047577857971191,
      "learning_rate": 9.995644842927465e-06,
      "loss": 0.0117,
      "step": 14435
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6464779376983643,
      "learning_rate": 9.990200897845237e-06,
      "loss": 0.0144,
      "step": 14440
    },
    {
      "epoch": 4.51,
      "grad_norm": 1.0427998304367065,
      "learning_rate": 9.98475695566713e-06,
      "loss": 0.012,
      "step": 14445
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.5257377028465271,
      "learning_rate": 9.979313018006533e-06,
      "loss": 0.0124,
      "step": 14450
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.6395034193992615,
      "learning_rate": 9.973869086476846e-06,
      "loss": 0.0113,
      "step": 14455
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.4360012114048004,
      "learning_rate": 9.96842516269146e-06,
      "loss": 0.016,
      "step": 14460
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.7435383796691895,
      "learning_rate": 9.962981248263768e-06,
      "loss": 0.0168,
      "step": 14465
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.5809354186058044,
      "learning_rate": 9.957537344807157e-06,
      "loss": 0.0106,
      "step": 14470
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.5905982851982117,
      "learning_rate": 9.952093453935017e-06,
      "loss": 0.0156,
      "step": 14475
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.5396180152893066,
      "learning_rate": 9.946649577260724e-06,
      "loss": 0.0172,
      "step": 14480
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.2679448425769806,
      "learning_rate": 9.941205716397654e-06,
      "loss": 0.0146,
      "step": 14485
    },
    {
      "epoch": 4.52,
      "grad_norm": 1.4850140810012817,
      "learning_rate": 9.935761872959185e-06,
      "loss": 0.0123,
      "step": 14490
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.47272995114326477,
      "learning_rate": 9.930318048558684e-06,
      "loss": 0.0155,
      "step": 14495
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.8253952264785767,
      "learning_rate": 9.924874244809511e-06,
      "loss": 0.0182,
      "step": 14500
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.6212565898895264,
      "learning_rate": 9.91943046332502e-06,
      "loss": 0.015,
      "step": 14505
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.7062631249427795,
      "learning_rate": 9.913986705718567e-06,
      "loss": 0.0171,
      "step": 14510
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.49691250920295715,
      "learning_rate": 9.908542973603485e-06,
      "loss": 0.0179,
      "step": 14515
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.7071046233177185,
      "learning_rate": 9.903099268593115e-06,
      "loss": 0.0212,
      "step": 14520
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.8758559823036194,
      "learning_rate": 9.897655592300779e-06,
      "loss": 0.015,
      "step": 14525
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.5860288143157959,
      "learning_rate": 9.8922119463398e-06,
      "loss": 0.0157,
      "step": 14530
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.65015709400177,
      "learning_rate": 9.886768332323483e-06,
      "loss": 0.0103,
      "step": 14535
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.5659227967262268,
      "learning_rate": 9.88132475186513e-06,
      "loss": 0.0202,
      "step": 14540
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.4392716884613037,
      "learning_rate": 9.875881206578032e-06,
      "loss": 0.0164,
      "step": 14545
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6249778270721436,
      "learning_rate": 9.870437698075464e-06,
      "loss": 0.0146,
      "step": 14550
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.5201079249382019,
      "learning_rate": 9.864994227970695e-06,
      "loss": 0.0145,
      "step": 14555
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.3231590986251831,
      "learning_rate": 9.859550797876984e-06,
      "loss": 0.0118,
      "step": 14560
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.5004443526268005,
      "learning_rate": 9.854107409407575e-06,
      "loss": 0.0194,
      "step": 14565
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.6936674118041992,
      "learning_rate": 9.848664064175698e-06,
      "loss": 0.0151,
      "step": 14570
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.3963569104671478,
      "learning_rate": 9.843220763794578e-06,
      "loss": 0.0112,
      "step": 14575
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.5949503779411316,
      "learning_rate": 9.837777509877425e-06,
      "loss": 0.0202,
      "step": 14580
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.6990585327148438,
      "learning_rate": 9.832334304037418e-06,
      "loss": 0.013,
      "step": 14585
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.7839829325675964,
      "learning_rate": 9.826891147887746e-06,
      "loss": 0.0103,
      "step": 14590
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.536649227142334,
      "learning_rate": 9.821448043041571e-06,
      "loss": 0.0096,
      "step": 14595
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.6189597845077515,
      "learning_rate": 9.816004991112042e-06,
      "loss": 0.0148,
      "step": 14600
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.5167118310928345,
      "learning_rate": 9.81056199371229e-06,
      "loss": 0.0158,
      "step": 14605
    },
    {
      "epoch": 4.56,
      "grad_norm": 1.288644790649414,
      "learning_rate": 9.805119052455433e-06,
      "loss": 0.0127,
      "step": 14610
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.5421487092971802,
      "learning_rate": 9.799676168954568e-06,
      "loss": 0.0171,
      "step": 14615
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.44585683941841125,
      "learning_rate": 9.79423334482279e-06,
      "loss": 0.0118,
      "step": 14620
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.49607759714126587,
      "learning_rate": 9.78879058167315e-06,
      "loss": 0.0122,
      "step": 14625
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.4464818835258484,
      "learning_rate": 9.783347881118698e-06,
      "loss": 0.0126,
      "step": 14630
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.638231098651886,
      "learning_rate": 9.77790524477247e-06,
      "loss": 0.0156,
      "step": 14635
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.5741913914680481,
      "learning_rate": 9.77246267424747e-06,
      "loss": 0.0175,
      "step": 14640
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.6646221876144409,
      "learning_rate": 9.76702017115669e-06,
      "loss": 0.0173,
      "step": 14645
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.5025471448898315,
      "learning_rate": 9.7615777371131e-06,
      "loss": 0.0142,
      "step": 14650
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.6054807901382446,
      "learning_rate": 9.756135373729654e-06,
      "loss": 0.0217,
      "step": 14655
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.6085841655731201,
      "learning_rate": 9.750693082619274e-06,
      "loss": 0.0111,
      "step": 14660
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.6730016469955444,
      "learning_rate": 9.74525086539487e-06,
      "loss": 0.0175,
      "step": 14665
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.7598947286605835,
      "learning_rate": 9.739808723669327e-06,
      "loss": 0.0182,
      "step": 14670
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.47956424951553345,
      "learning_rate": 9.73436665905551e-06,
      "loss": 0.0088,
      "step": 14675
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.5771499872207642,
      "learning_rate": 9.728924673166258e-06,
      "loss": 0.0159,
      "step": 14680
    },
    {
      "epoch": 4.58,
      "grad_norm": 1.1014750003814697,
      "learning_rate": 9.723482767614391e-06,
      "loss": 0.0154,
      "step": 14685
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6661223769187927,
      "learning_rate": 9.7180409440127e-06,
      "loss": 0.017,
      "step": 14690
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.7477891445159912,
      "learning_rate": 9.712599203973953e-06,
      "loss": 0.0209,
      "step": 14695
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6131916046142578,
      "learning_rate": 9.707157549110894e-06,
      "loss": 0.0165,
      "step": 14700
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.636452317237854,
      "learning_rate": 9.701715981036243e-06,
      "loss": 0.0114,
      "step": 14705
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5207667946815491,
      "learning_rate": 9.696274501362694e-06,
      "loss": 0.0169,
      "step": 14710
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.3022514581680298,
      "learning_rate": 9.69083311170291e-06,
      "loss": 0.0102,
      "step": 14715
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5009319186210632,
      "learning_rate": 9.685391813669537e-06,
      "loss": 0.0117,
      "step": 14720
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.5763520002365112,
      "learning_rate": 9.67995060887519e-06,
      "loss": 0.0137,
      "step": 14725
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.8080492615699768,
      "learning_rate": 9.674509498932443e-06,
      "loss": 0.0115,
      "step": 14730
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.8563107848167419,
      "learning_rate": 9.669068485453863e-06,
      "loss": 0.0134,
      "step": 14735
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5381313562393188,
      "learning_rate": 9.663627570051975e-06,
      "loss": 0.013,
      "step": 14740
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.7833681106567383,
      "learning_rate": 9.658186754339283e-06,
      "loss": 0.0181,
      "step": 14745
    },
    {
      "epoch": 4.6,
      "grad_norm": 1.1692765951156616,
      "learning_rate": 9.652746039928253e-06,
      "loss": 0.0174,
      "step": 14750
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5598737001419067,
      "learning_rate": 9.647305428431327e-06,
      "loss": 0.0122,
      "step": 14755
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.7559071183204651,
      "learning_rate": 9.641864921460916e-06,
      "loss": 0.018,
      "step": 14760
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.501556932926178,
      "learning_rate": 9.636424520629394e-06,
      "loss": 0.0131,
      "step": 14765
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.5668889880180359,
      "learning_rate": 9.630984227549112e-06,
      "loss": 0.0176,
      "step": 14770
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.5442070364952087,
      "learning_rate": 9.625544043832383e-06,
      "loss": 0.0134,
      "step": 14775
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.7403197884559631,
      "learning_rate": 9.620103971091493e-06,
      "loss": 0.0166,
      "step": 14780
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6325731873512268,
      "learning_rate": 9.61466401093869e-06,
      "loss": 0.0133,
      "step": 14785
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6614811420440674,
      "learning_rate": 9.609224164986192e-06,
      "loss": 0.0162,
      "step": 14790
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6733828783035278,
      "learning_rate": 9.603784434846182e-06,
      "loss": 0.0169,
      "step": 14795
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5302515625953674,
      "learning_rate": 9.598344822130808e-06,
      "loss": 0.0169,
      "step": 14800
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5816189646720886,
      "learning_rate": 9.592905328452182e-06,
      "loss": 0.0145,
      "step": 14805
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.4304620027542114,
      "learning_rate": 9.587465955422384e-06,
      "loss": 0.0179,
      "step": 14810
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5317957401275635,
      "learning_rate": 9.582026704653455e-06,
      "loss": 0.0148,
      "step": 14815
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.6904965043067932,
      "learning_rate": 9.576587577757403e-06,
      "loss": 0.0142,
      "step": 14820
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.7760544419288635,
      "learning_rate": 9.571148576346198e-06,
      "loss": 0.0174,
      "step": 14825
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.7780206203460693,
      "learning_rate": 9.565709702031773e-06,
      "loss": 0.0176,
      "step": 14830
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.5674579739570618,
      "learning_rate": 9.560270956426023e-06,
      "loss": 0.0157,
      "step": 14835
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.5733282566070557,
      "learning_rate": 9.554832341140801e-06,
      "loss": 0.0183,
      "step": 14840
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.320640504360199,
      "learning_rate": 9.54939385778793e-06,
      "loss": 0.0144,
      "step": 14845
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.533781111240387,
      "learning_rate": 9.543955507979185e-06,
      "loss": 0.0188,
      "step": 14850
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.8794890642166138,
      "learning_rate": 9.538517293326307e-06,
      "loss": 0.0153,
      "step": 14855
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.38584229350090027,
      "learning_rate": 9.533079215440995e-06,
      "loss": 0.0128,
      "step": 14860
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.48441481590270996,
      "learning_rate": 9.527641275934908e-06,
      "loss": 0.0142,
      "step": 14865
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.4725772738456726,
      "learning_rate": 9.52220347641967e-06,
      "loss": 0.0124,
      "step": 14870
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.5872272253036499,
      "learning_rate": 9.516765818506844e-06,
      "loss": 0.0118,
      "step": 14875
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.7887629866600037,
      "learning_rate": 9.511328303807977e-06,
      "loss": 0.0136,
      "step": 14880
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.5740352869033813,
      "learning_rate": 9.505890933934553e-06,
      "loss": 0.0118,
      "step": 14885
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.689967691898346,
      "learning_rate": 9.500453710498024e-06,
      "loss": 0.0135,
      "step": 14890
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.4478163421154022,
      "learning_rate": 9.495016635109795e-06,
      "loss": 0.0113,
      "step": 14895
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.653808057308197,
      "learning_rate": 9.489579709381229e-06,
      "loss": 0.0167,
      "step": 14900
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.3178090453147888,
      "learning_rate": 9.484142934923645e-06,
      "loss": 0.0112,
      "step": 14905
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.8894709348678589,
      "learning_rate": 9.47870631334831e-06,
      "loss": 0.0202,
      "step": 14910
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.4782661497592926,
      "learning_rate": 9.473269846266453e-06,
      "loss": 0.0131,
      "step": 14915
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.8897411227226257,
      "learning_rate": 9.467833535289257e-06,
      "loss": 0.0137,
      "step": 14920
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7093350291252136,
      "learning_rate": 9.462397382027857e-06,
      "loss": 0.0151,
      "step": 14925
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.6557932496070862,
      "learning_rate": 9.456961388093343e-06,
      "loss": 0.0102,
      "step": 14930
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.48611825704574585,
      "learning_rate": 9.451525555096753e-06,
      "loss": 0.0122,
      "step": 14935
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.5567466616630554,
      "learning_rate": 9.446089884649085e-06,
      "loss": 0.0171,
      "step": 14940
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.3401321470737457,
      "learning_rate": 9.44065437836128e-06,
      "loss": 0.0137,
      "step": 14945
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.40684616565704346,
      "learning_rate": 9.435219037844234e-06,
      "loss": 0.0118,
      "step": 14950
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.8801999688148499,
      "learning_rate": 9.429783864708798e-06,
      "loss": 0.018,
      "step": 14955
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.6474644541740417,
      "learning_rate": 9.42434886056577e-06,
      "loss": 0.0166,
      "step": 14960
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.8113958835601807,
      "learning_rate": 9.418914027025891e-06,
      "loss": 0.0158,
      "step": 14965
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.44780680537223816,
      "learning_rate": 9.413479365699868e-06,
      "loss": 0.0163,
      "step": 14970
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5623879432678223,
      "learning_rate": 9.408044878198343e-06,
      "loss": 0.014,
      "step": 14975
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5711343288421631,
      "learning_rate": 9.402610566131912e-06,
      "loss": 0.0155,
      "step": 14980
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5924063324928284,
      "learning_rate": 9.397176431111115e-06,
      "loss": 0.0153,
      "step": 14985
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.7692981958389282,
      "learning_rate": 9.391742474746444e-06,
      "loss": 0.0155,
      "step": 14990
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.7145710587501526,
      "learning_rate": 9.386308698648337e-06,
      "loss": 0.0171,
      "step": 14995
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.9549365639686584,
      "learning_rate": 9.380875104427175e-06,
      "loss": 0.0167,
      "step": 15000
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.5520309209823608,
      "learning_rate": 9.375441693693291e-06,
      "loss": 0.0128,
      "step": 15005
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.5013812780380249,
      "learning_rate": 9.37000846805696e-06,
      "loss": 0.0114,
      "step": 15010
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.8319398164749146,
      "learning_rate": 9.364575429128407e-06,
      "loss": 0.0156,
      "step": 15015
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.48659223318099976,
      "learning_rate": 9.359142578517787e-06,
      "loss": 0.014,
      "step": 15020
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.5164733529090881,
      "learning_rate": 9.353709917835218e-06,
      "loss": 0.0164,
      "step": 15025
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.4213215112686157,
      "learning_rate": 9.34827744869075e-06,
      "loss": 0.0124,
      "step": 15030
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.3633331060409546,
      "learning_rate": 9.342845172694383e-06,
      "loss": 0.012,
      "step": 15035
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.576378345489502,
      "learning_rate": 9.337413091456052e-06,
      "loss": 0.0174,
      "step": 15040
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.42284902930259705,
      "learning_rate": 9.331981206585642e-06,
      "loss": 0.021,
      "step": 15045
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.773980438709259,
      "learning_rate": 9.326549519692977e-06,
      "loss": 0.0167,
      "step": 15050
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.44876566529273987,
      "learning_rate": 9.321118032387816e-06,
      "loss": 0.0152,
      "step": 15055
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.48520439863204956,
      "learning_rate": 9.31568674627987e-06,
      "loss": 0.0151,
      "step": 15060
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.7521637678146362,
      "learning_rate": 9.310255662978783e-06,
      "loss": 0.0148,
      "step": 15065
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.6264703869819641,
      "learning_rate": 9.30482478409414e-06,
      "loss": 0.0158,
      "step": 15070
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.5164471864700317,
      "learning_rate": 9.29939411123547e-06,
      "loss": 0.0123,
      "step": 15075
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.4082270562648773,
      "learning_rate": 9.293963646012234e-06,
      "loss": 0.0101,
      "step": 15080
    },
    {
      "epoch": 4.7,
      "grad_norm": 1.352462887763977,
      "learning_rate": 9.288533390033838e-06,
      "loss": 0.0215,
      "step": 15085
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.31743505597114563,
      "learning_rate": 9.283103344909618e-06,
      "loss": 0.0114,
      "step": 15090
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.6116018891334534,
      "learning_rate": 9.277673512248855e-06,
      "loss": 0.0151,
      "step": 15095
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.5463698506355286,
      "learning_rate": 9.272243893660763e-06,
      "loss": 0.0124,
      "step": 15100
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.6346951723098755,
      "learning_rate": 9.266814490754494e-06,
      "loss": 0.0182,
      "step": 15105
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.543600857257843,
      "learning_rate": 9.261385305139135e-06,
      "loss": 0.0122,
      "step": 15110
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.7223466634750366,
      "learning_rate": 9.255956338423714e-06,
      "loss": 0.0171,
      "step": 15115
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.7255980372428894,
      "learning_rate": 9.250527592217183e-06,
      "loss": 0.0158,
      "step": 15120
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.43430185317993164,
      "learning_rate": 9.245099068128443e-06,
      "loss": 0.0161,
      "step": 15125
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.42145824432373047,
      "learning_rate": 9.239670767766315e-06,
      "loss": 0.0123,
      "step": 15130
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.3683522939682007,
      "learning_rate": 9.23424269273956e-06,
      "loss": 0.0122,
      "step": 15135
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5747255682945251,
      "learning_rate": 9.228814844656878e-06,
      "loss": 0.0147,
      "step": 15140
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.7251515984535217,
      "learning_rate": 9.22338722512689e-06,
      "loss": 0.0146,
      "step": 15145
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.39844369888305664,
      "learning_rate": 9.217959835758156e-06,
      "loss": 0.0143,
      "step": 15150
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.5769985318183899,
      "learning_rate": 9.212532678159173e-06,
      "loss": 0.0126,
      "step": 15155
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.45943641662597656,
      "learning_rate": 9.207105753938362e-06,
      "loss": 0.0206,
      "step": 15160
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.6093705296516418,
      "learning_rate": 9.201679064704067e-06,
      "loss": 0.0114,
      "step": 15165
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.801814079284668,
      "learning_rate": 9.196252612064582e-06,
      "loss": 0.0133,
      "step": 15170
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.41251370310783386,
      "learning_rate": 9.190826397628116e-06,
      "loss": 0.0133,
      "step": 15175
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.6013932228088379,
      "learning_rate": 9.185400423002813e-06,
      "loss": 0.0183,
      "step": 15180
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.7565571665763855,
      "learning_rate": 9.179974689796743e-06,
      "loss": 0.0185,
      "step": 15185
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.7211235761642456,
      "learning_rate": 9.17454919961791e-06,
      "loss": 0.016,
      "step": 15190
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.6997960209846497,
      "learning_rate": 9.169123954074242e-06,
      "loss": 0.0141,
      "step": 15195
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.4590340554714203,
      "learning_rate": 9.16369895477359e-06,
      "loss": 0.0119,
      "step": 15200
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.5132920145988464,
      "learning_rate": 9.158274203323739e-06,
      "loss": 0.0163,
      "step": 15205
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.590715765953064,
      "learning_rate": 9.152849701332398e-06,
      "loss": 0.0205,
      "step": 15210
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.45706048607826233,
      "learning_rate": 9.147425450407205e-06,
      "loss": 0.0119,
      "step": 15215
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.4636910557746887,
      "learning_rate": 9.14200145215572e-06,
      "loss": 0.0148,
      "step": 15220
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.9079941511154175,
      "learning_rate": 9.136577708185428e-06,
      "loss": 0.014,
      "step": 15225
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.912060022354126,
      "learning_rate": 9.131154220103743e-06,
      "loss": 0.009,
      "step": 15230
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.35609185695648193,
      "learning_rate": 9.125730989517994e-06,
      "loss": 0.0142,
      "step": 15235
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.6244210600852966,
      "learning_rate": 9.120308018035444e-06,
      "loss": 0.0134,
      "step": 15240
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.602467954158783,
      "learning_rate": 9.114885307263272e-06,
      "loss": 0.015,
      "step": 15245
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6154845952987671,
      "learning_rate": 9.109462858808586e-06,
      "loss": 0.0184,
      "step": 15250
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.5339671969413757,
      "learning_rate": 9.104040674278408e-06,
      "loss": 0.0133,
      "step": 15255
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.6302767395973206,
      "learning_rate": 9.098618755279693e-06,
      "loss": 0.0169,
      "step": 15260
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.762374758720398,
      "learning_rate": 9.093197103419309e-06,
      "loss": 0.0131,
      "step": 15265
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.5994053483009338,
      "learning_rate": 9.087775720304041e-06,
      "loss": 0.0152,
      "step": 15270
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.46510350704193115,
      "learning_rate": 9.082354607540606e-06,
      "loss": 0.0112,
      "step": 15275
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.32651185989379883,
      "learning_rate": 9.076933766735631e-06,
      "loss": 0.0123,
      "step": 15280
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.8065100908279419,
      "learning_rate": 9.071513199495671e-06,
      "loss": 0.0155,
      "step": 15285
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.653430700302124,
      "learning_rate": 9.066092907427194e-06,
      "loss": 0.0199,
      "step": 15290
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.4591868221759796,
      "learning_rate": 9.06067289213658e-06,
      "loss": 0.0131,
      "step": 15295
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.4707903563976288,
      "learning_rate": 9.055253155230147e-06,
      "loss": 0.0156,
      "step": 15300
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.5545510649681091,
      "learning_rate": 9.049833698314116e-06,
      "loss": 0.012,
      "step": 15305
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.5353935956954956,
      "learning_rate": 9.044414522994618e-06,
      "loss": 0.0197,
      "step": 15310
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.4159187078475952,
      "learning_rate": 9.038995630877716e-06,
      "loss": 0.0149,
      "step": 15315
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.5113319754600525,
      "learning_rate": 9.033577023569383e-06,
      "loss": 0.0129,
      "step": 15320
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.6067572832107544,
      "learning_rate": 9.028158702675505e-06,
      "loss": 0.0192,
      "step": 15325
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.422309011220932,
      "learning_rate": 9.022740669801888e-06,
      "loss": 0.0153,
      "step": 15330
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.5361863970756531,
      "learning_rate": 9.017322926554251e-06,
      "loss": 0.015,
      "step": 15335
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.28277361392974854,
      "learning_rate": 9.011905474538224e-06,
      "loss": 0.0131,
      "step": 15340
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.9252229332923889,
      "learning_rate": 9.006488315359355e-06,
      "loss": 0.0146,
      "step": 15345
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.7406449317932129,
      "learning_rate": 9.0010714506231e-06,
      "loss": 0.0134,
      "step": 15350
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.6453289985656738,
      "learning_rate": 8.995654881934832e-06,
      "loss": 0.0132,
      "step": 15355
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.7026208639144897,
      "learning_rate": 8.990238610899838e-06,
      "loss": 0.0181,
      "step": 15360
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.6112136244773865,
      "learning_rate": 8.98482263912331e-06,
      "loss": 0.0123,
      "step": 15365
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.5226003527641296,
      "learning_rate": 8.97940696821036e-06,
      "loss": 0.0135,
      "step": 15370
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.41499006748199463,
      "learning_rate": 8.973991599766005e-06,
      "loss": 0.0174,
      "step": 15375
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.7187453508377075,
      "learning_rate": 8.968576535395172e-06,
      "loss": 0.0132,
      "step": 15380
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.5334180593490601,
      "learning_rate": 8.963161776702698e-06,
      "loss": 0.0163,
      "step": 15385
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.5039098858833313,
      "learning_rate": 8.957747325293333e-06,
      "loss": 0.0167,
      "step": 15390
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.9802941083908081,
      "learning_rate": 8.952333182771734e-06,
      "loss": 0.0129,
      "step": 15395
    },
    {
      "epoch": 4.8,
      "grad_norm": 1.0111184120178223,
      "learning_rate": 8.946919350742462e-06,
      "loss": 0.0111,
      "step": 15400
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.33556801080703735,
      "learning_rate": 8.941505830809996e-06,
      "loss": 0.01,
      "step": 15405
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.4566008746623993,
      "learning_rate": 8.93609262457872e-06,
      "loss": 0.0147,
      "step": 15410
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.6226485371589661,
      "learning_rate": 8.930679733652908e-06,
      "loss": 0.0123,
      "step": 15415
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.47943395376205444,
      "learning_rate": 8.925267159636766e-06,
      "loss": 0.0138,
      "step": 15420
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.7245271801948547,
      "learning_rate": 8.91985490413439e-06,
      "loss": 0.0179,
      "step": 15425
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.3484051525592804,
      "learning_rate": 8.914442968749787e-06,
      "loss": 0.0144,
      "step": 15430
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.6594104766845703,
      "learning_rate": 8.909031355086868e-06,
      "loss": 0.0169,
      "step": 15435
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.19373483955860138,
      "learning_rate": 8.903620064749447e-06,
      "loss": 0.0121,
      "step": 15440
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.6843541860580444,
      "learning_rate": 8.898209099341246e-06,
      "loss": 0.0151,
      "step": 15445
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.30797871947288513,
      "learning_rate": 8.892798460465894e-06,
      "loss": 0.0097,
      "step": 15450
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.7222657799720764,
      "learning_rate": 8.887388149726907e-06,
      "loss": 0.0144,
      "step": 15455
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.6683160662651062,
      "learning_rate": 8.881978168727717e-06,
      "loss": 0.0116,
      "step": 15460
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.3965126872062683,
      "learning_rate": 8.876568519071661e-06,
      "loss": 0.0117,
      "step": 15465
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.47046273946762085,
      "learning_rate": 8.87115920236197e-06,
      "loss": 0.0136,
      "step": 15470
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.7667253613471985,
      "learning_rate": 8.86575022020178e-06,
      "loss": 0.0152,
      "step": 15475
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.9530908465385437,
      "learning_rate": 8.860341574194125e-06,
      "loss": 0.0174,
      "step": 15480
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.7011138796806335,
      "learning_rate": 8.854933265941946e-06,
      "loss": 0.0201,
      "step": 15485
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.6812859773635864,
      "learning_rate": 8.849525297048072e-06,
      "loss": 0.0135,
      "step": 15490
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.7322203516960144,
      "learning_rate": 8.844117669115242e-06,
      "loss": 0.0143,
      "step": 15495
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.5292491912841797,
      "learning_rate": 8.83871038374609e-06,
      "loss": 0.0162,
      "step": 15500
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.4952675700187683,
      "learning_rate": 8.833303442543148e-06,
      "loss": 0.0111,
      "step": 15505
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.770052969455719,
      "learning_rate": 8.827896847108851e-06,
      "loss": 0.0154,
      "step": 15510
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.6179236769676208,
      "learning_rate": 8.822490599045524e-06,
      "loss": 0.013,
      "step": 15515
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.39276185631752014,
      "learning_rate": 8.817084699955397e-06,
      "loss": 0.0131,
      "step": 15520
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.6833389401435852,
      "learning_rate": 8.811679151440585e-06,
      "loss": 0.0162,
      "step": 15525
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.504504382610321,
      "learning_rate": 8.806273955103111e-06,
      "loss": 0.0147,
      "step": 15530
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.38780394196510315,
      "learning_rate": 8.800869112544887e-06,
      "loss": 0.0119,
      "step": 15535
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.18057525157928467,
      "learning_rate": 8.795464625367724e-06,
      "loss": 0.013,
      "step": 15540
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.5500834584236145,
      "learning_rate": 8.790060495173321e-06,
      "loss": 0.0167,
      "step": 15545
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.708239734172821,
      "learning_rate": 8.784656723563283e-06,
      "loss": 0.0158,
      "step": 15550
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.4792191684246063,
      "learning_rate": 8.779253312139102e-06,
      "loss": 0.017,
      "step": 15555
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.517033576965332,
      "learning_rate": 8.773850262502152e-06,
      "loss": 0.0129,
      "step": 15560
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.6025055646896362,
      "learning_rate": 8.76844757625372e-06,
      "loss": 0.0144,
      "step": 15565
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.868748664855957,
      "learning_rate": 8.763045254994973e-06,
      "loss": 0.0184,
      "step": 15570
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.5168037414550781,
      "learning_rate": 8.757643300326975e-06,
      "loss": 0.0128,
      "step": 15575
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.6352051496505737,
      "learning_rate": 8.752241713850675e-06,
      "loss": 0.0196,
      "step": 15580
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.451431006193161,
      "learning_rate": 8.74684049716692e-06,
      "loss": 0.0122,
      "step": 15585
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.6849236488342285,
      "learning_rate": 8.741439651876449e-06,
      "loss": 0.0191,
      "step": 15590
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.7148830890655518,
      "learning_rate": 8.736039179579876e-06,
      "loss": 0.0154,
      "step": 15595
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.5371099710464478,
      "learning_rate": 8.730639081877722e-06,
      "loss": 0.015,
      "step": 15600
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.5085660219192505,
      "learning_rate": 8.725239360370385e-06,
      "loss": 0.0134,
      "step": 15605
    },
    {
      "epoch": 4.87,
      "grad_norm": 1.143683671951294,
      "learning_rate": 8.719840016658162e-06,
      "loss": 0.0181,
      "step": 15610
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.539717435836792,
      "learning_rate": 8.71444105234123e-06,
      "loss": 0.011,
      "step": 15615
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.49247002601623535,
      "learning_rate": 8.709042469019655e-06,
      "loss": 0.014,
      "step": 15620
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.8581780195236206,
      "learning_rate": 8.703644268293391e-06,
      "loss": 0.0174,
      "step": 15625
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.5282270312309265,
      "learning_rate": 8.698246451762283e-06,
      "loss": 0.0144,
      "step": 15630
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.8629797697067261,
      "learning_rate": 8.692849021026051e-06,
      "loss": 0.009,
      "step": 15635
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.8482663035392761,
      "learning_rate": 8.68745197768431e-06,
      "loss": 0.0156,
      "step": 15640
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.4563334286212921,
      "learning_rate": 8.682055323336557e-06,
      "loss": 0.0098,
      "step": 15645
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.5153648853302002,
      "learning_rate": 8.676659059582177e-06,
      "loss": 0.0128,
      "step": 15650
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.524944543838501,
      "learning_rate": 8.671263188020433e-06,
      "loss": 0.0151,
      "step": 15655
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.4763322174549103,
      "learning_rate": 8.66586771025048e-06,
      "loss": 0.0139,
      "step": 15660
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.3153722882270813,
      "learning_rate": 8.66047262787135e-06,
      "loss": 0.0118,
      "step": 15665
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.5332691073417664,
      "learning_rate": 8.655077942481956e-06,
      "loss": 0.0115,
      "step": 15670
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.7228777408599854,
      "learning_rate": 8.649683655681098e-06,
      "loss": 0.0178,
      "step": 15675
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.6176360249519348,
      "learning_rate": 8.644289769067459e-06,
      "loss": 0.0122,
      "step": 15680
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.5154833197593689,
      "learning_rate": 8.6388962842396e-06,
      "loss": 0.0131,
      "step": 15685
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.4984290301799774,
      "learning_rate": 8.63350320279596e-06,
      "loss": 0.0138,
      "step": 15690
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.4621444046497345,
      "learning_rate": 8.62811052633487e-06,
      "loss": 0.0147,
      "step": 15695
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.476084440946579,
      "learning_rate": 8.622718256454535e-06,
      "loss": 0.0144,
      "step": 15700
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.3840945363044739,
      "learning_rate": 8.617326394753023e-06,
      "loss": 0.0155,
      "step": 15705
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.40964987874031067,
      "learning_rate": 8.611934942828309e-06,
      "loss": 0.0167,
      "step": 15710
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.9059739708900452,
      "learning_rate": 8.60654390227823e-06,
      "loss": 0.0193,
      "step": 15715
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.5482441186904907,
      "learning_rate": 8.601153274700504e-06,
      "loss": 0.0132,
      "step": 15720
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.5010052919387817,
      "learning_rate": 8.595763061692729e-06,
      "loss": 0.0167,
      "step": 15725
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.5105224847793579,
      "learning_rate": 8.590373264852376e-06,
      "loss": 0.0143,
      "step": 15730
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.4789036512374878,
      "learning_rate": 8.5849838857768e-06,
      "loss": 0.0167,
      "step": 15735
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.28978851437568665,
      "learning_rate": 8.579594926063219e-06,
      "loss": 0.0093,
      "step": 15740
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.5114542245864868,
      "learning_rate": 8.57420638730874e-06,
      "loss": 0.0107,
      "step": 15745
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.5879108309745789,
      "learning_rate": 8.568818271110338e-06,
      "loss": 0.0133,
      "step": 15750
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.5466969013214111,
      "learning_rate": 8.563430579064866e-06,
      "loss": 0.015,
      "step": 15755
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.5947604179382324,
      "learning_rate": 8.558043312769053e-06,
      "loss": 0.0143,
      "step": 15760
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.5738160014152527,
      "learning_rate": 8.552656473819495e-06,
      "loss": 0.0182,
      "step": 15765
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.31728726625442505,
      "learning_rate": 8.547270063812669e-06,
      "loss": 0.0096,
      "step": 15770
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.4764464497566223,
      "learning_rate": 8.541884084344918e-06,
      "loss": 0.0186,
      "step": 15775
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.38102155923843384,
      "learning_rate": 8.536498537012461e-06,
      "loss": 0.0113,
      "step": 15780
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.42733028531074524,
      "learning_rate": 8.531113423411388e-06,
      "loss": 0.0103,
      "step": 15785
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.31331437826156616,
      "learning_rate": 8.525728745137662e-06,
      "loss": 0.0093,
      "step": 15790
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.3956007957458496,
      "learning_rate": 8.520344503787119e-06,
      "loss": 0.0126,
      "step": 15795
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.46345624327659607,
      "learning_rate": 8.514960700955457e-06,
      "loss": 0.0116,
      "step": 15800
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.4458674192428589,
      "learning_rate": 8.509577338238255e-06,
      "loss": 0.0129,
      "step": 15805
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.5788841843605042,
      "learning_rate": 8.504194417230955e-06,
      "loss": 0.0106,
      "step": 15810
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.527131974697113,
      "learning_rate": 8.498811939528866e-06,
      "loss": 0.0156,
      "step": 15815
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.8801189064979553,
      "learning_rate": 8.49342990672717e-06,
      "loss": 0.0135,
      "step": 15820
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.5658692121505737,
      "learning_rate": 8.48804832042092e-06,
      "loss": 0.0155,
      "step": 15825
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.5295402407646179,
      "learning_rate": 8.482667182205024e-06,
      "loss": 0.0127,
      "step": 15830
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.5799888968467712,
      "learning_rate": 8.477286493674272e-06,
      "loss": 0.0153,
      "step": 15835
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.4363563060760498,
      "learning_rate": 8.471906256423315e-06,
      "loss": 0.0123,
      "step": 15840
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.8872877359390259,
      "learning_rate": 8.466526472046675e-06,
      "loss": 0.0085,
      "step": 15845
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.5638665556907654,
      "learning_rate": 8.46114714213872e-06,
      "loss": 0.0132,
      "step": 15850
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.42812150716781616,
      "learning_rate": 8.455768268293709e-06,
      "loss": 0.0082,
      "step": 15855
    },
    {
      "epoch": 4.95,
      "grad_norm": 1.0099304914474487,
      "learning_rate": 8.45038985210575e-06,
      "loss": 0.024,
      "step": 15860
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.5413514971733093,
      "learning_rate": 8.445011895168826e-06,
      "loss": 0.0207,
      "step": 15865
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.6191872358322144,
      "learning_rate": 8.439634399076772e-06,
      "loss": 0.0148,
      "step": 15870
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.48538586497306824,
      "learning_rate": 8.434257365423296e-06,
      "loss": 0.0151,
      "step": 15875
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.42932063341140747,
      "learning_rate": 8.428880795801965e-06,
      "loss": 0.0157,
      "step": 15880
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.3916996121406555,
      "learning_rate": 8.423504691806209e-06,
      "loss": 0.0109,
      "step": 15885
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.24138857424259186,
      "learning_rate": 8.418129055029316e-06,
      "loss": 0.0124,
      "step": 15890
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.2233244627714157,
      "learning_rate": 8.412753887064441e-06,
      "loss": 0.0084,
      "step": 15895
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.5459782481193542,
      "learning_rate": 8.407379189504602e-06,
      "loss": 0.0167,
      "step": 15900
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.5292870402336121,
      "learning_rate": 8.402004963942672e-06,
      "loss": 0.0161,
      "step": 15905
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.5536977648735046,
      "learning_rate": 8.396631211971386e-06,
      "loss": 0.0138,
      "step": 15910
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.6610172986984253,
      "learning_rate": 8.391257935183342e-06,
      "loss": 0.0172,
      "step": 15915
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.5862566232681274,
      "learning_rate": 8.38588513517099e-06,
      "loss": 0.0158,
      "step": 15920
    },
    {
      "epoch": 4.97,
      "grad_norm": 1.3602510690689087,
      "learning_rate": 8.380512813526641e-06,
      "loss": 0.0151,
      "step": 15925
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.3232061266899109,
      "learning_rate": 8.375140971842471e-06,
      "loss": 0.0097,
      "step": 15930
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.41499772667884827,
      "learning_rate": 8.369769611710504e-06,
      "loss": 0.0144,
      "step": 15935
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.504601001739502,
      "learning_rate": 8.36439873472263e-06,
      "loss": 0.0119,
      "step": 15940
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.5243173837661743,
      "learning_rate": 8.35902834247059e-06,
      "loss": 0.0126,
      "step": 15945
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.8670143485069275,
      "learning_rate": 8.353658436545983e-06,
      "loss": 0.0158,
      "step": 15950
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.7447572946548462,
      "learning_rate": 8.348289018540268e-06,
      "loss": 0.0206,
      "step": 15955
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.4693433344364166,
      "learning_rate": 8.342920090044748e-06,
      "loss": 0.012,
      "step": 15960
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.7901753783226013,
      "learning_rate": 8.337551652650592e-06,
      "loss": 0.0199,
      "step": 15965
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.5879650712013245,
      "learning_rate": 8.332183707948822e-06,
      "loss": 0.0142,
      "step": 15970
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.5871224403381348,
      "learning_rate": 8.326816257530308e-06,
      "loss": 0.0129,
      "step": 15975
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.6463751792907715,
      "learning_rate": 8.32144930298578e-06,
      "loss": 0.0136,
      "step": 15980
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.4785372316837311,
      "learning_rate": 8.316082845905815e-06,
      "loss": 0.0163,
      "step": 15985
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.4490886330604553,
      "learning_rate": 8.310716887880856e-06,
      "loss": 0.0157,
      "step": 15990
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.9457346796989441,
      "learning_rate": 8.305351430501171e-06,
      "loss": 0.0127,
      "step": 15995
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.34795841574668884,
      "learning_rate": 8.29998647535691e-06,
      "loss": 0.0121,
      "step": 16000
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.3530838191509247,
      "learning_rate": 8.294622024038055e-06,
      "loss": 0.0157,
      "step": 16005
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.6324779391288757,
      "learning_rate": 8.289258078134446e-06,
      "loss": 0.014,
      "step": 16010
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.674531102180481,
      "learning_rate": 8.283894639235773e-06,
      "loss": 0.0165,
      "step": 16015
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5302462577819824,
      "learning_rate": 8.278531708931573e-06,
      "loss": 0.0169,
      "step": 16020
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5076550245285034,
      "learning_rate": 8.273169288811236e-06,
      "loss": 0.0164,
      "step": 16025
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5246521234512329,
      "learning_rate": 8.267807380463993e-06,
      "loss": 0.012,
      "step": 16030
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.37405821681022644,
      "learning_rate": 8.262445985478933e-06,
      "loss": 0.0099,
      "step": 16035
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4800686836242676,
      "learning_rate": 8.257085105444985e-06,
      "loss": 0.0066,
      "step": 16040
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4580146372318268,
      "learning_rate": 8.251724741950934e-06,
      "loss": 0.0065,
      "step": 16045
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.4707932770252228,
      "learning_rate": 8.246364896585404e-06,
      "loss": 0.0094,
      "step": 16050
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.23127494752407074,
      "learning_rate": 8.241005570936871e-06,
      "loss": 0.0072,
      "step": 16055
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.7896751761436462,
      "learning_rate": 8.235646766593654e-06,
      "loss": 0.0086,
      "step": 16060
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.3771972358226776,
      "learning_rate": 8.230288485143913e-06,
      "loss": 0.0096,
      "step": 16065
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.31021884083747864,
      "learning_rate": 8.224930728175663e-06,
      "loss": 0.0079,
      "step": 16070
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.5711778998374939,
      "learning_rate": 8.219573497276756e-06,
      "loss": 0.0109,
      "step": 16075
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.3094363510608673,
      "learning_rate": 8.214216794034889e-06,
      "loss": 0.0123,
      "step": 16080
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.5012419819831848,
      "learning_rate": 8.208860620037608e-06,
      "loss": 0.0093,
      "step": 16085
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.26737749576568604,
      "learning_rate": 8.203504976872298e-06,
      "loss": 0.0066,
      "step": 16090
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.5392455458641052,
      "learning_rate": 8.19814986612619e-06,
      "loss": 0.0064,
      "step": 16095
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.4574849009513855,
      "learning_rate": 8.192795289386343e-06,
      "loss": 0.0084,
      "step": 16100
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.614558756351471,
      "learning_rate": 8.187441248239681e-06,
      "loss": 0.0085,
      "step": 16105
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.3227488696575165,
      "learning_rate": 8.18208774427295e-06,
      "loss": 0.0082,
      "step": 16110
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.3623669147491455,
      "learning_rate": 8.176734779072747e-06,
      "loss": 0.0071,
      "step": 16115
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.20663155615329742,
      "learning_rate": 8.171382354225504e-06,
      "loss": 0.0074,
      "step": 16120
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.4030447006225586,
      "learning_rate": 8.166030471317497e-06,
      "loss": 0.0109,
      "step": 16125
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.3899942934513092,
      "learning_rate": 8.160679131934842e-06,
      "loss": 0.0081,
      "step": 16130
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.5883195400238037,
      "learning_rate": 8.15532833766349e-06,
      "loss": 0.0062,
      "step": 16135
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.509239137172699,
      "learning_rate": 8.149978090089228e-06,
      "loss": 0.0059,
      "step": 16140
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.18122568726539612,
      "learning_rate": 8.144628390797688e-06,
      "loss": 0.0079,
      "step": 16145
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.346282422542572,
      "learning_rate": 8.139279241374338e-06,
      "loss": 0.0067,
      "step": 16150
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.5716591477394104,
      "learning_rate": 8.133930643404479e-06,
      "loss": 0.0069,
      "step": 16155
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.3420288860797882,
      "learning_rate": 8.128582598473254e-06,
      "loss": 0.0071,
      "step": 16160
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.4361545145511627,
      "learning_rate": 8.123235108165636e-06,
      "loss": 0.009,
      "step": 16165
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.3243066072463989,
      "learning_rate": 8.117888174066439e-06,
      "loss": 0.0087,
      "step": 16170
    },
    {
      "epoch": 5.04,
      "grad_norm": 0.2942553460597992,
      "learning_rate": 8.11254179776031e-06,
      "loss": 0.0061,
      "step": 16175
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.24585643410682678,
      "learning_rate": 8.107195980831726e-06,
      "loss": 0.0094,
      "step": 16180
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.2915532886981964,
      "learning_rate": 8.101850724865005e-06,
      "loss": 0.0049,
      "step": 16185
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.25451603531837463,
      "learning_rate": 8.096506031444302e-06,
      "loss": 0.0079,
      "step": 16190
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.315114289522171,
      "learning_rate": 8.091161902153594e-06,
      "loss": 0.0079,
      "step": 16195
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.4226455092430115,
      "learning_rate": 8.085818338576702e-06,
      "loss": 0.0066,
      "step": 16200
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.4661683440208435,
      "learning_rate": 8.08047534229727e-06,
      "loss": 0.0088,
      "step": 16205
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.4152611494064331,
      "learning_rate": 8.075132914898776e-06,
      "loss": 0.0053,
      "step": 16210
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.23992493748664856,
      "learning_rate": 8.069791057964535e-06,
      "loss": 0.0048,
      "step": 16215
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.4227517545223236,
      "learning_rate": 8.06444977307769e-06,
      "loss": 0.0073,
      "step": 16220
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.176507830619812,
      "learning_rate": 8.059109061821208e-06,
      "loss": 0.0058,
      "step": 16225
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.4533865749835968,
      "learning_rate": 8.053768925777897e-06,
      "loss": 0.0088,
      "step": 16230
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.4857219159603119,
      "learning_rate": 8.048429366530387e-06,
      "loss": 0.0086,
      "step": 16235
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.4926903247833252,
      "learning_rate": 8.043090385661147e-06,
      "loss": 0.0075,
      "step": 16240
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.26682066917419434,
      "learning_rate": 8.037751984752453e-06,
      "loss": 0.0061,
      "step": 16245
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.44272875785827637,
      "learning_rate": 8.032414165386433e-06,
      "loss": 0.0075,
      "step": 16250
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.3670891523361206,
      "learning_rate": 8.027076929145028e-06,
      "loss": 0.0086,
      "step": 16255
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.3133547902107239,
      "learning_rate": 8.021740277610015e-06,
      "loss": 0.0058,
      "step": 16260
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.43501976132392883,
      "learning_rate": 8.016404212362989e-06,
      "loss": 0.0054,
      "step": 16265
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.41824793815612793,
      "learning_rate": 8.011068734985377e-06,
      "loss": 0.0088,
      "step": 16270
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.5613520741462708,
      "learning_rate": 8.005733847058438e-06,
      "loss": 0.0073,
      "step": 16275
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.4292289912700653,
      "learning_rate": 8.000399550163237e-06,
      "loss": 0.0079,
      "step": 16280
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.5372112989425659,
      "learning_rate": 7.99506584588068e-06,
      "loss": 0.0129,
      "step": 16285
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.37449565529823303,
      "learning_rate": 7.989732735791495e-06,
      "loss": 0.0078,
      "step": 16290
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.26634928584098816,
      "learning_rate": 7.984400221476233e-06,
      "loss": 0.0056,
      "step": 16295
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.3031364977359772,
      "learning_rate": 7.979068304515265e-06,
      "loss": 0.0039,
      "step": 16300
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.6217584609985352,
      "learning_rate": 7.973736986488789e-06,
      "loss": 0.0059,
      "step": 16305
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.40330415964126587,
      "learning_rate": 7.96840626897682e-06,
      "loss": 0.007,
      "step": 16310
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.30831795930862427,
      "learning_rate": 7.963076153559207e-06,
      "loss": 0.0043,
      "step": 16315
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.4585668444633484,
      "learning_rate": 7.957746641815603e-06,
      "loss": 0.0068,
      "step": 16320
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.25789645314216614,
      "learning_rate": 7.952417735325497e-06,
      "loss": 0.0073,
      "step": 16325
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.5011563897132874,
      "learning_rate": 7.94708943566819e-06,
      "loss": 0.0117,
      "step": 16330
    },
    {
      "epoch": 5.09,
      "grad_norm": 0.5686031579971313,
      "learning_rate": 7.941761744422811e-06,
      "loss": 0.0107,
      "step": 16335
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.4478742778301239,
      "learning_rate": 7.9364346631683e-06,
      "loss": 0.0066,
      "step": 16340
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.4688143730163574,
      "learning_rate": 7.931108193483423e-06,
      "loss": 0.0128,
      "step": 16345
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.34905996918678284,
      "learning_rate": 7.925782336946764e-06,
      "loss": 0.0089,
      "step": 16350
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.3972848355770111,
      "learning_rate": 7.920457095136715e-06,
      "loss": 0.0091,
      "step": 16355
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.18540921807289124,
      "learning_rate": 7.915132469631501e-06,
      "loss": 0.006,
      "step": 16360
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.46762675046920776,
      "learning_rate": 7.909808462009156e-06,
      "loss": 0.007,
      "step": 16365
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.1555611789226532,
      "learning_rate": 7.90448507384753e-06,
      "loss": 0.005,
      "step": 16370
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.6360225081443787,
      "learning_rate": 7.899162306724293e-06,
      "loss": 0.0076,
      "step": 16375
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.5464015007019043,
      "learning_rate": 7.893840162216933e-06,
      "loss": 0.0081,
      "step": 16380
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.37012243270874023,
      "learning_rate": 7.88851864190275e-06,
      "loss": 0.0061,
      "step": 16385
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.30513015389442444,
      "learning_rate": 7.883197747358851e-06,
      "loss": 0.0057,
      "step": 16390
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.500765323638916,
      "learning_rate": 7.877877480162173e-06,
      "loss": 0.0061,
      "step": 16395
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.5255210399627686,
      "learning_rate": 7.872557841889459e-06,
      "loss": 0.006,
      "step": 16400
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.43981844186782837,
      "learning_rate": 7.867238834117266e-06,
      "loss": 0.0068,
      "step": 16405
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.4827001392841339,
      "learning_rate": 7.861920458421963e-06,
      "loss": 0.0082,
      "step": 16410
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.5710332989692688,
      "learning_rate": 7.856602716379734e-06,
      "loss": 0.0095,
      "step": 16415
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.3762331008911133,
      "learning_rate": 7.85128560956658e-06,
      "loss": 0.0058,
      "step": 16420
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.7430204749107361,
      "learning_rate": 7.8459691395583e-06,
      "loss": 0.009,
      "step": 16425
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.47367557883262634,
      "learning_rate": 7.840653307930511e-06,
      "loss": 0.0082,
      "step": 16430
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.7471944093704224,
      "learning_rate": 7.83533811625865e-06,
      "loss": 0.0095,
      "step": 16435
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.3423488438129425,
      "learning_rate": 7.830023566117954e-06,
      "loss": 0.006,
      "step": 16440
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.6812673807144165,
      "learning_rate": 7.824709659083471e-06,
      "loss": 0.0068,
      "step": 16445
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.2431548535823822,
      "learning_rate": 7.819396396730062e-06,
      "loss": 0.0069,
      "step": 16450
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.22807416319847107,
      "learning_rate": 7.814083780632392e-06,
      "loss": 0.0073,
      "step": 16455
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.495912104845047,
      "learning_rate": 7.808771812364942e-06,
      "loss": 0.0071,
      "step": 16460
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.37738263607025146,
      "learning_rate": 7.80346049350199e-06,
      "loss": 0.0069,
      "step": 16465
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.275280624628067,
      "learning_rate": 7.798149825617632e-06,
      "loss": 0.006,
      "step": 16470
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.23672664165496826,
      "learning_rate": 7.792839810285762e-06,
      "loss": 0.0075,
      "step": 16475
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.2902403771877289,
      "learning_rate": 7.787530449080092e-06,
      "loss": 0.0057,
      "step": 16480
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.41451671719551086,
      "learning_rate": 7.782221743574131e-06,
      "loss": 0.0067,
      "step": 16485
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.5114469528198242,
      "learning_rate": 7.776913695341195e-06,
      "loss": 0.0067,
      "step": 16490
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.7483134269714355,
      "learning_rate": 7.771606305954413e-06,
      "loss": 0.0073,
      "step": 16495
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.6692242622375488,
      "learning_rate": 7.766299576986702e-06,
      "loss": 0.006,
      "step": 16500
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.3661095201969147,
      "learning_rate": 7.7609935100108e-06,
      "loss": 0.0071,
      "step": 16505
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.6875879764556885,
      "learning_rate": 7.75568810659924e-06,
      "loss": 0.0059,
      "step": 16510
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.6691876649856567,
      "learning_rate": 7.750383368324362e-06,
      "loss": 0.0083,
      "step": 16515
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.4257379472255707,
      "learning_rate": 7.745079296758305e-06,
      "loss": 0.0063,
      "step": 16520
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.509445071220398,
      "learning_rate": 7.739775893473018e-06,
      "loss": 0.0089,
      "step": 16525
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.48248153924942017,
      "learning_rate": 7.734473160040248e-06,
      "loss": 0.0086,
      "step": 16530
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.311449259519577,
      "learning_rate": 7.729171098031532e-06,
      "loss": 0.0056,
      "step": 16535
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.5185677409172058,
      "learning_rate": 7.723869709018226e-06,
      "loss": 0.0065,
      "step": 16540
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.3155025243759155,
      "learning_rate": 7.718568994571479e-06,
      "loss": 0.007,
      "step": 16545
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.6853141784667969,
      "learning_rate": 7.713268956262237e-06,
      "loss": 0.0074,
      "step": 16550
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.40242108702659607,
      "learning_rate": 7.707969595661253e-06,
      "loss": 0.0061,
      "step": 16555
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.35108885169029236,
      "learning_rate": 7.702670914339067e-06,
      "loss": 0.0073,
      "step": 16560
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.5612503290176392,
      "learning_rate": 7.697372913866041e-06,
      "loss": 0.0081,
      "step": 16565
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.19322334229946136,
      "learning_rate": 7.692075595812303e-06,
      "loss": 0.0072,
      "step": 16570
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.4028273820877075,
      "learning_rate": 7.686778961747799e-06,
      "loss": 0.0103,
      "step": 16575
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.47766733169555664,
      "learning_rate": 7.681483013242275e-06,
      "loss": 0.0092,
      "step": 16580
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.3629557490348816,
      "learning_rate": 7.676187751865266e-06,
      "loss": 0.0084,
      "step": 16585
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.30036285519599915,
      "learning_rate": 7.6708931791861e-06,
      "loss": 0.0088,
      "step": 16590
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.5219706296920776,
      "learning_rate": 7.665599296773911e-06,
      "loss": 0.0074,
      "step": 16595
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.4829447567462921,
      "learning_rate": 7.660306106197623e-06,
      "loss": 0.0044,
      "step": 16600
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.3575102984905243,
      "learning_rate": 7.655013609025952e-06,
      "loss": 0.0078,
      "step": 16605
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.36241811513900757,
      "learning_rate": 7.649721806827413e-06,
      "loss": 0.0064,
      "step": 16610
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.4477905035018921,
      "learning_rate": 7.644430701170314e-06,
      "loss": 0.0054,
      "step": 16615
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.5018873810768127,
      "learning_rate": 7.639140293622755e-06,
      "loss": 0.0085,
      "step": 16620
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.1742485612630844,
      "learning_rate": 7.633850585752634e-06,
      "loss": 0.0064,
      "step": 16625
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.408871591091156,
      "learning_rate": 7.628561579127635e-06,
      "loss": 0.0069,
      "step": 16630
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.4098290801048279,
      "learning_rate": 7.623273275315237e-06,
      "loss": 0.0051,
      "step": 16635
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.3893098533153534,
      "learning_rate": 7.617985675882716e-06,
      "loss": 0.0071,
      "step": 16640
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.2370014637708664,
      "learning_rate": 7.612698782397128e-06,
      "loss": 0.0065,
      "step": 16645
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.3227474093437195,
      "learning_rate": 7.607412596425327e-06,
      "loss": 0.0053,
      "step": 16650
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.3771594166755676,
      "learning_rate": 7.60212711953396e-06,
      "loss": 0.0074,
      "step": 16655
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.28053686022758484,
      "learning_rate": 7.596842353289455e-06,
      "loss": 0.0074,
      "step": 16660
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.30859634280204773,
      "learning_rate": 7.5915582992580375e-06,
      "loss": 0.0062,
      "step": 16665
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.3582562506198883,
      "learning_rate": 7.586274959005719e-06,
      "loss": 0.0077,
      "step": 16670
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.5159679055213928,
      "learning_rate": 7.580992334098305e-06,
      "loss": 0.0093,
      "step": 16675
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.3801780343055725,
      "learning_rate": 7.57571042610137e-06,
      "loss": 0.0064,
      "step": 16680
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.3430151045322418,
      "learning_rate": 7.5704292365803e-06,
      "loss": 0.0072,
      "step": 16685
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.25590091943740845,
      "learning_rate": 7.565148767100254e-06,
      "loss": 0.0068,
      "step": 16690
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.42597323656082153,
      "learning_rate": 7.559869019226182e-06,
      "loss": 0.0068,
      "step": 16695
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.19612255692481995,
      "learning_rate": 7.55458999452282e-06,
      "loss": 0.0068,
      "step": 16700
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.16416482627391815,
      "learning_rate": 7.549311694554684e-06,
      "loss": 0.0055,
      "step": 16705
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.4428281784057617,
      "learning_rate": 7.5440341208860915e-06,
      "loss": 0.009,
      "step": 16710
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.3866334855556488,
      "learning_rate": 7.538757275081121e-06,
      "loss": 0.0053,
      "step": 16715
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.5221220850944519,
      "learning_rate": 7.5334811587036536e-06,
      "loss": 0.0095,
      "step": 16720
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.5159658193588257,
      "learning_rate": 7.528205773317345e-06,
      "loss": 0.0079,
      "step": 16725
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.36763742566108704,
      "learning_rate": 7.522931120485643e-06,
      "loss": 0.005,
      "step": 16730
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.7494738101959229,
      "learning_rate": 7.517657201771768e-06,
      "loss": 0.0051,
      "step": 16735
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.41462403535842896,
      "learning_rate": 7.512384018738731e-06,
      "loss": 0.0089,
      "step": 16740
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.36195388436317444,
      "learning_rate": 7.507111572949323e-06,
      "loss": 0.0053,
      "step": 16745
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.48771071434020996,
      "learning_rate": 7.501839865966109e-06,
      "loss": 0.008,
      "step": 16750
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.4612463414669037,
      "learning_rate": 7.496568899351445e-06,
      "loss": 0.01,
      "step": 16755
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.19453437626361847,
      "learning_rate": 7.491298674667464e-06,
      "loss": 0.0054,
      "step": 16760
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.5014907121658325,
      "learning_rate": 7.486029193476075e-06,
      "loss": 0.0068,
      "step": 16765
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.7443044781684875,
      "learning_rate": 7.480760457338978e-06,
      "loss": 0.0096,
      "step": 16770
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.2169743776321411,
      "learning_rate": 7.47549246781764e-06,
      "loss": 0.0117,
      "step": 16775
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.5546299815177917,
      "learning_rate": 7.470225226473311e-06,
      "loss": 0.0072,
      "step": 16780
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.8305453658103943,
      "learning_rate": 7.464958734867025e-06,
      "loss": 0.0061,
      "step": 16785
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.44384169578552246,
      "learning_rate": 7.459692994559582e-06,
      "loss": 0.008,
      "step": 16790
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.44968289136886597,
      "learning_rate": 7.4544280071115684e-06,
      "loss": 0.0084,
      "step": 16795
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.5661505460739136,
      "learning_rate": 7.449163774083345e-06,
      "loss": 0.0059,
      "step": 16800
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.669439435005188,
      "learning_rate": 7.443900297035049e-06,
      "loss": 0.0097,
      "step": 16805
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.5095635056495667,
      "learning_rate": 7.4386375775265916e-06,
      "loss": 0.0069,
      "step": 16810
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.3544565439224243,
      "learning_rate": 7.433375617117666e-06,
      "loss": 0.0067,
      "step": 16815
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.48801660537719727,
      "learning_rate": 7.428114417367738e-06,
      "loss": 0.0079,
      "step": 16820
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.4711840748786926,
      "learning_rate": 7.422853979836035e-06,
      "loss": 0.0078,
      "step": 16825
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.43352481722831726,
      "learning_rate": 7.417594306081577e-06,
      "loss": 0.0075,
      "step": 16830
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.3749412000179291,
      "learning_rate": 7.412335397663149e-06,
      "loss": 0.0089,
      "step": 16835
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.17959199845790863,
      "learning_rate": 7.407077256139309e-06,
      "loss": 0.0073,
      "step": 16840
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.40305107831954956,
      "learning_rate": 7.4018198830683905e-06,
      "loss": 0.0074,
      "step": 16845
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.30933234095573425,
      "learning_rate": 7.396563280008494e-06,
      "loss": 0.0094,
      "step": 16850
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.267681360244751,
      "learning_rate": 7.391307448517506e-06,
      "loss": 0.0074,
      "step": 16855
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.33332905173301697,
      "learning_rate": 7.38605239015306e-06,
      "loss": 0.0093,
      "step": 16860
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.18610811233520508,
      "learning_rate": 7.380798106472582e-06,
      "loss": 0.0094,
      "step": 16865
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.5006966590881348,
      "learning_rate": 7.375544599033255e-06,
      "loss": 0.0074,
      "step": 16870
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.37795162200927734,
      "learning_rate": 7.370291869392045e-06,
      "loss": 0.0091,
      "step": 16875
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.5025229454040527,
      "learning_rate": 7.365039919105675e-06,
      "loss": 0.0085,
      "step": 16880
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.3851081430912018,
      "learning_rate": 7.3597887497306455e-06,
      "loss": 0.0092,
      "step": 16885
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.2460136115550995,
      "learning_rate": 7.35453836282322e-06,
      "loss": 0.0071,
      "step": 16890
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.3819717764854431,
      "learning_rate": 7.34928875993943e-06,
      "loss": 0.0067,
      "step": 16895
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.40334203839302063,
      "learning_rate": 7.34403994263508e-06,
      "loss": 0.0101,
      "step": 16900
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.22943516075611115,
      "learning_rate": 7.338791912465737e-06,
      "loss": 0.0094,
      "step": 16905
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.2664775848388672,
      "learning_rate": 7.333544670986734e-06,
      "loss": 0.0081,
      "step": 16910
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.3135198950767517,
      "learning_rate": 7.328298219753177e-06,
      "loss": 0.0081,
      "step": 16915
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.5245127081871033,
      "learning_rate": 7.323052560319932e-06,
      "loss": 0.0089,
      "step": 16920
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.3170236647129059,
      "learning_rate": 7.317807694241635e-06,
      "loss": 0.0057,
      "step": 16925
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.3144915997982025,
      "learning_rate": 7.3125636230726745e-06,
      "loss": 0.0076,
      "step": 16930
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.4207755923271179,
      "learning_rate": 7.307320348367219e-06,
      "loss": 0.0059,
      "step": 16935
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.6968371868133545,
      "learning_rate": 7.302077871679194e-06,
      "loss": 0.0104,
      "step": 16940
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.3698020577430725,
      "learning_rate": 7.296836194562287e-06,
      "loss": 0.0128,
      "step": 16945
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.5103458166122437,
      "learning_rate": 7.291595318569951e-06,
      "loss": 0.007,
      "step": 16950
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.2294590026140213,
      "learning_rate": 7.286355245255399e-06,
      "loss": 0.0065,
      "step": 16955
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.8034264445304871,
      "learning_rate": 7.281115976171612e-06,
      "loss": 0.0068,
      "step": 16960
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.5444835424423218,
      "learning_rate": 7.2758775128713325e-06,
      "loss": 0.006,
      "step": 16965
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.4064411520957947,
      "learning_rate": 7.270639856907047e-06,
      "loss": 0.0084,
      "step": 16970
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.17976729571819305,
      "learning_rate": 7.265403009831027e-06,
      "loss": 0.0062,
      "step": 16975
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.5045095086097717,
      "learning_rate": 7.260166973195292e-06,
      "loss": 0.0076,
      "step": 16980
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.32540497183799744,
      "learning_rate": 7.254931748551619e-06,
      "loss": 0.0069,
      "step": 16985
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.5425432920455933,
      "learning_rate": 7.2496973374515535e-06,
      "loss": 0.0081,
      "step": 16990
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.34949442744255066,
      "learning_rate": 7.244463741446387e-06,
      "loss": 0.0074,
      "step": 16995
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.37236133217811584,
      "learning_rate": 7.239230962087189e-06,
      "loss": 0.0093,
      "step": 17000
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.3425101339817047,
      "learning_rate": 7.233999000924763e-06,
      "loss": 0.0067,
      "step": 17005
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.5776107311248779,
      "learning_rate": 7.2287678595096865e-06,
      "loss": 0.0098,
      "step": 17010
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.39168134331703186,
      "learning_rate": 7.223537539392287e-06,
      "loss": 0.0114,
      "step": 17015
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.5078431963920593,
      "learning_rate": 7.218308042122655e-06,
      "loss": 0.0099,
      "step": 17020
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.19572730362415314,
      "learning_rate": 7.213079369250632e-06,
      "loss": 0.006,
      "step": 17025
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.2532220780849457,
      "learning_rate": 7.2078515223258165e-06,
      "loss": 0.0066,
      "step": 17030
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.38014352321624756,
      "learning_rate": 7.202624502897564e-06,
      "loss": 0.0078,
      "step": 17035
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.39346054196357727,
      "learning_rate": 7.197398312514978e-06,
      "loss": 0.0074,
      "step": 17040
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.3673287630081177,
      "learning_rate": 7.192172952726923e-06,
      "loss": 0.0074,
      "step": 17045
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.42266541719436646,
      "learning_rate": 7.186948425082017e-06,
      "loss": 0.0072,
      "step": 17050
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.7222540378570557,
      "learning_rate": 7.181724731128627e-06,
      "loss": 0.0065,
      "step": 17055
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.35174280405044556,
      "learning_rate": 7.176501872414881e-06,
      "loss": 0.0091,
      "step": 17060
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.591694712638855,
      "learning_rate": 7.1712798504886506e-06,
      "loss": 0.0071,
      "step": 17065
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.6140367984771729,
      "learning_rate": 7.166058666897565e-06,
      "loss": 0.0107,
      "step": 17070
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.2917439341545105,
      "learning_rate": 7.160838323189001e-06,
      "loss": 0.0056,
      "step": 17075
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.4179360866546631,
      "learning_rate": 7.1556188209100884e-06,
      "loss": 0.0087,
      "step": 17080
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.5467280149459839,
      "learning_rate": 7.150400161607709e-06,
      "loss": 0.0091,
      "step": 17085
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.7273814082145691,
      "learning_rate": 7.145182346828492e-06,
      "loss": 0.0076,
      "step": 17090
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.42867380380630493,
      "learning_rate": 7.139965378118821e-06,
      "loss": 0.0068,
      "step": 17095
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.2701338529586792,
      "learning_rate": 7.13474925702482e-06,
      "loss": 0.0088,
      "step": 17100
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.5208498239517212,
      "learning_rate": 7.129533985092378e-06,
      "loss": 0.0063,
      "step": 17105
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.47218087315559387,
      "learning_rate": 7.124319563867112e-06,
      "loss": 0.0056,
      "step": 17110
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.6032540202140808,
      "learning_rate": 7.119105994894398e-06,
      "loss": 0.0091,
      "step": 17115
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.2994047701358795,
      "learning_rate": 7.113893279719359e-06,
      "loss": 0.0071,
      "step": 17120
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.4658941924571991,
      "learning_rate": 7.108681419886867e-06,
      "loss": 0.0075,
      "step": 17125
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.17439764738082886,
      "learning_rate": 7.103470416941536e-06,
      "loss": 0.0064,
      "step": 17130
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.3298243582248688,
      "learning_rate": 7.098260272427726e-06,
      "loss": 0.0063,
      "step": 17135
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.5897051095962524,
      "learning_rate": 7.093050987889547e-06,
      "loss": 0.0052,
      "step": 17140
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.3087965250015259,
      "learning_rate": 7.087842564870851e-06,
      "loss": 0.0057,
      "step": 17145
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.19054736196994781,
      "learning_rate": 7.08263500491523e-06,
      "loss": 0.0102,
      "step": 17150
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.3062085211277008,
      "learning_rate": 7.077428309566033e-06,
      "loss": 0.0068,
      "step": 17155
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.3566223382949829,
      "learning_rate": 7.072222480366335e-06,
      "loss": 0.0059,
      "step": 17160
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.33739137649536133,
      "learning_rate": 7.067017518858975e-06,
      "loss": 0.0079,
      "step": 17165
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.3130680322647095,
      "learning_rate": 7.061813426586519e-06,
      "loss": 0.0075,
      "step": 17170
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.15110720694065094,
      "learning_rate": 7.056610205091281e-06,
      "loss": 0.0073,
      "step": 17175
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.5237319469451904,
      "learning_rate": 7.05140785591532e-06,
      "loss": 0.0064,
      "step": 17180
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.5438427925109863,
      "learning_rate": 7.0462063806004255e-06,
      "loss": 0.0072,
      "step": 17185
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.3356355130672455,
      "learning_rate": 7.041005780688142e-06,
      "loss": 0.006,
      "step": 17190
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.338467538356781,
      "learning_rate": 7.035806057719743e-06,
      "loss": 0.0068,
      "step": 17195
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.17586380243301392,
      "learning_rate": 7.030607213236249e-06,
      "loss": 0.0064,
      "step": 17200
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.47041377425193787,
      "learning_rate": 7.025409248778421e-06,
      "loss": 0.0071,
      "step": 17205
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.37697193026542664,
      "learning_rate": 7.020212165886755e-06,
      "loss": 0.0067,
      "step": 17210
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.4052976965904236,
      "learning_rate": 7.015015966101488e-06,
      "loss": 0.0092,
      "step": 17215
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.5404878258705139,
      "learning_rate": 7.009820650962593e-06,
      "loss": 0.0097,
      "step": 17220
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.3760913908481598,
      "learning_rate": 7.004626222009782e-06,
      "loss": 0.0066,
      "step": 17225
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.5538203120231628,
      "learning_rate": 6.9994326807825045e-06,
      "loss": 0.0085,
      "step": 17230
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.5948095917701721,
      "learning_rate": 6.994240028819947e-06,
      "loss": 0.0079,
      "step": 17235
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.5411348938941956,
      "learning_rate": 6.989048267661036e-06,
      "loss": 0.0085,
      "step": 17240
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.4167710244655609,
      "learning_rate": 6.983857398844425e-06,
      "loss": 0.0067,
      "step": 17245
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.5900302529335022,
      "learning_rate": 6.978667423908517e-06,
      "loss": 0.0066,
      "step": 17250
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.3182935118675232,
      "learning_rate": 6.973478344391433e-06,
      "loss": 0.0097,
      "step": 17255
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.4234834313392639,
      "learning_rate": 6.9682901618310385e-06,
      "loss": 0.0087,
      "step": 17260
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.34514597058296204,
      "learning_rate": 6.963102877764936e-06,
      "loss": 0.0066,
      "step": 17265
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.3347316086292267,
      "learning_rate": 6.9579164937304544e-06,
      "loss": 0.0072,
      "step": 17270
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.364444375038147,
      "learning_rate": 6.952731011264662e-06,
      "loss": 0.0048,
      "step": 17275
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.3441850543022156,
      "learning_rate": 6.9475464319043554e-06,
      "loss": 0.0068,
      "step": 17280
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.5691609978675842,
      "learning_rate": 6.942362757186065e-06,
      "loss": 0.0094,
      "step": 17285
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.5621780157089233,
      "learning_rate": 6.937179988646056e-06,
      "loss": 0.0072,
      "step": 17290
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.4421490430831909,
      "learning_rate": 6.931998127820316e-06,
      "loss": 0.0089,
      "step": 17295
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.5314481258392334,
      "learning_rate": 6.926817176244576e-06,
      "loss": 0.0112,
      "step": 17300
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.2780129313468933,
      "learning_rate": 6.921637135454286e-06,
      "loss": 0.0075,
      "step": 17305
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.5449675917625427,
      "learning_rate": 6.916458006984639e-06,
      "loss": 0.0063,
      "step": 17310
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.5825093388557434,
      "learning_rate": 6.9112797923705445e-06,
      "loss": 0.009,
      "step": 17315
    },
    {
      "epoch": 5.4,
      "grad_norm": 1.126198410987854,
      "learning_rate": 6.906102493146648e-06,
      "loss": 0.0096,
      "step": 17320
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.5498197078704834,
      "learning_rate": 6.9009261108473255e-06,
      "loss": 0.0053,
      "step": 17325
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.38959771394729614,
      "learning_rate": 6.895750647006674e-06,
      "loss": 0.0081,
      "step": 17330
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.4668501913547516,
      "learning_rate": 6.890576103158524e-06,
      "loss": 0.0076,
      "step": 17335
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.4254097640514374,
      "learning_rate": 6.885402480836431e-06,
      "loss": 0.008,
      "step": 17340
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.38289356231689453,
      "learning_rate": 6.880229781573678e-06,
      "loss": 0.0083,
      "step": 17345
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.3994235694408417,
      "learning_rate": 6.875058006903279e-06,
      "loss": 0.0076,
      "step": 17350
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.33938008546829224,
      "learning_rate": 6.8698871583579666e-06,
      "loss": 0.008,
      "step": 17355
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.5851308703422546,
      "learning_rate": 6.864717237470205e-06,
      "loss": 0.0071,
      "step": 17360
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.5873516798019409,
      "learning_rate": 6.859548245772174e-06,
      "loss": 0.0088,
      "step": 17365
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.1811552196741104,
      "learning_rate": 6.854380184795789e-06,
      "loss": 0.0054,
      "step": 17370
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.8535553812980652,
      "learning_rate": 6.849213056072685e-06,
      "loss": 0.0057,
      "step": 17375
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.2805171608924866,
      "learning_rate": 6.844046861134221e-06,
      "loss": 0.0073,
      "step": 17380
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.388770192861557,
      "learning_rate": 6.838881601511478e-06,
      "loss": 0.006,
      "step": 17385
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.7432388663291931,
      "learning_rate": 6.83371727873526e-06,
      "loss": 0.008,
      "step": 17390
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.41411373019218445,
      "learning_rate": 6.828553894336102e-06,
      "loss": 0.0083,
      "step": 17395
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.3691873550415039,
      "learning_rate": 6.823391449844242e-06,
      "loss": 0.0092,
      "step": 17400
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.21740001440048218,
      "learning_rate": 6.818229946789654e-06,
      "loss": 0.0069,
      "step": 17405
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.20824594795703888,
      "learning_rate": 6.8130693867020325e-06,
      "loss": 0.0049,
      "step": 17410
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.1659618467092514,
      "learning_rate": 6.807909771110788e-06,
      "loss": 0.0071,
      "step": 17415
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.71225905418396,
      "learning_rate": 6.802751101545053e-06,
      "loss": 0.0084,
      "step": 17420
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.3749540448188782,
      "learning_rate": 6.7975933795336805e-06,
      "loss": 0.0067,
      "step": 17425
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.45130300521850586,
      "learning_rate": 6.792436606605242e-06,
      "loss": 0.006,
      "step": 17430
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.4924355745315552,
      "learning_rate": 6.787280784288022e-06,
      "loss": 0.0095,
      "step": 17435
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.5896020531654358,
      "learning_rate": 6.7821259141100326e-06,
      "loss": 0.009,
      "step": 17440
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.43783894181251526,
      "learning_rate": 6.776971997598999e-06,
      "loss": 0.0106,
      "step": 17445
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.5060574412345886,
      "learning_rate": 6.771819036282362e-06,
      "loss": 0.0075,
      "step": 17450
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.4454652667045593,
      "learning_rate": 6.766667031687286e-06,
      "loss": 0.0075,
      "step": 17455
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.17911115288734436,
      "learning_rate": 6.761515985340647e-06,
      "loss": 0.0086,
      "step": 17460
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.4743806719779968,
      "learning_rate": 6.756365898769035e-06,
      "loss": 0.0078,
      "step": 17465
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.8190107941627502,
      "learning_rate": 6.751216773498762e-06,
      "loss": 0.0089,
      "step": 17470
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.19955362379550934,
      "learning_rate": 6.746068611055847e-06,
      "loss": 0.0078,
      "step": 17475
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.265946626663208,
      "learning_rate": 6.740921412966029e-06,
      "loss": 0.0067,
      "step": 17480
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.22414854168891907,
      "learning_rate": 6.73577518075476e-06,
      "loss": 0.0084,
      "step": 17485
    },
    {
      "epoch": 5.46,
      "grad_norm": 1.177064299583435,
      "learning_rate": 6.7306299159472045e-06,
      "loss": 0.0066,
      "step": 17490
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.3832259476184845,
      "learning_rate": 6.7254856200682465e-06,
      "loss": 0.0085,
      "step": 17495
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.281180202960968,
      "learning_rate": 6.720342294642474e-06,
      "loss": 0.0067,
      "step": 17500
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.3122906982898712,
      "learning_rate": 6.715199941194196e-06,
      "loss": 0.0092,
      "step": 17505
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.33859917521476746,
      "learning_rate": 6.71005856124742e-06,
      "loss": 0.0065,
      "step": 17510
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.3163743019104004,
      "learning_rate": 6.70491815632588e-06,
      "loss": 0.0075,
      "step": 17515
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.39969271421432495,
      "learning_rate": 6.699778727953015e-06,
      "loss": 0.0059,
      "step": 17520
    },
    {
      "epoch": 5.47,
      "grad_norm": 0.31698331236839294,
      "learning_rate": 6.694640277651973e-06,
      "loss": 0.0094,
      "step": 17525
    },
    {
      "epoch": 5.47,
      "grad_norm": 0.393361359834671,
      "learning_rate": 6.6895028069456135e-06,
      "loss": 0.008,
      "step": 17530
    },
    {
      "epoch": 5.47,
      "grad_norm": 0.3493746817111969,
      "learning_rate": 6.684366317356503e-06,
      "loss": 0.0056,
      "step": 17535
    },
    {
      "epoch": 5.47,
      "grad_norm": 0.25108638405799866,
      "learning_rate": 6.67923081040693e-06,
      "loss": 0.0044,
      "step": 17540
    },
    {
      "epoch": 5.47,
      "grad_norm": 0.5176618695259094,
      "learning_rate": 6.674096287618869e-06,
      "loss": 0.0055,
      "step": 17545
    },
    {
      "epoch": 5.47,
      "grad_norm": 0.3980843424797058,
      "learning_rate": 6.668962750514022e-06,
      "loss": 0.0088,
      "step": 17550
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.37212374806404114,
      "learning_rate": 6.663830200613792e-06,
      "loss": 0.0067,
      "step": 17555
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.334362655878067,
      "learning_rate": 6.658698639439289e-06,
      "loss": 0.0099,
      "step": 17560
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.6859321594238281,
      "learning_rate": 6.6535680685113325e-06,
      "loss": 0.0096,
      "step": 17565
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.7242924571037292,
      "learning_rate": 6.648438489350445e-06,
      "loss": 0.0092,
      "step": 17570
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.35120564699172974,
      "learning_rate": 6.64330990347686e-06,
      "loss": 0.0072,
      "step": 17575
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.27487900853157043,
      "learning_rate": 6.638182312410509e-06,
      "loss": 0.0047,
      "step": 17580
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.8260858654975891,
      "learning_rate": 6.633055717671035e-06,
      "loss": 0.0079,
      "step": 17585
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.2853456139564514,
      "learning_rate": 6.627930120777786e-06,
      "loss": 0.0062,
      "step": 17590
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.4730463922023773,
      "learning_rate": 6.622805523249807e-06,
      "loss": 0.0071,
      "step": 17595
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.4280153214931488,
      "learning_rate": 6.617681926605859e-06,
      "loss": 0.0101,
      "step": 17600
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.4409542679786682,
      "learning_rate": 6.612559332364395e-06,
      "loss": 0.0078,
      "step": 17605
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.47959208488464355,
      "learning_rate": 6.607437742043577e-06,
      "loss": 0.0087,
      "step": 17610
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.32082483172416687,
      "learning_rate": 6.602317157161269e-06,
      "loss": 0.0072,
      "step": 17615
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.4335188865661621,
      "learning_rate": 6.597197579235033e-06,
      "loss": 0.0092,
      "step": 17620
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.3355504870414734,
      "learning_rate": 6.592079009782135e-06,
      "loss": 0.0067,
      "step": 17625
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.4183447062969208,
      "learning_rate": 6.586961450319544e-06,
      "loss": 0.0061,
      "step": 17630
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.4092869162559509,
      "learning_rate": 6.58184490236393e-06,
      "loss": 0.0084,
      "step": 17635
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.7920547127723694,
      "learning_rate": 6.576729367431658e-06,
      "loss": 0.0067,
      "step": 17640
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.35771897435188293,
      "learning_rate": 6.571614847038798e-06,
      "loss": 0.0113,
      "step": 17645
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.46013301610946655,
      "learning_rate": 6.566501342701124e-06,
      "loss": 0.009,
      "step": 17650
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.39704209566116333,
      "learning_rate": 6.561388855934088e-06,
      "loss": 0.006,
      "step": 17655
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.4999745190143585,
      "learning_rate": 6.5562773882528655e-06,
      "loss": 0.0085,
      "step": 17660
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.4186748266220093,
      "learning_rate": 6.5511669411723165e-06,
      "loss": 0.006,
      "step": 17665
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.5002567172050476,
      "learning_rate": 6.546057516207002e-06,
      "loss": 0.0109,
      "step": 17670
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.5059384107589722,
      "learning_rate": 6.540949114871178e-06,
      "loss": 0.01,
      "step": 17675
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.36181607842445374,
      "learning_rate": 6.535841738678798e-06,
      "loss": 0.0081,
      "step": 17680
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.4996640086174011,
      "learning_rate": 6.53073538914352e-06,
      "loss": 0.0096,
      "step": 17685
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.430021196603775,
      "learning_rate": 6.525630067778679e-06,
      "loss": 0.0076,
      "step": 17690
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.5806915163993835,
      "learning_rate": 6.52052577609732e-06,
      "loss": 0.0095,
      "step": 17695
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.16090747714042664,
      "learning_rate": 6.515422515612181e-06,
      "loss": 0.0054,
      "step": 17700
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.35832875967025757,
      "learning_rate": 6.510320287835692e-06,
      "loss": 0.0078,
      "step": 17705
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.5224231481552124,
      "learning_rate": 6.5052190942799774e-06,
      "loss": 0.0077,
      "step": 17710
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.6822590827941895,
      "learning_rate": 6.500118936456854e-06,
      "loss": 0.0063,
      "step": 17715
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.22760039567947388,
      "learning_rate": 6.495019815877836e-06,
      "loss": 0.0086,
      "step": 17720
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.47752487659454346,
      "learning_rate": 6.489921734054122e-06,
      "loss": 0.0079,
      "step": 17725
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.44208142161369324,
      "learning_rate": 6.484824692496611e-06,
      "loss": 0.0086,
      "step": 17730
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.3971075713634491,
      "learning_rate": 6.479728692715889e-06,
      "loss": 0.0091,
      "step": 17735
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.48173531889915466,
      "learning_rate": 6.474633736222234e-06,
      "loss": 0.0104,
      "step": 17740
    },
    {
      "epoch": 5.53,
      "grad_norm": 0.3530065417289734,
      "learning_rate": 6.469539824525617e-06,
      "loss": 0.0077,
      "step": 17745
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.6809261441230774,
      "learning_rate": 6.4644469591357e-06,
      "loss": 0.0073,
      "step": 17750
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.2195349484682083,
      "learning_rate": 6.45935514156183e-06,
      "loss": 0.006,
      "step": 17755
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.3710613548755646,
      "learning_rate": 6.454264373313046e-06,
      "loss": 0.0092,
      "step": 17760
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.37743639945983887,
      "learning_rate": 6.449174655898076e-06,
      "loss": 0.0073,
      "step": 17765
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.28303930163383484,
      "learning_rate": 6.444085990825338e-06,
      "loss": 0.0072,
      "step": 17770
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.5262421369552612,
      "learning_rate": 6.4389983796029365e-06,
      "loss": 0.0106,
      "step": 17775
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.4436008036136627,
      "learning_rate": 6.433911823738663e-06,
      "loss": 0.0103,
      "step": 17780
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.4982847571372986,
      "learning_rate": 6.428826324739999e-06,
      "loss": 0.0065,
      "step": 17785
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.1745348423719406,
      "learning_rate": 6.42374188411411e-06,
      "loss": 0.0067,
      "step": 17790
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.8984200358390808,
      "learning_rate": 6.418658503367854e-06,
      "loss": 0.009,
      "step": 17795
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.415552020072937,
      "learning_rate": 6.41357618400776e-06,
      "loss": 0.0089,
      "step": 17800
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.3550618886947632,
      "learning_rate": 6.4084949275400585e-06,
      "loss": 0.0062,
      "step": 17805
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.32670262455940247,
      "learning_rate": 6.403414735470659e-06,
      "loss": 0.0093,
      "step": 17810
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.18545816838741302,
      "learning_rate": 6.398335609305153e-06,
      "loss": 0.0064,
      "step": 17815
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.3328353464603424,
      "learning_rate": 6.39325755054882e-06,
      "loss": 0.008,
      "step": 17820
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.2967495620250702,
      "learning_rate": 6.388180560706619e-06,
      "loss": 0.0054,
      "step": 17825
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.2916224002838135,
      "learning_rate": 6.383104641283203e-06,
      "loss": 0.0063,
      "step": 17830
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.4712311923503876,
      "learning_rate": 6.378029793782889e-06,
      "loss": 0.0073,
      "step": 17835
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.712516725063324,
      "learning_rate": 6.37295601970969e-06,
      "loss": 0.0078,
      "step": 17840
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.32948067784309387,
      "learning_rate": 6.3678833205673004e-06,
      "loss": 0.0076,
      "step": 17845
    },
    {
      "epoch": 5.57,
      "grad_norm": 1.3793821334838867,
      "learning_rate": 6.362811697859093e-06,
      "loss": 0.0141,
      "step": 17850
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.2590395212173462,
      "learning_rate": 6.357741153088119e-06,
      "loss": 0.007,
      "step": 17855
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.37518686056137085,
      "learning_rate": 6.352671687757116e-06,
      "loss": 0.0074,
      "step": 17860
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.4013412594795227,
      "learning_rate": 6.347603303368499e-06,
      "loss": 0.008,
      "step": 17865
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.5398988723754883,
      "learning_rate": 6.342536001424361e-06,
      "loss": 0.0092,
      "step": 17870
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.4342360496520996,
      "learning_rate": 6.337469783426474e-06,
      "loss": 0.0111,
      "step": 17875
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.3777037560939789,
      "learning_rate": 6.332404650876294e-06,
      "loss": 0.0052,
      "step": 17880
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.3465896248817444,
      "learning_rate": 6.327340605274945e-06,
      "loss": 0.0076,
      "step": 17885
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.43914374709129333,
      "learning_rate": 6.322277648123245e-06,
      "loss": 0.0079,
      "step": 17890
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.1666235476732254,
      "learning_rate": 6.317215780921675e-06,
      "loss": 0.0077,
      "step": 17895
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.3394373059272766,
      "learning_rate": 6.3121550051704e-06,
      "loss": 0.0067,
      "step": 17900
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.37623217701911926,
      "learning_rate": 6.3070953223692554e-06,
      "loss": 0.01,
      "step": 17905
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.49671798944473267,
      "learning_rate": 6.30203673401776e-06,
      "loss": 0.01,
      "step": 17910
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.6104849576950073,
      "learning_rate": 6.296979241615105e-06,
      "loss": 0.0113,
      "step": 17915
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.4916791617870331,
      "learning_rate": 6.291922846660154e-06,
      "loss": 0.0092,
      "step": 17920
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.32638782262802124,
      "learning_rate": 6.28686755065145e-06,
      "loss": 0.0064,
      "step": 17925
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.44251349568367004,
      "learning_rate": 6.281813355087207e-06,
      "loss": 0.0058,
      "step": 17930
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.35004180669784546,
      "learning_rate": 6.276760261465325e-06,
      "loss": 0.0108,
      "step": 17935
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.526656985282898,
      "learning_rate": 6.27170827128335e-06,
      "loss": 0.007,
      "step": 17940
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.928345799446106,
      "learning_rate": 6.266657386038526e-06,
      "loss": 0.0077,
      "step": 17945
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.5705955624580383,
      "learning_rate": 6.261607607227761e-06,
      "loss": 0.0095,
      "step": 17950
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.5848002433776855,
      "learning_rate": 6.256558936347636e-06,
      "loss": 0.0098,
      "step": 17955
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.2608962655067444,
      "learning_rate": 6.251511374894403e-06,
      "loss": 0.0077,
      "step": 17960
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.2794182300567627,
      "learning_rate": 6.246464924363983e-06,
      "loss": 0.0049,
      "step": 17965
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.4642208218574524,
      "learning_rate": 6.24141958625197e-06,
      "loss": 0.0089,
      "step": 17970
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.5016393661499023,
      "learning_rate": 6.236375362053635e-06,
      "loss": 0.007,
      "step": 17975
    },
    {
      "epoch": 5.61,
      "grad_norm": 1.0015122890472412,
      "learning_rate": 6.231332253263904e-06,
      "loss": 0.0093,
      "step": 17980
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.5549787282943726,
      "learning_rate": 6.226290261377381e-06,
      "loss": 0.0065,
      "step": 17985
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.456961452960968,
      "learning_rate": 6.221249387888343e-06,
      "loss": 0.0084,
      "step": 17990
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.4469737708568573,
      "learning_rate": 6.216209634290729e-06,
      "loss": 0.0056,
      "step": 17995
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.3027060329914093,
      "learning_rate": 6.211171002078149e-06,
      "loss": 0.0042,
      "step": 18000
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.2818998396396637,
      "learning_rate": 6.2061334927438784e-06,
      "loss": 0.0073,
      "step": 18005
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.31544992327690125,
      "learning_rate": 6.201097107780865e-06,
      "loss": 0.0076,
      "step": 18010
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.2734914720058441,
      "learning_rate": 6.1960618486817145e-06,
      "loss": 0.0065,
      "step": 18015
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.21664875745773315,
      "learning_rate": 6.191027716938706e-06,
      "loss": 0.0061,
      "step": 18020
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.5399858951568604,
      "learning_rate": 6.185994714043786e-06,
      "loss": 0.0076,
      "step": 18025
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.5281535387039185,
      "learning_rate": 6.180962841488556e-06,
      "loss": 0.0101,
      "step": 18030
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.4660307765007019,
      "learning_rate": 6.175932100764296e-06,
      "loss": 0.0087,
      "step": 18035
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.42953482270240784,
      "learning_rate": 6.170902493361941e-06,
      "loss": 0.0063,
      "step": 18040
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.21518750488758087,
      "learning_rate": 6.1658740207720975e-06,
      "loss": 0.009,
      "step": 18045
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.35465386509895325,
      "learning_rate": 6.160846684485027e-06,
      "loss": 0.0077,
      "step": 18050
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.8630399107933044,
      "learning_rate": 6.155820485990658e-06,
      "loss": 0.0074,
      "step": 18055
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.34728437662124634,
      "learning_rate": 6.1507954267785865e-06,
      "loss": 0.006,
      "step": 18060
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.7103444933891296,
      "learning_rate": 6.145771508338063e-06,
      "loss": 0.0077,
      "step": 18065
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.4810100793838501,
      "learning_rate": 6.140748732158005e-06,
      "loss": 0.0091,
      "step": 18070
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.2638356685638428,
      "learning_rate": 6.1357270997269894e-06,
      "loss": 0.0059,
      "step": 18075
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.2502276599407196,
      "learning_rate": 6.130706612533262e-06,
      "loss": 0.0073,
      "step": 18080
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.4023531675338745,
      "learning_rate": 6.125687272064713e-06,
      "loss": 0.0055,
      "step": 18085
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.4165780544281006,
      "learning_rate": 6.120669079808902e-06,
      "loss": 0.0063,
      "step": 18090
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.3536226451396942,
      "learning_rate": 6.115652037253054e-06,
      "loss": 0.0084,
      "step": 18095
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.2316283881664276,
      "learning_rate": 6.110636145884045e-06,
      "loss": 0.0072,
      "step": 18100
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.09715107083320618,
      "learning_rate": 6.105621407188411e-06,
      "loss": 0.0048,
      "step": 18105
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.6509740948677063,
      "learning_rate": 6.100607822652351e-06,
      "loss": 0.0071,
      "step": 18110
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.48909834027290344,
      "learning_rate": 6.095595393761713e-06,
      "loss": 0.0058,
      "step": 18115
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.44172388315200806,
      "learning_rate": 6.090584122002018e-06,
      "loss": 0.0085,
      "step": 18120
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.5661274790763855,
      "learning_rate": 6.085574008858423e-06,
      "loss": 0.006,
      "step": 18125
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.5164355635643005,
      "learning_rate": 6.0805650558157595e-06,
      "loss": 0.0064,
      "step": 18130
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.35191330313682556,
      "learning_rate": 6.075557264358504e-06,
      "loss": 0.0055,
      "step": 18135
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.46775275468826294,
      "learning_rate": 6.070550635970798e-06,
      "loss": 0.0096,
      "step": 18140
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.3156464993953705,
      "learning_rate": 6.065545172136431e-06,
      "loss": 0.0073,
      "step": 18145
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.27590519189834595,
      "learning_rate": 6.060540874338849e-06,
      "loss": 0.0068,
      "step": 18150
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.28529655933380127,
      "learning_rate": 6.05553774406116e-06,
      "loss": 0.0088,
      "step": 18155
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.48952776193618774,
      "learning_rate": 6.050535782786111e-06,
      "loss": 0.007,
      "step": 18160
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.587226390838623,
      "learning_rate": 6.045534991996114e-06,
      "loss": 0.0072,
      "step": 18165
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.2564910650253296,
      "learning_rate": 6.0405353731732296e-06,
      "loss": 0.0084,
      "step": 18170
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.5341393351554871,
      "learning_rate": 6.035536927799173e-06,
      "loss": 0.0096,
      "step": 18175
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.4170636832714081,
      "learning_rate": 6.0305396573553145e-06,
      "loss": 0.0063,
      "step": 18180
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.5394135117530823,
      "learning_rate": 6.025543563322671e-06,
      "loss": 0.0093,
      "step": 18185
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.14204931259155273,
      "learning_rate": 6.0205486471819115e-06,
      "loss": 0.0068,
      "step": 18190
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.34632179141044617,
      "learning_rate": 6.015554910413358e-06,
      "loss": 0.0086,
      "step": 18195
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.3683947026729584,
      "learning_rate": 6.010562354496981e-06,
      "loss": 0.0065,
      "step": 18200
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.47035160660743713,
      "learning_rate": 6.005570980912401e-06,
      "loss": 0.0046,
      "step": 18205
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.4578624665737152,
      "learning_rate": 6.000580791138891e-06,
      "loss": 0.0073,
      "step": 18210
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.28652694821357727,
      "learning_rate": 5.99559178665537e-06,
      "loss": 0.0085,
      "step": 18215
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.5046875476837158,
      "learning_rate": 5.990603968940406e-06,
      "loss": 0.0065,
      "step": 18220
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.45822709798812866,
      "learning_rate": 5.985617339472225e-06,
      "loss": 0.0067,
      "step": 18225
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.4730364680290222,
      "learning_rate": 5.980631899728679e-06,
      "loss": 0.0116,
      "step": 18230
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.6081570386886597,
      "learning_rate": 5.975647651187285e-06,
      "loss": 0.0111,
      "step": 18235
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.47355684638023376,
      "learning_rate": 5.970664595325206e-06,
      "loss": 0.0071,
      "step": 18240
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.4323113262653351,
      "learning_rate": 5.965682733619246e-06,
      "loss": 0.0056,
      "step": 18245
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.5472829937934875,
      "learning_rate": 5.960702067545856e-06,
      "loss": 0.0084,
      "step": 18250
    },
    {
      "epoch": 5.69,
      "grad_norm": 0.5550527572631836,
      "learning_rate": 5.955722598581135e-06,
      "loss": 0.0122,
      "step": 18255
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.40815654397010803,
      "learning_rate": 5.9507443282008285e-06,
      "loss": 0.0065,
      "step": 18260
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.4066593647003174,
      "learning_rate": 5.94576725788032e-06,
      "loss": 0.0071,
      "step": 18265
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.3816416263580322,
      "learning_rate": 5.94079138909464e-06,
      "loss": 0.0064,
      "step": 18270
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.5269685983657837,
      "learning_rate": 5.93581672331847e-06,
      "loss": 0.006,
      "step": 18275
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.7609176635742188,
      "learning_rate": 5.930843262026122e-06,
      "loss": 0.0066,
      "step": 18280
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.6766870617866516,
      "learning_rate": 5.925871006691568e-06,
      "loss": 0.0062,
      "step": 18285
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.4741158187389374,
      "learning_rate": 5.920899958788408e-06,
      "loss": 0.0094,
      "step": 18290
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.21361322700977325,
      "learning_rate": 5.915930119789889e-06,
      "loss": 0.007,
      "step": 18295
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.5186450481414795,
      "learning_rate": 5.910961491168903e-06,
      "loss": 0.0086,
      "step": 18300
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.23191529512405396,
      "learning_rate": 5.905994074397975e-06,
      "loss": 0.0104,
      "step": 18305
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.2721780240535736,
      "learning_rate": 5.9010278709492796e-06,
      "loss": 0.0087,
      "step": 18310
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.6018416285514832,
      "learning_rate": 5.896062882294628e-06,
      "loss": 0.0071,
      "step": 18315
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.20528219640254974,
      "learning_rate": 5.891099109905468e-06,
      "loss": 0.0077,
      "step": 18320
    },
    {
      "epoch": 5.72,
      "grad_norm": 0.41088053584098816,
      "learning_rate": 5.886136555252897e-06,
      "loss": 0.0089,
      "step": 18325
    },
    {
      "epoch": 5.72,
      "grad_norm": 0.29491591453552246,
      "learning_rate": 5.881175219807642e-06,
      "loss": 0.0074,
      "step": 18330
    },
    {
      "epoch": 5.72,
      "grad_norm": 0.9241437911987305,
      "learning_rate": 5.876215105040074e-06,
      "loss": 0.0081,
      "step": 18335
    },
    {
      "epoch": 5.72,
      "grad_norm": 0.5419356822967529,
      "learning_rate": 5.871256212420195e-06,
      "loss": 0.0097,
      "step": 18340
    },
    {
      "epoch": 5.72,
      "grad_norm": 0.8657314777374268,
      "learning_rate": 5.8662985434176545e-06,
      "loss": 0.0096,
      "step": 18345
    },
    {
      "epoch": 5.72,
      "grad_norm": 0.5296528339385986,
      "learning_rate": 5.861342099501732e-06,
      "loss": 0.0066,
      "step": 18350
    },
    {
      "epoch": 5.72,
      "grad_norm": 0.3924877643585205,
      "learning_rate": 5.856386882141349e-06,
      "loss": 0.0068,
      "step": 18355
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.5159767866134644,
      "learning_rate": 5.851432892805058e-06,
      "loss": 0.0042,
      "step": 18360
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.39266231656074524,
      "learning_rate": 5.84648013296105e-06,
      "loss": 0.0093,
      "step": 18365
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.7416058778762817,
      "learning_rate": 5.841528604077162e-06,
      "loss": 0.0075,
      "step": 18370
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.6685163378715515,
      "learning_rate": 5.836578307620845e-06,
      "loss": 0.0079,
      "step": 18375
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.9518123269081116,
      "learning_rate": 5.8316292450592e-06,
      "loss": 0.0093,
      "step": 18380
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.47857388854026794,
      "learning_rate": 5.8266814178589595e-06,
      "loss": 0.0066,
      "step": 18385
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.5052703022956848,
      "learning_rate": 5.821734827486487e-06,
      "loss": 0.0093,
      "step": 18390
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.35116446018218994,
      "learning_rate": 5.81678947540778e-06,
      "loss": 0.0082,
      "step": 18395
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.3625754117965698,
      "learning_rate": 5.811845363088477e-06,
      "loss": 0.0069,
      "step": 18400
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.6056107878684998,
      "learning_rate": 5.8069024919938425e-06,
      "loss": 0.0099,
      "step": 18405
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.3495260179042816,
      "learning_rate": 5.801960863588765e-06,
      "loss": 0.0073,
      "step": 18410
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.4010307490825653,
      "learning_rate": 5.797020479337777e-06,
      "loss": 0.0067,
      "step": 18415
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.16561423242092133,
      "learning_rate": 5.792081340705038e-06,
      "loss": 0.0069,
      "step": 18420
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.4529399871826172,
      "learning_rate": 5.78714344915434e-06,
      "loss": 0.008,
      "step": 18425
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.3863770067691803,
      "learning_rate": 5.7822068061491e-06,
      "loss": 0.0045,
      "step": 18430
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.8393487930297852,
      "learning_rate": 5.777271413152369e-06,
      "loss": 0.0086,
      "step": 18435
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.37971916794776917,
      "learning_rate": 5.772337271626833e-06,
      "loss": 0.0088,
      "step": 18440
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.46120578050613403,
      "learning_rate": 5.7674043830348016e-06,
      "loss": 0.0077,
      "step": 18445
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.18467526137828827,
      "learning_rate": 5.762472748838206e-06,
      "loss": 0.0055,
      "step": 18450
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.36761900782585144,
      "learning_rate": 5.757542370498614e-06,
      "loss": 0.007,
      "step": 18455
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.6084608435630798,
      "learning_rate": 5.752613249477223e-06,
      "loss": 0.0077,
      "step": 18460
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.4849078357219696,
      "learning_rate": 5.747685387234855e-06,
      "loss": 0.0076,
      "step": 18465
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.44836604595184326,
      "learning_rate": 5.742758785231954e-06,
      "loss": 0.0075,
      "step": 18470
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.6127037405967712,
      "learning_rate": 5.7378334449286e-06,
      "loss": 0.0075,
      "step": 18475
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.43997830152511597,
      "learning_rate": 5.732909367784492e-06,
      "loss": 0.0081,
      "step": 18480
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.28263017535209656,
      "learning_rate": 5.727986555258955e-06,
      "loss": 0.0084,
      "step": 18485
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.33662813901901245,
      "learning_rate": 5.723065008810942e-06,
      "loss": 0.0099,
      "step": 18490
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.2875825762748718,
      "learning_rate": 5.718144729899031e-06,
      "loss": 0.0058,
      "step": 18495
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.20810118317604065,
      "learning_rate": 5.713225719981424e-06,
      "loss": 0.0068,
      "step": 18500
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.3628634214401245,
      "learning_rate": 5.708307980515943e-06,
      "loss": 0.007,
      "step": 18505
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.7062156796455383,
      "learning_rate": 5.703391512960037e-06,
      "loss": 0.0057,
      "step": 18510
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.43663591146469116,
      "learning_rate": 5.698476318770778e-06,
      "loss": 0.0087,
      "step": 18515
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.3063855767250061,
      "learning_rate": 5.693562399404862e-06,
      "loss": 0.0072,
      "step": 18520
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.3865858316421509,
      "learning_rate": 5.688649756318603e-06,
      "loss": 0.0093,
      "step": 18525
    },
    {
      "epoch": 5.78,
      "grad_norm": 1.110260248184204,
      "learning_rate": 5.68373839096794e-06,
      "loss": 0.009,
      "step": 18530
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.5447413921356201,
      "learning_rate": 5.678828304808433e-06,
      "loss": 0.0067,
      "step": 18535
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.44069159030914307,
      "learning_rate": 5.673919499295259e-06,
      "loss": 0.0087,
      "step": 18540
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.8088098764419556,
      "learning_rate": 5.669011975883223e-06,
      "loss": 0.0091,
      "step": 18545
    },
    {
      "epoch": 5.79,
      "grad_norm": 1.602562427520752,
      "learning_rate": 5.664105736026747e-06,
      "loss": 0.0063,
      "step": 18550
    },
    {
      "epoch": 5.79,
      "grad_norm": 0.43349578976631165,
      "learning_rate": 5.659200781179861e-06,
      "loss": 0.0061,
      "step": 18555
    },
    {
      "epoch": 5.79,
      "grad_norm": 0.4047665596008301,
      "learning_rate": 5.6542971127962354e-06,
      "loss": 0.0071,
      "step": 18560
    },
    {
      "epoch": 5.79,
      "grad_norm": 0.7945146560668945,
      "learning_rate": 5.649394732329146e-06,
      "loss": 0.0075,
      "step": 18565
    },
    {
      "epoch": 5.79,
      "grad_norm": 0.4201711118221283,
      "learning_rate": 5.6444936412314865e-06,
      "loss": 0.0096,
      "step": 18570
    },
    {
      "epoch": 5.79,
      "grad_norm": 0.41688671708106995,
      "learning_rate": 5.639593840955774e-06,
      "loss": 0.005,
      "step": 18575
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.7006680965423584,
      "learning_rate": 5.6346953329541394e-06,
      "loss": 0.0088,
      "step": 18580
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.27040454745292664,
      "learning_rate": 5.629798118678335e-06,
      "loss": 0.0081,
      "step": 18585
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.32489171624183655,
      "learning_rate": 5.624902199579721e-06,
      "loss": 0.0071,
      "step": 18590
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.5387438535690308,
      "learning_rate": 5.620007577109281e-06,
      "loss": 0.008,
      "step": 18595
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.6050368547439575,
      "learning_rate": 5.6151142527176074e-06,
      "loss": 0.0101,
      "step": 18600
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.23941612243652344,
      "learning_rate": 5.6102222278549226e-06,
      "loss": 0.0047,
      "step": 18605
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.6153684258460999,
      "learning_rate": 5.6053315039710495e-06,
      "loss": 0.0099,
      "step": 18610
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.35738781094551086,
      "learning_rate": 5.6004420825154295e-06,
      "loss": 0.0072,
      "step": 18615
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.5066648721694946,
      "learning_rate": 5.59555396493712e-06,
      "loss": 0.0067,
      "step": 18620
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.3617473244667053,
      "learning_rate": 5.590667152684792e-06,
      "loss": 0.0081,
      "step": 18625
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.6097965240478516,
      "learning_rate": 5.585781647206724e-06,
      "loss": 0.0059,
      "step": 18630
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.5452948808670044,
      "learning_rate": 5.580897449950813e-06,
      "loss": 0.0082,
      "step": 18635
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.5515166521072388,
      "learning_rate": 5.576014562364562e-06,
      "loss": 0.0079,
      "step": 18640
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.3684811294078827,
      "learning_rate": 5.5711329858951e-06,
      "loss": 0.0082,
      "step": 18645
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.5962668061256409,
      "learning_rate": 5.5662527219891535e-06,
      "loss": 0.0077,
      "step": 18650
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.2659561038017273,
      "learning_rate": 5.561373772093065e-06,
      "loss": 0.0088,
      "step": 18655
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.34423190355300903,
      "learning_rate": 5.5564961376527904e-06,
      "loss": 0.007,
      "step": 18660
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.40950068831443787,
      "learning_rate": 5.551619820113885e-06,
      "loss": 0.0049,
      "step": 18665
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.3321197032928467,
      "learning_rate": 5.5467448209215246e-06,
      "loss": 0.0079,
      "step": 18670
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.4530077576637268,
      "learning_rate": 5.541871141520492e-06,
      "loss": 0.0099,
      "step": 18675
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.485211580991745,
      "learning_rate": 5.536998783355175e-06,
      "loss": 0.0058,
      "step": 18680
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.5379996299743652,
      "learning_rate": 5.532127747869572e-06,
      "loss": 0.01,
      "step": 18685
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.564601719379425,
      "learning_rate": 5.527258036507297e-06,
      "loss": 0.0078,
      "step": 18690
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.29679882526397705,
      "learning_rate": 5.522389650711564e-06,
      "loss": 0.0068,
      "step": 18695
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.29927995800971985,
      "learning_rate": 5.5175225919251866e-06,
      "loss": 0.0066,
      "step": 18700
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.5875388383865356,
      "learning_rate": 5.512656861590597e-06,
      "loss": 0.0058,
      "step": 18705
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.3787662982940674,
      "learning_rate": 5.5077924611498305e-06,
      "loss": 0.0086,
      "step": 18710
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.41421857476234436,
      "learning_rate": 5.502929392044528e-06,
      "loss": 0.0056,
      "step": 18715
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.16817878186702728,
      "learning_rate": 5.498067655715935e-06,
      "loss": 0.0075,
      "step": 18720
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.46259284019470215,
      "learning_rate": 5.4932072536049045e-06,
      "loss": 0.0088,
      "step": 18725
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.5328652858734131,
      "learning_rate": 5.488348187151888e-06,
      "loss": 0.0063,
      "step": 18730
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.3828234076499939,
      "learning_rate": 5.483490457796949e-06,
      "loss": 0.0083,
      "step": 18735
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.29909276962280273,
      "learning_rate": 5.478634066979751e-06,
      "loss": 0.0084,
      "step": 18740
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.37078049778938293,
      "learning_rate": 5.473779016139557e-06,
      "loss": 0.0109,
      "step": 18745
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.603055477142334,
      "learning_rate": 5.468925306715243e-06,
      "loss": 0.008,
      "step": 18750
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.7020739316940308,
      "learning_rate": 5.464072940145274e-06,
      "loss": 0.0077,
      "step": 18755
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.8944587707519531,
      "learning_rate": 5.459221917867729e-06,
      "loss": 0.0058,
      "step": 18760
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.306552916765213,
      "learning_rate": 5.454372241320283e-06,
      "loss": 0.0087,
      "step": 18765
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.35202908515930176,
      "learning_rate": 5.449523911940211e-06,
      "loss": 0.006,
      "step": 18770
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.4559305012226105,
      "learning_rate": 5.444676931164393e-06,
      "loss": 0.0082,
      "step": 18775
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.2316230684518814,
      "learning_rate": 5.4398313004293056e-06,
      "loss": 0.0056,
      "step": 18780
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.5010417699813843,
      "learning_rate": 5.4349870211710285e-06,
      "loss": 0.0096,
      "step": 18785
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.5433796048164368,
      "learning_rate": 5.430144094825238e-06,
      "loss": 0.0107,
      "step": 18790
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.40989163517951965,
      "learning_rate": 5.425302522827209e-06,
      "loss": 0.0061,
      "step": 18795
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.3090606927871704,
      "learning_rate": 5.420462306611818e-06,
      "loss": 0.0061,
      "step": 18800
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.5436885356903076,
      "learning_rate": 5.4156234476135385e-06,
      "loss": 0.0105,
      "step": 18805
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.6983522176742554,
      "learning_rate": 5.410785947266442e-06,
      "loss": 0.0068,
      "step": 18810
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.24617521464824677,
      "learning_rate": 5.405949807004196e-06,
      "loss": 0.0071,
      "step": 18815
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.5607662200927734,
      "learning_rate": 5.401115028260064e-06,
      "loss": 0.0064,
      "step": 18820
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.6128187775611877,
      "learning_rate": 5.396281612466912e-06,
      "loss": 0.0073,
      "step": 18825
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.46771490573883057,
      "learning_rate": 5.391449561057194e-06,
      "loss": 0.0109,
      "step": 18830
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.19731804728507996,
      "learning_rate": 5.386618875462963e-06,
      "loss": 0.0082,
      "step": 18835
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.39063313603401184,
      "learning_rate": 5.381789557115877e-06,
      "loss": 0.0079,
      "step": 18840
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.2991364300251007,
      "learning_rate": 5.376961607447162e-06,
      "loss": 0.0076,
      "step": 18845
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.3813847005367279,
      "learning_rate": 5.37213502788767e-06,
      "loss": 0.0089,
      "step": 18850
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.46439966559410095,
      "learning_rate": 5.36730981986783e-06,
      "loss": 0.0108,
      "step": 18855
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.4384636878967285,
      "learning_rate": 5.3624859848176645e-06,
      "loss": 0.0098,
      "step": 18860
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.2956441342830658,
      "learning_rate": 5.357663524166794e-06,
      "loss": 0.008,
      "step": 18865
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.3376927971839905,
      "learning_rate": 5.35284243934443e-06,
      "loss": 0.0066,
      "step": 18870
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.6353251934051514,
      "learning_rate": 5.348022731779377e-06,
      "loss": 0.0064,
      "step": 18875
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.651057779788971,
      "learning_rate": 5.3432044029000255e-06,
      "loss": 0.0079,
      "step": 18880
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.3579331338405609,
      "learning_rate": 5.338387454134364e-06,
      "loss": 0.0077,
      "step": 18885
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.1430189609527588,
      "learning_rate": 5.3335718869099674e-06,
      "loss": 0.008,
      "step": 18890
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.538234531879425,
      "learning_rate": 5.328757702654011e-06,
      "loss": 0.0102,
      "step": 18895
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.5153356194496155,
      "learning_rate": 5.323944902793248e-06,
      "loss": 0.0084,
      "step": 18900
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.3739708960056305,
      "learning_rate": 5.3191334887540295e-06,
      "loss": 0.0077,
      "step": 18905
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.6305508613586426,
      "learning_rate": 5.314323461962294e-06,
      "loss": 0.0055,
      "step": 18910
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.4070693254470825,
      "learning_rate": 5.30951482384356e-06,
      "loss": 0.0077,
      "step": 18915
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.5026772618293762,
      "learning_rate": 5.304707575822947e-06,
      "loss": 0.0099,
      "step": 18920
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.4167676866054535,
      "learning_rate": 5.299901719325158e-06,
      "loss": 0.0074,
      "step": 18925
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.3035011291503906,
      "learning_rate": 5.295097255774477e-06,
      "loss": 0.0056,
      "step": 18930
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.4963723421096802,
      "learning_rate": 5.290294186594791e-06,
      "loss": 0.0086,
      "step": 18935
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.38422855734825134,
      "learning_rate": 5.28549251320956e-06,
      "loss": 0.0088,
      "step": 18940
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.39815953373908997,
      "learning_rate": 5.280692237041836e-06,
      "loss": 0.0086,
      "step": 18945
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.36266201734542847,
      "learning_rate": 5.275893359514256e-06,
      "loss": 0.0072,
      "step": 18950
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.20838455855846405,
      "learning_rate": 5.271095882049035e-06,
      "loss": 0.0063,
      "step": 18955
    },
    {
      "epoch": 5.91,
      "grad_norm": 1.448961853981018,
      "learning_rate": 5.266299806067987e-06,
      "loss": 0.0086,
      "step": 18960
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.29663020372390747,
      "learning_rate": 5.261505132992499e-06,
      "loss": 0.0077,
      "step": 18965
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.8893621563911438,
      "learning_rate": 5.2567118642435515e-06,
      "loss": 0.0088,
      "step": 18970
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.4229799211025238,
      "learning_rate": 5.251920001241695e-06,
      "loss": 0.0095,
      "step": 18975
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.8327857255935669,
      "learning_rate": 5.247129545407085e-06,
      "loss": 0.0091,
      "step": 18980
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.147551029920578,
      "learning_rate": 5.242340498159444e-06,
      "loss": 0.0077,
      "step": 18985
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.40383464097976685,
      "learning_rate": 5.237552860918076e-06,
      "loss": 0.0079,
      "step": 18990
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.16825222969055176,
      "learning_rate": 5.232766635101872e-06,
      "loss": 0.0059,
      "step": 18995
    },
    {
      "epoch": 5.93,
      "grad_norm": 0.8230346441268921,
      "learning_rate": 5.227981822129307e-06,
      "loss": 0.0074,
      "step": 19000
    },
    {
      "epoch": 5.93,
      "grad_norm": 0.3880881071090698,
      "learning_rate": 5.223198423418434e-06,
      "loss": 0.0077,
      "step": 19005
    },
    {
      "epoch": 5.93,
      "grad_norm": 0.38730117678642273,
      "learning_rate": 5.218416440386888e-06,
      "loss": 0.0074,
      "step": 19010
    },
    {
      "epoch": 5.93,
      "grad_norm": 0.38056811690330505,
      "learning_rate": 5.213635874451881e-06,
      "loss": 0.0081,
      "step": 19015
    },
    {
      "epoch": 5.93,
      "grad_norm": 0.21690897643566132,
      "learning_rate": 5.208856727030212e-06,
      "loss": 0.0058,
      "step": 19020
    },
    {
      "epoch": 5.93,
      "grad_norm": 0.46819308400154114,
      "learning_rate": 5.2040789995382535e-06,
      "loss": 0.0088,
      "step": 19025
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.30583274364471436,
      "learning_rate": 5.199302693391958e-06,
      "loss": 0.0063,
      "step": 19030
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.5496878027915955,
      "learning_rate": 5.194527810006862e-06,
      "loss": 0.0078,
      "step": 19035
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.38829562067985535,
      "learning_rate": 5.1897543507980715e-06,
      "loss": 0.0076,
      "step": 19040
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.5420771241188049,
      "learning_rate": 5.184982317180278e-06,
      "loss": 0.0082,
      "step": 19045
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.7623363137245178,
      "learning_rate": 5.180211710567749e-06,
      "loss": 0.0094,
      "step": 19050
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.3645646870136261,
      "learning_rate": 5.175442532374324e-06,
      "loss": 0.0074,
      "step": 19055
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.2958046793937683,
      "learning_rate": 5.170674784013425e-06,
      "loss": 0.0043,
      "step": 19060
    },
    {
      "epoch": 5.95,
      "grad_norm": 0.20783288776874542,
      "learning_rate": 5.1659084668980505e-06,
      "loss": 0.006,
      "step": 19065
    },
    {
      "epoch": 5.95,
      "grad_norm": 0.5735898613929749,
      "learning_rate": 5.16114358244077e-06,
      "loss": 0.0074,
      "step": 19070
    },
    {
      "epoch": 5.95,
      "grad_norm": 0.3978206217288971,
      "learning_rate": 5.156380132053731e-06,
      "loss": 0.007,
      "step": 19075
    },
    {
      "epoch": 5.95,
      "grad_norm": 0.15805083513259888,
      "learning_rate": 5.1516181171486555e-06,
      "loss": 0.007,
      "step": 19080
    },
    {
      "epoch": 5.95,
      "grad_norm": 0.33696630597114563,
      "learning_rate": 5.1468575391368424e-06,
      "loss": 0.008,
      "step": 19085
    },
    {
      "epoch": 5.95,
      "grad_norm": 0.5924596190452576,
      "learning_rate": 5.142098399429167e-06,
      "loss": 0.008,
      "step": 19090
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.5464776754379272,
      "learning_rate": 5.1373406994360605e-06,
      "loss": 0.0076,
      "step": 19095
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.5945857167243958,
      "learning_rate": 5.132584440567554e-06,
      "loss": 0.0081,
      "step": 19100
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.28529831767082214,
      "learning_rate": 5.127829624233236e-06,
      "loss": 0.0071,
      "step": 19105
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.3450835943222046,
      "learning_rate": 5.123076251842267e-06,
      "loss": 0.0084,
      "step": 19110
    },
    {
      "epoch": 5.96,
      "grad_norm": 1.0156694650650024,
      "learning_rate": 5.118324324803386e-06,
      "loss": 0.0066,
      "step": 19115
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.6432920098304749,
      "learning_rate": 5.113573844524899e-06,
      "loss": 0.0056,
      "step": 19120
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.20642875134944916,
      "learning_rate": 5.1088248124146835e-06,
      "loss": 0.0056,
      "step": 19125
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.39436250925064087,
      "learning_rate": 5.1040772298801946e-06,
      "loss": 0.0073,
      "step": 19130
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.2230510264635086,
      "learning_rate": 5.099331098328439e-06,
      "loss": 0.0088,
      "step": 19135
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.4602661430835724,
      "learning_rate": 5.09458641916602e-06,
      "loss": 0.0064,
      "step": 19140
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.29295966029167175,
      "learning_rate": 5.089843193799091e-06,
      "loss": 0.0054,
      "step": 19145
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.5813841819763184,
      "learning_rate": 5.085101423633383e-06,
      "loss": 0.012,
      "step": 19150
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.3695999085903168,
      "learning_rate": 5.08036111007419e-06,
      "loss": 0.0067,
      "step": 19155
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.3806309998035431,
      "learning_rate": 5.07562225452638e-06,
      "loss": 0.0075,
      "step": 19160
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.36197715997695923,
      "learning_rate": 5.0708848583943895e-06,
      "loss": 0.0082,
      "step": 19165
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.39807888865470886,
      "learning_rate": 5.066148923082211e-06,
      "loss": 0.0084,
      "step": 19170
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.32840362191200256,
      "learning_rate": 5.061414449993417e-06,
      "loss": 0.0049,
      "step": 19175
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.5978043675422668,
      "learning_rate": 5.0566814405311394e-06,
      "loss": 0.0101,
      "step": 19180
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.3670240342617035,
      "learning_rate": 5.051949896098085e-06,
      "loss": 0.0069,
      "step": 19185
    },
    {
      "epoch": 5.99,
      "grad_norm": 0.379352331161499,
      "learning_rate": 5.047219818096516e-06,
      "loss": 0.0052,
      "step": 19190
    },
    {
      "epoch": 5.99,
      "grad_norm": 0.4825327694416046,
      "learning_rate": 5.042491207928267e-06,
      "loss": 0.0075,
      "step": 19195
    },
    {
      "epoch": 5.99,
      "grad_norm": 0.21042650938034058,
      "learning_rate": 5.037764066994738e-06,
      "loss": 0.0072,
      "step": 19200
    },
    {
      "epoch": 5.99,
      "grad_norm": 0.18937867879867554,
      "learning_rate": 5.033038396696882e-06,
      "loss": 0.009,
      "step": 19205
    },
    {
      "epoch": 5.99,
      "grad_norm": 1.988599419593811,
      "learning_rate": 5.028314198435229e-06,
      "loss": 0.0098,
      "step": 19210
    },
    {
      "epoch": 5.99,
      "grad_norm": 0.38489198684692383,
      "learning_rate": 5.02359147360987e-06,
      "loss": 0.0075,
      "step": 19215
    },
    {
      "epoch": 5.99,
      "grad_norm": 0.5206966996192932,
      "learning_rate": 5.0188702236204555e-06,
      "loss": 0.0114,
      "step": 19220
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.477316290140152,
      "learning_rate": 5.014150449866197e-06,
      "loss": 0.0052,
      "step": 19225
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.34607836604118347,
      "learning_rate": 5.0094321537458836e-06,
      "loss": 0.0084,
      "step": 19230
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.21586057543754578,
      "learning_rate": 5.004715336657855e-06,
      "loss": 0.0065,
      "step": 19235
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.41035139560699463,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.0059,
      "step": 19240
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.13773435354232788,
      "learning_rate": 4.995286145169796e-06,
      "loss": 0.0041,
      "step": 19245
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.1690952032804489,
      "learning_rate": 4.9905737735642575e-06,
      "loss": 0.0043,
      "step": 19250
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.15634578466415405,
      "learning_rate": 4.985862886579974e-06,
      "loss": 0.0034,
      "step": 19255
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.14361192286014557,
      "learning_rate": 4.981153485613088e-06,
      "loss": 0.0048,
      "step": 19260
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.21888059377670288,
      "learning_rate": 4.976445572059301e-06,
      "loss": 0.0042,
      "step": 19265
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.07288442552089691,
      "learning_rate": 4.971739147313883e-06,
      "loss": 0.0024,
      "step": 19270
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.8259077072143555,
      "learning_rate": 4.967034212771656e-06,
      "loss": 0.0044,
      "step": 19275
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.28250551223754883,
      "learning_rate": 4.962330769826995e-06,
      "loss": 0.0045,
      "step": 19280
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.1364012509584427,
      "learning_rate": 4.957628819873838e-06,
      "loss": 0.003,
      "step": 19285
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.24540041387081146,
      "learning_rate": 4.952928364305686e-06,
      "loss": 0.005,
      "step": 19290
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.31267842650413513,
      "learning_rate": 4.948229404515588e-06,
      "loss": 0.004,
      "step": 19295
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.25432685017585754,
      "learning_rate": 4.943531941896155e-06,
      "loss": 0.0036,
      "step": 19300
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.38814499974250793,
      "learning_rate": 4.938835977839552e-06,
      "loss": 0.0026,
      "step": 19305
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.21670231223106384,
      "learning_rate": 4.934141513737504e-06,
      "loss": 0.0034,
      "step": 19310
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.295858234167099,
      "learning_rate": 4.929448550981284e-06,
      "loss": 0.0036,
      "step": 19315
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.29953882098197937,
      "learning_rate": 4.924757090961728e-06,
      "loss": 0.0021,
      "step": 19320
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.17823810875415802,
      "learning_rate": 4.920067135069221e-06,
      "loss": 0.0034,
      "step": 19325
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.1693033128976822,
      "learning_rate": 4.915378684693705e-06,
      "loss": 0.004,
      "step": 19330
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.1305498629808426,
      "learning_rate": 4.910691741224675e-06,
      "loss": 0.0031,
      "step": 19335
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.30055421590805054,
      "learning_rate": 4.90600630605118e-06,
      "loss": 0.003,
      "step": 19340
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.5468003153800964,
      "learning_rate": 4.90132238056182e-06,
      "loss": 0.004,
      "step": 19345
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.17088231444358826,
      "learning_rate": 4.896639966144751e-06,
      "loss": 0.0032,
      "step": 19350
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.5591351985931396,
      "learning_rate": 4.891959064187678e-06,
      "loss": 0.0047,
      "step": 19355
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.27993714809417725,
      "learning_rate": 4.8872796760778594e-06,
      "loss": 0.0036,
      "step": 19360
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.336864709854126,
      "learning_rate": 4.882601803202104e-06,
      "loss": 0.0045,
      "step": 19365
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.1561504602432251,
      "learning_rate": 4.877925446946772e-06,
      "loss": 0.0034,
      "step": 19370
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.21346211433410645,
      "learning_rate": 4.873250608697775e-06,
      "loss": 0.0052,
      "step": 19375
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.12421856820583344,
      "learning_rate": 4.868577289840577e-06,
      "loss": 0.0037,
      "step": 19380
    },
    {
      "epoch": 6.05,
      "grad_norm": 0.2269437611103058,
      "learning_rate": 4.863905491760178e-06,
      "loss": 0.0042,
      "step": 19385
    },
    {
      "epoch": 6.05,
      "grad_norm": 0.36611998081207275,
      "learning_rate": 4.859235215841148e-06,
      "loss": 0.0048,
      "step": 19390
    },
    {
      "epoch": 6.05,
      "grad_norm": 0.28265631198883057,
      "learning_rate": 4.854566463467594e-06,
      "loss": 0.0033,
      "step": 19395
    },
    {
      "epoch": 6.05,
      "grad_norm": 0.26229220628738403,
      "learning_rate": 4.8498992360231716e-06,
      "loss": 0.0035,
      "step": 19400
    },
    {
      "epoch": 6.05,
      "grad_norm": 0.1218532919883728,
      "learning_rate": 4.845233534891086e-06,
      "loss": 0.0033,
      "step": 19405
    },
    {
      "epoch": 6.05,
      "grad_norm": 0.31334492564201355,
      "learning_rate": 4.840569361454092e-06,
      "loss": 0.0043,
      "step": 19410
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.22502319514751434,
      "learning_rate": 4.835906717094491e-06,
      "loss": 0.0031,
      "step": 19415
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.3406829237937927,
      "learning_rate": 4.831245603194125e-06,
      "loss": 0.0029,
      "step": 19420
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.16087867319583893,
      "learning_rate": 4.826586021134385e-06,
      "loss": 0.0032,
      "step": 19425
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.4197031259536743,
      "learning_rate": 4.821927972296221e-06,
      "loss": 0.0033,
      "step": 19430
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.34231555461883545,
      "learning_rate": 4.817271458060112e-06,
      "loss": 0.0042,
      "step": 19435
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.3761197030544281,
      "learning_rate": 4.8126164798060895e-06,
      "loss": 0.0035,
      "step": 19440
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.2574174702167511,
      "learning_rate": 4.807963038913728e-06,
      "loss": 0.0039,
      "step": 19445
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.3793030083179474,
      "learning_rate": 4.8033111367621475e-06,
      "loss": 0.0037,
      "step": 19450
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.10669315606355667,
      "learning_rate": 4.798660774730014e-06,
      "loss": 0.0041,
      "step": 19455
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.11876823753118515,
      "learning_rate": 4.794011954195528e-06,
      "loss": 0.0038,
      "step": 19460
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.49461352825164795,
      "learning_rate": 4.789364676536444e-06,
      "loss": 0.0036,
      "step": 19465
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.12926629185676575,
      "learning_rate": 4.784718943130051e-06,
      "loss": 0.0027,
      "step": 19470
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.26242661476135254,
      "learning_rate": 4.78007475535319e-06,
      "loss": 0.0037,
      "step": 19475
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.27047592401504517,
      "learning_rate": 4.7754321145822355e-06,
      "loss": 0.0037,
      "step": 19480
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.3222498595714569,
      "learning_rate": 4.7707910221931065e-06,
      "loss": 0.004,
      "step": 19485
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.10554604232311249,
      "learning_rate": 4.766151479561267e-06,
      "loss": 0.0033,
      "step": 19490
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.8339462876319885,
      "learning_rate": 4.761513488061709e-06,
      "loss": 0.0027,
      "step": 19495
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.20556354522705078,
      "learning_rate": 4.756877049068977e-06,
      "loss": 0.003,
      "step": 19500
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.4011286199092865,
      "learning_rate": 4.752242163957151e-06,
      "loss": 0.0034,
      "step": 19505
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.1158180832862854,
      "learning_rate": 4.747608834099854e-06,
      "loss": 0.0029,
      "step": 19510
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.33003300428390503,
      "learning_rate": 4.7429770608702395e-06,
      "loss": 0.0037,
      "step": 19515
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.266290545463562,
      "learning_rate": 4.738346845641013e-06,
      "loss": 0.005,
      "step": 19520
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.38621214032173157,
      "learning_rate": 4.733718189784411e-06,
      "loss": 0.0065,
      "step": 19525
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.38451021909713745,
      "learning_rate": 4.729091094672199e-06,
      "loss": 0.004,
      "step": 19530
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.46946078538894653,
      "learning_rate": 4.7244655616756944e-06,
      "loss": 0.0064,
      "step": 19535
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.10865452140569687,
      "learning_rate": 4.719841592165743e-06,
      "loss": 0.0044,
      "step": 19540
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.23039183020591736,
      "learning_rate": 4.715219187512733e-06,
      "loss": 0.0056,
      "step": 19545
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.29119929671287537,
      "learning_rate": 4.710598349086584e-06,
      "loss": 0.0028,
      "step": 19550
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.25930312275886536,
      "learning_rate": 4.7059790782567484e-06,
      "loss": 0.0037,
      "step": 19555
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.24252845346927643,
      "learning_rate": 4.701361376392231e-06,
      "loss": 0.0028,
      "step": 19560
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.4317205548286438,
      "learning_rate": 4.696745244861552e-06,
      "loss": 0.0053,
      "step": 19565
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.35169634222984314,
      "learning_rate": 4.692130685032771e-06,
      "loss": 0.0039,
      "step": 19570
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.3276447653770447,
      "learning_rate": 4.687517698273488e-06,
      "loss": 0.0026,
      "step": 19575
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.2100311517715454,
      "learning_rate": 4.682906285950834e-06,
      "loss": 0.0035,
      "step": 19580
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.30278119444847107,
      "learning_rate": 4.67829644943147e-06,
      "loss": 0.0039,
      "step": 19585
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.34402674436569214,
      "learning_rate": 4.673688190081594e-06,
      "loss": 0.0041,
      "step": 19590
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.3510453402996063,
      "learning_rate": 4.669081509266936e-06,
      "loss": 0.0035,
      "step": 19595
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.11933588981628418,
      "learning_rate": 4.664476408352757e-06,
      "loss": 0.0031,
      "step": 19600
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.23259113729000092,
      "learning_rate": 4.659872888703849e-06,
      "loss": 0.0042,
      "step": 19605
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.31132641434669495,
      "learning_rate": 4.6552709516845385e-06,
      "loss": 0.0051,
      "step": 19610
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.3782808780670166,
      "learning_rate": 4.650670598658679e-06,
      "loss": 0.0038,
      "step": 19615
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.3041776716709137,
      "learning_rate": 4.646071830989658e-06,
      "loss": 0.0026,
      "step": 19620
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.09638538956642151,
      "learning_rate": 4.6414746500403904e-06,
      "loss": 0.0023,
      "step": 19625
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.38467106223106384,
      "learning_rate": 4.6368790571733234e-06,
      "loss": 0.0032,
      "step": 19630
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.15978440642356873,
      "learning_rate": 4.632285053750432e-06,
      "loss": 0.0033,
      "step": 19635
    },
    {
      "epoch": 6.13,
      "grad_norm": 0.4148300588130951,
      "learning_rate": 4.627692641133219e-06,
      "loss": 0.0062,
      "step": 19640
    },
    {
      "epoch": 6.13,
      "grad_norm": 0.29224053025245667,
      "learning_rate": 4.62310182068272e-06,
      "loss": 0.0028,
      "step": 19645
    },
    {
      "epoch": 6.13,
      "grad_norm": 0.13670198619365692,
      "learning_rate": 4.618512593759493e-06,
      "loss": 0.0038,
      "step": 19650
    },
    {
      "epoch": 6.13,
      "grad_norm": 0.279555082321167,
      "learning_rate": 4.613924961723629e-06,
      "loss": 0.003,
      "step": 19655
    },
    {
      "epoch": 6.13,
      "grad_norm": 0.4272875487804413,
      "learning_rate": 4.609338925934743e-06,
      "loss": 0.0034,
      "step": 19660
    },
    {
      "epoch": 6.13,
      "grad_norm": 0.18500889837741852,
      "learning_rate": 4.604754487751978e-06,
      "loss": 0.003,
      "step": 19665
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.184945747256279,
      "learning_rate": 4.600171648534008e-06,
      "loss": 0.003,
      "step": 19670
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.28593793511390686,
      "learning_rate": 4.595590409639014e-06,
      "loss": 0.0035,
      "step": 19675
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.6415123343467712,
      "learning_rate": 4.591010772424733e-06,
      "loss": 0.0033,
      "step": 19680
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.2268776297569275,
      "learning_rate": 4.586432738248405e-06,
      "loss": 0.0029,
      "step": 19685
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.1716085523366928,
      "learning_rate": 4.581856308466804e-06,
      "loss": 0.0039,
      "step": 19690
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.6511481404304504,
      "learning_rate": 4.577281484436223e-06,
      "loss": 0.0028,
      "step": 19695
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.35789069533348083,
      "learning_rate": 4.572708267512484e-06,
      "loss": 0.0059,
      "step": 19700
    },
    {
      "epoch": 6.15,
      "grad_norm": 0.2436264157295227,
      "learning_rate": 4.568136659050933e-06,
      "loss": 0.0032,
      "step": 19705
    },
    {
      "epoch": 6.15,
      "grad_norm": 0.49384281039237976,
      "learning_rate": 4.563566660406433e-06,
      "loss": 0.0056,
      "step": 19710
    },
    {
      "epoch": 6.15,
      "grad_norm": 0.2979516386985779,
      "learning_rate": 4.558998272933375e-06,
      "loss": 0.0039,
      "step": 19715
    },
    {
      "epoch": 6.15,
      "grad_norm": 0.23512761294841766,
      "learning_rate": 4.554431497985667e-06,
      "loss": 0.0064,
      "step": 19720
    },
    {
      "epoch": 6.15,
      "grad_norm": 0.137066051363945,
      "learning_rate": 4.549866336916755e-06,
      "loss": 0.0039,
      "step": 19725
    },
    {
      "epoch": 6.15,
      "grad_norm": 0.13359549641609192,
      "learning_rate": 4.545302791079588e-06,
      "loss": 0.0034,
      "step": 19730
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.17334330081939697,
      "learning_rate": 4.540740861826644e-06,
      "loss": 0.002,
      "step": 19735
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.2552857995033264,
      "learning_rate": 4.536180550509928e-06,
      "loss": 0.0035,
      "step": 19740
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.6206842660903931,
      "learning_rate": 4.531621858480949e-06,
      "loss": 0.0043,
      "step": 19745
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.30290666222572327,
      "learning_rate": 4.527064787090752e-06,
      "loss": 0.0033,
      "step": 19750
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.2942062020301819,
      "learning_rate": 4.522509337689893e-06,
      "loss": 0.0039,
      "step": 19755
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.28894585371017456,
      "learning_rate": 4.517955511628449e-06,
      "loss": 0.0041,
      "step": 19760
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.3294394612312317,
      "learning_rate": 4.513403310256023e-06,
      "loss": 0.0052,
      "step": 19765
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.17340120673179626,
      "learning_rate": 4.5088527349217286e-06,
      "loss": 0.0035,
      "step": 19770
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.789074182510376,
      "learning_rate": 4.5043037869741965e-06,
      "loss": 0.0036,
      "step": 19775
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.5654256343841553,
      "learning_rate": 4.499756467761585e-06,
      "loss": 0.0036,
      "step": 19780
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.11048149317502975,
      "learning_rate": 4.495210778631553e-06,
      "loss": 0.0032,
      "step": 19785
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.09231693297624588,
      "learning_rate": 4.490666720931293e-06,
      "loss": 0.0028,
      "step": 19790
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.17293661832809448,
      "learning_rate": 4.486124296007502e-06,
      "loss": 0.0038,
      "step": 19795
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.24658824503421783,
      "learning_rate": 4.481583505206403e-06,
      "loss": 0.0041,
      "step": 19800
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.18947665393352509,
      "learning_rate": 4.4770443498737224e-06,
      "loss": 0.0037,
      "step": 19805
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.40546736121177673,
      "learning_rate": 4.47250683135472e-06,
      "loss": 0.0043,
      "step": 19810
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.5654178857803345,
      "learning_rate": 4.4679709509941585e-06,
      "loss": 0.0042,
      "step": 19815
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.4417353868484497,
      "learning_rate": 4.463436710136308e-06,
      "loss": 0.0061,
      "step": 19820
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.4045957922935486,
      "learning_rate": 4.458904110124965e-06,
      "loss": 0.0042,
      "step": 19825
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.054574161767959595,
      "learning_rate": 4.454373152303439e-06,
      "loss": 0.0041,
      "step": 19830
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.10774843394756317,
      "learning_rate": 4.449843838014546e-06,
      "loss": 0.0036,
      "step": 19835
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.3569006621837616,
      "learning_rate": 4.445316168600621e-06,
      "loss": 0.0045,
      "step": 19840
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.12751227617263794,
      "learning_rate": 4.440790145403503e-06,
      "loss": 0.0043,
      "step": 19845
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.43211138248443604,
      "learning_rate": 4.4362657697645624e-06,
      "loss": 0.004,
      "step": 19850
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.3285439908504486,
      "learning_rate": 4.431743043024658e-06,
      "loss": 0.0043,
      "step": 19855
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.4181636571884155,
      "learning_rate": 4.427221966524171e-06,
      "loss": 0.0044,
      "step": 19860
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.26568159461021423,
      "learning_rate": 4.422702541602994e-06,
      "loss": 0.0045,
      "step": 19865
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.13646848499774933,
      "learning_rate": 4.41818476960053e-06,
      "loss": 0.0032,
      "step": 19870
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.45093557238578796,
      "learning_rate": 4.4136686518556884e-06,
      "loss": 0.0033,
      "step": 19875
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.22524970769882202,
      "learning_rate": 4.409154189706892e-06,
      "loss": 0.0025,
      "step": 19880
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.6135339140892029,
      "learning_rate": 4.404641384492073e-06,
      "loss": 0.0035,
      "step": 19885
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.29468056559562683,
      "learning_rate": 4.400130237548669e-06,
      "loss": 0.0035,
      "step": 19890
    },
    {
      "epoch": 6.21,
      "grad_norm": 0.2552112340927124,
      "learning_rate": 4.39562075021363e-06,
      "loss": 0.007,
      "step": 19895
    },
    {
      "epoch": 6.21,
      "grad_norm": 0.38863444328308105,
      "learning_rate": 4.391112923823413e-06,
      "loss": 0.0044,
      "step": 19900
    },
    {
      "epoch": 6.21,
      "grad_norm": 0.36713364720344543,
      "learning_rate": 4.38660675971398e-06,
      "loss": 0.0036,
      "step": 19905
    },
    {
      "epoch": 6.21,
      "grad_norm": 0.7707926034927368,
      "learning_rate": 4.382102259220805e-06,
      "loss": 0.0032,
      "step": 19910
    },
    {
      "epoch": 6.21,
      "grad_norm": 0.1946425884962082,
      "learning_rate": 4.377599423678867e-06,
      "loss": 0.0042,
      "step": 19915
    },
    {
      "epoch": 6.21,
      "grad_norm": 0.1757972240447998,
      "learning_rate": 4.373098254422649e-06,
      "loss": 0.0056,
      "step": 19920
    },
    {
      "epoch": 6.21,
      "grad_norm": 0.21779265999794006,
      "learning_rate": 4.368598752786143e-06,
      "loss": 0.0027,
      "step": 19925
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.24872009456157684,
      "learning_rate": 4.364100920102847e-06,
      "loss": 0.0043,
      "step": 19930
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.08531658351421356,
      "learning_rate": 4.359604757705764e-06,
      "loss": 0.0028,
      "step": 19935
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.17155839502811432,
      "learning_rate": 4.3551102669274e-06,
      "loss": 0.0028,
      "step": 19940
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.136628195643425,
      "learning_rate": 4.350617449099769e-06,
      "loss": 0.0039,
      "step": 19945
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.6224244236946106,
      "learning_rate": 4.346126305554385e-06,
      "loss": 0.0022,
      "step": 19950
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.41527363657951355,
      "learning_rate": 4.341636837622272e-06,
      "loss": 0.0033,
      "step": 19955
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.25839653611183167,
      "learning_rate": 4.337149046633953e-06,
      "loss": 0.0027,
      "step": 19960
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.23557740449905396,
      "learning_rate": 4.332662933919445e-06,
      "loss": 0.0045,
      "step": 19965
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.2542274296283722,
      "learning_rate": 4.328178500808289e-06,
      "loss": 0.005,
      "step": 19970
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.16733460128307343,
      "learning_rate": 4.3236957486295115e-06,
      "loss": 0.0034,
      "step": 19975
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.28059637546539307,
      "learning_rate": 4.319214678711646e-06,
      "loss": 0.0037,
      "step": 19980
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.3275239169597626,
      "learning_rate": 4.314735292382729e-06,
      "loss": 0.0031,
      "step": 19985
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.15801599621772766,
      "learning_rate": 4.310257590970294e-06,
      "loss": 0.0035,
      "step": 19990
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.07836773246526718,
      "learning_rate": 4.305781575801381e-06,
      "loss": 0.0035,
      "step": 19995
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.16531780362129211,
      "learning_rate": 4.301307248202521e-06,
      "loss": 0.0027,
      "step": 20000
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.16912704706192017,
      "learning_rate": 4.296834609499752e-06,
      "loss": 0.0032,
      "step": 20005
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.3261689245700836,
      "learning_rate": 4.2923636610186084e-06,
      "loss": 0.0042,
      "step": 20010
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.19443699717521667,
      "learning_rate": 4.287894404084131e-06,
      "loss": 0.0043,
      "step": 20015
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.23142337799072266,
      "learning_rate": 4.28342684002085e-06,
      "loss": 0.0033,
      "step": 20020
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.2110389918088913,
      "learning_rate": 4.278960970152798e-06,
      "loss": 0.0034,
      "step": 20025
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.47038784623146057,
      "learning_rate": 4.274496795803509e-06,
      "loss": 0.0033,
      "step": 20030
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.5487150549888611,
      "learning_rate": 4.2700343182960016e-06,
      "loss": 0.003,
      "step": 20035
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.21167035400867462,
      "learning_rate": 4.265573538952805e-06,
      "loss": 0.0046,
      "step": 20040
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.17947198450565338,
      "learning_rate": 4.26111445909594e-06,
      "loss": 0.0044,
      "step": 20045
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.373744934797287,
      "learning_rate": 4.256657080046921e-06,
      "loss": 0.0037,
      "step": 20050
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.3738728165626526,
      "learning_rate": 4.252201403126768e-06,
      "loss": 0.0041,
      "step": 20055
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.44824865460395813,
      "learning_rate": 4.247747429655987e-06,
      "loss": 0.0067,
      "step": 20060
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.23117795586585999,
      "learning_rate": 4.243295160954586e-06,
      "loss": 0.0049,
      "step": 20065
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.2192441076040268,
      "learning_rate": 4.238844598342056e-06,
      "loss": 0.0032,
      "step": 20070
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.27675142884254456,
      "learning_rate": 4.234395743137395e-06,
      "loss": 0.0041,
      "step": 20075
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.1432168185710907,
      "learning_rate": 4.2299485966590895e-06,
      "loss": 0.0031,
      "step": 20080
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.18300987780094147,
      "learning_rate": 4.225503160225119e-06,
      "loss": 0.0028,
      "step": 20085
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.34145647287368774,
      "learning_rate": 4.221059435152961e-06,
      "loss": 0.0042,
      "step": 20090
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.4050016403198242,
      "learning_rate": 4.216617422759577e-06,
      "loss": 0.0047,
      "step": 20095
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.2938024401664734,
      "learning_rate": 4.212177124361433e-06,
      "loss": 0.0026,
      "step": 20100
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.13880467414855957,
      "learning_rate": 4.207738541274482e-06,
      "loss": 0.0033,
      "step": 20105
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.2138069123029709,
      "learning_rate": 4.203301674814158e-06,
      "loss": 0.0018,
      "step": 20110
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.3359643816947937,
      "learning_rate": 4.1988665262954e-06,
      "loss": 0.0045,
      "step": 20115
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.2553737163543701,
      "learning_rate": 4.194433097032632e-06,
      "loss": 0.0041,
      "step": 20120
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.4595305919647217,
      "learning_rate": 4.190001388339771e-06,
      "loss": 0.0037,
      "step": 20125
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.11691176146268845,
      "learning_rate": 4.185571401530223e-06,
      "loss": 0.0046,
      "step": 20130
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.29672908782958984,
      "learning_rate": 4.1811431379168824e-06,
      "loss": 0.0039,
      "step": 20135
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.30570900440216064,
      "learning_rate": 4.176716598812133e-06,
      "loss": 0.0027,
      "step": 20140
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.2994268238544464,
      "learning_rate": 4.17229178552785e-06,
      "loss": 0.0032,
      "step": 20145
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.284789115190506,
      "learning_rate": 4.167868699375394e-06,
      "loss": 0.0033,
      "step": 20150
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.33170077204704285,
      "learning_rate": 4.1634473416656175e-06,
      "loss": 0.0035,
      "step": 20155
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.22531452775001526,
      "learning_rate": 4.159027713708858e-06,
      "loss": 0.0036,
      "step": 20160
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.3662478029727936,
      "learning_rate": 4.154609816814938e-06,
      "loss": 0.0035,
      "step": 20165
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.24185219407081604,
      "learning_rate": 4.150193652293173e-06,
      "loss": 0.0045,
      "step": 20170
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.1317419558763504,
      "learning_rate": 4.1457792214523606e-06,
      "loss": 0.0036,
      "step": 20175
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.25156107544898987,
      "learning_rate": 4.141366525600786e-06,
      "loss": 0.0038,
      "step": 20180
    },
    {
      "epoch": 6.3,
      "grad_norm": 0.4421023428440094,
      "learning_rate": 4.136955566046221e-06,
      "loss": 0.0041,
      "step": 20185
    },
    {
      "epoch": 6.3,
      "grad_norm": 0.09637313336133957,
      "learning_rate": 4.132546344095922e-06,
      "loss": 0.003,
      "step": 20190
    },
    {
      "epoch": 6.3,
      "grad_norm": 0.3749695420265198,
      "learning_rate": 4.128138861056629e-06,
      "loss": 0.0058,
      "step": 20195
    },
    {
      "epoch": 6.3,
      "grad_norm": 0.13544899225234985,
      "learning_rate": 4.123733118234569e-06,
      "loss": 0.0033,
      "step": 20200
    },
    {
      "epoch": 6.3,
      "grad_norm": 0.289870023727417,
      "learning_rate": 4.119329116935451e-06,
      "loss": 0.0041,
      "step": 20205
    },
    {
      "epoch": 6.3,
      "grad_norm": 0.18580834567546844,
      "learning_rate": 4.114926858464474e-06,
      "loss": 0.0035,
      "step": 20210
    },
    {
      "epoch": 6.3,
      "grad_norm": 0.18573562800884247,
      "learning_rate": 4.110526344126306e-06,
      "loss": 0.0047,
      "step": 20215
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.4938793480396271,
      "learning_rate": 4.106127575225116e-06,
      "loss": 0.0043,
      "step": 20220
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.22334390878677368,
      "learning_rate": 4.101730553064546e-06,
      "loss": 0.0038,
      "step": 20225
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.25191017985343933,
      "learning_rate": 4.097335278947721e-06,
      "loss": 0.0033,
      "step": 20230
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.27087491750717163,
      "learning_rate": 4.0929417541772474e-06,
      "loss": 0.0034,
      "step": 20235
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.32818400859832764,
      "learning_rate": 4.088549980055218e-06,
      "loss": 0.0034,
      "step": 20240
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.36982041597366333,
      "learning_rate": 4.084159957883202e-06,
      "loss": 0.0048,
      "step": 20245
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.18681477010250092,
      "learning_rate": 4.079771688962248e-06,
      "loss": 0.0025,
      "step": 20250
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.18717291951179504,
      "learning_rate": 4.0753851745928856e-06,
      "loss": 0.0053,
      "step": 20255
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.4687967002391815,
      "learning_rate": 4.071000416075133e-06,
      "loss": 0.0038,
      "step": 20260
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.2571309804916382,
      "learning_rate": 4.0666174147084796e-06,
      "loss": 0.0043,
      "step": 20265
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.1658802479505539,
      "learning_rate": 4.0622361717918935e-06,
      "loss": 0.0052,
      "step": 20270
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.14285005629062653,
      "learning_rate": 4.057856688623827e-06,
      "loss": 0.0042,
      "step": 20275
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.40070050954818726,
      "learning_rate": 4.053478966502205e-06,
      "loss": 0.0034,
      "step": 20280
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.1511838585138321,
      "learning_rate": 4.049103006724439e-06,
      "loss": 0.0029,
      "step": 20285
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.40002161264419556,
      "learning_rate": 4.044728810587406e-06,
      "loss": 0.0057,
      "step": 20290
    },
    {
      "epoch": 6.33,
      "grad_norm": 1.249631643295288,
      "learning_rate": 4.040356379387469e-06,
      "loss": 0.0044,
      "step": 20295
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.3465880751609802,
      "learning_rate": 4.0359857144204625e-06,
      "loss": 0.0031,
      "step": 20300
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.25544366240501404,
      "learning_rate": 4.031616816981708e-06,
      "loss": 0.0037,
      "step": 20305
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.38923880457878113,
      "learning_rate": 4.027249688365994e-06,
      "loss": 0.0041,
      "step": 20310
    },
    {
      "epoch": 6.34,
      "grad_norm": 0.1947145313024521,
      "learning_rate": 4.022884329867586e-06,
      "loss": 0.0034,
      "step": 20315
    },
    {
      "epoch": 6.34,
      "grad_norm": 0.154936745762825,
      "learning_rate": 4.018520742780227e-06,
      "loss": 0.0025,
      "step": 20320
    },
    {
      "epoch": 6.34,
      "grad_norm": 0.29328685998916626,
      "learning_rate": 4.01415892839713e-06,
      "loss": 0.0039,
      "step": 20325
    },
    {
      "epoch": 6.34,
      "grad_norm": 0.2891095280647278,
      "learning_rate": 4.00979888801099e-06,
      "loss": 0.0031,
      "step": 20330
    },
    {
      "epoch": 6.34,
      "grad_norm": 0.15493731200695038,
      "learning_rate": 4.0054406229139706e-06,
      "loss": 0.003,
      "step": 20335
    },
    {
      "epoch": 6.34,
      "grad_norm": 0.17908723652362823,
      "learning_rate": 4.001084134397708e-06,
      "loss": 0.0036,
      "step": 20340
    },
    {
      "epoch": 6.35,
      "grad_norm": 0.5203627943992615,
      "learning_rate": 3.996729423753324e-06,
      "loss": 0.0056,
      "step": 20345
    },
    {
      "epoch": 6.35,
      "grad_norm": 0.2771199345588684,
      "learning_rate": 3.9923764922714e-06,
      "loss": 0.0037,
      "step": 20350
    },
    {
      "epoch": 6.35,
      "grad_norm": 0.3039291501045227,
      "learning_rate": 3.988025341241999e-06,
      "loss": 0.0042,
      "step": 20355
    },
    {
      "epoch": 6.35,
      "grad_norm": 0.22865301370620728,
      "learning_rate": 3.983675971954643e-06,
      "loss": 0.0045,
      "step": 20360
    },
    {
      "epoch": 6.35,
      "grad_norm": 0.47996705770492554,
      "learning_rate": 3.979328385698342e-06,
      "loss": 0.0047,
      "step": 20365
    },
    {
      "epoch": 6.35,
      "grad_norm": 0.41893118619918823,
      "learning_rate": 3.974982583761568e-06,
      "loss": 0.0026,
      "step": 20370
    },
    {
      "epoch": 6.35,
      "grad_norm": 0.6908000111579895,
      "learning_rate": 3.970638567432269e-06,
      "loss": 0.0033,
      "step": 20375
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.1881827861070633,
      "learning_rate": 3.966296337997859e-06,
      "loss": 0.0041,
      "step": 20380
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.2765253782272339,
      "learning_rate": 3.961955896745224e-06,
      "loss": 0.0043,
      "step": 20385
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.5398893356323242,
      "learning_rate": 3.957617244960732e-06,
      "loss": 0.0044,
      "step": 20390
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.13423627614974976,
      "learning_rate": 3.953280383930197e-06,
      "loss": 0.0028,
      "step": 20395
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.11971479654312134,
      "learning_rate": 3.94894531493892e-06,
      "loss": 0.0025,
      "step": 20400
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.17684215307235718,
      "learning_rate": 3.944612039271666e-06,
      "loss": 0.0032,
      "step": 20405
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.37312769889831543,
      "learning_rate": 3.940280558212669e-06,
      "loss": 0.004,
      "step": 20410
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.45159024000167847,
      "learning_rate": 3.935950873045629e-06,
      "loss": 0.0041,
      "step": 20415
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.08123624324798584,
      "learning_rate": 3.931622985053717e-06,
      "loss": 0.0031,
      "step": 20420
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.2570103406906128,
      "learning_rate": 3.927296895519569e-06,
      "loss": 0.0055,
      "step": 20425
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.2280176728963852,
      "learning_rate": 3.922972605725291e-06,
      "loss": 0.0026,
      "step": 20430
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.34249600768089294,
      "learning_rate": 3.91865011695245e-06,
      "loss": 0.0051,
      "step": 20435
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.5048180818557739,
      "learning_rate": 3.914329430482083e-06,
      "loss": 0.0038,
      "step": 20440
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.32122793793678284,
      "learning_rate": 3.910010547594695e-06,
      "loss": 0.0035,
      "step": 20445
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.039529550820589066,
      "learning_rate": 3.905693469570253e-06,
      "loss": 0.0047,
      "step": 20450
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.23486529290676117,
      "learning_rate": 3.901378197688191e-06,
      "loss": 0.0039,
      "step": 20455
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.12477610260248184,
      "learning_rate": 3.897064733227404e-06,
      "loss": 0.0023,
      "step": 20460
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.30318892002105713,
      "learning_rate": 3.892753077466256e-06,
      "loss": 0.0034,
      "step": 20465
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.10283812880516052,
      "learning_rate": 3.888443231682573e-06,
      "loss": 0.0028,
      "step": 20470
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.252560019493103,
      "learning_rate": 3.884135197153645e-06,
      "loss": 0.0047,
      "step": 20475
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.27102649211883545,
      "learning_rate": 3.879828975156225e-06,
      "loss": 0.0032,
      "step": 20480
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.4798802137374878,
      "learning_rate": 3.875524566966529e-06,
      "loss": 0.0053,
      "step": 20485
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.5022309422492981,
      "learning_rate": 3.871221973860234e-06,
      "loss": 0.0035,
      "step": 20490
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.2926425039768219,
      "learning_rate": 3.866921197112481e-06,
      "loss": 0.0052,
      "step": 20495
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.28833624720573425,
      "learning_rate": 3.862622237997875e-06,
      "loss": 0.004,
      "step": 20500
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.24428056180477142,
      "learning_rate": 3.85832509779047e-06,
      "loss": 0.0047,
      "step": 20505
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.2550247013568878,
      "learning_rate": 3.854029777763799e-06,
      "loss": 0.0032,
      "step": 20510
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.5598758459091187,
      "learning_rate": 3.849736279190846e-06,
      "loss": 0.0028,
      "step": 20515
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.14802761375904083,
      "learning_rate": 3.845444603344053e-06,
      "loss": 0.0038,
      "step": 20520
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.23852317035198212,
      "learning_rate": 3.841154751495326e-06,
      "loss": 0.0039,
      "step": 20525
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.17386658489704132,
      "learning_rate": 3.836866724916029e-06,
      "loss": 0.0031,
      "step": 20530
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.2971315383911133,
      "learning_rate": 3.8325805248769896e-06,
      "loss": 0.0038,
      "step": 20535
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.2956434190273285,
      "learning_rate": 3.8282961526484815e-06,
      "loss": 0.0049,
      "step": 20540
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.07207696884870529,
      "learning_rate": 3.8240136095002465e-06,
      "loss": 0.003,
      "step": 20545
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.09542746841907501,
      "learning_rate": 3.819732896701488e-06,
      "loss": 0.0033,
      "step": 20550
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.1593623012304306,
      "learning_rate": 3.81545401552086e-06,
      "loss": 0.0034,
      "step": 20555
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.13820824027061462,
      "learning_rate": 3.811176967226473e-06,
      "loss": 0.0037,
      "step": 20560
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.37225475907325745,
      "learning_rate": 3.8069017530859e-06,
      "loss": 0.0051,
      "step": 20565
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.07505133002996445,
      "learning_rate": 3.8026283743661673e-06,
      "loss": 0.0033,
      "step": 20570
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.2258065938949585,
      "learning_rate": 3.7983568323337518e-06,
      "loss": 0.0031,
      "step": 20575
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.27532634139060974,
      "learning_rate": 3.7940871282545945e-06,
      "loss": 0.005,
      "step": 20580
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.24397745728492737,
      "learning_rate": 3.789819263394088e-06,
      "loss": 0.0042,
      "step": 20585
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.34583353996276855,
      "learning_rate": 3.785553239017078e-06,
      "loss": 0.0038,
      "step": 20590
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.331437349319458,
      "learning_rate": 3.7812890563878736e-06,
      "loss": 0.0045,
      "step": 20595
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.14943893253803253,
      "learning_rate": 3.777026716770227e-06,
      "loss": 0.0026,
      "step": 20600
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.2038610726594925,
      "learning_rate": 3.7727662214273496e-06,
      "loss": 0.0036,
      "step": 20605
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.4535045623779297,
      "learning_rate": 3.768507571621909e-06,
      "loss": 0.0032,
      "step": 20610
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.1757279336452484,
      "learning_rate": 3.7642507686160155e-06,
      "loss": 0.0033,
      "step": 20615
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.2368738204240799,
      "learning_rate": 3.759995813671241e-06,
      "loss": 0.004,
      "step": 20620
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.211713045835495,
      "learning_rate": 3.7557427080486098e-06,
      "loss": 0.0039,
      "step": 20625
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.25176042318344116,
      "learning_rate": 3.751491453008592e-06,
      "loss": 0.0029,
      "step": 20630
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.4962671399116516,
      "learning_rate": 3.747242049811114e-06,
      "loss": 0.0036,
      "step": 20635
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.2933531105518341,
      "learning_rate": 3.7429944997155565e-06,
      "loss": 0.0053,
      "step": 20640
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.3499653935432434,
      "learning_rate": 3.7387488039807473e-06,
      "loss": 0.0041,
      "step": 20645
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.806042492389679,
      "learning_rate": 3.7345049638649577e-06,
      "loss": 0.005,
      "step": 20650
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.36196035146713257,
      "learning_rate": 3.7302629806259184e-06,
      "loss": 0.0041,
      "step": 20655
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.14944805204868317,
      "learning_rate": 3.7268707318307895e-06,
      "loss": 0.0033,
      "step": 20660
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.2249460369348526,
      "learning_rate": 3.7226320941376224e-06,
      "loss": 0.0046,
      "step": 20665
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.36011096835136414,
      "learning_rate": 3.71839531683991e-06,
      "loss": 0.0042,
      "step": 20670
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.09397727996110916,
      "learning_rate": 3.7141604011932965e-06,
      "loss": 0.0033,
      "step": 20675
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.15068700909614563,
      "learning_rate": 3.709927348452861e-06,
      "loss": 0.0037,
      "step": 20680
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.28642067313194275,
      "learning_rate": 3.7056961598731335e-06,
      "loss": 0.004,
      "step": 20685
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.09968727827072144,
      "learning_rate": 3.701466836708094e-06,
      "loss": 0.0032,
      "step": 20690
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.17942407727241516,
      "learning_rate": 3.697239380211167e-06,
      "loss": 0.0032,
      "step": 20695
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.21173350512981415,
      "learning_rate": 3.6930137916352258e-06,
      "loss": 0.0047,
      "step": 20700
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.2591868042945862,
      "learning_rate": 3.688790072232592e-06,
      "loss": 0.0043,
      "step": 20705
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.16367943584918976,
      "learning_rate": 3.684568223255024e-06,
      "loss": 0.0032,
      "step": 20710
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.3665323257446289,
      "learning_rate": 3.680348245953732e-06,
      "loss": 0.0024,
      "step": 20715
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.31445151567459106,
      "learning_rate": 3.6761301415793784e-06,
      "loss": 0.0032,
      "step": 20720
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.30521151423454285,
      "learning_rate": 3.671913911382061e-06,
      "loss": 0.0034,
      "step": 20725
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.4186500608921051,
      "learning_rate": 3.667699556611325e-06,
      "loss": 0.0034,
      "step": 20730
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.13586103916168213,
      "learning_rate": 3.6634870785161604e-06,
      "loss": 0.0058,
      "step": 20735
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.28456878662109375,
      "learning_rate": 3.659276478345002e-06,
      "loss": 0.0046,
      "step": 20740
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.7787228226661682,
      "learning_rate": 3.6550677573457218e-06,
      "loss": 0.0032,
      "step": 20745
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.4641948640346527,
      "learning_rate": 3.650860916765643e-06,
      "loss": 0.0044,
      "step": 20750
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.3779393136501312,
      "learning_rate": 3.6466559578515226e-06,
      "loss": 0.0041,
      "step": 20755
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.18350698053836823,
      "learning_rate": 3.6424528818495752e-06,
      "loss": 0.0039,
      "step": 20760
    },
    {
      "epoch": 6.48,
      "grad_norm": 0.18837334215641022,
      "learning_rate": 3.638251690005442e-06,
      "loss": 0.0041,
      "step": 20765
    },
    {
      "epoch": 6.48,
      "grad_norm": 0.1566227376461029,
      "learning_rate": 3.6340523835642114e-06,
      "loss": 0.0043,
      "step": 20770
    },
    {
      "epoch": 6.48,
      "grad_norm": 0.36541804671287537,
      "learning_rate": 3.6298549637704184e-06,
      "loss": 0.0041,
      "step": 20775
    },
    {
      "epoch": 6.48,
      "grad_norm": 0.20943257212638855,
      "learning_rate": 3.6256594318680247e-06,
      "loss": 0.0028,
      "step": 20780
    },
    {
      "epoch": 6.48,
      "grad_norm": 0.17792825400829315,
      "learning_rate": 3.621465789100447e-06,
      "loss": 0.0041,
      "step": 20785
    },
    {
      "epoch": 6.48,
      "grad_norm": 0.7338669300079346,
      "learning_rate": 3.617274036710534e-06,
      "loss": 0.0042,
      "step": 20790
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.19768068194389343,
      "learning_rate": 3.6130841759405776e-06,
      "loss": 0.0037,
      "step": 20795
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.21714939177036285,
      "learning_rate": 3.6088962080323055e-06,
      "loss": 0.0031,
      "step": 20800
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.2594071626663208,
      "learning_rate": 3.604710134226892e-06,
      "loss": 0.0036,
      "step": 20805
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.3518996834754944,
      "learning_rate": 3.6005259557649473e-06,
      "loss": 0.0032,
      "step": 20810
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.12475259602069855,
      "learning_rate": 3.59634367388651e-06,
      "loss": 0.003,
      "step": 20815
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.3000490367412567,
      "learning_rate": 3.5921632898310666e-06,
      "loss": 0.0039,
      "step": 20820
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.8753775954246521,
      "learning_rate": 3.58798480483754e-06,
      "loss": 0.0041,
      "step": 20825
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.2704598009586334,
      "learning_rate": 3.583808220144288e-06,
      "loss": 0.0054,
      "step": 20830
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.40405184030532837,
      "learning_rate": 3.5796335369891076e-06,
      "loss": 0.0032,
      "step": 20835
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.321799099445343,
      "learning_rate": 3.5754607566092313e-06,
      "loss": 0.0044,
      "step": 20840
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.3521104156970978,
      "learning_rate": 3.5712898802413266e-06,
      "loss": 0.006,
      "step": 20845
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.17451238632202148,
      "learning_rate": 3.5671209091214963e-06,
      "loss": 0.0025,
      "step": 20850
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.2520848512649536,
      "learning_rate": 3.562953844485282e-06,
      "loss": 0.0028,
      "step": 20855
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.41883954405784607,
      "learning_rate": 3.558788687567656e-06,
      "loss": 0.0046,
      "step": 20860
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.30903568863868713,
      "learning_rate": 3.554625439603029e-06,
      "loss": 0.0025,
      "step": 20865
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.6085941791534424,
      "learning_rate": 3.550464101825243e-06,
      "loss": 0.0053,
      "step": 20870
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.1391933262348175,
      "learning_rate": 3.5463046754675744e-06,
      "loss": 0.004,
      "step": 20875
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.22042496502399445,
      "learning_rate": 3.5421471617627356e-06,
      "loss": 0.0028,
      "step": 20880
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.2457202672958374,
      "learning_rate": 3.5379915619428697e-06,
      "loss": 0.0041,
      "step": 20885
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.07952205091714859,
      "learning_rate": 3.5338378772395532e-06,
      "loss": 0.002,
      "step": 20890
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.163711816072464,
      "learning_rate": 3.529686108883794e-06,
      "loss": 0.0022,
      "step": 20895
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.136203795671463,
      "learning_rate": 3.525536258106035e-06,
      "loss": 0.0026,
      "step": 20900
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.21723493933677673,
      "learning_rate": 3.521388326136148e-06,
      "loss": 0.0037,
      "step": 20905
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.5146577954292297,
      "learning_rate": 3.5180713629280194e-06,
      "loss": 0.0039,
      "step": 20910
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.271622896194458,
      "learning_rate": 3.5139268879097656e-06,
      "loss": 0.0048,
      "step": 20915
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.434649795293808,
      "learning_rate": 3.509784335140001e-06,
      "loss": 0.0049,
      "step": 20920
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.5571953058242798,
      "learning_rate": 3.505643705846432e-06,
      "loss": 0.005,
      "step": 20925
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.33921700716018677,
      "learning_rate": 3.5015050012562047e-06,
      "loss": 0.0034,
      "step": 20930
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.3331754505634308,
      "learning_rate": 3.497368222595886e-06,
      "loss": 0.005,
      "step": 20935
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.21557758748531342,
      "learning_rate": 3.493233371091477e-06,
      "loss": 0.0052,
      "step": 20940
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.23702630400657654,
      "learning_rate": 3.4891004479683986e-06,
      "loss": 0.003,
      "step": 20945
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.2295365035533905,
      "learning_rate": 3.484969454451511e-06,
      "loss": 0.0037,
      "step": 20950
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.41207176446914673,
      "learning_rate": 3.480840391765098e-06,
      "loss": 0.0046,
      "step": 20955
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.15661972761154175,
      "learning_rate": 3.47671326113287e-06,
      "loss": 0.004,
      "step": 20960
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.12692229449748993,
      "learning_rate": 3.472588063777964e-06,
      "loss": 0.0039,
      "step": 20965
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.11728105694055557,
      "learning_rate": 3.468464800922954e-06,
      "loss": 0.0028,
      "step": 20970
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.12148414552211761,
      "learning_rate": 3.4643434737898286e-06,
      "loss": 0.0047,
      "step": 20975
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.19293716549873352,
      "learning_rate": 3.460224083600009e-06,
      "loss": 0.0026,
      "step": 20980
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.22977282106876373,
      "learning_rate": 3.4561066315743365e-06,
      "loss": 0.0026,
      "step": 20985
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.30549490451812744,
      "learning_rate": 3.4519911189330834e-06,
      "loss": 0.0048,
      "step": 20990
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.2782987952232361,
      "learning_rate": 3.447877546895946e-06,
      "loss": 0.0042,
      "step": 20995
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.29214322566986084,
      "learning_rate": 3.4437659166820457e-06,
      "loss": 0.0048,
      "step": 21000
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.15185247361660004,
      "learning_rate": 3.439656229509929e-06,
      "loss": 0.0049,
      "step": 21005
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.14987649023532867,
      "learning_rate": 3.4355484865975596e-06,
      "loss": 0.0038,
      "step": 21010
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.2198924869298935,
      "learning_rate": 3.4314426891623432e-06,
      "loss": 0.0043,
      "step": 21015
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.37174761295318604,
      "learning_rate": 3.4273388384210858e-06,
      "loss": 0.0039,
      "step": 21020
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.38870739936828613,
      "learning_rate": 3.423236935590031e-06,
      "loss": 0.0034,
      "step": 21025
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.2492036074399948,
      "learning_rate": 3.4191369818848396e-06,
      "loss": 0.0033,
      "step": 21030
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.17112156748771667,
      "learning_rate": 3.4150389785205996e-06,
      "loss": 0.0048,
      "step": 21035
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.2990744113922119,
      "learning_rate": 3.4109429267118155e-06,
      "loss": 0.0042,
      "step": 21040
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.3892471492290497,
      "learning_rate": 3.4068488276724166e-06,
      "loss": 0.005,
      "step": 21045
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.26357969641685486,
      "learning_rate": 3.402756682615752e-06,
      "loss": 0.0036,
      "step": 21050
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.19580000638961792,
      "learning_rate": 3.3986664927545943e-06,
      "loss": 0.0046,
      "step": 21055
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.2106107473373413,
      "learning_rate": 3.3945782593011334e-06,
      "loss": 0.0041,
      "step": 21060
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.16694635152816772,
      "learning_rate": 3.3904919834669813e-06,
      "loss": 0.0031,
      "step": 21065
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.08712832629680634,
      "learning_rate": 3.3864076664631694e-06,
      "loss": 0.0022,
      "step": 21070
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.12394944578409195,
      "learning_rate": 3.3823253095001476e-06,
      "loss": 0.0023,
      "step": 21075
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.18365685641765594,
      "learning_rate": 3.378244913787787e-06,
      "loss": 0.0034,
      "step": 21080
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.05024956166744232,
      "learning_rate": 3.3741664805353746e-06,
      "loss": 0.0044,
      "step": 21085
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.15042851865291595,
      "learning_rate": 3.3700900109516188e-06,
      "loss": 0.0029,
      "step": 21090
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.27349185943603516,
      "learning_rate": 3.366015506244644e-06,
      "loss": 0.0033,
      "step": 21095
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.25597962737083435,
      "learning_rate": 3.3619429676219937e-06,
      "loss": 0.0051,
      "step": 21100
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.11255136132240295,
      "learning_rate": 3.357872396290628e-06,
      "loss": 0.0044,
      "step": 21105
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.2635140120983124,
      "learning_rate": 3.3538037934569223e-06,
      "loss": 0.0043,
      "step": 21110
    },
    {
      "epoch": 6.59,
      "grad_norm": 0.35105404257774353,
      "learning_rate": 3.3497371603266714e-06,
      "loss": 0.0053,
      "step": 21115
    },
    {
      "epoch": 6.59,
      "grad_norm": 0.12501968443393707,
      "learning_rate": 3.345672498105088e-06,
      "loss": 0.0023,
      "step": 21120
    },
    {
      "epoch": 6.59,
      "grad_norm": 0.24185575544834137,
      "learning_rate": 3.341609807996793e-06,
      "loss": 0.0048,
      "step": 21125
    },
    {
      "epoch": 6.59,
      "grad_norm": 0.5106483101844788,
      "learning_rate": 3.337549091205826e-06,
      "loss": 0.0033,
      "step": 21130
    },
    {
      "epoch": 6.59,
      "grad_norm": 0.2359280288219452,
      "learning_rate": 3.3334903489356496e-06,
      "loss": 0.0022,
      "step": 21135
    },
    {
      "epoch": 6.59,
      "grad_norm": 0.2506748139858246,
      "learning_rate": 3.329433582389133e-06,
      "loss": 0.0033,
      "step": 21140
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.1529313325881958,
      "learning_rate": 3.3253787927685622e-06,
      "loss": 0.0027,
      "step": 21145
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.18696674704551697,
      "learning_rate": 3.321325981275636e-06,
      "loss": 0.0053,
      "step": 21150
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.6021310091018677,
      "learning_rate": 3.3172751491114686e-06,
      "loss": 0.0055,
      "step": 21155
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.3504422903060913,
      "learning_rate": 3.3132262974765906e-06,
      "loss": 0.0034,
      "step": 21160
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.15375612676143646,
      "learning_rate": 3.3091794275709345e-06,
      "loss": 0.0032,
      "step": 21165
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.12261166423559189,
      "learning_rate": 3.305134540593854e-06,
      "loss": 0.0026,
      "step": 21170
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.14208921790122986,
      "learning_rate": 3.3010916377441207e-06,
      "loss": 0.0032,
      "step": 21175
    },
    {
      "epoch": 6.61,
      "grad_norm": 0.47172245383262634,
      "learning_rate": 3.2970507202199077e-06,
      "loss": 0.0047,
      "step": 21180
    },
    {
      "epoch": 6.61,
      "grad_norm": 0.2029520571231842,
      "learning_rate": 3.293011789218803e-06,
      "loss": 0.0034,
      "step": 21185
    },
    {
      "epoch": 6.61,
      "grad_norm": 0.2652159333229065,
      "learning_rate": 3.288974845937809e-06,
      "loss": 0.003,
      "step": 21190
    },
    {
      "epoch": 6.61,
      "grad_norm": 0.22905781865119934,
      "learning_rate": 3.2849398915733376e-06,
      "loss": 0.0038,
      "step": 21195
    },
    {
      "epoch": 6.61,
      "grad_norm": 0.20727509260177612,
      "learning_rate": 3.2809069273212047e-06,
      "loss": 0.0036,
      "step": 21200
    },
    {
      "epoch": 6.61,
      "grad_norm": 0.12545184791088104,
      "learning_rate": 3.276875954376644e-06,
      "loss": 0.0024,
      "step": 21205
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.13865812122821808,
      "learning_rate": 3.2728469739342984e-06,
      "loss": 0.0035,
      "step": 21210
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.17256861925125122,
      "learning_rate": 3.2688199871882155e-06,
      "loss": 0.0032,
      "step": 21215
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.24499864876270294,
      "learning_rate": 3.26479499533186e-06,
      "loss": 0.0037,
      "step": 21220
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.47705909609794617,
      "learning_rate": 3.2607719995580998e-06,
      "loss": 0.004,
      "step": 21225
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.48891186714172363,
      "learning_rate": 3.256751001059214e-06,
      "loss": 0.0036,
      "step": 21230
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.19393102824687958,
      "learning_rate": 3.2527320010268816e-06,
      "loss": 0.0043,
      "step": 21235
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.1835356056690216,
      "learning_rate": 3.2487150006521994e-06,
      "loss": 0.0036,
      "step": 21240
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.26606136560440063,
      "learning_rate": 3.244700001125668e-06,
      "loss": 0.0038,
      "step": 21245
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.41781163215637207,
      "learning_rate": 3.2406870036371942e-06,
      "loss": 0.0033,
      "step": 21250
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.20610876381397247,
      "learning_rate": 3.2366760093760885e-06,
      "loss": 0.0037,
      "step": 21255
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.04123732075095177,
      "learning_rate": 3.232667019531078e-06,
      "loss": 0.0022,
      "step": 21260
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.1551009863615036,
      "learning_rate": 3.228660035290291e-06,
      "loss": 0.0029,
      "step": 21265
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.32036370038986206,
      "learning_rate": 3.2246550578412517e-06,
      "loss": 0.0051,
      "step": 21270
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.22211022675037384,
      "learning_rate": 3.220652088370899e-06,
      "loss": 0.0042,
      "step": 21275
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.1752278208732605,
      "learning_rate": 3.2166511280655775e-06,
      "loss": 0.0026,
      "step": 21280
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.23283261060714722,
      "learning_rate": 3.212652178111033e-06,
      "loss": 0.0049,
      "step": 21285
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.07854650169610977,
      "learning_rate": 3.2086552396924166e-06,
      "loss": 0.004,
      "step": 21290
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.1206623762845993,
      "learning_rate": 3.2046603139942823e-06,
      "loss": 0.0037,
      "step": 21295
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.16894768178462982,
      "learning_rate": 3.2006674022005857e-06,
      "loss": 0.0024,
      "step": 21300
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.2977941036224365,
      "learning_rate": 3.196676505494698e-06,
      "loss": 0.0031,
      "step": 21305
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.9024665951728821,
      "learning_rate": 3.192687625059373e-06,
      "loss": 0.0039,
      "step": 21310
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.41954556107521057,
      "learning_rate": 3.188700762076782e-06,
      "loss": 0.0037,
      "step": 21315
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.6290353536605835,
      "learning_rate": 3.1847159177284926e-06,
      "loss": 0.0047,
      "step": 21320
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.17991706728935242,
      "learning_rate": 3.180733093195476e-06,
      "loss": 0.0052,
      "step": 21325
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.2722851634025574,
      "learning_rate": 3.176752289658104e-06,
      "loss": 0.0024,
      "step": 21330
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.2850939929485321,
      "learning_rate": 3.1727735082961485e-06,
      "loss": 0.0039,
      "step": 21335
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.08436838537454605,
      "learning_rate": 3.168796750288784e-06,
      "loss": 0.0028,
      "step": 21340
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.5143541693687439,
      "learning_rate": 3.164822016814585e-06,
      "loss": 0.004,
      "step": 21345
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.08529569953680038,
      "learning_rate": 3.1608493090515245e-06,
      "loss": 0.0039,
      "step": 21350
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.36727142333984375,
      "learning_rate": 3.156878628176977e-06,
      "loss": 0.0058,
      "step": 21355
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.2226283848285675,
      "learning_rate": 3.152909975367715e-06,
      "loss": 0.004,
      "step": 21360
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.26580193638801575,
      "learning_rate": 3.1489433517999102e-06,
      "loss": 0.0045,
      "step": 21365
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.08652780950069427,
      "learning_rate": 3.144978758649133e-06,
      "loss": 0.0046,
      "step": 21370
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.2908978760242462,
      "learning_rate": 3.141016197090353e-06,
      "loss": 0.0033,
      "step": 21375
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.13436950743198395,
      "learning_rate": 3.137055668297936e-06,
      "loss": 0.0023,
      "step": 21380
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.27255043387413025,
      "learning_rate": 3.133097173445646e-06,
      "loss": 0.0033,
      "step": 21385
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.2276810258626938,
      "learning_rate": 3.1291407137066453e-06,
      "loss": 0.0048,
      "step": 21390
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.29353004693984985,
      "learning_rate": 3.1251862902534926e-06,
      "loss": 0.0025,
      "step": 21395
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.30397000908851624,
      "learning_rate": 3.121233904258141e-06,
      "loss": 0.0032,
      "step": 21400
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.30446457862854004,
      "learning_rate": 3.1172835568919425e-06,
      "loss": 0.0051,
      "step": 21405
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.1925806999206543,
      "learning_rate": 3.1133352493256475e-06,
      "loss": 0.0023,
      "step": 21410
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.26819539070129395,
      "learning_rate": 3.109388982729391e-06,
      "loss": 0.003,
      "step": 21415
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.42312636971473694,
      "learning_rate": 3.1054447582727098e-06,
      "loss": 0.0037,
      "step": 21420
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.23445560038089752,
      "learning_rate": 3.1015025771245446e-06,
      "loss": 0.0024,
      "step": 21425
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.47701147198677063,
      "learning_rate": 3.0975624404532167e-06,
      "loss": 0.0042,
      "step": 21430
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.24110835790634155,
      "learning_rate": 3.0936243494264485e-06,
      "loss": 0.0042,
      "step": 21435
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.14478828012943268,
      "learning_rate": 3.0896883052113526e-06,
      "loss": 0.0031,
      "step": 21440
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.19179287552833557,
      "learning_rate": 3.0857543089744424e-06,
      "loss": 0.003,
      "step": 21445
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.13948087394237518,
      "learning_rate": 3.08182236188161e-06,
      "loss": 0.0021,
      "step": 21450
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.17611132562160492,
      "learning_rate": 3.0778924650981536e-06,
      "loss": 0.0034,
      "step": 21455
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.2598230540752411,
      "learning_rate": 3.073964619788755e-06,
      "loss": 0.0028,
      "step": 21460
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.21520379185676575,
      "learning_rate": 3.0700388271174985e-06,
      "loss": 0.0027,
      "step": 21465
    },
    {
      "epoch": 6.7,
      "grad_norm": 0.6616179347038269,
      "learning_rate": 3.066115088247852e-06,
      "loss": 0.0032,
      "step": 21470
    },
    {
      "epoch": 6.7,
      "grad_norm": 0.15049836039543152,
      "learning_rate": 3.062193404342676e-06,
      "loss": 0.002,
      "step": 21475
    },
    {
      "epoch": 6.7,
      "grad_norm": 0.3814724087715149,
      "learning_rate": 3.058273776564221e-06,
      "loss": 0.004,
      "step": 21480
    },
    {
      "epoch": 6.7,
      "grad_norm": 0.39669278264045715,
      "learning_rate": 3.0543562060741337e-06,
      "loss": 0.0047,
      "step": 21485
    },
    {
      "epoch": 6.7,
      "grad_norm": 0.09728311747312546,
      "learning_rate": 3.0504406940334418e-06,
      "loss": 0.0039,
      "step": 21490
    },
    {
      "epoch": 6.7,
      "grad_norm": 0.19073884189128876,
      "learning_rate": 3.0465272416025695e-06,
      "loss": 0.003,
      "step": 21495
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.20341838896274567,
      "learning_rate": 3.0426158499413294e-06,
      "loss": 0.0032,
      "step": 21500
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.25398117303848267,
      "learning_rate": 3.0387065202089204e-06,
      "loss": 0.0034,
      "step": 21505
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.2423587143421173,
      "learning_rate": 3.034799253563939e-06,
      "loss": 0.0034,
      "step": 21510
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.2504832446575165,
      "learning_rate": 3.0308940511643627e-06,
      "loss": 0.0039,
      "step": 21515
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.07413171231746674,
      "learning_rate": 3.026990914167559e-06,
      "loss": 0.0029,
      "step": 21520
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.4909358322620392,
      "learning_rate": 3.02308984373028e-06,
      "loss": 0.003,
      "step": 21525
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.319919615983963,
      "learning_rate": 3.019190841008671e-06,
      "loss": 0.0033,
      "step": 21530
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.2074088305234909,
      "learning_rate": 3.01529390715826e-06,
      "loss": 0.0042,
      "step": 21535
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.24591241776943207,
      "learning_rate": 3.011399043333967e-06,
      "loss": 0.0032,
      "step": 21540
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.33461540937423706,
      "learning_rate": 3.007506250690092e-06,
      "loss": 0.0038,
      "step": 21545
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.27527114748954773,
      "learning_rate": 3.003615530380324e-06,
      "loss": 0.0024,
      "step": 21550
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.10472951829433441,
      "learning_rate": 2.9997268835577454e-06,
      "loss": 0.0031,
      "step": 21555
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.22347362339496613,
      "learning_rate": 2.9958403113748104e-06,
      "loss": 0.0047,
      "step": 21560
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.3212512135505676,
      "learning_rate": 2.991955814983367e-06,
      "loss": 0.0057,
      "step": 21565
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.1527688354253769,
      "learning_rate": 2.988073395534645e-06,
      "loss": 0.0027,
      "step": 21570
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.10932507365942001,
      "learning_rate": 2.984193054179262e-06,
      "loss": 0.0034,
      "step": 21575
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.3322175443172455,
      "learning_rate": 2.980314792067215e-06,
      "loss": 0.0048,
      "step": 21580
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.2838209271430969,
      "learning_rate": 2.976438610347887e-06,
      "loss": 0.0058,
      "step": 21585
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.266382098197937,
      "learning_rate": 2.9725645101700473e-06,
      "loss": 0.0045,
      "step": 21590
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.1659298539161682,
      "learning_rate": 2.968692492681844e-06,
      "loss": 0.0038,
      "step": 21595
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.19411054253578186,
      "learning_rate": 2.96482255903081e-06,
      "loss": 0.0044,
      "step": 21600
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.28912949562072754,
      "learning_rate": 2.9609547103638603e-06,
      "loss": 0.0032,
      "step": 21605
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.3964943289756775,
      "learning_rate": 2.9570889478272914e-06,
      "loss": 0.0043,
      "step": 21610
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.18239091336727142,
      "learning_rate": 2.953225272566782e-06,
      "loss": 0.0023,
      "step": 21615
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.21042893826961517,
      "learning_rate": 2.949363685727392e-06,
      "loss": 0.0044,
      "step": 21620
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.3126097023487091,
      "learning_rate": 2.945504188453564e-06,
      "loss": 0.0021,
      "step": 21625
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.2078578770160675,
      "learning_rate": 2.941646781889119e-06,
      "loss": 0.0034,
      "step": 21630
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.29286807775497437,
      "learning_rate": 2.93779146717726e-06,
      "loss": 0.0042,
      "step": 21635
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.34374094009399414,
      "learning_rate": 2.933938245460568e-06,
      "loss": 0.0044,
      "step": 21640
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.33126339316368103,
      "learning_rate": 2.9300871178810065e-06,
      "loss": 0.0045,
      "step": 21645
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.29246968030929565,
      "learning_rate": 2.9262380855799165e-06,
      "loss": 0.0029,
      "step": 21650
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.7621261477470398,
      "learning_rate": 2.9223911496980184e-06,
      "loss": 0.0027,
      "step": 21655
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.2603248655796051,
      "learning_rate": 2.918546311375411e-06,
      "loss": 0.003,
      "step": 21660
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.3011554777622223,
      "learning_rate": 2.9147035717515725e-06,
      "loss": 0.0036,
      "step": 21665
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.27007898688316345,
      "learning_rate": 2.910862931965358e-06,
      "loss": 0.004,
      "step": 21670
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.2060871124267578,
      "learning_rate": 2.9070243931550013e-06,
      "loss": 0.0032,
      "step": 21675
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.6070762872695923,
      "learning_rate": 2.9031879564581135e-06,
      "loss": 0.006,
      "step": 21680
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.3831270933151245,
      "learning_rate": 2.899353623011679e-06,
      "loss": 0.0042,
      "step": 21685
    },
    {
      "epoch": 6.77,
      "grad_norm": 0.2492065280675888,
      "learning_rate": 2.8955213939520664e-06,
      "loss": 0.003,
      "step": 21690
    },
    {
      "epoch": 6.77,
      "grad_norm": 0.456470787525177,
      "learning_rate": 2.8916912704150136e-06,
      "loss": 0.0037,
      "step": 21695
    },
    {
      "epoch": 6.77,
      "grad_norm": 0.2670809328556061,
      "learning_rate": 2.887863253535641e-06,
      "loss": 0.0038,
      "step": 21700
    },
    {
      "epoch": 6.77,
      "grad_norm": 0.3004983961582184,
      "learning_rate": 2.8840373444484347e-06,
      "loss": 0.0032,
      "step": 21705
    },
    {
      "epoch": 6.77,
      "grad_norm": 0.18564684689044952,
      "learning_rate": 2.8802135442872627e-06,
      "loss": 0.0036,
      "step": 21710
    },
    {
      "epoch": 6.77,
      "grad_norm": 0.4440821409225464,
      "learning_rate": 2.876391854185373e-06,
      "loss": 0.0046,
      "step": 21715
    },
    {
      "epoch": 6.77,
      "grad_norm": 0.2004132866859436,
      "learning_rate": 2.8725722752753795e-06,
      "loss": 0.0042,
      "step": 21720
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.3126099109649658,
      "learning_rate": 2.8687548086892734e-06,
      "loss": 0.0041,
      "step": 21725
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.3466210961341858,
      "learning_rate": 2.864939455558423e-06,
      "loss": 0.0026,
      "step": 21730
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.18223507702350616,
      "learning_rate": 2.8611262170135666e-06,
      "loss": 0.0035,
      "step": 21735
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.5391222834587097,
      "learning_rate": 2.8573150941848138e-06,
      "loss": 0.0043,
      "step": 21740
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.30332350730895996,
      "learning_rate": 2.853506088201652e-06,
      "loss": 0.003,
      "step": 21745
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.20412799715995789,
      "learning_rate": 2.8496992001929347e-06,
      "loss": 0.0034,
      "step": 21750
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.5126622319221497,
      "learning_rate": 2.8458944312868998e-06,
      "loss": 0.0035,
      "step": 21755
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.25375452637672424,
      "learning_rate": 2.8420917826111473e-06,
      "loss": 0.0042,
      "step": 21760
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.21945394575595856,
      "learning_rate": 2.8382912552926488e-06,
      "loss": 0.003,
      "step": 21765
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.33257466554641724,
      "learning_rate": 2.8344928504577563e-06,
      "loss": 0.0034,
      "step": 21770
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.3007264733314514,
      "learning_rate": 2.8306965692321764e-06,
      "loss": 0.0038,
      "step": 21775
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.3190644085407257,
      "learning_rate": 2.8269024127410004e-06,
      "loss": 0.0032,
      "step": 21780
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.13521669805049896,
      "learning_rate": 2.8231103821086856e-06,
      "loss": 0.0029,
      "step": 21785
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.25221729278564453,
      "learning_rate": 2.81932047845906e-06,
      "loss": 0.0045,
      "step": 21790
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.21736393868923187,
      "learning_rate": 2.8155327029153155e-06,
      "loss": 0.0029,
      "step": 21795
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.1783631443977356,
      "learning_rate": 2.811747056600026e-06,
      "loss": 0.0037,
      "step": 21800
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.45112115144729614,
      "learning_rate": 2.8079635406351226e-06,
      "loss": 0.0039,
      "step": 21805
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.2695203423500061,
      "learning_rate": 2.8041821561419126e-06,
      "loss": 0.004,
      "step": 21810
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.19870398938655853,
      "learning_rate": 2.8004029042410608e-06,
      "loss": 0.0021,
      "step": 21815
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.5985007882118225,
      "learning_rate": 2.7966257860526115e-06,
      "loss": 0.0035,
      "step": 21820
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.23408366739749908,
      "learning_rate": 2.792850802695971e-06,
      "loss": 0.0038,
      "step": 21825
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.19258131086826324,
      "learning_rate": 2.789077955289916e-06,
      "loss": 0.0036,
      "step": 21830
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.28733906149864197,
      "learning_rate": 2.7853072449525875e-06,
      "loss": 0.0041,
      "step": 21835
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.07309596240520477,
      "learning_rate": 2.78153867280149e-06,
      "loss": 0.0033,
      "step": 21840
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.2608742117881775,
      "learning_rate": 2.77777223995351e-06,
      "loss": 0.0038,
      "step": 21845
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.29153764247894287,
      "learning_rate": 2.7740079475248773e-06,
      "loss": 0.0021,
      "step": 21850
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.3125184178352356,
      "learning_rate": 2.770245796631201e-06,
      "loss": 0.0038,
      "step": 21855
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.2416171431541443,
      "learning_rate": 2.766485788387455e-06,
      "loss": 0.0042,
      "step": 21860
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.07885327935218811,
      "learning_rate": 2.762727923907974e-06,
      "loss": 0.0036,
      "step": 21865
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.2367239147424698,
      "learning_rate": 2.7589722043064613e-06,
      "loss": 0.0037,
      "step": 21870
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.2745641767978668,
      "learning_rate": 2.7552186306959815e-06,
      "loss": 0.0046,
      "step": 21875
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.5628076195716858,
      "learning_rate": 2.7514672041889657e-06,
      "loss": 0.0032,
      "step": 21880
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.2814357578754425,
      "learning_rate": 2.7477179258972053e-06,
      "loss": 0.0035,
      "step": 21885
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.2630438208580017,
      "learning_rate": 2.7439707969318585e-06,
      "loss": 0.0037,
      "step": 21890
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.3329617381095886,
      "learning_rate": 2.740225818403446e-06,
      "loss": 0.0032,
      "step": 21895
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.1600303202867508,
      "learning_rate": 2.7364829914218493e-06,
      "loss": 0.0027,
      "step": 21900
    },
    {
      "epoch": 6.83,
      "grad_norm": 1.125551462173462,
      "learning_rate": 2.732742317096313e-06,
      "loss": 0.006,
      "step": 21905
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.2530844509601593,
      "learning_rate": 2.7290037965354445e-06,
      "loss": 0.0038,
      "step": 21910
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.18102207779884338,
      "learning_rate": 2.7252674308472115e-06,
      "loss": 0.0036,
      "step": 21915
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.678311824798584,
      "learning_rate": 2.7215332211389454e-06,
      "loss": 0.0048,
      "step": 21920
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.24131645262241364,
      "learning_rate": 2.717801168517337e-06,
      "loss": 0.0049,
      "step": 21925
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.1348024308681488,
      "learning_rate": 2.714071274088438e-06,
      "loss": 0.004,
      "step": 21930
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.18559803068637848,
      "learning_rate": 2.710343538957659e-06,
      "loss": 0.0046,
      "step": 21935
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.25179290771484375,
      "learning_rate": 2.706617964229775e-06,
      "loss": 0.0041,
      "step": 21940
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.15272609889507294,
      "learning_rate": 2.7028945510089157e-06,
      "loss": 0.0046,
      "step": 21945
    },
    {
      "epoch": 6.85,
      "grad_norm": 0.12700574100017548,
      "learning_rate": 2.6991733003985752e-06,
      "loss": 0.0031,
      "step": 21950
    },
    {
      "epoch": 6.85,
      "grad_norm": 0.1488189995288849,
      "learning_rate": 2.695454213501597e-06,
      "loss": 0.0021,
      "step": 21955
    },
    {
      "epoch": 6.85,
      "grad_norm": 0.27931275963783264,
      "learning_rate": 2.6917372914201976e-06,
      "loss": 0.0042,
      "step": 21960
    },
    {
      "epoch": 6.85,
      "grad_norm": 0.28481337428092957,
      "learning_rate": 2.688022535255942e-06,
      "loss": 0.0049,
      "step": 21965
    },
    {
      "epoch": 6.85,
      "grad_norm": 0.39484626054763794,
      "learning_rate": 2.6843099461097566e-06,
      "loss": 0.0041,
      "step": 21970
    },
    {
      "epoch": 6.85,
      "grad_norm": 0.11301475763320923,
      "learning_rate": 2.680599525081924e-06,
      "loss": 0.003,
      "step": 21975
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.32207757234573364,
      "learning_rate": 2.676891273272084e-06,
      "loss": 0.0056,
      "step": 21980
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.20087608695030212,
      "learning_rate": 2.673185191779237e-06,
      "loss": 0.0041,
      "step": 21985
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.10096064954996109,
      "learning_rate": 2.669481281701739e-06,
      "loss": 0.0039,
      "step": 21990
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.13234908878803253,
      "learning_rate": 2.6657795441372967e-06,
      "loss": 0.0031,
      "step": 21995
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.4627056419849396,
      "learning_rate": 2.6620799801829766e-06,
      "loss": 0.0042,
      "step": 22000
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.28749316930770874,
      "learning_rate": 2.658382590935208e-06,
      "loss": 0.0029,
      "step": 22005
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.21665598452091217,
      "learning_rate": 2.6546873774897663e-06,
      "loss": 0.0033,
      "step": 22010
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.33342960476875305,
      "learning_rate": 2.6509943409417858e-06,
      "loss": 0.0032,
      "step": 22015
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.32708919048309326,
      "learning_rate": 2.647303482385756e-06,
      "loss": 0.0043,
      "step": 22020
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.20584945380687714,
      "learning_rate": 2.6436148029155216e-06,
      "loss": 0.0029,
      "step": 22025
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.1696053147315979,
      "learning_rate": 2.639928303624275e-06,
      "loss": 0.0038,
      "step": 22030
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.07086633145809174,
      "learning_rate": 2.6362439856045707e-06,
      "loss": 0.0026,
      "step": 22035
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.20959654450416565,
      "learning_rate": 2.632561849948313e-06,
      "loss": 0.0035,
      "step": 22040
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.4945712685585022,
      "learning_rate": 2.6288818977467568e-06,
      "loss": 0.004,
      "step": 22045
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.21944449841976166,
      "learning_rate": 2.6252041300905197e-06,
      "loss": 0.0034,
      "step": 22050
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.2166629284620285,
      "learning_rate": 2.6215285480695618e-06,
      "loss": 0.0031,
      "step": 22055
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.2644397020339966,
      "learning_rate": 2.617855152773202e-06,
      "loss": 0.0046,
      "step": 22060
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.08215432614088058,
      "learning_rate": 2.614183945290103e-06,
      "loss": 0.0034,
      "step": 22065
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.09645352512598038,
      "learning_rate": 2.610514926708285e-06,
      "loss": 0.0033,
      "step": 22070
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.49109041690826416,
      "learning_rate": 2.606848098115121e-06,
      "loss": 0.0034,
      "step": 22075
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.3283572494983673,
      "learning_rate": 2.6031834605973315e-06,
      "loss": 0.0033,
      "step": 22080
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.10526677221059799,
      "learning_rate": 2.5995210152409855e-06,
      "loss": 0.0033,
      "step": 22085
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.4159996509552002,
      "learning_rate": 2.5958607631315113e-06,
      "loss": 0.0031,
      "step": 22090
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.42396876215934753,
      "learning_rate": 2.5922027053536823e-06,
      "loss": 0.0039,
      "step": 22095
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.6997884511947632,
      "learning_rate": 2.588546842991615e-06,
      "loss": 0.0041,
      "step": 22100
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.1890052706003189,
      "learning_rate": 2.584893177128782e-06,
      "loss": 0.0035,
      "step": 22105
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.3274327516555786,
      "learning_rate": 2.5812417088480067e-06,
      "loss": 0.0037,
      "step": 22110
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.12549816071987152,
      "learning_rate": 2.5775924392314567e-06,
      "loss": 0.0031,
      "step": 22115
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.21259631216526031,
      "learning_rate": 2.57394536936065e-06,
      "loss": 0.0034,
      "step": 22120
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.13228316605091095,
      "learning_rate": 2.570300500316452e-06,
      "loss": 0.003,
      "step": 22125
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.3453643023967743,
      "learning_rate": 2.5666578331790735e-06,
      "loss": 0.0044,
      "step": 22130
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.29129546880722046,
      "learning_rate": 2.5630173690280856e-06,
      "loss": 0.0062,
      "step": 22135
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.4494245946407318,
      "learning_rate": 2.559379108942386e-06,
      "loss": 0.0049,
      "step": 22140
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.2582186460494995,
      "learning_rate": 2.555743054000234e-06,
      "loss": 0.0044,
      "step": 22145
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.24486398696899414,
      "learning_rate": 2.5521092052792286e-06,
      "loss": 0.0033,
      "step": 22150
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.1669885218143463,
      "learning_rate": 2.548477563856319e-06,
      "loss": 0.003,
      "step": 22155
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.3093135952949524,
      "learning_rate": 2.5448481308077976e-06,
      "loss": 0.0041,
      "step": 22160
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.3074520528316498,
      "learning_rate": 2.541220907209304e-06,
      "loss": 0.0036,
      "step": 22165
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.1470278799533844,
      "learning_rate": 2.5375958941358224e-06,
      "loss": 0.0031,
      "step": 22170
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.2791728377342224,
      "learning_rate": 2.5339730926616823e-06,
      "loss": 0.0048,
      "step": 22175
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.1116819679737091,
      "learning_rate": 2.530352503860556e-06,
      "loss": 0.0025,
      "step": 22180
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.2816881835460663,
      "learning_rate": 2.52673412880546e-06,
      "loss": 0.0032,
      "step": 22185
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.35410192608833313,
      "learning_rate": 2.523117968568759e-06,
      "loss": 0.0044,
      "step": 22190
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.3361516296863556,
      "learning_rate": 2.519504024222156e-06,
      "loss": 0.0038,
      "step": 22195
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.3186194598674774,
      "learning_rate": 2.5158922968366995e-06,
      "loss": 0.0038,
      "step": 22200
    },
    {
      "epoch": 6.93,
      "grad_norm": 0.156856507062912,
      "learning_rate": 2.5122827874827825e-06,
      "loss": 0.004,
      "step": 22205
    },
    {
      "epoch": 6.93,
      "grad_norm": 0.07649488747119904,
      "learning_rate": 2.5086754972301386e-06,
      "loss": 0.0028,
      "step": 22210
    },
    {
      "epoch": 6.93,
      "grad_norm": 0.4045189917087555,
      "learning_rate": 2.505070427147843e-06,
      "loss": 0.0042,
      "step": 22215
    },
    {
      "epoch": 6.93,
      "grad_norm": 0.41002410650253296,
      "learning_rate": 2.5014675783043165e-06,
      "loss": 0.0048,
      "step": 22220
    },
    {
      "epoch": 6.93,
      "grad_norm": 0.3689776062965393,
      "learning_rate": 2.497866951767317e-06,
      "loss": 0.0056,
      "step": 22225
    },
    {
      "epoch": 6.93,
      "grad_norm": 0.3803596496582031,
      "learning_rate": 2.4942685486039487e-06,
      "loss": 0.0023,
      "step": 22230
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.08025051653385162,
      "learning_rate": 2.490672369880651e-06,
      "loss": 0.004,
      "step": 22235
    },
    {
      "epoch": 6.94,
      "grad_norm": 1.6498674154281616,
      "learning_rate": 2.4870784166632123e-06,
      "loss": 0.0066,
      "step": 22240
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.2917429506778717,
      "learning_rate": 2.4834866900167478e-06,
      "loss": 0.0032,
      "step": 22245
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.4845334589481354,
      "learning_rate": 2.479897191005729e-06,
      "loss": 0.004,
      "step": 22250
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.2840844988822937,
      "learning_rate": 2.476309920693958e-06,
      "loss": 0.004,
      "step": 22255
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.19367815554141998,
      "learning_rate": 2.4727248801445768e-06,
      "loss": 0.0047,
      "step": 22260
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.20746980607509613,
      "learning_rate": 2.4691420704200696e-06,
      "loss": 0.0024,
      "step": 22265
    },
    {
      "epoch": 6.95,
      "grad_norm": 0.5077010989189148,
      "learning_rate": 2.4655614925822558e-06,
      "loss": 0.0032,
      "step": 22270
    },
    {
      "epoch": 6.95,
      "grad_norm": 0.28518030047416687,
      "learning_rate": 2.4619831476922994e-06,
      "loss": 0.0045,
      "step": 22275
    },
    {
      "epoch": 6.95,
      "grad_norm": 0.19493155181407928,
      "learning_rate": 2.458407036810693e-06,
      "loss": 0.004,
      "step": 22280
    },
    {
      "epoch": 6.95,
      "grad_norm": 0.1684066355228424,
      "learning_rate": 2.454833160997273e-06,
      "loss": 0.0034,
      "step": 22285
    },
    {
      "epoch": 6.95,
      "grad_norm": 0.1775389462709427,
      "learning_rate": 2.4512615213112133e-06,
      "loss": 0.0037,
      "step": 22290
    },
    {
      "epoch": 6.95,
      "grad_norm": 0.4576650857925415,
      "learning_rate": 2.4476921188110293e-06,
      "loss": 0.0025,
      "step": 22295
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.35909825563430786,
      "learning_rate": 2.4441249545545653e-06,
      "loss": 0.0043,
      "step": 22300
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.23739157617092133,
      "learning_rate": 2.440560029599006e-06,
      "loss": 0.0027,
      "step": 22305
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.11318200081586838,
      "learning_rate": 2.4369973450008734e-06,
      "loss": 0.0033,
      "step": 22310
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.3762189745903015,
      "learning_rate": 2.4334369018160262e-06,
      "loss": 0.0037,
      "step": 22315
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.15604671835899353,
      "learning_rate": 2.4298787010996517e-06,
      "loss": 0.0033,
      "step": 22320
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.25424784421920776,
      "learning_rate": 2.42632274390628e-06,
      "loss": 0.0031,
      "step": 22325
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.38101381063461304,
      "learning_rate": 2.4227690312897754e-06,
      "loss": 0.0028,
      "step": 22330
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.25989487767219543,
      "learning_rate": 2.419217564303332e-06,
      "loss": 0.0037,
      "step": 22335
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.12496748566627502,
      "learning_rate": 2.41566834399949e-06,
      "loss": 0.0031,
      "step": 22340
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.28795716166496277,
      "learning_rate": 2.4121213714301096e-06,
      "loss": 0.0028,
      "step": 22345
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.13154681026935577,
      "learning_rate": 2.408576647646397e-06,
      "loss": 0.0032,
      "step": 22350
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.11087685078382492,
      "learning_rate": 2.4050341736988793e-06,
      "loss": 0.0042,
      "step": 22355
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.23004786670207977,
      "learning_rate": 2.4014939506374267e-06,
      "loss": 0.0043,
      "step": 22360
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.28975680470466614,
      "learning_rate": 2.3979559795112384e-06,
      "loss": 0.0048,
      "step": 22365
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.12339877337217331,
      "learning_rate": 2.394420261368848e-06,
      "loss": 0.0031,
      "step": 22370
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.15025168657302856,
      "learning_rate": 2.3908867972581186e-06,
      "loss": 0.0035,
      "step": 22375
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.3735179305076599,
      "learning_rate": 2.387355588226251e-06,
      "loss": 0.0035,
      "step": 22380
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.1667015254497528,
      "learning_rate": 2.383826635319775e-06,
      "loss": 0.0031,
      "step": 22385
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.17676593363285065,
      "learning_rate": 2.380299939584544e-06,
      "loss": 0.0038,
      "step": 22390
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.30443859100341797,
      "learning_rate": 2.3767755020657523e-06,
      "loss": 0.0026,
      "step": 22395
    },
    {
      "epoch": 6.99,
      "grad_norm": 0.3084462583065033,
      "learning_rate": 2.373253323807921e-06,
      "loss": 0.0029,
      "step": 22400
    },
    {
      "epoch": 6.99,
      "grad_norm": 0.32841742038726807,
      "learning_rate": 2.3697334058549047e-06,
      "loss": 0.0035,
      "step": 22405
    },
    {
      "epoch": 6.99,
      "grad_norm": 0.28747549653053284,
      "learning_rate": 2.3662157492498838e-06,
      "loss": 0.0049,
      "step": 22410
    },
    {
      "epoch": 6.99,
      "grad_norm": 0.15251532196998596,
      "learning_rate": 2.36270035503537e-06,
      "loss": 0.0041,
      "step": 22415
    },
    {
      "epoch": 6.99,
      "grad_norm": 0.1651524305343628,
      "learning_rate": 2.3591872242532066e-06,
      "loss": 0.0024,
      "step": 22420
    },
    {
      "epoch": 6.99,
      "grad_norm": 0.27897897362709045,
      "learning_rate": 2.355676357944563e-06,
      "loss": 0.004,
      "step": 22425
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2509746253490448,
      "learning_rate": 2.35216775714994e-06,
      "loss": 0.0043,
      "step": 22430
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.39022454619407654,
      "learning_rate": 2.3486614229091654e-06,
      "loss": 0.0048,
      "step": 22435
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2762365937232971,
      "learning_rate": 2.345157356261395e-06,
      "loss": 0.0035,
      "step": 22440
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1320386826992035,
      "learning_rate": 2.341655558245115e-06,
      "loss": 0.0041,
      "step": 22445
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.12403634190559387,
      "learning_rate": 2.338156029898138e-06,
      "loss": 0.0023,
      "step": 22450
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.10191764682531357,
      "learning_rate": 2.3346587722576007e-06,
      "loss": 0.0022,
      "step": 22455
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.08048174530267715,
      "learning_rate": 2.3311637863599724e-06,
      "loss": 0.002,
      "step": 22460
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.0919155478477478,
      "learning_rate": 2.327671073241046e-06,
      "loss": 0.0017,
      "step": 22465
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.08900649100542068,
      "learning_rate": 2.3241806339359397e-06,
      "loss": 0.0019,
      "step": 22470
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.17529134452342987,
      "learning_rate": 2.320692469479101e-06,
      "loss": 0.0024,
      "step": 22475
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.03381878510117531,
      "learning_rate": 2.317206580904301e-06,
      "loss": 0.0015,
      "step": 22480
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.09003644436597824,
      "learning_rate": 2.3137229692446373e-06,
      "loss": 0.0014,
      "step": 22485
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.18930363655090332,
      "learning_rate": 2.3102416355325307e-06,
      "loss": 0.0013,
      "step": 22490
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.05011576786637306,
      "learning_rate": 2.306762580799732e-06,
      "loss": 0.0013,
      "step": 22495
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.13884611427783966,
      "learning_rate": 2.3032858060773087e-06,
      "loss": 0.0013,
      "step": 22500
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.13446393609046936,
      "learning_rate": 2.2998113123956613e-06,
      "loss": 0.0021,
      "step": 22505
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.2293945550918579,
      "learning_rate": 2.296339100784506e-06,
      "loss": 0.0034,
      "step": 22510
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.09749463200569153,
      "learning_rate": 2.292869172272889e-06,
      "loss": 0.0013,
      "step": 22515
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.24164359271526337,
      "learning_rate": 2.289401527889178e-06,
      "loss": 0.0019,
      "step": 22520
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.0538080595433712,
      "learning_rate": 2.2859361686610617e-06,
      "loss": 0.0012,
      "step": 22525
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.08052002638578415,
      "learning_rate": 2.282473095615556e-06,
      "loss": 0.002,
      "step": 22530
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.020526913926005363,
      "learning_rate": 2.279012309778992e-06,
      "loss": 0.0012,
      "step": 22535
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.5815721154212952,
      "learning_rate": 2.275553812177026e-06,
      "loss": 0.0017,
      "step": 22540
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.08703240007162094,
      "learning_rate": 2.272097603834643e-06,
      "loss": 0.0025,
      "step": 22545
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.12789790332317352,
      "learning_rate": 2.268643685776143e-06,
      "loss": 0.0014,
      "step": 22550
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.15096785128116608,
      "learning_rate": 2.2651920590251463e-06,
      "loss": 0.0022,
      "step": 22555
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.09342913329601288,
      "learning_rate": 2.2617427246045976e-06,
      "loss": 0.0012,
      "step": 22560
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.0484846793115139,
      "learning_rate": 2.2582956835367632e-06,
      "loss": 0.0021,
      "step": 22565
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.04795683175325394,
      "learning_rate": 2.25485093684322e-06,
      "loss": 0.0019,
      "step": 22570
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.033159200102090836,
      "learning_rate": 2.251408485544877e-06,
      "loss": 0.0018,
      "step": 22575
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.22821782529354095,
      "learning_rate": 2.2479683306619537e-06,
      "loss": 0.0019,
      "step": 22580
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.12525643408298492,
      "learning_rate": 2.2445304732140005e-06,
      "loss": 0.0014,
      "step": 22585
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.1769990772008896,
      "learning_rate": 2.2410949142198766e-06,
      "loss": 0.0019,
      "step": 22590
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.19607287645339966,
      "learning_rate": 2.2376616546977635e-06,
      "loss": 0.0022,
      "step": 22595
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.06251411139965057,
      "learning_rate": 2.234230695665163e-06,
      "loss": 0.0017,
      "step": 22600
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.15792545676231384,
      "learning_rate": 2.2308020381388874e-06,
      "loss": 0.0014,
      "step": 22605
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.18448935449123383,
      "learning_rate": 2.2273756831350755e-06,
      "loss": 0.001,
      "step": 22610
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.04084780439734459,
      "learning_rate": 2.2239516316691812e-06,
      "loss": 0.0012,
      "step": 22615
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.21777600049972534,
      "learning_rate": 2.2205298847559762e-06,
      "loss": 0.0016,
      "step": 22620
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.22892186045646667,
      "learning_rate": 2.2171104434095436e-06,
      "loss": 0.0016,
      "step": 22625
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.36050963401794434,
      "learning_rate": 2.2136933086432956e-06,
      "loss": 0.0022,
      "step": 22630
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.0779469683766365,
      "learning_rate": 2.2102784814699483e-06,
      "loss": 0.0015,
      "step": 22635
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.17603763937950134,
      "learning_rate": 2.2068659629015433e-06,
      "loss": 0.0025,
      "step": 22640
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.15554018318653107,
      "learning_rate": 2.203455753949426e-06,
      "loss": 0.002,
      "step": 22645
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.06508053839206696,
      "learning_rate": 2.20004785562427e-06,
      "loss": 0.0019,
      "step": 22650
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.10713350772857666,
      "learning_rate": 2.196642268936057e-06,
      "loss": 0.0016,
      "step": 22655
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.15166723728179932,
      "learning_rate": 2.193238994894086e-06,
      "loss": 0.0015,
      "step": 22660
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.07224074751138687,
      "learning_rate": 2.1898380345069703e-06,
      "loss": 0.0015,
      "step": 22665
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.024913206696510315,
      "learning_rate": 2.1864393887826352e-06,
      "loss": 0.0015,
      "step": 22670
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.4481954276561737,
      "learning_rate": 2.183043058728329e-06,
      "loss": 0.003,
      "step": 22675
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.32068902254104614,
      "learning_rate": 2.179649045350599e-06,
      "loss": 0.0019,
      "step": 22680
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.07001259177923203,
      "learning_rate": 2.1762573496553176e-06,
      "loss": 0.0017,
      "step": 22685
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.11420892924070358,
      "learning_rate": 2.172867972647664e-06,
      "loss": 0.0022,
      "step": 22690
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.17854531109333038,
      "learning_rate": 2.169480915332136e-06,
      "loss": 0.0021,
      "step": 22695
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.21439093351364136,
      "learning_rate": 2.166096178712539e-06,
      "loss": 0.002,
      "step": 22700
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.1033504381775856,
      "learning_rate": 2.1627137637919916e-06,
      "loss": 0.0013,
      "step": 22705
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.059978220611810684,
      "learning_rate": 2.1593336715729264e-06,
      "loss": 0.0022,
      "step": 22710
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.12991735339164734,
      "learning_rate": 2.155955903057084e-06,
      "loss": 0.0018,
      "step": 22715
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.14945195615291595,
      "learning_rate": 2.1525804592455226e-06,
      "loss": 0.0017,
      "step": 22720
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.10099814832210541,
      "learning_rate": 2.149207341138603e-06,
      "loss": 0.0016,
      "step": 22725
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.04819072037935257,
      "learning_rate": 2.145836549736003e-06,
      "loss": 0.0018,
      "step": 22730
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.14807431399822235,
      "learning_rate": 2.1424680860367086e-06,
      "loss": 0.0018,
      "step": 22735
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.1759273111820221,
      "learning_rate": 2.1391019510390177e-06,
      "loss": 0.0015,
      "step": 22740
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.11430688947439194,
      "learning_rate": 2.135738145740536e-06,
      "loss": 0.0023,
      "step": 22745
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.09345373511314392,
      "learning_rate": 2.1323766711381776e-06,
      "loss": 0.0013,
      "step": 22750
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.1269146054983139,
      "learning_rate": 2.12901752822817e-06,
      "loss": 0.0027,
      "step": 22755
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.22538627684116364,
      "learning_rate": 2.1256607180060464e-06,
      "loss": 0.0015,
      "step": 22760
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.14450940489768982,
      "learning_rate": 2.122306241466651e-06,
      "loss": 0.0018,
      "step": 22765
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.06322424858808517,
      "learning_rate": 2.1189540996041314e-06,
      "loss": 0.0012,
      "step": 22770
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.14859558641910553,
      "learning_rate": 2.1156042934119514e-06,
      "loss": 0.0022,
      "step": 22775
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.08851836621761322,
      "learning_rate": 2.112256823882878e-06,
      "loss": 0.0015,
      "step": 22780
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.08810979872941971,
      "learning_rate": 2.108911692008978e-06,
      "loss": 0.0018,
      "step": 22785
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.3387254476547241,
      "learning_rate": 2.1055688987816414e-06,
      "loss": 0.0021,
      "step": 22790
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.02320532500743866,
      "learning_rate": 2.1022284451915554e-06,
      "loss": 0.0014,
      "step": 22795
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.09600692987442017,
      "learning_rate": 2.0988903322287137e-06,
      "loss": 0.0013,
      "step": 22800
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.41077396273612976,
      "learning_rate": 2.095554560882418e-06,
      "loss": 0.0041,
      "step": 22805
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.08172262459993362,
      "learning_rate": 2.092221132141278e-06,
      "loss": 0.0018,
      "step": 22810
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.16185802221298218,
      "learning_rate": 2.088890046993204e-06,
      "loss": 0.0015,
      "step": 22815
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.44276872277259827,
      "learning_rate": 2.0855613064254198e-06,
      "loss": 0.0022,
      "step": 22820
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.169739231467247,
      "learning_rate": 2.0822349114244434e-06,
      "loss": 0.0016,
      "step": 22825
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.06298299133777618,
      "learning_rate": 2.0789108629761033e-06,
      "loss": 0.0009,
      "step": 22830
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.10760955512523651,
      "learning_rate": 2.075589162065538e-06,
      "loss": 0.0021,
      "step": 22835
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.17187266051769257,
      "learning_rate": 2.072269809677183e-06,
      "loss": 0.0028,
      "step": 22840
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.07554920017719269,
      "learning_rate": 2.0689528067947805e-06,
      "loss": 0.0015,
      "step": 22845
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.20058053731918335,
      "learning_rate": 2.0656381544013736e-06,
      "loss": 0.002,
      "step": 22850
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.1426592618227005,
      "learning_rate": 2.0623258534793167e-06,
      "loss": 0.0033,
      "step": 22855
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.15479739010334015,
      "learning_rate": 2.0590159050102543e-06,
      "loss": 0.0018,
      "step": 22860
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.13816289603710175,
      "learning_rate": 2.055708309975144e-06,
      "loss": 0.002,
      "step": 22865
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.29392099380493164,
      "learning_rate": 2.05240306935424e-06,
      "loss": 0.0024,
      "step": 22870
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.15190020203590393,
      "learning_rate": 2.0491001841271073e-06,
      "loss": 0.0017,
      "step": 22875
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.07080844044685364,
      "learning_rate": 2.0457996552726058e-06,
      "loss": 0.0011,
      "step": 22880
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.04252322018146515,
      "learning_rate": 2.042501483768897e-06,
      "loss": 0.002,
      "step": 22885
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.06752241402864456,
      "learning_rate": 2.039205670593448e-06,
      "loss": 0.0019,
      "step": 22890
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.11792897433042526,
      "learning_rate": 2.0359122167230194e-06,
      "loss": 0.0017,
      "step": 22895
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.24622458219528198,
      "learning_rate": 2.03262112313368e-06,
      "loss": 0.0014,
      "step": 22900
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.08839282393455505,
      "learning_rate": 2.0293323908007957e-06,
      "loss": 0.0018,
      "step": 22905
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.16991908848285675,
      "learning_rate": 2.026046020699035e-06,
      "loss": 0.0024,
      "step": 22910
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.055997684597969055,
      "learning_rate": 2.0227620138023605e-06,
      "loss": 0.0021,
      "step": 22915
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.25313636660575867,
      "learning_rate": 2.019480371084045e-06,
      "loss": 0.0016,
      "step": 22920
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.35071924328804016,
      "learning_rate": 2.016201093516653e-06,
      "loss": 0.002,
      "step": 22925
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.10866184532642365,
      "learning_rate": 2.0129241820720457e-06,
      "loss": 0.0024,
      "step": 22930
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.04748568683862686,
      "learning_rate": 2.009649637721388e-06,
      "loss": 0.0014,
      "step": 22935
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.13335078954696655,
      "learning_rate": 2.0063774614351427e-06,
      "loss": 0.0025,
      "step": 22940
    },
    {
      "epoch": 7.16,
      "grad_norm": 0.08357970416545868,
      "learning_rate": 2.0031076541830696e-06,
      "loss": 0.0015,
      "step": 22945
    },
    {
      "epoch": 7.16,
      "grad_norm": 0.12237964570522308,
      "learning_rate": 1.9998402169342266e-06,
      "loss": 0.0019,
      "step": 22950
    },
    {
      "epoch": 7.16,
      "grad_norm": 0.15404756367206573,
      "learning_rate": 1.996575150656971e-06,
      "loss": 0.0021,
      "step": 22955
    },
    {
      "epoch": 7.16,
      "grad_norm": 0.09013190865516663,
      "learning_rate": 1.9933124563189523e-06,
      "loss": 0.0016,
      "step": 22960
    },
    {
      "epoch": 7.16,
      "grad_norm": 0.09105832129716873,
      "learning_rate": 1.9900521348871273e-06,
      "loss": 0.0019,
      "step": 22965
    },
    {
      "epoch": 7.16,
      "grad_norm": 0.05199562385678291,
      "learning_rate": 1.9867941873277365e-06,
      "loss": 0.0017,
      "step": 22970
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.0494653545320034,
      "learning_rate": 1.9835386146063263e-06,
      "loss": 0.0012,
      "step": 22975
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.1796211451292038,
      "learning_rate": 1.980285417687735e-06,
      "loss": 0.0014,
      "step": 22980
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.03675558790564537,
      "learning_rate": 1.977034597536098e-06,
      "loss": 0.002,
      "step": 22985
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.04541827738285065,
      "learning_rate": 1.9737861551148474e-06,
      "loss": 0.0011,
      "step": 22990
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.1748480200767517,
      "learning_rate": 1.970540091386709e-06,
      "loss": 0.0013,
      "step": 22995
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.2021503746509552,
      "learning_rate": 1.9672964073137037e-06,
      "loss": 0.0025,
      "step": 23000
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.09402123093605042,
      "learning_rate": 1.9640551038571476e-06,
      "loss": 0.0016,
      "step": 23005
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.0501425638794899,
      "learning_rate": 1.9608161819776527e-06,
      "loss": 0.0021,
      "step": 23010
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.04663001373410225,
      "learning_rate": 1.9575796426351233e-06,
      "loss": 0.0011,
      "step": 23015
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.10955876111984253,
      "learning_rate": 1.9543454867887568e-06,
      "loss": 0.0016,
      "step": 23020
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.24366258084774017,
      "learning_rate": 1.9511137153970473e-06,
      "loss": 0.0024,
      "step": 23025
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.04715977981686592,
      "learning_rate": 1.9478843294177785e-06,
      "loss": 0.0012,
      "step": 23030
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.2500501871109009,
      "learning_rate": 1.9446573298080308e-06,
      "loss": 0.0022,
      "step": 23035
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.07603548467159271,
      "learning_rate": 1.9414327175241755e-06,
      "loss": 0.0011,
      "step": 23040
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.08409067988395691,
      "learning_rate": 1.938210493521875e-06,
      "loss": 0.0017,
      "step": 23045
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.3819182217121124,
      "learning_rate": 1.934990658756086e-06,
      "loss": 0.0022,
      "step": 23050
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.10924136638641357,
      "learning_rate": 1.9317732141810587e-06,
      "loss": 0.0014,
      "step": 23055
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.11690499633550644,
      "learning_rate": 1.92855816075033e-06,
      "loss": 0.0023,
      "step": 23060
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.1321161836385727,
      "learning_rate": 1.925345499416732e-06,
      "loss": 0.0015,
      "step": 23065
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.12577715516090393,
      "learning_rate": 1.9221352311323893e-06,
      "loss": 0.0017,
      "step": 23070
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.1400633454322815,
      "learning_rate": 1.9189273568487068e-06,
      "loss": 0.0023,
      "step": 23075
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.16724464297294617,
      "learning_rate": 1.915721877516397e-06,
      "loss": 0.0018,
      "step": 23080
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.29586324095726013,
      "learning_rate": 1.9125187940854484e-06,
      "loss": 0.0022,
      "step": 23085
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.1833949089050293,
      "learning_rate": 1.909318107505146e-06,
      "loss": 0.0013,
      "step": 23090
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.10151824355125427,
      "learning_rate": 1.9061198187240626e-06,
      "loss": 0.0015,
      "step": 23095
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.2685157358646393,
      "learning_rate": 1.9029239286900614e-06,
      "loss": 0.0019,
      "step": 23100
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.13566923141479492,
      "learning_rate": 1.8997304383502956e-06,
      "loss": 0.0016,
      "step": 23105
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.03023446351289749,
      "learning_rate": 1.8965393486512019e-06,
      "loss": 0.0011,
      "step": 23110
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.10596200823783875,
      "learning_rate": 1.893350660538511e-06,
      "loss": 0.0016,
      "step": 23115
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.15463019907474518,
      "learning_rate": 1.8901643749572373e-06,
      "loss": 0.0022,
      "step": 23120
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.9097394347190857,
      "learning_rate": 1.886980492851691e-06,
      "loss": 0.0032,
      "step": 23125
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.10497675836086273,
      "learning_rate": 1.8837990151654639e-06,
      "loss": 0.0016,
      "step": 23130
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.04598972573876381,
      "learning_rate": 1.880619942841435e-06,
      "loss": 0.0023,
      "step": 23135
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.15432889759540558,
      "learning_rate": 1.877443276821771e-06,
      "loss": 0.0019,
      "step": 23140
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.08259057253599167,
      "learning_rate": 1.8742690180479307e-06,
      "loss": 0.0017,
      "step": 23145
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.05540916323661804,
      "learning_rate": 1.8710971674606493e-06,
      "loss": 0.0011,
      "step": 23150
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.15354594588279724,
      "learning_rate": 1.867927725999956e-06,
      "loss": 0.0012,
      "step": 23155
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.13194885849952698,
      "learning_rate": 1.8647606946051634e-06,
      "loss": 0.0025,
      "step": 23160
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.1791795939207077,
      "learning_rate": 1.8615960742148699e-06,
      "loss": 0.0015,
      "step": 23165
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.09989296644926071,
      "learning_rate": 1.8584338657669631e-06,
      "loss": 0.0019,
      "step": 23170
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.32999667525291443,
      "learning_rate": 1.8552740701986106e-06,
      "loss": 0.0017,
      "step": 23175
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.11784721165895462,
      "learning_rate": 1.8521166884462693e-06,
      "loss": 0.0011,
      "step": 23180
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.1490192860364914,
      "learning_rate": 1.8489617214456746e-06,
      "loss": 0.002,
      "step": 23185
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.05966825410723686,
      "learning_rate": 1.8458091701318504e-06,
      "loss": 0.0016,
      "step": 23190
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.08577445149421692,
      "learning_rate": 1.842659035439106e-06,
      "loss": 0.0013,
      "step": 23195
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.18630562722682953,
      "learning_rate": 1.8395113183010316e-06,
      "loss": 0.0026,
      "step": 23200
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.08478522300720215,
      "learning_rate": 1.8363660196504995e-06,
      "loss": 0.0014,
      "step": 23205
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.05588646978139877,
      "learning_rate": 1.8332231404196732e-06,
      "loss": 0.0013,
      "step": 23210
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.15565073490142822,
      "learning_rate": 1.8300826815399941e-06,
      "loss": 0.0013,
      "step": 23215
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.1549144834280014,
      "learning_rate": 1.826944643942181e-06,
      "loss": 0.0021,
      "step": 23220
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.25133708119392395,
      "learning_rate": 1.8238090285562416e-06,
      "loss": 0.002,
      "step": 23225
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.12042286247015,
      "learning_rate": 1.8206758363114652e-06,
      "loss": 0.0013,
      "step": 23230
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.06990408152341843,
      "learning_rate": 1.8175450681364216e-06,
      "loss": 0.0019,
      "step": 23235
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.1288883239030838,
      "learning_rate": 1.8144167249589629e-06,
      "loss": 0.0017,
      "step": 23240
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.13696298003196716,
      "learning_rate": 1.8112908077062208e-06,
      "loss": 0.0015,
      "step": 23245
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.1362457573413849,
      "learning_rate": 1.808167317304611e-06,
      "loss": 0.0019,
      "step": 23250
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.12248032540082932,
      "learning_rate": 1.8050462546798275e-06,
      "loss": 0.0015,
      "step": 23255
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.21087703108787537,
      "learning_rate": 1.8019276207568471e-06,
      "loss": 0.0018,
      "step": 23260
    },
    {
      "epoch": 7.26,
      "grad_norm": 0.09722114354372025,
      "learning_rate": 1.7988114164599236e-06,
      "loss": 0.002,
      "step": 23265
    },
    {
      "epoch": 7.26,
      "grad_norm": 0.07935526967048645,
      "learning_rate": 1.7956976427125927e-06,
      "loss": 0.0016,
      "step": 23270
    },
    {
      "epoch": 7.26,
      "grad_norm": 0.3887324929237366,
      "learning_rate": 1.7925863004376709e-06,
      "loss": 0.0021,
      "step": 23275
    },
    {
      "epoch": 7.26,
      "grad_norm": 0.18374229967594147,
      "learning_rate": 1.7894773905572517e-06,
      "loss": 0.002,
      "step": 23280
    },
    {
      "epoch": 7.26,
      "grad_norm": 0.05458586663007736,
      "learning_rate": 1.78637091399271e-06,
      "loss": 0.0016,
      "step": 23285
    },
    {
      "epoch": 7.26,
      "grad_norm": 0.26444464921951294,
      "learning_rate": 1.7832668716646961e-06,
      "loss": 0.0011,
      "step": 23290
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.15571677684783936,
      "learning_rate": 1.7801652644931433e-06,
      "loss": 0.0021,
      "step": 23295
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.1157768964767456,
      "learning_rate": 1.7770660933972606e-06,
      "loss": 0.0013,
      "step": 23300
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.25759342312812805,
      "learning_rate": 1.7739693592955353e-06,
      "loss": 0.0034,
      "step": 23305
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.12619584798812866,
      "learning_rate": 1.7708750631057325e-06,
      "loss": 0.0014,
      "step": 23310
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.23353789746761322,
      "learning_rate": 1.7677832057448952e-06,
      "loss": 0.0026,
      "step": 23315
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.11836845427751541,
      "learning_rate": 1.7646937881293425e-06,
      "loss": 0.002,
      "step": 23320
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.13385270535945892,
      "learning_rate": 1.761606811174672e-06,
      "loss": 0.0028,
      "step": 23325
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.1433686912059784,
      "learning_rate": 1.7585222757957576e-06,
      "loss": 0.002,
      "step": 23330
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.06003287434577942,
      "learning_rate": 1.755440182906748e-06,
      "loss": 0.0012,
      "step": 23335
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.04114048182964325,
      "learning_rate": 1.7523605334210713e-06,
      "loss": 0.0014,
      "step": 23340
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.2964211702346802,
      "learning_rate": 1.7492833282514278e-06,
      "loss": 0.002,
      "step": 23345
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.15680080652236938,
      "learning_rate": 1.7462085683097952e-06,
      "loss": 0.0024,
      "step": 23350
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.07513896375894547,
      "learning_rate": 1.743136254507427e-06,
      "loss": 0.0017,
      "step": 23355
    },
    {
      "epoch": 7.29,
      "grad_norm": 0.38666918873786926,
      "learning_rate": 1.740066387754853e-06,
      "loss": 0.0029,
      "step": 23360
    },
    {
      "epoch": 7.29,
      "grad_norm": 0.20412775874137878,
      "learning_rate": 1.7369989689618683e-06,
      "loss": 0.0017,
      "step": 23365
    },
    {
      "epoch": 7.29,
      "grad_norm": 0.18670332431793213,
      "learning_rate": 1.7339339990375592e-06,
      "loss": 0.0021,
      "step": 23370
    },
    {
      "epoch": 7.29,
      "grad_norm": 0.09254976361989975,
      "learning_rate": 1.7308714788902726e-06,
      "loss": 0.0025,
      "step": 23375
    },
    {
      "epoch": 7.29,
      "grad_norm": 0.22622790932655334,
      "learning_rate": 1.7278114094276343e-06,
      "loss": 0.0014,
      "step": 23380
    },
    {
      "epoch": 7.29,
      "grad_norm": 0.4574730098247528,
      "learning_rate": 1.7247537915565449e-06,
      "loss": 0.0026,
      "step": 23385
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.08487337082624435,
      "learning_rate": 1.7216986261831747e-06,
      "loss": 0.0016,
      "step": 23390
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.2237267643213272,
      "learning_rate": 1.718645914212973e-06,
      "loss": 0.0032,
      "step": 23395
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.08069462329149246,
      "learning_rate": 1.715595656550655e-06,
      "loss": 0.0018,
      "step": 23400
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.10751515626907349,
      "learning_rate": 1.7125478541002117e-06,
      "loss": 0.0034,
      "step": 23405
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.20734716951847076,
      "learning_rate": 1.7095025077649053e-06,
      "loss": 0.002,
      "step": 23410
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.13567696511745453,
      "learning_rate": 1.7064596184472759e-06,
      "loss": 0.0012,
      "step": 23415
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.17350482940673828,
      "learning_rate": 1.7034191870491301e-06,
      "loss": 0.0022,
      "step": 23420
    },
    {
      "epoch": 7.31,
      "grad_norm": 0.13199970126152039,
      "learning_rate": 1.7003812144715459e-06,
      "loss": 0.0016,
      "step": 23425
    },
    {
      "epoch": 7.31,
      "grad_norm": 0.10658343881368637,
      "learning_rate": 1.6973457016148764e-06,
      "loss": 0.0024,
      "step": 23430
    },
    {
      "epoch": 7.31,
      "grad_norm": 0.18849949538707733,
      "learning_rate": 1.6943126493787377e-06,
      "loss": 0.002,
      "step": 23435
    },
    {
      "epoch": 7.31,
      "grad_norm": 0.031828440725803375,
      "learning_rate": 1.6912820586620237e-06,
      "loss": 0.0012,
      "step": 23440
    },
    {
      "epoch": 7.31,
      "grad_norm": 0.06511376798152924,
      "learning_rate": 1.6882539303628987e-06,
      "loss": 0.0021,
      "step": 23445
    },
    {
      "epoch": 7.31,
      "grad_norm": 0.14078600704669952,
      "learning_rate": 1.6852282653787943e-06,
      "loss": 0.0021,
      "step": 23450
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.09009654074907303,
      "learning_rate": 1.6822050646064113e-06,
      "loss": 0.0011,
      "step": 23455
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.11348678171634674,
      "learning_rate": 1.6791843289417275e-06,
      "loss": 0.0021,
      "step": 23460
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.13618040084838867,
      "learning_rate": 1.6761660592799822e-06,
      "loss": 0.0029,
      "step": 23465
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.16246213018894196,
      "learning_rate": 1.6731502565156875e-06,
      "loss": 0.0018,
      "step": 23470
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.04980913922190666,
      "learning_rate": 1.670136921542621e-06,
      "loss": 0.0031,
      "step": 23475
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.22228434681892395,
      "learning_rate": 1.6671260552538315e-06,
      "loss": 0.0017,
      "step": 23480
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.07236877083778381,
      "learning_rate": 1.6641176585416375e-06,
      "loss": 0.001,
      "step": 23485
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.24225230515003204,
      "learning_rate": 1.6611117322976222e-06,
      "loss": 0.0024,
      "step": 23490
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.2151099145412445,
      "learning_rate": 1.658108277412639e-06,
      "loss": 0.002,
      "step": 23495
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.08060291409492493,
      "learning_rate": 1.655107294776811e-06,
      "loss": 0.0022,
      "step": 23500
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.08732081949710846,
      "learning_rate": 1.652108785279526e-06,
      "loss": 0.0016,
      "step": 23505
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.3340969979763031,
      "learning_rate": 1.649112749809435e-06,
      "loss": 0.0027,
      "step": 23510
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.1414659023284912,
      "learning_rate": 1.6461191892544615e-06,
      "loss": 0.0023,
      "step": 23515
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.062250565737485886,
      "learning_rate": 1.6431281045017945e-06,
      "loss": 0.0014,
      "step": 23520
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.18974672257900238,
      "learning_rate": 1.640139496437887e-06,
      "loss": 0.0015,
      "step": 23525
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.09417407959699631,
      "learning_rate": 1.6371533659484606e-06,
      "loss": 0.0013,
      "step": 23530
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.13603679835796356,
      "learning_rate": 1.6341697139185008e-06,
      "loss": 0.0017,
      "step": 23535
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.030808500945568085,
      "learning_rate": 1.6311885412322602e-06,
      "loss": 0.002,
      "step": 23540
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.024461254477500916,
      "learning_rate": 1.6282098487732545e-06,
      "loss": 0.0011,
      "step": 23545
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.17799636721611023,
      "learning_rate": 1.6252336374242649e-06,
      "loss": 0.0016,
      "step": 23550
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.13428995013237,
      "learning_rate": 1.6222599080673395e-06,
      "loss": 0.002,
      "step": 23555
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.07780268788337708,
      "learning_rate": 1.6192886615837888e-06,
      "loss": 0.0012,
      "step": 23560
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.38052231073379517,
      "learning_rate": 1.6163198988541862e-06,
      "loss": 0.002,
      "step": 23565
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.10485098510980606,
      "learning_rate": 1.6133536207583734e-06,
      "loss": 0.0011,
      "step": 23570
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.14786922931671143,
      "learning_rate": 1.6103898281754504e-06,
      "loss": 0.0016,
      "step": 23575
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.11026573926210403,
      "learning_rate": 1.6074285219837848e-06,
      "loss": 0.0011,
      "step": 23580
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.22885969281196594,
      "learning_rate": 1.6044697030610045e-06,
      "loss": 0.0021,
      "step": 23585
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.6174030900001526,
      "learning_rate": 1.6015133722840026e-06,
      "loss": 0.0017,
      "step": 23590
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.07516336441040039,
      "learning_rate": 1.5985595305289337e-06,
      "loss": 0.0018,
      "step": 23595
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.4718790054321289,
      "learning_rate": 1.5956081786712129e-06,
      "loss": 0.0026,
      "step": 23600
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.08145270496606827,
      "learning_rate": 1.592659317585521e-06,
      "loss": 0.0018,
      "step": 23605
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.14953170716762543,
      "learning_rate": 1.5897129481457995e-06,
      "loss": 0.0012,
      "step": 23610
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.07766649127006531,
      "learning_rate": 1.5867690712252459e-06,
      "loss": 0.0014,
      "step": 23615
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.13531377911567688,
      "learning_rate": 1.5838276876963288e-06,
      "loss": 0.003,
      "step": 23620
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.2835178077220917,
      "learning_rate": 1.580888798430772e-06,
      "loss": 0.0012,
      "step": 23625
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.07614782452583313,
      "learning_rate": 1.5779524042995597e-06,
      "loss": 0.002,
      "step": 23630
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.10155941545963287,
      "learning_rate": 1.5750185061729395e-06,
      "loss": 0.0016,
      "step": 23635
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.09648201614618301,
      "learning_rate": 1.5720871049204156e-06,
      "loss": 0.0019,
      "step": 23640
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.1506904512643814,
      "learning_rate": 1.5691582014107553e-06,
      "loss": 0.002,
      "step": 23645
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.053801387548446655,
      "learning_rate": 1.5662317965119877e-06,
      "loss": 0.0016,
      "step": 23650
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.37937116622924805,
      "learning_rate": 1.5633078910913934e-06,
      "loss": 0.0023,
      "step": 23655
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.3482864201068878,
      "learning_rate": 1.560386486015516e-06,
      "loss": 0.0019,
      "step": 23660
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.16894619166851044,
      "learning_rate": 1.557467582150165e-06,
      "loss": 0.0022,
      "step": 23665
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.11991281807422638,
      "learning_rate": 1.5545511803603997e-06,
      "loss": 0.0016,
      "step": 23670
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.18368425965309143,
      "learning_rate": 1.5516372815105429e-06,
      "loss": 0.0028,
      "step": 23675
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.12280163913965225,
      "learning_rate": 1.548725886464172e-06,
      "loss": 0.0022,
      "step": 23680
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.022240638732910156,
      "learning_rate": 1.5458169960841264e-06,
      "loss": 0.0018,
      "step": 23685
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.4456542134284973,
      "learning_rate": 1.5429106112324976e-06,
      "loss": 0.0019,
      "step": 23690
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.06904762983322144,
      "learning_rate": 1.5400067327706392e-06,
      "loss": 0.0014,
      "step": 23695
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.1527893841266632,
      "learning_rate": 1.537105361559158e-06,
      "loss": 0.0027,
      "step": 23700
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.07344002276659012,
      "learning_rate": 1.5342064984579263e-06,
      "loss": 0.0014,
      "step": 23705
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.10091239213943481,
      "learning_rate": 1.5313101443260637e-06,
      "loss": 0.0019,
      "step": 23710
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.0808711051940918,
      "learning_rate": 1.528416300021951e-06,
      "loss": 0.0017,
      "step": 23715
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.1359367072582245,
      "learning_rate": 1.5255249664032256e-06,
      "loss": 0.0026,
      "step": 23720
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.2536752223968506,
      "learning_rate": 1.5226361443267735e-06,
      "loss": 0.0016,
      "step": 23725
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.1135883554816246,
      "learning_rate": 1.519749834648746e-06,
      "loss": 0.0016,
      "step": 23730
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.101815365254879,
      "learning_rate": 1.516866038224545e-06,
      "loss": 0.0017,
      "step": 23735
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.06885939836502075,
      "learning_rate": 1.5139847559088283e-06,
      "loss": 0.0012,
      "step": 23740
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.3209657073020935,
      "learning_rate": 1.5111059885555057e-06,
      "loss": 0.0023,
      "step": 23745
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.22403649985790253,
      "learning_rate": 1.5082297370177502e-06,
      "loss": 0.0027,
      "step": 23750
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.20601578056812286,
      "learning_rate": 1.5053560021479829e-06,
      "loss": 0.0029,
      "step": 23755
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.07541821897029877,
      "learning_rate": 1.5024847847978763e-06,
      "loss": 0.0014,
      "step": 23760
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.08014453947544098,
      "learning_rate": 1.4996160858183617e-06,
      "loss": 0.0018,
      "step": 23765
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.10281440615653992,
      "learning_rate": 1.4967499060596224e-06,
      "loss": 0.0021,
      "step": 23770
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.1816490888595581,
      "learning_rate": 1.4938862463710956e-06,
      "loss": 0.0023,
      "step": 23775
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.05056086555123329,
      "learning_rate": 1.4910251076014714e-06,
      "loss": 0.0023,
      "step": 23780
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.05771959200501442,
      "learning_rate": 1.4881664905986903e-06,
      "loss": 0.0014,
      "step": 23785
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.06854867935180664,
      "learning_rate": 1.4853103962099558e-06,
      "loss": 0.0016,
      "step": 23790
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.1187528446316719,
      "learning_rate": 1.4824568252817062e-06,
      "loss": 0.0017,
      "step": 23795
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.01285573374480009,
      "learning_rate": 1.4796057786596464e-06,
      "loss": 0.0018,
      "step": 23800
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.16395248472690582,
      "learning_rate": 1.4767572571887268e-06,
      "loss": 0.0032,
      "step": 23805
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.08021502941846848,
      "learning_rate": 1.4739112617131523e-06,
      "loss": 0.0016,
      "step": 23810
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.07731685787439346,
      "learning_rate": 1.4710677930763773e-06,
      "loss": 0.0014,
      "step": 23815
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.17272412776947021,
      "learning_rate": 1.4682268521211075e-06,
      "loss": 0.0014,
      "step": 23820
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.1399872601032257,
      "learning_rate": 1.4653884396892992e-06,
      "loss": 0.002,
      "step": 23825
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.11964300274848938,
      "learning_rate": 1.462552556622162e-06,
      "loss": 0.001,
      "step": 23830
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.08896835148334503,
      "learning_rate": 1.4597192037601526e-06,
      "loss": 0.0019,
      "step": 23835
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.11974415928125381,
      "learning_rate": 1.4568883819429803e-06,
      "loss": 0.0016,
      "step": 23840
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.25031784176826477,
      "learning_rate": 1.4540600920096003e-06,
      "loss": 0.0027,
      "step": 23845
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.06933608651161194,
      "learning_rate": 1.451234334798224e-06,
      "loss": 0.0027,
      "step": 23850
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.16940376162528992,
      "learning_rate": 1.448411111146305e-06,
      "loss": 0.0016,
      "step": 23855
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.07995875924825668,
      "learning_rate": 1.4455904218905514e-06,
      "loss": 0.0023,
      "step": 23860
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.28575241565704346,
      "learning_rate": 1.4427722678669188e-06,
      "loss": 0.0022,
      "step": 23865
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.0801328793168068,
      "learning_rate": 1.4399566499106089e-06,
      "loss": 0.002,
      "step": 23870
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.059965942054986954,
      "learning_rate": 1.4371435688560753e-06,
      "loss": 0.0012,
      "step": 23875
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.27908244729042053,
      "learning_rate": 1.434333025537018e-06,
      "loss": 0.0026,
      "step": 23880
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.11046892404556274,
      "learning_rate": 1.431525020786385e-06,
      "loss": 0.0025,
      "step": 23885
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.2373012900352478,
      "learning_rate": 1.428719555436372e-06,
      "loss": 0.0016,
      "step": 23890
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.09902806580066681,
      "learning_rate": 1.4259166303184212e-06,
      "loss": 0.002,
      "step": 23895
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.07448076456785202,
      "learning_rate": 1.4231162462632276e-06,
      "loss": 0.0016,
      "step": 23900
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.1125267893075943,
      "learning_rate": 1.42031840410072e-06,
      "loss": 0.0023,
      "step": 23905
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.1669638454914093,
      "learning_rate": 1.4175231046600902e-06,
      "loss": 0.0022,
      "step": 23910
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.1995648890733719,
      "learning_rate": 1.4147303487697661e-06,
      "loss": 0.002,
      "step": 23915
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.15967807173728943,
      "learning_rate": 1.4119401372574226e-06,
      "loss": 0.0021,
      "step": 23920
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.16623800992965698,
      "learning_rate": 1.4091524709499849e-06,
      "loss": 0.0023,
      "step": 23925
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.047116223722696304,
      "learning_rate": 1.4063673506736198e-06,
      "loss": 0.0024,
      "step": 23930
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.06346458941698074,
      "learning_rate": 1.4035847772537426e-06,
      "loss": 0.002,
      "step": 23935
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.2038680911064148,
      "learning_rate": 1.400804751515008e-06,
      "loss": 0.0015,
      "step": 23940
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.046279482543468475,
      "learning_rate": 1.3980272742813228e-06,
      "loss": 0.0021,
      "step": 23945
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.12787197530269623,
      "learning_rate": 1.3952523463758327e-06,
      "loss": 0.0022,
      "step": 23950
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.11368878930807114,
      "learning_rate": 1.3924799686209357e-06,
      "loss": 0.0032,
      "step": 23955
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.04303009808063507,
      "learning_rate": 1.3897101418382664e-06,
      "loss": 0.0016,
      "step": 23960
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.23386818170547485,
      "learning_rate": 1.3869428668487061e-06,
      "loss": 0.0029,
      "step": 23965
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.06971883773803711,
      "learning_rate": 1.384178144472379e-06,
      "loss": 0.0017,
      "step": 23970
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.08481031656265259,
      "learning_rate": 1.381415975528656e-06,
      "loss": 0.0016,
      "step": 23975
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.14777174592018127,
      "learning_rate": 1.3786563608361447e-06,
      "loss": 0.0017,
      "step": 23980
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.16118240356445312,
      "learning_rate": 1.375899301212703e-06,
      "loss": 0.0023,
      "step": 23985
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.19115978479385376,
      "learning_rate": 1.3731447974754241e-06,
      "loss": 0.0015,
      "step": 23990
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.09597334265708923,
      "learning_rate": 1.3703928504406527e-06,
      "loss": 0.0012,
      "step": 23995
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.2174786627292633,
      "learning_rate": 1.3676434609239708e-06,
      "loss": 0.0015,
      "step": 24000
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.2653753161430359,
      "learning_rate": 1.3648966297402e-06,
      "loss": 0.0023,
      "step": 24005
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.15789607167243958,
      "learning_rate": 1.3621523577034102e-06,
      "loss": 0.0023,
      "step": 24010
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.34349653124809265,
      "learning_rate": 1.3594106456269052e-06,
      "loss": 0.0021,
      "step": 24015
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.07134407013654709,
      "learning_rate": 1.356671494323234e-06,
      "loss": 0.0013,
      "step": 24020
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.11368986964225769,
      "learning_rate": 1.3539349046041882e-06,
      "loss": 0.0014,
      "step": 24025
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.12296315282583237,
      "learning_rate": 1.3512008772807994e-06,
      "loss": 0.0017,
      "step": 24030
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.17269238829612732,
      "learning_rate": 1.3484694131633346e-06,
      "loss": 0.0016,
      "step": 24035
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.07103969156742096,
      "learning_rate": 1.345740513061311e-06,
      "loss": 0.0013,
      "step": 24040
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.2738235294818878,
      "learning_rate": 1.3430141777834815e-06,
      "loss": 0.0021,
      "step": 24045
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.10616691410541534,
      "learning_rate": 1.3402904081378333e-06,
      "loss": 0.0014,
      "step": 24050
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.07980950176715851,
      "learning_rate": 1.337569204931598e-06,
      "loss": 0.0017,
      "step": 24055
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.09553997963666916,
      "learning_rate": 1.3348505689712488e-06,
      "loss": 0.0015,
      "step": 24060
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.13931411504745483,
      "learning_rate": 1.3321345010624941e-06,
      "loss": 0.0013,
      "step": 24065
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.23361235857009888,
      "learning_rate": 1.3294210020102837e-06,
      "loss": 0.0019,
      "step": 24070
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.16457533836364746,
      "learning_rate": 1.3267100726188053e-06,
      "loss": 0.003,
      "step": 24075
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.125288724899292,
      "learning_rate": 1.3240017136914852e-06,
      "loss": 0.0017,
      "step": 24080
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.2684983015060425,
      "learning_rate": 1.3212959260309855e-06,
      "loss": 0.0019,
      "step": 24085
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.17238128185272217,
      "learning_rate": 1.3185927104392105e-06,
      "loss": 0.0017,
      "step": 24090
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.21431346237659454,
      "learning_rate": 1.3158920677172994e-06,
      "loss": 0.0018,
      "step": 24095
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.1683589220046997,
      "learning_rate": 1.3131939986656305e-06,
      "loss": 0.0019,
      "step": 24100
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.07739071547985077,
      "learning_rate": 1.310498504083817e-06,
      "loss": 0.0015,
      "step": 24105
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.07120134681463242,
      "learning_rate": 1.3078055847707116e-06,
      "loss": 0.0014,
      "step": 24110
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.09688661247491837,
      "learning_rate": 1.3051152415244016e-06,
      "loss": 0.0014,
      "step": 24115
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.15954263508319855,
      "learning_rate": 1.302427475142214e-06,
      "loss": 0.0028,
      "step": 24120
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.07054508477449417,
      "learning_rate": 1.2997422864207076e-06,
      "loss": 0.0016,
      "step": 24125
    },
    {
      "epoch": 7.53,
      "grad_norm": 0.1399802416563034,
      "learning_rate": 1.297059676155682e-06,
      "loss": 0.002,
      "step": 24130
    },
    {
      "epoch": 7.53,
      "grad_norm": 0.2348816841840744,
      "learning_rate": 1.2943796451421686e-06,
      "loss": 0.0022,
      "step": 24135
    },
    {
      "epoch": 7.53,
      "grad_norm": 0.1977377086877823,
      "learning_rate": 1.291702194174438e-06,
      "loss": 0.0026,
      "step": 24140
    },
    {
      "epoch": 7.53,
      "grad_norm": 0.41480663418769836,
      "learning_rate": 1.2890273240459938e-06,
      "loss": 0.0018,
      "step": 24145
    },
    {
      "epoch": 7.53,
      "grad_norm": 0.13588984310626984,
      "learning_rate": 1.2863550355495735e-06,
      "loss": 0.0017,
      "step": 24150
    },
    {
      "epoch": 7.53,
      "grad_norm": 0.2099747657775879,
      "learning_rate": 1.2836853294771545e-06,
      "loss": 0.0017,
      "step": 24155
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.13870127499103546,
      "learning_rate": 1.2810182066199429e-06,
      "loss": 0.0018,
      "step": 24160
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.21662606298923492,
      "learning_rate": 1.2783536677683828e-06,
      "loss": 0.0019,
      "step": 24165
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.11504267156124115,
      "learning_rate": 1.2756917137121527e-06,
      "loss": 0.0019,
      "step": 24170
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.1169959232211113,
      "learning_rate": 1.2730323452401627e-06,
      "loss": 0.0019,
      "step": 24175
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.07010991871356964,
      "learning_rate": 1.2703755631405589e-06,
      "loss": 0.0016,
      "step": 24180
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.1383741796016693,
      "learning_rate": 1.2677213682007195e-06,
      "loss": 0.0015,
      "step": 24185
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.22670747339725494,
      "learning_rate": 1.2650697612072582e-06,
      "loss": 0.0017,
      "step": 24190
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.06694146990776062,
      "learning_rate": 1.2624207429460134e-06,
      "loss": 0.0023,
      "step": 24195
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.13071152567863464,
      "learning_rate": 1.2597743142020703e-06,
      "loss": 0.0018,
      "step": 24200
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.06935352832078934,
      "learning_rate": 1.2571304757597357e-06,
      "loss": 0.0009,
      "step": 24205
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.06360245496034622,
      "learning_rate": 1.2544892284025523e-06,
      "loss": 0.0031,
      "step": 24210
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.17494170367717743,
      "learning_rate": 1.2518505729132969e-06,
      "loss": 0.0015,
      "step": 24215
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.08135698735713959,
      "learning_rate": 1.2492145100739728e-06,
      "loss": 0.0017,
      "step": 24220
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.2048783153295517,
      "learning_rate": 1.246581040665823e-06,
      "loss": 0.002,
      "step": 24225
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.1028202474117279,
      "learning_rate": 1.243950165469311e-06,
      "loss": 0.0014,
      "step": 24230
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.13023538887500763,
      "learning_rate": 1.2413218852641406e-06,
      "loss": 0.0018,
      "step": 24235
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.10343587398529053,
      "learning_rate": 1.2386962008292413e-06,
      "loss": 0.0024,
      "step": 24240
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.12373864650726318,
      "learning_rate": 1.2360731129427794e-06,
      "loss": 0.0029,
      "step": 24245
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.12057089060544968,
      "learning_rate": 1.2334526223821453e-06,
      "loss": 0.0017,
      "step": 24250
    },
    {
      "epoch": 7.57,
      "grad_norm": 0.1910213679075241,
      "learning_rate": 1.230834729923963e-06,
      "loss": 0.0022,
      "step": 24255
    },
    {
      "epoch": 7.57,
      "grad_norm": 0.4182658791542053,
      "learning_rate": 1.228219436344087e-06,
      "loss": 0.002,
      "step": 24260
    },
    {
      "epoch": 7.57,
      "grad_norm": 0.22596170008182526,
      "learning_rate": 1.2256067424175956e-06,
      "loss": 0.0027,
      "step": 24265
    },
    {
      "epoch": 7.57,
      "grad_norm": 0.09142956137657166,
      "learning_rate": 1.2229966489188038e-06,
      "loss": 0.0017,
      "step": 24270
    },
    {
      "epoch": 7.57,
      "grad_norm": 0.1430046558380127,
      "learning_rate": 1.2203891566212545e-06,
      "loss": 0.0023,
      "step": 24275
    },
    {
      "epoch": 7.57,
      "grad_norm": 0.1019957885146141,
      "learning_rate": 1.2177842662977136e-06,
      "loss": 0.0017,
      "step": 24280
    },
    {
      "epoch": 7.57,
      "grad_norm": 0.2015380859375,
      "learning_rate": 1.215181978720188e-06,
      "loss": 0.0019,
      "step": 24285
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.15892979502677917,
      "learning_rate": 1.212582294659902e-06,
      "loss": 0.0019,
      "step": 24290
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.11718596518039703,
      "learning_rate": 1.2099852148873136e-06,
      "loss": 0.0018,
      "step": 24295
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.08311883360147476,
      "learning_rate": 1.2073907401721086e-06,
      "loss": 0.0015,
      "step": 24300
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.1394026130437851,
      "learning_rate": 1.2047988712831959e-06,
      "loss": 0.0013,
      "step": 24305
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.20909558236598969,
      "learning_rate": 1.2022096089887191e-06,
      "loss": 0.0019,
      "step": 24310
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.17596064507961273,
      "learning_rate": 1.1996229540560456e-06,
      "loss": 0.0018,
      "step": 24315
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.11002202332019806,
      "learning_rate": 1.1970389072517707e-06,
      "loss": 0.0025,
      "step": 24320
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.17328931391239166,
      "learning_rate": 1.1944574693417155e-06,
      "loss": 0.0017,
      "step": 24325
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.3190286159515381,
      "learning_rate": 1.1918786410909323e-06,
      "loss": 0.0015,
      "step": 24330
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.20087780058383942,
      "learning_rate": 1.1893024232636974e-06,
      "loss": 0.0025,
      "step": 24335
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.24992135167121887,
      "learning_rate": 1.1867288166235091e-06,
      "loss": 0.0016,
      "step": 24340
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.14544323086738586,
      "learning_rate": 1.1841578219330974e-06,
      "loss": 0.0027,
      "step": 24345
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.04758672043681145,
      "learning_rate": 1.181589439954416e-06,
      "loss": 0.0017,
      "step": 24350
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.1231321468949318,
      "learning_rate": 1.1790236714486459e-06,
      "loss": 0.0021,
      "step": 24355
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.14308707416057587,
      "learning_rate": 1.1764605171761923e-06,
      "loss": 0.0015,
      "step": 24360
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.24447542428970337,
      "learning_rate": 1.1738999778966842e-06,
      "loss": 0.0018,
      "step": 24365
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.14588113129138947,
      "learning_rate": 1.171342054368979e-06,
      "loss": 0.0016,
      "step": 24370
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.17098912596702576,
      "learning_rate": 1.168786747351156e-06,
      "loss": 0.0014,
      "step": 24375
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.10740490257740021,
      "learning_rate": 1.1662340576005215e-06,
      "loss": 0.0023,
      "step": 24380
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.1715841144323349,
      "learning_rate": 1.1636839858736027e-06,
      "loss": 0.0013,
      "step": 24385
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.2061709761619568,
      "learning_rate": 1.1611365329261549e-06,
      "loss": 0.0014,
      "step": 24390
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.11730609089136124,
      "learning_rate": 1.158591699513154e-06,
      "loss": 0.0016,
      "step": 24395
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.09845942258834839,
      "learning_rate": 1.1560494863888006e-06,
      "loss": 0.0014,
      "step": 24400
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.18577240407466888,
      "learning_rate": 1.1535098943065204e-06,
      "loss": 0.0018,
      "step": 24405
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.1618165522813797,
      "learning_rate": 1.1509729240189605e-06,
      "loss": 0.0018,
      "step": 24410
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.11084864288568497,
      "learning_rate": 1.1484385762779914e-06,
      "loss": 0.0017,
      "step": 24415
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.23109754920005798,
      "learning_rate": 1.1459068518347062e-06,
      "loss": 0.0021,
      "step": 24420
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.04823751002550125,
      "learning_rate": 1.1433777514394206e-06,
      "loss": 0.0013,
      "step": 24425
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.08719401061534882,
      "learning_rate": 1.1408512758416735e-06,
      "loss": 0.0015,
      "step": 24430
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.08335618674755096,
      "learning_rate": 1.1383274257902244e-06,
      "loss": 0.002,
      "step": 24435
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.29434606432914734,
      "learning_rate": 1.1358062020330574e-06,
      "loss": 0.0028,
      "step": 24440
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.19055098295211792,
      "learning_rate": 1.1332876053173714e-06,
      "loss": 0.0017,
      "step": 24445
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.05965162441134453,
      "learning_rate": 1.130771636389596e-06,
      "loss": 0.0019,
      "step": 24450
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.18123778700828552,
      "learning_rate": 1.1282582959953769e-06,
      "loss": 0.0015,
      "step": 24455
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.027868177741765976,
      "learning_rate": 1.1257475848795818e-06,
      "loss": 0.0018,
      "step": 24460
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.14499281346797943,
      "learning_rate": 1.1232395037862974e-06,
      "loss": 0.0023,
      "step": 24465
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.16413047909736633,
      "learning_rate": 1.120734053458834e-06,
      "loss": 0.0022,
      "step": 24470
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.13355393707752228,
      "learning_rate": 1.1182312346397205e-06,
      "loss": 0.0021,
      "step": 24475
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.16823016107082367,
      "learning_rate": 1.1157310480707073e-06,
      "loss": 0.001,
      "step": 24480
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.1044691801071167,
      "learning_rate": 1.1132334944927582e-06,
      "loss": 0.0014,
      "step": 24485
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.13820968568325043,
      "learning_rate": 1.1107385746460686e-06,
      "loss": 0.0021,
      "step": 24490
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.18508176505565643,
      "learning_rate": 1.108246289270043e-06,
      "loss": 0.0033,
      "step": 24495
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.05543508380651474,
      "learning_rate": 1.1057566391033092e-06,
      "loss": 0.0016,
      "step": 24500
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.07221606373786926,
      "learning_rate": 1.103269624883715e-06,
      "loss": 0.0018,
      "step": 24505
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.08910844475030899,
      "learning_rate": 1.100785247348325e-06,
      "loss": 0.0011,
      "step": 24510
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.16187050938606262,
      "learning_rate": 1.0983035072334248e-06,
      "loss": 0.0018,
      "step": 24515
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.1790592521429062,
      "learning_rate": 1.0958244052745127e-06,
      "loss": 0.0022,
      "step": 24520
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.0456264354288578,
      "learning_rate": 1.093347942206312e-06,
      "loss": 0.0013,
      "step": 24525
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.09067858755588531,
      "learning_rate": 1.090874118762757e-06,
      "loss": 0.0021,
      "step": 24530
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.08108704537153244,
      "learning_rate": 1.08840293567701e-06,
      "loss": 0.0023,
      "step": 24535
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.10392729192972183,
      "learning_rate": 1.0859343936814416e-06,
      "loss": 0.0012,
      "step": 24540
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.3361983299255371,
      "learning_rate": 1.083468493507641e-06,
      "loss": 0.0018,
      "step": 24545
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.07768458127975464,
      "learning_rate": 1.08100523588642e-06,
      "loss": 0.0018,
      "step": 24550
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.11710520833730698,
      "learning_rate": 1.078544621547799e-06,
      "loss": 0.001,
      "step": 24555
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.08064442873001099,
      "learning_rate": 1.07608665122102e-06,
      "loss": 0.0022,
      "step": 24560
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.11553987860679626,
      "learning_rate": 1.073631325634542e-06,
      "loss": 0.0014,
      "step": 24565
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.22669553756713867,
      "learning_rate": 1.0711786455160389e-06,
      "loss": 0.0014,
      "step": 24570
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.13220682740211487,
      "learning_rate": 1.0687286115923979e-06,
      "loss": 0.0016,
      "step": 24575
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.2977820336818695,
      "learning_rate": 1.066281224589728e-06,
      "loss": 0.0026,
      "step": 24580
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.04379139095544815,
      "learning_rate": 1.0638364852333516e-06,
      "loss": 0.0015,
      "step": 24585
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.18310536444187164,
      "learning_rate": 1.0613943942478e-06,
      "loss": 0.0022,
      "step": 24590
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.12095720320940018,
      "learning_rate": 1.0589549523568287e-06,
      "loss": 0.002,
      "step": 24595
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.12880194187164307,
      "learning_rate": 1.0565181602834017e-06,
      "loss": 0.0029,
      "step": 24600
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.1454593986272812,
      "learning_rate": 1.0540840187497026e-06,
      "loss": 0.0021,
      "step": 24605
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.15364618599414825,
      "learning_rate": 1.0516525284771252e-06,
      "loss": 0.0022,
      "step": 24610
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.23326414823532104,
      "learning_rate": 1.0492236901862774e-06,
      "loss": 0.0024,
      "step": 24615
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.17910537123680115,
      "learning_rate": 1.0467975045969902e-06,
      "loss": 0.0014,
      "step": 24620
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.12565740942955017,
      "learning_rate": 1.0443739724282954e-06,
      "loss": 0.0023,
      "step": 24625
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.0883830189704895,
      "learning_rate": 1.0419530943984458e-06,
      "loss": 0.0014,
      "step": 24630
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.0909748524427414,
      "learning_rate": 1.0395348712249064e-06,
      "loss": 0.0019,
      "step": 24635
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.10346776992082596,
      "learning_rate": 1.0371193036243554e-06,
      "loss": 0.0012,
      "step": 24640
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.18627353012561798,
      "learning_rate": 1.0347063923126832e-06,
      "loss": 0.0023,
      "step": 24645
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.11687027662992477,
      "learning_rate": 1.0322961380049935e-06,
      "loss": 0.0026,
      "step": 24650
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.17399488389492035,
      "learning_rate": 1.0298885414156034e-06,
      "loss": 0.0019,
      "step": 24655
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.1125384047627449,
      "learning_rate": 1.0274836032580416e-06,
      "loss": 0.0008,
      "step": 24660
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.11277636140584946,
      "learning_rate": 1.0250813242450474e-06,
      "loss": 0.0024,
      "step": 24665
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.17943529784679413,
      "learning_rate": 1.022681705088575e-06,
      "loss": 0.0017,
      "step": 24670
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.037092261016368866,
      "learning_rate": 1.020284746499788e-06,
      "loss": 0.0014,
      "step": 24675
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.1271539330482483,
      "learning_rate": 1.0178904491890628e-06,
      "loss": 0.0019,
      "step": 24680
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.08718859404325485,
      "learning_rate": 1.015498813865986e-06,
      "loss": 0.0019,
      "step": 24685
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.1267717331647873,
      "learning_rate": 1.013109841239356e-06,
      "loss": 0.0016,
      "step": 24690
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.2325306236743927,
      "learning_rate": 1.0107235320171826e-06,
      "loss": 0.0026,
      "step": 24695
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.06632918864488602,
      "learning_rate": 1.0083398869066852e-06,
      "loss": 0.0022,
      "step": 24700
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.21014709770679474,
      "learning_rate": 1.005958906614294e-06,
      "loss": 0.0018,
      "step": 24705
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.17265684902668,
      "learning_rate": 1.0035805918456487e-06,
      "loss": 0.004,
      "step": 24710
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.13958702981472015,
      "learning_rate": 1.0012049433056015e-06,
      "loss": 0.0022,
      "step": 24715
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.14307720959186554,
      "learning_rate": 9.988319616982112e-07,
      "loss": 0.0025,
      "step": 24720
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.18551354110240936,
      "learning_rate": 9.964616477267486e-07,
      "loss": 0.0031,
      "step": 24725
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.16873350739479065,
      "learning_rate": 9.940940020936952e-07,
      "loss": 0.0017,
      "step": 24730
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.10657307505607605,
      "learning_rate": 9.917290255007327e-07,
      "loss": 0.0019,
      "step": 24735
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.09567416459321976,
      "learning_rate": 9.893667186487654e-07,
      "loss": 0.002,
      "step": 24740
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.05228479579091072,
      "learning_rate": 9.870070822378974e-07,
      "loss": 0.0015,
      "step": 24745
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.13476140797138214,
      "learning_rate": 9.84650116967444e-07,
      "loss": 0.0027,
      "step": 24750
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.12251874804496765,
      "learning_rate": 9.82295823535927e-07,
      "loss": 0.0018,
      "step": 24755
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.08515207469463348,
      "learning_rate": 9.799442026410788e-07,
      "loss": 0.0011,
      "step": 24760
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.6399232745170593,
      "learning_rate": 9.775952549798406e-07,
      "loss": 0.0038,
      "step": 24765
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.19654053449630737,
      "learning_rate": 9.752489812483567e-07,
      "loss": 0.0031,
      "step": 24770
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.09422720223665237,
      "learning_rate": 9.729053821419797e-07,
      "loss": 0.002,
      "step": 24775
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.2208988070487976,
      "learning_rate": 9.705644583552775e-07,
      "loss": 0.001,
      "step": 24780
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.09688408672809601,
      "learning_rate": 9.682262105820161e-07,
      "loss": 0.0024,
      "step": 24785
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.07366550713777542,
      "learning_rate": 9.658906395151713e-07,
      "loss": 0.0023,
      "step": 24790
    },
    {
      "epoch": 7.73,
      "grad_norm": 0.17863616347312927,
      "learning_rate": 9.63557745846927e-07,
      "loss": 0.0021,
      "step": 24795
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.08455432206392288,
      "learning_rate": 9.612275302686714e-07,
      "loss": 0.0021,
      "step": 24800
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.23154126107692719,
      "learning_rate": 9.58899993471003e-07,
      "loss": 0.0019,
      "step": 24805
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.11111778020858765,
      "learning_rate": 9.56575136143718e-07,
      "loss": 0.0014,
      "step": 24810
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.1214340329170227,
      "learning_rate": 9.542529589758276e-07,
      "loss": 0.0023,
      "step": 24815
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.13460777699947357,
      "learning_rate": 9.519334626555421e-07,
      "loss": 0.0016,
      "step": 24820
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.16119147837162018,
      "learning_rate": 9.496166478702851e-07,
      "loss": 0.0021,
      "step": 24825
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.09674394875764847,
      "learning_rate": 9.473025153066773e-07,
      "loss": 0.0019,
      "step": 24830
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.15101002156734467,
      "learning_rate": 9.44991065650549e-07,
      "loss": 0.0025,
      "step": 24835
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.06771969050168991,
      "learning_rate": 9.426822995869367e-07,
      "loss": 0.0013,
      "step": 24840
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.12194545567035675,
      "learning_rate": 9.403762178000731e-07,
      "loss": 0.0021,
      "step": 24845
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.10539015382528305,
      "learning_rate": 9.380728209734047e-07,
      "loss": 0.0018,
      "step": 24850
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.10646653920412064,
      "learning_rate": 9.357721097895778e-07,
      "loss": 0.0018,
      "step": 24855
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.35590100288391113,
      "learning_rate": 9.334740849304457e-07,
      "loss": 0.0021,
      "step": 24860
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.09334748238325119,
      "learning_rate": 9.311787470770606e-07,
      "loss": 0.0021,
      "step": 24865
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.3298105001449585,
      "learning_rate": 9.288860969096858e-07,
      "loss": 0.0031,
      "step": 24870
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.0878896489739418,
      "learning_rate": 9.265961351077835e-07,
      "loss": 0.002,
      "step": 24875
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.05391193926334381,
      "learning_rate": 9.243088623500163e-07,
      "loss": 0.0014,
      "step": 24880
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.137955442070961,
      "learning_rate": 9.22024279314253e-07,
      "loss": 0.0019,
      "step": 24885
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.13348926603794098,
      "learning_rate": 9.197423866775679e-07,
      "loss": 0.0017,
      "step": 24890
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.604273796081543,
      "learning_rate": 9.17463185116233e-07,
      "loss": 0.0019,
      "step": 24895
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.21288979053497314,
      "learning_rate": 9.15186675305727e-07,
      "loss": 0.0015,
      "step": 24900
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.2926914095878601,
      "learning_rate": 9.129128579207258e-07,
      "loss": 0.0025,
      "step": 24905
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.0603574700653553,
      "learning_rate": 9.106417336351159e-07,
      "loss": 0.0013,
      "step": 24910
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.15772417187690735,
      "learning_rate": 9.083733031219765e-07,
      "loss": 0.0018,
      "step": 24915
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.04270298406481743,
      "learning_rate": 9.061075670535913e-07,
      "loss": 0.0011,
      "step": 24920
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.11631837487220764,
      "learning_rate": 9.038445261014484e-07,
      "loss": 0.0022,
      "step": 24925
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.18365252017974854,
      "learning_rate": 9.015841809362347e-07,
      "loss": 0.0021,
      "step": 24930
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.1408557891845703,
      "learning_rate": 8.993265322278389e-07,
      "loss": 0.0018,
      "step": 24935
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.10175014287233353,
      "learning_rate": 8.97071580645349e-07,
      "loss": 0.0021,
      "step": 24940
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.048780281096696854,
      "learning_rate": 8.94819326857056e-07,
      "loss": 0.001,
      "step": 24945
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.15094539523124695,
      "learning_rate": 8.925697715304504e-07,
      "loss": 0.0017,
      "step": 24950
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.1805548220872879,
      "learning_rate": 8.903229153322212e-07,
      "loss": 0.0018,
      "step": 24955
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.07314382493495941,
      "learning_rate": 8.880787589282613e-07,
      "loss": 0.0021,
      "step": 24960
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.12884323298931122,
      "learning_rate": 8.858373029836609e-07,
      "loss": 0.0007,
      "step": 24965
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.19089964032173157,
      "learning_rate": 8.835985481627097e-07,
      "loss": 0.0017,
      "step": 24970
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.13515068590641022,
      "learning_rate": 8.813624951288969e-07,
      "loss": 0.0013,
      "step": 24975
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.2111157923936844,
      "learning_rate": 8.791291445449124e-07,
      "loss": 0.0023,
      "step": 24980
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.1934300810098648,
      "learning_rate": 8.768984970726435e-07,
      "loss": 0.0022,
      "step": 24985
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.15313348174095154,
      "learning_rate": 8.746705533731792e-07,
      "loss": 0.0019,
      "step": 24990
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.07578973472118378,
      "learning_rate": 8.724453141068023e-07,
      "loss": 0.0011,
      "step": 24995
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.09974081814289093,
      "learning_rate": 8.702227799329999e-07,
      "loss": 0.002,
      "step": 25000
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.03508811444044113,
      "learning_rate": 8.680029515104516e-07,
      "loss": 0.0013,
      "step": 25005
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.14680829644203186,
      "learning_rate": 8.657858294970411e-07,
      "loss": 0.0026,
      "step": 25010
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.19211848080158234,
      "learning_rate": 8.635714145498442e-07,
      "loss": 0.0024,
      "step": 25015
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.12090691179037094,
      "learning_rate": 8.613597073251412e-07,
      "loss": 0.0017,
      "step": 25020
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.1656752973794937,
      "learning_rate": 8.59150708478399e-07,
      "loss": 0.002,
      "step": 25025
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.20714865624904633,
      "learning_rate": 8.56944418664295e-07,
      "loss": 0.0021,
      "step": 25030
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.25840499997138977,
      "learning_rate": 8.547408385366951e-07,
      "loss": 0.0013,
      "step": 25035
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.1592530459165573,
      "learning_rate": 8.525399687486635e-07,
      "loss": 0.0021,
      "step": 25040
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.182389497756958,
      "learning_rate": 8.503418099524641e-07,
      "loss": 0.0018,
      "step": 25045
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.040837742388248444,
      "learning_rate": 8.481463627995545e-07,
      "loss": 0.0021,
      "step": 25050
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.10436588525772095,
      "learning_rate": 8.459536279405922e-07,
      "loss": 0.0026,
      "step": 25055
    },
    {
      "epoch": 7.82,
      "grad_norm": 0.18838675320148468,
      "learning_rate": 8.437636060254228e-07,
      "loss": 0.0019,
      "step": 25060
    },
    {
      "epoch": 7.82,
      "grad_norm": 0.2783716917037964,
      "learning_rate": 8.415762977030961e-07,
      "loss": 0.0024,
      "step": 25065
    },
    {
      "epoch": 7.82,
      "grad_norm": 0.3167223036289215,
      "learning_rate": 8.393917036218535e-07,
      "loss": 0.0019,
      "step": 25070
    },
    {
      "epoch": 7.82,
      "grad_norm": 0.11779921501874924,
      "learning_rate": 8.372098244291371e-07,
      "loss": 0.0022,
      "step": 25075
    },
    {
      "epoch": 7.82,
      "grad_norm": 0.2449861466884613,
      "learning_rate": 8.350306607715774e-07,
      "loss": 0.0022,
      "step": 25080
    },
    {
      "epoch": 7.82,
      "grad_norm": 0.08655525743961334,
      "learning_rate": 8.328542132950046e-07,
      "loss": 0.0014,
      "step": 25085
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.06086605787277222,
      "learning_rate": 8.306804826444448e-07,
      "loss": 0.0011,
      "step": 25090
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.0736154168844223,
      "learning_rate": 8.285094694641127e-07,
      "loss": 0.002,
      "step": 25095
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.38189685344696045,
      "learning_rate": 8.263411743974226e-07,
      "loss": 0.003,
      "step": 25100
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.02803908847272396,
      "learning_rate": 8.241755980869837e-07,
      "loss": 0.0013,
      "step": 25105
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.10847136378288269,
      "learning_rate": 8.220127411745971e-07,
      "loss": 0.0016,
      "step": 25110
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.2512924373149872,
      "learning_rate": 8.198526043012611e-07,
      "loss": 0.0017,
      "step": 25115
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.06729435920715332,
      "learning_rate": 8.176951881071638e-07,
      "loss": 0.002,
      "step": 25120
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.1283143311738968,
      "learning_rate": 8.155404932316902e-07,
      "loss": 0.0015,
      "step": 25125
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.2080482542514801,
      "learning_rate": 8.133885203134184e-07,
      "loss": 0.0011,
      "step": 25130
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.21707133948802948,
      "learning_rate": 8.112392699901172e-07,
      "loss": 0.0018,
      "step": 25135
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.5315659642219543,
      "learning_rate": 8.090927428987494e-07,
      "loss": 0.0019,
      "step": 25140
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.1359284371137619,
      "learning_rate": 8.069489396754738e-07,
      "loss": 0.0025,
      "step": 25145
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.13661302626132965,
      "learning_rate": 8.048078609556387e-07,
      "loss": 0.0021,
      "step": 25150
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.2878587245941162,
      "learning_rate": 8.026695073737845e-07,
      "loss": 0.0023,
      "step": 25155
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.09561365097761154,
      "learning_rate": 8.005338795636497e-07,
      "loss": 0.0019,
      "step": 25160
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.1620081216096878,
      "learning_rate": 7.984009781581592e-07,
      "loss": 0.0022,
      "step": 25165
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.08829033374786377,
      "learning_rate": 7.962708037894296e-07,
      "loss": 0.0012,
      "step": 25170
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.17688344419002533,
      "learning_rate": 7.941433570887714e-07,
      "loss": 0.0025,
      "step": 25175
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.18411220610141754,
      "learning_rate": 7.92018638686688e-07,
      "loss": 0.0019,
      "step": 25180
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.24484209716320038,
      "learning_rate": 7.898966492128724e-07,
      "loss": 0.002,
      "step": 25185
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.18548721075057983,
      "learning_rate": 7.87777389296207e-07,
      "loss": 0.001,
      "step": 25190
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.5427772402763367,
      "learning_rate": 7.856608595647686e-07,
      "loss": 0.0025,
      "step": 25195
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.13414634764194489,
      "learning_rate": 7.835470606458273e-07,
      "loss": 0.0022,
      "step": 25200
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.11223254352807999,
      "learning_rate": 7.814359931658356e-07,
      "loss": 0.0013,
      "step": 25205
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.10069742798805237,
      "learning_rate": 7.793276577504428e-07,
      "loss": 0.0012,
      "step": 25210
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.4649175703525543,
      "learning_rate": 7.772220550244858e-07,
      "loss": 0.0024,
      "step": 25215
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.25964003801345825,
      "learning_rate": 7.751191856119933e-07,
      "loss": 0.0013,
      "step": 25220
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.15606264770030975,
      "learning_rate": 7.730190501361845e-07,
      "loss": 0.0015,
      "step": 25225
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.12841029465198517,
      "learning_rate": 7.70921649219466e-07,
      "loss": 0.0029,
      "step": 25230
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.13047796487808228,
      "learning_rate": 7.688269834834349e-07,
      "loss": 0.0013,
      "step": 25235
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.08756542950868607,
      "learning_rate": 7.667350535488782e-07,
      "loss": 0.0018,
      "step": 25240
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.12574045360088348,
      "learning_rate": 7.64645860035772e-07,
      "loss": 0.0015,
      "step": 25245
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.08015564829111099,
      "learning_rate": 7.625594035632821e-07,
      "loss": 0.002,
      "step": 25250
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.10074813663959503,
      "learning_rate": 7.604756847497596e-07,
      "loss": 0.0024,
      "step": 25255
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.1459551900625229,
      "learning_rate": 7.583947042127504e-07,
      "loss": 0.0021,
      "step": 25260
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.13360217213630676,
      "learning_rate": 7.563164625689823e-07,
      "loss": 0.0023,
      "step": 25265
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.16135920584201813,
      "learning_rate": 7.54240960434377e-07,
      "loss": 0.0023,
      "step": 25270
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.19439354538917542,
      "learning_rate": 7.521681984240403e-07,
      "loss": 0.002,
      "step": 25275
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.09518568962812424,
      "learning_rate": 7.500981771522675e-07,
      "loss": 0.0018,
      "step": 25280
    },
    {
      "epoch": 7.89,
      "grad_norm": 0.17542310059070587,
      "learning_rate": 7.480308972325423e-07,
      "loss": 0.0023,
      "step": 25285
    },
    {
      "epoch": 7.89,
      "grad_norm": 0.03590959310531616,
      "learning_rate": 7.459663592775335e-07,
      "loss": 0.0014,
      "step": 25290
    },
    {
      "epoch": 7.89,
      "grad_norm": 0.1374923437833786,
      "learning_rate": 7.439045638991016e-07,
      "loss": 0.0016,
      "step": 25295
    },
    {
      "epoch": 7.89,
      "grad_norm": 0.18914851546287537,
      "learning_rate": 7.418455117082901e-07,
      "loss": 0.0013,
      "step": 25300
    },
    {
      "epoch": 7.89,
      "grad_norm": 0.13574159145355225,
      "learning_rate": 7.397892033153309e-07,
      "loss": 0.002,
      "step": 25305
    },
    {
      "epoch": 7.89,
      "grad_norm": 0.12614206969738007,
      "learning_rate": 7.377356393296442e-07,
      "loss": 0.0026,
      "step": 25310
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.3813757598400116,
      "learning_rate": 7.356848203598321e-07,
      "loss": 0.0023,
      "step": 25315
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.18741311132907867,
      "learning_rate": 7.336367470136906e-07,
      "loss": 0.0017,
      "step": 25320
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.10486096888780594,
      "learning_rate": 7.315914198981977e-07,
      "loss": 0.002,
      "step": 25325
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.20126007497310638,
      "learning_rate": 7.295488396195161e-07,
      "loss": 0.0023,
      "step": 25330
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.1647581160068512,
      "learning_rate": 7.275090067829971e-07,
      "loss": 0.0017,
      "step": 25335
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.12988118827342987,
      "learning_rate": 7.254719219931761e-07,
      "loss": 0.0015,
      "step": 25340
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.19390501081943512,
      "learning_rate": 7.234375858537768e-07,
      "loss": 0.0019,
      "step": 25345
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.2685386836528778,
      "learning_rate": 7.214059989677036e-07,
      "loss": 0.0021,
      "step": 25350
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.13090895116329193,
      "learning_rate": 7.193771619370504e-07,
      "loss": 0.0018,
      "step": 25355
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.1114158183336258,
      "learning_rate": 7.173510753630919e-07,
      "loss": 0.0024,
      "step": 25360
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.03875642642378807,
      "learning_rate": 7.153277398462955e-07,
      "loss": 0.0025,
      "step": 25365
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.17294959723949432,
      "learning_rate": 7.133071559863048e-07,
      "loss": 0.0022,
      "step": 25370
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.25580188632011414,
      "learning_rate": 7.112893243819508e-07,
      "loss": 0.0025,
      "step": 25375
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.17490358650684357,
      "learning_rate": 7.092742456312518e-07,
      "loss": 0.0014,
      "step": 25380
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.12375898659229279,
      "learning_rate": 7.072619203314057e-07,
      "loss": 0.0016,
      "step": 25385
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.11281836777925491,
      "learning_rate": 7.052523490787954e-07,
      "loss": 0.0022,
      "step": 25390
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.15400633215904236,
      "learning_rate": 7.032455324689902e-07,
      "loss": 0.0022,
      "step": 25395
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.0841684639453888,
      "learning_rate": 7.012414710967386e-07,
      "loss": 0.0016,
      "step": 25400
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.09495735913515091,
      "learning_rate": 6.9924016555598e-07,
      "loss": 0.001,
      "step": 25405
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.10059231519699097,
      "learning_rate": 6.972416164398299e-07,
      "loss": 0.0014,
      "step": 25410
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.39723658561706543,
      "learning_rate": 6.952458243405913e-07,
      "loss": 0.0016,
      "step": 25415
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.15220953524112701,
      "learning_rate": 6.932527898497443e-07,
      "loss": 0.002,
      "step": 25420
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.06055636703968048,
      "learning_rate": 6.912625135579587e-07,
      "loss": 0.0018,
      "step": 25425
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.486737459897995,
      "learning_rate": 6.892749960550815e-07,
      "loss": 0.0028,
      "step": 25430
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.09291423112154007,
      "learning_rate": 6.872902379301461e-07,
      "loss": 0.0013,
      "step": 25435
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.07296492159366608,
      "learning_rate": 6.853082397713662e-07,
      "loss": 0.0024,
      "step": 25440
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.057733118534088135,
      "learning_rate": 6.833290021661354e-07,
      "loss": 0.0017,
      "step": 25445
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.09245966374874115,
      "learning_rate": 6.813525257010378e-07,
      "loss": 0.0019,
      "step": 25450
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.14521659910678864,
      "learning_rate": 6.79378810961826e-07,
      "loss": 0.002,
      "step": 25455
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.2611040472984314,
      "learning_rate": 6.774078585334443e-07,
      "loss": 0.0026,
      "step": 25460
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.08619168400764465,
      "learning_rate": 6.754396690000142e-07,
      "loss": 0.0012,
      "step": 25465
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.12047097831964493,
      "learning_rate": 6.734742429448394e-07,
      "loss": 0.0018,
      "step": 25470
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.14527033269405365,
      "learning_rate": 6.715115809504047e-07,
      "loss": 0.0014,
      "step": 25475
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.08826136589050293,
      "learning_rate": 6.695516835983751e-07,
      "loss": 0.0018,
      "step": 25480
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.11847338825464249,
      "learning_rate": 6.67594551469597e-07,
      "loss": 0.0013,
      "step": 25485
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.12033139169216156,
      "learning_rate": 6.656401851440963e-07,
      "loss": 0.0018,
      "step": 25490
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.2235201746225357,
      "learning_rate": 6.636885852010799e-07,
      "loss": 0.0014,
      "step": 25495
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.21314990520477295,
      "learning_rate": 6.617397522189361e-07,
      "loss": 0.0019,
      "step": 25500
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.036323484033346176,
      "learning_rate": 6.597936867752297e-07,
      "loss": 0.0013,
      "step": 25505
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.1177893653512001,
      "learning_rate": 6.578503894467092e-07,
      "loss": 0.0016,
      "step": 25510
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.09829962998628616,
      "learning_rate": 6.559098608092996e-07,
      "loss": 0.0019,
      "step": 25515
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.11559101194143295,
      "learning_rate": 6.539721014381073e-07,
      "loss": 0.0013,
      "step": 25520
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.03609297797083855,
      "learning_rate": 6.520371119074187e-07,
      "loss": 0.0027,
      "step": 25525
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.10191293805837631,
      "learning_rate": 6.50104892790695e-07,
      "loss": 0.0034,
      "step": 25530
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.23136061429977417,
      "learning_rate": 6.481754446605825e-07,
      "loss": 0.0019,
      "step": 25535
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.1560174822807312,
      "learning_rate": 6.462487680889007e-07,
      "loss": 0.0015,
      "step": 25540
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.0929771214723587,
      "learning_rate": 6.443248636466515e-07,
      "loss": 0.0017,
      "step": 25545
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.08027144521474838,
      "learning_rate": 6.424037319040122e-07,
      "loss": 0.0021,
      "step": 25550
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.09332633018493652,
      "learning_rate": 6.404853734303429e-07,
      "loss": 0.0017,
      "step": 25555
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.07819889485836029,
      "learning_rate": 6.385697887941778e-07,
      "loss": 0.0015,
      "step": 25560
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.06993052363395691,
      "learning_rate": 6.36656978563227e-07,
      "loss": 0.0012,
      "step": 25565
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.23164385557174683,
      "learning_rate": 6.347469433043852e-07,
      "loss": 0.002,
      "step": 25570
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.4761601686477661,
      "learning_rate": 6.328396835837203e-07,
      "loss": 0.0018,
      "step": 25575
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.07373147457838058,
      "learning_rate": 6.309351999664781e-07,
      "loss": 0.0016,
      "step": 25580
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.14360737800598145,
      "learning_rate": 6.290334930170816e-07,
      "loss": 0.0017,
      "step": 25585
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.15142029523849487,
      "learning_rate": 6.271345632991321e-07,
      "loss": 0.0024,
      "step": 25590
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.11168301105499268,
      "learning_rate": 6.252384113754073e-07,
      "loss": 0.0018,
      "step": 25595
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.085924431681633,
      "learning_rate": 6.233450378078576e-07,
      "loss": 0.0016,
      "step": 25600
    },
    {
      "epoch": 7.99,
      "grad_norm": 0.1271122246980667,
      "learning_rate": 6.214544431576163e-07,
      "loss": 0.0023,
      "step": 25605
    },
    {
      "epoch": 7.99,
      "grad_norm": 0.06772235035896301,
      "learning_rate": 6.19566627984991e-07,
      "loss": 0.0014,
      "step": 25610
    },
    {
      "epoch": 7.99,
      "grad_norm": 0.18694095313549042,
      "learning_rate": 6.176815928494639e-07,
      "loss": 0.0019,
      "step": 25615
    },
    {
      "epoch": 7.99,
      "grad_norm": 0.8700315356254578,
      "learning_rate": 6.15799338309696e-07,
      "loss": 0.0025,
      "step": 25620
    },
    {
      "epoch": 7.99,
      "grad_norm": 0.08709261566400528,
      "learning_rate": 6.139198649235212e-07,
      "loss": 0.0025,
      "step": 25625
    },
    {
      "epoch": 7.99,
      "grad_norm": 0.11673015356063843,
      "learning_rate": 6.120431732479515e-07,
      "loss": 0.0022,
      "step": 25630
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.3640889823436737,
      "learning_rate": 6.101692638391732e-07,
      "loss": 0.0024,
      "step": 25635
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.18475152552127838,
      "learning_rate": 6.082981372525487e-07,
      "loss": 0.0017,
      "step": 25640
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.0432017557322979,
      "learning_rate": 6.064297940426144e-07,
      "loss": 0.0008,
      "step": 25645
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.19196128845214844,
      "learning_rate": 6.04564234763082e-07,
      "loss": 0.0016,
      "step": 25650
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.11198998987674713,
      "learning_rate": 6.027014599668423e-07,
      "loss": 0.0012,
      "step": 25655
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.028984561562538147,
      "learning_rate": 6.008414702059551e-07,
      "loss": 0.0007,
      "step": 25660
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.10032795369625092,
      "learning_rate": 5.989842660316591e-07,
      "loss": 0.001,
      "step": 25665
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.051030367612838745,
      "learning_rate": 5.971298479943644e-07,
      "loss": 0.0015,
      "step": 25670
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.07989995181560516,
      "learning_rate": 5.952782166436566e-07,
      "loss": 0.0015,
      "step": 25675
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.0383356548845768,
      "learning_rate": 5.934293725282947e-07,
      "loss": 0.0012,
      "step": 25680
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.053743164986371994,
      "learning_rate": 5.915833161962126e-07,
      "loss": 0.0008,
      "step": 25685
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.04896370321512222,
      "learning_rate": 5.89740048194517e-07,
      "loss": 0.0014,
      "step": 25690
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.09783028066158295,
      "learning_rate": 5.878995690694922e-07,
      "loss": 0.0015,
      "step": 25695
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.09386777132749557,
      "learning_rate": 5.860618793665907e-07,
      "loss": 0.0011,
      "step": 25700
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.09913419932126999,
      "learning_rate": 5.842269796304423e-07,
      "loss": 0.0015,
      "step": 25705
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.05181480199098587,
      "learning_rate": 5.823948704048443e-07,
      "loss": 0.0014,
      "step": 25710
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.060823723673820496,
      "learning_rate": 5.805655522327725e-07,
      "loss": 0.0013,
      "step": 25715
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.12706847488880157,
      "learning_rate": 5.78739025656374e-07,
      "loss": 0.0012,
      "step": 25720
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.02192131243646145,
      "learning_rate": 5.769152912169685e-07,
      "loss": 0.001,
      "step": 25725
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.03676176071166992,
      "learning_rate": 5.750943494550465e-07,
      "loss": 0.0006,
      "step": 25730
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.07616845518350601,
      "learning_rate": 5.732762009102732e-07,
      "loss": 0.0012,
      "step": 25735
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.09130016714334488,
      "learning_rate": 5.714608461214888e-07,
      "loss": 0.0012,
      "step": 25740
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.018413733690977097,
      "learning_rate": 5.69648285626696e-07,
      "loss": 0.0011,
      "step": 25745
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.06552843004465103,
      "learning_rate": 5.678385199630787e-07,
      "loss": 0.0013,
      "step": 25750
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.05143953859806061,
      "learning_rate": 5.660315496669888e-07,
      "loss": 0.0009,
      "step": 25755
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.043731290847063065,
      "learning_rate": 5.64227375273948e-07,
      "loss": 0.001,
      "step": 25760
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.09969484806060791,
      "learning_rate": 5.624259973186541e-07,
      "loss": 0.0012,
      "step": 25765
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.0643591657280922,
      "learning_rate": 5.606274163349712e-07,
      "loss": 0.0007,
      "step": 25770
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.12061813473701477,
      "learning_rate": 5.588316328559385e-07,
      "loss": 0.0017,
      "step": 25775
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.07178698480129242,
      "learning_rate": 5.570386474137624e-07,
      "loss": 0.0017,
      "step": 25780
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.052433501929044724,
      "learning_rate": 5.552484605398234e-07,
      "loss": 0.0023,
      "step": 25785
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.053316082805395126,
      "learning_rate": 5.534610727646716e-07,
      "loss": 0.0008,
      "step": 25790
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.11161698400974274,
      "learning_rate": 5.516764846180256e-07,
      "loss": 0.0012,
      "step": 25795
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.05765370652079582,
      "learning_rate": 5.498946966287777e-07,
      "loss": 0.0011,
      "step": 25800
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.021409546956419945,
      "learning_rate": 5.481157093249866e-07,
      "loss": 0.0009,
      "step": 25805
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.16189897060394287,
      "learning_rate": 5.46339523233883e-07,
      "loss": 0.0012,
      "step": 25810
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.07351277023553848,
      "learning_rate": 5.44566138881868e-07,
      "loss": 0.0006,
      "step": 25815
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.09774326533079147,
      "learning_rate": 5.42795556794512e-07,
      "loss": 0.0008,
      "step": 25820
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.029827700927853584,
      "learning_rate": 5.410277774965533e-07,
      "loss": 0.0012,
      "step": 25825
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.05322696641087532,
      "learning_rate": 5.392628015119017e-07,
      "loss": 0.001,
      "step": 25830
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.0880359560251236,
      "learning_rate": 5.375006293636342e-07,
      "loss": 0.001,
      "step": 25835
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.02932639606297016,
      "learning_rate": 5.357412615739999e-07,
      "loss": 0.0011,
      "step": 25840
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.09115365147590637,
      "learning_rate": 5.339846986644126e-07,
      "loss": 0.001,
      "step": 25845
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.057407550513744354,
      "learning_rate": 5.322309411554583e-07,
      "loss": 0.001,
      "step": 25850
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.057419128715991974,
      "learning_rate": 5.304799895668877e-07,
      "loss": 0.0006,
      "step": 25855
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.0982402116060257,
      "learning_rate": 5.287318444176259e-07,
      "loss": 0.0014,
      "step": 25860
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.08957464247941971,
      "learning_rate": 5.269865062257618e-07,
      "loss": 0.0012,
      "step": 25865
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.05921388417482376,
      "learning_rate": 5.252439755085536e-07,
      "loss": 0.0009,
      "step": 25870
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.02325126901268959,
      "learning_rate": 5.235042527824264e-07,
      "loss": 0.0013,
      "step": 25875
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.059985339641571045,
      "learning_rate": 5.21767338562974e-07,
      "loss": 0.0012,
      "step": 25880
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.08335838466882706,
      "learning_rate": 5.200332333649605e-07,
      "loss": 0.0011,
      "step": 25885
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.09611143916845322,
      "learning_rate": 5.183019377023113e-07,
      "loss": 0.0017,
      "step": 25890
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.03355991095304489,
      "learning_rate": 5.165734520881227e-07,
      "loss": 0.0008,
      "step": 25895
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.06265156716108322,
      "learning_rate": 5.148477770346616e-07,
      "loss": 0.0009,
      "step": 25900
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.07441845536231995,
      "learning_rate": 5.131249130533567e-07,
      "loss": 0.0014,
      "step": 25905
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.06214327737689018,
      "learning_rate": 5.114048606548061e-07,
      "loss": 0.001,
      "step": 25910
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.059696801006793976,
      "learning_rate": 5.096876203487733e-07,
      "loss": 0.001,
      "step": 25915
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.03962871804833412,
      "learning_rate": 5.079731926441911e-07,
      "loss": 0.0014,
      "step": 25920
    },
    {
      "epoch": 8.09,
      "grad_norm": 0.10176525264978409,
      "learning_rate": 5.062615780491531e-07,
      "loss": 0.001,
      "step": 25925
    },
    {
      "epoch": 8.09,
      "grad_norm": 0.07909473031759262,
      "learning_rate": 5.045527770709258e-07,
      "loss": 0.001,
      "step": 25930
    },
    {
      "epoch": 8.09,
      "grad_norm": 0.09313105791807175,
      "learning_rate": 5.028467902159373e-07,
      "loss": 0.0012,
      "step": 25935
    },
    {
      "epoch": 8.09,
      "grad_norm": 0.06423319876194,
      "learning_rate": 5.011436179897833e-07,
      "loss": 0.0007,
      "step": 25940
    },
    {
      "epoch": 8.09,
      "grad_norm": 0.08090867102146149,
      "learning_rate": 4.994432608972266e-07,
      "loss": 0.0008,
      "step": 25945
    },
    {
      "epoch": 8.09,
      "grad_norm": 0.04727114737033844,
      "learning_rate": 4.977457194421941e-07,
      "loss": 0.0008,
      "step": 25950
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.11219633370637894,
      "learning_rate": 4.960509941277792e-07,
      "loss": 0.0013,
      "step": 25955
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.09417059272527695,
      "learning_rate": 4.943590854562397e-07,
      "loss": 0.0016,
      "step": 25960
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.03398860618472099,
      "learning_rate": 4.92669993928997e-07,
      "loss": 0.0009,
      "step": 25965
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.06553380936384201,
      "learning_rate": 4.909837200466416e-07,
      "loss": 0.0017,
      "step": 25970
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.08675870299339294,
      "learning_rate": 4.893002643089262e-07,
      "loss": 0.0014,
      "step": 25975
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.05225129798054695,
      "learning_rate": 4.876196272147693e-07,
      "loss": 0.0008,
      "step": 25980
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.13486607372760773,
      "learning_rate": 4.859418092622525e-07,
      "loss": 0.0015,
      "step": 25985
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.08578328043222427,
      "learning_rate": 4.842668109486271e-07,
      "loss": 0.0011,
      "step": 25990
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.08599358797073364,
      "learning_rate": 4.825946327703024e-07,
      "loss": 0.0015,
      "step": 25995
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.04679775983095169,
      "learning_rate": 4.809252752228522e-07,
      "loss": 0.0007,
      "step": 26000
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.09810055792331696,
      "learning_rate": 4.792587388010195e-07,
      "loss": 0.0015,
      "step": 26005
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.028725087642669678,
      "learning_rate": 4.77595023998707e-07,
      "loss": 0.0014,
      "step": 26010
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.09821700304746628,
      "learning_rate": 4.759341313089838e-07,
      "loss": 0.0011,
      "step": 26015
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.054021578282117844,
      "learning_rate": 4.742760612240793e-07,
      "loss": 0.0011,
      "step": 26020
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.09391536563634872,
      "learning_rate": 4.7262081423538717e-07,
      "loss": 0.0008,
      "step": 26025
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.05493954196572304,
      "learning_rate": 4.7096839083347166e-07,
      "loss": 0.0009,
      "step": 26030
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.012998754158616066,
      "learning_rate": 4.693187915080477e-07,
      "loss": 0.001,
      "step": 26035
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.10303235054016113,
      "learning_rate": 4.676720167480031e-07,
      "loss": 0.0013,
      "step": 26040
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.05119301378726959,
      "learning_rate": 4.6602806704138304e-07,
      "loss": 0.0009,
      "step": 26045
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.043302636593580246,
      "learning_rate": 4.643869428753989e-07,
      "loss": 0.0017,
      "step": 26050
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.029808012768626213,
      "learning_rate": 4.6274864473642266e-07,
      "loss": 0.0007,
      "step": 26055
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.028385180979967117,
      "learning_rate": 4.611131731099905e-07,
      "loss": 0.0012,
      "step": 26060
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.05698021501302719,
      "learning_rate": 4.594805284807979e-07,
      "loss": 0.0012,
      "step": 26065
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.050949130207300186,
      "learning_rate": 4.578507113327069e-07,
      "loss": 0.0009,
      "step": 26070
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.04433879628777504,
      "learning_rate": 4.5622372214873644e-07,
      "loss": 0.0014,
      "step": 26075
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.052618153393268585,
      "learning_rate": 4.545995614110721e-07,
      "loss": 0.001,
      "step": 26080
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.061884429305791855,
      "learning_rate": 4.5297822960105877e-07,
      "loss": 0.0013,
      "step": 26085
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.05236772447824478,
      "learning_rate": 4.513597271992032e-07,
      "loss": 0.0009,
      "step": 26090
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.08672960102558136,
      "learning_rate": 4.497440546851728e-07,
      "loss": 0.0012,
      "step": 26095
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.02289748750627041,
      "learning_rate": 4.4813121253779787e-07,
      "loss": 0.0014,
      "step": 26100
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.0577850304543972,
      "learning_rate": 4.465212012350695e-07,
      "loss": 0.0012,
      "step": 26105
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.062029387801885605,
      "learning_rate": 4.4491402125413917e-07,
      "loss": 0.0019,
      "step": 26110
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.05040305480360985,
      "learning_rate": 4.4330967307132043e-07,
      "loss": 0.0011,
      "step": 26115
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.08229151368141174,
      "learning_rate": 4.4170815716208514e-07,
      "loss": 0.0012,
      "step": 26120
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.060748420655727386,
      "learning_rate": 4.401094740010692e-07,
      "loss": 0.0012,
      "step": 26125
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.07263582944869995,
      "learning_rate": 4.3851362406206575e-07,
      "loss": 0.0012,
      "step": 26130
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.020791111513972282,
      "learning_rate": 4.369206078180299e-07,
      "loss": 0.001,
      "step": 26135
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.08441232144832611,
      "learning_rate": 4.3533042574107953e-07,
      "loss": 0.0015,
      "step": 26140
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.13795146346092224,
      "learning_rate": 4.337430783024843e-07,
      "loss": 0.0012,
      "step": 26145
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.05792690068483353,
      "learning_rate": 4.3215856597268346e-07,
      "loss": 0.0014,
      "step": 26150
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.014939947985112667,
      "learning_rate": 4.305768892212714e-07,
      "loss": 0.0011,
      "step": 26155
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.12674427032470703,
      "learning_rate": 4.28998048517002e-07,
      "loss": 0.0014,
      "step": 26160
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.09565210342407227,
      "learning_rate": 4.274220443277888e-07,
      "loss": 0.0011,
      "step": 26165
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.10866738110780716,
      "learning_rate": 4.2584887712070587e-07,
      "loss": 0.0011,
      "step": 26170
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.04313337057828903,
      "learning_rate": 4.2427854736198594e-07,
      "loss": 0.0006,
      "step": 26175
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.06837520748376846,
      "learning_rate": 4.2271105551702e-07,
      "loss": 0.0011,
      "step": 26180
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.04312984645366669,
      "learning_rate": 4.211464020503564e-07,
      "loss": 0.0009,
      "step": 26185
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.058144159615039825,
      "learning_rate": 4.195845874257087e-07,
      "loss": 0.0005,
      "step": 26190
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.04488474503159523,
      "learning_rate": 4.180256121059423e-07,
      "loss": 0.0011,
      "step": 26195
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.05889725685119629,
      "learning_rate": 4.164694765530841e-07,
      "loss": 0.0007,
      "step": 26200
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.04025601968169212,
      "learning_rate": 4.149161812283209e-07,
      "loss": 0.0012,
      "step": 26205
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.29714614152908325,
      "learning_rate": 4.1336572659199435e-07,
      "loss": 0.0015,
      "step": 26210
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.2917931377887726,
      "learning_rate": 4.1181811310360476e-07,
      "loss": 0.0023,
      "step": 26215
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.04373086243867874,
      "learning_rate": 4.1027334122181297e-07,
      "loss": 0.0006,
      "step": 26220
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.012117493897676468,
      "learning_rate": 4.0873141140443496e-07,
      "loss": 0.0004,
      "step": 26225
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.10136426985263824,
      "learning_rate": 4.0719232410844414e-07,
      "loss": 0.0014,
      "step": 26230
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.2507619261741638,
      "learning_rate": 4.0565607978997777e-07,
      "loss": 0.0017,
      "step": 26235
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.0383005253970623,
      "learning_rate": 4.041226789043218e-07,
      "loss": 0.001,
      "step": 26240
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.06149076670408249,
      "learning_rate": 4.02592121905927e-07,
      "loss": 0.0012,
      "step": 26245
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.07256074249744415,
      "learning_rate": 4.0106440924839286e-07,
      "loss": 0.0014,
      "step": 26250
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.08970825374126434,
      "learning_rate": 3.995395413844838e-07,
      "loss": 0.0009,
      "step": 26255
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.08046100288629532,
      "learning_rate": 3.9801751876611616e-07,
      "loss": 0.0014,
      "step": 26260
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.06438751518726349,
      "learning_rate": 3.9649834184436797e-07,
      "loss": 0.0012,
      "step": 26265
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.03636743873357773,
      "learning_rate": 3.94982011069468e-07,
      "loss": 0.0012,
      "step": 26270
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.04495951533317566,
      "learning_rate": 3.9346852689080453e-07,
      "loss": 0.0008,
      "step": 26275
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.1346733570098877,
      "learning_rate": 3.919578897569254e-07,
      "loss": 0.0017,
      "step": 26280
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.09756915271282196,
      "learning_rate": 3.904501001155281e-07,
      "loss": 0.0015,
      "step": 26285
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.10456876456737518,
      "learning_rate": 3.8894515841347067e-07,
      "loss": 0.0008,
      "step": 26290
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.03999922797083855,
      "learning_rate": 3.8744306509676534e-07,
      "loss": 0.0006,
      "step": 26295
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.05138784274458885,
      "learning_rate": 3.8594382061058144e-07,
      "loss": 0.0009,
      "step": 26300
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.042921282351017,
      "learning_rate": 3.8444742539924364e-07,
      "loss": 0.0016,
      "step": 26305
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.04632921516895294,
      "learning_rate": 3.8295387990623156e-07,
      "loss": 0.0008,
      "step": 26310
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.13882958889007568,
      "learning_rate": 3.8146318457417784e-07,
      "loss": 0.0014,
      "step": 26315
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.07300465553998947,
      "learning_rate": 3.7997533984488024e-07,
      "loss": 0.0015,
      "step": 26320
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.05495378002524376,
      "learning_rate": 3.7849034615927706e-07,
      "loss": 0.0007,
      "step": 26325
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.06700440496206284,
      "learning_rate": 3.770082039574741e-07,
      "loss": 0.0009,
      "step": 26330
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.08584974706172943,
      "learning_rate": 3.7552891367872433e-07,
      "loss": 0.0009,
      "step": 26335
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.04152682423591614,
      "learning_rate": 3.7405247576144055e-07,
      "loss": 0.0016,
      "step": 26340
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.0706639513373375,
      "learning_rate": 3.7257889064318596e-07,
      "loss": 0.0018,
      "step": 26345
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.08724109083414078,
      "learning_rate": 3.711081587606824e-07,
      "loss": 0.0009,
      "step": 26350
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.014516488648951054,
      "learning_rate": 3.696402805498034e-07,
      "loss": 0.0009,
      "step": 26355
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.10170592367649078,
      "learning_rate": 3.681752564455765e-07,
      "loss": 0.0009,
      "step": 26360
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.0802757740020752,
      "learning_rate": 3.6671308688218556e-07,
      "loss": 0.0013,
      "step": 26365
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.20823164284229279,
      "learning_rate": 3.652537722929672e-07,
      "loss": 0.0017,
      "step": 26370
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.09853595495223999,
      "learning_rate": 3.637973131104111e-07,
      "loss": 0.0008,
      "step": 26375
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.09480353444814682,
      "learning_rate": 3.6234370976616196e-07,
      "loss": 0.0012,
      "step": 26380
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.1411803662776947,
      "learning_rate": 3.6089296269101736e-07,
      "loss": 0.0012,
      "step": 26385
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.1442306637763977,
      "learning_rate": 3.594450723149301e-07,
      "loss": 0.0012,
      "step": 26390
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.03480003401637077,
      "learning_rate": 3.580000390670035e-07,
      "loss": 0.0008,
      "step": 26395
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.059012580662965775,
      "learning_rate": 3.5655786337549626e-07,
      "loss": 0.001,
      "step": 26400
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.1240774467587471,
      "learning_rate": 3.5511854566781966e-07,
      "loss": 0.0013,
      "step": 26405
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.11328074336051941,
      "learning_rate": 3.53682086370537e-07,
      "loss": 0.0011,
      "step": 26410
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.03457697853446007,
      "learning_rate": 3.522484859093667e-07,
      "loss": 0.0005,
      "step": 26415
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.0656462237238884,
      "learning_rate": 3.5081774470917763e-07,
      "loss": 0.0009,
      "step": 26420
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.054810818284749985,
      "learning_rate": 3.4938986319399295e-07,
      "loss": 0.0015,
      "step": 26425
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.20973767340183258,
      "learning_rate": 3.479648417869863e-07,
      "loss": 0.0011,
      "step": 26430
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.058540064841508865,
      "learning_rate": 3.4654268091048324e-07,
      "loss": 0.0012,
      "step": 26435
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.059108808636665344,
      "learning_rate": 3.451233809859678e-07,
      "loss": 0.0013,
      "step": 26440
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.1766173541545868,
      "learning_rate": 3.437069424340678e-07,
      "loss": 0.0014,
      "step": 26445
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.07533004879951477,
      "learning_rate": 3.4229336567456974e-07,
      "loss": 0.001,
      "step": 26450
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.1083541065454483,
      "learning_rate": 3.408826511264063e-07,
      "loss": 0.0009,
      "step": 26455
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.1199837252497673,
      "learning_rate": 3.394747992076663e-07,
      "loss": 0.0012,
      "step": 26460
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.06192876398563385,
      "learning_rate": 3.380698103355884e-07,
      "loss": 0.0007,
      "step": 26465
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.11188165098428726,
      "learning_rate": 3.3666768492656264e-07,
      "loss": 0.0009,
      "step": 26470
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.023300036787986755,
      "learning_rate": 3.352684233961301e-07,
      "loss": 0.0009,
      "step": 26475
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.08266016840934753,
      "learning_rate": 3.3387202615898226e-07,
      "loss": 0.001,
      "step": 26480
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.04422038048505783,
      "learning_rate": 3.3247849362896713e-07,
      "loss": 0.0008,
      "step": 26485
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.09365659952163696,
      "learning_rate": 3.3108782621907643e-07,
      "loss": 0.0012,
      "step": 26490
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.020148232579231262,
      "learning_rate": 3.2970002434145833e-07,
      "loss": 0.001,
      "step": 26495
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.09790826588869095,
      "learning_rate": 3.283150884074082e-07,
      "loss": 0.0015,
      "step": 26500
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.033936236053705215,
      "learning_rate": 3.269330188273734e-07,
      "loss": 0.0008,
      "step": 26505
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.17181117832660675,
      "learning_rate": 3.2555381601095057e-07,
      "loss": 0.0018,
      "step": 26510
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.09773623198270798,
      "learning_rate": 3.241774803668907e-07,
      "loss": 0.0008,
      "step": 26515
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.08878730237483978,
      "learning_rate": 3.2280401230308845e-07,
      "loss": 0.0013,
      "step": 26520
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.09682467579841614,
      "learning_rate": 3.2143341222659495e-07,
      "loss": 0.0006,
      "step": 26525
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.07581570744514465,
      "learning_rate": 3.2006568054360974e-07,
      "loss": 0.0009,
      "step": 26530
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.10100431740283966,
      "learning_rate": 3.187008176594797e-07,
      "loss": 0.0011,
      "step": 26535
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.07675759494304657,
      "learning_rate": 3.173388239787034e-07,
      "loss": 0.0012,
      "step": 26540
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.07665850967168808,
      "learning_rate": 3.1597969990492804e-07,
      "loss": 0.0005,
      "step": 26545
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.030014360323548317,
      "learning_rate": 3.146234458409525e-07,
      "loss": 0.001,
      "step": 26550
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.09450791031122208,
      "learning_rate": 3.1327006218872303e-07,
      "loss": 0.001,
      "step": 26555
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.08888278901576996,
      "learning_rate": 3.1191954934933657e-07,
      "loss": 0.0012,
      "step": 26560
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.06696563959121704,
      "learning_rate": 3.1057190772303626e-07,
      "loss": 0.0018,
      "step": 26565
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.046966444700956345,
      "learning_rate": 3.0922713770922155e-07,
      "loss": 0.0008,
      "step": 26570
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.029699940234422684,
      "learning_rate": 3.0788523970643027e-07,
      "loss": 0.0013,
      "step": 26575
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.07722587883472443,
      "learning_rate": 3.0654621411235874e-07,
      "loss": 0.0011,
      "step": 26580
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.10917071253061295,
      "learning_rate": 3.0521006132384513e-07,
      "loss": 0.0007,
      "step": 26585
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.02738550864160061,
      "learning_rate": 3.038767817368815e-07,
      "loss": 0.0009,
      "step": 26590
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.11273742467164993,
      "learning_rate": 3.0254637574660406e-07,
      "loss": 0.0012,
      "step": 26595
    },
    {
      "epoch": 8.3,
      "grad_norm": 0.024482088163495064,
      "learning_rate": 3.012188437473007e-07,
      "loss": 0.0008,
      "step": 26600
    },
    {
      "epoch": 8.3,
      "grad_norm": 0.0887916088104248,
      "learning_rate": 2.998941861324056e-07,
      "loss": 0.0009,
      "step": 26605
    },
    {
      "epoch": 8.3,
      "grad_norm": 0.04568292200565338,
      "learning_rate": 2.985724032945003e-07,
      "loss": 0.0008,
      "step": 26610
    },
    {
      "epoch": 8.3,
      "grad_norm": 0.07197744399309158,
      "learning_rate": 2.972534956253181e-07,
      "loss": 0.0016,
      "step": 26615
    },
    {
      "epoch": 8.3,
      "grad_norm": 0.11375932395458221,
      "learning_rate": 2.959374635157364e-07,
      "loss": 0.0011,
      "step": 26620
    },
    {
      "epoch": 8.3,
      "grad_norm": 0.07851669937372208,
      "learning_rate": 2.946243073557831e-07,
      "loss": 0.0008,
      "step": 26625
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.15448705852031708,
      "learning_rate": 2.933140275346291e-07,
      "loss": 0.0009,
      "step": 26630
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.06736727058887482,
      "learning_rate": 2.9200662444060056e-07,
      "loss": 0.0009,
      "step": 26635
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.19965125620365143,
      "learning_rate": 2.9070209846116303e-07,
      "loss": 0.0014,
      "step": 26640
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.037576086819171906,
      "learning_rate": 2.89400449982935e-07,
      "loss": 0.0007,
      "step": 26645
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.0317411944270134,
      "learning_rate": 2.881016793916791e-07,
      "loss": 0.0006,
      "step": 26650
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.07414986193180084,
      "learning_rate": 2.868057870723073e-07,
      "loss": 0.001,
      "step": 26655
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.33764782547950745,
      "learning_rate": 2.8551277340887696e-07,
      "loss": 0.0011,
      "step": 26660
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.02444450557231903,
      "learning_rate": 2.8422263878459256e-07,
      "loss": 0.0007,
      "step": 26665
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.0931888222694397,
      "learning_rate": 2.8293538358180607e-07,
      "loss": 0.0015,
      "step": 26670
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.053565751761198044,
      "learning_rate": 2.816510081820145e-07,
      "loss": 0.001,
      "step": 26675
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.1053864061832428,
      "learning_rate": 2.8036951296586346e-07,
      "loss": 0.0007,
      "step": 26680
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.249299556016922,
      "learning_rate": 2.790908983131435e-07,
      "loss": 0.0017,
      "step": 26685
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.07624279707670212,
      "learning_rate": 2.7781516460279154e-07,
      "loss": 0.0013,
      "step": 26690
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.06324794143438339,
      "learning_rate": 2.7654231221289187e-07,
      "loss": 0.0012,
      "step": 26695
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.0402330718934536,
      "learning_rate": 2.752723415206737e-07,
      "loss": 0.0013,
      "step": 26700
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.0825943797826767,
      "learning_rate": 2.740052529025128e-07,
      "loss": 0.0013,
      "step": 26705
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.08207705616950989,
      "learning_rate": 2.727410467339309e-07,
      "loss": 0.0011,
      "step": 26710
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.07926084846258163,
      "learning_rate": 2.714797233895949e-07,
      "loss": 0.0014,
      "step": 26715
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.02722475863993168,
      "learning_rate": 2.70221283243316e-07,
      "loss": 0.0012,
      "step": 26720
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.017125660553574562,
      "learning_rate": 2.689657266680534e-07,
      "loss": 0.001,
      "step": 26725
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.070092111825943,
      "learning_rate": 2.6771305403591184e-07,
      "loss": 0.001,
      "step": 26730
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.1087091714143753,
      "learning_rate": 2.6646326571813984e-07,
      "loss": 0.0009,
      "step": 26735
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.08016277104616165,
      "learning_rate": 2.652163620851311e-07,
      "loss": 0.001,
      "step": 26740
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.07137833535671234,
      "learning_rate": 2.639723435064245e-07,
      "loss": 0.0009,
      "step": 26745
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.031105132773518562,
      "learning_rate": 2.627312103507063e-07,
      "loss": 0.0011,
      "step": 26750
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.0440361313521862,
      "learning_rate": 2.614929629858043e-07,
      "loss": 0.0011,
      "step": 26755
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.007057182025164366,
      "learning_rate": 2.6025760177869063e-07,
      "loss": 0.0012,
      "step": 26760
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.07112181931734085,
      "learning_rate": 2.590251270954858e-07,
      "loss": 0.0012,
      "step": 26765
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.11220315098762512,
      "learning_rate": 2.5779553930145194e-07,
      "loss": 0.002,
      "step": 26770
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.09323884546756744,
      "learning_rate": 2.565688387609966e-07,
      "loss": 0.0011,
      "step": 26775
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.046231284737586975,
      "learning_rate": 2.553450258376733e-07,
      "loss": 0.0009,
      "step": 26780
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.11686007678508759,
      "learning_rate": 2.5412410089417525e-07,
      "loss": 0.0011,
      "step": 26785
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.07207207381725311,
      "learning_rate": 2.5290606429234534e-07,
      "loss": 0.0012,
      "step": 26790
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.008922003209590912,
      "learning_rate": 2.516909163931658e-07,
      "loss": 0.001,
      "step": 26795
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.0872945487499237,
      "learning_rate": 2.5047865755676413e-07,
      "loss": 0.0012,
      "step": 26800
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.12244457006454468,
      "learning_rate": 2.4926928814241305e-07,
      "loss": 0.0012,
      "step": 26805
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.18743190169334412,
      "learning_rate": 2.480628085085257e-07,
      "loss": 0.0009,
      "step": 26810
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.24648462235927582,
      "learning_rate": 2.4685921901266505e-07,
      "loss": 0.0008,
      "step": 26815
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.03587842360138893,
      "learning_rate": 2.456585200115313e-07,
      "loss": 0.0007,
      "step": 26820
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.08156446367502213,
      "learning_rate": 2.444607118609721e-07,
      "loss": 0.0008,
      "step": 26825
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.20609742403030396,
      "learning_rate": 2.4326579491597335e-07,
      "loss": 0.0015,
      "step": 26830
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.0752704069018364,
      "learning_rate": 2.4207376953066853e-07,
      "loss": 0.0017,
      "step": 26835
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.1400754451751709,
      "learning_rate": 2.408846360583339e-07,
      "loss": 0.0013,
      "step": 26840
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.023851117119193077,
      "learning_rate": 2.3969839485138644e-07,
      "loss": 0.0008,
      "step": 26845
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.028327608481049538,
      "learning_rate": 2.385150462613883e-07,
      "loss": 0.001,
      "step": 26850
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.10897330194711685,
      "learning_rate": 2.3733459063904009e-07,
      "loss": 0.0009,
      "step": 26855
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.047635432332754135,
      "learning_rate": 2.3615702833419297e-07,
      "loss": 0.0014,
      "step": 26860
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.19598473608493805,
      "learning_rate": 2.349823596958334e-07,
      "loss": 0.0017,
      "step": 26865
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.051575757563114166,
      "learning_rate": 2.3381058507209175e-07,
      "loss": 0.0012,
      "step": 26870
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.08787445724010468,
      "learning_rate": 2.326417048102425e-07,
      "loss": 0.0006,
      "step": 26875
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.20478199422359467,
      "learning_rate": 2.3147571925670187e-07,
      "loss": 0.0015,
      "step": 26880
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.11912625283002853,
      "learning_rate": 2.3031262875702787e-07,
      "loss": 0.0017,
      "step": 26885
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.11537455767393112,
      "learning_rate": 2.2915243365591923e-07,
      "loss": 0.0011,
      "step": 26890
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.18906544148921967,
      "learning_rate": 2.2799513429721865e-07,
      "loss": 0.0014,
      "step": 26895
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.07237179577350616,
      "learning_rate": 2.268407310239107e-07,
      "loss": 0.0008,
      "step": 26900
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.15105654299259186,
      "learning_rate": 2.2568922417811835e-07,
      "loss": 0.001,
      "step": 26905
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.0675887018442154,
      "learning_rate": 2.2454061410111083e-07,
      "loss": 0.0008,
      "step": 26910
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.14444920420646667,
      "learning_rate": 2.2339490113329475e-07,
      "loss": 0.002,
      "step": 26915
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.0659114420413971,
      "learning_rate": 2.2225208561422185e-07,
      "loss": 0.001,
      "step": 26920
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.09269290417432785,
      "learning_rate": 2.2111216788258117e-07,
      "loss": 0.0013,
      "step": 26925
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.15672466158866882,
      "learning_rate": 2.1997514827620693e-07,
      "loss": 0.0014,
      "step": 26930
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.0962071493268013,
      "learning_rate": 2.1884102713207177e-07,
      "loss": 0.001,
      "step": 26935
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.13821633160114288,
      "learning_rate": 2.177098047862891e-07,
      "loss": 0.0011,
      "step": 26940
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.025947265326976776,
      "learning_rate": 2.1658148157411628e-07,
      "loss": 0.0007,
      "step": 26945
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.04380210489034653,
      "learning_rate": 2.1545605782994693e-07,
      "loss": 0.0006,
      "step": 26950
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.04115161672234535,
      "learning_rate": 2.143335338873198e-07,
      "loss": 0.0013,
      "step": 26955
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.11357036978006363,
      "learning_rate": 2.1321391007891211e-07,
      "loss": 0.0012,
      "step": 26960
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.09247150272130966,
      "learning_rate": 2.1209718673654177e-07,
      "loss": 0.0012,
      "step": 26965
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.039926059544086456,
      "learning_rate": 2.1098336419116628e-07,
      "loss": 0.0009,
      "step": 26970
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.029094960540533066,
      "learning_rate": 2.0987244277288377e-07,
      "loss": 0.0009,
      "step": 26975
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.05168910324573517,
      "learning_rate": 2.0876442281093423e-07,
      "loss": 0.0007,
      "step": 26980
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.04371324181556702,
      "learning_rate": 2.076593046336972e-07,
      "loss": 0.0012,
      "step": 26985
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.03896841034293175,
      "learning_rate": 2.065570885686896e-07,
      "loss": 0.0012,
      "step": 26990
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.07638699561357498,
      "learning_rate": 2.0545777494257235e-07,
      "loss": 0.001,
      "step": 26995
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.0764116644859314,
      "learning_rate": 2.043613640811426e-07,
      "loss": 0.0009,
      "step": 27000
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.07070528715848923,
      "learning_rate": 2.0326785630934043e-07,
      "loss": 0.0009,
      "step": 27005
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.0403977669775486,
      "learning_rate": 2.0217725195124104e-07,
      "loss": 0.0014,
      "step": 27010
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.12113960087299347,
      "learning_rate": 2.010895513300637e-07,
      "loss": 0.0013,
      "step": 27015
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.06756232678890228,
      "learning_rate": 2.0000475476816495e-07,
      "loss": 0.0008,
      "step": 27020
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.1345250904560089,
      "learning_rate": 1.989228625870432e-07,
      "loss": 0.0011,
      "step": 27025
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.05166436731815338,
      "learning_rate": 1.9784387510733083e-07,
      "loss": 0.0009,
      "step": 27030
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.07139864563941956,
      "learning_rate": 1.967677926488043e-07,
      "loss": 0.001,
      "step": 27035
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.06834730505943298,
      "learning_rate": 1.956946155303785e-07,
      "loss": 0.001,
      "step": 27040
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.15904085338115692,
      "learning_rate": 1.9462434407010345e-07,
      "loss": 0.0014,
      "step": 27045
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.07280880957841873,
      "learning_rate": 1.9355697858517098e-07,
      "loss": 0.0013,
      "step": 27050
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.04626739025115967,
      "learning_rate": 1.924925193919136e-07,
      "loss": 0.001,
      "step": 27055
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.05268017202615738,
      "learning_rate": 1.9143096680579787e-07,
      "loss": 0.0008,
      "step": 27060
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.07796227931976318,
      "learning_rate": 1.9037232114143323e-07,
      "loss": 0.0008,
      "step": 27065
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.33021658658981323,
      "learning_rate": 1.8931658271256426e-07,
      "loss": 0.0014,
      "step": 27070
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.077763132750988,
      "learning_rate": 1.8826375183207847e-07,
      "loss": 0.0011,
      "step": 27075
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.17922165989875793,
      "learning_rate": 1.8721382881199512e-07,
      "loss": 0.0017,
      "step": 27080
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.05037185922265053,
      "learning_rate": 1.8616681396347646e-07,
      "loss": 0.0006,
      "step": 27085
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.06594012677669525,
      "learning_rate": 1.8512270759682204e-07,
      "loss": 0.0006,
      "step": 27090
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.07595254480838776,
      "learning_rate": 1.8408151002146878e-07,
      "loss": 0.0008,
      "step": 27095
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.06146182119846344,
      "learning_rate": 1.8304322154599097e-07,
      "loss": 0.0011,
      "step": 27100
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.06487392634153366,
      "learning_rate": 1.8200784247810243e-07,
      "loss": 0.0012,
      "step": 27105
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.11371538043022156,
      "learning_rate": 1.8097537312465442e-07,
      "loss": 0.0008,
      "step": 27110
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.08306070417165756,
      "learning_rate": 1.7994581379163323e-07,
      "loss": 0.0007,
      "step": 27115
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.09612597525119781,
      "learning_rate": 1.7891916478416593e-07,
      "loss": 0.0008,
      "step": 27120
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.037405677139759064,
      "learning_rate": 1.7789542640651692e-07,
      "loss": 0.0011,
      "step": 27125
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.12324880063533783,
      "learning_rate": 1.768745989620857e-07,
      "loss": 0.0006,
      "step": 27130
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.048658937215805054,
      "learning_rate": 1.7606003308179187e-07,
      "loss": 0.001,
      "step": 27135
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.03540576621890068,
      "learning_rate": 1.7504444607896797e-07,
      "loss": 0.0013,
      "step": 27140
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.06150227412581444,
      "learning_rate": 1.74031770854296e-07,
      "loss": 0.001,
      "step": 27145
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.1833973377943039,
      "learning_rate": 1.730220077078959e-07,
      "loss": 0.0018,
      "step": 27150
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.23827354609966278,
      "learning_rate": 1.7201515693902605e-07,
      "loss": 0.0011,
      "step": 27155
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.058667294681072235,
      "learning_rate": 1.710112188460844e-07,
      "loss": 0.0009,
      "step": 27160
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.07912705093622208,
      "learning_rate": 1.7001019372660298e-07,
      "loss": 0.0013,
      "step": 27165
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.09708718210458755,
      "learning_rate": 1.6901208187725115e-07,
      "loss": 0.0013,
      "step": 27170
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.1299380362033844,
      "learning_rate": 1.680168835938334e-07,
      "loss": 0.0012,
      "step": 27175
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.07040596753358841,
      "learning_rate": 1.6702459917129376e-07,
      "loss": 0.0011,
      "step": 27180
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.09709855914115906,
      "learning_rate": 1.6603522890370927e-07,
      "loss": 0.0013,
      "step": 27185
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.027560172602534294,
      "learning_rate": 1.6504877308429757e-07,
      "loss": 0.0009,
      "step": 27190
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.10199445486068726,
      "learning_rate": 1.6406523200540926e-07,
      "loss": 0.0009,
      "step": 27195
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.08776000142097473,
      "learning_rate": 1.6308460595853227e-07,
      "loss": 0.0007,
      "step": 27200
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.046178627759218216,
      "learning_rate": 1.6210689523428968e-07,
      "loss": 0.0008,
      "step": 27205
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.025471661239862442,
      "learning_rate": 1.6113210012244196e-07,
      "loss": 0.0011,
      "step": 27210
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.2755342125892639,
      "learning_rate": 1.6016022091188354e-07,
      "loss": 0.0016,
      "step": 27215
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.1328873336315155,
      "learning_rate": 1.5919125789064627e-07,
      "loss": 0.0013,
      "step": 27220
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.06281613558530807,
      "learning_rate": 1.5822521134589819e-07,
      "loss": 0.0011,
      "step": 27225
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.11914748698472977,
      "learning_rate": 1.5726208156394252e-07,
      "loss": 0.0016,
      "step": 27230
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.04355286806821823,
      "learning_rate": 1.5630186883021647e-07,
      "loss": 0.0011,
      "step": 27235
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.13519792258739471,
      "learning_rate": 1.5534457342929465e-07,
      "loss": 0.0011,
      "step": 27240
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.08627419173717499,
      "learning_rate": 1.5439019564488677e-07,
      "loss": 0.0016,
      "step": 27245
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.14163166284561157,
      "learning_rate": 1.534387357598366e-07,
      "loss": 0.0012,
      "step": 27250
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.0620029978454113,
      "learning_rate": 1.524901940561252e-07,
      "loss": 0.0011,
      "step": 27255
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.11000042408704758,
      "learning_rate": 1.5154457081486663e-07,
      "loss": 0.0011,
      "step": 27260
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.13251151144504547,
      "learning_rate": 1.5060186631631112e-07,
      "loss": 0.0014,
      "step": 27265
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.09971214830875397,
      "learning_rate": 1.4966208083984302e-07,
      "loss": 0.0012,
      "step": 27270
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.08008839935064316,
      "learning_rate": 1.4872521466398392e-07,
      "loss": 0.0011,
      "step": 27275
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.07394137978553772,
      "learning_rate": 1.4779126806638954e-07,
      "loss": 0.0009,
      "step": 27280
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.03994069993495941,
      "learning_rate": 1.4686024132384736e-07,
      "loss": 0.0012,
      "step": 27285
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.13227759301662445,
      "learning_rate": 1.4593213471228218e-07,
      "loss": 0.0014,
      "step": 27290
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.08159665763378143,
      "learning_rate": 1.450069485067518e-07,
      "loss": 0.0009,
      "step": 27295
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.12798109650611877,
      "learning_rate": 1.4408468298145238e-07,
      "loss": 0.0014,
      "step": 27300
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.09440769255161285,
      "learning_rate": 1.431653384097087e-07,
      "loss": 0.0014,
      "step": 27305
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.12320124357938766,
      "learning_rate": 1.4224891506398386e-07,
      "loss": 0.0013,
      "step": 27310
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.19550946354866028,
      "learning_rate": 1.4133541321587509e-07,
      "loss": 0.0022,
      "step": 27315
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.05926552414894104,
      "learning_rate": 1.4042483313611133e-07,
      "loss": 0.0011,
      "step": 27320
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.08274723589420319,
      "learning_rate": 1.3951717509455788e-07,
      "loss": 0.0007,
      "step": 27325
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.022885218262672424,
      "learning_rate": 1.3861243936021286e-07,
      "loss": 0.001,
      "step": 27330
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.01137836929410696,
      "learning_rate": 1.3771062620120955e-07,
      "loss": 0.0014,
      "step": 27335
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.12840622663497925,
      "learning_rate": 1.3681173588481423e-07,
      "loss": 0.0014,
      "step": 27340
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.047768354415893555,
      "learning_rate": 1.3591576867742594e-07,
      "loss": 0.0013,
      "step": 27345
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.024363325908780098,
      "learning_rate": 1.3502272484458013e-07,
      "loss": 0.001,
      "step": 27350
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.029072320088744164,
      "learning_rate": 1.341326046509428e-07,
      "loss": 0.0013,
      "step": 27355
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.0950700044631958,
      "learning_rate": 1.332454083603152e-07,
      "loss": 0.0013,
      "step": 27360
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.06451021879911423,
      "learning_rate": 1.323611362356314e-07,
      "loss": 0.0012,
      "step": 27365
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.1161188930273056,
      "learning_rate": 1.314797885389607e-07,
      "loss": 0.002,
      "step": 27370
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.0990685299038887,
      "learning_rate": 1.3060136553150194e-07,
      "loss": 0.0009,
      "step": 27375
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.07793032377958298,
      "learning_rate": 1.297258674735913e-07,
      "loss": 0.0009,
      "step": 27380
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.060773540288209915,
      "learning_rate": 1.2885329462469564e-07,
      "loss": 0.0017,
      "step": 27385
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.07595169544219971,
      "learning_rate": 1.2798364724341483e-07,
      "loss": 0.0012,
      "step": 27390
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.08787562698125839,
      "learning_rate": 1.2711692558748268e-07,
      "loss": 0.0013,
      "step": 27395
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.05152950808405876,
      "learning_rate": 1.2625312991376702e-07,
      "loss": 0.0016,
      "step": 27400
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.04772697389125824,
      "learning_rate": 1.2539226047826424e-07,
      "loss": 0.0008,
      "step": 27405
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.16844956576824188,
      "learning_rate": 1.2453431753610913e-07,
      "loss": 0.0009,
      "step": 27410
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.06177696958184242,
      "learning_rate": 1.236793013415649e-07,
      "loss": 0.0008,
      "step": 27415
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.021597489714622498,
      "learning_rate": 1.2282721214803007e-07,
      "loss": 0.001,
      "step": 27420
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.11656033992767334,
      "learning_rate": 1.2197805020803365e-07,
      "loss": 0.0011,
      "step": 27425
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.06125600263476372,
      "learning_rate": 1.2113181577323662e-07,
      "loss": 0.0015,
      "step": 27430
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.11162044107913971,
      "learning_rate": 1.2028850909443612e-07,
      "loss": 0.0016,
      "step": 27435
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.10665519535541534,
      "learning_rate": 1.194481304215589e-07,
      "loss": 0.0012,
      "step": 27440
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.09857645630836487,
      "learning_rate": 1.1861068000366238e-07,
      "loss": 0.0008,
      "step": 27445
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.11146674305200577,
      "learning_rate": 1.1777615808894027e-07,
      "loss": 0.0011,
      "step": 27450
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.10069559514522552,
      "learning_rate": 1.1694456492471362e-07,
      "loss": 0.0009,
      "step": 27455
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.07771911472082138,
      "learning_rate": 1.1611590075744083e-07,
      "loss": 0.0012,
      "step": 27460
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.13914179801940918,
      "learning_rate": 1.1529016583270657e-07,
      "loss": 0.0012,
      "step": 27465
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.11388552188873291,
      "learning_rate": 1.1446736039523176e-07,
      "loss": 0.0014,
      "step": 27470
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.038466423749923706,
      "learning_rate": 1.1364748468886688e-07,
      "loss": 0.0008,
      "step": 27475
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.06610891222953796,
      "learning_rate": 1.1283053895659645e-07,
      "loss": 0.0006,
      "step": 27480
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.02082733064889908,
      "learning_rate": 1.1201652344053349e-07,
      "loss": 0.0007,
      "step": 27485
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.12689946591854095,
      "learning_rate": 1.1120543838192388e-07,
      "loss": 0.0016,
      "step": 27490
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.07294479012489319,
      "learning_rate": 1.1039728402114757e-07,
      "loss": 0.0013,
      "step": 27495
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.06044188141822815,
      "learning_rate": 1.0959206059771077e-07,
      "loss": 0.0007,
      "step": 27500
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.0841093435883522,
      "learning_rate": 1.0878976835025478e-07,
      "loss": 0.0008,
      "step": 27505
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.09894666820764542,
      "learning_rate": 1.0799040751655166e-07,
      "loss": 0.0014,
      "step": 27510
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.1542588472366333,
      "learning_rate": 1.071939783335041e-07,
      "loss": 0.0012,
      "step": 27515
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.044521354138851166,
      "learning_rate": 1.0640048103714773e-07,
      "loss": 0.001,
      "step": 27520
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.050902362912893295,
      "learning_rate": 1.0560991586264669e-07,
      "loss": 0.0013,
      "step": 27525
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.060398731380701065,
      "learning_rate": 1.0482228304429686e-07,
      "loss": 0.0012,
      "step": 27530
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.08845556527376175,
      "learning_rate": 1.0403758281552601e-07,
      "loss": 0.0011,
      "step": 27535
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.06878877431154251,
      "learning_rate": 1.0325581540889251e-07,
      "loss": 0.0009,
      "step": 27540
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.05974522605538368,
      "learning_rate": 1.0247698105608439e-07,
      "loss": 0.0008,
      "step": 27545
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.07035703212022781,
      "learning_rate": 1.0170107998792145e-07,
      "loss": 0.0013,
      "step": 27550
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.04951699450612068,
      "learning_rate": 1.0092811243435529e-07,
      "loss": 0.001,
      "step": 27555
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.07079406082630157,
      "learning_rate": 1.0015807862446381e-07,
      "loss": 0.0008,
      "step": 27560
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.046903353184461594,
      "learning_rate": 9.93909787864622e-08,
      "loss": 0.0009,
      "step": 27565
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.036676373332738876,
      "learning_rate": 9.862681314769196e-08,
      "loss": 0.0014,
      "step": 27570
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.0619211420416832,
      "learning_rate": 9.786558193462304e-08,
      "loss": 0.0011,
      "step": 27575
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.061969440430402756,
      "learning_rate": 9.710728537285829e-08,
      "loss": 0.0013,
      "step": 27580
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.08273687213659286,
      "learning_rate": 9.635192368713354e-08,
      "loss": 0.001,
      "step": 27585
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.13152478635311127,
      "learning_rate": 9.559949710130855e-08,
      "loss": 0.001,
      "step": 27590
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.08596333116292953,
      "learning_rate": 9.485000583837944e-08,
      "loss": 0.001,
      "step": 27595
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.07673116028308868,
      "learning_rate": 9.41034501204674e-08,
      "loss": 0.0015,
      "step": 27600
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.14104962348937988,
      "learning_rate": 9.335983016882766e-08,
      "loss": 0.0014,
      "step": 27605
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.10371097177267075,
      "learning_rate": 9.261914620384171e-08,
      "loss": 0.0013,
      "step": 27610
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.015523822978138924,
      "learning_rate": 9.188139844502508e-08,
      "loss": 0.0009,
      "step": 27615
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.08790897578001022,
      "learning_rate": 9.114658711101842e-08,
      "loss": 0.0015,
      "step": 27620
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.05163196474313736,
      "learning_rate": 9.041471241959531e-08,
      "loss": 0.0008,
      "step": 27625
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.08871548622846603,
      "learning_rate": 8.968577458765893e-08,
      "loss": 0.001,
      "step": 27630
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.049854159355163574,
      "learning_rate": 8.895977383123977e-08,
      "loss": 0.0012,
      "step": 27635
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.06681330502033234,
      "learning_rate": 8.82367103655013e-08,
      "loss": 0.0007,
      "step": 27640
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.1309957057237625,
      "learning_rate": 8.75165844047332e-08,
      "loss": 0.001,
      "step": 27645
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.11755676567554474,
      "learning_rate": 8.6799396162357e-08,
      "loss": 0.0008,
      "step": 27650
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.08385858684778214,
      "learning_rate": 8.608514585092153e-08,
      "loss": 0.0013,
      "step": 27655
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.05883089825510979,
      "learning_rate": 8.537383368210639e-08,
      "loss": 0.0011,
      "step": 27660
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.16626568138599396,
      "learning_rate": 8.46654598667207e-08,
      "loss": 0.0014,
      "step": 27665
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.1791432499885559,
      "learning_rate": 8.396002461470098e-08,
      "loss": 0.0014,
      "step": 27670
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.037083640694618225,
      "learning_rate": 8.325752813511445e-08,
      "loss": 0.0017,
      "step": 27675
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.05854767933487892,
      "learning_rate": 8.255797063615567e-08,
      "loss": 0.0012,
      "step": 27680
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.020023584365844727,
      "learning_rate": 8.186135232515213e-08,
      "loss": 0.0007,
      "step": 27685
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.06901459395885468,
      "learning_rate": 8.116767340855425e-08,
      "loss": 0.001,
      "step": 27690
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.10090130567550659,
      "learning_rate": 8.047693409194757e-08,
      "loss": 0.0009,
      "step": 27695
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.016971636563539505,
      "learning_rate": 7.978913458004167e-08,
      "loss": 0.0009,
      "step": 27700
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.0443294420838356,
      "learning_rate": 7.910427507667573e-08,
      "loss": 0.0007,
      "step": 27705
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.08994010090827942,
      "learning_rate": 7.842235578482182e-08,
      "loss": 0.0012,
      "step": 27710
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.11785995960235596,
      "learning_rate": 7.774337690657274e-08,
      "loss": 0.0012,
      "step": 27715
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.08951311558485031,
      "learning_rate": 7.706733864315752e-08,
      "loss": 0.0011,
      "step": 27720
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.08338700979948044,
      "learning_rate": 7.639424119493033e-08,
      "loss": 0.0013,
      "step": 27725
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.07392533868551254,
      "learning_rate": 7.572408476137383e-08,
      "loss": 0.0011,
      "step": 27730
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.09654698520898819,
      "learning_rate": 7.505686954109915e-08,
      "loss": 0.0009,
      "step": 27735
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.02619049698114395,
      "learning_rate": 7.439259573184476e-08,
      "loss": 0.0007,
      "step": 27740
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.06405049562454224,
      "learning_rate": 7.373126353048099e-08,
      "loss": 0.0013,
      "step": 27745
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.08028136193752289,
      "learning_rate": 7.307287313300216e-08,
      "loss": 0.0015,
      "step": 27750
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.048656243830919266,
      "learning_rate": 7.241742473453217e-08,
      "loss": 0.0012,
      "step": 27755
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.04180813208222389,
      "learning_rate": 7.176491852932455e-08,
      "loss": 0.0008,
      "step": 27760
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.05267529562115669,
      "learning_rate": 7.111535471075792e-08,
      "loss": 0.0015,
      "step": 27765
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.08153411746025085,
      "learning_rate": 7.046873347134275e-08,
      "loss": 0.0009,
      "step": 27770
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.05059076100587845,
      "learning_rate": 6.982505500271464e-08,
      "loss": 0.0012,
      "step": 27775
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.0719943568110466,
      "learning_rate": 6.918431949563654e-08,
      "loss": 0.0008,
      "step": 27780
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.045789770781993866,
      "learning_rate": 6.85465271400021e-08,
      "loss": 0.0013,
      "step": 27785
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.05941304191946983,
      "learning_rate": 6.791167812483013e-08,
      "loss": 0.0007,
      "step": 27790
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.10284554958343506,
      "learning_rate": 6.72797726382668e-08,
      "loss": 0.0006,
      "step": 27795
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.07532849162817001,
      "learning_rate": 6.665081086758896e-08,
      "loss": 0.0009,
      "step": 27800
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.04503500089049339,
      "learning_rate": 6.602479299919861e-08,
      "loss": 0.0008,
      "step": 27805
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.06832282245159149,
      "learning_rate": 6.540171921862515e-08,
      "loss": 0.0008,
      "step": 27810
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.050845157355070114,
      "learning_rate": 6.47815897105264e-08,
      "loss": 0.0008,
      "step": 27815
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.08400414139032364,
      "learning_rate": 6.416440465868979e-08,
      "loss": 0.0012,
      "step": 27820
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.0506439134478569,
      "learning_rate": 6.355016424602346e-08,
      "loss": 0.0009,
      "step": 27825
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.13978984951972961,
      "learning_rate": 6.293886865457067e-08,
      "loss": 0.0012,
      "step": 27830
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.05353309586644173,
      "learning_rate": 6.233051806549651e-08,
      "loss": 0.001,
      "step": 27835
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.057319726794958115,
      "learning_rate": 6.172511265909565e-08,
      "loss": 0.0013,
      "step": 27840
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.09258401393890381,
      "learning_rate": 6.1122652614789e-08,
      "loss": 0.0011,
      "step": 27845
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.07895422726869583,
      "learning_rate": 6.052313811112486e-08,
      "loss": 0.0013,
      "step": 27850
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.2097400575876236,
      "learning_rate": 5.992656932578e-08,
      "loss": 0.0016,
      "step": 27855
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.06631385535001755,
      "learning_rate": 5.9332946435556314e-08,
      "loss": 0.0013,
      "step": 27860
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.11734557151794434,
      "learning_rate": 5.8742269616381965e-08,
      "loss": 0.0011,
      "step": 27865
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.037774424999952316,
      "learning_rate": 5.815453904331359e-08,
      "loss": 0.0008,
      "step": 27870
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.07216952741146088,
      "learning_rate": 5.7569754890535176e-08,
      "loss": 0.001,
      "step": 27875
    },
    {
      "epoch": 8.7,
      "grad_norm": 0.10858375579118729,
      "learning_rate": 5.698791733135589e-08,
      "loss": 0.0012,
      "step": 27880
    },
    {
      "epoch": 8.7,
      "grad_norm": 0.011037354357540607,
      "learning_rate": 5.640902653821334e-08,
      "loss": 0.0013,
      "step": 27885
    },
    {
      "epoch": 8.7,
      "grad_norm": 0.11795692145824432,
      "learning_rate": 5.583308268266918e-08,
      "loss": 0.0011,
      "step": 27890
    },
    {
      "epoch": 8.7,
      "grad_norm": 0.06190943345427513,
      "learning_rate": 5.526008593541465e-08,
      "loss": 0.0011,
      "step": 27895
    },
    {
      "epoch": 8.7,
      "grad_norm": 0.023181727156043053,
      "learning_rate": 5.469003646626503e-08,
      "loss": 0.0012,
      "step": 27900
    },
    {
      "epoch": 8.7,
      "grad_norm": 0.08947614580392838,
      "learning_rate": 5.412293444416405e-08,
      "loss": 0.0008,
      "step": 27905
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.17211729288101196,
      "learning_rate": 5.3558780037181735e-08,
      "loss": 0.0017,
      "step": 27910
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.0541967898607254,
      "learning_rate": 5.299757341251321e-08,
      "loss": 0.001,
      "step": 27915
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.05166836455464363,
      "learning_rate": 5.2439314736481004e-08,
      "loss": 0.001,
      "step": 27920
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.06777120381593704,
      "learning_rate": 5.1884004174533876e-08,
      "loss": 0.0014,
      "step": 27925
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.0960666760802269,
      "learning_rate": 5.1331641891246844e-08,
      "loss": 0.0011,
      "step": 27930
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.048764344304800034,
      "learning_rate": 5.078222805032007e-08,
      "loss": 0.0013,
      "step": 27935
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.07357207685709,
      "learning_rate": 5.02357628145822e-08,
      "loss": 0.0008,
      "step": 27940
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.05783311277627945,
      "learning_rate": 4.9692246345985905e-08,
      "loss": 0.0007,
      "step": 27945
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.07628779858350754,
      "learning_rate": 4.915167880561122e-08,
      "loss": 0.0017,
      "step": 27950
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.042343996465206146,
      "learning_rate": 4.861406035366334e-08,
      "loss": 0.0007,
      "step": 27955
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.10046465694904327,
      "learning_rate": 4.8079391149473686e-08,
      "loss": 0.001,
      "step": 27960
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.06461944431066513,
      "learning_rate": 4.754767135149996e-08,
      "loss": 0.0011,
      "step": 27965
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.05853043869137764,
      "learning_rate": 4.70189011173261e-08,
      "loss": 0.001,
      "step": 27970
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.08283140510320663,
      "learning_rate": 4.649308060366009e-08,
      "loss": 0.0012,
      "step": 27975
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.12241838127374649,
      "learning_rate": 4.597020996633839e-08,
      "loss": 0.0012,
      "step": 27980
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.057213619351387024,
      "learning_rate": 4.545028936032147e-08,
      "loss": 0.0009,
      "step": 27985
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.06766070425510406,
      "learning_rate": 4.4933318939694995e-08,
      "loss": 0.0006,
      "step": 27990
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.05963273346424103,
      "learning_rate": 4.441929885767304e-08,
      "loss": 0.0013,
      "step": 27995
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.11648065596818924,
      "learning_rate": 4.390822926659155e-08,
      "loss": 0.0017,
      "step": 28000
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.0579272024333477,
      "learning_rate": 4.340011031791491e-08,
      "loss": 0.0007,
      "step": 28005
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.08932799845933914,
      "learning_rate": 4.2894942162231555e-08,
      "loss": 0.0011,
      "step": 28010
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.08602336794137955,
      "learning_rate": 4.239272494925617e-08,
      "loss": 0.0011,
      "step": 28015
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.11151181161403656,
      "learning_rate": 4.189345882782858e-08,
      "loss": 0.0007,
      "step": 28020
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.06075332313776016,
      "learning_rate": 4.1397143945913766e-08,
      "loss": 0.0011,
      "step": 28025
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.0862547978758812,
      "learning_rate": 4.090378045060406e-08,
      "loss": 0.0012,
      "step": 28030
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.07240119576454163,
      "learning_rate": 4.04133684881125e-08,
      "loss": 0.0014,
      "step": 28035
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.06106055900454521,
      "learning_rate": 3.992590820378284e-08,
      "loss": 0.0011,
      "step": 28040
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.12147120386362076,
      "learning_rate": 3.944139974208061e-08,
      "loss": 0.0007,
      "step": 28045
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.17372600734233856,
      "learning_rate": 3.8959843246596516e-08,
      "loss": 0.0013,
      "step": 28050
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.050082284957170486,
      "learning_rate": 3.848123886004751e-08,
      "loss": 0.0009,
      "step": 28055
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.06022435426712036,
      "learning_rate": 3.800558672427679e-08,
      "loss": 0.001,
      "step": 28060
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.054569125175476074,
      "learning_rate": 3.7532886980250484e-08,
      "loss": 0.0007,
      "step": 28065
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.09262100607156754,
      "learning_rate": 3.7063139768060976e-08,
      "loss": 0.0014,
      "step": 28070
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.09075400233268738,
      "learning_rate": 3.659634522692357e-08,
      "loss": 0.0008,
      "step": 28075
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.09641103446483612,
      "learning_rate": 3.6132503495182046e-08,
      "loss": 0.0018,
      "step": 28080
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.038265615701675415,
      "learning_rate": 3.5671614710302e-08,
      "loss": 0.0013,
      "step": 28085
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.07987763732671738,
      "learning_rate": 3.521367900887418e-08,
      "loss": 0.0014,
      "step": 28090
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.13263726234436035,
      "learning_rate": 3.4758696526617783e-08,
      "loss": 0.0012,
      "step": 28095
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.10438277572393417,
      "learning_rate": 3.430666739837052e-08,
      "loss": 0.001,
      "step": 28100
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.05294902250170708,
      "learning_rate": 3.385759175809966e-08,
      "loss": 0.0012,
      "step": 28105
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.15452563762664795,
      "learning_rate": 3.341146973889764e-08,
      "loss": 0.0013,
      "step": 28110
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.09602925926446915,
      "learning_rate": 3.296830147297647e-08,
      "loss": 0.0013,
      "step": 28115
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.06384595483541489,
      "learning_rate": 3.252808709167776e-08,
      "loss": 0.0009,
      "step": 28120
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.035283368080854416,
      "learning_rate": 3.209082672546604e-08,
      "loss": 0.0017,
      "step": 28125
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.1074545606970787,
      "learning_rate": 3.1656520503928755e-08,
      "loss": 0.0009,
      "step": 28130
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.04797278344631195,
      "learning_rate": 3.1225168555780725e-08,
      "loss": 0.0011,
      "step": 28135
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.07991745322942734,
      "learning_rate": 3.0796771008858586e-08,
      "loss": 0.0014,
      "step": 28140
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.05039117485284805,
      "learning_rate": 3.037132799012632e-08,
      "loss": 0.0009,
      "step": 28145
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.10603125393390656,
      "learning_rate": 2.9948839625669745e-08,
      "loss": 0.0006,
      "step": 28150
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.08775552362203598,
      "learning_rate": 2.9529306040698703e-08,
      "loss": 0.0011,
      "step": 28155
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.0653933510184288,
      "learning_rate": 2.9112727359550398e-08,
      "loss": 0.001,
      "step": 28160
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.1549730896949768,
      "learning_rate": 2.869910370568274e-08,
      "loss": 0.0009,
      "step": 28165
    },
    {
      "epoch": 8.79,
      "grad_norm": 0.04075611010193825,
      "learning_rate": 2.82884352016799e-08,
      "loss": 0.001,
      "step": 28170
    },
    {
      "epoch": 8.79,
      "grad_norm": 0.14746034145355225,
      "learning_rate": 2.788072196925118e-08,
      "loss": 0.0011,
      "step": 28175
    },
    {
      "epoch": 8.79,
      "grad_norm": 0.03397578001022339,
      "learning_rate": 2.7475964129227706e-08,
      "loss": 0.0007,
      "step": 28180
    },
    {
      "epoch": 8.79,
      "grad_norm": 0.03829203546047211,
      "learning_rate": 2.7074161801564635e-08,
      "loss": 0.0009,
      "step": 28185
    },
    {
      "epoch": 8.79,
      "grad_norm": 0.11898770928382874,
      "learning_rate": 2.6675315105344492e-08,
      "loss": 0.001,
      "step": 28190
    },
    {
      "epoch": 8.79,
      "grad_norm": 0.09407939016819,
      "learning_rate": 2.6279424158770495e-08,
      "loss": 0.0012,
      "step": 28195
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.09977072477340698,
      "learning_rate": 2.5886489079171016e-08,
      "loss": 0.0009,
      "step": 28200
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.05728138983249664,
      "learning_rate": 2.5496509982998463e-08,
      "loss": 0.0012,
      "step": 28205
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.17846325039863586,
      "learning_rate": 2.510948698583038e-08,
      "loss": 0.0011,
      "step": 28210
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.20595858991146088,
      "learning_rate": 2.4725420202365015e-08,
      "loss": 0.0011,
      "step": 28215
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.04241137206554413,
      "learning_rate": 2.4344309746427984e-08,
      "loss": 0.0012,
      "step": 28220
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.06812385469675064,
      "learning_rate": 2.396615573096783e-08,
      "loss": 0.0018,
      "step": 28225
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.05665101483464241,
      "learning_rate": 2.35909582680538e-08,
      "loss": 0.0008,
      "step": 28230
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.0710788443684578,
      "learning_rate": 2.321871746888471e-08,
      "loss": 0.001,
      "step": 28235
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.03182736039161682,
      "learning_rate": 2.2849433443776768e-08,
      "loss": 0.001,
      "step": 28240
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.04013293981552124,
      "learning_rate": 2.2483106302175762e-08,
      "loss": 0.0015,
      "step": 28245
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.22291040420532227,
      "learning_rate": 2.2119736152647064e-08,
      "loss": 0.0009,
      "step": 28250
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.07432419061660767,
      "learning_rate": 2.17593231028812e-08,
      "loss": 0.0014,
      "step": 28255
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.11221190541982651,
      "learning_rate": 2.1401867259691622e-08,
      "loss": 0.0015,
      "step": 28260
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.10827355086803436,
      "learning_rate": 2.1047368729018025e-08,
      "loss": 0.0011,
      "step": 28265
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.09072833508253098,
      "learning_rate": 2.0695827615918595e-08,
      "loss": 0.0006,
      "step": 28270
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.07544603198766708,
      "learning_rate": 2.034724402457999e-08,
      "loss": 0.0007,
      "step": 28275
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.025641784071922302,
      "learning_rate": 2.000161805830958e-08,
      "loss": 0.0009,
      "step": 28280
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.21241925656795502,
      "learning_rate": 1.9658949819539864e-08,
      "loss": 0.0014,
      "step": 28285
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.10191631317138672,
      "learning_rate": 1.9319239409825165e-08,
      "loss": 0.0017,
      "step": 28290
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.0852564126253128,
      "learning_rate": 1.8982486929843835e-08,
      "loss": 0.0015,
      "step": 28295
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.01896956004202366,
      "learning_rate": 1.8648692479398266e-08,
      "loss": 0.0011,
      "step": 28300
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.046149298548698425,
      "learning_rate": 1.8317856157412662e-08,
      "loss": 0.0007,
      "step": 28305
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.07430528849363327,
      "learning_rate": 1.7989978061936363e-08,
      "loss": 0.0011,
      "step": 28310
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.13385100662708282,
      "learning_rate": 1.7665058290140537e-08,
      "loss": 0.0011,
      "step": 28315
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.0955640971660614,
      "learning_rate": 1.734309693832037e-08,
      "loss": 0.0013,
      "step": 28320
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.13716626167297363,
      "learning_rate": 1.7024094101895093e-08,
      "loss": 0.0008,
      "step": 28325
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.06619204580783844,
      "learning_rate": 1.670804987540575e-08,
      "loss": 0.0012,
      "step": 28330
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.07766672223806381,
      "learning_rate": 1.6394964352515198e-08,
      "loss": 0.0007,
      "step": 28335
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.08800486475229263,
      "learning_rate": 1.6084837626013648e-08,
      "loss": 0.0012,
      "step": 28340
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.023620784282684326,
      "learning_rate": 1.5777669787810923e-08,
      "loss": 0.001,
      "step": 28345
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.16947531700134277,
      "learning_rate": 1.5473460928939753e-08,
      "loss": 0.0011,
      "step": 28350
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.08615032583475113,
      "learning_rate": 1.5172211139559133e-08,
      "loss": 0.001,
      "step": 28355
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.05784229934215546,
      "learning_rate": 1.4873920508948758e-08,
      "loss": 0.0007,
      "step": 28360
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.11840718239545822,
      "learning_rate": 1.4578589125511245e-08,
      "loss": 0.0019,
      "step": 28365
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.052449408918619156,
      "learning_rate": 1.4286217076772136e-08,
      "loss": 0.0014,
      "step": 28370
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.029246846213936806,
      "learning_rate": 1.3996804449381007e-08,
      "loss": 0.0005,
      "step": 28375
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.041718367487192154,
      "learning_rate": 1.3710351329109251e-08,
      "loss": 0.0011,
      "step": 28380
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.07578592747449875,
      "learning_rate": 1.3426857800853398e-08,
      "loss": 0.001,
      "step": 28385
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.11366725713014603,
      "learning_rate": 1.314632394862847e-08,
      "loss": 0.0013,
      "step": 28390
    },
    {
      "epoch": 8.86,
      "grad_norm": 0.06276487559080124,
      "learning_rate": 1.2868749855577955e-08,
      "loss": 0.0015,
      "step": 28395
    },
    {
      "epoch": 8.86,
      "grad_norm": 0.06069565564393997,
      "learning_rate": 1.259413560396272e-08,
      "loss": 0.0014,
      "step": 28400
    },
    {
      "epoch": 8.86,
      "grad_norm": 0.3192262351512909,
      "learning_rate": 1.2322481275169885e-08,
      "loss": 0.0015,
      "step": 28405
    },
    {
      "epoch": 8.86,
      "grad_norm": 0.10366208851337433,
      "learning_rate": 1.205378694970949e-08,
      "loss": 0.0007,
      "step": 28410
    },
    {
      "epoch": 8.86,
      "grad_norm": 0.027929887175559998,
      "learning_rate": 1.1788052707212283e-08,
      "loss": 0.0011,
      "step": 28415
    },
    {
      "epoch": 8.86,
      "grad_norm": 0.16714242100715637,
      "learning_rate": 1.1525278626431935e-08,
      "loss": 0.0008,
      "step": 28420
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.08561515808105469,
      "learning_rate": 1.126546478524726e-08,
      "loss": 0.0013,
      "step": 28425
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.08352944254875183,
      "learning_rate": 1.1008611260656666e-08,
      "loss": 0.0013,
      "step": 28430
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.18621785938739777,
      "learning_rate": 1.0754718128783704e-08,
      "loss": 0.0015,
      "step": 28435
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.08013457804918289,
      "learning_rate": 1.050378546487374e-08,
      "loss": 0.0008,
      "step": 28440
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.0238870307803154,
      "learning_rate": 1.0255813343292843e-08,
      "loss": 0.0009,
      "step": 28445
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.04448197782039642,
      "learning_rate": 1.0010801837533334e-08,
      "loss": 0.0011,
      "step": 28450
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.11206652969121933,
      "learning_rate": 9.768751020208245e-09,
      "loss": 0.0011,
      "step": 28455
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.1502913534641266,
      "learning_rate": 9.529660963051302e-09,
      "loss": 0.0013,
      "step": 28460
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.10655543953180313,
      "learning_rate": 9.29353173692138e-09,
      "loss": 0.0015,
      "step": 28465
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.061149418354034424,
      "learning_rate": 9.060363411799167e-09,
      "loss": 0.0013,
      "step": 28470
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.03373811021447182,
      "learning_rate": 8.830156056788276e-09,
      "loss": 0.0005,
      "step": 28475
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.07910972088575363,
      "learning_rate": 8.602909740114129e-09,
      "loss": 0.0014,
      "step": 28480
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.10172421485185623,
      "learning_rate": 8.378624529123968e-09,
      "loss": 0.0009,
      "step": 28485
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.1482880711555481,
      "learning_rate": 8.157300490287956e-09,
      "loss": 0.0018,
      "step": 28490
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.047668956220149994,
      "learning_rate": 7.938937689199178e-09,
      "loss": 0.0013,
      "step": 28495
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.10796379297971725,
      "learning_rate": 7.723536190573645e-09,
      "loss": 0.001,
      "step": 28500
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.085243821144104,
      "learning_rate": 7.511096058249178e-09,
      "loss": 0.0008,
      "step": 28505
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.04042333364486694,
      "learning_rate": 7.301617355184309e-09,
      "loss": 0.0013,
      "step": 28510
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.07351209968328476,
      "learning_rate": 7.095100143461598e-09,
      "loss": 0.0006,
      "step": 28515
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.09849730879068375,
      "learning_rate": 6.891544484286527e-09,
      "loss": 0.0009,
      "step": 28520
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.20490317046642303,
      "learning_rate": 6.6909504379852885e-09,
      "loss": 0.0016,
      "step": 28525
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.02441880851984024,
      "learning_rate": 6.493318064006993e-09,
      "loss": 0.0006,
      "step": 28530
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.07980853319168091,
      "learning_rate": 6.337344608285945e-09,
      "loss": 0.0011,
      "step": 28535
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.10102265328168869,
      "learning_rate": 6.1450433915000165e-09,
      "loss": 0.0008,
      "step": 28540
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.5179263949394226,
      "learning_rate": 5.955704008824281e-09,
      "loss": 0.0014,
      "step": 28545
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.11356153339147568,
      "learning_rate": 5.769326516373852e-09,
      "loss": 0.0011,
      "step": 28550
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.01838083378970623,
      "learning_rate": 5.585910969384545e-09,
      "loss": 0.001,
      "step": 28555
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.0684712678194046,
      "learning_rate": 5.405457422212879e-09,
      "loss": 0.0015,
      "step": 28560
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.10153453797101974,
      "learning_rate": 5.227965928341627e-09,
      "loss": 0.0017,
      "step": 28565
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.09030450135469437,
      "learning_rate": 5.053436540370938e-09,
      "loss": 0.0013,
      "step": 28570
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.04882200434803963,
      "learning_rate": 4.881869310024989e-09,
      "loss": 0.0007,
      "step": 28575
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.11226236820220947,
      "learning_rate": 4.713264288153107e-09,
      "loss": 0.0014,
      "step": 28580
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.22697344422340393,
      "learning_rate": 4.5476215247219905e-09,
      "loss": 0.001,
      "step": 28585
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.0568106546998024,
      "learning_rate": 4.384941068822368e-09,
      "loss": 0.0013,
      "step": 28590
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.12141884118318558,
      "learning_rate": 4.225222968667897e-09,
      "loss": 0.001,
      "step": 28595
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.11962190270423889,
      "learning_rate": 4.068467271592935e-09,
      "loss": 0.0014,
      "step": 28600
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.07136228680610657,
      "learning_rate": 3.9146740240547655e-09,
      "loss": 0.0012,
      "step": 28605
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.04858747869729996,
      "learning_rate": 3.763843271631373e-09,
      "loss": 0.0012,
      "step": 28610
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.0264420323073864,
      "learning_rate": 3.615975059025889e-09,
      "loss": 0.001,
      "step": 28615
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.061483949422836304,
      "learning_rate": 3.471069430061036e-09,
      "loss": 0.0017,
      "step": 28620
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.08160971850156784,
      "learning_rate": 3.3291264276791302e-09,
      "loss": 0.0011,
      "step": 28625
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.04612981900572777,
      "learning_rate": 3.190146093950963e-09,
      "loss": 0.0011,
      "step": 28630
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.012312815524637699,
      "learning_rate": 3.0541284700624783e-09,
      "loss": 0.0011,
      "step": 28635
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.03374966233968735,
      "learning_rate": 2.921073596325874e-09,
      "loss": 0.001,
      "step": 28640
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.15238814055919647,
      "learning_rate": 2.7909815121740512e-09,
      "loss": 0.0009,
      "step": 28645
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.11198405921459198,
      "learning_rate": 2.6638522561617254e-09,
      "loss": 0.001,
      "step": 28650
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.10427577048540115,
      "learning_rate": 2.5396858659665345e-09,
      "loss": 0.0006,
      "step": 28655
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.07450409978628159,
      "learning_rate": 2.418482378384601e-09,
      "loss": 0.001,
      "step": 28660
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.5595073103904724,
      "learning_rate": 2.3002418293394115e-09,
      "loss": 0.0016,
      "step": 28665
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.03920646756887436,
      "learning_rate": 2.184964253871824e-09,
      "loss": 0.0013,
      "step": 28670
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.0426039844751358,
      "learning_rate": 2.0726496861456224e-09,
      "loss": 0.0009,
      "step": 28675
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.050191134214401245,
      "learning_rate": 1.963298159447513e-09,
      "loss": 0.0006,
      "step": 28680
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.0812898799777031,
      "learning_rate": 1.8569097061871266e-09,
      "loss": 0.001,
      "step": 28685
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.04886721447110176,
      "learning_rate": 1.7534843578914662e-09,
      "loss": 0.001,
      "step": 28690
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.08173787593841553,
      "learning_rate": 1.6530221452148998e-09,
      "loss": 0.001,
      "step": 28695
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.04918965697288513,
      "learning_rate": 1.555523097929168e-09,
      "loss": 0.0008,
      "step": 28700
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.0397033616900444,
      "learning_rate": 1.4609872449300455e-09,
      "loss": 0.001,
      "step": 28705
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.08010255545377731,
      "learning_rate": 1.3694146142351205e-09,
      "loss": 0.0012,
      "step": 28710
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.047962285578250885,
      "learning_rate": 1.2808052329826848e-09,
      "loss": 0.0007,
      "step": 28715
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.03405216336250305,
      "learning_rate": 1.1951591274339535e-09,
      "loss": 0.0009,
      "step": 28720
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.21903178095817566,
      "learning_rate": 1.112476322971956e-09,
      "loss": 0.001,
      "step": 28725
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.030231649056077003,
      "learning_rate": 1.0327568441004243e-09,
      "loss": 0.0007,
      "step": 28730
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.02870813198387623,
      "learning_rate": 9.560007144449045e-10,
      "loss": 0.0011,
      "step": 28735
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.05557563900947571,
      "learning_rate": 8.822079567538666e-10,
      "loss": 0.001,
      "step": 28740
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.06193048134446144,
      "learning_rate": 8.113785928975937e-10,
      "loss": 0.0012,
      "step": 28745
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.09998735785484314,
      "learning_rate": 7.435126438670726e-10,
      "loss": 0.0011,
      "step": 28750
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.0637703463435173,
      "learning_rate": 6.786101297751035e-10,
      "loss": 0.002,
      "step": 28755
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.09788890182971954,
      "learning_rate": 6.166710698563005e-10,
      "loss": 0.001,
      "step": 28760
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.043414145708084106,
      "learning_rate": 5.576954824682012e-10,
      "loss": 0.0008,
      "step": 28765
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.07151433825492859,
      "learning_rate": 5.016833850879366e-10,
      "loss": 0.0012,
      "step": 28770
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.04484909400343895,
      "learning_rate": 4.4863479431667136e-10,
      "loss": 0.0004,
      "step": 28775
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.08398403972387314,
      "learning_rate": 3.9854972587627383e-10,
      "loss": 0.0011,
      "step": 28780
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.09407573193311691,
      "learning_rate": 3.514281946082054e-10,
      "loss": 0.0007,
      "step": 28785
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.16476760804653168,
      "learning_rate": 3.07270214480182e-10,
      "loss": 0.0013,
      "step": 28790
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.04877808690071106,
      "learning_rate": 2.6607579857840236e-10,
      "loss": 0.001,
      "step": 28795
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.04684771969914436,
      "learning_rate": 2.2784495910976866e-10,
      "loss": 0.0017,
      "step": 28800
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.06821387261152267,
      "learning_rate": 1.9257770740743753e-10,
      "loss": 0.0012,
      "step": 28805
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.134215846657753,
      "learning_rate": 1.6027405392082807e-10,
      "loss": 0.0009,
      "step": 28810
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.06247800588607788,
      "learning_rate": 1.3093400822450364e-10,
      "loss": 0.0012,
      "step": 28815
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.099249467253685,
      "learning_rate": 1.045575790148412e-10,
      "loss": 0.001,
      "step": 28820
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.045093026012182236,
      "learning_rate": 8.114477410781085e-11,
      "loss": 0.0016,
      "step": 28825
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.0406327061355114,
      "learning_rate": 6.069560044341671e-11,
      "loss": 0.001,
      "step": 28830
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.04194846376776695,
      "learning_rate": 4.321006408014583e-11,
      "loss": 0.0006,
      "step": 28835
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.09402500838041306,
      "learning_rate": 2.8688170200519282e-11,
      "loss": 0.0013,
      "step": 28840
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.02617935836315155,
      "learning_rate": 1.712992310998196e-11,
      "loss": 0.0009,
      "step": 28845
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.06014137715101242,
      "learning_rate": 8.535326232461671e-12,
      "loss": 0.0011,
      "step": 28850
    },
    {
      "epoch": 9.0,
      "step": 28854,
      "total_flos": 5.673215034528458e+19,
      "train_loss": 0.032082520662370544,
      "train_runtime": 172613.4265,
      "train_samples_per_second": 5.349,
      "train_steps_per_second": 0.167
    }
  ],
  "logging_steps": 5,
  "max_steps": 28854,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 13000,
  "total_flos": 5.673215034528458e+19,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}