{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2408,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004152823920265781,
      "grad_norm": 25.422981813437236,
      "learning_rate": 4.1493775933609963e-08,
      "loss": 1.3975,
      "step": 1
    },
    {
      "epoch": 0.0020764119601328905,
      "grad_norm": 23.65282908395334,
      "learning_rate": 2.074688796680498e-07,
      "loss": 1.4281,
      "step": 5
    },
    {
      "epoch": 0.004152823920265781,
      "grad_norm": 16.38973942245371,
      "learning_rate": 4.149377593360996e-07,
      "loss": 1.3933,
      "step": 10
    },
    {
      "epoch": 0.006229235880398671,
      "grad_norm": 8.620332321861904,
      "learning_rate": 6.224066390041494e-07,
      "loss": 1.2986,
      "step": 15
    },
    {
      "epoch": 0.008305647840531562,
      "grad_norm": 10.289897317705874,
      "learning_rate": 8.298755186721992e-07,
      "loss": 1.1565,
      "step": 20
    },
    {
      "epoch": 0.010382059800664452,
      "grad_norm": 4.429779856244459,
      "learning_rate": 1.037344398340249e-06,
      "loss": 1.051,
      "step": 25
    },
    {
      "epoch": 0.012458471760797342,
      "grad_norm": 3.3098208738585213,
      "learning_rate": 1.2448132780082988e-06,
      "loss": 0.9902,
      "step": 30
    },
    {
      "epoch": 0.014534883720930232,
      "grad_norm": 3.4349888460346687,
      "learning_rate": 1.4522821576763488e-06,
      "loss": 0.9652,
      "step": 35
    },
    {
      "epoch": 0.016611295681063124,
      "grad_norm": 3.1515624301454133,
      "learning_rate": 1.6597510373443984e-06,
      "loss": 0.9415,
      "step": 40
    },
    {
      "epoch": 0.018687707641196014,
      "grad_norm": 3.1235312209606505,
      "learning_rate": 1.8672199170124482e-06,
      "loss": 0.93,
      "step": 45
    },
    {
      "epoch": 0.020764119601328904,
      "grad_norm": 3.1741829648141926,
      "learning_rate": 2.074688796680498e-06,
      "loss": 0.9238,
      "step": 50
    },
    {
      "epoch": 0.022840531561461794,
      "grad_norm": 3.232116295196654,
      "learning_rate": 2.282157676348548e-06,
      "loss": 0.9123,
      "step": 55
    },
    {
      "epoch": 0.024916943521594685,
      "grad_norm": 3.1515595029223396,
      "learning_rate": 2.4896265560165977e-06,
      "loss": 0.9031,
      "step": 60
    },
    {
      "epoch": 0.026993355481727575,
      "grad_norm": 3.1003061370301617,
      "learning_rate": 2.6970954356846475e-06,
      "loss": 0.8947,
      "step": 65
    },
    {
      "epoch": 0.029069767441860465,
      "grad_norm": 2.9767060692194844,
      "learning_rate": 2.9045643153526977e-06,
      "loss": 0.8919,
      "step": 70
    },
    {
      "epoch": 0.031146179401993355,
      "grad_norm": 3.0759553041103205,
      "learning_rate": 3.112033195020747e-06,
      "loss": 0.8702,
      "step": 75
    },
    {
      "epoch": 0.03322259136212625,
      "grad_norm": 3.285827319776166,
      "learning_rate": 3.319502074688797e-06,
      "loss": 0.8727,
      "step": 80
    },
    {
      "epoch": 0.03529900332225914,
      "grad_norm": 3.3462993523967186,
      "learning_rate": 3.526970954356847e-06,
      "loss": 0.8736,
      "step": 85
    },
    {
      "epoch": 0.03737541528239203,
      "grad_norm": 3.256004424550593,
      "learning_rate": 3.7344398340248965e-06,
      "loss": 0.8858,
      "step": 90
    },
    {
      "epoch": 0.03945182724252492,
      "grad_norm": 3.159488005717498,
      "learning_rate": 3.941908713692946e-06,
      "loss": 0.8559,
      "step": 95
    },
    {
      "epoch": 0.04152823920265781,
      "grad_norm": 2.9634363451500114,
      "learning_rate": 4.149377593360996e-06,
      "loss": 0.8586,
      "step": 100
    },
    {
      "epoch": 0.0436046511627907,
      "grad_norm": 3.159728031235876,
      "learning_rate": 4.356846473029046e-06,
      "loss": 0.8674,
      "step": 105
    },
    {
      "epoch": 0.04568106312292359,
      "grad_norm": 2.970452415217835,
      "learning_rate": 4.564315352697096e-06,
      "loss": 0.8542,
      "step": 110
    },
    {
      "epoch": 0.04775747508305648,
      "grad_norm": 3.1788047641427513,
      "learning_rate": 4.771784232365146e-06,
      "loss": 0.8701,
      "step": 115
    },
    {
      "epoch": 0.04983388704318937,
      "grad_norm": 3.451301944053267,
      "learning_rate": 4.979253112033195e-06,
      "loss": 0.8714,
      "step": 120
    },
    {
      "epoch": 0.05191029900332226,
      "grad_norm": 3.388326009403783,
      "learning_rate": 5.1867219917012455e-06,
      "loss": 0.8483,
      "step": 125
    },
    {
      "epoch": 0.05398671096345515,
      "grad_norm": 2.9959540593135645,
      "learning_rate": 5.394190871369295e-06,
      "loss": 0.8481,
      "step": 130
    },
    {
      "epoch": 0.05606312292358804,
      "grad_norm": 3.1213953476841856,
      "learning_rate": 5.601659751037345e-06,
      "loss": 0.8387,
      "step": 135
    },
    {
      "epoch": 0.05813953488372093,
      "grad_norm": 3.403963416369247,
      "learning_rate": 5.809128630705395e-06,
      "loss": 0.8399,
      "step": 140
    },
    {
      "epoch": 0.06021594684385382,
      "grad_norm": 3.0893053330914695,
      "learning_rate": 6.016597510373444e-06,
      "loss": 0.8386,
      "step": 145
    },
    {
      "epoch": 0.06229235880398671,
      "grad_norm": 3.166492177328262,
      "learning_rate": 6.224066390041494e-06,
      "loss": 0.8457,
      "step": 150
    },
    {
      "epoch": 0.0643687707641196,
      "grad_norm": 3.2851556967703117,
      "learning_rate": 6.431535269709544e-06,
      "loss": 0.8421,
      "step": 155
    },
    {
      "epoch": 0.0664451827242525,
      "grad_norm": 2.9899317822541454,
      "learning_rate": 6.639004149377594e-06,
      "loss": 0.8373,
      "step": 160
    },
    {
      "epoch": 0.06852159468438539,
      "grad_norm": 3.0509892785590456,
      "learning_rate": 6.846473029045644e-06,
      "loss": 0.8334,
      "step": 165
    },
    {
      "epoch": 0.07059800664451828,
      "grad_norm": 3.00742757115455,
      "learning_rate": 7.053941908713694e-06,
      "loss": 0.8233,
      "step": 170
    },
    {
      "epoch": 0.07267441860465117,
      "grad_norm": 3.0518393701751485,
      "learning_rate": 7.261410788381743e-06,
      "loss": 0.8296,
      "step": 175
    },
    {
      "epoch": 0.07475083056478406,
      "grad_norm": 3.1984146233667263,
      "learning_rate": 7.468879668049793e-06,
      "loss": 0.8155,
      "step": 180
    },
    {
      "epoch": 0.07682724252491695,
      "grad_norm": 3.05629449726749,
      "learning_rate": 7.676348547717844e-06,
      "loss": 0.8377,
      "step": 185
    },
    {
      "epoch": 0.07890365448504984,
      "grad_norm": 3.249088059891964,
      "learning_rate": 7.883817427385892e-06,
      "loss": 0.8432,
      "step": 190
    },
    {
      "epoch": 0.08098006644518273,
      "grad_norm": 3.0028481508425515,
      "learning_rate": 8.091286307053943e-06,
      "loss": 0.8173,
      "step": 195
    },
    {
      "epoch": 0.08305647840531562,
      "grad_norm": 3.059733445916786,
      "learning_rate": 8.298755186721992e-06,
      "loss": 0.8227,
      "step": 200
    },
    {
      "epoch": 0.08513289036544851,
      "grad_norm": 3.0867633236533365,
      "learning_rate": 8.506224066390042e-06,
      "loss": 0.8181,
      "step": 205
    },
    {
      "epoch": 0.0872093023255814,
      "grad_norm": 2.997953986592159,
      "learning_rate": 8.713692946058093e-06,
      "loss": 0.821,
      "step": 210
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 3.2351659520743072,
      "learning_rate": 8.921161825726142e-06,
      "loss": 0.8294,
      "step": 215
    },
    {
      "epoch": 0.09136212624584718,
      "grad_norm": 3.1494481731597586,
      "learning_rate": 9.128630705394191e-06,
      "loss": 0.8261,
      "step": 220
    },
    {
      "epoch": 0.09343853820598007,
      "grad_norm": 3.105511823234228,
      "learning_rate": 9.33609958506224e-06,
      "loss": 0.8165,
      "step": 225
    },
    {
      "epoch": 0.09551495016611296,
      "grad_norm": 3.023901781664328,
      "learning_rate": 9.543568464730292e-06,
      "loss": 0.8123,
      "step": 230
    },
    {
      "epoch": 0.09759136212624585,
      "grad_norm": 3.4303556589177187,
      "learning_rate": 9.751037344398341e-06,
      "loss": 0.8093,
      "step": 235
    },
    {
      "epoch": 0.09966777408637874,
      "grad_norm": 3.6054989714255408,
      "learning_rate": 9.95850622406639e-06,
      "loss": 0.8201,
      "step": 240
    },
    {
      "epoch": 0.10174418604651163,
      "grad_norm": 2.990225009601177,
      "learning_rate": 9.999915930067828e-06,
      "loss": 0.8208,
      "step": 245
    },
    {
      "epoch": 0.10382059800664452,
      "grad_norm": 2.9957103647324264,
      "learning_rate": 9.999574400813641e-06,
      "loss": 0.816,
      "step": 250
    },
    {
      "epoch": 0.10589700996677741,
      "grad_norm": 2.8988415018010287,
      "learning_rate": 9.998970175798065e-06,
      "loss": 0.8044,
      "step": 255
    },
    {
      "epoch": 0.1079734219269103,
      "grad_norm": 2.893907971746992,
      "learning_rate": 9.998103286769267e-06,
      "loss": 0.799,
      "step": 260
    },
    {
      "epoch": 0.11004983388704319,
      "grad_norm": 2.898946354458808,
      "learning_rate": 9.996973779276743e-06,
      "loss": 0.8113,
      "step": 265
    },
    {
      "epoch": 0.11212624584717608,
      "grad_norm": 3.0485697591450998,
      "learning_rate": 9.99558171266891e-06,
      "loss": 0.8194,
      "step": 270
    },
    {
      "epoch": 0.11420265780730897,
      "grad_norm": 2.933613250090363,
      "learning_rate": 9.993927160089991e-06,
      "loss": 0.7981,
      "step": 275
    },
    {
      "epoch": 0.11627906976744186,
      "grad_norm": 2.900283777987733,
      "learning_rate": 9.992010208476178e-06,
      "loss": 0.8114,
      "step": 280
    },
    {
      "epoch": 0.11835548172757475,
      "grad_norm": 2.869639926652705,
      "learning_rate": 9.989830958551058e-06,
      "loss": 0.8026,
      "step": 285
    },
    {
      "epoch": 0.12043189368770764,
      "grad_norm": 3.0764284732072236,
      "learning_rate": 9.98738952482032e-06,
      "loss": 0.7816,
      "step": 290
    },
    {
      "epoch": 0.12250830564784053,
      "grad_norm": 2.872848930860205,
      "learning_rate": 9.984686035565742e-06,
      "loss": 0.7851,
      "step": 295
    },
    {
      "epoch": 0.12458471760797342,
      "grad_norm": 2.7170384439590367,
      "learning_rate": 9.98172063283845e-06,
      "loss": 0.8054,
      "step": 300
    },
    {
      "epoch": 0.12666112956810632,
      "grad_norm": 2.785739578421159,
      "learning_rate": 9.978493472451451e-06,
      "loss": 0.7824,
      "step": 305
    },
    {
      "epoch": 0.1287375415282392,
      "grad_norm": 2.955753943035507,
      "learning_rate": 9.975004723971452e-06,
      "loss": 0.7788,
      "step": 310
    },
    {
      "epoch": 0.1308139534883721,
      "grad_norm": 2.7566534229071378,
      "learning_rate": 9.971254570709939e-06,
      "loss": 0.7804,
      "step": 315
    },
    {
      "epoch": 0.132890365448505,
      "grad_norm": 3.0399050026271945,
      "learning_rate": 9.967243209713563e-06,
      "loss": 0.7712,
      "step": 320
    },
    {
      "epoch": 0.13496677740863788,
      "grad_norm": 3.227011718605211,
      "learning_rate": 9.962970851753767e-06,
      "loss": 0.7852,
      "step": 325
    },
    {
      "epoch": 0.13704318936877077,
      "grad_norm": 2.894940556652265,
      "learning_rate": 9.95843772131573e-06,
      "loss": 0.767,
      "step": 330
    },
    {
      "epoch": 0.13911960132890366,
      "grad_norm": 3.137972193410393,
      "learning_rate": 9.95364405658655e-06,
      "loss": 0.77,
      "step": 335
    },
    {
      "epoch": 0.14119601328903655,
      "grad_norm": 2.7913612546678426,
      "learning_rate": 9.948590109442755e-06,
      "loss": 0.7768,
      "step": 340
    },
    {
      "epoch": 0.14327242524916944,
      "grad_norm": 2.893979747266515,
      "learning_rate": 9.94327614543704e-06,
      "loss": 0.7827,
      "step": 345
    },
    {
      "epoch": 0.14534883720930233,
      "grad_norm": 2.665071280290936,
      "learning_rate": 9.937702443784343e-06,
      "loss": 0.7474,
      "step": 350
    },
    {
      "epoch": 0.14742524916943522,
      "grad_norm": 2.741350083908129,
      "learning_rate": 9.931869297347146e-06,
      "loss": 0.7638,
      "step": 355
    },
    {
      "epoch": 0.14950166112956811,
      "grad_norm": 2.9878149207237357,
      "learning_rate": 9.925777012620111e-06,
      "loss": 0.7419,
      "step": 360
    },
    {
      "epoch": 0.151578073089701,
      "grad_norm": 2.801227928713699,
      "learning_rate": 9.919425909713958e-06,
      "loss": 0.769,
      "step": 365
    },
    {
      "epoch": 0.1536544850498339,
      "grad_norm": 3.023770968839729,
      "learning_rate": 9.912816322338659e-06,
      "loss": 0.7447,
      "step": 370
    },
    {
      "epoch": 0.15573089700996678,
      "grad_norm": 2.9927287523796715,
      "learning_rate": 9.905948597785888e-06,
      "loss": 0.754,
      "step": 375
    },
    {
      "epoch": 0.15780730897009967,
      "grad_norm": 11.785492453222856,
      "learning_rate": 9.89882309691079e-06,
      "loss": 0.7497,
      "step": 380
    },
    {
      "epoch": 0.15988372093023256,
      "grad_norm": 2.8752234411604682,
      "learning_rate": 9.891440194113008e-06,
      "loss": 0.7427,
      "step": 385
    },
    {
      "epoch": 0.16196013289036545,
      "grad_norm": 3.097207390376622,
      "learning_rate": 9.88380027731702e-06,
      "loss": 0.7542,
      "step": 390
    },
    {
      "epoch": 0.16403654485049834,
      "grad_norm": 2.921991118334764,
      "learning_rate": 9.875903747951742e-06,
      "loss": 0.7621,
      "step": 395
    },
    {
      "epoch": 0.16611295681063123,
      "grad_norm": 2.8395297947865963,
      "learning_rate": 9.867751020929454e-06,
      "loss": 0.735,
      "step": 400
    },
    {
      "epoch": 0.16818936877076412,
      "grad_norm": 2.726116425089643,
      "learning_rate": 9.859342524623985e-06,
      "loss": 0.7124,
      "step": 405
    },
    {
      "epoch": 0.17026578073089702,
      "grad_norm": 3.2173444091652943,
      "learning_rate": 9.850678700848208e-06,
      "loss": 0.7374,
      "step": 410
    },
    {
      "epoch": 0.1723421926910299,
      "grad_norm": 2.716930762983964,
      "learning_rate": 9.84176000483083e-06,
      "loss": 0.7138,
      "step": 415
    },
    {
      "epoch": 0.1744186046511628,
      "grad_norm": 2.985441779621083,
      "learning_rate": 9.832586905192469e-06,
      "loss": 0.731,
      "step": 420
    },
    {
      "epoch": 0.17649501661129569,
      "grad_norm": 3.032790315651323,
      "learning_rate": 9.823159883921028e-06,
      "loss": 0.7215,
      "step": 425
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 2.6988344818168155,
      "learning_rate": 9.813479436346378e-06,
      "loss": 0.7183,
      "step": 430
    },
    {
      "epoch": 0.18064784053156147,
      "grad_norm": 2.973146607192177,
      "learning_rate": 9.803546071114323e-06,
      "loss": 0.7311,
      "step": 435
    },
    {
      "epoch": 0.18272425249169436,
      "grad_norm": 2.9093506646801344,
      "learning_rate": 9.793360310159878e-06,
      "loss": 0.7049,
      "step": 440
    },
    {
      "epoch": 0.18480066445182725,
      "grad_norm": 3.01100096145872,
      "learning_rate": 9.782922688679847e-06,
      "loss": 0.7118,
      "step": 445
    },
    {
      "epoch": 0.18687707641196014,
      "grad_norm": 2.716470652939527,
      "learning_rate": 9.772233755104695e-06,
      "loss": 0.7277,
      "step": 450
    },
    {
      "epoch": 0.18895348837209303,
      "grad_norm": 2.7134248053870165,
      "learning_rate": 9.761294071069736e-06,
      "loss": 0.7205,
      "step": 455
    },
    {
      "epoch": 0.19102990033222592,
      "grad_norm": 2.6251507638777163,
      "learning_rate": 9.750104211385625e-06,
      "loss": 0.7152,
      "step": 460
    },
    {
      "epoch": 0.1931063122923588,
      "grad_norm": 2.8023948010803483,
      "learning_rate": 9.738664764008149e-06,
      "loss": 0.7233,
      "step": 465
    },
    {
      "epoch": 0.1951827242524917,
      "grad_norm": 3.714290449563204,
      "learning_rate": 9.726976330007341e-06,
      "loss": 0.6998,
      "step": 470
    },
    {
      "epoch": 0.1972591362126246,
      "grad_norm": 2.8670419197216512,
      "learning_rate": 9.71503952353589e-06,
      "loss": 0.6985,
      "step": 475
    },
    {
      "epoch": 0.19933554817275748,
      "grad_norm": 3.1683988394439107,
      "learning_rate": 9.702854971796876e-06,
      "loss": 0.7089,
      "step": 480
    },
    {
      "epoch": 0.20141196013289037,
      "grad_norm": 3.2223078839261166,
      "learning_rate": 9.690423315010814e-06,
      "loss": 0.7053,
      "step": 485
    },
    {
      "epoch": 0.20348837209302326,
      "grad_norm": 2.77875488832717,
      "learning_rate": 9.677745206382014e-06,
      "loss": 0.7271,
      "step": 490
    },
    {
      "epoch": 0.20556478405315615,
      "grad_norm": 2.888271933836237,
      "learning_rate": 9.664821312064258e-06,
      "loss": 0.7018,
      "step": 495
    },
    {
      "epoch": 0.20764119601328904,
      "grad_norm": 3.2746008040723815,
      "learning_rate": 9.651652311125803e-06,
      "loss": 0.6991,
      "step": 500
    },
    {
      "epoch": 0.20971760797342193,
      "grad_norm": 2.76622547311742,
      "learning_rate": 9.638238895513687e-06,
      "loss": 0.7075,
      "step": 505
    },
    {
      "epoch": 0.21179401993355482,
      "grad_norm": 2.9972446036957114,
      "learning_rate": 9.624581770017392e-06,
      "loss": 0.6857,
      "step": 510
    },
    {
      "epoch": 0.2138704318936877,
      "grad_norm": 2.869516499460042,
      "learning_rate": 9.610681652231794e-06,
      "loss": 0.6916,
      "step": 515
    },
    {
      "epoch": 0.2159468438538206,
      "grad_norm": 2.742923434452921,
      "learning_rate": 9.596539272519468e-06,
      "loss": 0.6811,
      "step": 520
    },
    {
      "epoch": 0.2180232558139535,
      "grad_norm": 2.8482023108565677,
      "learning_rate": 9.582155373972303e-06,
      "loss": 0.6744,
      "step": 525
    },
    {
      "epoch": 0.22009966777408638,
      "grad_norm": 2.9348099403663124,
      "learning_rate": 9.56753071237247e-06,
      "loss": 0.6776,
      "step": 530
    },
    {
      "epoch": 0.22217607973421927,
      "grad_norm": 2.786772996017183,
      "learning_rate": 9.552666056152704e-06,
      "loss": 0.6798,
      "step": 535
    },
    {
      "epoch": 0.22425249169435216,
      "grad_norm": 2.92722689041533,
      "learning_rate": 9.537562186355918e-06,
      "loss": 0.6843,
      "step": 540
    },
    {
      "epoch": 0.22632890365448505,
      "grad_norm": 2.7694998172195207,
      "learning_rate": 9.52221989659418e-06,
      "loss": 0.6938,
      "step": 545
    },
    {
      "epoch": 0.22840531561461794,
      "grad_norm": 2.9300442858036244,
      "learning_rate": 9.506639993007012e-06,
      "loss": 0.6944,
      "step": 550
    },
    {
      "epoch": 0.23048172757475083,
      "grad_norm": 3.1035204783454993,
      "learning_rate": 9.490823294219015e-06,
      "loss": 0.672,
      "step": 555
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 2.6193387690961245,
      "learning_rate": 9.474770631296882e-06,
      "loss": 0.6561,
      "step": 560
    },
    {
      "epoch": 0.2346345514950166,
      "grad_norm": 2.61646550507026,
      "learning_rate": 9.458482847705705e-06,
      "loss": 0.6576,
      "step": 565
    },
    {
      "epoch": 0.2367109634551495,
      "grad_norm": 2.756473668019519,
      "learning_rate": 9.441960799264678e-06,
      "loss": 0.6851,
      "step": 570
    },
    {
      "epoch": 0.2387873754152824,
      "grad_norm": 2.6995089678231614,
      "learning_rate": 9.425205354102111e-06,
      "loss": 0.6648,
      "step": 575
    },
    {
      "epoch": 0.24086378737541528,
      "grad_norm": 2.7140254791209677,
      "learning_rate": 9.408217392609831e-06,
      "loss": 0.6451,
      "step": 580
    },
    {
      "epoch": 0.24294019933554817,
      "grad_norm": 2.607599787114018,
      "learning_rate": 9.390997807396912e-06,
      "loss": 0.67,
      "step": 585
    },
    {
      "epoch": 0.24501661129568106,
      "grad_norm": 2.8420050898692764,
      "learning_rate": 9.373547503242775e-06,
      "loss": 0.6657,
      "step": 590
    },
    {
      "epoch": 0.24709302325581395,
      "grad_norm": 2.9228965685399095,
      "learning_rate": 9.355867397049658e-06,
      "loss": 0.6566,
      "step": 595
    },
    {
      "epoch": 0.24916943521594684,
      "grad_norm": 2.8048600929777403,
      "learning_rate": 9.337958417794425e-06,
      "loss": 0.6457,
      "step": 600
    },
    {
      "epoch": 0.25124584717607973,
      "grad_norm": 2.6983485281997415,
      "learning_rate": 9.319821506479762e-06,
      "loss": 0.6376,
      "step": 605
    },
    {
      "epoch": 0.25332225913621265,
      "grad_norm": 2.801805288954333,
      "learning_rate": 9.301457616084733e-06,
      "loss": 0.6523,
      "step": 610
    },
    {
      "epoch": 0.2553986710963455,
      "grad_norm": 2.820864396273499,
      "learning_rate": 9.282867711514703e-06,
      "loss": 0.6365,
      "step": 615
    },
    {
      "epoch": 0.2574750830564784,
      "grad_norm": 2.9932167823643043,
      "learning_rate": 9.264052769550643e-06,
      "loss": 0.6425,
      "step": 620
    },
    {
      "epoch": 0.2595514950166113,
      "grad_norm": 2.6556108045628544,
      "learning_rate": 9.245013778797802e-06,
      "loss": 0.6562,
      "step": 625
    },
    {
      "epoch": 0.2616279069767442,
      "grad_norm": 2.676416816690246,
      "learning_rate": 9.225751739633772e-06,
      "loss": 0.6387,
      "step": 630
    },
    {
      "epoch": 0.26370431893687707,
      "grad_norm": 2.702226526508375,
      "learning_rate": 9.206267664155906e-06,
      "loss": 0.6348,
      "step": 635
    },
    {
      "epoch": 0.26578073089701,
      "grad_norm": 2.637563222880754,
      "learning_rate": 9.186562576128159e-06,
      "loss": 0.6263,
      "step": 640
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 2.7815352111724603,
      "learning_rate": 9.16663751092728e-06,
      "loss": 0.6362,
      "step": 645
    },
    {
      "epoch": 0.26993355481727577,
      "grad_norm": 2.8822755136904528,
      "learning_rate": 9.146493515488418e-06,
      "loss": 0.6164,
      "step": 650
    },
    {
      "epoch": 0.27200996677740863,
      "grad_norm": 2.5755107274498146,
      "learning_rate": 9.126131648250112e-06,
      "loss": 0.6342,
      "step": 655
    },
    {
      "epoch": 0.27408637873754155,
      "grad_norm": 2.584492766117294,
      "learning_rate": 9.105552979098675e-06,
      "loss": 0.6329,
      "step": 660
    },
    {
      "epoch": 0.2761627906976744,
      "grad_norm": 2.6805388863449036,
      "learning_rate": 9.084758589311977e-06,
      "loss": 0.6307,
      "step": 665
    },
    {
      "epoch": 0.2782392026578073,
      "grad_norm": 2.7584115266730693,
      "learning_rate": 9.063749571502633e-06,
      "loss": 0.6374,
      "step": 670
    },
    {
      "epoch": 0.2803156146179402,
      "grad_norm": 2.8092430217085145,
      "learning_rate": 9.04252702956059e-06,
      "loss": 0.6282,
      "step": 675
    },
    {
      "epoch": 0.2823920265780731,
      "grad_norm": 2.6353604501522168,
      "learning_rate": 9.021092078595132e-06,
      "loss": 0.6332,
      "step": 680
    },
    {
      "epoch": 0.28446843853820597,
      "grad_norm": 2.7859177417571486,
      "learning_rate": 8.999445844876276e-06,
      "loss": 0.6381,
      "step": 685
    },
    {
      "epoch": 0.2865448504983389,
      "grad_norm": 2.6603634875986457,
      "learning_rate": 8.977589465775607e-06,
      "loss": 0.6312,
      "step": 690
    },
    {
      "epoch": 0.28862126245847175,
      "grad_norm": 2.6293766795824354,
      "learning_rate": 8.955524089706506e-06,
      "loss": 0.5999,
      "step": 695
    },
    {
      "epoch": 0.29069767441860467,
      "grad_norm": 2.8986723382239967,
      "learning_rate": 8.933250876063815e-06,
      "loss": 0.6297,
      "step": 700
    },
    {
      "epoch": 0.29277408637873753,
      "grad_norm": 2.6589365161649834,
      "learning_rate": 8.910770995162913e-06,
      "loss": 0.6303,
      "step": 705
    },
    {
      "epoch": 0.29485049833887045,
      "grad_norm": 2.64992234535583,
      "learning_rate": 8.88808562817823e-06,
      "loss": 0.6114,
      "step": 710
    },
    {
      "epoch": 0.2969269102990033,
      "grad_norm": 2.7322760412568776,
      "learning_rate": 8.865195967081174e-06,
      "loss": 0.6215,
      "step": 715
    },
    {
      "epoch": 0.29900332225913623,
      "grad_norm": 2.576473302210113,
      "learning_rate": 8.842103214577511e-06,
      "loss": 0.6147,
      "step": 720
    },
    {
      "epoch": 0.3010797342192691,
      "grad_norm": 2.507546434543662,
      "learning_rate": 8.818808584044163e-06,
      "loss": 0.6089,
      "step": 725
    },
    {
      "epoch": 0.303156146179402,
      "grad_norm": 2.953501799132662,
      "learning_rate": 8.795313299465455e-06,
      "loss": 0.6147,
      "step": 730
    },
    {
      "epoch": 0.30523255813953487,
      "grad_norm": 2.58266860044093,
      "learning_rate": 8.771618595368806e-06,
      "loss": 0.6024,
      "step": 735
    },
    {
      "epoch": 0.3073089700996678,
      "grad_norm": 2.7291039422306613,
      "learning_rate": 8.747725716759859e-06,
      "loss": 0.6152,
      "step": 740
    },
    {
      "epoch": 0.30938538205980065,
      "grad_norm": 2.696653736904745,
      "learning_rate": 8.723635919057058e-06,
      "loss": 0.6082,
      "step": 745
    },
    {
      "epoch": 0.31146179401993357,
      "grad_norm": 2.639188973608746,
      "learning_rate": 8.699350468025699e-06,
      "loss": 0.5924,
      "step": 750
    },
    {
      "epoch": 0.31353820598006643,
      "grad_norm": 2.5960120065556294,
      "learning_rate": 8.674870639711403e-06,
      "loss": 0.5871,
      "step": 755
    },
    {
      "epoch": 0.31561461794019935,
      "grad_norm": 2.691098687645451,
      "learning_rate": 8.650197720373091e-06,
      "loss": 0.5937,
      "step": 760
    },
    {
      "epoch": 0.3176910299003322,
      "grad_norm": 2.7922815680081947,
      "learning_rate": 8.625333006415372e-06,
      "loss": 0.5806,
      "step": 765
    },
    {
      "epoch": 0.31976744186046513,
      "grad_norm": 2.5989983221444635,
      "learning_rate": 8.600277804320452e-06,
      "loss": 0.5889,
      "step": 770
    },
    {
      "epoch": 0.321843853820598,
      "grad_norm": 2.7500580415708553,
      "learning_rate": 8.575033430579465e-06,
      "loss": 0.5929,
      "step": 775
    },
    {
      "epoch": 0.3239202657807309,
      "grad_norm": 2.9863748696055485,
      "learning_rate": 8.549601211623316e-06,
      "loss": 0.5905,
      "step": 780
    },
    {
      "epoch": 0.32599667774086377,
      "grad_norm": 2.7128601524461966,
      "learning_rate": 8.523982483752973e-06,
      "loss": 0.5838,
      "step": 785
    },
    {
      "epoch": 0.3280730897009967,
      "grad_norm": 2.6273588590853727,
      "learning_rate": 8.498178593069262e-06,
      "loss": 0.579,
      "step": 790
    },
    {
      "epoch": 0.33014950166112955,
      "grad_norm": 2.6424251208940714,
      "learning_rate": 8.472190895402131e-06,
      "loss": 0.568,
      "step": 795
    },
    {
      "epoch": 0.33222591362126247,
      "grad_norm": 2.774060760650428,
      "learning_rate": 8.446020756239418e-06,
      "loss": 0.5881,
      "step": 800
    },
    {
      "epoch": 0.33430232558139533,
      "grad_norm": 2.7429673227633193,
      "learning_rate": 8.419669550655093e-06,
      "loss": 0.5807,
      "step": 805
    },
    {
      "epoch": 0.33637873754152825,
      "grad_norm": 2.4588138685140164,
      "learning_rate": 8.393138663237015e-06,
      "loss": 0.5699,
      "step": 810
    },
    {
      "epoch": 0.3384551495016611,
      "grad_norm": 2.8894345220890845,
      "learning_rate": 8.366429488014178e-06,
      "loss": 0.5644,
      "step": 815
    },
    {
      "epoch": 0.34053156146179403,
      "grad_norm": 2.6417969175920253,
      "learning_rate": 8.339543428383467e-06,
      "loss": 0.577,
      "step": 820
    },
    {
      "epoch": 0.3426079734219269,
      "grad_norm": 2.639049529021501,
      "learning_rate": 8.312481897035906e-06,
      "loss": 0.5835,
      "step": 825
    },
    {
      "epoch": 0.3446843853820598,
      "grad_norm": 2.791601353912272,
      "learning_rate": 8.285246315882448e-06,
      "loss": 0.5873,
      "step": 830
    },
    {
      "epoch": 0.3467607973421927,
      "grad_norm": 2.760486538247162,
      "learning_rate": 8.257838115979244e-06,
      "loss": 0.5743,
      "step": 835
    },
    {
      "epoch": 0.3488372093023256,
      "grad_norm": 2.6084506349864114,
      "learning_rate": 8.230258737452473e-06,
      "loss": 0.5835,
      "step": 840
    },
    {
      "epoch": 0.35091362126245845,
      "grad_norm": 2.568077365967415,
      "learning_rate": 8.202509629422647e-06,
      "loss": 0.5663,
      "step": 845
    },
    {
      "epoch": 0.35299003322259137,
      "grad_norm": 3.338586543406698,
      "learning_rate": 8.17459224992849e-06,
      "loss": 0.561,
      "step": 850
    },
    {
      "epoch": 0.35506644518272423,
      "grad_norm": 2.550936924190995,
      "learning_rate": 8.14650806585031e-06,
      "loss": 0.5748,
      "step": 855
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 2.730568567607308,
      "learning_rate": 8.118258552832945e-06,
      "loss": 0.5526,
      "step": 860
    },
    {
      "epoch": 0.35921926910299,
      "grad_norm": 2.7922640713365765,
      "learning_rate": 8.0898451952082e-06,
      "loss": 0.5636,
      "step": 865
    },
    {
      "epoch": 0.36129568106312293,
      "grad_norm": 2.4817520439108782,
      "learning_rate": 8.061269485916881e-06,
      "loss": 0.565,
      "step": 870
    },
    {
      "epoch": 0.3633720930232558,
      "grad_norm": 2.5897869437416814,
      "learning_rate": 8.032532926430335e-06,
      "loss": 0.5718,
      "step": 875
    },
    {
      "epoch": 0.3654485049833887,
      "grad_norm": 2.6233407361081196,
      "learning_rate": 8.003637026671558e-06,
      "loss": 0.5495,
      "step": 880
    },
    {
      "epoch": 0.3675249169435216,
      "grad_norm": 2.590608968830393,
      "learning_rate": 7.974583304935867e-06,
      "loss": 0.5701,
      "step": 885
    },
    {
      "epoch": 0.3696013289036545,
      "grad_norm": 2.676185626796156,
      "learning_rate": 7.945373287811116e-06,
      "loss": 0.5476,
      "step": 890
    },
    {
      "epoch": 0.37167774086378735,
      "grad_norm": 2.588749653152642,
      "learning_rate": 7.916008510097483e-06,
      "loss": 0.5363,
      "step": 895
    },
    {
      "epoch": 0.37375415282392027,
      "grad_norm": 2.648109565452331,
      "learning_rate": 7.88649051472683e-06,
      "loss": 0.5566,
      "step": 900
    },
    {
      "epoch": 0.37583056478405313,
      "grad_norm": 2.6123078212762567,
      "learning_rate": 7.856820852681634e-06,
      "loss": 0.5481,
      "step": 905
    },
    {
      "epoch": 0.37790697674418605,
      "grad_norm": 2.5715025577779107,
      "learning_rate": 7.82700108291348e-06,
      "loss": 0.5554,
      "step": 910
    },
    {
      "epoch": 0.3799833887043189,
      "grad_norm": 2.6810117688521333,
      "learning_rate": 7.797032772261164e-06,
      "loss": 0.5396,
      "step": 915
    },
    {
      "epoch": 0.38205980066445183,
      "grad_norm": 2.828001329589521,
      "learning_rate": 7.766917495368356e-06,
      "loss": 0.549,
      "step": 920
    },
    {
      "epoch": 0.3841362126245847,
      "grad_norm": 2.6073301891312455,
      "learning_rate": 7.736656834600866e-06,
      "loss": 0.5403,
      "step": 925
    },
    {
      "epoch": 0.3862126245847176,
      "grad_norm": 2.7467154847057107,
      "learning_rate": 7.706252379963498e-06,
      "loss": 0.5395,
      "step": 930
    },
    {
      "epoch": 0.3882890365448505,
      "grad_norm": 2.6418072073420067,
      "learning_rate": 7.675705729016508e-06,
      "loss": 0.5363,
      "step": 935
    },
    {
      "epoch": 0.3903654485049834,
      "grad_norm": 2.632007372607857,
      "learning_rate": 7.645018486791664e-06,
      "loss": 0.5377,
      "step": 940
    },
    {
      "epoch": 0.39244186046511625,
      "grad_norm": 2.4652302347093364,
      "learning_rate": 7.6141922657079045e-06,
      "loss": 0.5321,
      "step": 945
    },
    {
      "epoch": 0.3945182724252492,
      "grad_norm": 2.5492866422631764,
      "learning_rate": 7.583228685486623e-06,
      "loss": 0.5433,
      "step": 950
    },
    {
      "epoch": 0.39659468438538203,
      "grad_norm": 2.4794671881341936,
      "learning_rate": 7.552129373066565e-06,
      "loss": 0.5423,
      "step": 955
    },
    {
      "epoch": 0.39867109634551495,
      "grad_norm": 2.565377450639672,
      "learning_rate": 7.520895962518329e-06,
      "loss": 0.5357,
      "step": 960
    },
    {
      "epoch": 0.4007475083056478,
      "grad_norm": 2.7376349329000504,
      "learning_rate": 7.489530094958521e-06,
      "loss": 0.5529,
      "step": 965
    },
    {
      "epoch": 0.40282392026578073,
      "grad_norm": 2.5470062145134778,
      "learning_rate": 7.458033418463517e-06,
      "loss": 0.5167,
      "step": 970
    },
    {
      "epoch": 0.4049003322259136,
      "grad_norm": 2.5915393940286724,
      "learning_rate": 7.426407587982869e-06,
      "loss": 0.5359,
      "step": 975
    },
    {
      "epoch": 0.4069767441860465,
      "grad_norm": 2.5521473612501118,
      "learning_rate": 7.394654265252348e-06,
      "loss": 0.5448,
      "step": 980
    },
    {
      "epoch": 0.4090531561461794,
      "grad_norm": 2.540390049884069,
      "learning_rate": 7.362775118706627e-06,
      "loss": 0.5224,
      "step": 985
    },
    {
      "epoch": 0.4111295681063123,
      "grad_norm": 2.67106563437947,
      "learning_rate": 7.330771823391622e-06,
      "loss": 0.547,
      "step": 990
    },
    {
      "epoch": 0.41320598006644516,
      "grad_norm": 2.5844286453504752,
      "learning_rate": 7.298646060876473e-06,
      "loss": 0.5245,
      "step": 995
    },
    {
      "epoch": 0.4152823920265781,
      "grad_norm": 2.489462893225223,
      "learning_rate": 7.266399519165193e-06,
      "loss": 0.5177,
      "step": 1000
    },
    {
      "epoch": 0.417358803986711,
      "grad_norm": 2.548885028848683,
      "learning_rate": 7.234033892607969e-06,
      "loss": 0.5285,
      "step": 1005
    },
    {
      "epoch": 0.41943521594684385,
      "grad_norm": 2.515732979636329,
      "learning_rate": 7.201550881812138e-06,
      "loss": 0.5295,
      "step": 1010
    },
    {
      "epoch": 0.42151162790697677,
      "grad_norm": 2.580813201220608,
      "learning_rate": 7.168952193552831e-06,
      "loss": 0.5144,
      "step": 1015
    },
    {
      "epoch": 0.42358803986710963,
      "grad_norm": 2.8605769340325544,
      "learning_rate": 7.136239540683297e-06,
      "loss": 0.5189,
      "step": 1020
    },
    {
      "epoch": 0.42566445182724255,
      "grad_norm": 2.7042921962644773,
      "learning_rate": 7.103414642044888e-06,
      "loss": 0.516,
      "step": 1025
    },
    {
      "epoch": 0.4277408637873754,
      "grad_norm": 2.5935305392513475,
      "learning_rate": 7.070479222376765e-06,
      "loss": 0.5273,
      "step": 1030
    },
    {
      "epoch": 0.42981727574750833,
      "grad_norm": 2.521806447567166,
      "learning_rate": 7.037435012225259e-06,
      "loss": 0.514,
      "step": 1035
    },
    {
      "epoch": 0.4318936877076412,
      "grad_norm": 2.4922095571026808,
      "learning_rate": 7.00428374785295e-06,
      "loss": 0.5191,
      "step": 1040
    },
    {
      "epoch": 0.4339700996677741,
      "grad_norm": 2.53445755137843,
      "learning_rate": 6.971027171147436e-06,
      "loss": 0.5175,
      "step": 1045
    },
    {
      "epoch": 0.436046511627907,
      "grad_norm": 2.5854663493896815,
      "learning_rate": 6.937667029529803e-06,
      "loss": 0.5052,
      "step": 1050
    },
    {
      "epoch": 0.4381229235880399,
      "grad_norm": 2.6149256231235767,
      "learning_rate": 6.904205075862816e-06,
      "loss": 0.5155,
      "step": 1055
    },
    {
      "epoch": 0.44019933554817275,
      "grad_norm": 2.5728069972099643,
      "learning_rate": 6.870643068358813e-06,
      "loss": 0.5164,
      "step": 1060
    },
    {
      "epoch": 0.44227574750830567,
      "grad_norm": 2.610034601385569,
      "learning_rate": 6.8369827704873225e-06,
      "loss": 0.515,
      "step": 1065
    },
    {
      "epoch": 0.44435215946843853,
      "grad_norm": 2.559653943614866,
      "learning_rate": 6.803225950882407e-06,
      "loss": 0.5103,
      "step": 1070
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 2.744659999074845,
      "learning_rate": 6.769374383249728e-06,
      "loss": 0.5144,
      "step": 1075
    },
    {
      "epoch": 0.4485049833887043,
      "grad_norm": 2.500834722382555,
      "learning_rate": 6.735429846273356e-06,
      "loss": 0.509,
      "step": 1080
    },
    {
      "epoch": 0.45058139534883723,
      "grad_norm": 2.571303478772175,
      "learning_rate": 6.701394123522303e-06,
      "loss": 0.5061,
      "step": 1085
    },
    {
      "epoch": 0.4526578073089701,
      "grad_norm": 2.6726371126474042,
      "learning_rate": 6.667269003356815e-06,
      "loss": 0.4872,
      "step": 1090
    },
    {
      "epoch": 0.454734219269103,
      "grad_norm": 2.314624945694432,
      "learning_rate": 6.633056278834403e-06,
      "loss": 0.4978,
      "step": 1095
    },
    {
      "epoch": 0.4568106312292359,
      "grad_norm": 2.5660125412801986,
      "learning_rate": 6.598757747615625e-06,
      "loss": 0.4873,
      "step": 1100
    },
    {
      "epoch": 0.4588870431893688,
      "grad_norm": 2.5055302944005655,
      "learning_rate": 6.564375211869638e-06,
      "loss": 0.4955,
      "step": 1105
    },
    {
      "epoch": 0.46096345514950166,
      "grad_norm": 2.3161654964295963,
      "learning_rate": 6.529910478179499e-06,
      "loss": 0.4996,
      "step": 1110
    },
    {
      "epoch": 0.4630398671096346,
      "grad_norm": 2.713583584390501,
      "learning_rate": 6.495365357447242e-06,
      "loss": 0.4837,
      "step": 1115
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 2.6986080979156597,
      "learning_rate": 6.4607416647987285e-06,
      "loss": 0.503,
      "step": 1120
    },
    {
      "epoch": 0.46719269102990035,
      "grad_norm": 2.3758745672703614,
      "learning_rate": 6.426041219488275e-06,
      "loss": 0.4917,
      "step": 1125
    },
    {
      "epoch": 0.4692691029900332,
      "grad_norm": 2.468317610874025,
      "learning_rate": 6.39126584480306e-06,
      "loss": 0.4947,
      "step": 1130
    },
    {
      "epoch": 0.47134551495016613,
      "grad_norm": 2.672466601805675,
      "learning_rate": 6.3564173679673225e-06,
      "loss": 0.4956,
      "step": 1135
    },
    {
      "epoch": 0.473421926910299,
      "grad_norm": 2.686387722109422,
      "learning_rate": 6.321497620046353e-06,
      "loss": 0.4958,
      "step": 1140
    },
    {
      "epoch": 0.4754983388704319,
      "grad_norm": 2.4115883144762105,
      "learning_rate": 6.286508435850282e-06,
      "loss": 0.4884,
      "step": 1145
    },
    {
      "epoch": 0.4775747508305648,
      "grad_norm": 2.473062095275494,
      "learning_rate": 6.251451653837679e-06,
      "loss": 0.4873,
      "step": 1150
    },
    {
      "epoch": 0.4796511627906977,
      "grad_norm": 2.4611172122096034,
      "learning_rate": 6.216329116018943e-06,
      "loss": 0.4828,
      "step": 1155
    },
    {
      "epoch": 0.48172757475083056,
      "grad_norm": 2.438501558434762,
      "learning_rate": 6.181142667859521e-06,
      "loss": 0.4743,
      "step": 1160
    },
    {
      "epoch": 0.4838039867109635,
      "grad_norm": 2.4623748153401586,
      "learning_rate": 6.145894158182945e-06,
      "loss": 0.4813,
      "step": 1165
    },
    {
      "epoch": 0.48588039867109634,
      "grad_norm": 2.5841330806095093,
      "learning_rate": 6.11058543907368e-06,
      "loss": 0.4757,
      "step": 1170
    },
    {
      "epoch": 0.48795681063122925,
      "grad_norm": 2.420645551171905,
      "learning_rate": 6.075218365779814e-06,
      "loss": 0.4717,
      "step": 1175
    },
    {
      "epoch": 0.4900332225913621,
      "grad_norm": 2.41753538282735,
      "learning_rate": 6.039794796615575e-06,
      "loss": 0.4683,
      "step": 1180
    },
    {
      "epoch": 0.49210963455149503,
      "grad_norm": 2.6345922483315993,
      "learning_rate": 6.004316592863693e-06,
      "loss": 0.4758,
      "step": 1185
    },
    {
      "epoch": 0.4941860465116279,
      "grad_norm": 2.580357854248359,
      "learning_rate": 5.96878561867759e-06,
      "loss": 0.4923,
      "step": 1190
    },
    {
      "epoch": 0.4962624584717608,
      "grad_norm": 2.3693846881679463,
      "learning_rate": 5.9332037409834466e-06,
      "loss": 0.4732,
      "step": 1195
    },
    {
      "epoch": 0.4983388704318937,
      "grad_norm": 2.769567429139866,
      "learning_rate": 5.89757282938209e-06,
      "loss": 0.4713,
      "step": 1200
    },
    {
      "epoch": 0.5004152823920266,
      "grad_norm": 2.41622785319668,
      "learning_rate": 5.86189475605077e-06,
      "loss": 0.476,
      "step": 1205
    },
    {
      "epoch": 0.5024916943521595,
      "grad_norm": 2.499791289384567,
      "learning_rate": 5.826171395644786e-06,
      "loss": 0.4749,
      "step": 1210
    },
    {
      "epoch": 0.5045681063122923,
      "grad_norm": 2.417525944289692,
      "learning_rate": 5.790404625198982e-06,
      "loss": 0.4726,
      "step": 1215
    },
    {
      "epoch": 0.5066445182724253,
      "grad_norm": 2.5878334687029114,
      "learning_rate": 5.754596324029125e-06,
      "loss": 0.4761,
      "step": 1220
    },
    {
      "epoch": 0.5087209302325582,
      "grad_norm": 2.4962102663667043,
      "learning_rate": 5.7187483736331554e-06,
      "loss": 0.4578,
      "step": 1225
    },
    {
      "epoch": 0.510797342192691,
      "grad_norm": 2.6263564446414636,
      "learning_rate": 5.682862657592327e-06,
      "loss": 0.4825,
      "step": 1230
    },
    {
      "epoch": 0.5128737541528239,
      "grad_norm": 2.880797119411763,
      "learning_rate": 5.646941061472242e-06,
      "loss": 0.469,
      "step": 1235
    },
    {
      "epoch": 0.5149501661129569,
      "grad_norm": 2.555965100494747,
      "learning_rate": 5.610985472723764e-06,
      "loss": 0.4712,
      "step": 1240
    },
    {
      "epoch": 0.5170265780730897,
      "grad_norm": 2.502236357284136,
      "learning_rate": 5.5749977805838615e-06,
      "loss": 0.4681,
      "step": 1245
    },
    {
      "epoch": 0.5191029900332226,
      "grad_norm": 2.4360635002482347,
      "learning_rate": 5.538979875976324e-06,
      "loss": 0.4636,
      "step": 1250
    },
    {
      "epoch": 0.5211794019933554,
      "grad_norm": 2.488011716508302,
      "learning_rate": 5.502933651412417e-06,
      "loss": 0.4699,
      "step": 1255
    },
    {
      "epoch": 0.5232558139534884,
      "grad_norm": 2.3770436189443696,
      "learning_rate": 5.466861000891439e-06,
      "loss": 0.4592,
      "step": 1260
    },
    {
      "epoch": 0.5253322259136213,
      "grad_norm": 2.7541846157024876,
      "learning_rate": 5.430763819801205e-06,
      "loss": 0.4692,
      "step": 1265
    },
    {
      "epoch": 0.5274086378737541,
      "grad_norm": 2.7287082031019745,
      "learning_rate": 5.394644004818452e-06,
      "loss": 0.4745,
      "step": 1270
    },
    {
      "epoch": 0.529485049833887,
      "grad_norm": 2.5164954994115094,
      "learning_rate": 5.3585034538091885e-06,
      "loss": 0.4525,
      "step": 1275
    },
    {
      "epoch": 0.53156146179402,
      "grad_norm": 2.347205777105881,
      "learning_rate": 5.322344065728964e-06,
      "loss": 0.4689,
      "step": 1280
    },
    {
      "epoch": 0.5336378737541528,
      "grad_norm": 2.582827989286747,
      "learning_rate": 5.286167740523099e-06,
      "loss": 0.4691,
      "step": 1285
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 2.5061090934097843,
      "learning_rate": 5.249976379026851e-06,
      "loss": 0.4436,
      "step": 1290
    },
    {
      "epoch": 0.5377906976744186,
      "grad_norm": 2.4524559965169748,
      "learning_rate": 5.213771882865538e-06,
      "loss": 0.4643,
      "step": 1295
    },
    {
      "epoch": 0.5398671096345515,
      "grad_norm": 2.560097527019471,
      "learning_rate": 5.177556154354622e-06,
      "loss": 0.4464,
      "step": 1300
    },
    {
      "epoch": 0.5419435215946844,
      "grad_norm": 2.397260026201424,
      "learning_rate": 5.141331096399755e-06,
      "loss": 0.4501,
      "step": 1305
    },
    {
      "epoch": 0.5440199335548173,
      "grad_norm": 2.351541148312247,
      "learning_rate": 5.1050986123967884e-06,
      "loss": 0.4398,
      "step": 1310
    },
    {
      "epoch": 0.5460963455149501,
      "grad_norm": 2.452194040455103,
      "learning_rate": 5.068860606131766e-06,
      "loss": 0.4516,
      "step": 1315
    },
    {
      "epoch": 0.5481727574750831,
      "grad_norm": 2.593569889967618,
      "learning_rate": 5.032618981680893e-06,
      "loss": 0.4534,
      "step": 1320
    },
    {
      "epoch": 0.550249169435216,
      "grad_norm": 2.491194365967403,
      "learning_rate": 4.9963756433104875e-06,
      "loss": 0.4561,
      "step": 1325
    },
    {
      "epoch": 0.5523255813953488,
      "grad_norm": 2.5315048028501432,
      "learning_rate": 4.960132495376919e-06,
      "loss": 0.4387,
      "step": 1330
    },
    {
      "epoch": 0.5544019933554817,
      "grad_norm": 2.4221610492026566,
      "learning_rate": 4.923891442226554e-06,
      "loss": 0.4526,
      "step": 1335
    },
    {
      "epoch": 0.5564784053156147,
      "grad_norm": 2.4574741459986043,
      "learning_rate": 4.887654388095691e-06,
      "loss": 0.4388,
      "step": 1340
    },
    {
      "epoch": 0.5585548172757475,
      "grad_norm": 2.5581004359073565,
      "learning_rate": 4.851423237010504e-06,
      "loss": 0.4512,
      "step": 1345
    },
    {
      "epoch": 0.5606312292358804,
      "grad_norm": 2.5084567945271634,
      "learning_rate": 4.815199892687006e-06,
      "loss": 0.464,
      "step": 1350
    },
    {
      "epoch": 0.5627076411960132,
      "grad_norm": 2.4656070255557294,
      "learning_rate": 4.778986258431005e-06,
      "loss": 0.4471,
      "step": 1355
    },
    {
      "epoch": 0.5647840531561462,
      "grad_norm": 2.494517722129321,
      "learning_rate": 4.742784237038113e-06,
      "loss": 0.4352,
      "step": 1360
    },
    {
      "epoch": 0.5668604651162791,
      "grad_norm": 2.5383042319953995,
      "learning_rate": 4.70659573069376e-06,
      "loss": 0.421,
      "step": 1365
    },
    {
      "epoch": 0.5689368770764119,
      "grad_norm": 2.3933135171603936,
      "learning_rate": 4.670422640873242e-06,
      "loss": 0.4379,
      "step": 1370
    },
    {
      "epoch": 0.5710132890365448,
      "grad_norm": 2.4020680375977133,
      "learning_rate": 4.63426686824182e-06,
      "loss": 0.4323,
      "step": 1375
    },
    {
      "epoch": 0.5730897009966778,
      "grad_norm": 2.43581294994139,
      "learning_rate": 4.598130312554843e-06,
      "loss": 0.4397,
      "step": 1380
    },
    {
      "epoch": 0.5751661129568106,
      "grad_norm": 2.5772706634163027,
      "learning_rate": 4.562014872557936e-06,
      "loss": 0.4362,
      "step": 1385
    },
    {
      "epoch": 0.5772425249169435,
      "grad_norm": 2.448863408768738,
      "learning_rate": 4.525922445887224e-06,
      "loss": 0.4349,
      "step": 1390
    },
    {
      "epoch": 0.5793189368770764,
      "grad_norm": 2.535308434878213,
      "learning_rate": 4.489854928969635e-06,
      "loss": 0.4516,
      "step": 1395
    },
    {
      "epoch": 0.5813953488372093,
      "grad_norm": 2.3973615256768768,
      "learning_rate": 4.453814216923242e-06,
      "loss": 0.4336,
      "step": 1400
    },
    {
      "epoch": 0.5834717607973422,
      "grad_norm": 2.3119199540164965,
      "learning_rate": 4.4178022034576976e-06,
      "loss": 0.4226,
      "step": 1405
    },
    {
      "epoch": 0.5855481727574751,
      "grad_norm": 2.3014825037296633,
      "learning_rate": 4.381820780774724e-06,
      "loss": 0.4322,
      "step": 1410
    },
    {
      "epoch": 0.5876245847176079,
      "grad_norm": 2.5351337278959556,
      "learning_rate": 4.345871839468694e-06,
      "loss": 0.4055,
      "step": 1415
    },
    {
      "epoch": 0.5897009966777409,
      "grad_norm": 2.611286820208639,
      "learning_rate": 4.309957268427292e-06,
      "loss": 0.4216,
      "step": 1420
    },
    {
      "epoch": 0.5917774086378738,
      "grad_norm": 2.3889570520642684,
      "learning_rate": 4.274078954732262e-06,
      "loss": 0.4427,
      "step": 1425
    },
    {
      "epoch": 0.5938538205980066,
      "grad_norm": 2.384724624598042,
      "learning_rate": 4.2382387835602565e-06,
      "loss": 0.4246,
      "step": 1430
    },
    {
      "epoch": 0.5959302325581395,
      "grad_norm": 2.3536762842777126,
      "learning_rate": 4.20243863808378e-06,
      "loss": 0.4352,
      "step": 1435
    },
    {
      "epoch": 0.5980066445182725,
      "grad_norm": 2.367560729519929,
      "learning_rate": 4.166680399372248e-06,
      "loss": 0.4226,
      "step": 1440
    },
    {
      "epoch": 0.6000830564784053,
      "grad_norm": 2.401186140827422,
      "learning_rate": 4.130965946293135e-06,
      "loss": 0.4529,
      "step": 1445
    },
    {
      "epoch": 0.6021594684385382,
      "grad_norm": 2.3503805374006457,
      "learning_rate": 4.095297155413264e-06,
      "loss": 0.4213,
      "step": 1450
    },
    {
      "epoch": 0.604235880398671,
      "grad_norm": 2.404199762232402,
      "learning_rate": 4.059675900900199e-06,
      "loss": 0.4309,
      "step": 1455
    },
    {
      "epoch": 0.606312292358804,
      "grad_norm": 2.5304024582625053,
      "learning_rate": 4.024104054423772e-06,
      "loss": 0.4215,
      "step": 1460
    },
    {
      "epoch": 0.6083887043189369,
      "grad_norm": 2.4035116235125473,
      "learning_rate": 3.9885834850577375e-06,
      "loss": 0.4282,
      "step": 1465
    },
    {
      "epoch": 0.6104651162790697,
      "grad_norm": 2.3499844076305156,
      "learning_rate": 3.953116059181563e-06,
      "loss": 0.422,
      "step": 1470
    },
    {
      "epoch": 0.6125415282392026,
      "grad_norm": 2.5288170114153585,
      "learning_rate": 3.9177036403823645e-06,
      "loss": 0.4329,
      "step": 1475
    },
    {
      "epoch": 0.6146179401993356,
      "grad_norm": 2.3290974062316057,
      "learning_rate": 3.882348089356992e-06,
      "loss": 0.4137,
      "step": 1480
    },
    {
      "epoch": 0.6166943521594684,
      "grad_norm": 2.4328677326588894,
      "learning_rate": 3.84705126381425e-06,
      "loss": 0.4297,
      "step": 1485
    },
    {
      "epoch": 0.6187707641196013,
      "grad_norm": 2.3908310630477954,
      "learning_rate": 3.8118150183772974e-06,
      "loss": 0.4293,
      "step": 1490
    },
    {
      "epoch": 0.6208471760797342,
      "grad_norm": 2.4893827738846808,
      "learning_rate": 3.776641204486191e-06,
      "loss": 0.4214,
      "step": 1495
    },
    {
      "epoch": 0.6229235880398671,
      "grad_norm": 2.3486377563484133,
      "learning_rate": 3.7415316703006116e-06,
      "loss": 0.405,
      "step": 1500
    },
    {
      "epoch": 0.625,
      "grad_norm": 2.466506888817687,
      "learning_rate": 3.7064882606027497e-06,
      "loss": 0.426,
      "step": 1505
    },
    {
      "epoch": 0.6270764119601329,
      "grad_norm": 2.496662130115367,
      "learning_rate": 3.671512816700375e-06,
      "loss": 0.4201,
      "step": 1510
    },
    {
      "epoch": 0.6291528239202658,
      "grad_norm": 2.265163717312505,
      "learning_rate": 3.636607176330088e-06,
      "loss": 0.4205,
      "step": 1515
    },
    {
      "epoch": 0.6312292358803987,
      "grad_norm": 2.2703878574783163,
      "learning_rate": 3.60177317356076e-06,
      "loss": 0.4101,
      "step": 1520
    },
    {
      "epoch": 0.6333056478405316,
      "grad_norm": 2.423443407995488,
      "learning_rate": 3.5670126386971625e-06,
      "loss": 0.4171,
      "step": 1525
    },
    {
      "epoch": 0.6353820598006644,
      "grad_norm": 2.44608682526587,
      "learning_rate": 3.5323273981837965e-06,
      "loss": 0.416,
      "step": 1530
    },
    {
      "epoch": 0.6374584717607974,
      "grad_norm": 2.2051417207338173,
      "learning_rate": 3.497719274508925e-06,
      "loss": 0.4019,
      "step": 1535
    },
    {
      "epoch": 0.6395348837209303,
      "grad_norm": 2.4800578989548034,
      "learning_rate": 3.4631900861088132e-06,
      "loss": 0.4029,
      "step": 1540
    },
    {
      "epoch": 0.6416112956810631,
      "grad_norm": 2.3268282845100035,
      "learning_rate": 3.4287416472721795e-06,
      "loss": 0.4111,
      "step": 1545
    },
    {
      "epoch": 0.643687707641196,
      "grad_norm": 2.3872453059218532,
      "learning_rate": 3.3943757680448697e-06,
      "loss": 0.4061,
      "step": 1550
    },
    {
      "epoch": 0.645764119601329,
      "grad_norm": 2.42558490404232,
      "learning_rate": 3.360094254134746e-06,
      "loss": 0.403,
      "step": 1555
    },
    {
      "epoch": 0.6478405315614618,
      "grad_norm": 2.441847356983534,
      "learning_rate": 3.3258989068168123e-06,
      "loss": 0.417,
      "step": 1560
    },
    {
      "epoch": 0.6499169435215947,
      "grad_norm": 2.356616246546388,
      "learning_rate": 3.2917915228385676e-06,
      "loss": 0.4008,
      "step": 1565
    },
    {
      "epoch": 0.6519933554817275,
      "grad_norm": 2.457529466848808,
      "learning_rate": 3.257773894325599e-06,
      "loss": 0.4166,
      "step": 1570
    },
    {
      "epoch": 0.6540697674418605,
      "grad_norm": 2.5688010790796154,
      "learning_rate": 3.223847808687415e-06,
      "loss": 0.3982,
      "step": 1575
    },
    {
      "epoch": 0.6561461794019934,
      "grad_norm": 2.2695295812005836,
      "learning_rate": 3.190015048523528e-06,
      "loss": 0.3912,
      "step": 1580
    },
    {
      "epoch": 0.6582225913621262,
      "grad_norm": 2.5664307243505227,
      "learning_rate": 3.156277391529796e-06,
      "loss": 0.4044,
      "step": 1585
    },
    {
      "epoch": 0.6602990033222591,
      "grad_norm": 2.421377162101449,
      "learning_rate": 3.1226366104050067e-06,
      "loss": 0.4061,
      "step": 1590
    },
    {
      "epoch": 0.6623754152823921,
      "grad_norm": 2.50702313044333,
      "learning_rate": 3.089094472757742e-06,
      "loss": 0.3986,
      "step": 1595
    },
    {
      "epoch": 0.6644518272425249,
      "grad_norm": 2.2015982709846122,
      "learning_rate": 3.055652741013497e-06,
      "loss": 0.3773,
      "step": 1600
    },
    {
      "epoch": 0.6665282392026578,
      "grad_norm": 2.484025604844624,
      "learning_rate": 3.0223131723220756e-06,
      "loss": 0.4043,
      "step": 1605
    },
    {
      "epoch": 0.6686046511627907,
      "grad_norm": 2.2673450694224426,
      "learning_rate": 2.9890775184652666e-06,
      "loss": 0.3975,
      "step": 1610
    },
    {
      "epoch": 0.6706810631229236,
      "grad_norm": 2.411243052140437,
      "learning_rate": 2.955947525764796e-06,
      "loss": 0.4162,
      "step": 1615
    },
    {
      "epoch": 0.6727574750830565,
      "grad_norm": 2.467788088547966,
      "learning_rate": 2.9229249349905686e-06,
      "loss": 0.3905,
      "step": 1620
    },
    {
      "epoch": 0.6748338870431894,
      "grad_norm": 2.441034044229084,
      "learning_rate": 2.890011481269204e-06,
      "loss": 0.404,
      "step": 1625
    },
    {
      "epoch": 0.6769102990033222,
      "grad_norm": 2.4310426686498507,
      "learning_rate": 2.8572088939928623e-06,
      "loss": 0.3985,
      "step": 1630
    },
    {
      "epoch": 0.6789867109634552,
      "grad_norm": 2.5154739727394397,
      "learning_rate": 2.824518896728386e-06,
      "loss": 0.3972,
      "step": 1635
    },
    {
      "epoch": 0.6810631229235881,
      "grad_norm": 2.4239374759188066,
      "learning_rate": 2.7919432071267212e-06,
      "loss": 0.3986,
      "step": 1640
    },
    {
      "epoch": 0.6831395348837209,
      "grad_norm": 2.401230714452262,
      "learning_rate": 2.759483536832682e-06,
      "loss": 0.3961,
      "step": 1645
    },
    {
      "epoch": 0.6852159468438538,
      "grad_norm": 2.3945770626194425,
      "learning_rate": 2.7271415913950027e-06,
      "loss": 0.3987,
      "step": 1650
    },
    {
      "epoch": 0.6872923588039868,
      "grad_norm": 2.5083750676716248,
      "learning_rate": 2.6949190701767323e-06,
      "loss": 0.3987,
      "step": 1655
    },
    {
      "epoch": 0.6893687707641196,
      "grad_norm": 2.359597868105036,
      "learning_rate": 2.662817666265932e-06,
      "loss": 0.3992,
      "step": 1660
    },
    {
      "epoch": 0.6914451827242525,
      "grad_norm": 2.3950900870588305,
      "learning_rate": 2.6308390663867247e-06,
      "loss": 0.3755,
      "step": 1665
    },
    {
      "epoch": 0.6935215946843853,
      "grad_norm": 2.2726643843793783,
      "learning_rate": 2.5989849508106663e-06,
      "loss": 0.3788,
      "step": 1670
    },
    {
      "epoch": 0.6955980066445183,
      "grad_norm": 2.3688642141053644,
      "learning_rate": 2.5672569932684486e-06,
      "loss": 0.3923,
      "step": 1675
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 2.4674555381530543,
      "learning_rate": 2.5356568608619737e-06,
      "loss": 0.3784,
      "step": 1680
    },
    {
      "epoch": 0.699750830564784,
      "grad_norm": 2.348080957902949,
      "learning_rate": 2.504186213976736e-06,
      "loss": 0.3888,
      "step": 1685
    },
    {
      "epoch": 0.7018272425249169,
      "grad_norm": 2.2245908133987506,
      "learning_rate": 2.4728467061946017e-06,
      "loss": 0.383,
      "step": 1690
    },
    {
      "epoch": 0.7039036544850499,
      "grad_norm": 2.308262964854599,
      "learning_rate": 2.441639984206903e-06,
      "loss": 0.3873,
      "step": 1695
    },
    {
      "epoch": 0.7059800664451827,
      "grad_norm": 2.3316191201720726,
      "learning_rate": 2.4105676877279376e-06,
      "loss": 0.3764,
      "step": 1700
    },
    {
      "epoch": 0.7080564784053156,
      "grad_norm": 2.2575654898253363,
      "learning_rate": 2.379631449408788e-06,
      "loss": 0.3857,
      "step": 1705
    },
    {
      "epoch": 0.7101328903654485,
      "grad_norm": 2.295434521334263,
      "learning_rate": 2.3488328947515566e-06,
      "loss": 0.3825,
      "step": 1710
    },
    {
      "epoch": 0.7122093023255814,
      "grad_norm": 2.3045365012329704,
      "learning_rate": 2.318173642023939e-06,
      "loss": 0.3851,
      "step": 1715
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 2.3117392889776665,
      "learning_rate": 2.287655302174208e-06,
      "loss": 0.3897,
      "step": 1720
    },
    {
      "epoch": 0.7163621262458472,
      "grad_norm": 2.422532892044474,
      "learning_rate": 2.257279478746564e-06,
      "loss": 0.3799,
      "step": 1725
    },
    {
      "epoch": 0.71843853820598,
      "grad_norm": 2.2839185079742514,
      "learning_rate": 2.2270477677968727e-06,
      "loss": 0.3703,
      "step": 1730
    },
    {
      "epoch": 0.720514950166113,
      "grad_norm": 2.7279247585921786,
      "learning_rate": 2.196961757808813e-06,
      "loss": 0.3794,
      "step": 1735
    },
    {
      "epoch": 0.7225913621262459,
      "grad_norm": 2.396361579385602,
      "learning_rate": 2.167023029610402e-06,
      "loss": 0.3642,
      "step": 1740
    },
    {
      "epoch": 0.7246677740863787,
      "grad_norm": 2.340856081292544,
      "learning_rate": 2.1372331562909453e-06,
      "loss": 0.372,
      "step": 1745
    },
    {
      "epoch": 0.7267441860465116,
      "grad_norm": 2.413915292833693,
      "learning_rate": 2.1075937031183636e-06,
      "loss": 0.3767,
      "step": 1750
    },
    {
      "epoch": 0.7288205980066446,
      "grad_norm": 2.2094868525489386,
      "learning_rate": 2.0781062274569657e-06,
      "loss": 0.3713,
      "step": 1755
    },
    {
      "epoch": 0.7308970099667774,
      "grad_norm": 2.2242377702402663,
      "learning_rate": 2.0487722786856107e-06,
      "loss": 0.3808,
      "step": 1760
    },
    {
      "epoch": 0.7329734219269103,
      "grad_norm": 2.451226818715509,
      "learning_rate": 2.019593398116292e-06,
      "loss": 0.3752,
      "step": 1765
    },
    {
      "epoch": 0.7350498338870431,
      "grad_norm": 2.5070300923436006,
      "learning_rate": 1.990571118913166e-06,
      "loss": 0.3754,
      "step": 1770
    },
    {
      "epoch": 0.7371262458471761,
      "grad_norm": 2.4891905395473963,
      "learning_rate": 1.961706966011978e-06,
      "loss": 0.3877,
      "step": 1775
    },
    {
      "epoch": 0.739202657807309,
      "grad_norm": 2.4842650358701905,
      "learning_rate": 1.9330024560399507e-06,
      "loss": 0.3836,
      "step": 1780
    },
    {
      "epoch": 0.7412790697674418,
      "grad_norm": 2.250133568783516,
      "learning_rate": 1.9044590972360822e-06,
      "loss": 0.3725,
      "step": 1785
    },
    {
      "epoch": 0.7433554817275747,
      "grad_norm": 2.341904795212687,
      "learning_rate": 1.876078389371911e-06,
      "loss": 0.3679,
      "step": 1790
    },
    {
      "epoch": 0.7454318936877077,
      "grad_norm": 2.3068998565270746,
      "learning_rate": 1.8478618236726992e-06,
      "loss": 0.3757,
      "step": 1795
    },
    {
      "epoch": 0.7475083056478405,
      "grad_norm": 2.2619310866276203,
      "learning_rate": 1.8198108827390892e-06,
      "loss": 0.3742,
      "step": 1800
    },
    {
      "epoch": 0.7495847176079734,
      "grad_norm": 2.406091048606607,
      "learning_rate": 1.791927040469198e-06,
      "loss": 0.3805,
      "step": 1805
    },
    {
      "epoch": 0.7516611295681063,
      "grad_norm": 2.3430777426784077,
      "learning_rate": 1.7642117619811672e-06,
      "loss": 0.3744,
      "step": 1810
    },
    {
      "epoch": 0.7537375415282392,
      "grad_norm": 2.309496934162411,
      "learning_rate": 1.7366665035361947e-06,
      "loss": 0.3856,
      "step": 1815
    },
    {
      "epoch": 0.7558139534883721,
      "grad_norm": 2.3680236136606085,
      "learning_rate": 1.7092927124620007e-06,
      "loss": 0.3747,
      "step": 1820
    },
    {
      "epoch": 0.757890365448505,
      "grad_norm": 2.3303370070854066,
      "learning_rate": 1.682091827076796e-06,
      "loss": 0.3724,
      "step": 1825
    },
    {
      "epoch": 0.7599667774086378,
      "grad_norm": 2.308665058379731,
      "learning_rate": 1.6550652766136932e-06,
      "loss": 0.3701,
      "step": 1830
    },
    {
      "epoch": 0.7620431893687708,
      "grad_norm": 2.423141151726278,
      "learning_rate": 1.6282144811456196e-06,
      "loss": 0.3749,
      "step": 1835
    },
    {
      "epoch": 0.7641196013289037,
      "grad_norm": 2.310790310097539,
      "learning_rate": 1.6015408515107e-06,
      "loss": 0.3649,
      "step": 1840
    },
    {
      "epoch": 0.7661960132890365,
      "grad_norm": 2.350953218186428,
      "learning_rate": 1.5750457892381183e-06,
      "loss": 0.3766,
      "step": 1845
    },
    {
      "epoch": 0.7682724252491694,
      "grad_norm": 2.3685044215677826,
      "learning_rate": 1.5487306864744878e-06,
      "loss": 0.3626,
      "step": 1850
    },
    {
      "epoch": 0.7703488372093024,
      "grad_norm": 2.4283396349263384,
      "learning_rate": 1.5225969259106909e-06,
      "loss": 0.358,
      "step": 1855
    },
    {
      "epoch": 0.7724252491694352,
      "grad_norm": 2.515904865078178,
      "learning_rate": 1.4966458807092404e-06,
      "loss": 0.3703,
      "step": 1860
    },
    {
      "epoch": 0.7745016611295681,
      "grad_norm": 2.369156818267499,
      "learning_rate": 1.470878914432115e-06,
      "loss": 0.3628,
      "step": 1865
    },
    {
      "epoch": 0.776578073089701,
      "grad_norm": 2.3898762463795302,
      "learning_rate": 1.4452973809691245e-06,
      "loss": 0.3491,
      "step": 1870
    },
    {
      "epoch": 0.7786544850498339,
      "grad_norm": 2.307405290268551,
      "learning_rate": 1.4199026244667636e-06,
      "loss": 0.3715,
      "step": 1875
    },
    {
      "epoch": 0.7807308970099668,
      "grad_norm": 2.3679557325362808,
      "learning_rate": 1.3946959792575915e-06,
      "loss": 0.3716,
      "step": 1880
    },
    {
      "epoch": 0.7828073089700996,
      "grad_norm": 2.368304219604154,
      "learning_rate": 1.3696787697901131e-06,
      "loss": 0.3661,
      "step": 1885
    },
    {
      "epoch": 0.7848837209302325,
      "grad_norm": 2.337789695422565,
      "learning_rate": 1.3448523105591976e-06,
      "loss": 0.3605,
      "step": 1890
    },
    {
      "epoch": 0.7869601328903655,
      "grad_norm": 2.611258973451232,
      "learning_rate": 1.3202179060370041e-06,
      "loss": 0.3699,
      "step": 1895
    },
    {
      "epoch": 0.7890365448504983,
      "grad_norm": 2.437657572808606,
      "learning_rate": 1.2957768506044383e-06,
      "loss": 0.3651,
      "step": 1900
    },
    {
      "epoch": 0.7911129568106312,
      "grad_norm": 2.388228690853508,
      "learning_rate": 1.2715304284831492e-06,
      "loss": 0.3664,
      "step": 1905
    },
    {
      "epoch": 0.7931893687707641,
      "grad_norm": 2.2860587085497235,
      "learning_rate": 1.2474799136680394e-06,
      "loss": 0.3577,
      "step": 1910
    },
    {
      "epoch": 0.795265780730897,
      "grad_norm": 2.2178621526275077,
      "learning_rate": 1.223626569860339e-06,
      "loss": 0.3441,
      "step": 1915
    },
    {
      "epoch": 0.7973421926910299,
      "grad_norm": 2.518415787103085,
      "learning_rate": 1.1999716504011917e-06,
      "loss": 0.3673,
      "step": 1920
    },
    {
      "epoch": 0.7994186046511628,
      "grad_norm": 2.359475880122496,
      "learning_rate": 1.1765163982058109e-06,
      "loss": 0.3567,
      "step": 1925
    },
    {
      "epoch": 0.8014950166112956,
      "grad_norm": 2.407404285602653,
      "learning_rate": 1.1532620456981685e-06,
      "loss": 0.3476,
      "step": 1930
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 2.562334088122669,
      "learning_rate": 1.1302098147462348e-06,
      "loss": 0.3658,
      "step": 1935
    },
    {
      "epoch": 0.8056478405315615,
      "grad_norm": 2.4467720130350163,
      "learning_rate": 1.1073609165977866e-06,
      "loss": 0.348,
      "step": 1940
    },
    {
      "epoch": 0.8077242524916943,
      "grad_norm": 2.3514873698583574,
      "learning_rate": 1.0847165518167513e-06,
      "loss": 0.3601,
      "step": 1945
    },
    {
      "epoch": 0.8098006644518272,
      "grad_norm": 2.258063143891622,
      "learning_rate": 1.062277910220138e-06,
      "loss": 0.3548,
      "step": 1950
    },
    {
      "epoch": 0.8118770764119602,
      "grad_norm": 2.3377988411022246,
      "learning_rate": 1.0400461708155095e-06,
      "loss": 0.3591,
      "step": 1955
    },
    {
      "epoch": 0.813953488372093,
      "grad_norm": 2.4485426821221004,
      "learning_rate": 1.0180225017390416e-06,
      "loss": 0.3583,
      "step": 1960
    },
    {
      "epoch": 0.8160299003322259,
      "grad_norm": 2.3726559534317797,
      "learning_rate": 9.962080601941365e-07,
      "loss": 0.3426,
      "step": 1965
    },
    {
      "epoch": 0.8181063122923588,
      "grad_norm": 2.2417751776494543,
      "learning_rate": 9.746039923906258e-07,
      "loss": 0.343,
      "step": 1970
    },
    {
      "epoch": 0.8201827242524917,
      "grad_norm": 2.5294843157217906,
      "learning_rate": 9.532114334845444e-07,
      "loss": 0.3664,
      "step": 1975
    },
    {
      "epoch": 0.8222591362126246,
      "grad_norm": 2.5572851406694235,
      "learning_rate": 9.320315075184771e-07,
      "loss": 0.3483,
      "step": 1980
    },
    {
      "epoch": 0.8243355481727574,
      "grad_norm": 2.4014306355585973,
      "learning_rate": 9.110653273625103e-07,
      "loss": 0.3454,
      "step": 1985
    },
    {
      "epoch": 0.8264119601328903,
      "grad_norm": 2.3699223457500715,
      "learning_rate": 8.903139946557437e-07,
      "loss": 0.3527,
      "step": 1990
    },
    {
      "epoch": 0.8284883720930233,
      "grad_norm": 2.4489197804834197,
      "learning_rate": 8.697785997484198e-07,
      "loss": 0.3535,
      "step": 1995
    },
    {
      "epoch": 0.8305647840531561,
      "grad_norm": 2.4381698669696044,
      "learning_rate": 8.494602216446213e-07,
      "loss": 0.3522,
      "step": 2000
    },
    {
      "epoch": 0.832641196013289,
      "grad_norm": 2.373612659548005,
      "learning_rate": 8.293599279455838e-07,
      "loss": 0.352,
      "step": 2005
    },
    {
      "epoch": 0.834717607973422,
      "grad_norm": 2.5001126967401763,
      "learning_rate": 8.094787747935995e-07,
      "loss": 0.3533,
      "step": 2010
    },
    {
      "epoch": 0.8367940199335548,
      "grad_norm": 2.4033229472375637,
      "learning_rate": 7.898178068165175e-07,
      "loss": 0.3569,
      "step": 2015
    },
    {
      "epoch": 0.8388704318936877,
      "grad_norm": 2.34177766700727,
      "learning_rate": 7.703780570728637e-07,
      "loss": 0.3485,
      "step": 2020
    },
    {
      "epoch": 0.8409468438538206,
      "grad_norm": 2.345211689975521,
      "learning_rate": 7.511605469975524e-07,
      "loss": 0.3541,
      "step": 2025
    },
    {
      "epoch": 0.8430232558139535,
      "grad_norm": 2.491346976334481,
      "learning_rate": 7.321662863482248e-07,
      "loss": 0.357,
      "step": 2030
    },
    {
      "epoch": 0.8450996677740864,
      "grad_norm": 2.4991193300068515,
      "learning_rate": 7.133962731521837e-07,
      "loss": 0.3504,
      "step": 2035
    },
    {
      "epoch": 0.8471760797342193,
      "grad_norm": 2.4131651786978376,
      "learning_rate": 6.948514936539596e-07,
      "loss": 0.3413,
      "step": 2040
    },
    {
      "epoch": 0.8492524916943521,
      "grad_norm": 2.4158508388648046,
      "learning_rate": 6.765329222634892e-07,
      "loss": 0.3368,
      "step": 2045
    },
    {
      "epoch": 0.8513289036544851,
      "grad_norm": 2.444048773418729,
      "learning_rate": 6.584415215049145e-07,
      "loss": 0.3478,
      "step": 2050
    },
    {
      "epoch": 0.853405315614618,
      "grad_norm": 2.3067727734077854,
      "learning_rate": 6.405782419660073e-07,
      "loss": 0.3539,
      "step": 2055
    },
    {
      "epoch": 0.8554817275747508,
      "grad_norm": 2.389540542776719,
      "learning_rate": 6.229440222482258e-07,
      "loss": 0.3568,
      "step": 2060
    },
    {
      "epoch": 0.8575581395348837,
      "grad_norm": 2.490728442827626,
      "learning_rate": 6.055397889173947e-07,
      "loss": 0.3425,
      "step": 2065
    },
    {
      "epoch": 0.8596345514950167,
      "grad_norm": 2.4309142506564116,
      "learning_rate": 5.88366456455019e-07,
      "loss": 0.3556,
      "step": 2070
    },
    {
      "epoch": 0.8617109634551495,
      "grad_norm": 2.577695548294538,
      "learning_rate": 5.714249272102368e-07,
      "loss": 0.3479,
      "step": 2075
    },
    {
      "epoch": 0.8637873754152824,
      "grad_norm": 2.3780994980865513,
      "learning_rate": 5.547160913524024e-07,
      "loss": 0.3407,
      "step": 2080
    },
    {
      "epoch": 0.8658637873754153,
      "grad_norm": 2.3471940728385645,
      "learning_rate": 5.382408268243194e-07,
      "loss": 0.327,
      "step": 2085
    },
    {
      "epoch": 0.8679401993355482,
      "grad_norm": 2.5308209588235964,
      "learning_rate": 5.219999992961044e-07,
      "loss": 0.3486,
      "step": 2090
    },
    {
      "epoch": 0.8700166112956811,
      "grad_norm": 2.347529844497377,
      "learning_rate": 5.05994462119705e-07,
      "loss": 0.3507,
      "step": 2095
    },
    {
      "epoch": 0.872093023255814,
      "grad_norm": 2.4490768218202428,
      "learning_rate": 4.902250562840622e-07,
      "loss": 0.3484,
      "step": 2100
    },
    {
      "epoch": 0.8741694352159468,
      "grad_norm": 2.4607053819399227,
      "learning_rate": 4.7469261037091765e-07,
      "loss": 0.355,
      "step": 2105
    },
    {
      "epoch": 0.8762458471760798,
      "grad_norm": 2.37905091425431,
      "learning_rate": 4.5939794051128363e-07,
      "loss": 0.3544,
      "step": 2110
    },
    {
      "epoch": 0.8783222591362126,
      "grad_norm": 2.3898177002048397,
      "learning_rate": 4.443418503425517e-07,
      "loss": 0.3459,
      "step": 2115
    },
    {
      "epoch": 0.8803986710963455,
      "grad_norm": 2.3945638825763336,
      "learning_rate": 4.295251309662768e-07,
      "loss": 0.3475,
      "step": 2120
    },
    {
      "epoch": 0.8824750830564784,
      "grad_norm": 2.376437633901908,
      "learning_rate": 4.149485609066001e-07,
      "loss": 0.3448,
      "step": 2125
    },
    {
      "epoch": 0.8845514950166113,
      "grad_norm": 2.4682795986451884,
      "learning_rate": 4.0061290606935145e-07,
      "loss": 0.3501,
      "step": 2130
    },
    {
      "epoch": 0.8866279069767442,
      "grad_norm": 2.307696986215917,
      "learning_rate": 3.8651891970179876e-07,
      "loss": 0.3509,
      "step": 2135
    },
    {
      "epoch": 0.8887043189368771,
      "grad_norm": 2.2638655900879323,
      "learning_rate": 3.7266734235307357e-07,
      "loss": 0.3494,
      "step": 2140
    },
    {
      "epoch": 0.8907807308970099,
      "grad_norm": 2.4074516319355865,
      "learning_rate": 3.5905890183525916e-07,
      "loss": 0.3381,
      "step": 2145
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 2.4580735039851263,
      "learning_rate": 3.4569431318514647e-07,
      "loss": 0.3506,
      "step": 2150
    },
    {
      "epoch": 0.8949335548172758,
      "grad_norm": 2.223651003352099,
      "learning_rate": 3.3257427862666894e-07,
      "loss": 0.3426,
      "step": 2155
    },
    {
      "epoch": 0.8970099667774086,
      "grad_norm": 2.5240054200803925,
      "learning_rate": 3.196994875339976e-07,
      "loss": 0.3394,
      "step": 2160
    },
    {
      "epoch": 0.8990863787375415,
      "grad_norm": 2.650418412385108,
      "learning_rate": 3.0707061639532687e-07,
      "loss": 0.3469,
      "step": 2165
    },
    {
      "epoch": 0.9011627906976745,
      "grad_norm": 2.5283079967315256,
      "learning_rate": 2.946883287773211e-07,
      "loss": 0.3572,
      "step": 2170
    },
    {
      "epoch": 0.9032392026578073,
      "grad_norm": 2.482824449172331,
      "learning_rate": 2.82553275290256e-07,
      "loss": 0.3469,
      "step": 2175
    },
    {
      "epoch": 0.9053156146179402,
      "grad_norm": 2.42162117653704,
      "learning_rate": 2.706660935538297e-07,
      "loss": 0.3522,
      "step": 2180
    },
    {
      "epoch": 0.907392026578073,
      "grad_norm": 2.610628055343181,
      "learning_rate": 2.590274081636568e-07,
      "loss": 0.3326,
      "step": 2185
    },
    {
      "epoch": 0.909468438538206,
      "grad_norm": 2.337754822501405,
      "learning_rate": 2.476378306584576e-07,
      "loss": 0.3472,
      "step": 2190
    },
    {
      "epoch": 0.9115448504983389,
      "grad_norm": 2.422013772805342,
      "learning_rate": 2.3649795948791744e-07,
      "loss": 0.3291,
      "step": 2195
    },
    {
      "epoch": 0.9136212624584718,
      "grad_norm": 2.5260012444754865,
      "learning_rate": 2.2560837998124862e-07,
      "loss": 0.3443,
      "step": 2200
    },
    {
      "epoch": 0.9156976744186046,
      "grad_norm": 2.5167784300702203,
      "learning_rate": 2.1496966431642895e-07,
      "loss": 0.344,
      "step": 2205
    },
    {
      "epoch": 0.9177740863787376,
      "grad_norm": 2.5184080924547976,
      "learning_rate": 2.0458237149014347e-07,
      "loss": 0.3431,
      "step": 2210
    },
    {
      "epoch": 0.9198504983388704,
      "grad_norm": 2.6121850478268915,
      "learning_rate": 1.944470472884097e-07,
      "loss": 0.3469,
      "step": 2215
    },
    {
      "epoch": 0.9219269102990033,
      "grad_norm": 2.4250182138955987,
      "learning_rate": 1.8456422425789822e-07,
      "loss": 0.346,
      "step": 2220
    },
    {
      "epoch": 0.9240033222591362,
      "grad_norm": 2.4126854578567056,
      "learning_rate": 1.7493442167795526e-07,
      "loss": 0.3394,
      "step": 2225
    },
    {
      "epoch": 0.9260797342192691,
      "grad_norm": 2.2732400743037546,
      "learning_rate": 1.6555814553331328e-07,
      "loss": 0.3474,
      "step": 2230
    },
    {
      "epoch": 0.928156146179402,
      "grad_norm": 2.4576436036196867,
      "learning_rate": 1.5643588848750944e-07,
      "loss": 0.3455,
      "step": 2235
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 2.417373647969096,
      "learning_rate": 1.4756812985699364e-07,
      "loss": 0.3389,
      "step": 2240
    },
    {
      "epoch": 0.9323089700996677,
      "grad_norm": 2.314864797926019,
      "learning_rate": 1.3895533558594853e-07,
      "loss": 0.3307,
      "step": 2245
    },
    {
      "epoch": 0.9343853820598007,
      "grad_norm": 2.4942438872944375,
      "learning_rate": 1.305979582218042e-07,
      "loss": 0.3413,
      "step": 2250
    },
    {
      "epoch": 0.9364617940199336,
      "grad_norm": 2.4271492623044733,
      "learning_rate": 1.224964368914622e-07,
      "loss": 0.3533,
      "step": 2255
    },
    {
      "epoch": 0.9385382059800664,
      "grad_norm": 2.404072393255019,
      "learning_rate": 1.1465119727821828e-07,
      "loss": 0.3388,
      "step": 2260
    },
    {
      "epoch": 0.9406146179401993,
      "grad_norm": 2.4291366569357233,
      "learning_rate": 1.0706265159939944e-07,
      "loss": 0.329,
      "step": 2265
    },
    {
      "epoch": 0.9426910299003323,
      "grad_norm": 2.370319609790916,
      "learning_rate": 9.973119858470326e-08,
      "loss": 0.3435,
      "step": 2270
    },
    {
      "epoch": 0.9447674418604651,
      "grad_norm": 2.612518036597659,
      "learning_rate": 9.265722345524475e-08,
      "loss": 0.3544,
      "step": 2275
    },
    {
      "epoch": 0.946843853820598,
      "grad_norm": 2.325383175606347,
      "learning_rate": 8.584109790331918e-08,
      "loss": 0.334,
      "step": 2280
    },
    {
      "epoch": 0.9489202657807309,
      "grad_norm": 2.483650038896797,
      "learning_rate": 7.92831800728705e-08,
      "loss": 0.3495,
      "step": 2285
    },
    {
      "epoch": 0.9509966777408638,
      "grad_norm": 2.3917415303858323,
      "learning_rate": 7.29838145406725e-08,
      "loss": 0.3525,
      "step": 2290
    },
    {
      "epoch": 0.9530730897009967,
      "grad_norm": 2.3491361755345297,
      "learning_rate": 6.69433322982238e-08,
      "loss": 0.3261,
      "step": 2295
    },
    {
      "epoch": 0.9551495016611296,
      "grad_norm": 2.395546616132027,
      "learning_rate": 6.116205073435632e-08,
      "loss": 0.3572,
      "step": 2300
    },
    {
      "epoch": 0.9572259136212624,
      "grad_norm": 2.38127790933904,
      "learning_rate": 5.5640273618560724e-08,
      "loss": 0.3477,
      "step": 2305
    },
    {
      "epoch": 0.9593023255813954,
      "grad_norm": 2.437345577309693,
      "learning_rate": 5.0378291085020905e-08,
      "loss": 0.3498,
      "step": 2310
    },
    {
      "epoch": 0.9613787375415282,
      "grad_norm": 2.3743576771009125,
      "learning_rate": 4.537637961737285e-08,
      "loss": 0.3537,
      "step": 2315
    },
    {
      "epoch": 0.9634551495016611,
      "grad_norm": 2.4186159686143816,
      "learning_rate": 4.063480203417625e-08,
      "loss": 0.3491,
      "step": 2320
    },
    {
      "epoch": 0.965531561461794,
      "grad_norm": 2.43658746364112,
      "learning_rate": 3.6153807475103886e-08,
      "loss": 0.3372,
      "step": 2325
    },
    {
      "epoch": 0.967607973421927,
      "grad_norm": 2.433100952556644,
      "learning_rate": 3.1933631387853215e-08,
      "loss": 0.34,
      "step": 2330
    },
    {
      "epoch": 0.9696843853820598,
      "grad_norm": 2.3533082714101288,
      "learning_rate": 2.7974495515772915e-08,
      "loss": 0.3478,
      "step": 2335
    },
    {
      "epoch": 0.9717607973421927,
      "grad_norm": 2.3042817476032296,
      "learning_rate": 2.427660788621222e-08,
      "loss": 0.3522,
      "step": 2340
    },
    {
      "epoch": 0.9738372093023255,
      "grad_norm": 2.5758246509298184,
      "learning_rate": 2.0840162799591335e-08,
      "loss": 0.3518,
      "step": 2345
    },
    {
      "epoch": 0.9759136212624585,
      "grad_norm": 2.2839484862848254,
      "learning_rate": 1.7665340819192356e-08,
      "loss": 0.3412,
      "step": 2350
    },
    {
      "epoch": 0.9779900332225914,
      "grad_norm": 2.3825796160738184,
      "learning_rate": 1.475230876166911e-08,
      "loss": 0.3484,
      "step": 2355
    },
    {
      "epoch": 0.9800664451827242,
      "grad_norm": 2.436946151591597,
      "learning_rate": 1.2101219688285815e-08,
      "loss": 0.3406,
      "step": 2360
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 2.3004422772721385,
      "learning_rate": 9.712212896871854e-09,
      "loss": 0.3483,
      "step": 2365
    },
    {
      "epoch": 0.9842192691029901,
      "grad_norm": 2.243191260776767,
      "learning_rate": 7.585413914503182e-09,
      "loss": 0.3279,
      "step": 2370
    },
    {
      "epoch": 0.9862956810631229,
      "grad_norm": 2.4579139874339213,
      "learning_rate": 5.720934490907604e-09,
      "loss": 0.3539,
      "step": 2375
    },
    {
      "epoch": 0.9883720930232558,
      "grad_norm": 2.2560463637497885,
      "learning_rate": 4.118872592592804e-09,
      "loss": 0.3376,
      "step": 2380
    },
    {
      "epoch": 0.9904485049833887,
      "grad_norm": 2.4374426699588327,
      "learning_rate": 2.7793123976976866e-09,
      "loss": 0.337,
      "step": 2385
    },
    {
      "epoch": 0.9925249169435216,
      "grad_norm": 2.4021295260594466,
      "learning_rate": 1.7023242915703563e-09,
      "loss": 0.3422,
      "step": 2390
    },
    {
      "epoch": 0.9946013289036545,
      "grad_norm": 2.3647029145641847,
      "learning_rate": 8.879648630705229e-10,
      "loss": 0.3402,
      "step": 2395
    },
    {
      "epoch": 0.9966777408637874,
      "grad_norm": 2.377691718973852,
      "learning_rate": 3.362769015941014e-10,
      "loss": 0.3437,
      "step": 2400
    },
    {
      "epoch": 0.9987541528239202,
      "grad_norm": 2.277353937646912,
      "learning_rate": 4.7289394825567046e-11,
      "loss": 0.3486,
      "step": 2405
    },
    {
      "epoch": 1.0,
      "eval_runtime": 3.4135,
      "eval_samples_per_second": 2.93,
      "eval_steps_per_second": 0.879,
      "step": 2408
    },
    {
      "epoch": 1.0,
      "step": 2408,
      "total_flos": 252093105438720.0,
      "train_loss": 0.0,
      "train_runtime": 0.0085,
      "train_samples_per_second": 4527521.09,
      "train_steps_per_second": 283028.837
    }
  ],
  "logging_steps": 5,
  "max_steps": 2408,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 252093105438720.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}