{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9999607119003654,
  "eval_steps": 500,
  "global_step": 6363,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00015715239853848268,
      "grad_norm": 29.15457534790039,
      "learning_rate": 1.5698587127158556e-06,
      "loss": 5.9252,
      "step": 1
    },
    {
      "epoch": 0.0007857619926924135,
      "grad_norm": 24.992176055908203,
      "learning_rate": 7.849293563579277e-06,
      "loss": 5.9618,
      "step": 5
    },
    {
      "epoch": 0.001571523985384827,
      "grad_norm": 25.485349655151367,
      "learning_rate": 1.5698587127158555e-05,
      "loss": 5.9265,
      "step": 10
    },
    {
      "epoch": 0.0023572859780772405,
      "grad_norm": 20.05711555480957,
      "learning_rate": 2.3547880690737836e-05,
      "loss": 5.9391,
      "step": 15
    },
    {
      "epoch": 0.003143047970769654,
      "grad_norm": 8.419732093811035,
      "learning_rate": 3.139717425431711e-05,
      "loss": 5.6583,
      "step": 20
    },
    {
      "epoch": 0.003928809963462067,
      "grad_norm": 7.963233470916748,
      "learning_rate": 3.924646781789639e-05,
      "loss": 5.4406,
      "step": 25
    },
    {
      "epoch": 0.004714571956154481,
      "grad_norm": 4.564691066741943,
      "learning_rate": 4.709576138147567e-05,
      "loss": 5.1726,
      "step": 30
    },
    {
      "epoch": 0.005500333948846894,
      "grad_norm": 4.1633124351501465,
      "learning_rate": 5.4945054945054945e-05,
      "loss": 4.9458,
      "step": 35
    },
    {
      "epoch": 0.006286095941539308,
      "grad_norm": 3.9529995918273926,
      "learning_rate": 6.279434850863422e-05,
      "loss": 4.6674,
      "step": 40
    },
    {
      "epoch": 0.0070718579342317215,
      "grad_norm": 3.318066120147705,
      "learning_rate": 7.06436420722135e-05,
      "loss": 4.4785,
      "step": 45
    },
    {
      "epoch": 0.007857619926924134,
      "grad_norm": 3.131993055343628,
      "learning_rate": 7.849293563579278e-05,
      "loss": 4.3329,
      "step": 50
    },
    {
      "epoch": 0.008643381919616548,
      "grad_norm": 2.6008388996124268,
      "learning_rate": 8.634222919937205e-05,
      "loss": 4.1992,
      "step": 55
    },
    {
      "epoch": 0.009429143912308962,
      "grad_norm": 2.183000326156616,
      "learning_rate": 9.419152276295134e-05,
      "loss": 4.0823,
      "step": 60
    },
    {
      "epoch": 0.010214905905001376,
      "grad_norm": 1.9434701204299927,
      "learning_rate": 0.00010204081632653062,
      "loss": 3.8471,
      "step": 65
    },
    {
      "epoch": 0.011000667897693788,
      "grad_norm": 1.625468134880066,
      "learning_rate": 0.00010989010989010989,
      "loss": 3.6757,
      "step": 70
    },
    {
      "epoch": 0.011786429890386202,
      "grad_norm": 1.4538207054138184,
      "learning_rate": 0.00011773940345368916,
      "loss": 3.5552,
      "step": 75
    },
    {
      "epoch": 0.012572191883078615,
      "grad_norm": 1.18235182762146,
      "learning_rate": 0.00012558869701726844,
      "loss": 3.5154,
      "step": 80
    },
    {
      "epoch": 0.01335795387577103,
      "grad_norm": 0.815110981464386,
      "learning_rate": 0.00013343799058084774,
      "loss": 3.3802,
      "step": 85
    },
    {
      "epoch": 0.014143715868463443,
      "grad_norm": 0.662001371383667,
      "learning_rate": 0.000141287284144427,
      "loss": 3.2555,
      "step": 90
    },
    {
      "epoch": 0.014929477861155857,
      "grad_norm": 0.5505642890930176,
      "learning_rate": 0.0001491365777080063,
      "loss": 3.0703,
      "step": 95
    },
    {
      "epoch": 0.01571523985384827,
      "grad_norm": 0.4972209334373474,
      "learning_rate": 0.00015698587127158556,
      "loss": 3.0455,
      "step": 100
    },
    {
      "epoch": 0.016501001846540683,
      "grad_norm": 0.427781879901886,
      "learning_rate": 0.00016483516483516484,
      "loss": 2.9642,
      "step": 105
    },
    {
      "epoch": 0.017286763839233096,
      "grad_norm": 0.3651106655597687,
      "learning_rate": 0.0001726844583987441,
      "loss": 2.8522,
      "step": 110
    },
    {
      "epoch": 0.01807252583192551,
      "grad_norm": 0.3085882067680359,
      "learning_rate": 0.00018053375196232338,
      "loss": 2.8798,
      "step": 115
    },
    {
      "epoch": 0.018858287824617924,
      "grad_norm": 0.277054101228714,
      "learning_rate": 0.00018838304552590268,
      "loss": 2.8542,
      "step": 120
    },
    {
      "epoch": 0.019644049817310338,
      "grad_norm": 0.2966115474700928,
      "learning_rate": 0.00019623233908948196,
      "loss": 2.8472,
      "step": 125
    },
    {
      "epoch": 0.02042981181000275,
      "grad_norm": 0.3365791440010071,
      "learning_rate": 0.00020408163265306123,
      "loss": 2.7396,
      "step": 130
    },
    {
      "epoch": 0.021215573802695165,
      "grad_norm": 0.2897559702396393,
      "learning_rate": 0.0002119309262166405,
      "loss": 2.7275,
      "step": 135
    },
    {
      "epoch": 0.022001335795387576,
      "grad_norm": 0.3298715651035309,
      "learning_rate": 0.00021978021978021978,
      "loss": 2.6553,
      "step": 140
    },
    {
      "epoch": 0.02278709778807999,
      "grad_norm": 0.4270775020122528,
      "learning_rate": 0.00022762951334379905,
      "loss": 2.6166,
      "step": 145
    },
    {
      "epoch": 0.023572859780772403,
      "grad_norm": 0.29401084780693054,
      "learning_rate": 0.00023547880690737833,
      "loss": 2.5642,
      "step": 150
    },
    {
      "epoch": 0.024358621773464817,
      "grad_norm": 0.2512624263763428,
      "learning_rate": 0.00024332810047095763,
      "loss": 2.5249,
      "step": 155
    },
    {
      "epoch": 0.02514438376615723,
      "grad_norm": 0.27092209458351135,
      "learning_rate": 0.0002511773940345369,
      "loss": 2.5108,
      "step": 160
    },
    {
      "epoch": 0.025930145758849645,
      "grad_norm": 0.2265215367078781,
      "learning_rate": 0.0002590266875981162,
      "loss": 2.5287,
      "step": 165
    },
    {
      "epoch": 0.02671590775154206,
      "grad_norm": 0.2427075356245041,
      "learning_rate": 0.0002668759811616955,
      "loss": 2.5689,
      "step": 170
    },
    {
      "epoch": 0.027501669744234472,
      "grad_norm": 0.20640349388122559,
      "learning_rate": 0.0002747252747252748,
      "loss": 2.5005,
      "step": 175
    },
    {
      "epoch": 0.028287431736926886,
      "grad_norm": 0.2663057744503021,
      "learning_rate": 0.000282574568288854,
      "loss": 2.4339,
      "step": 180
    },
    {
      "epoch": 0.0290731937296193,
      "grad_norm": 0.29693251848220825,
      "learning_rate": 0.00029042386185243333,
      "loss": 2.4894,
      "step": 185
    },
    {
      "epoch": 0.029858955722311713,
      "grad_norm": 0.2689124643802643,
      "learning_rate": 0.0002982731554160126,
      "loss": 2.525,
      "step": 190
    },
    {
      "epoch": 0.030644717715004124,
      "grad_norm": 0.23312485218048096,
      "learning_rate": 0.0003061224489795919,
      "loss": 2.4748,
      "step": 195
    },
    {
      "epoch": 0.03143047970769654,
      "grad_norm": 0.21752777695655823,
      "learning_rate": 0.0003139717425431711,
      "loss": 2.4346,
      "step": 200
    },
    {
      "epoch": 0.032216241700388955,
      "grad_norm": 0.18333539366722107,
      "learning_rate": 0.0003218210361067504,
      "loss": 2.3613,
      "step": 205
    },
    {
      "epoch": 0.033002003693081365,
      "grad_norm": 0.2482461780309677,
      "learning_rate": 0.00032967032967032967,
      "loss": 2.3628,
      "step": 210
    },
    {
      "epoch": 0.03378776568577378,
      "grad_norm": 0.36294370889663696,
      "learning_rate": 0.00033751962323390897,
      "loss": 2.4492,
      "step": 215
    },
    {
      "epoch": 0.03457352767846619,
      "grad_norm": 0.2805122137069702,
      "learning_rate": 0.0003453689167974882,
      "loss": 2.376,
      "step": 220
    },
    {
      "epoch": 0.0353592896711586,
      "grad_norm": 0.2519840598106384,
      "learning_rate": 0.0003532182103610675,
      "loss": 2.309,
      "step": 225
    },
    {
      "epoch": 0.03614505166385102,
      "grad_norm": 0.19297201931476593,
      "learning_rate": 0.00036106750392464677,
      "loss": 2.299,
      "step": 230
    },
    {
      "epoch": 0.03693081365654343,
      "grad_norm": 0.18939702212810516,
      "learning_rate": 0.00036891679748822607,
      "loss": 2.4143,
      "step": 235
    },
    {
      "epoch": 0.03771657564923585,
      "grad_norm": 0.20321442186832428,
      "learning_rate": 0.00037676609105180537,
      "loss": 2.3443,
      "step": 240
    },
    {
      "epoch": 0.03850233764192826,
      "grad_norm": 0.1922532320022583,
      "learning_rate": 0.00038461538461538467,
      "loss": 2.2984,
      "step": 245
    },
    {
      "epoch": 0.039288099634620675,
      "grad_norm": 0.2797047197818756,
      "learning_rate": 0.0003924646781789639,
      "loss": 2.2656,
      "step": 250
    },
    {
      "epoch": 0.040073861627313086,
      "grad_norm": 0.21200767159461975,
      "learning_rate": 0.0004003139717425432,
      "loss": 2.285,
      "step": 255
    },
    {
      "epoch": 0.0408596236200055,
      "grad_norm": 0.21136586368083954,
      "learning_rate": 0.00040816326530612246,
      "loss": 2.284,
      "step": 260
    },
    {
      "epoch": 0.04164538561269791,
      "grad_norm": 0.16410328447818756,
      "learning_rate": 0.00041601255886970177,
      "loss": 2.1946,
      "step": 265
    },
    {
      "epoch": 0.04243114760539033,
      "grad_norm": 0.2457076907157898,
      "learning_rate": 0.000423861852433281,
      "loss": 2.332,
      "step": 270
    },
    {
      "epoch": 0.04321690959808274,
      "grad_norm": 0.23471523821353912,
      "learning_rate": 0.0004317111459968603,
      "loss": 2.2363,
      "step": 275
    },
    {
      "epoch": 0.04400267159077515,
      "grad_norm": 0.1952430009841919,
      "learning_rate": 0.00043956043956043956,
      "loss": 2.2182,
      "step": 280
    },
    {
      "epoch": 0.04478843358346757,
      "grad_norm": 0.23113080859184265,
      "learning_rate": 0.00044740973312401886,
      "loss": 2.2092,
      "step": 285
    },
    {
      "epoch": 0.04557419557615998,
      "grad_norm": 0.18108302354812622,
      "learning_rate": 0.0004552590266875981,
      "loss": 2.2655,
      "step": 290
    },
    {
      "epoch": 0.046359957568852396,
      "grad_norm": 0.15540535748004913,
      "learning_rate": 0.0004631083202511774,
      "loss": 2.1482,
      "step": 295
    },
    {
      "epoch": 0.047145719561544806,
      "grad_norm": 0.18835408985614777,
      "learning_rate": 0.00047095761381475666,
      "loss": 2.2144,
      "step": 300
    },
    {
      "epoch": 0.047931481554237224,
      "grad_norm": 0.2090383619070053,
      "learning_rate": 0.000478806907378336,
      "loss": 2.1736,
      "step": 305
    },
    {
      "epoch": 0.048717243546929634,
      "grad_norm": 0.2130228579044342,
      "learning_rate": 0.00048665620094191526,
      "loss": 2.2297,
      "step": 310
    },
    {
      "epoch": 0.04950300553962205,
      "grad_norm": 0.2176492065191269,
      "learning_rate": 0.0004945054945054945,
      "loss": 2.171,
      "step": 315
    },
    {
      "epoch": 0.05028876753231446,
      "grad_norm": 0.15854674577713013,
      "learning_rate": 0.0005023547880690738,
      "loss": 2.2807,
      "step": 320
    },
    {
      "epoch": 0.05107452952500688,
      "grad_norm": 0.2400057166814804,
      "learning_rate": 0.0005102040816326531,
      "loss": 2.2298,
      "step": 325
    },
    {
      "epoch": 0.05186029151769929,
      "grad_norm": 0.18969252705574036,
      "learning_rate": 0.0005180533751962324,
      "loss": 2.1822,
      "step": 330
    },
    {
      "epoch": 0.0526460535103917,
      "grad_norm": 0.35932809114456177,
      "learning_rate": 0.0005259026687598116,
      "loss": 2.2482,
      "step": 335
    },
    {
      "epoch": 0.05343181550308412,
      "grad_norm": 0.21439987421035767,
      "learning_rate": 0.000533751962323391,
      "loss": 2.2391,
      "step": 340
    },
    {
      "epoch": 0.05421757749577653,
      "grad_norm": 0.2086169719696045,
      "learning_rate": 0.0005416012558869702,
      "loss": 2.1001,
      "step": 345
    },
    {
      "epoch": 0.055003339488468944,
      "grad_norm": 0.20357146859169006,
      "learning_rate": 0.0005494505494505496,
      "loss": 2.1642,
      "step": 350
    },
    {
      "epoch": 0.055789101481161354,
      "grad_norm": 0.25846609473228455,
      "learning_rate": 0.0005572998430141287,
      "loss": 2.2399,
      "step": 355
    },
    {
      "epoch": 0.05657486347385377,
      "grad_norm": 0.22251342236995697,
      "learning_rate": 0.000565149136577708,
      "loss": 2.1899,
      "step": 360
    },
    {
      "epoch": 0.05736062546654618,
      "grad_norm": 0.15347051620483398,
      "learning_rate": 0.0005729984301412873,
      "loss": 2.202,
      "step": 365
    },
    {
      "epoch": 0.0581463874592386,
      "grad_norm": 0.2482958287000656,
      "learning_rate": 0.0005808477237048667,
      "loss": 2.181,
      "step": 370
    },
    {
      "epoch": 0.05893214945193101,
      "grad_norm": 0.18029803037643433,
      "learning_rate": 0.0005886970172684458,
      "loss": 2.1423,
      "step": 375
    },
    {
      "epoch": 0.05971791144462343,
      "grad_norm": 0.20792289078235626,
      "learning_rate": 0.0005965463108320251,
      "loss": 2.1015,
      "step": 380
    },
    {
      "epoch": 0.06050367343731584,
      "grad_norm": 0.1740165799856186,
      "learning_rate": 0.0006043956043956044,
      "loss": 2.0581,
      "step": 385
    },
    {
      "epoch": 0.06128943543000825,
      "grad_norm": 0.22941261529922485,
      "learning_rate": 0.0006122448979591838,
      "loss": 2.1631,
      "step": 390
    },
    {
      "epoch": 0.062075197422700665,
      "grad_norm": 0.1646764874458313,
      "learning_rate": 0.0006200941915227629,
      "loss": 2.142,
      "step": 395
    },
    {
      "epoch": 0.06286095941539308,
      "grad_norm": 0.2238229662179947,
      "learning_rate": 0.0006279434850863422,
      "loss": 2.1254,
      "step": 400
    },
    {
      "epoch": 0.06364672140808549,
      "grad_norm": 0.19127151370048523,
      "learning_rate": 0.0006357927786499215,
      "loss": 2.1019,
      "step": 405
    },
    {
      "epoch": 0.06443248340077791,
      "grad_norm": 0.19444602727890015,
      "learning_rate": 0.0006436420722135008,
      "loss": 2.0932,
      "step": 410
    },
    {
      "epoch": 0.06521824539347032,
      "grad_norm": 0.20162132382392883,
      "learning_rate": 0.0006514913657770801,
      "loss": 2.0709,
      "step": 415
    },
    {
      "epoch": 0.06600400738616273,
      "grad_norm": 0.24843716621398926,
      "learning_rate": 0.0006593406593406593,
      "loss": 2.0435,
      "step": 420
    },
    {
      "epoch": 0.06678976937885514,
      "grad_norm": 0.19505095481872559,
      "learning_rate": 0.0006671899529042387,
      "loss": 2.1971,
      "step": 425
    },
    {
      "epoch": 0.06757553137154756,
      "grad_norm": 0.16956081986427307,
      "learning_rate": 0.0006750392464678179,
      "loss": 2.0889,
      "step": 430
    },
    {
      "epoch": 0.06836129336423998,
      "grad_norm": 0.16742552816867828,
      "learning_rate": 0.0006828885400313972,
      "loss": 2.0928,
      "step": 435
    },
    {
      "epoch": 0.06914705535693239,
      "grad_norm": 0.20322011411190033,
      "learning_rate": 0.0006907378335949764,
      "loss": 2.0823,
      "step": 440
    },
    {
      "epoch": 0.0699328173496248,
      "grad_norm": 0.20075753331184387,
      "learning_rate": 0.0006985871271585558,
      "loss": 2.0312,
      "step": 445
    },
    {
      "epoch": 0.0707185793423172,
      "grad_norm": 0.26476067304611206,
      "learning_rate": 0.000706436420722135,
      "loss": 2.1047,
      "step": 450
    },
    {
      "epoch": 0.07150434133500963,
      "grad_norm": 0.1607220470905304,
      "learning_rate": 0.0007142857142857143,
      "loss": 2.0503,
      "step": 455
    },
    {
      "epoch": 0.07229010332770204,
      "grad_norm": 0.25718483328819275,
      "learning_rate": 0.0007221350078492935,
      "loss": 2.0814,
      "step": 460
    },
    {
      "epoch": 0.07307586532039445,
      "grad_norm": 0.19512560963630676,
      "learning_rate": 0.0007299843014128729,
      "loss": 2.0276,
      "step": 465
    },
    {
      "epoch": 0.07386162731308686,
      "grad_norm": 0.5019248127937317,
      "learning_rate": 0.0007378335949764521,
      "loss": 1.9824,
      "step": 470
    },
    {
      "epoch": 0.07464738930577929,
      "grad_norm": 0.21145927906036377,
      "learning_rate": 0.0007456828885400314,
      "loss": 2.0159,
      "step": 475
    },
    {
      "epoch": 0.0754331512984717,
      "grad_norm": 0.19788451492786407,
      "learning_rate": 0.0007535321821036107,
      "loss": 2.1007,
      "step": 480
    },
    {
      "epoch": 0.0762189132911641,
      "grad_norm": 0.1825549602508545,
      "learning_rate": 0.00076138147566719,
      "loss": 2.0288,
      "step": 485
    },
    {
      "epoch": 0.07700467528385652,
      "grad_norm": 0.1614513248205185,
      "learning_rate": 0.0007692307692307693,
      "loss": 2.0782,
      "step": 490
    },
    {
      "epoch": 0.07779043727654893,
      "grad_norm": 0.15320807695388794,
      "learning_rate": 0.0007770800627943485,
      "loss": 1.9095,
      "step": 495
    },
    {
      "epoch": 0.07857619926924135,
      "grad_norm": 0.18029426038265228,
      "learning_rate": 0.0007849293563579278,
      "loss": 2.0567,
      "step": 500
    },
    {
      "epoch": 0.07936196126193376,
      "grad_norm": 0.23825803399085999,
      "learning_rate": 0.0007927786499215071,
      "loss": 1.9964,
      "step": 505
    },
    {
      "epoch": 0.08014772325462617,
      "grad_norm": 0.18158216774463654,
      "learning_rate": 0.0008006279434850864,
      "loss": 1.9442,
      "step": 510
    },
    {
      "epoch": 0.08093348524731858,
      "grad_norm": 0.22512096166610718,
      "learning_rate": 0.0008084772370486656,
      "loss": 2.0562,
      "step": 515
    },
    {
      "epoch": 0.081719247240011,
      "grad_norm": 0.19693829119205475,
      "learning_rate": 0.0008163265306122449,
      "loss": 2.0231,
      "step": 520
    },
    {
      "epoch": 0.08250500923270342,
      "grad_norm": 0.17301593720912933,
      "learning_rate": 0.0008241758241758242,
      "loss": 1.9871,
      "step": 525
    },
    {
      "epoch": 0.08329077122539583,
      "grad_norm": 0.22018562257289886,
      "learning_rate": 0.0008320251177394035,
      "loss": 2.0008,
      "step": 530
    },
    {
      "epoch": 0.08407653321808824,
      "grad_norm": 0.23574061691761017,
      "learning_rate": 0.0008398744113029827,
      "loss": 1.9259,
      "step": 535
    },
    {
      "epoch": 0.08486229521078066,
      "grad_norm": 0.16205668449401855,
      "learning_rate": 0.000847723704866562,
      "loss": 1.9598,
      "step": 540
    },
    {
      "epoch": 0.08564805720347307,
      "grad_norm": 0.19473537802696228,
      "learning_rate": 0.0008555729984301414,
      "loss": 1.9231,
      "step": 545
    },
    {
      "epoch": 0.08643381919616548,
      "grad_norm": 0.205925852060318,
      "learning_rate": 0.0008634222919937206,
      "loss": 1.9964,
      "step": 550
    },
    {
      "epoch": 0.08721958118885789,
      "grad_norm": 0.27399733662605286,
      "learning_rate": 0.0008712715855572999,
      "loss": 1.9949,
      "step": 555
    },
    {
      "epoch": 0.0880053431815503,
      "grad_norm": 0.18213149905204773,
      "learning_rate": 0.0008791208791208791,
      "loss": 1.9875,
      "step": 560
    },
    {
      "epoch": 0.08879110517424273,
      "grad_norm": 0.27155083417892456,
      "learning_rate": 0.0008869701726844585,
      "loss": 1.9682,
      "step": 565
    },
    {
      "epoch": 0.08957686716693514,
      "grad_norm": 0.163107231259346,
      "learning_rate": 0.0008948194662480377,
      "loss": 1.9654,
      "step": 570
    },
    {
      "epoch": 0.09036262915962755,
      "grad_norm": 0.31603917479515076,
      "learning_rate": 0.000902668759811617,
      "loss": 1.9086,
      "step": 575
    },
    {
      "epoch": 0.09114839115231996,
      "grad_norm": 0.3522120416164398,
      "learning_rate": 0.0009105180533751962,
      "loss": 2.0186,
      "step": 580
    },
    {
      "epoch": 0.09193415314501238,
      "grad_norm": 0.16749052703380585,
      "learning_rate": 0.0009183673469387756,
      "loss": 1.9922,
      "step": 585
    },
    {
      "epoch": 0.09271991513770479,
      "grad_norm": 0.18135912716388702,
      "learning_rate": 0.0009262166405023548,
      "loss": 2.018,
      "step": 590
    },
    {
      "epoch": 0.0935056771303972,
      "grad_norm": 0.26006466150283813,
      "learning_rate": 0.0009340659340659341,
      "loss": 1.9165,
      "step": 595
    },
    {
      "epoch": 0.09429143912308961,
      "grad_norm": 0.1768893599510193,
      "learning_rate": 0.0009419152276295133,
      "loss": 1.9371,
      "step": 600
    },
    {
      "epoch": 0.09507720111578202,
      "grad_norm": 0.2588590681552887,
      "learning_rate": 0.0009497645211930927,
      "loss": 1.9935,
      "step": 605
    },
    {
      "epoch": 0.09586296310847445,
      "grad_norm": 0.20219220221042633,
      "learning_rate": 0.000957613814756672,
      "loss": 1.9188,
      "step": 610
    },
    {
      "epoch": 0.09664872510116686,
      "grad_norm": 0.21311190724372864,
      "learning_rate": 0.0009654631083202512,
      "loss": 1.9969,
      "step": 615
    },
    {
      "epoch": 0.09743448709385927,
      "grad_norm": 0.2351716160774231,
      "learning_rate": 0.0009733124018838305,
      "loss": 1.9984,
      "step": 620
    },
    {
      "epoch": 0.09822024908655168,
      "grad_norm": 0.1859682947397232,
      "learning_rate": 0.0009811616954474097,
      "loss": 1.9381,
      "step": 625
    },
    {
      "epoch": 0.0990060110792441,
      "grad_norm": 0.15183129906654358,
      "learning_rate": 0.000989010989010989,
      "loss": 1.9573,
      "step": 630
    },
    {
      "epoch": 0.09979177307193651,
      "grad_norm": 0.25219467282295227,
      "learning_rate": 0.0009968602825745684,
      "loss": 1.9229,
      "step": 635
    },
    {
      "epoch": 0.10057753506462892,
      "grad_norm": 0.1858813762664795,
      "learning_rate": 0.0009999993227024916,
      "loss": 1.9226,
      "step": 640
    },
    {
      "epoch": 0.10136329705732133,
      "grad_norm": 0.16762517392635345,
      "learning_rate": 0.0009999951836688061,
      "loss": 1.9119,
      "step": 645
    },
    {
      "epoch": 0.10214905905001376,
      "grad_norm": 0.28066954016685486,
      "learning_rate": 0.00099998728190894,
      "loss": 2.0073,
      "step": 650
    },
    {
      "epoch": 0.10293482104270617,
      "grad_norm": 0.21832288801670074,
      "learning_rate": 0.0009999756174823573,
      "loss": 1.9843,
      "step": 655
    },
    {
      "epoch": 0.10372058303539858,
      "grad_norm": 0.1832871437072754,
      "learning_rate": 0.0009999601904768399,
      "loss": 1.8482,
      "step": 660
    },
    {
      "epoch": 0.10450634502809099,
      "grad_norm": 0.18006999790668488,
      "learning_rate": 0.0009999410010084833,
      "loss": 1.846,
      "step": 665
    },
    {
      "epoch": 0.1052921070207834,
      "grad_norm": 0.18294790387153625,
      "learning_rate": 0.0009999180492216988,
      "loss": 1.9233,
      "step": 670
    },
    {
      "epoch": 0.10607786901347582,
      "grad_norm": 0.17135338485240936,
      "learning_rate": 0.0009998913352892106,
      "loss": 1.8711,
      "step": 675
    },
    {
      "epoch": 0.10686363100616823,
      "grad_norm": 0.1738177239894867,
      "learning_rate": 0.0009998608594120547,
      "loss": 1.872,
      "step": 680
    },
    {
      "epoch": 0.10764939299886064,
      "grad_norm": 0.16996446251869202,
      "learning_rate": 0.0009998266218195786,
      "loss": 1.9671,
      "step": 685
    },
    {
      "epoch": 0.10843515499155305,
      "grad_norm": 0.17967042326927185,
      "learning_rate": 0.000999788622769438,
      "loss": 1.8942,
      "step": 690
    },
    {
      "epoch": 0.10922091698424548,
      "grad_norm": 0.2092641294002533,
      "learning_rate": 0.0009997468625475953,
      "loss": 1.919,
      "step": 695
    },
    {
      "epoch": 0.11000667897693789,
      "grad_norm": 0.16578802466392517,
      "learning_rate": 0.0009997013414683184,
      "loss": 1.8921,
      "step": 700
    },
    {
      "epoch": 0.1107924409696303,
      "grad_norm": 0.15281111001968384,
      "learning_rate": 0.0009996520598741774,
      "loss": 1.9357,
      "step": 705
    },
    {
      "epoch": 0.11157820296232271,
      "grad_norm": 0.16658322513103485,
      "learning_rate": 0.000999599018136042,
      "loss": 1.97,
      "step": 710
    },
    {
      "epoch": 0.11236396495501512,
      "grad_norm": 0.18053396046161652,
      "learning_rate": 0.0009995422166530791,
      "loss": 1.8677,
      "step": 715
    },
    {
      "epoch": 0.11314972694770754,
      "grad_norm": 0.1864767223596573,
      "learning_rate": 0.0009994816558527497,
      "loss": 1.8493,
      "step": 720
    },
    {
      "epoch": 0.11393548894039995,
      "grad_norm": 0.27378034591674805,
      "learning_rate": 0.000999417336190806,
      "loss": 1.9325,
      "step": 725
    },
    {
      "epoch": 0.11472125093309236,
      "grad_norm": 0.29733896255493164,
      "learning_rate": 0.0009993492581512864,
      "loss": 1.8759,
      "step": 730
    },
    {
      "epoch": 0.11550701292578477,
      "grad_norm": 0.25246742367744446,
      "learning_rate": 0.0009992774222465147,
      "loss": 1.9363,
      "step": 735
    },
    {
      "epoch": 0.1162927749184772,
      "grad_norm": 0.17535606026649475,
      "learning_rate": 0.0009992018290170932,
      "loss": 1.848,
      "step": 740
    },
    {
      "epoch": 0.11707853691116961,
      "grad_norm": 0.16660931706428528,
      "learning_rate": 0.0009991224790319008,
      "loss": 1.8419,
      "step": 745
    },
    {
      "epoch": 0.11786429890386202,
      "grad_norm": 0.18775136768817902,
      "learning_rate": 0.000999039372888088,
      "loss": 1.8606,
      "step": 750
    },
    {
      "epoch": 0.11865006089655443,
      "grad_norm": 0.16369999945163727,
      "learning_rate": 0.000998952511211072,
      "loss": 1.9784,
      "step": 755
    },
    {
      "epoch": 0.11943582288924685,
      "grad_norm": 0.18088851869106293,
      "learning_rate": 0.0009988618946545325,
      "loss": 1.8971,
      "step": 760
    },
    {
      "epoch": 0.12022158488193926,
      "grad_norm": 0.15194383263587952,
      "learning_rate": 0.0009987675239004066,
      "loss": 1.8897,
      "step": 765
    },
    {
      "epoch": 0.12100734687463167,
      "grad_norm": 0.16550414264202118,
      "learning_rate": 0.0009986693996588837,
      "loss": 1.9305,
      "step": 770
    },
    {
      "epoch": 0.12179310886732408,
      "grad_norm": 0.1574241816997528,
      "learning_rate": 0.0009985675226684004,
      "loss": 1.905,
      "step": 775
    },
    {
      "epoch": 0.1225788708600165,
      "grad_norm": 0.1789279580116272,
      "learning_rate": 0.0009984618936956344,
      "loss": 1.9603,
      "step": 780
    },
    {
      "epoch": 0.12336463285270892,
      "grad_norm": 0.17177298665046692,
      "learning_rate": 0.0009983525135354994,
      "loss": 1.9059,
      "step": 785
    },
    {
      "epoch": 0.12415039484540133,
      "grad_norm": 1.0995100736618042,
      "learning_rate": 0.000998239383011138,
      "loss": 1.8245,
      "step": 790
    },
    {
      "epoch": 0.12493615683809374,
      "grad_norm": 0.1880197674036026,
      "learning_rate": 0.0009981225029739172,
      "loss": 1.8176,
      "step": 795
    },
    {
      "epoch": 0.12572191883078615,
      "grad_norm": 0.21664464473724365,
      "learning_rate": 0.0009980018743034208,
      "loss": 1.8941,
      "step": 800
    },
    {
      "epoch": 0.12650768082347857,
      "grad_norm": 0.18087147176265717,
      "learning_rate": 0.0009978774979074422,
      "loss": 1.8696,
      "step": 805
    },
    {
      "epoch": 0.12729344281617097,
      "grad_norm": 0.31465378403663635,
      "learning_rate": 0.0009977493747219794,
      "loss": 1.8643,
      "step": 810
    },
    {
      "epoch": 0.1280792048088634,
      "grad_norm": 0.15885500609874725,
      "learning_rate": 0.000997617505711227,
      "loss": 1.8692,
      "step": 815
    },
    {
      "epoch": 0.12886496680155582,
      "grad_norm": 0.19823002815246582,
      "learning_rate": 0.0009974818918675678,
      "loss": 1.8476,
      "step": 820
    },
    {
      "epoch": 0.12965072879424822,
      "grad_norm": 0.33955878019332886,
      "learning_rate": 0.0009973425342115678,
      "loss": 1.8371,
      "step": 825
    },
    {
      "epoch": 0.13043649078694064,
      "grad_norm": 0.15947683155536652,
      "learning_rate": 0.0009971994337919662,
      "loss": 1.8525,
      "step": 830
    },
    {
      "epoch": 0.13122225277963304,
      "grad_norm": 0.22582250833511353,
      "learning_rate": 0.0009970525916856686,
      "loss": 1.8344,
      "step": 835
    },
    {
      "epoch": 0.13200801477232546,
      "grad_norm": 0.16655315458774567,
      "learning_rate": 0.0009969020089977392,
      "loss": 1.8229,
      "step": 840
    },
    {
      "epoch": 0.13279377676501788,
      "grad_norm": 0.22109279036521912,
      "learning_rate": 0.0009967476868613916,
      "loss": 1.8169,
      "step": 845
    },
    {
      "epoch": 0.13357953875771028,
      "grad_norm": 0.23577183485031128,
      "learning_rate": 0.0009965896264379811,
      "loss": 1.8485,
      "step": 850
    },
    {
      "epoch": 0.1343653007504027,
      "grad_norm": 0.16854895651340485,
      "learning_rate": 0.000996427828916995,
      "loss": 1.8947,
      "step": 855
    },
    {
      "epoch": 0.13515106274309513,
      "grad_norm": 0.2264672815799713,
      "learning_rate": 0.000996262295516045,
      "loss": 1.8915,
      "step": 860
    },
    {
      "epoch": 0.13593682473578753,
      "grad_norm": 0.22935688495635986,
      "learning_rate": 0.0009960930274808563,
      "loss": 1.836,
      "step": 865
    },
    {
      "epoch": 0.13672258672847995,
      "grad_norm": 0.15960639715194702,
      "learning_rate": 0.0009959200260852602,
      "loss": 1.8687,
      "step": 870
    },
    {
      "epoch": 0.13750834872117235,
      "grad_norm": 0.17692284286022186,
      "learning_rate": 0.0009957432926311824,
      "loss": 1.8443,
      "step": 875
    },
    {
      "epoch": 0.13829411071386477,
      "grad_norm": 0.190086230635643,
      "learning_rate": 0.0009955628284486356,
      "loss": 1.8876,
      "step": 880
    },
    {
      "epoch": 0.1390798727065572,
      "grad_norm": 0.16075825691223145,
      "learning_rate": 0.000995378634895707,
      "loss": 1.8239,
      "step": 885
    },
    {
      "epoch": 0.1398656346992496,
      "grad_norm": 0.16710132360458374,
      "learning_rate": 0.0009951907133585501,
      "loss": 1.8647,
      "step": 890
    },
    {
      "epoch": 0.14065139669194202,
      "grad_norm": 0.16804201900959015,
      "learning_rate": 0.000994999065251373,
      "loss": 1.7629,
      "step": 895
    },
    {
      "epoch": 0.1414371586846344,
      "grad_norm": 0.1883011907339096,
      "learning_rate": 0.0009948036920164282,
      "loss": 1.8401,
      "step": 900
    },
    {
      "epoch": 0.14222292067732684,
      "grad_norm": 0.1541757583618164,
      "learning_rate": 0.0009946045951240016,
      "loss": 1.8263,
      "step": 905
    },
    {
      "epoch": 0.14300868267001926,
      "grad_norm": 0.1607087254524231,
      "learning_rate": 0.000994401776072402,
      "loss": 1.8431,
      "step": 910
    },
    {
      "epoch": 0.14379444466271166,
      "grad_norm": 0.155876025557518,
      "learning_rate": 0.0009941952363879496,
      "loss": 1.835,
      "step": 915
    },
    {
      "epoch": 0.14458020665540408,
      "grad_norm": 0.1786189079284668,
      "learning_rate": 0.0009939849776249635,
      "loss": 1.8537,
      "step": 920
    },
    {
      "epoch": 0.1453659686480965,
      "grad_norm": 0.1885564625263214,
      "learning_rate": 0.0009937710013657513,
      "loss": 1.8781,
      "step": 925
    },
    {
      "epoch": 0.1461517306407889,
      "grad_norm": 0.20302869379520416,
      "learning_rate": 0.000993553309220597,
      "loss": 1.7989,
      "step": 930
    },
    {
      "epoch": 0.14693749263348133,
      "grad_norm": 0.15116022527217865,
      "learning_rate": 0.0009933319028277475,
      "loss": 1.7587,
      "step": 935
    },
    {
      "epoch": 0.14772325462617372,
      "grad_norm": 0.20895975828170776,
      "learning_rate": 0.0009931067838534029,
      "loss": 1.901,
      "step": 940
    },
    {
      "epoch": 0.14850901661886615,
      "grad_norm": 0.19565752148628235,
      "learning_rate": 0.000992877953991701,
      "loss": 1.8388,
      "step": 945
    },
    {
      "epoch": 0.14929477861155857,
      "grad_norm": 0.233185276389122,
      "learning_rate": 0.0009926454149647072,
      "loss": 1.8021,
      "step": 950
    },
    {
      "epoch": 0.15008054060425097,
      "grad_norm": 0.19039976596832275,
      "learning_rate": 0.0009924091685223995,
      "loss": 1.8105,
      "step": 955
    },
    {
      "epoch": 0.1508663025969434,
      "grad_norm": 0.15231332182884216,
      "learning_rate": 0.0009921692164426561,
      "loss": 1.8455,
      "step": 960
    },
    {
      "epoch": 0.1516520645896358,
      "grad_norm": 0.16343708336353302,
      "learning_rate": 0.0009919255605312428,
      "loss": 1.9254,
      "step": 965
    },
    {
      "epoch": 0.1524378265823282,
      "grad_norm": 0.19997046887874603,
      "learning_rate": 0.0009916782026217976,
      "loss": 1.7809,
      "step": 970
    },
    {
      "epoch": 0.15322358857502064,
      "grad_norm": 0.15831786394119263,
      "learning_rate": 0.0009914271445758192,
      "loss": 1.8139,
      "step": 975
    },
    {
      "epoch": 0.15400935056771303,
      "grad_norm": 0.23234544694423676,
      "learning_rate": 0.000991172388282651,
      "loss": 1.8541,
      "step": 980
    },
    {
      "epoch": 0.15479511256040546,
      "grad_norm": 0.16235429048538208,
      "learning_rate": 0.0009909139356594678,
      "loss": 1.8377,
      "step": 985
    },
    {
      "epoch": 0.15558087455309785,
      "grad_norm": 0.1860172301530838,
      "learning_rate": 0.000990651788651261,
      "loss": 1.8644,
      "step": 990
    },
    {
      "epoch": 0.15636663654579028,
      "grad_norm": 0.18624064326286316,
      "learning_rate": 0.0009903859492308247,
      "loss": 1.8254,
      "step": 995
    },
    {
      "epoch": 0.1571523985384827,
      "grad_norm": 0.2140810191631317,
      "learning_rate": 0.00099011641939874,
      "loss": 1.7965,
      "step": 1000
    },
    {
      "epoch": 0.1579381605311751,
      "grad_norm": 0.17189514636993408,
      "learning_rate": 0.0009898432011833601,
      "loss": 1.7997,
      "step": 1005
    },
    {
      "epoch": 0.15872392252386752,
      "grad_norm": 0.16160257160663605,
      "learning_rate": 0.000989566296640796,
      "loss": 1.7915,
      "step": 1010
    },
    {
      "epoch": 0.15950968451655995,
      "grad_norm": 0.15080223977565765,
      "learning_rate": 0.0009892857078548994,
      "loss": 1.8035,
      "step": 1015
    },
    {
      "epoch": 0.16029544650925234,
      "grad_norm": 0.20072196424007416,
      "learning_rate": 0.0009890014369372483,
      "loss": 1.7741,
      "step": 1020
    },
    {
      "epoch": 0.16108120850194477,
      "grad_norm": 0.16227015852928162,
      "learning_rate": 0.0009887134860271302,
      "loss": 1.7972,
      "step": 1025
    },
    {
      "epoch": 0.16186697049463716,
      "grad_norm": 0.19584600627422333,
      "learning_rate": 0.0009884218572915272,
      "loss": 1.7683,
      "step": 1030
    },
    {
      "epoch": 0.1626527324873296,
      "grad_norm": 0.2638975977897644,
      "learning_rate": 0.0009881265529250985,
      "loss": 1.8308,
      "step": 1035
    },
    {
      "epoch": 0.163438494480022,
      "grad_norm": 0.15571711957454681,
      "learning_rate": 0.0009878275751501644,
      "loss": 1.8231,
      "step": 1040
    },
    {
      "epoch": 0.1642242564727144,
      "grad_norm": 0.2547709047794342,
      "learning_rate": 0.0009875249262166898,
      "loss": 1.7788,
      "step": 1045
    },
    {
      "epoch": 0.16501001846540683,
      "grad_norm": 0.2403130978345871,
      "learning_rate": 0.0009872186084022663,
      "loss": 1.8067,
      "step": 1050
    },
    {
      "epoch": 0.16579578045809923,
      "grad_norm": 0.16266390681266785,
      "learning_rate": 0.0009869086240120967,
      "loss": 1.8301,
      "step": 1055
    },
    {
      "epoch": 0.16658154245079165,
      "grad_norm": 0.18285565078258514,
      "learning_rate": 0.0009865949753789759,
      "loss": 1.7974,
      "step": 1060
    },
    {
      "epoch": 0.16736730444348408,
      "grad_norm": 0.1687275618314743,
      "learning_rate": 0.0009862776648632745,
      "loss": 1.7562,
      "step": 1065
    },
    {
      "epoch": 0.16815306643617647,
      "grad_norm": 0.1526845097541809,
      "learning_rate": 0.000985956694852921,
      "loss": 1.7842,
      "step": 1070
    },
    {
      "epoch": 0.1689388284288689,
      "grad_norm": 0.22343283891677856,
      "learning_rate": 0.000985632067763383,
      "loss": 1.8555,
      "step": 1075
    },
    {
      "epoch": 0.16972459042156132,
      "grad_norm": 0.15908358991146088,
      "learning_rate": 0.0009853037860376496,
      "loss": 1.7357,
      "step": 1080
    },
    {
      "epoch": 0.17051035241425372,
      "grad_norm": 0.19639401137828827,
      "learning_rate": 0.0009849718521462133,
      "loss": 1.7991,
      "step": 1085
    },
    {
      "epoch": 0.17129611440694614,
      "grad_norm": 0.23364387452602386,
      "learning_rate": 0.0009846362685870505,
      "loss": 1.8042,
      "step": 1090
    },
    {
      "epoch": 0.17208187639963854,
      "grad_norm": 0.1824749857187271,
      "learning_rate": 0.0009842970378856042,
      "loss": 1.8007,
      "step": 1095
    },
    {
      "epoch": 0.17286763839233096,
      "grad_norm": 0.17124874889850616,
      "learning_rate": 0.000983954162594763,
      "loss": 1.8384,
      "step": 1100
    },
    {
      "epoch": 0.1736534003850234,
      "grad_norm": 0.16191472113132477,
      "learning_rate": 0.0009836076452948434,
      "loss": 1.826,
      "step": 1105
    },
    {
      "epoch": 0.17443916237771578,
      "grad_norm": 0.315621942281723,
      "learning_rate": 0.0009832574885935703,
      "loss": 1.7842,
      "step": 1110
    },
    {
      "epoch": 0.1752249243704082,
      "grad_norm": 0.18572475016117096,
      "learning_rate": 0.0009829036951260566,
      "loss": 1.7614,
      "step": 1115
    },
    {
      "epoch": 0.1760106863631006,
      "grad_norm": 0.1333106905221939,
      "learning_rate": 0.0009825462675547834,
      "loss": 1.7805,
      "step": 1120
    },
    {
      "epoch": 0.17679644835579303,
      "grad_norm": 0.1603097766637802,
      "learning_rate": 0.0009821852085695813,
      "loss": 1.7832,
      "step": 1125
    },
    {
      "epoch": 0.17758221034848545,
      "grad_norm": 0.21313336491584778,
      "learning_rate": 0.0009818205208876082,
      "loss": 1.8232,
      "step": 1130
    },
    {
      "epoch": 0.17836797234117785,
      "grad_norm": 0.23421534895896912,
      "learning_rate": 0.0009814522072533309,
      "loss": 1.7626,
      "step": 1135
    },
    {
      "epoch": 0.17915373433387027,
      "grad_norm": 0.16554996371269226,
      "learning_rate": 0.0009810802704385022,
      "loss": 1.8293,
      "step": 1140
    },
    {
      "epoch": 0.1799394963265627,
      "grad_norm": 0.1454770267009735,
      "learning_rate": 0.0009807047132421422,
      "loss": 1.9289,
      "step": 1145
    },
    {
      "epoch": 0.1807252583192551,
      "grad_norm": 0.21001407504081726,
      "learning_rate": 0.0009803255384905158,
      "loss": 1.7769,
      "step": 1150
    },
    {
      "epoch": 0.18151102031194752,
      "grad_norm": 0.2452252209186554,
      "learning_rate": 0.000979942749037112,
      "loss": 1.7492,
      "step": 1155
    },
    {
      "epoch": 0.18229678230463991,
      "grad_norm": 0.22114233672618866,
      "learning_rate": 0.0009795563477626224,
      "loss": 1.8087,
      "step": 1160
    },
    {
      "epoch": 0.18308254429733234,
      "grad_norm": 0.1857612133026123,
      "learning_rate": 0.0009791663375749195,
      "loss": 1.7559,
      "step": 1165
    },
    {
      "epoch": 0.18386830629002476,
      "grad_norm": 0.19770672917366028,
      "learning_rate": 0.0009787727214090344,
      "loss": 1.7924,
      "step": 1170
    },
    {
      "epoch": 0.18465406828271716,
      "grad_norm": 0.1833062618970871,
      "learning_rate": 0.0009783755022271356,
      "loss": 1.8486,
      "step": 1175
    },
    {
      "epoch": 0.18543983027540958,
      "grad_norm": 0.2029699981212616,
      "learning_rate": 0.0009779746830185056,
      "loss": 1.809,
      "step": 1180
    },
    {
      "epoch": 0.18622559226810198,
      "grad_norm": 0.1918937712907791,
      "learning_rate": 0.0009775702667995196,
      "loss": 1.7214,
      "step": 1185
    },
    {
      "epoch": 0.1870113542607944,
      "grad_norm": 0.15816162526607513,
      "learning_rate": 0.000977162256613622,
      "loss": 1.7447,
      "step": 1190
    },
    {
      "epoch": 0.18779711625348683,
      "grad_norm": 0.17059017717838287,
      "learning_rate": 0.0009767506555313033,
      "loss": 1.7853,
      "step": 1195
    },
    {
      "epoch": 0.18858287824617923,
      "grad_norm": 0.15515488386154175,
      "learning_rate": 0.0009763354666500778,
      "loss": 1.8019,
      "step": 1200
    },
    {
      "epoch": 0.18936864023887165,
      "grad_norm": 0.17095063626766205,
      "learning_rate": 0.0009759166930944596,
      "loss": 1.7827,
      "step": 1205
    },
    {
      "epoch": 0.19015440223156405,
      "grad_norm": 0.20768775045871735,
      "learning_rate": 0.0009754943380159398,
      "loss": 1.7911,
      "step": 1210
    },
    {
      "epoch": 0.19094016422425647,
      "grad_norm": 0.2060774266719818,
      "learning_rate": 0.0009750684045929618,
      "loss": 1.7821,
      "step": 1215
    },
    {
      "epoch": 0.1917259262169489,
      "grad_norm": 0.1782405525445938,
      "learning_rate": 0.0009746388960308982,
      "loss": 1.7665,
      "step": 1220
    },
    {
      "epoch": 0.1925116882096413,
      "grad_norm": 0.17600089311599731,
      "learning_rate": 0.0009742058155620266,
      "loss": 1.7827,
      "step": 1225
    },
    {
      "epoch": 0.19329745020233371,
      "grad_norm": 0.14907889068126678,
      "learning_rate": 0.0009737691664455045,
      "loss": 1.7102,
      "step": 1230
    },
    {
      "epoch": 0.19408321219502614,
      "grad_norm": 0.16940507292747498,
      "learning_rate": 0.0009733289519673459,
      "loss": 1.7302,
      "step": 1235
    },
    {
      "epoch": 0.19486897418771854,
      "grad_norm": 0.17307011783123016,
      "learning_rate": 0.0009728851754403955,
      "loss": 1.8605,
      "step": 1240
    },
    {
      "epoch": 0.19565473618041096,
      "grad_norm": 0.1481655091047287,
      "learning_rate": 0.0009724378402043049,
      "loss": 1.8172,
      "step": 1245
    },
    {
      "epoch": 0.19644049817310336,
      "grad_norm": 0.16137413680553436,
      "learning_rate": 0.0009719869496255065,
      "loss": 1.831,
      "step": 1250
    },
    {
      "epoch": 0.19722626016579578,
      "grad_norm": 0.14303524792194366,
      "learning_rate": 0.0009715325070971883,
      "loss": 1.7464,
      "step": 1255
    },
    {
      "epoch": 0.1980120221584882,
      "grad_norm": 0.14343352615833282,
      "learning_rate": 0.0009710745160392692,
      "loss": 1.7807,
      "step": 1260
    },
    {
      "epoch": 0.1987977841511806,
      "grad_norm": 0.14091046154499054,
      "learning_rate": 0.000970612979898372,
      "loss": 1.8111,
      "step": 1265
    },
    {
      "epoch": 0.19958354614387303,
      "grad_norm": 0.22847269475460052,
      "learning_rate": 0.0009701479021477986,
      "loss": 1.8072,
      "step": 1270
    },
    {
      "epoch": 0.20036930813656542,
      "grad_norm": 0.1330152153968811,
      "learning_rate": 0.000969679286287503,
      "loss": 1.7495,
      "step": 1275
    },
    {
      "epoch": 0.20115507012925785,
      "grad_norm": 0.16856741905212402,
      "learning_rate": 0.0009692071358440656,
      "loss": 1.7604,
      "step": 1280
    },
    {
      "epoch": 0.20194083212195027,
      "grad_norm": 0.16036243736743927,
      "learning_rate": 0.0009687314543706662,
      "loss": 1.7445,
      "step": 1285
    },
    {
      "epoch": 0.20272659411464267,
      "grad_norm": 0.1447385549545288,
      "learning_rate": 0.0009682522454470577,
      "loss": 1.7842,
      "step": 1290
    },
    {
      "epoch": 0.2035123561073351,
      "grad_norm": 0.19771301746368408,
      "learning_rate": 0.0009677695126795387,
      "loss": 1.7434,
      "step": 1295
    },
    {
      "epoch": 0.20429811810002751,
      "grad_norm": 0.17068088054656982,
      "learning_rate": 0.0009672832597009267,
      "loss": 1.7717,
      "step": 1300
    },
    {
      "epoch": 0.2050838800927199,
      "grad_norm": 0.17894063889980316,
      "learning_rate": 0.0009667934901705304,
      "loss": 1.6901,
      "step": 1305
    },
    {
      "epoch": 0.20586964208541234,
      "grad_norm": 0.12928830087184906,
      "learning_rate": 0.0009663002077741228,
      "loss": 1.7405,
      "step": 1310
    },
    {
      "epoch": 0.20665540407810473,
      "grad_norm": 0.18357087671756744,
      "learning_rate": 0.0009658034162239126,
      "loss": 1.748,
      "step": 1315
    },
    {
      "epoch": 0.20744116607079716,
      "grad_norm": 0.17646510899066925,
      "learning_rate": 0.000965303119258517,
      "loss": 1.7418,
      "step": 1320
    },
    {
      "epoch": 0.20822692806348958,
      "grad_norm": 0.14562882483005524,
      "learning_rate": 0.0009647993206429336,
      "loss": 1.734,
      "step": 1325
    },
    {
      "epoch": 0.20901269005618198,
      "grad_norm": 0.16555768251419067,
      "learning_rate": 0.0009642920241685109,
      "loss": 1.7481,
      "step": 1330
    },
    {
      "epoch": 0.2097984520488744,
      "grad_norm": 0.22527766227722168,
      "learning_rate": 0.0009637812336529214,
      "loss": 1.797,
      "step": 1335
    },
    {
      "epoch": 0.2105842140415668,
      "grad_norm": 0.2324640154838562,
      "learning_rate": 0.0009632669529401317,
      "loss": 1.7871,
      "step": 1340
    },
    {
      "epoch": 0.21136997603425922,
      "grad_norm": 0.18207474052906036,
      "learning_rate": 0.0009627491859003744,
      "loss": 1.7247,
      "step": 1345
    },
    {
      "epoch": 0.21215573802695165,
      "grad_norm": 0.21197597682476044,
      "learning_rate": 0.0009622279364301177,
      "loss": 1.7338,
      "step": 1350
    },
    {
      "epoch": 0.21294150001964404,
      "grad_norm": 0.15248093008995056,
      "learning_rate": 0.0009617032084520378,
      "loss": 1.6968,
      "step": 1355
    },
    {
      "epoch": 0.21372726201233647,
      "grad_norm": 0.15904445946216583,
      "learning_rate": 0.000961175005914988,
      "loss": 1.7187,
      "step": 1360
    },
    {
      "epoch": 0.2145130240050289,
      "grad_norm": 0.1640489548444748,
      "learning_rate": 0.0009606433327939697,
      "loss": 1.7054,
      "step": 1365
    },
    {
      "epoch": 0.2152987859977213,
      "grad_norm": 0.20775249600410461,
      "learning_rate": 0.0009601081930901024,
      "loss": 1.7597,
      "step": 1370
    },
    {
      "epoch": 0.2160845479904137,
      "grad_norm": 0.22332137823104858,
      "learning_rate": 0.0009595695908305927,
      "loss": 1.7071,
      "step": 1375
    },
    {
      "epoch": 0.2168703099831061,
      "grad_norm": 0.2273789495229721,
      "learning_rate": 0.0009590275300687057,
      "loss": 1.7413,
      "step": 1380
    },
    {
      "epoch": 0.21765607197579853,
      "grad_norm": 0.17828406393527985,
      "learning_rate": 0.0009584820148837331,
      "loss": 1.7059,
      "step": 1385
    },
    {
      "epoch": 0.21844183396849096,
      "grad_norm": 0.19548800587654114,
      "learning_rate": 0.0009579330493809629,
      "loss": 1.7752,
      "step": 1390
    },
    {
      "epoch": 0.21922759596118335,
      "grad_norm": 0.16092374920845032,
      "learning_rate": 0.0009573806376916486,
      "loss": 1.7438,
      "step": 1395
    },
    {
      "epoch": 0.22001335795387578,
      "grad_norm": 0.1342301070690155,
      "learning_rate": 0.0009568247839729782,
      "loss": 1.751,
      "step": 1400
    },
    {
      "epoch": 0.22079911994656817,
      "grad_norm": 0.1362721025943756,
      "learning_rate": 0.0009562654924080425,
      "loss": 1.717,
      "step": 1405
    },
    {
      "epoch": 0.2215848819392606,
      "grad_norm": 0.16024941205978394,
      "learning_rate": 0.0009557027672058043,
      "loss": 1.8112,
      "step": 1410
    },
    {
      "epoch": 0.22237064393195302,
      "grad_norm": 0.154706671833992,
      "learning_rate": 0.000955136612601066,
      "loss": 1.738,
      "step": 1415
    },
    {
      "epoch": 0.22315640592464542,
      "grad_norm": 0.1728077530860901,
      "learning_rate": 0.0009545670328544382,
      "loss": 1.7967,
      "step": 1420
    },
    {
      "epoch": 0.22394216791733784,
      "grad_norm": 0.15964308381080627,
      "learning_rate": 0.0009539940322523072,
      "loss": 1.8126,
      "step": 1425
    },
    {
      "epoch": 0.22472792991003024,
      "grad_norm": 0.1784270852804184,
      "learning_rate": 0.0009534176151068035,
      "loss": 1.7388,
      "step": 1430
    },
    {
      "epoch": 0.22551369190272266,
      "grad_norm": 0.18767036497592926,
      "learning_rate": 0.0009528377857557686,
      "loss": 1.6965,
      "step": 1435
    },
    {
      "epoch": 0.2262994538954151,
      "grad_norm": 0.21508271992206573,
      "learning_rate": 0.0009522545485627227,
      "loss": 1.7143,
      "step": 1440
    },
    {
      "epoch": 0.22708521588810748,
      "grad_norm": 0.30984237790107727,
      "learning_rate": 0.0009516679079168318,
      "loss": 1.6847,
      "step": 1445
    },
    {
      "epoch": 0.2278709778807999,
      "grad_norm": 0.12825290858745575,
      "learning_rate": 0.0009510778682328746,
      "loss": 1.749,
      "step": 1450
    },
    {
      "epoch": 0.22865673987349233,
      "grad_norm": 0.23394402861595154,
      "learning_rate": 0.0009504844339512095,
      "loss": 1.7347,
      "step": 1455
    },
    {
      "epoch": 0.22944250186618473,
      "grad_norm": 0.24663372337818146,
      "learning_rate": 0.0009498876095377409,
      "loss": 1.6935,
      "step": 1460
    },
    {
      "epoch": 0.23022826385887715,
      "grad_norm": 0.23490461707115173,
      "learning_rate": 0.0009492873994838858,
      "loss": 1.804,
      "step": 1465
    },
    {
      "epoch": 0.23101402585156955,
      "grad_norm": 0.20233745872974396,
      "learning_rate": 0.0009486838083065396,
      "loss": 1.7856,
      "step": 1470
    },
    {
      "epoch": 0.23179978784426197,
      "grad_norm": 0.18699239194393158,
      "learning_rate": 0.0009480768405480432,
      "loss": 1.6904,
      "step": 1475
    },
    {
      "epoch": 0.2325855498369544,
      "grad_norm": 0.27135419845581055,
      "learning_rate": 0.0009474665007761471,
      "loss": 1.755,
      "step": 1480
    },
    {
      "epoch": 0.2333713118296468,
      "grad_norm": 0.1718074530363083,
      "learning_rate": 0.0009468527935839787,
      "loss": 1.7456,
      "step": 1485
    },
    {
      "epoch": 0.23415707382233922,
      "grad_norm": 0.161013662815094,
      "learning_rate": 0.0009462357235900065,
      "loss": 1.7511,
      "step": 1490
    },
    {
      "epoch": 0.23494283581503161,
      "grad_norm": 0.20107895135879517,
      "learning_rate": 0.0009456152954380063,
      "loss": 1.7084,
      "step": 1495
    },
    {
      "epoch": 0.23572859780772404,
      "grad_norm": 0.14512185752391815,
      "learning_rate": 0.0009449915137970255,
      "loss": 1.7557,
      "step": 1500
    },
    {
      "epoch": 0.23651435980041646,
      "grad_norm": 0.1591528356075287,
      "learning_rate": 0.0009443643833613483,
      "loss": 1.7474,
      "step": 1505
    },
    {
      "epoch": 0.23730012179310886,
      "grad_norm": 0.22338952124118805,
      "learning_rate": 0.0009437339088504603,
      "loss": 1.7563,
      "step": 1510
    },
    {
      "epoch": 0.23808588378580128,
      "grad_norm": 0.1846311241388321,
      "learning_rate": 0.0009431000950090131,
      "loss": 1.7591,
      "step": 1515
    },
    {
      "epoch": 0.2388716457784937,
      "grad_norm": 0.1189197227358818,
      "learning_rate": 0.0009424629466067889,
      "loss": 1.7469,
      "step": 1520
    },
    {
      "epoch": 0.2396574077711861,
      "grad_norm": 0.19045637547969818,
      "learning_rate": 0.0009418224684386633,
      "loss": 1.6978,
      "step": 1525
    },
    {
      "epoch": 0.24044316976387853,
      "grad_norm": 0.1384202241897583,
      "learning_rate": 0.0009411786653245712,
      "loss": 1.7447,
      "step": 1530
    },
    {
      "epoch": 0.24122893175657092,
      "grad_norm": 0.19370493292808533,
      "learning_rate": 0.0009405315421094684,
      "loss": 1.777,
      "step": 1535
    },
    {
      "epoch": 0.24201469374926335,
      "grad_norm": 0.14141830801963806,
      "learning_rate": 0.0009398811036632973,
      "loss": 1.7378,
      "step": 1540
    },
    {
      "epoch": 0.24280045574195577,
      "grad_norm": 0.18611302971839905,
      "learning_rate": 0.000939227354880948,
      "loss": 1.7318,
      "step": 1545
    },
    {
      "epoch": 0.24358621773464817,
      "grad_norm": 0.18367739021778107,
      "learning_rate": 0.0009385703006822237,
      "loss": 1.7881,
      "step": 1550
    },
    {
      "epoch": 0.2443719797273406,
      "grad_norm": 0.17989496886730194,
      "learning_rate": 0.0009379099460118017,
      "loss": 1.7075,
      "step": 1555
    },
    {
      "epoch": 0.245157741720033,
      "grad_norm": 0.16730940341949463,
      "learning_rate": 0.0009372462958391978,
      "loss": 1.7174,
      "step": 1560
    },
    {
      "epoch": 0.24594350371272541,
      "grad_norm": 0.16663284599781036,
      "learning_rate": 0.0009365793551587278,
      "loss": 1.7412,
      "step": 1565
    },
    {
      "epoch": 0.24672926570541784,
      "grad_norm": 0.1993977576494217,
      "learning_rate": 0.0009359091289894702,
      "loss": 1.7547,
      "step": 1570
    },
    {
      "epoch": 0.24751502769811023,
      "grad_norm": 0.14924155175685883,
      "learning_rate": 0.000935235622375229,
      "loss": 1.6906,
      "step": 1575
    },
    {
      "epoch": 0.24830078969080266,
      "grad_norm": 0.1745729148387909,
      "learning_rate": 0.0009345588403844945,
      "loss": 1.6876,
      "step": 1580
    },
    {
      "epoch": 0.24908655168349506,
      "grad_norm": 0.17542210221290588,
      "learning_rate": 0.000933878788110407,
      "loss": 1.6807,
      "step": 1585
    },
    {
      "epoch": 0.24987231367618748,
      "grad_norm": 0.18530936539173126,
      "learning_rate": 0.0009331954706707162,
      "loss": 1.7146,
      "step": 1590
    },
    {
      "epoch": 0.2506580756688799,
      "grad_norm": 0.14683561027050018,
      "learning_rate": 0.0009325088932077448,
      "loss": 1.707,
      "step": 1595
    },
    {
      "epoch": 0.2514438376615723,
      "grad_norm": 0.16001202166080475,
      "learning_rate": 0.0009318190608883485,
      "loss": 1.7857,
      "step": 1600
    },
    {
      "epoch": 0.2522295996542647,
      "grad_norm": 0.16465690732002258,
      "learning_rate": 0.0009311259789038775,
      "loss": 1.6935,
      "step": 1605
    },
    {
      "epoch": 0.25301536164695715,
      "grad_norm": 0.20848460495471954,
      "learning_rate": 0.0009304296524701377,
      "loss": 1.7694,
      "step": 1610
    },
    {
      "epoch": 0.2538011236396496,
      "grad_norm": 0.14489056169986725,
      "learning_rate": 0.0009297300868273506,
      "loss": 1.7444,
      "step": 1615
    },
    {
      "epoch": 0.25458688563234194,
      "grad_norm": 0.14249326288700104,
      "learning_rate": 0.0009290272872401153,
      "loss": 1.7361,
      "step": 1620
    },
    {
      "epoch": 0.25537264762503437,
      "grad_norm": 0.15388713777065277,
      "learning_rate": 0.0009283212589973671,
      "loss": 1.7088,
      "step": 1625
    },
    {
      "epoch": 0.2561584096177268,
      "grad_norm": 0.21898041665554047,
      "learning_rate": 0.0009276120074123395,
      "loss": 1.792,
      "step": 1630
    },
    {
      "epoch": 0.2569441716104192,
      "grad_norm": 0.1465713083744049,
      "learning_rate": 0.0009268995378225228,
      "loss": 1.7808,
      "step": 1635
    },
    {
      "epoch": 0.25772993360311164,
      "grad_norm": 0.13897457718849182,
      "learning_rate": 0.0009261838555896245,
      "loss": 1.702,
      "step": 1640
    },
    {
      "epoch": 0.258515695595804,
      "grad_norm": 0.18026679754257202,
      "learning_rate": 0.000925464966099529,
      "loss": 1.7441,
      "step": 1645
    },
    {
      "epoch": 0.25930145758849643,
      "grad_norm": 0.13608209788799286,
      "learning_rate": 0.0009247428747622573,
      "loss": 1.6581,
      "step": 1650
    },
    {
      "epoch": 0.26008721958118886,
      "grad_norm": 0.21667583286762238,
      "learning_rate": 0.0009240175870119252,
      "loss": 1.7058,
      "step": 1655
    },
    {
      "epoch": 0.2608729815738813,
      "grad_norm": 0.19037890434265137,
      "learning_rate": 0.0009232891083067043,
      "loss": 1.7079,
      "step": 1660
    },
    {
      "epoch": 0.2616587435665737,
      "grad_norm": 0.14540046453475952,
      "learning_rate": 0.0009225574441287787,
      "loss": 1.7886,
      "step": 1665
    },
    {
      "epoch": 0.2624445055592661,
      "grad_norm": 0.15847225487232208,
      "learning_rate": 0.0009218225999843057,
      "loss": 1.7691,
      "step": 1670
    },
    {
      "epoch": 0.2632302675519585,
      "grad_norm": 0.1281496286392212,
      "learning_rate": 0.0009210845814033729,
      "loss": 1.7192,
      "step": 1675
    },
    {
      "epoch": 0.2640160295446509,
      "grad_norm": 0.14975865185260773,
      "learning_rate": 0.0009203433939399577,
      "loss": 1.6939,
      "step": 1680
    },
    {
      "epoch": 0.26480179153734335,
      "grad_norm": 0.23517081141471863,
      "learning_rate": 0.0009195990431718846,
      "loss": 1.7496,
      "step": 1685
    },
    {
      "epoch": 0.26558755353003577,
      "grad_norm": 0.24029028415679932,
      "learning_rate": 0.000918851534700784,
      "loss": 1.7212,
      "step": 1690
    },
    {
      "epoch": 0.26637331552272814,
      "grad_norm": 0.1643204241991043,
      "learning_rate": 0.0009181008741520494,
      "loss": 1.6901,
      "step": 1695
    },
    {
      "epoch": 0.26715907751542056,
      "grad_norm": 0.14907053112983704,
      "learning_rate": 0.0009173470671747952,
      "loss": 1.7005,
      "step": 1700
    },
    {
      "epoch": 0.267944839508113,
      "grad_norm": 0.14824387431144714,
      "learning_rate": 0.0009165901194418147,
      "loss": 1.7242,
      "step": 1705
    },
    {
      "epoch": 0.2687306015008054,
      "grad_norm": 0.15933027863502502,
      "learning_rate": 0.0009158300366495371,
      "loss": 1.7147,
      "step": 1710
    },
    {
      "epoch": 0.26951636349349783,
      "grad_norm": 0.15938429534435272,
      "learning_rate": 0.0009150668245179839,
      "loss": 1.755,
      "step": 1715
    },
    {
      "epoch": 0.27030212548619026,
      "grad_norm": 0.1357879489660263,
      "learning_rate": 0.0009143004887907273,
      "loss": 1.7154,
      "step": 1720
    },
    {
      "epoch": 0.2710878874788826,
      "grad_norm": 0.17757509648799896,
      "learning_rate": 0.0009135310352348458,
      "loss": 1.7224,
      "step": 1725
    },
    {
      "epoch": 0.27187364947157505,
      "grad_norm": 0.1572904735803604,
      "learning_rate": 0.0009127584696408813,
      "loss": 1.7091,
      "step": 1730
    },
    {
      "epoch": 0.2726594114642675,
      "grad_norm": 0.13565635681152344,
      "learning_rate": 0.0009119827978227952,
      "loss": 1.6385,
      "step": 1735
    },
    {
      "epoch": 0.2734451734569599,
      "grad_norm": 0.17699536681175232,
      "learning_rate": 0.0009112040256179253,
      "loss": 1.7315,
      "step": 1740
    },
    {
      "epoch": 0.2742309354496523,
      "grad_norm": 0.13635475933551788,
      "learning_rate": 0.0009104221588869407,
      "loss": 1.7882,
      "step": 1745
    },
    {
      "epoch": 0.2750166974423447,
      "grad_norm": 0.1719839721918106,
      "learning_rate": 0.0009096372035137995,
      "loss": 1.7376,
      "step": 1750
    },
    {
      "epoch": 0.2758024594350371,
      "grad_norm": 0.16986599564552307,
      "learning_rate": 0.0009088491654057025,
      "loss": 1.7442,
      "step": 1755
    },
    {
      "epoch": 0.27658822142772954,
      "grad_norm": 0.1694546341896057,
      "learning_rate": 0.0009080580504930503,
      "loss": 1.733,
      "step": 1760
    },
    {
      "epoch": 0.27737398342042197,
      "grad_norm": 0.16824652254581451,
      "learning_rate": 0.0009072638647293976,
      "loss": 1.726,
      "step": 1765
    },
    {
      "epoch": 0.2781597454131144,
      "grad_norm": 0.15378861129283905,
      "learning_rate": 0.0009064666140914093,
      "loss": 1.7501,
      "step": 1770
    },
    {
      "epoch": 0.27894550740580676,
      "grad_norm": 0.17473557591438293,
      "learning_rate": 0.0009056663045788147,
      "loss": 1.6769,
      "step": 1775
    },
    {
      "epoch": 0.2797312693984992,
      "grad_norm": 0.17376935482025146,
      "learning_rate": 0.000904862942214363,
      "loss": 1.6894,
      "step": 1780
    },
    {
      "epoch": 0.2805170313911916,
      "grad_norm": 0.1903456747531891,
      "learning_rate": 0.0009040565330437778,
      "loss": 1.6576,
      "step": 1785
    },
    {
      "epoch": 0.28130279338388403,
      "grad_norm": 0.21947069466114044,
      "learning_rate": 0.0009032470831357112,
      "loss": 1.7025,
      "step": 1790
    },
    {
      "epoch": 0.28208855537657646,
      "grad_norm": 0.12956298887729645,
      "learning_rate": 0.0009024345985816988,
      "loss": 1.7224,
      "step": 1795
    },
    {
      "epoch": 0.2828743173692688,
      "grad_norm": 0.17501172423362732,
      "learning_rate": 0.0009016190854961128,
      "loss": 1.7486,
      "step": 1800
    },
    {
      "epoch": 0.28366007936196125,
      "grad_norm": 0.13125917315483093,
      "learning_rate": 0.0009008005500161177,
      "loss": 1.7292,
      "step": 1805
    },
    {
      "epoch": 0.2844458413546537,
      "grad_norm": 0.13544058799743652,
      "learning_rate": 0.0008999789983016224,
      "loss": 1.7354,
      "step": 1810
    },
    {
      "epoch": 0.2852316033473461,
      "grad_norm": 0.18308532238006592,
      "learning_rate": 0.0008991544365352349,
      "loss": 1.6718,
      "step": 1815
    },
    {
      "epoch": 0.2860173653400385,
      "grad_norm": 0.1633603572845459,
      "learning_rate": 0.000898326870922215,
      "loss": 1.7,
      "step": 1820
    },
    {
      "epoch": 0.2868031273327309,
      "grad_norm": 0.1754382848739624,
      "learning_rate": 0.0008974963076904284,
      "loss": 1.7266,
      "step": 1825
    },
    {
      "epoch": 0.2875888893254233,
      "grad_norm": 0.2322840392589569,
      "learning_rate": 0.0008966627530902993,
      "loss": 1.6203,
      "step": 1830
    },
    {
      "epoch": 0.28837465131811574,
      "grad_norm": 0.1650850623846054,
      "learning_rate": 0.0008958262133947638,
      "loss": 1.7018,
      "step": 1835
    },
    {
      "epoch": 0.28916041331080816,
      "grad_norm": 0.18278241157531738,
      "learning_rate": 0.0008949866948992219,
      "loss": 1.7729,
      "step": 1840
    },
    {
      "epoch": 0.2899461753035006,
      "grad_norm": 0.1643027514219284,
      "learning_rate": 0.0008941442039214911,
      "loss": 1.7301,
      "step": 1845
    },
    {
      "epoch": 0.290731937296193,
      "grad_norm": 0.18282204866409302,
      "learning_rate": 0.000893298746801758,
      "loss": 1.7888,
      "step": 1850
    },
    {
      "epoch": 0.2915176992888854,
      "grad_norm": 0.16003000736236572,
      "learning_rate": 0.0008924503299025313,
      "loss": 1.7235,
      "step": 1855
    },
    {
      "epoch": 0.2923034612815778,
      "grad_norm": 0.16633334755897522,
      "learning_rate": 0.0008915989596085933,
      "loss": 1.7026,
      "step": 1860
    },
    {
      "epoch": 0.2930892232742702,
      "grad_norm": 0.29993849992752075,
      "learning_rate": 0.0008907446423269525,
      "loss": 1.7593,
      "step": 1865
    },
    {
      "epoch": 0.29387498526696265,
      "grad_norm": 0.15839068591594696,
      "learning_rate": 0.0008898873844867947,
      "loss": 1.6872,
      "step": 1870
    },
    {
      "epoch": 0.2946607472596551,
      "grad_norm": 0.21255196630954742,
      "learning_rate": 0.0008890271925394353,
      "loss": 1.8016,
      "step": 1875
    },
    {
      "epoch": 0.29544650925234744,
      "grad_norm": 0.1803053617477417,
      "learning_rate": 0.0008881640729582698,
      "loss": 1.7209,
      "step": 1880
    },
    {
      "epoch": 0.29623227124503987,
      "grad_norm": 0.20228391885757446,
      "learning_rate": 0.0008872980322387264,
      "loss": 1.6888,
      "step": 1885
    },
    {
      "epoch": 0.2970180332377323,
      "grad_norm": 0.1691243052482605,
      "learning_rate": 0.000886429076898216,
      "loss": 1.6535,
      "step": 1890
    },
    {
      "epoch": 0.2978037952304247,
      "grad_norm": 0.14024175703525543,
      "learning_rate": 0.0008855572134760834,
      "loss": 1.7665,
      "step": 1895
    },
    {
      "epoch": 0.29858955722311714,
      "grad_norm": 0.20091187953948975,
      "learning_rate": 0.0008846824485335587,
      "loss": 1.691,
      "step": 1900
    },
    {
      "epoch": 0.2993753192158095,
      "grad_norm": 0.12678097188472748,
      "learning_rate": 0.0008838047886537071,
      "loss": 1.6767,
      "step": 1905
    },
    {
      "epoch": 0.30016108120850193,
      "grad_norm": 0.14702141284942627,
      "learning_rate": 0.0008829242404413798,
      "loss": 1.6825,
      "step": 1910
    },
    {
      "epoch": 0.30094684320119436,
      "grad_norm": 0.17799334228038788,
      "learning_rate": 0.0008820408105231643,
      "loss": 1.7464,
      "step": 1915
    },
    {
      "epoch": 0.3017326051938868,
      "grad_norm": 0.1361367255449295,
      "learning_rate": 0.0008811545055473345,
      "loss": 1.665,
      "step": 1920
    },
    {
      "epoch": 0.3025183671865792,
      "grad_norm": 0.14883366227149963,
      "learning_rate": 0.0008802653321838003,
      "loss": 1.6717,
      "step": 1925
    },
    {
      "epoch": 0.3033041291792716,
      "grad_norm": 0.1452435553073883,
      "learning_rate": 0.0008793732971240582,
      "loss": 1.6707,
      "step": 1930
    },
    {
      "epoch": 0.304089891171964,
      "grad_norm": 0.14490064978599548,
      "learning_rate": 0.00087847840708114,
      "loss": 1.7017,
      "step": 1935
    },
    {
      "epoch": 0.3048756531646564,
      "grad_norm": 0.1717883199453354,
      "learning_rate": 0.0008775806687895631,
      "loss": 1.6796,
      "step": 1940
    },
    {
      "epoch": 0.30566141515734885,
      "grad_norm": 0.17518092691898346,
      "learning_rate": 0.0008766800890052793,
      "loss": 1.7223,
      "step": 1945
    },
    {
      "epoch": 0.30644717715004127,
      "grad_norm": 0.21119089424610138,
      "learning_rate": 0.0008757766745056241,
      "loss": 1.6164,
      "step": 1950
    },
    {
      "epoch": 0.30723293914273364,
      "grad_norm": 0.21860988438129425,
      "learning_rate": 0.0008748704320892658,
      "loss": 1.6693,
      "step": 1955
    },
    {
      "epoch": 0.30801870113542607,
      "grad_norm": 0.17688967287540436,
      "learning_rate": 0.000873961368576154,
      "loss": 1.6734,
      "step": 1960
    },
    {
      "epoch": 0.3088044631281185,
      "grad_norm": 0.15035995841026306,
      "learning_rate": 0.0008730494908074693,
      "loss": 1.6174,
      "step": 1965
    },
    {
      "epoch": 0.3095902251208109,
      "grad_norm": 0.131919726729393,
      "learning_rate": 0.00087213480564557,
      "loss": 1.7046,
      "step": 1970
    },
    {
      "epoch": 0.31037598711350334,
      "grad_norm": 0.16669556498527527,
      "learning_rate": 0.0008712173199739424,
      "loss": 1.6493,
      "step": 1975
    },
    {
      "epoch": 0.3111617491061957,
      "grad_norm": 0.14696942269802094,
      "learning_rate": 0.0008702970406971473,
      "loss": 1.6731,
      "step": 1980
    },
    {
      "epoch": 0.31194751109888813,
      "grad_norm": 0.12769675254821777,
      "learning_rate": 0.0008693739747407697,
      "loss": 1.665,
      "step": 1985
    },
    {
      "epoch": 0.31273327309158055,
      "grad_norm": 0.15159061551094055,
      "learning_rate": 0.000868448129051365,
      "loss": 1.6186,
      "step": 1990
    },
    {
      "epoch": 0.313519035084273,
      "grad_norm": 0.17533285915851593,
      "learning_rate": 0.0008675195105964083,
      "loss": 1.677,
      "step": 1995
    },
    {
      "epoch": 0.3143047970769654,
      "grad_norm": 0.19997474551200867,
      "learning_rate": 0.0008665881263642407,
      "loss": 1.6715,
      "step": 2000
    },
    {
      "epoch": 0.3150905590696578,
      "grad_norm": 0.191118523478508,
      "learning_rate": 0.0008656539833640174,
      "loss": 1.6473,
      "step": 2005
    },
    {
      "epoch": 0.3158763210623502,
      "grad_norm": 0.15149670839309692,
      "learning_rate": 0.0008647170886256546,
      "loss": 1.6804,
      "step": 2010
    },
    {
      "epoch": 0.3166620830550426,
      "grad_norm": 0.14794859290122986,
      "learning_rate": 0.0008637774491997773,
      "loss": 1.7138,
      "step": 2015
    },
    {
      "epoch": 0.31744784504773504,
      "grad_norm": 0.1326620876789093,
      "learning_rate": 0.000862835072157665,
      "loss": 1.6957,
      "step": 2020
    },
    {
      "epoch": 0.31823360704042747,
      "grad_norm": 0.16334739327430725,
      "learning_rate": 0.0008618899645911997,
      "loss": 1.7162,
      "step": 2025
    },
    {
      "epoch": 0.3190193690331199,
      "grad_norm": 0.17086929082870483,
      "learning_rate": 0.0008609421336128121,
      "loss": 1.7613,
      "step": 2030
    },
    {
      "epoch": 0.31980513102581226,
      "grad_norm": 0.14249901473522186,
      "learning_rate": 0.0008599915863554273,
      "loss": 1.7397,
      "step": 2035
    },
    {
      "epoch": 0.3205908930185047,
      "grad_norm": 0.21984586119651794,
      "learning_rate": 0.0008590383299724127,
      "loss": 1.651,
      "step": 2040
    },
    {
      "epoch": 0.3213766550111971,
      "grad_norm": 0.12290716916322708,
      "learning_rate": 0.0008580823716375227,
      "loss": 1.6888,
      "step": 2045
    },
    {
      "epoch": 0.32216241700388953,
      "grad_norm": 0.13404521346092224,
      "learning_rate": 0.0008571237185448456,
      "loss": 1.7038,
      "step": 2050
    },
    {
      "epoch": 0.32294817899658196,
      "grad_norm": 0.24460573494434357,
      "learning_rate": 0.000856162377908749,
      "loss": 1.6997,
      "step": 2055
    },
    {
      "epoch": 0.3237339409892743,
      "grad_norm": 0.1533568948507309,
      "learning_rate": 0.000855198356963826,
      "loss": 1.6101,
      "step": 2060
    },
    {
      "epoch": 0.32451970298196675,
      "grad_norm": 0.14102007448673248,
      "learning_rate": 0.0008542316629648398,
      "loss": 1.6884,
      "step": 2065
    },
    {
      "epoch": 0.3253054649746592,
      "grad_norm": 0.17003892362117767,
      "learning_rate": 0.0008532623031866704,
      "loss": 1.7216,
      "step": 2070
    },
    {
      "epoch": 0.3260912269673516,
      "grad_norm": 0.17711204290390015,
      "learning_rate": 0.0008522902849242587,
      "loss": 1.637,
      "step": 2075
    },
    {
      "epoch": 0.326876988960044,
      "grad_norm": 0.1561163067817688,
      "learning_rate": 0.0008513156154925523,
      "loss": 1.7727,
      "step": 2080
    },
    {
      "epoch": 0.3276627509527364,
      "grad_norm": 0.1554763913154602,
      "learning_rate": 0.0008503383022264506,
      "loss": 1.6895,
      "step": 2085
    },
    {
      "epoch": 0.3284485129454288,
      "grad_norm": 0.16475364565849304,
      "learning_rate": 0.0008493583524807484,
      "loss": 1.6817,
      "step": 2090
    },
    {
      "epoch": 0.32923427493812124,
      "grad_norm": 0.15485644340515137,
      "learning_rate": 0.0008483757736300822,
      "loss": 1.7206,
      "step": 2095
    },
    {
      "epoch": 0.33002003693081366,
      "grad_norm": 0.15525245666503906,
      "learning_rate": 0.0008473905730688733,
      "loss": 1.6805,
      "step": 2100
    },
    {
      "epoch": 0.3308057989235061,
      "grad_norm": 0.17805235087871552,
      "learning_rate": 0.0008464027582112732,
      "loss": 1.7431,
      "step": 2105
    },
    {
      "epoch": 0.33159156091619846,
      "grad_norm": 0.16209563612937927,
      "learning_rate": 0.000845412336491107,
      "loss": 1.7098,
      "step": 2110
    },
    {
      "epoch": 0.3323773229088909,
      "grad_norm": 0.1910310685634613,
      "learning_rate": 0.0008444193153618182,
      "loss": 1.6679,
      "step": 2115
    },
    {
      "epoch": 0.3331630849015833,
      "grad_norm": 0.202640563249588,
      "learning_rate": 0.0008434237022964117,
      "loss": 1.7205,
      "step": 2120
    },
    {
      "epoch": 0.33394884689427573,
      "grad_norm": 0.3347494602203369,
      "learning_rate": 0.0008424255047873986,
      "loss": 1.7269,
      "step": 2125
    },
    {
      "epoch": 0.33473460888696815,
      "grad_norm": 0.1840268224477768,
      "learning_rate": 0.0008414247303467389,
      "loss": 1.6667,
      "step": 2130
    },
    {
      "epoch": 0.3355203708796605,
      "grad_norm": 0.17524029314517975,
      "learning_rate": 0.0008404213865057857,
      "loss": 1.7353,
      "step": 2135
    },
    {
      "epoch": 0.33630613287235295,
      "grad_norm": 0.15382514894008636,
      "learning_rate": 0.0008394154808152278,
      "loss": 1.6428,
      "step": 2140
    },
    {
      "epoch": 0.33709189486504537,
      "grad_norm": 0.2421923577785492,
      "learning_rate": 0.0008384070208450338,
      "loss": 1.7266,
      "step": 2145
    },
    {
      "epoch": 0.3378776568577378,
      "grad_norm": 0.16263365745544434,
      "learning_rate": 0.0008373960141843938,
      "loss": 1.7025,
      "step": 2150
    },
    {
      "epoch": 0.3386634188504302,
      "grad_norm": 0.17603865265846252,
      "learning_rate": 0.000836382468441664,
      "loss": 1.6223,
      "step": 2155
    },
    {
      "epoch": 0.33944918084312264,
      "grad_norm": 0.6742755770683289,
      "learning_rate": 0.0008353663912443079,
      "loss": 1.7571,
      "step": 2160
    },
    {
      "epoch": 0.340234942835815,
      "grad_norm": 0.16411545872688293,
      "learning_rate": 0.0008343477902388395,
      "loss": 1.7097,
      "step": 2165
    },
    {
      "epoch": 0.34102070482850744,
      "grad_norm": 0.1652558147907257,
      "learning_rate": 0.0008333266730907663,
      "loss": 1.7458,
      "step": 2170
    },
    {
      "epoch": 0.34180646682119986,
      "grad_norm": 0.1978401392698288,
      "learning_rate": 0.0008323030474845306,
      "loss": 1.6904,
      "step": 2175
    },
    {
      "epoch": 0.3425922288138923,
      "grad_norm": 0.17765507102012634,
      "learning_rate": 0.0008312769211234524,
      "loss": 1.6717,
      "step": 2180
    },
    {
      "epoch": 0.3433779908065847,
      "grad_norm": 0.1868458241224289,
      "learning_rate": 0.0008302483017296711,
      "loss": 1.7436,
      "step": 2185
    },
    {
      "epoch": 0.3441637527992771,
      "grad_norm": 0.15795573592185974,
      "learning_rate": 0.0008292171970440876,
      "loss": 1.7093,
      "step": 2190
    },
    {
      "epoch": 0.3449495147919695,
      "grad_norm": 0.14316695928573608,
      "learning_rate": 0.0008281836148263057,
      "loss": 1.6806,
      "step": 2195
    },
    {
      "epoch": 0.3457352767846619,
      "grad_norm": 0.17996078729629517,
      "learning_rate": 0.0008271475628545741,
      "loss": 1.6856,
      "step": 2200
    },
    {
      "epoch": 0.34652103877735435,
      "grad_norm": 0.13540180027484894,
      "learning_rate": 0.0008261090489257277,
      "loss": 1.7469,
      "step": 2205
    },
    {
      "epoch": 0.3473068007700468,
      "grad_norm": 0.22064046561717987,
      "learning_rate": 0.000825068080855129,
      "loss": 1.7472,
      "step": 2210
    },
    {
      "epoch": 0.34809256276273914,
      "grad_norm": 0.1410595029592514,
      "learning_rate": 0.0008240246664766087,
      "loss": 1.6685,
      "step": 2215
    },
    {
      "epoch": 0.34887832475543157,
      "grad_norm": 0.16196031868457794,
      "learning_rate": 0.0008229788136424081,
      "loss": 1.6643,
      "step": 2220
    },
    {
      "epoch": 0.349664086748124,
      "grad_norm": 0.13669522106647491,
      "learning_rate": 0.0008219305302231185,
      "loss": 1.6631,
      "step": 2225
    },
    {
      "epoch": 0.3504498487408164,
      "grad_norm": 0.15855997800827026,
      "learning_rate": 0.0008208798241076227,
      "loss": 1.6319,
      "step": 2230
    },
    {
      "epoch": 0.35123561073350884,
      "grad_norm": 0.13129590451717377,
      "learning_rate": 0.000819826703203036,
      "loss": 1.6641,
      "step": 2235
    },
    {
      "epoch": 0.3520213727262012,
      "grad_norm": 0.17638882994651794,
      "learning_rate": 0.0008187711754346456,
      "loss": 1.676,
      "step": 2240
    },
    {
      "epoch": 0.35280713471889363,
      "grad_norm": 0.17996013164520264,
      "learning_rate": 0.0008177132487458521,
      "loss": 1.6299,
      "step": 2245
    },
    {
      "epoch": 0.35359289671158606,
      "grad_norm": 0.18002741038799286,
      "learning_rate": 0.0008166529310981092,
      "loss": 1.6424,
      "step": 2250
    },
    {
      "epoch": 0.3543786587042785,
      "grad_norm": 0.14587530493736267,
      "learning_rate": 0.0008155902304708634,
      "loss": 1.6697,
      "step": 2255
    },
    {
      "epoch": 0.3551644206969709,
      "grad_norm": 0.1491909921169281,
      "learning_rate": 0.000814525154861495,
      "loss": 1.6469,
      "step": 2260
    },
    {
      "epoch": 0.3559501826896633,
      "grad_norm": 0.16428177058696747,
      "learning_rate": 0.0008134577122852569,
      "loss": 1.7014,
      "step": 2265
    },
    {
      "epoch": 0.3567359446823557,
      "grad_norm": 0.12703630328178406,
      "learning_rate": 0.0008123879107752146,
      "loss": 1.6804,
      "step": 2270
    },
    {
      "epoch": 0.3575217066750481,
      "grad_norm": 0.22857950627803802,
      "learning_rate": 0.0008113157583821861,
      "loss": 1.6736,
      "step": 2275
    },
    {
      "epoch": 0.35830746866774055,
      "grad_norm": 0.144062340259552,
      "learning_rate": 0.0008102412631746807,
      "loss": 1.5896,
      "step": 2280
    },
    {
      "epoch": 0.35909323066043297,
      "grad_norm": 0.13251659274101257,
      "learning_rate": 0.000809164433238839,
      "loss": 1.6734,
      "step": 2285
    },
    {
      "epoch": 0.3598789926531254,
      "grad_norm": 0.20809276401996613,
      "learning_rate": 0.0008080852766783712,
      "loss": 1.7509,
      "step": 2290
    },
    {
      "epoch": 0.36066475464581776,
      "grad_norm": 0.20248106122016907,
      "learning_rate": 0.0008070038016144972,
      "loss": 1.6447,
      "step": 2295
    },
    {
      "epoch": 0.3614505166385102,
      "grad_norm": 0.2282155454158783,
      "learning_rate": 0.0008059200161858841,
      "loss": 1.6913,
      "step": 2300
    },
    {
      "epoch": 0.3622362786312026,
      "grad_norm": 0.15200570225715637,
      "learning_rate": 0.0008048339285485862,
      "loss": 1.6474,
      "step": 2305
    },
    {
      "epoch": 0.36302204062389504,
      "grad_norm": 0.16428659856319427,
      "learning_rate": 0.0008037455468759831,
      "loss": 1.6909,
      "step": 2310
    },
    {
      "epoch": 0.36380780261658746,
      "grad_norm": 0.1449277251958847,
      "learning_rate": 0.0008026548793587181,
      "loss": 1.6456,
      "step": 2315
    },
    {
      "epoch": 0.36459356460927983,
      "grad_norm": 0.12401038408279419,
      "learning_rate": 0.0008015619342046365,
      "loss": 1.6657,
      "step": 2320
    },
    {
      "epoch": 0.36537932660197225,
      "grad_norm": 0.18149249255657196,
      "learning_rate": 0.0008004667196387245,
      "loss": 1.6201,
      "step": 2325
    },
    {
      "epoch": 0.3661650885946647,
      "grad_norm": 0.1270531415939331,
      "learning_rate": 0.0007993692439030463,
      "loss": 1.7276,
      "step": 2330
    },
    {
      "epoch": 0.3669508505873571,
      "grad_norm": 0.11993835866451263,
      "learning_rate": 0.000798269515256683,
      "loss": 1.6833,
      "step": 2335
    },
    {
      "epoch": 0.3677366125800495,
      "grad_norm": 0.13166651129722595,
      "learning_rate": 0.0007971675419756698,
      "loss": 1.6036,
      "step": 2340
    },
    {
      "epoch": 0.3685223745727419,
      "grad_norm": 0.18711446225643158,
      "learning_rate": 0.0007960633323529341,
      "loss": 1.6782,
      "step": 2345
    },
    {
      "epoch": 0.3693081365654343,
      "grad_norm": 0.1776147037744522,
      "learning_rate": 0.0007949568946982325,
      "loss": 1.6903,
      "step": 2350
    },
    {
      "epoch": 0.37009389855812674,
      "grad_norm": 0.14657479524612427,
      "learning_rate": 0.0007938482373380895,
      "loss": 1.7061,
      "step": 2355
    },
    {
      "epoch": 0.37087966055081917,
      "grad_norm": 0.1647639274597168,
      "learning_rate": 0.0007927373686157334,
      "loss": 1.6486,
      "step": 2360
    },
    {
      "epoch": 0.3716654225435116,
      "grad_norm": 0.13776181638240814,
      "learning_rate": 0.0007916242968910345,
      "loss": 1.5636,
      "step": 2365
    },
    {
      "epoch": 0.37245118453620396,
      "grad_norm": 0.1941072791814804,
      "learning_rate": 0.0007905090305404417,
      "loss": 1.638,
      "step": 2370
    },
    {
      "epoch": 0.3732369465288964,
      "grad_norm": 0.15426956117153168,
      "learning_rate": 0.0007893915779569194,
      "loss": 1.6899,
      "step": 2375
    },
    {
      "epoch": 0.3740227085215888,
      "grad_norm": 0.15892522037029266,
      "learning_rate": 0.0007882719475498851,
      "loss": 1.6904,
      "step": 2380
    },
    {
      "epoch": 0.37480847051428123,
      "grad_norm": 0.15704740583896637,
      "learning_rate": 0.0007871501477451451,
      "loss": 1.6176,
      "step": 2385
    },
    {
      "epoch": 0.37559423250697366,
      "grad_norm": 0.14114616811275482,
      "learning_rate": 0.0007860261869848318,
      "loss": 1.6511,
      "step": 2390
    },
    {
      "epoch": 0.376379994499666,
      "grad_norm": 0.1459956020116806,
      "learning_rate": 0.0007849000737273397,
      "loss": 1.6805,
      "step": 2395
    },
    {
      "epoch": 0.37716575649235845,
      "grad_norm": 0.14372305572032928,
      "learning_rate": 0.0007837718164472623,
      "loss": 1.6114,
      "step": 2400
    },
    {
      "epoch": 0.3779515184850509,
      "grad_norm": 0.15054304897785187,
      "learning_rate": 0.0007826414236353276,
      "loss": 1.7266,
      "step": 2405
    },
    {
      "epoch": 0.3787372804777433,
      "grad_norm": 0.22728455066680908,
      "learning_rate": 0.000781508903798335,
      "loss": 1.7173,
      "step": 2410
    },
    {
      "epoch": 0.3795230424704357,
      "grad_norm": 0.15423768758773804,
      "learning_rate": 0.0007803742654590906,
      "loss": 1.63,
      "step": 2415
    },
    {
      "epoch": 0.3803088044631281,
      "grad_norm": 0.21122123301029205,
      "learning_rate": 0.0007792375171563433,
      "loss": 1.6727,
      "step": 2420
    },
    {
      "epoch": 0.3810945664558205,
      "grad_norm": 0.16940170526504517,
      "learning_rate": 0.0007780986674447208,
      "loss": 1.7145,
      "step": 2425
    },
    {
      "epoch": 0.38188032844851294,
      "grad_norm": 0.13977159559726715,
      "learning_rate": 0.0007769577248946649,
      "loss": 1.6329,
      "step": 2430
    },
    {
      "epoch": 0.38266609044120536,
      "grad_norm": 0.11926419287919998,
      "learning_rate": 0.000775814698092367,
      "loss": 1.7181,
      "step": 2435
    },
    {
      "epoch": 0.3834518524338978,
      "grad_norm": 0.1658172905445099,
      "learning_rate": 0.0007746695956397041,
      "loss": 1.5932,
      "step": 2440
    },
    {
      "epoch": 0.3842376144265902,
      "grad_norm": 0.11511421203613281,
      "learning_rate": 0.0007735224261541727,
      "loss": 1.6356,
      "step": 2445
    },
    {
      "epoch": 0.3850233764192826,
      "grad_norm": 0.15368768572807312,
      "learning_rate": 0.0007723731982688255,
      "loss": 1.6138,
      "step": 2450
    },
    {
      "epoch": 0.385809138411975,
      "grad_norm": 0.1516408920288086,
      "learning_rate": 0.0007712219206322056,
      "loss": 1.6715,
      "step": 2455
    },
    {
      "epoch": 0.38659490040466743,
      "grad_norm": 0.15503014624118805,
      "learning_rate": 0.0007700686019082812,
      "loss": 1.5931,
      "step": 2460
    },
    {
      "epoch": 0.38738066239735985,
      "grad_norm": 0.18041059374809265,
      "learning_rate": 0.0007689132507763812,
      "loss": 1.6274,
      "step": 2465
    },
    {
      "epoch": 0.3881664243900523,
      "grad_norm": 0.14432255923748016,
      "learning_rate": 0.0007677558759311291,
      "loss": 1.6361,
      "step": 2470
    },
    {
      "epoch": 0.38895218638274465,
      "grad_norm": 0.14074422419071198,
      "learning_rate": 0.0007665964860823784,
      "loss": 1.6276,
      "step": 2475
    },
    {
      "epoch": 0.38973794837543707,
      "grad_norm": 0.15996317565441132,
      "learning_rate": 0.0007654350899551458,
      "loss": 1.6351,
      "step": 2480
    },
    {
      "epoch": 0.3905237103681295,
      "grad_norm": 0.1606004387140274,
      "learning_rate": 0.0007642716962895472,
      "loss": 1.6978,
      "step": 2485
    },
    {
      "epoch": 0.3913094723608219,
      "grad_norm": 0.15753324329853058,
      "learning_rate": 0.0007631063138407301,
      "loss": 1.6589,
      "step": 2490
    },
    {
      "epoch": 0.39209523435351434,
      "grad_norm": 0.14514195919036865,
      "learning_rate": 0.0007619389513788093,
      "loss": 1.6412,
      "step": 2495
    },
    {
      "epoch": 0.3928809963462067,
      "grad_norm": 0.17876259982585907,
      "learning_rate": 0.0007607696176888002,
      "loss": 1.7265,
      "step": 2500
    },
    {
      "epoch": 0.39366675833889914,
      "grad_norm": 0.1770378202199936,
      "learning_rate": 0.000759598321570552,
      "loss": 1.6525,
      "step": 2505
    },
    {
      "epoch": 0.39445252033159156,
      "grad_norm": 0.13470305502414703,
      "learning_rate": 0.0007584250718386832,
      "loss": 1.5947,
      "step": 2510
    },
    {
      "epoch": 0.395238282324284,
      "grad_norm": 0.16835801303386688,
      "learning_rate": 0.0007572498773225136,
      "loss": 1.6635,
      "step": 2515
    },
    {
      "epoch": 0.3960240443169764,
      "grad_norm": 0.18890178203582764,
      "learning_rate": 0.0007560727468659988,
      "loss": 1.58,
      "step": 2520
    },
    {
      "epoch": 0.3968098063096688,
      "grad_norm": 0.1397354155778885,
      "learning_rate": 0.0007548936893276634,
      "loss": 1.6523,
      "step": 2525
    },
    {
      "epoch": 0.3975955683023612,
      "grad_norm": 0.1338605135679245,
      "learning_rate": 0.0007537127135805341,
      "loss": 1.6461,
      "step": 2530
    },
    {
      "epoch": 0.3983813302950536,
      "grad_norm": 0.1662890762090683,
      "learning_rate": 0.0007525298285120733,
      "loss": 1.6881,
      "step": 2535
    },
    {
      "epoch": 0.39916709228774605,
      "grad_norm": 0.14534525573253632,
      "learning_rate": 0.000751345043024112,
      "loss": 1.6404,
      "step": 2540
    },
    {
      "epoch": 0.3999528542804385,
      "grad_norm": 0.16143235564231873,
      "learning_rate": 0.0007501583660327827,
      "loss": 1.6824,
      "step": 2545
    },
    {
      "epoch": 0.40073861627313084,
      "grad_norm": 0.1411980539560318,
      "learning_rate": 0.0007489698064684527,
      "loss": 1.6082,
      "step": 2550
    },
    {
      "epoch": 0.40152437826582327,
      "grad_norm": 0.1654977947473526,
      "learning_rate": 0.0007477793732756564,
      "loss": 1.6116,
      "step": 2555
    },
    {
      "epoch": 0.4023101402585157,
      "grad_norm": 0.1615074872970581,
      "learning_rate": 0.0007465870754130286,
      "loss": 1.613,
      "step": 2560
    },
    {
      "epoch": 0.4030959022512081,
      "grad_norm": 0.14305217564105988,
      "learning_rate": 0.0007453929218532365,
      "loss": 1.6736,
      "step": 2565
    },
    {
      "epoch": 0.40388166424390054,
      "grad_norm": 0.13158391416072845,
      "learning_rate": 0.0007441969215829122,
      "loss": 1.6553,
      "step": 2570
    },
    {
      "epoch": 0.4046674262365929,
      "grad_norm": 0.30756956338882446,
      "learning_rate": 0.0007429990836025855,
      "loss": 1.6628,
      "step": 2575
    },
    {
      "epoch": 0.40545318822928533,
      "grad_norm": 0.127960205078125,
      "learning_rate": 0.0007417994169266159,
      "loss": 1.6054,
      "step": 2580
    },
    {
      "epoch": 0.40623895022197776,
      "grad_norm": 0.19319772720336914,
      "learning_rate": 0.000740597930583125,
      "loss": 1.631,
      "step": 2585
    },
    {
      "epoch": 0.4070247122146702,
      "grad_norm": 0.14012780785560608,
      "learning_rate": 0.0007393946336139278,
      "loss": 1.5878,
      "step": 2590
    },
    {
      "epoch": 0.4078104742073626,
      "grad_norm": 0.12992408871650696,
      "learning_rate": 0.0007381895350744656,
      "loss": 1.7209,
      "step": 2595
    },
    {
      "epoch": 0.40859623620005503,
      "grad_norm": 0.17100664973258972,
      "learning_rate": 0.0007369826440337378,
      "loss": 1.658,
      "step": 2600
    },
    {
      "epoch": 0.4093819981927474,
      "grad_norm": 0.13439995050430298,
      "learning_rate": 0.0007357739695742324,
      "loss": 1.6069,
      "step": 2605
    },
    {
      "epoch": 0.4101677601854398,
      "grad_norm": 0.15416887402534485,
      "learning_rate": 0.0007345635207918594,
      "loss": 1.7172,
      "step": 2610
    },
    {
      "epoch": 0.41095352217813225,
      "grad_norm": 0.1403576135635376,
      "learning_rate": 0.0007333513067958812,
      "loss": 1.7471,
      "step": 2615
    },
    {
      "epoch": 0.41173928417082467,
      "grad_norm": 0.1455819308757782,
      "learning_rate": 0.0007321373367088442,
      "loss": 1.6624,
      "step": 2620
    },
    {
      "epoch": 0.4125250461635171,
      "grad_norm": 0.17243732511997223,
      "learning_rate": 0.0007309216196665105,
      "loss": 1.5789,
      "step": 2625
    },
    {
      "epoch": 0.41331080815620946,
      "grad_norm": 0.16030173003673553,
      "learning_rate": 0.0007297041648177889,
      "loss": 1.7102,
      "step": 2630
    },
    {
      "epoch": 0.4140965701489019,
      "grad_norm": 0.14231261610984802,
      "learning_rate": 0.0007284849813246662,
      "loss": 1.707,
      "step": 2635
    },
    {
      "epoch": 0.4148823321415943,
      "grad_norm": 0.14213086664676666,
      "learning_rate": 0.000727264078362138,
      "loss": 1.6524,
      "step": 2640
    },
    {
      "epoch": 0.41566809413428674,
      "grad_norm": 0.16257204115390778,
      "learning_rate": 0.0007260414651181399,
      "loss": 1.6361,
      "step": 2645
    },
    {
      "epoch": 0.41645385612697916,
      "grad_norm": 0.1615426391363144,
      "learning_rate": 0.0007248171507934785,
      "loss": 1.6094,
      "step": 2650
    },
    {
      "epoch": 0.41723961811967153,
      "grad_norm": 0.16172373294830322,
      "learning_rate": 0.0007235911446017612,
      "loss": 1.6127,
      "step": 2655
    },
    {
      "epoch": 0.41802538011236395,
      "grad_norm": 0.22370944917201996,
      "learning_rate": 0.0007223634557693286,
      "loss": 1.6398,
      "step": 2660
    },
    {
      "epoch": 0.4188111421050564,
      "grad_norm": 0.1495995670557022,
      "learning_rate": 0.0007211340935351831,
      "loss": 1.6818,
      "step": 2665
    },
    {
      "epoch": 0.4195969040977488,
      "grad_norm": 0.13977926969528198,
      "learning_rate": 0.0007199030671509209,
      "loss": 1.7055,
      "step": 2670
    },
    {
      "epoch": 0.4203826660904412,
      "grad_norm": 0.15699009597301483,
      "learning_rate": 0.0007186703858806617,
      "loss": 1.6536,
      "step": 2675
    },
    {
      "epoch": 0.4211684280831336,
      "grad_norm": 0.148645281791687,
      "learning_rate": 0.0007174360590009787,
      "loss": 1.5893,
      "step": 2680
    },
    {
      "epoch": 0.421954190075826,
      "grad_norm": 0.14828044176101685,
      "learning_rate": 0.0007162000958008301,
      "loss": 1.6163,
      "step": 2685
    },
    {
      "epoch": 0.42273995206851844,
      "grad_norm": 0.15765506029129028,
      "learning_rate": 0.0007149625055814872,
      "loss": 1.6033,
      "step": 2690
    },
    {
      "epoch": 0.42352571406121087,
      "grad_norm": 0.18456675112247467,
      "learning_rate": 0.0007137232976564663,
      "loss": 1.6514,
      "step": 2695
    },
    {
      "epoch": 0.4243114760539033,
      "grad_norm": 0.16911190748214722,
      "learning_rate": 0.0007124824813514571,
      "loss": 1.6484,
      "step": 2700
    },
    {
      "epoch": 0.42509723804659566,
      "grad_norm": 0.1826726347208023,
      "learning_rate": 0.0007112400660042537,
      "loss": 1.6445,
      "step": 2705
    },
    {
      "epoch": 0.4258830000392881,
      "grad_norm": 0.14041639864444733,
      "learning_rate": 0.0007099960609646838,
      "loss": 1.6464,
      "step": 2710
    },
    {
      "epoch": 0.4266687620319805,
      "grad_norm": 0.12943454086780548,
      "learning_rate": 0.000708750475594538,
      "loss": 1.6154,
      "step": 2715
    },
    {
      "epoch": 0.42745452402467293,
      "grad_norm": 0.15895868837833405,
      "learning_rate": 0.0007075033192675,
      "loss": 1.5816,
      "step": 2720
    },
    {
      "epoch": 0.42824028601736536,
      "grad_norm": 0.17042267322540283,
      "learning_rate": 0.0007062546013690759,
      "loss": 1.5853,
      "step": 2725
    },
    {
      "epoch": 0.4290260480100578,
      "grad_norm": 0.20396976172924042,
      "learning_rate": 0.0007050043312965232,
      "loss": 1.6317,
      "step": 2730
    },
    {
      "epoch": 0.42981181000275015,
      "grad_norm": 0.18434785306453705,
      "learning_rate": 0.0007037525184587803,
      "loss": 1.6722,
      "step": 2735
    },
    {
      "epoch": 0.4305975719954426,
      "grad_norm": 0.12577559053897858,
      "learning_rate": 0.000702499172276396,
      "loss": 1.7135,
      "step": 2740
    },
    {
      "epoch": 0.431383333988135,
      "grad_norm": 0.20810461044311523,
      "learning_rate": 0.0007012443021814581,
      "loss": 1.7251,
      "step": 2745
    },
    {
      "epoch": 0.4321690959808274,
      "grad_norm": 0.22052597999572754,
      "learning_rate": 0.0006999879176175226,
      "loss": 1.6398,
      "step": 2750
    },
    {
      "epoch": 0.43295485797351985,
      "grad_norm": 0.1631423681974411,
      "learning_rate": 0.0006987300280395427,
      "loss": 1.6847,
      "step": 2755
    },
    {
      "epoch": 0.4337406199662122,
      "grad_norm": 0.166092187166214,
      "learning_rate": 0.0006974706429137978,
      "loss": 1.6606,
      "step": 2760
    },
    {
      "epoch": 0.43452638195890464,
      "grad_norm": 0.16192929446697235,
      "learning_rate": 0.0006962097717178221,
      "loss": 1.6697,
      "step": 2765
    },
    {
      "epoch": 0.43531214395159706,
      "grad_norm": 0.11831378191709518,
      "learning_rate": 0.0006949474239403328,
      "loss": 1.5835,
      "step": 2770
    },
    {
      "epoch": 0.4360979059442895,
      "grad_norm": 0.15122415125370026,
      "learning_rate": 0.0006936836090811598,
      "loss": 1.6469,
      "step": 2775
    },
    {
      "epoch": 0.4368836679369819,
      "grad_norm": 0.18945452570915222,
      "learning_rate": 0.000692418336651173,
      "loss": 1.5554,
      "step": 2780
    },
    {
      "epoch": 0.4376694299296743,
      "grad_norm": 0.1292644888162613,
      "learning_rate": 0.0006911516161722115,
      "loss": 1.6987,
      "step": 2785
    },
    {
      "epoch": 0.4384551919223667,
      "grad_norm": 0.16247059404850006,
      "learning_rate": 0.000689883457177012,
      "loss": 1.6198,
      "step": 2790
    },
    {
      "epoch": 0.43924095391505913,
      "grad_norm": 0.1748296171426773,
      "learning_rate": 0.0006886138692091362,
      "loss": 1.691,
      "step": 2795
    },
    {
      "epoch": 0.44002671590775155,
      "grad_norm": 0.12950074672698975,
      "learning_rate": 0.0006873428618229003,
      "loss": 1.6407,
      "step": 2800
    },
    {
      "epoch": 0.440812477900444,
      "grad_norm": 0.17089992761611938,
      "learning_rate": 0.0006860704445833015,
      "loss": 1.641,
      "step": 2805
    },
    {
      "epoch": 0.44159823989313635,
      "grad_norm": 0.1477111577987671,
      "learning_rate": 0.0006847966270659478,
      "loss": 1.678,
      "step": 2810
    },
    {
      "epoch": 0.44238400188582877,
      "grad_norm": 0.14656804502010345,
      "learning_rate": 0.0006835214188569843,
      "loss": 1.6195,
      "step": 2815
    },
    {
      "epoch": 0.4431697638785212,
      "grad_norm": 0.15418322384357452,
      "learning_rate": 0.0006822448295530221,
      "loss": 1.6286,
      "step": 2820
    },
    {
      "epoch": 0.4439555258712136,
      "grad_norm": 0.13586872816085815,
      "learning_rate": 0.000680966868761066,
      "loss": 1.6822,
      "step": 2825
    },
    {
      "epoch": 0.44474128786390604,
      "grad_norm": 0.14607511460781097,
      "learning_rate": 0.0006796875460984414,
      "loss": 1.6614,
      "step": 2830
    },
    {
      "epoch": 0.4455270498565984,
      "grad_norm": 0.15407826006412506,
      "learning_rate": 0.0006784068711927229,
      "loss": 1.6135,
      "step": 2835
    },
    {
      "epoch": 0.44631281184929084,
      "grad_norm": 0.1422751098871231,
      "learning_rate": 0.0006771248536816613,
      "loss": 1.6104,
      "step": 2840
    },
    {
      "epoch": 0.44709857384198326,
      "grad_norm": 0.1250545233488083,
      "learning_rate": 0.0006758415032131112,
      "loss": 1.6461,
      "step": 2845
    },
    {
      "epoch": 0.4478843358346757,
      "grad_norm": 0.1620786488056183,
      "learning_rate": 0.0006745568294449587,
      "loss": 1.6642,
      "step": 2850
    },
    {
      "epoch": 0.4486700978273681,
      "grad_norm": 0.2060784548521042,
      "learning_rate": 0.0006732708420450477,
      "loss": 1.666,
      "step": 2855
    },
    {
      "epoch": 0.4494558598200605,
      "grad_norm": 0.1400783210992813,
      "learning_rate": 0.0006719835506911088,
      "loss": 1.6238,
      "step": 2860
    },
    {
      "epoch": 0.4502416218127529,
      "grad_norm": 0.1502895951271057,
      "learning_rate": 0.0006706949650706849,
      "loss": 1.5815,
      "step": 2865
    },
    {
      "epoch": 0.4510273838054453,
      "grad_norm": 0.16864760220050812,
      "learning_rate": 0.0006694050948810592,
      "loss": 1.6644,
      "step": 2870
    },
    {
      "epoch": 0.45181314579813775,
      "grad_norm": 0.14266374707221985,
      "learning_rate": 0.0006681139498291815,
      "loss": 1.6552,
      "step": 2875
    },
    {
      "epoch": 0.4525989077908302,
      "grad_norm": 0.25216013193130493,
      "learning_rate": 0.0006668215396315964,
      "loss": 1.5835,
      "step": 2880
    },
    {
      "epoch": 0.4533846697835226,
      "grad_norm": 0.15556307137012482,
      "learning_rate": 0.0006655278740143689,
      "loss": 1.6315,
      "step": 2885
    },
    {
      "epoch": 0.45417043177621497,
      "grad_norm": 0.14055636525154114,
      "learning_rate": 0.0006642329627130115,
      "loss": 1.6428,
      "step": 2890
    },
    {
      "epoch": 0.4549561937689074,
      "grad_norm": 0.16647887229919434,
      "learning_rate": 0.0006629368154724116,
      "loss": 1.6541,
      "step": 2895
    },
    {
      "epoch": 0.4557419557615998,
      "grad_norm": 0.17561818659305573,
      "learning_rate": 0.0006616394420467575,
      "loss": 1.6565,
      "step": 2900
    },
    {
      "epoch": 0.45652771775429224,
      "grad_norm": 0.1490733027458191,
      "learning_rate": 0.000660340852199465,
      "loss": 1.6201,
      "step": 2905
    },
    {
      "epoch": 0.45731347974698466,
      "grad_norm": 0.16749203205108643,
      "learning_rate": 0.0006590410557031044,
      "loss": 1.6306,
      "step": 2910
    },
    {
      "epoch": 0.45809924173967703,
      "grad_norm": 0.12608392536640167,
      "learning_rate": 0.0006577400623393266,
      "loss": 1.6271,
      "step": 2915
    },
    {
      "epoch": 0.45888500373236946,
      "grad_norm": 0.1351223587989807,
      "learning_rate": 0.0006564378818987893,
      "loss": 1.6072,
      "step": 2920
    },
    {
      "epoch": 0.4596707657250619,
      "grad_norm": 0.14112618565559387,
      "learning_rate": 0.0006551345241810837,
      "loss": 1.6559,
      "step": 2925
    },
    {
      "epoch": 0.4604565277177543,
      "grad_norm": 0.12047524750232697,
      "learning_rate": 0.000653829998994661,
      "loss": 1.6756,
      "step": 2930
    },
    {
      "epoch": 0.46124228971044673,
      "grad_norm": 0.16356752812862396,
      "learning_rate": 0.0006525243161567576,
      "loss": 1.6884,
      "step": 2935
    },
    {
      "epoch": 0.4620280517031391,
      "grad_norm": 0.14315475523471832,
      "learning_rate": 0.0006512174854933224,
      "loss": 1.6488,
      "step": 2940
    },
    {
      "epoch": 0.4628138136958315,
      "grad_norm": 0.12611393630504608,
      "learning_rate": 0.0006499095168389419,
      "loss": 1.6481,
      "step": 2945
    },
    {
      "epoch": 0.46359957568852395,
      "grad_norm": 0.12579138576984406,
      "learning_rate": 0.0006486004200367669,
      "loss": 1.6586,
      "step": 2950
    },
    {
      "epoch": 0.46438533768121637,
      "grad_norm": 0.18551719188690186,
      "learning_rate": 0.0006472902049384377,
      "loss": 1.5887,
      "step": 2955
    },
    {
      "epoch": 0.4651710996739088,
      "grad_norm": 0.15424126386642456,
      "learning_rate": 0.0006459788814040105,
      "loss": 1.6598,
      "step": 2960
    },
    {
      "epoch": 0.46595686166660116,
      "grad_norm": 0.1597795933485031,
      "learning_rate": 0.0006446664593018834,
      "loss": 1.6753,
      "step": 2965
    },
    {
      "epoch": 0.4667426236592936,
      "grad_norm": 0.14304637908935547,
      "learning_rate": 0.0006433529485087214,
      "loss": 1.6558,
      "step": 2970
    },
    {
      "epoch": 0.467528385651986,
      "grad_norm": 0.15329056978225708,
      "learning_rate": 0.0006420383589093827,
      "loss": 1.6326,
      "step": 2975
    },
    {
      "epoch": 0.46831414764467844,
      "grad_norm": 0.1391008198261261,
      "learning_rate": 0.000640722700396844,
      "loss": 1.6108,
      "step": 2980
    },
    {
      "epoch": 0.46909990963737086,
      "grad_norm": 0.14664943516254425,
      "learning_rate": 0.0006394059828721261,
      "loss": 1.6063,
      "step": 2985
    },
    {
      "epoch": 0.46988567163006323,
      "grad_norm": 0.22576557099819183,
      "learning_rate": 0.0006380882162442196,
      "loss": 1.6035,
      "step": 2990
    },
    {
      "epoch": 0.47067143362275565,
      "grad_norm": 0.1459265500307083,
      "learning_rate": 0.0006367694104300097,
      "loss": 1.6023,
      "step": 2995
    },
    {
      "epoch": 0.4714571956154481,
      "grad_norm": 0.12152886390686035,
      "learning_rate": 0.0006354495753542027,
      "loss": 1.5558,
      "step": 3000
    },
    {
      "epoch": 0.4722429576081405,
      "grad_norm": 0.13670937716960907,
      "learning_rate": 0.0006341287209492498,
      "loss": 1.5578,
      "step": 3005
    },
    {
      "epoch": 0.4730287196008329,
      "grad_norm": 0.14993344247341156,
      "learning_rate": 0.000632806857155274,
      "loss": 1.636,
      "step": 3010
    },
    {
      "epoch": 0.4738144815935253,
      "grad_norm": 0.20564016699790955,
      "learning_rate": 0.000631483993919994,
      "loss": 1.6322,
      "step": 3015
    },
    {
      "epoch": 0.4746002435862177,
      "grad_norm": 0.1213807612657547,
      "learning_rate": 0.0006301601411986502,
      "loss": 1.6361,
      "step": 3020
    },
    {
      "epoch": 0.47538600557891014,
      "grad_norm": 0.12629488110542297,
      "learning_rate": 0.0006288353089539289,
      "loss": 1.6376,
      "step": 3025
    },
    {
      "epoch": 0.47617176757160257,
      "grad_norm": 0.14332091808319092,
      "learning_rate": 0.0006275095071558881,
      "loss": 1.5986,
      "step": 3030
    },
    {
      "epoch": 0.476957529564295,
      "grad_norm": 0.16367210447788239,
      "learning_rate": 0.0006261827457818822,
      "loss": 1.6127,
      "step": 3035
    },
    {
      "epoch": 0.4777432915569874,
      "grad_norm": 0.1383843868970871,
      "learning_rate": 0.0006248550348164868,
      "loss": 1.5363,
      "step": 3040
    },
    {
      "epoch": 0.4785290535496798,
      "grad_norm": 0.14265769720077515,
      "learning_rate": 0.0006235263842514242,
      "loss": 1.5952,
      "step": 3045
    },
    {
      "epoch": 0.4793148155423722,
      "grad_norm": 0.13620123267173767,
      "learning_rate": 0.0006221968040854867,
      "loss": 1.5788,
      "step": 3050
    },
    {
      "epoch": 0.48010057753506463,
      "grad_norm": 0.17570005357265472,
      "learning_rate": 0.0006208663043244631,
      "loss": 1.6889,
      "step": 3055
    },
    {
      "epoch": 0.48088633952775706,
      "grad_norm": 0.18797026574611664,
      "learning_rate": 0.0006195348949810625,
      "loss": 1.7221,
      "step": 3060
    },
    {
      "epoch": 0.4816721015204495,
      "grad_norm": 0.14126017689704895,
      "learning_rate": 0.0006182025860748389,
      "loss": 1.5377,
      "step": 3065
    },
    {
      "epoch": 0.48245786351314185,
      "grad_norm": 0.18811370432376862,
      "learning_rate": 0.0006168693876321163,
      "loss": 1.6591,
      "step": 3070
    },
    {
      "epoch": 0.4832436255058343,
      "grad_norm": 0.20044410228729248,
      "learning_rate": 0.0006155353096859124,
      "loss": 1.6215,
      "step": 3075
    },
    {
      "epoch": 0.4840293874985267,
      "grad_norm": 0.13966962695121765,
      "learning_rate": 0.0006142003622758641,
      "loss": 1.6331,
      "step": 3080
    },
    {
      "epoch": 0.4848151494912191,
      "grad_norm": 0.1432633399963379,
      "learning_rate": 0.0006128645554481512,
      "loss": 1.6516,
      "step": 3085
    },
    {
      "epoch": 0.48560091148391155,
      "grad_norm": 0.1459830105304718,
      "learning_rate": 0.0006115278992554214,
      "loss": 1.5967,
      "step": 3090
    },
    {
      "epoch": 0.4863866734766039,
      "grad_norm": 0.1282690018415451,
      "learning_rate": 0.000610190403756714,
      "loss": 1.6412,
      "step": 3095
    },
    {
      "epoch": 0.48717243546929634,
      "grad_norm": 0.1409655064344406,
      "learning_rate": 0.0006088520790173843,
      "loss": 1.6607,
      "step": 3100
    },
    {
      "epoch": 0.48795819746198876,
      "grad_norm": 0.1343081146478653,
      "learning_rate": 0.0006075129351090291,
      "loss": 1.6391,
      "step": 3105
    },
    {
      "epoch": 0.4887439594546812,
      "grad_norm": 0.13972026109695435,
      "learning_rate": 0.0006061729821094085,
      "loss": 1.5868,
      "step": 3110
    },
    {
      "epoch": 0.4895297214473736,
      "grad_norm": 0.14654341340065002,
      "learning_rate": 0.0006048322301023723,
      "loss": 1.612,
      "step": 3115
    },
    {
      "epoch": 0.490315483440066,
      "grad_norm": 0.17553551495075226,
      "learning_rate": 0.0006034906891777831,
      "loss": 1.613,
      "step": 3120
    },
    {
      "epoch": 0.4911012454327584,
      "grad_norm": 0.18809862434864044,
      "learning_rate": 0.0006021483694314406,
      "loss": 1.6353,
      "step": 3125
    },
    {
      "epoch": 0.49188700742545083,
      "grad_norm": 0.17778019607067108,
      "learning_rate": 0.0006008052809650052,
      "loss": 1.6269,
      "step": 3130
    },
    {
      "epoch": 0.49267276941814325,
      "grad_norm": 0.13343100249767303,
      "learning_rate": 0.0005994614338859224,
      "loss": 1.6214,
      "step": 3135
    },
    {
      "epoch": 0.4934585314108357,
      "grad_norm": 0.13967198133468628,
      "learning_rate": 0.0005981168383073472,
      "loss": 1.6184,
      "step": 3140
    },
    {
      "epoch": 0.49424429340352805,
      "grad_norm": 0.12899260222911835,
      "learning_rate": 0.0005967715043480667,
      "loss": 1.5902,
      "step": 3145
    },
    {
      "epoch": 0.49503005539622047,
      "grad_norm": 0.1432379186153412,
      "learning_rate": 0.0005954254421324253,
      "loss": 1.6336,
      "step": 3150
    },
    {
      "epoch": 0.4958158173889129,
      "grad_norm": 0.15734805166721344,
      "learning_rate": 0.0005940786617902477,
      "loss": 1.6614,
      "step": 3155
    },
    {
      "epoch": 0.4966015793816053,
      "grad_norm": 0.17404969036579132,
      "learning_rate": 0.0005927311734567624,
      "loss": 1.6685,
      "step": 3160
    },
    {
      "epoch": 0.49738734137429774,
      "grad_norm": 0.21907399594783783,
      "learning_rate": 0.0005913829872725269,
      "loss": 1.5632,
      "step": 3165
    },
    {
      "epoch": 0.4981731033669901,
      "grad_norm": 0.1369909644126892,
      "learning_rate": 0.0005900341133833491,
      "loss": 1.623,
      "step": 3170
    },
    {
      "epoch": 0.49895886535968254,
      "grad_norm": 0.14243030548095703,
      "learning_rate": 0.0005886845619402138,
      "loss": 1.5829,
      "step": 3175
    },
    {
      "epoch": 0.49974462735237496,
      "grad_norm": 0.1607036292552948,
      "learning_rate": 0.0005873343430992034,
      "loss": 1.5537,
      "step": 3180
    },
    {
      "epoch": 0.5005303893450673,
      "grad_norm": 0.12607838213443756,
      "learning_rate": 0.0005859834670214236,
      "loss": 1.6441,
      "step": 3185
    },
    {
      "epoch": 0.5013161513377598,
      "grad_norm": 0.13709622621536255,
      "learning_rate": 0.0005846319438729257,
      "loss": 1.627,
      "step": 3190
    },
    {
      "epoch": 0.5021019133304522,
      "grad_norm": 0.1329515129327774,
      "learning_rate": 0.000583279783824631,
      "loss": 1.6167,
      "step": 3195
    },
    {
      "epoch": 0.5028876753231446,
      "grad_norm": 0.15467658638954163,
      "learning_rate": 0.0005819269970522532,
      "loss": 1.6389,
      "step": 3200
    },
    {
      "epoch": 0.503673437315837,
      "grad_norm": 0.19440630078315735,
      "learning_rate": 0.0005805735937362231,
      "loss": 1.5974,
      "step": 3205
    },
    {
      "epoch": 0.5044591993085294,
      "grad_norm": 0.1676415205001831,
      "learning_rate": 0.0005792195840616108,
      "loss": 1.6102,
      "step": 3210
    },
    {
      "epoch": 0.5052449613012219,
      "grad_norm": 0.19758261740207672,
      "learning_rate": 0.0005778649782180497,
      "loss": 1.6061,
      "step": 3215
    },
    {
      "epoch": 0.5060307232939143,
      "grad_norm": 0.15707242488861084,
      "learning_rate": 0.00057650978639966,
      "loss": 1.5979,
      "step": 3220
    },
    {
      "epoch": 0.5068164852866067,
      "grad_norm": 0.15218037366867065,
      "learning_rate": 0.0005751540188049713,
      "loss": 1.5923,
      "step": 3225
    },
    {
      "epoch": 0.5076022472792991,
      "grad_norm": 0.149359792470932,
      "learning_rate": 0.000573797685636846,
      "loss": 1.5835,
      "step": 3230
    },
    {
      "epoch": 0.5083880092719915,
      "grad_norm": 0.18229952454566956,
      "learning_rate": 0.0005724407971024037,
      "loss": 1.6207,
      "step": 3235
    },
    {
      "epoch": 0.5091737712646839,
      "grad_norm": 0.16412632167339325,
      "learning_rate": 0.0005710833634129424,
      "loss": 1.6654,
      "step": 3240
    },
    {
      "epoch": 0.5099595332573763,
      "grad_norm": 0.15745171904563904,
      "learning_rate": 0.0005697253947838631,
      "loss": 1.543,
      "step": 3245
    },
    {
      "epoch": 0.5107452952500687,
      "grad_norm": 0.13623344898223877,
      "learning_rate": 0.0005683669014345924,
      "loss": 1.582,
      "step": 3250
    },
    {
      "epoch": 0.5115310572427612,
      "grad_norm": 0.1499263346195221,
      "learning_rate": 0.0005670078935885058,
      "loss": 1.5902,
      "step": 3255
    },
    {
      "epoch": 0.5123168192354536,
      "grad_norm": 0.14522770047187805,
      "learning_rate": 0.0005656483814728508,
      "loss": 1.6872,
      "step": 3260
    },
    {
      "epoch": 0.513102581228146,
      "grad_norm": 0.1918184757232666,
      "learning_rate": 0.0005642883753186693,
      "loss": 1.6782,
      "step": 3265
    },
    {
      "epoch": 0.5138883432208384,
      "grad_norm": 0.18836189806461334,
      "learning_rate": 0.0005629278853607218,
      "loss": 1.6462,
      "step": 3270
    },
    {
      "epoch": 0.5146741052135309,
      "grad_norm": 0.12512125074863434,
      "learning_rate": 0.000561566921837409,
      "loss": 1.6126,
      "step": 3275
    },
    {
      "epoch": 0.5154598672062233,
      "grad_norm": 0.14574217796325684,
      "learning_rate": 0.0005602054949906957,
      "loss": 1.6821,
      "step": 3280
    },
    {
      "epoch": 0.5162456291989157,
      "grad_norm": 0.17824433743953705,
      "learning_rate": 0.0005588436150660336,
      "loss": 1.62,
      "step": 3285
    },
    {
      "epoch": 0.517031391191608,
      "grad_norm": 0.21210236847400665,
      "learning_rate": 0.000557481292312284,
      "loss": 1.6722,
      "step": 3290
    },
    {
      "epoch": 0.5178171531843004,
      "grad_norm": 0.169542133808136,
      "learning_rate": 0.0005561185369816404,
      "loss": 1.5686,
      "step": 3295
    },
    {
      "epoch": 0.5186029151769929,
      "grad_norm": 0.15672409534454346,
      "learning_rate": 0.0005547553593295521,
      "loss": 1.6275,
      "step": 3300
    },
    {
      "epoch": 0.5193886771696853,
      "grad_norm": 0.1227111890912056,
      "learning_rate": 0.0005533917696146464,
      "loss": 1.6133,
      "step": 3305
    },
    {
      "epoch": 0.5201744391623777,
      "grad_norm": 0.12742657959461212,
      "learning_rate": 0.0005520277780986514,
      "loss": 1.6471,
      "step": 3310
    },
    {
      "epoch": 0.5209602011550701,
      "grad_norm": 1.2764840126037598,
      "learning_rate": 0.0005506633950463195,
      "loss": 1.6583,
      "step": 3315
    },
    {
      "epoch": 0.5217459631477626,
      "grad_norm": 0.26217857003211975,
      "learning_rate": 0.0005492986307253489,
      "loss": 1.5793,
      "step": 3320
    },
    {
      "epoch": 0.522531725140455,
      "grad_norm": 0.2518414258956909,
      "learning_rate": 0.0005479334954063076,
      "loss": 1.6404,
      "step": 3325
    },
    {
      "epoch": 0.5233174871331474,
      "grad_norm": 0.14766822755336761,
      "learning_rate": 0.0005465679993625552,
      "loss": 1.6244,
      "step": 3330
    },
    {
      "epoch": 0.5241032491258398,
      "grad_norm": 0.15620769560337067,
      "learning_rate": 0.0005452021528701663,
      "loss": 1.5799,
      "step": 3335
    },
    {
      "epoch": 0.5248890111185321,
      "grad_norm": 0.1620067059993744,
      "learning_rate": 0.0005438359662078528,
      "loss": 1.6273,
      "step": 3340
    },
    {
      "epoch": 0.5256747731112246,
      "grad_norm": 0.13544130325317383,
      "learning_rate": 0.0005424694496568859,
      "loss": 1.5757,
      "step": 3345
    },
    {
      "epoch": 0.526460535103917,
      "grad_norm": 0.13699087500572205,
      "learning_rate": 0.0005411026135010203,
      "loss": 1.7186,
      "step": 3350
    },
    {
      "epoch": 0.5272462970966094,
      "grad_norm": 0.1648073047399521,
      "learning_rate": 0.0005397354680264151,
      "loss": 1.5614,
      "step": 3355
    },
    {
      "epoch": 0.5280320590893018,
      "grad_norm": 0.15802352130413055,
      "learning_rate": 0.0005383680235215579,
      "loss": 1.7054,
      "step": 3360
    },
    {
      "epoch": 0.5288178210819943,
      "grad_norm": 0.16547876596450806,
      "learning_rate": 0.0005370002902771861,
      "loss": 1.6021,
      "step": 3365
    },
    {
      "epoch": 0.5296035830746867,
      "grad_norm": 0.19628289341926575,
      "learning_rate": 0.0005356322785862102,
      "loss": 1.6354,
      "step": 3370
    },
    {
      "epoch": 0.5303893450673791,
      "grad_norm": 0.18181757628917694,
      "learning_rate": 0.0005342639987436363,
      "loss": 1.6311,
      "step": 3375
    },
    {
      "epoch": 0.5311751070600715,
      "grad_norm": 0.13773776590824127,
      "learning_rate": 0.0005328954610464881,
      "loss": 1.6258,
      "step": 3380
    },
    {
      "epoch": 0.531960869052764,
      "grad_norm": 0.16505779325962067,
      "learning_rate": 0.0005315266757937305,
      "loss": 1.6317,
      "step": 3385
    },
    {
      "epoch": 0.5327466310454563,
      "grad_norm": 0.18392467498779297,
      "learning_rate": 0.0005301576532861905,
      "loss": 1.5434,
      "step": 3390
    },
    {
      "epoch": 0.5335323930381487,
      "grad_norm": 0.4494565725326538,
      "learning_rate": 0.0005287884038264813,
      "loss": 1.6195,
      "step": 3395
    },
    {
      "epoch": 0.5343181550308411,
      "grad_norm": 0.11773408949375153,
      "learning_rate": 0.0005274189377189235,
      "loss": 1.6681,
      "step": 3400
    },
    {
      "epoch": 0.5351039170235335,
      "grad_norm": 0.12879732251167297,
      "learning_rate": 0.0005260492652694687,
      "loss": 1.576,
      "step": 3405
    },
    {
      "epoch": 0.535889679016226,
      "grad_norm": 0.18784938752651215,
      "learning_rate": 0.0005246793967856206,
      "loss": 1.6229,
      "step": 3410
    },
    {
      "epoch": 0.5366754410089184,
      "grad_norm": 0.17286457121372223,
      "learning_rate": 0.0005233093425763586,
      "loss": 1.5689,
      "step": 3415
    },
    {
      "epoch": 0.5374612030016108,
      "grad_norm": 0.14577491581439972,
      "learning_rate": 0.0005219391129520597,
      "loss": 1.563,
      "step": 3420
    },
    {
      "epoch": 0.5382469649943032,
      "grad_norm": 0.11635435372591019,
      "learning_rate": 0.0005205687182244211,
      "loss": 1.5521,
      "step": 3425
    },
    {
      "epoch": 0.5390327269869957,
      "grad_norm": 0.17694814503192902,
      "learning_rate": 0.0005191981687063822,
      "loss": 1.6428,
      "step": 3430
    },
    {
      "epoch": 0.5398184889796881,
      "grad_norm": 0.13338413834571838,
      "learning_rate": 0.0005178274747120477,
      "loss": 1.5941,
      "step": 3435
    },
    {
      "epoch": 0.5406042509723805,
      "grad_norm": 0.16258469223976135,
      "learning_rate": 0.0005164566465566092,
      "loss": 1.5991,
      "step": 3440
    },
    {
      "epoch": 0.5413900129650728,
      "grad_norm": 0.14882458746433258,
      "learning_rate": 0.0005150856945562681,
      "loss": 1.5826,
      "step": 3445
    },
    {
      "epoch": 0.5421757749577653,
      "grad_norm": 0.15069830417633057,
      "learning_rate": 0.0005137146290281574,
      "loss": 1.6133,
      "step": 3450
    },
    {
      "epoch": 0.5429615369504577,
      "grad_norm": 0.16154706478118896,
      "learning_rate": 0.0005123434602902655,
      "loss": 1.6148,
      "step": 3455
    },
    {
      "epoch": 0.5437472989431501,
      "grad_norm": 0.14857438206672668,
      "learning_rate": 0.0005109721986613561,
      "loss": 1.6026,
      "step": 3460
    },
    {
      "epoch": 0.5445330609358425,
      "grad_norm": 0.180572509765625,
      "learning_rate": 0.0005096008544608931,
      "loss": 1.6101,
      "step": 3465
    },
    {
      "epoch": 0.545318822928535,
      "grad_norm": 0.2038494050502777,
      "learning_rate": 0.0005082294380089613,
      "loss": 1.5987,
      "step": 3470
    },
    {
      "epoch": 0.5461045849212274,
      "grad_norm": 0.15037231147289276,
      "learning_rate": 0.000506857959626189,
      "loss": 1.5964,
      "step": 3475
    },
    {
      "epoch": 0.5468903469139198,
      "grad_norm": 0.12556269764900208,
      "learning_rate": 0.0005054864296336714,
      "loss": 1.5537,
      "step": 3480
    },
    {
      "epoch": 0.5476761089066122,
      "grad_norm": 0.1907692849636078,
      "learning_rate": 0.0005041148583528912,
      "loss": 1.6699,
      "step": 3485
    },
    {
      "epoch": 0.5484618708993046,
      "grad_norm": 0.15658941864967346,
      "learning_rate": 0.0005027432561056421,
      "loss": 1.6159,
      "step": 3490
    },
    {
      "epoch": 0.549247632891997,
      "grad_norm": 0.1611129492521286,
      "learning_rate": 0.0005013716332139509,
      "loss": 1.6474,
      "step": 3495
    },
    {
      "epoch": 0.5500333948846894,
      "grad_norm": 0.17045730352401733,
      "learning_rate": 0.0005,
      "loss": 1.5635,
      "step": 3500
    },
    {
      "epoch": 0.5508191568773818,
      "grad_norm": 0.1496753841638565,
      "learning_rate": 0.0004986283667860492,
      "loss": 1.679,
      "step": 3505
    },
    {
      "epoch": 0.5516049188700742,
      "grad_norm": 0.13515827059745789,
      "learning_rate": 0.0004972567438943581,
      "loss": 1.6049,
      "step": 3510
    },
    {
      "epoch": 0.5523906808627667,
      "grad_norm": 0.12685181200504303,
      "learning_rate": 0.0004958851416471089,
      "loss": 1.6869,
      "step": 3515
    },
    {
      "epoch": 0.5531764428554591,
      "grad_norm": 0.14106754958629608,
      "learning_rate": 0.0004945135703663286,
      "loss": 1.6025,
      "step": 3520
    },
    {
      "epoch": 0.5539622048481515,
      "grad_norm": 0.1331792026758194,
      "learning_rate": 0.000493142040373811,
      "loss": 1.6071,
      "step": 3525
    },
    {
      "epoch": 0.5547479668408439,
      "grad_norm": 0.14535002410411835,
      "learning_rate": 0.0004917705619910389,
      "loss": 1.656,
      "step": 3530
    },
    {
      "epoch": 0.5555337288335364,
      "grad_norm": 0.12771806120872498,
      "learning_rate": 0.0004903991455391071,
      "loss": 1.5305,
      "step": 3535
    },
    {
      "epoch": 0.5563194908262288,
      "grad_norm": 0.15101516246795654,
      "learning_rate": 0.0004890278013386439,
      "loss": 1.6659,
      "step": 3540
    },
    {
      "epoch": 0.5571052528189212,
      "grad_norm": 0.13331681489944458,
      "learning_rate": 0.00048765653970973463,
      "loss": 1.6045,
      "step": 3545
    },
    {
      "epoch": 0.5578910148116135,
      "grad_norm": 0.11692263185977936,
      "learning_rate": 0.00048628537097184253,
      "loss": 1.5341,
      "step": 3550
    },
    {
      "epoch": 0.5586767768043059,
      "grad_norm": 0.12343940883874893,
      "learning_rate": 0.00048491430544373213,
      "loss": 1.5635,
      "step": 3555
    },
    {
      "epoch": 0.5594625387969984,
      "grad_norm": 0.17898155748844147,
      "learning_rate": 0.00048354335344339084,
      "loss": 1.5985,
      "step": 3560
    },
    {
      "epoch": 0.5602483007896908,
      "grad_norm": 0.14213447272777557,
      "learning_rate": 0.0004821725252879523,
      "loss": 1.6057,
      "step": 3565
    },
    {
      "epoch": 0.5610340627823832,
      "grad_norm": 0.12153229862451553,
      "learning_rate": 0.0004808018312936178,
      "loss": 1.6717,
      "step": 3570
    },
    {
      "epoch": 0.5618198247750756,
      "grad_norm": 0.12813471257686615,
      "learning_rate": 0.00047943128177557905,
      "loss": 1.6458,
      "step": 3575
    },
    {
      "epoch": 0.5626055867677681,
      "grad_norm": 0.15701644122600555,
      "learning_rate": 0.00047806088704794046,
      "loss": 1.6205,
      "step": 3580
    },
    {
      "epoch": 0.5633913487604605,
      "grad_norm": 0.1378682404756546,
      "learning_rate": 0.00047669065742364145,
      "loss": 1.643,
      "step": 3585
    },
    {
      "epoch": 0.5641771107531529,
      "grad_norm": 0.14088793098926544,
      "learning_rate": 0.0004753206032143795,
      "loss": 1.7,
      "step": 3590
    },
    {
      "epoch": 0.5649628727458453,
      "grad_norm": 0.15259869396686554,
      "learning_rate": 0.0004739507347305314,
      "loss": 1.6119,
      "step": 3595
    },
    {
      "epoch": 0.5657486347385376,
      "grad_norm": 0.13838313519954681,
      "learning_rate": 0.0004725810622810765,
      "loss": 1.609,
      "step": 3600
    },
    {
      "epoch": 0.5665343967312301,
      "grad_norm": 0.19349806010723114,
      "learning_rate": 0.00047121159617351884,
      "loss": 1.5298,
      "step": 3605
    },
    {
      "epoch": 0.5673201587239225,
      "grad_norm": 0.14952600002288818,
      "learning_rate": 0.00046984234671380956,
      "loss": 1.5832,
      "step": 3610
    },
    {
      "epoch": 0.5681059207166149,
      "grad_norm": 0.1487891525030136,
      "learning_rate": 0.00046847332420626964,
      "loss": 1.58,
      "step": 3615
    },
    {
      "epoch": 0.5688916827093073,
      "grad_norm": 0.12130414694547653,
      "learning_rate": 0.00046710453895351195,
      "loss": 1.6952,
      "step": 3620
    },
    {
      "epoch": 0.5696774447019998,
      "grad_norm": 0.13457755744457245,
      "learning_rate": 0.00046573600125636395,
      "loss": 1.5882,
      "step": 3625
    },
    {
      "epoch": 0.5704632066946922,
      "grad_norm": 0.11782009154558182,
      "learning_rate": 0.0004643677214137898,
      "loss": 1.606,
      "step": 3630
    },
    {
      "epoch": 0.5712489686873846,
      "grad_norm": 0.17926670610904694,
      "learning_rate": 0.00046299970972281396,
      "loss": 1.6567,
      "step": 3635
    },
    {
      "epoch": 0.572034730680077,
      "grad_norm": 0.14689800143241882,
      "learning_rate": 0.0004616319764784421,
      "loss": 1.5803,
      "step": 3640
    },
    {
      "epoch": 0.5728204926727695,
      "grad_norm": 0.17201411724090576,
      "learning_rate": 0.0004602645319735849,
      "loss": 1.5905,
      "step": 3645
    },
    {
      "epoch": 0.5736062546654618,
      "grad_norm": 0.15453019738197327,
      "learning_rate": 0.00045889738649897984,
      "loss": 1.613,
      "step": 3650
    },
    {
      "epoch": 0.5743920166581542,
      "grad_norm": 0.19608797132968903,
      "learning_rate": 0.00045753055034311406,
      "loss": 1.5823,
      "step": 3655
    },
    {
      "epoch": 0.5751777786508466,
      "grad_norm": 0.13336700201034546,
      "learning_rate": 0.00045616403379214725,
      "loss": 1.633,
      "step": 3660
    },
    {
      "epoch": 0.575963540643539,
      "grad_norm": 0.1499382108449936,
      "learning_rate": 0.0004547978471298336,
      "loss": 1.5769,
      "step": 3665
    },
    {
      "epoch": 0.5767493026362315,
      "grad_norm": 0.1264503449201584,
      "learning_rate": 0.00045343200063744483,
      "loss": 1.581,
      "step": 3670
    },
    {
      "epoch": 0.5775350646289239,
      "grad_norm": 0.15569479763507843,
      "learning_rate": 0.00045206650459369256,
      "loss": 1.5404,
      "step": 3675
    },
    {
      "epoch": 0.5783208266216163,
      "grad_norm": 0.12474947422742844,
      "learning_rate": 0.00045070136927465124,
      "loss": 1.6176,
      "step": 3680
    },
    {
      "epoch": 0.5791065886143087,
      "grad_norm": 0.12202292680740356,
      "learning_rate": 0.0004493366049536806,
      "loss": 1.649,
      "step": 3685
    },
    {
      "epoch": 0.5798923506070012,
      "grad_norm": 0.1440856158733368,
      "learning_rate": 0.00044797222190134863,
      "loss": 1.6267,
      "step": 3690
    },
    {
      "epoch": 0.5806781125996936,
      "grad_norm": 0.14805102348327637,
      "learning_rate": 0.00044660823038535373,
      "loss": 1.568,
      "step": 3695
    },
    {
      "epoch": 0.581463874592386,
      "grad_norm": 0.16469161212444305,
      "learning_rate": 0.0004452446406704478,
      "loss": 1.6439,
      "step": 3700
    },
    {
      "epoch": 0.5822496365850783,
      "grad_norm": 0.1361335963010788,
      "learning_rate": 0.0004438814630183595,
      "loss": 1.5802,
      "step": 3705
    },
    {
      "epoch": 0.5830353985777708,
      "grad_norm": 0.14041323959827423,
      "learning_rate": 0.0004425187076877161,
      "loss": 1.6259,
      "step": 3710
    },
    {
      "epoch": 0.5838211605704632,
      "grad_norm": 0.173885777592659,
      "learning_rate": 0.00044115638493396643,
      "loss": 1.6512,
      "step": 3715
    },
    {
      "epoch": 0.5846069225631556,
      "grad_norm": 0.13919731974601746,
      "learning_rate": 0.0004397945050093044,
      "loss": 1.582,
      "step": 3720
    },
    {
      "epoch": 0.585392684555848,
      "grad_norm": 0.13981099426746368,
      "learning_rate": 0.0004384330781625911,
      "loss": 1.5468,
      "step": 3725
    },
    {
      "epoch": 0.5861784465485405,
      "grad_norm": 0.1628427505493164,
      "learning_rate": 0.0004370721146392783,
      "loss": 1.6533,
      "step": 3730
    },
    {
      "epoch": 0.5869642085412329,
      "grad_norm": 0.1454862356185913,
      "learning_rate": 0.00043571162468133073,
      "loss": 1.611,
      "step": 3735
    },
    {
      "epoch": 0.5877499705339253,
      "grad_norm": 0.127084419131279,
      "learning_rate": 0.0004343516185271494,
      "loss": 1.5406,
      "step": 3740
    },
    {
      "epoch": 0.5885357325266177,
      "grad_norm": 0.12855832278728485,
      "learning_rate": 0.00043299210641149435,
      "loss": 1.5789,
      "step": 3745
    },
    {
      "epoch": 0.5893214945193102,
      "grad_norm": 0.12993024289608002,
      "learning_rate": 0.0004316330985654077,
      "loss": 1.6774,
      "step": 3750
    },
    {
      "epoch": 0.5901072565120025,
      "grad_norm": 0.13693836331367493,
      "learning_rate": 0.000430274605216137,
      "loss": 1.6085,
      "step": 3755
    },
    {
      "epoch": 0.5908930185046949,
      "grad_norm": 0.1504506915807724,
      "learning_rate": 0.0004289166365870577,
      "loss": 1.5573,
      "step": 3760
    },
    {
      "epoch": 0.5916787804973873,
      "grad_norm": 0.17335031926631927,
      "learning_rate": 0.00042755920289759634,
      "loss": 1.626,
      "step": 3765
    },
    {
      "epoch": 0.5924645424900797,
      "grad_norm": 0.13067643344402313,
      "learning_rate": 0.0004262023143631538,
      "loss": 1.5954,
      "step": 3770
    },
    {
      "epoch": 0.5932503044827722,
      "grad_norm": 0.15987832844257355,
      "learning_rate": 0.0004248459811950288,
      "loss": 1.6007,
      "step": 3775
    },
    {
      "epoch": 0.5940360664754646,
      "grad_norm": 0.240760937333107,
      "learning_rate": 0.00042349021360034005,
      "loss": 1.6095,
      "step": 3780
    },
    {
      "epoch": 0.594821828468157,
      "grad_norm": 0.19699476659297943,
      "learning_rate": 0.0004221350217819504,
      "loss": 1.6421,
      "step": 3785
    },
    {
      "epoch": 0.5956075904608494,
      "grad_norm": 0.13763698935508728,
      "learning_rate": 0.00042078041593838946,
      "loss": 1.5438,
      "step": 3790
    },
    {
      "epoch": 0.5963933524535419,
      "grad_norm": 0.13714967668056488,
      "learning_rate": 0.00041942640626377694,
      "loss": 1.5082,
      "step": 3795
    },
    {
      "epoch": 0.5971791144462343,
      "grad_norm": 0.1486494392156601,
      "learning_rate": 0.0004180730029477468,
      "loss": 1.6616,
      "step": 3800
    },
    {
      "epoch": 0.5979648764389266,
      "grad_norm": 0.14766061305999756,
      "learning_rate": 0.00041672021617536914,
      "loss": 1.6168,
      "step": 3805
    },
    {
      "epoch": 0.598750638431619,
      "grad_norm": 0.13385340571403503,
      "learning_rate": 0.00041536805612707434,
      "loss": 1.5723,
      "step": 3810
    },
    {
      "epoch": 0.5995364004243114,
      "grad_norm": 0.17751716077327728,
      "learning_rate": 0.00041401653297857654,
      "loss": 1.5451,
      "step": 3815
    },
    {
      "epoch": 0.6003221624170039,
      "grad_norm": 0.14219363033771515,
      "learning_rate": 0.0004126656569007966,
      "loss": 1.6025,
      "step": 3820
    },
    {
      "epoch": 0.6011079244096963,
      "grad_norm": 0.13877145946025848,
      "learning_rate": 0.0004113154380597863,
      "loss": 1.6305,
      "step": 3825
    },
    {
      "epoch": 0.6018936864023887,
      "grad_norm": 0.1523716002702713,
      "learning_rate": 0.0004099658866166509,
      "loss": 1.5868,
      "step": 3830
    },
    {
      "epoch": 0.6026794483950811,
      "grad_norm": 0.13320079445838928,
      "learning_rate": 0.00040861701272747343,
      "loss": 1.581,
      "step": 3835
    },
    {
      "epoch": 0.6034652103877736,
      "grad_norm": 0.13998959958553314,
      "learning_rate": 0.00040726882654323757,
      "loss": 1.6086,
      "step": 3840
    },
    {
      "epoch": 0.604250972380466,
      "grad_norm": 0.1749476045370102,
      "learning_rate": 0.00040592133820975247,
      "loss": 1.5971,
      "step": 3845
    },
    {
      "epoch": 0.6050367343731584,
      "grad_norm": 0.14331944286823273,
      "learning_rate": 0.00040457455786757466,
      "loss": 1.6309,
      "step": 3850
    },
    {
      "epoch": 0.6058224963658508,
      "grad_norm": 0.1268499195575714,
      "learning_rate": 0.00040322849565193334,
      "loss": 1.6688,
      "step": 3855
    },
    {
      "epoch": 0.6066082583585432,
      "grad_norm": 0.1139456257224083,
      "learning_rate": 0.00040188316169265293,
      "loss": 1.5303,
      "step": 3860
    },
    {
      "epoch": 0.6073940203512356,
      "grad_norm": 0.1379457414150238,
      "learning_rate": 0.0004005385661140775,
      "loss": 1.643,
      "step": 3865
    },
    {
      "epoch": 0.608179782343928,
      "grad_norm": 0.15720854699611664,
      "learning_rate": 0.0003991947190349949,
      "loss": 1.6458,
      "step": 3870
    },
    {
      "epoch": 0.6089655443366204,
      "grad_norm": 0.13279718160629272,
      "learning_rate": 0.00039785163056855956,
      "loss": 1.6413,
      "step": 3875
    },
    {
      "epoch": 0.6097513063293128,
      "grad_norm": 0.16527603566646576,
      "learning_rate": 0.000396509310822217,
      "loss": 1.6411,
      "step": 3880
    },
    {
      "epoch": 0.6105370683220053,
      "grad_norm": 0.11751680821180344,
      "learning_rate": 0.00039516776989762783,
      "loss": 1.5949,
      "step": 3885
    },
    {
      "epoch": 0.6113228303146977,
      "grad_norm": 0.1423860639333725,
      "learning_rate": 0.0003938270178905916,
      "loss": 1.5761,
      "step": 3890
    },
    {
      "epoch": 0.6121085923073901,
      "grad_norm": 0.13618935644626617,
      "learning_rate": 0.0003924870648909711,
      "loss": 1.6124,
      "step": 3895
    },
    {
      "epoch": 0.6128943543000825,
      "grad_norm": 0.16317065060138702,
      "learning_rate": 0.00039114792098261564,
      "loss": 1.5755,
      "step": 3900
    },
    {
      "epoch": 0.613680116292775,
      "grad_norm": 0.13522252440452576,
      "learning_rate": 0.0003898095962432862,
      "loss": 1.6334,
      "step": 3905
    },
    {
      "epoch": 0.6144658782854673,
      "grad_norm": 0.1667720079421997,
      "learning_rate": 0.00038847210074457854,
      "loss": 1.5734,
      "step": 3910
    },
    {
      "epoch": 0.6152516402781597,
      "grad_norm": 0.14179572463035583,
      "learning_rate": 0.0003871354445518487,
      "loss": 1.5733,
      "step": 3915
    },
    {
      "epoch": 0.6160374022708521,
      "grad_norm": 0.13804243505001068,
      "learning_rate": 0.00038579963772413596,
      "loss": 1.6097,
      "step": 3920
    },
    {
      "epoch": 0.6168231642635446,
      "grad_norm": 0.14240044355392456,
      "learning_rate": 0.0003844646903140877,
      "loss": 1.6263,
      "step": 3925
    },
    {
      "epoch": 0.617608926256237,
      "grad_norm": 0.15428559482097626,
      "learning_rate": 0.0003831306123678838,
      "loss": 1.6567,
      "step": 3930
    },
    {
      "epoch": 0.6183946882489294,
      "grad_norm": 0.16212952136993408,
      "learning_rate": 0.00038179741392516104,
      "loss": 1.6562,
      "step": 3935
    },
    {
      "epoch": 0.6191804502416218,
      "grad_norm": 0.1477675437927246,
      "learning_rate": 0.00038046510501893747,
      "loss": 1.5078,
      "step": 3940
    },
    {
      "epoch": 0.6199662122343143,
      "grad_norm": 0.2077953964471817,
      "learning_rate": 0.00037913369567553696,
      "loss": 1.5593,
      "step": 3945
    },
    {
      "epoch": 0.6207519742270067,
      "grad_norm": 0.14560724794864655,
      "learning_rate": 0.00037780319591451347,
      "loss": 1.5988,
      "step": 3950
    },
    {
      "epoch": 0.6215377362196991,
      "grad_norm": 0.15922938287258148,
      "learning_rate": 0.0003764736157485761,
      "loss": 1.585,
      "step": 3955
    },
    {
      "epoch": 0.6223234982123914,
      "grad_norm": 0.15636377036571503,
      "learning_rate": 0.0003751449651835131,
      "loss": 1.6747,
      "step": 3960
    },
    {
      "epoch": 0.6231092602050838,
      "grad_norm": 0.17764806747436523,
      "learning_rate": 0.00037381725421811783,
      "loss": 1.6357,
      "step": 3965
    },
    {
      "epoch": 0.6238950221977763,
      "grad_norm": 0.18865743279457092,
      "learning_rate": 0.000372490492844112,
      "loss": 1.5983,
      "step": 3970
    },
    {
      "epoch": 0.6246807841904687,
      "grad_norm": 0.25516217947006226,
      "learning_rate": 0.0003711646910460713,
      "loss": 1.6183,
      "step": 3975
    },
    {
      "epoch": 0.6254665461831611,
      "grad_norm": 0.20700369775295258,
      "learning_rate": 0.00036983985880134987,
      "loss": 1.5452,
      "step": 3980
    },
    {
      "epoch": 0.6262523081758535,
      "grad_norm": 0.16521629691123962,
      "learning_rate": 0.0003685160060800059,
      "loss": 1.5794,
      "step": 3985
    },
    {
      "epoch": 0.627038070168546,
      "grad_norm": 0.13135258853435516,
      "learning_rate": 0.00036719314284472604,
      "loss": 1.6298,
      "step": 3990
    },
    {
      "epoch": 0.6278238321612384,
      "grad_norm": 0.13091085851192474,
      "learning_rate": 0.00036587127905075037,
      "loss": 1.553,
      "step": 3995
    },
    {
      "epoch": 0.6286095941539308,
      "grad_norm": 0.11903288215398788,
      "learning_rate": 0.0003645504246457976,
      "loss": 1.617,
      "step": 4000
    },
    {
      "epoch": 0.6293953561466232,
      "grad_norm": 0.21342940628528595,
      "learning_rate": 0.00036323058956999024,
      "loss": 1.6411,
      "step": 4005
    },
    {
      "epoch": 0.6301811181393157,
      "grad_norm": 0.15517936646938324,
      "learning_rate": 0.00036191178375578046,
      "loss": 1.5769,
      "step": 4010
    },
    {
      "epoch": 0.630966880132008,
      "grad_norm": 0.15112268924713135,
      "learning_rate": 0.00036059401712787393,
      "loss": 1.6215,
      "step": 4015
    },
    {
      "epoch": 0.6317526421247004,
      "grad_norm": 0.12736016511917114,
      "learning_rate": 0.000359277299603156,
      "loss": 1.5587,
      "step": 4020
    },
    {
      "epoch": 0.6325384041173928,
      "grad_norm": 0.11462604254484177,
      "learning_rate": 0.0003579616410906174,
      "loss": 1.6249,
      "step": 4025
    },
    {
      "epoch": 0.6333241661100852,
      "grad_norm": 0.12737910449504852,
      "learning_rate": 0.00035664705149127856,
      "loss": 1.6519,
      "step": 4030
    },
    {
      "epoch": 0.6341099281027777,
      "grad_norm": 0.13220828771591187,
      "learning_rate": 0.00035533354069811665,
      "loss": 1.5816,
      "step": 4035
    },
    {
      "epoch": 0.6348956900954701,
      "grad_norm": 0.18805982172489166,
      "learning_rate": 0.00035402111859598963,
      "loss": 1.5487,
      "step": 4040
    },
    {
      "epoch": 0.6356814520881625,
      "grad_norm": 0.1409517526626587,
      "learning_rate": 0.0003527097950615625,
      "loss": 1.6414,
      "step": 4045
    },
    {
      "epoch": 0.6364672140808549,
      "grad_norm": 0.12021984159946442,
      "learning_rate": 0.0003513995799632332,
      "loss": 1.5564,
      "step": 4050
    },
    {
      "epoch": 0.6372529760735474,
      "grad_norm": 0.12560012936592102,
      "learning_rate": 0.00035009048316105815,
      "loss": 1.6305,
      "step": 4055
    },
    {
      "epoch": 0.6380387380662398,
      "grad_norm": 0.11739975959062576,
      "learning_rate": 0.00034878251450667767,
      "loss": 1.6374,
      "step": 4060
    },
    {
      "epoch": 0.6388245000589321,
      "grad_norm": 0.1290743499994278,
      "learning_rate": 0.00034747568384324253,
      "loss": 1.5489,
      "step": 4065
    },
    {
      "epoch": 0.6396102620516245,
      "grad_norm": 0.1409534215927124,
      "learning_rate": 0.00034617000100533923,
      "loss": 1.5066,
      "step": 4070
    },
    {
      "epoch": 0.640396024044317,
      "grad_norm": 0.11737538129091263,
      "learning_rate": 0.00034486547581891625,
      "loss": 1.6869,
      "step": 4075
    },
    {
      "epoch": 0.6411817860370094,
      "grad_norm": 0.1106194406747818,
      "learning_rate": 0.00034356211810121083,
      "loss": 1.6172,
      "step": 4080
    },
    {
      "epoch": 0.6419675480297018,
      "grad_norm": 0.14032632112503052,
      "learning_rate": 0.0003422599376606735,
      "loss": 1.6105,
      "step": 4085
    },
    {
      "epoch": 0.6427533100223942,
      "grad_norm": 0.14203083515167236,
      "learning_rate": 0.00034095894429689554,
      "loss": 1.5882,
      "step": 4090
    },
    {
      "epoch": 0.6435390720150866,
      "grad_norm": 0.13212530314922333,
      "learning_rate": 0.00033965914780053496,
      "loss": 1.6069,
      "step": 4095
    },
    {
      "epoch": 0.6443248340077791,
      "grad_norm": 0.12459713220596313,
      "learning_rate": 0.0003383605579532425,
      "loss": 1.6644,
      "step": 4100
    },
    {
      "epoch": 0.6451105960004715,
      "grad_norm": 0.1313340663909912,
      "learning_rate": 0.0003370631845275883,
      "loss": 1.5869,
      "step": 4105
    },
    {
      "epoch": 0.6458963579931639,
      "grad_norm": 0.13839416205883026,
      "learning_rate": 0.00033576703728698855,
      "loss": 1.5567,
      "step": 4110
    },
    {
      "epoch": 0.6466821199858562,
      "grad_norm": 0.11876852810382843,
      "learning_rate": 0.00033447212598563126,
      "loss": 1.6101,
      "step": 4115
    },
    {
      "epoch": 0.6474678819785487,
      "grad_norm": 0.13081587851047516,
      "learning_rate": 0.0003331784603684035,
      "loss": 1.5729,
      "step": 4120
    },
    {
      "epoch": 0.6482536439712411,
      "grad_norm": 0.15403269231319427,
      "learning_rate": 0.0003318860501708184,
      "loss": 1.5955,
      "step": 4125
    },
    {
      "epoch": 0.6490394059639335,
      "grad_norm": 0.1455393135547638,
      "learning_rate": 0.0003305949051189409,
      "loss": 1.5778,
      "step": 4130
    },
    {
      "epoch": 0.6498251679566259,
      "grad_norm": 0.13471262156963348,
      "learning_rate": 0.00032930503492931514,
      "loss": 1.5595,
      "step": 4135
    },
    {
      "epoch": 0.6506109299493184,
      "grad_norm": 0.15387079119682312,
      "learning_rate": 0.0003280164493088912,
      "loss": 1.5661,
      "step": 4140
    },
    {
      "epoch": 0.6513966919420108,
      "grad_norm": 0.15407784283161163,
      "learning_rate": 0.00032672915795495223,
      "loss": 1.5295,
      "step": 4145
    },
    {
      "epoch": 0.6521824539347032,
      "grad_norm": 0.13979849219322205,
      "learning_rate": 0.0003254431705550414,
      "loss": 1.5533,
      "step": 4150
    },
    {
      "epoch": 0.6529682159273956,
      "grad_norm": 0.15691180527210236,
      "learning_rate": 0.0003241584967868888,
      "loss": 1.5126,
      "step": 4155
    },
    {
      "epoch": 0.653753977920088,
      "grad_norm": 0.1667742133140564,
      "learning_rate": 0.0003228751463183388,
      "loss": 1.5099,
      "step": 4160
    },
    {
      "epoch": 0.6545397399127805,
      "grad_norm": 0.1104639545083046,
      "learning_rate": 0.00032159312880727724,
      "loss": 1.5985,
      "step": 4165
    },
    {
      "epoch": 0.6553255019054728,
      "grad_norm": 0.11984073370695114,
      "learning_rate": 0.00032031245390155854,
      "loss": 1.6355,
      "step": 4170
    },
    {
      "epoch": 0.6561112638981652,
      "grad_norm": 0.14178301393985748,
      "learning_rate": 0.0003190331312389341,
      "loss": 1.6044,
      "step": 4175
    },
    {
      "epoch": 0.6568970258908576,
      "grad_norm": 0.13143818080425262,
      "learning_rate": 0.0003177551704469779,
      "loss": 1.6314,
      "step": 4180
    },
    {
      "epoch": 0.6576827878835501,
      "grad_norm": 0.12936004996299744,
      "learning_rate": 0.00031647858114301585,
      "loss": 1.5972,
      "step": 4185
    },
    {
      "epoch": 0.6584685498762425,
      "grad_norm": 0.13757792115211487,
      "learning_rate": 0.00031520337293405236,
      "loss": 1.647,
      "step": 4190
    },
    {
      "epoch": 0.6592543118689349,
      "grad_norm": 0.12552618980407715,
      "learning_rate": 0.00031392955541669843,
      "loss": 1.6214,
      "step": 4195
    },
    {
      "epoch": 0.6600400738616273,
      "grad_norm": 0.14673569798469543,
      "learning_rate": 0.0003126571381770998,
      "loss": 1.6282,
      "step": 4200
    },
    {
      "epoch": 0.6608258358543198,
      "grad_norm": 0.13930362462997437,
      "learning_rate": 0.0003113861307908638,
      "loss": 1.6094,
      "step": 4205
    },
    {
      "epoch": 0.6616115978470122,
      "grad_norm": 0.15072229504585266,
      "learning_rate": 0.00031011654282298815,
      "loss": 1.5527,
      "step": 4210
    },
    {
      "epoch": 0.6623973598397046,
      "grad_norm": 0.18607932329177856,
      "learning_rate": 0.0003088483838277885,
      "loss": 1.6465,
      "step": 4215
    },
    {
      "epoch": 0.6631831218323969,
      "grad_norm": 0.17094473540782928,
      "learning_rate": 0.0003075816633488271,
      "loss": 1.5388,
      "step": 4220
    },
    {
      "epoch": 0.6639688838250893,
      "grad_norm": 0.12629753351211548,
      "learning_rate": 0.0003063163909188403,
      "loss": 1.5959,
      "step": 4225
    },
    {
      "epoch": 0.6647546458177818,
      "grad_norm": 0.15863820910453796,
      "learning_rate": 0.0003050525760596673,
      "loss": 1.6186,
      "step": 4230
    },
    {
      "epoch": 0.6655404078104742,
      "grad_norm": 0.16218873858451843,
      "learning_rate": 0.00030379022828217806,
      "loss": 1.5959,
      "step": 4235
    },
    {
      "epoch": 0.6663261698031666,
      "grad_norm": 0.12031345069408417,
      "learning_rate": 0.0003025293570862021,
      "loss": 1.6435,
      "step": 4240
    },
    {
      "epoch": 0.667111931795859,
      "grad_norm": 0.13391123712062836,
      "learning_rate": 0.0003012699719604573,
      "loss": 1.5797,
      "step": 4245
    },
    {
      "epoch": 0.6678976937885515,
      "grad_norm": 0.1528289020061493,
      "learning_rate": 0.0003000120823824775,
      "loss": 1.5509,
      "step": 4250
    },
    {
      "epoch": 0.6686834557812439,
      "grad_norm": 0.13060513138771057,
      "learning_rate": 0.00029875569781854204,
      "loss": 1.5022,
      "step": 4255
    },
    {
      "epoch": 0.6694692177739363,
      "grad_norm": 0.12339182198047638,
      "learning_rate": 0.0002975008277236041,
      "loss": 1.5875,
      "step": 4260
    },
    {
      "epoch": 0.6702549797666287,
      "grad_norm": 0.13095946609973907,
      "learning_rate": 0.00029624748154121974,
      "loss": 1.507,
      "step": 4265
    },
    {
      "epoch": 0.671040741759321,
      "grad_norm": 0.1417776495218277,
      "learning_rate": 0.0002949956687034769,
      "loss": 1.5078,
      "step": 4270
    },
    {
      "epoch": 0.6718265037520135,
      "grad_norm": 0.1152493953704834,
      "learning_rate": 0.0002937453986309242,
      "loss": 1.5462,
      "step": 4275
    },
    {
      "epoch": 0.6726122657447059,
      "grad_norm": 0.11423389613628387,
      "learning_rate": 0.00029249668073250015,
      "loss": 1.5613,
      "step": 4280
    },
    {
      "epoch": 0.6733980277373983,
      "grad_norm": 0.15614832937717438,
      "learning_rate": 0.000291249524405462,
      "loss": 1.5854,
      "step": 4285
    },
    {
      "epoch": 0.6741837897300907,
      "grad_norm": 0.11558407545089722,
      "learning_rate": 0.0002900039390353164,
      "loss": 1.5662,
      "step": 4290
    },
    {
      "epoch": 0.6749695517227832,
      "grad_norm": 0.11645932495594025,
      "learning_rate": 0.00028875993399574634,
      "loss": 1.532,
      "step": 4295
    },
    {
      "epoch": 0.6757553137154756,
      "grad_norm": 0.18881000578403473,
      "learning_rate": 0.00028751751864854316,
      "loss": 1.5883,
      "step": 4300
    },
    {
      "epoch": 0.676541075708168,
      "grad_norm": 0.12830057740211487,
      "learning_rate": 0.0002862767023435339,
      "loss": 1.7049,
      "step": 4305
    },
    {
      "epoch": 0.6773268377008604,
      "grad_norm": 0.16305623948574066,
      "learning_rate": 0.0002850374944185128,
      "loss": 1.5801,
      "step": 4310
    },
    {
      "epoch": 0.6781125996935529,
      "grad_norm": 0.13387204706668854,
      "learning_rate": 0.0002837999041991699,
      "loss": 1.5547,
      "step": 4315
    },
    {
      "epoch": 0.6788983616862453,
      "grad_norm": 0.14606694877147675,
      "learning_rate": 0.00028256394099902127,
      "loss": 1.5591,
      "step": 4320
    },
    {
      "epoch": 0.6796841236789376,
      "grad_norm": 0.1575789600610733,
      "learning_rate": 0.0002813296141193384,
      "loss": 1.5756,
      "step": 4325
    },
    {
      "epoch": 0.68046988567163,
      "grad_norm": 0.17807775735855103,
      "learning_rate": 0.00028009693284907926,
      "loss": 1.6095,
      "step": 4330
    },
    {
      "epoch": 0.6812556476643225,
      "grad_norm": 0.1488974243402481,
      "learning_rate": 0.00027886590646481705,
      "loss": 1.5749,
      "step": 4335
    },
    {
      "epoch": 0.6820414096570149,
      "grad_norm": 0.14755655825138092,
      "learning_rate": 0.00027763654423067143,
      "loss": 1.5712,
      "step": 4340
    },
    {
      "epoch": 0.6828271716497073,
      "grad_norm": 0.1276608258485794,
      "learning_rate": 0.0002764088553982388,
      "loss": 1.5585,
      "step": 4345
    },
    {
      "epoch": 0.6836129336423997,
      "grad_norm": 0.18238218128681183,
      "learning_rate": 0.00027518284920652157,
      "loss": 1.6127,
      "step": 4350
    },
    {
      "epoch": 0.6843986956350921,
      "grad_norm": 0.10505139082670212,
      "learning_rate": 0.00027395853488186,
      "loss": 1.6235,
      "step": 4355
    },
    {
      "epoch": 0.6851844576277846,
      "grad_norm": 0.14508095383644104,
      "learning_rate": 0.0002727359216378621,
      "loss": 1.5853,
      "step": 4360
    },
    {
      "epoch": 0.685970219620477,
      "grad_norm": 0.1283094584941864,
      "learning_rate": 0.0002715150186753339,
      "loss": 1.5724,
      "step": 4365
    },
    {
      "epoch": 0.6867559816131694,
      "grad_norm": 0.13297708332538605,
      "learning_rate": 0.00027029583518221135,
      "loss": 1.6299,
      "step": 4370
    },
    {
      "epoch": 0.6875417436058617,
      "grad_norm": 0.1485053300857544,
      "learning_rate": 0.0002690783803334897,
      "loss": 1.5592,
      "step": 4375
    },
    {
      "epoch": 0.6883275055985542,
      "grad_norm": 0.14514178037643433,
      "learning_rate": 0.00026786266329115595,
      "loss": 1.6329,
      "step": 4380
    },
    {
      "epoch": 0.6891132675912466,
      "grad_norm": 0.12720349431037903,
      "learning_rate": 0.0002666486932041188,
      "loss": 1.5256,
      "step": 4385
    },
    {
      "epoch": 0.689899029583939,
      "grad_norm": 0.14566215872764587,
      "learning_rate": 0.0002654364792081407,
      "loss": 1.57,
      "step": 4390
    },
    {
      "epoch": 0.6906847915766314,
      "grad_norm": 0.1420929729938507,
      "learning_rate": 0.00026422603042576765,
      "loss": 1.5619,
      "step": 4395
    },
    {
      "epoch": 0.6914705535693239,
      "grad_norm": 0.15932297706604004,
      "learning_rate": 0.0002630173559662624,
      "loss": 1.5294,
      "step": 4400
    },
    {
      "epoch": 0.6922563155620163,
      "grad_norm": 0.10718350857496262,
      "learning_rate": 0.0002618104649255344,
      "loss": 1.5715,
      "step": 4405
    },
    {
      "epoch": 0.6930420775547087,
      "grad_norm": 0.1438807249069214,
      "learning_rate": 0.0002606053663860722,
      "loss": 1.5939,
      "step": 4410
    },
    {
      "epoch": 0.6938278395474011,
      "grad_norm": 0.1529790461063385,
      "learning_rate": 0.0002594020694168753,
      "loss": 1.5308,
      "step": 4415
    },
    {
      "epoch": 0.6946136015400936,
      "grad_norm": 0.1177016943693161,
      "learning_rate": 0.0002582005830733841,
      "loss": 1.4905,
      "step": 4420
    },
    {
      "epoch": 0.695399363532786,
      "grad_norm": 0.14389681816101074,
      "learning_rate": 0.0002570009163974145,
      "loss": 1.5644,
      "step": 4425
    },
    {
      "epoch": 0.6961851255254783,
      "grad_norm": 0.13227079808712006,
      "learning_rate": 0.00025580307841708785,
      "loss": 1.5765,
      "step": 4430
    },
    {
      "epoch": 0.6969708875181707,
      "grad_norm": 0.14656752347946167,
      "learning_rate": 0.00025460707814676365,
      "loss": 1.5437,
      "step": 4435
    },
    {
      "epoch": 0.6977566495108631,
      "grad_norm": 0.14903421700000763,
      "learning_rate": 0.00025341292458697134,
      "loss": 1.6126,
      "step": 4440
    },
    {
      "epoch": 0.6985424115035556,
      "grad_norm": 0.13473910093307495,
      "learning_rate": 0.00025222062672434364,
      "loss": 1.5527,
      "step": 4445
    },
    {
      "epoch": 0.699328173496248,
      "grad_norm": 0.1363900601863861,
      "learning_rate": 0.0002510301935315474,
      "loss": 1.6206,
      "step": 4450
    },
    {
      "epoch": 0.7001139354889404,
      "grad_norm": 0.140158548951149,
      "learning_rate": 0.00024984163396721737,
      "loss": 1.5959,
      "step": 4455
    },
    {
      "epoch": 0.7008996974816328,
      "grad_norm": 0.13826651871204376,
      "learning_rate": 0.0002486549569758882,
      "loss": 1.5772,
      "step": 4460
    },
    {
      "epoch": 0.7016854594743253,
      "grad_norm": 0.13384497165679932,
      "learning_rate": 0.0002474701714879268,
      "loss": 1.5658,
      "step": 4465
    },
    {
      "epoch": 0.7024712214670177,
      "grad_norm": 0.14882899820804596,
      "learning_rate": 0.0002462872864194661,
      "loss": 1.56,
      "step": 4470
    },
    {
      "epoch": 0.7032569834597101,
      "grad_norm": 0.1305229365825653,
      "learning_rate": 0.0002451063106723366,
      "loss": 1.5813,
      "step": 4475
    },
    {
      "epoch": 0.7040427454524024,
      "grad_norm": 0.1169796884059906,
      "learning_rate": 0.00024392725313400127,
      "loss": 1.6154,
      "step": 4480
    },
    {
      "epoch": 0.7048285074450948,
      "grad_norm": 0.2215767800807953,
      "learning_rate": 0.00024275012267748646,
      "loss": 1.5724,
      "step": 4485
    },
    {
      "epoch": 0.7056142694377873,
      "grad_norm": 0.12483309954404831,
      "learning_rate": 0.000241574928161317,
      "loss": 1.5052,
      "step": 4490
    },
    {
      "epoch": 0.7064000314304797,
      "grad_norm": 0.16134008765220642,
      "learning_rate": 0.0002404016784294481,
      "loss": 1.5418,
      "step": 4495
    },
    {
      "epoch": 0.7071857934231721,
      "grad_norm": 0.13525766134262085,
      "learning_rate": 0.00023923038231119992,
      "loss": 1.5272,
      "step": 4500
    },
    {
      "epoch": 0.7079715554158645,
      "grad_norm": 0.13309256732463837,
      "learning_rate": 0.0002380610486211907,
      "loss": 1.5805,
      "step": 4505
    },
    {
      "epoch": 0.708757317408557,
      "grad_norm": 0.14293818175792694,
      "learning_rate": 0.00023689368615926988,
      "loss": 1.5433,
      "step": 4510
    },
    {
      "epoch": 0.7095430794012494,
      "grad_norm": 0.18992508947849274,
      "learning_rate": 0.0002357283037104529,
      "loss": 1.5574,
      "step": 4515
    },
    {
      "epoch": 0.7103288413939418,
      "grad_norm": 0.1394491195678711,
      "learning_rate": 0.00023456491004485413,
      "loss": 1.5914,
      "step": 4520
    },
    {
      "epoch": 0.7111146033866342,
      "grad_norm": 0.10986452549695969,
      "learning_rate": 0.0002334035139176216,
      "loss": 1.543,
      "step": 4525
    },
    {
      "epoch": 0.7119003653793265,
      "grad_norm": 0.1387847661972046,
      "learning_rate": 0.00023224412406887098,
      "loss": 1.4953,
      "step": 4530
    },
    {
      "epoch": 0.712686127372019,
      "grad_norm": 0.13508880138397217,
      "learning_rate": 0.00023108674922361894,
      "loss": 1.5373,
      "step": 4535
    },
    {
      "epoch": 0.7134718893647114,
      "grad_norm": 0.14578190445899963,
      "learning_rate": 0.0002299313980917191,
      "loss": 1.5914,
      "step": 4540
    },
    {
      "epoch": 0.7142576513574038,
      "grad_norm": 0.15806812047958374,
      "learning_rate": 0.0002287780793677945,
      "loss": 1.5966,
      "step": 4545
    },
    {
      "epoch": 0.7150434133500962,
      "grad_norm": 0.14135576784610748,
      "learning_rate": 0.00022762680173117456,
      "loss": 1.5991,
      "step": 4550
    },
    {
      "epoch": 0.7158291753427887,
      "grad_norm": 0.11881235241889954,
      "learning_rate": 0.00022647757384582735,
      "loss": 1.5315,
      "step": 4555
    },
    {
      "epoch": 0.7166149373354811,
      "grad_norm": 0.14931359887123108,
      "learning_rate": 0.00022533040436029613,
      "loss": 1.6535,
      "step": 4560
    },
    {
      "epoch": 0.7174006993281735,
      "grad_norm": 0.14255623519420624,
      "learning_rate": 0.000224185301907633,
      "loss": 1.5585,
      "step": 4565
    },
    {
      "epoch": 0.7181864613208659,
      "grad_norm": 0.16669470071792603,
      "learning_rate": 0.00022304227510533514,
      "loss": 1.5423,
      "step": 4570
    },
    {
      "epoch": 0.7189722233135584,
      "grad_norm": 0.16331365704536438,
      "learning_rate": 0.0002219013325552794,
      "loss": 1.5725,
      "step": 4575
    },
    {
      "epoch": 0.7197579853062508,
      "grad_norm": 0.14862756431102753,
      "learning_rate": 0.0002207624828436568,
      "loss": 1.5734,
      "step": 4580
    },
    {
      "epoch": 0.7205437472989431,
      "grad_norm": 0.13208940625190735,
      "learning_rate": 0.00021962573454090966,
      "loss": 1.5893,
      "step": 4585
    },
    {
      "epoch": 0.7213295092916355,
      "grad_norm": 0.11863264441490173,
      "learning_rate": 0.0002184910962016649,
      "loss": 1.5707,
      "step": 4590
    },
    {
      "epoch": 0.722115271284328,
      "grad_norm": 0.11269424110651016,
      "learning_rate": 0.00021735857636467237,
      "loss": 1.5872,
      "step": 4595
    },
    {
      "epoch": 0.7229010332770204,
      "grad_norm": 0.12873615324497223,
      "learning_rate": 0.00021622818355273766,
      "loss": 1.6142,
      "step": 4600
    },
    {
      "epoch": 0.7236867952697128,
      "grad_norm": 0.13957516849040985,
      "learning_rate": 0.00021509992627266033,
      "loss": 1.5326,
      "step": 4605
    },
    {
      "epoch": 0.7244725572624052,
      "grad_norm": 0.13987894356250763,
      "learning_rate": 0.00021397381301516822,
      "loss": 1.5646,
      "step": 4610
    },
    {
      "epoch": 0.7252583192550976,
      "grad_norm": 0.12960903346538544,
      "learning_rate": 0.00021284985225485486,
      "loss": 1.5412,
      "step": 4615
    },
    {
      "epoch": 0.7260440812477901,
      "grad_norm": 0.14691345393657684,
      "learning_rate": 0.000211728052450115,
      "loss": 1.6413,
      "step": 4620
    },
    {
      "epoch": 0.7268298432404825,
      "grad_norm": 0.17346711456775665,
      "learning_rate": 0.00021060842204308062,
      "loss": 1.5836,
      "step": 4625
    },
    {
      "epoch": 0.7276156052331749,
      "grad_norm": 0.11949850618839264,
      "learning_rate": 0.00020949096945955859,
      "loss": 1.5568,
      "step": 4630
    },
    {
      "epoch": 0.7284013672258672,
      "grad_norm": 0.16478979587554932,
      "learning_rate": 0.00020837570310896543,
      "loss": 1.5923,
      "step": 4635
    },
    {
      "epoch": 0.7291871292185597,
      "grad_norm": 0.16687503457069397,
      "learning_rate": 0.0002072626313842666,
      "loss": 1.5882,
      "step": 4640
    },
    {
      "epoch": 0.7299728912112521,
      "grad_norm": 0.13315603137016296,
      "learning_rate": 0.0002061517626619105,
      "loss": 1.5726,
      "step": 4645
    },
    {
      "epoch": 0.7307586532039445,
      "grad_norm": 0.16446514427661896,
      "learning_rate": 0.00020504310530176757,
      "loss": 1.5051,
      "step": 4650
    },
    {
      "epoch": 0.7315444151966369,
      "grad_norm": 0.14502394199371338,
      "learning_rate": 0.0002039366676470661,
      "loss": 1.528,
      "step": 4655
    },
    {
      "epoch": 0.7323301771893294,
      "grad_norm": 0.14425790309906006,
      "learning_rate": 0.0002028324580243302,
      "loss": 1.5396,
      "step": 4660
    },
    {
      "epoch": 0.7331159391820218,
      "grad_norm": 0.14922231435775757,
      "learning_rate": 0.00020173048474331705,
      "loss": 1.5766,
      "step": 4665
    },
    {
      "epoch": 0.7339017011747142,
      "grad_norm": 0.1422736942768097,
      "learning_rate": 0.0002006307560969537,
      "loss": 1.5823,
      "step": 4670
    },
    {
      "epoch": 0.7346874631674066,
      "grad_norm": 0.1642303317785263,
      "learning_rate": 0.00019953328036127565,
      "loss": 1.583,
      "step": 4675
    },
    {
      "epoch": 0.735473225160099,
      "grad_norm": 0.1379195898771286,
      "learning_rate": 0.00019843806579536354,
      "loss": 1.5402,
      "step": 4680
    },
    {
      "epoch": 0.7362589871527914,
      "grad_norm": 0.14833813905715942,
      "learning_rate": 0.00019734512064128196,
      "loss": 1.5992,
      "step": 4685
    },
    {
      "epoch": 0.7370447491454838,
      "grad_norm": 0.13516221940517426,
      "learning_rate": 0.00019625445312401697,
      "loss": 1.5905,
      "step": 4690
    },
    {
      "epoch": 0.7378305111381762,
      "grad_norm": 0.1159839853644371,
      "learning_rate": 0.0001951660714514138,
      "loss": 1.5938,
      "step": 4695
    },
    {
      "epoch": 0.7386162731308686,
      "grad_norm": 0.12499277293682098,
      "learning_rate": 0.00019407998381411603,
      "loss": 1.5649,
      "step": 4700
    },
    {
      "epoch": 0.7394020351235611,
      "grad_norm": 0.12141992896795273,
      "learning_rate": 0.0001929961983855027,
      "loss": 1.5515,
      "step": 4705
    },
    {
      "epoch": 0.7401877971162535,
      "grad_norm": 0.13888101279735565,
      "learning_rate": 0.00019191472332162873,
      "loss": 1.5339,
      "step": 4710
    },
    {
      "epoch": 0.7409735591089459,
      "grad_norm": 0.1663873940706253,
      "learning_rate": 0.000190835566761161,
      "loss": 1.5977,
      "step": 4715
    },
    {
      "epoch": 0.7417593211016383,
      "grad_norm": 0.15261366963386536,
      "learning_rate": 0.00018975873682531942,
      "loss": 1.5611,
      "step": 4720
    },
    {
      "epoch": 0.7425450830943308,
      "grad_norm": 0.12267375737428665,
      "learning_rate": 0.000188684241617814,
      "loss": 1.5463,
      "step": 4725
    },
    {
      "epoch": 0.7433308450870232,
      "grad_norm": 0.1139058992266655,
      "learning_rate": 0.0001876120892247854,
      "loss": 1.5242,
      "step": 4730
    },
    {
      "epoch": 0.7441166070797156,
      "grad_norm": 0.13267825543880463,
      "learning_rate": 0.00018654228771474324,
      "loss": 1.5563,
      "step": 4735
    },
    {
      "epoch": 0.7449023690724079,
      "grad_norm": 0.10687974095344543,
      "learning_rate": 0.00018547484513850506,
      "loss": 1.5425,
      "step": 4740
    },
    {
      "epoch": 0.7456881310651003,
      "grad_norm": 0.13263630867004395,
      "learning_rate": 0.00018440976952913674,
      "loss": 1.5726,
      "step": 4745
    },
    {
      "epoch": 0.7464738930577928,
      "grad_norm": 0.16454148292541504,
      "learning_rate": 0.00018334706890189102,
      "loss": 1.6489,
      "step": 4750
    },
    {
      "epoch": 0.7472596550504852,
      "grad_norm": 0.12462660670280457,
      "learning_rate": 0.00018228675125414796,
      "loss": 1.5117,
      "step": 4755
    },
    {
      "epoch": 0.7480454170431776,
      "grad_norm": 0.13918079435825348,
      "learning_rate": 0.00018122882456535438,
      "loss": 1.5798,
      "step": 4760
    },
    {
      "epoch": 0.74883117903587,
      "grad_norm": 0.12236585468053818,
      "learning_rate": 0.00018017329679696415,
      "loss": 1.5931,
      "step": 4765
    },
    {
      "epoch": 0.7496169410285625,
      "grad_norm": 0.15193891525268555,
      "learning_rate": 0.00017912017589237728,
      "loss": 1.5679,
      "step": 4770
    },
    {
      "epoch": 0.7504027030212549,
      "grad_norm": 0.1298915594816208,
      "learning_rate": 0.00017806946977688148,
      "loss": 1.6734,
      "step": 4775
    },
    {
      "epoch": 0.7511884650139473,
      "grad_norm": 0.12884870171546936,
      "learning_rate": 0.00017702118635759195,
      "loss": 1.5389,
      "step": 4780
    },
    {
      "epoch": 0.7519742270066397,
      "grad_norm": 0.12693621218204498,
      "learning_rate": 0.00017597533352339123,
      "loss": 1.6321,
      "step": 4785
    },
    {
      "epoch": 0.752759988999332,
      "grad_norm": 0.15435655415058136,
      "learning_rate": 0.0001749319191448712,
      "loss": 1.5707,
      "step": 4790
    },
    {
      "epoch": 0.7535457509920245,
      "grad_norm": 0.12875737249851227,
      "learning_rate": 0.00017389095107427237,
      "loss": 1.4873,
      "step": 4795
    },
    {
      "epoch": 0.7543315129847169,
      "grad_norm": 0.11512191593647003,
      "learning_rate": 0.00017285243714542593,
      "loss": 1.5503,
      "step": 4800
    },
    {
      "epoch": 0.7551172749774093,
      "grad_norm": 0.1425315886735916,
      "learning_rate": 0.0001718163851736943,
      "loss": 1.673,
      "step": 4805
    },
    {
      "epoch": 0.7559030369701017,
      "grad_norm": 0.1565292924642563,
      "learning_rate": 0.00017078280295591253,
      "loss": 1.5589,
      "step": 4810
    },
    {
      "epoch": 0.7566887989627942,
      "grad_norm": 0.12783244252204895,
      "learning_rate": 0.00016975169827032887,
      "loss": 1.5735,
      "step": 4815
    },
    {
      "epoch": 0.7574745609554866,
      "grad_norm": 0.17071731388568878,
      "learning_rate": 0.0001687230788765477,
      "loss": 1.5362,
      "step": 4820
    },
    {
      "epoch": 0.758260322948179,
      "grad_norm": 0.13069270551204681,
      "learning_rate": 0.00016769695251546947,
      "loss": 1.5619,
      "step": 4825
    },
    {
      "epoch": 0.7590460849408714,
      "grad_norm": 0.16330239176750183,
      "learning_rate": 0.0001666733269092337,
      "loss": 1.573,
      "step": 4830
    },
    {
      "epoch": 0.7598318469335639,
      "grad_norm": 0.1309252232313156,
      "learning_rate": 0.00016565220976116058,
      "loss": 1.6176,
      "step": 4835
    },
    {
      "epoch": 0.7606176089262562,
      "grad_norm": 0.14916756749153137,
      "learning_rate": 0.00016463360875569223,
      "loss": 1.5321,
      "step": 4840
    },
    {
      "epoch": 0.7614033709189486,
      "grad_norm": 0.1265818029642105,
      "learning_rate": 0.00016361753155833596,
      "loss": 1.6256,
      "step": 4845
    },
    {
      "epoch": 0.762189132911641,
      "grad_norm": 0.17490801215171814,
      "learning_rate": 0.0001626039858156062,
      "loss": 1.5349,
      "step": 4850
    },
    {
      "epoch": 0.7629748949043335,
      "grad_norm": 0.14255455136299133,
      "learning_rate": 0.0001615929791549663,
      "loss": 1.5207,
      "step": 4855
    },
    {
      "epoch": 0.7637606568970259,
      "grad_norm": 0.13435468077659607,
      "learning_rate": 0.00016058451918477223,
      "loss": 1.5818,
      "step": 4860
    },
    {
      "epoch": 0.7645464188897183,
      "grad_norm": 0.14869236946105957,
      "learning_rate": 0.00015957861349421437,
      "loss": 1.5473,
      "step": 4865
    },
    {
      "epoch": 0.7653321808824107,
      "grad_norm": 0.1301611363887787,
      "learning_rate": 0.00015857526965326108,
      "loss": 1.5311,
      "step": 4870
    },
    {
      "epoch": 0.7661179428751032,
      "grad_norm": 0.1497713327407837,
      "learning_rate": 0.00015757449521260143,
      "loss": 1.6037,
      "step": 4875
    },
    {
      "epoch": 0.7669037048677956,
      "grad_norm": 0.12209343910217285,
      "learning_rate": 0.00015657629770358838,
      "loss": 1.6328,
      "step": 4880
    },
    {
      "epoch": 0.767689466860488,
      "grad_norm": 0.13366752862930298,
      "learning_rate": 0.00015558068463818192,
      "loss": 1.6426,
      "step": 4885
    },
    {
      "epoch": 0.7684752288531804,
      "grad_norm": 0.1502147614955902,
      "learning_rate": 0.00015458766350889315,
      "loss": 1.5698,
      "step": 4890
    },
    {
      "epoch": 0.7692609908458727,
      "grad_norm": 0.13302424550056458,
      "learning_rate": 0.00015359724178872693,
      "loss": 1.5695,
      "step": 4895
    },
    {
      "epoch": 0.7700467528385652,
      "grad_norm": 0.1549658626317978,
      "learning_rate": 0.00015260942693112674,
      "loss": 1.6054,
      "step": 4900
    },
    {
      "epoch": 0.7708325148312576,
      "grad_norm": 0.15308880805969238,
      "learning_rate": 0.00015162422636991795,
      "loss": 1.5842,
      "step": 4905
    },
    {
      "epoch": 0.77161827682395,
      "grad_norm": 0.18410451710224152,
      "learning_rate": 0.00015064164751925165,
      "loss": 1.5244,
      "step": 4910
    },
    {
      "epoch": 0.7724040388166424,
      "grad_norm": 0.1323505938053131,
      "learning_rate": 0.0001496616977735496,
      "loss": 1.5401,
      "step": 4915
    },
    {
      "epoch": 0.7731898008093349,
      "grad_norm": 0.1685735434293747,
      "learning_rate": 0.00014868438450744753,
      "loss": 1.6259,
      "step": 4920
    },
    {
      "epoch": 0.7739755628020273,
      "grad_norm": 0.12531651556491852,
      "learning_rate": 0.00014770971507574137,
      "loss": 1.5469,
      "step": 4925
    },
    {
      "epoch": 0.7747613247947197,
      "grad_norm": 0.13574694097042084,
      "learning_rate": 0.00014673769681332967,
      "loss": 1.5455,
      "step": 4930
    },
    {
      "epoch": 0.7755470867874121,
      "grad_norm": 0.14784350991249084,
      "learning_rate": 0.00014576833703516034,
      "loss": 1.5476,
      "step": 4935
    },
    {
      "epoch": 0.7763328487801046,
      "grad_norm": 0.1291940063238144,
      "learning_rate": 0.0001448016430361741,
      "loss": 1.5855,
      "step": 4940
    },
    {
      "epoch": 0.7771186107727969,
      "grad_norm": 0.13253316283226013,
      "learning_rate": 0.00014383762209125095,
      "loss": 1.5273,
      "step": 4945
    },
    {
      "epoch": 0.7779043727654893,
      "grad_norm": 0.12694372236728668,
      "learning_rate": 0.00014287628145515452,
      "loss": 1.5332,
      "step": 4950
    },
    {
      "epoch": 0.7786901347581817,
      "grad_norm": 0.12803617119789124,
      "learning_rate": 0.00014191762836247734,
      "loss": 1.5241,
      "step": 4955
    },
    {
      "epoch": 0.7794758967508741,
      "grad_norm": 0.14353320002555847,
      "learning_rate": 0.00014096167002758747,
      "loss": 1.6188,
      "step": 4960
    },
    {
      "epoch": 0.7802616587435666,
      "grad_norm": 0.12240661680698395,
      "learning_rate": 0.00014000841364457267,
      "loss": 1.5319,
      "step": 4965
    },
    {
      "epoch": 0.781047420736259,
      "grad_norm": 0.11028309911489487,
      "learning_rate": 0.00013905786638718803,
      "loss": 1.5787,
      "step": 4970
    },
    {
      "epoch": 0.7818331827289514,
      "grad_norm": 0.12622590363025665,
      "learning_rate": 0.00013811003540880023,
      "loss": 1.5253,
      "step": 4975
    },
    {
      "epoch": 0.7826189447216438,
      "grad_norm": 0.1264607012271881,
      "learning_rate": 0.00013716492784233508,
      "loss": 1.5129,
      "step": 4980
    },
    {
      "epoch": 0.7834047067143363,
      "grad_norm": 0.12151068449020386,
      "learning_rate": 0.00013622255080022279,
      "loss": 1.6467,
      "step": 4985
    },
    {
      "epoch": 0.7841904687070287,
      "grad_norm": 0.15833619236946106,
      "learning_rate": 0.0001352829113743453,
      "loss": 1.5941,
      "step": 4990
    },
    {
      "epoch": 0.784976230699721,
      "grad_norm": 0.13726761937141418,
      "learning_rate": 0.00013434601663598272,
      "loss": 1.5248,
      "step": 4995
    },
    {
      "epoch": 0.7857619926924134,
      "grad_norm": 0.13140997290611267,
      "learning_rate": 0.00013341187363575936,
      "loss": 1.6044,
      "step": 5000
    },
    {
      "epoch": 0.7865477546851058,
      "grad_norm": 0.12105604261159897,
      "learning_rate": 0.0001324804894035918,
      "loss": 1.5391,
      "step": 5005
    },
    {
      "epoch": 0.7873335166777983,
      "grad_norm": 0.13176581263542175,
      "learning_rate": 0.00013155187094863498,
      "loss": 1.5185,
      "step": 5010
    },
    {
      "epoch": 0.7881192786704907,
      "grad_norm": 0.16288596391677856,
      "learning_rate": 0.00013062602525923034,
      "loss": 1.5822,
      "step": 5015
    },
    {
      "epoch": 0.7889050406631831,
      "grad_norm": 0.14374810457229614,
      "learning_rate": 0.00012970295930285274,
      "loss": 1.5183,
      "step": 5020
    },
    {
      "epoch": 0.7896908026558755,
      "grad_norm": 0.13987670838832855,
      "learning_rate": 0.00012878268002605776,
      "loss": 1.5824,
      "step": 5025
    },
    {
      "epoch": 0.790476564648568,
      "grad_norm": 0.1393355280160904,
      "learning_rate": 0.0001278651943544301,
      "loss": 1.5475,
      "step": 5030
    },
    {
      "epoch": 0.7912623266412604,
      "grad_norm": 0.14403095841407776,
      "learning_rate": 0.0001269505091925307,
      "loss": 1.5717,
      "step": 5035
    },
    {
      "epoch": 0.7920480886339528,
      "grad_norm": 0.1370304971933365,
      "learning_rate": 0.00012603863142384596,
      "loss": 1.5685,
      "step": 5040
    },
    {
      "epoch": 0.7928338506266452,
      "grad_norm": 0.12681813538074493,
      "learning_rate": 0.0001251295679107343,
      "loss": 1.5349,
      "step": 5045
    },
    {
      "epoch": 0.7936196126193376,
      "grad_norm": 0.12516118586063385,
      "learning_rate": 0.0001242233254943761,
      "loss": 1.5502,
      "step": 5050
    },
    {
      "epoch": 0.79440537461203,
      "grad_norm": 0.10329622775316238,
      "learning_rate": 0.00012331991099472084,
      "loss": 1.5565,
      "step": 5055
    },
    {
      "epoch": 0.7951911366047224,
      "grad_norm": 0.13856732845306396,
      "learning_rate": 0.00012241933121043692,
      "loss": 1.6043,
      "step": 5060
    },
    {
      "epoch": 0.7959768985974148,
      "grad_norm": 0.11811645328998566,
      "learning_rate": 0.00012152159291886012,
      "loss": 1.5341,
      "step": 5065
    },
    {
      "epoch": 0.7967626605901073,
      "grad_norm": 0.1346331089735031,
      "learning_rate": 0.0001206267028759419,
      "loss": 1.5075,
      "step": 5070
    },
    {
      "epoch": 0.7975484225827997,
      "grad_norm": 0.12564124166965485,
      "learning_rate": 0.00011973466781619984,
      "loss": 1.5282,
      "step": 5075
    },
    {
      "epoch": 0.7983341845754921,
      "grad_norm": 0.11750493198633194,
      "learning_rate": 0.00011884549445266552,
      "loss": 1.5096,
      "step": 5080
    },
    {
      "epoch": 0.7991199465681845,
      "grad_norm": 0.13426090776920319,
      "learning_rate": 0.00011795918947683576,
      "loss": 1.5373,
      "step": 5085
    },
    {
      "epoch": 0.799905708560877,
      "grad_norm": 0.14929060637950897,
      "learning_rate": 0.00011707575955862021,
      "loss": 1.5751,
      "step": 5090
    },
    {
      "epoch": 0.8006914705535694,
      "grad_norm": 0.13136523962020874,
      "learning_rate": 0.00011619521134629301,
      "loss": 1.5116,
      "step": 5095
    },
    {
      "epoch": 0.8014772325462617,
      "grad_norm": 0.14692793786525726,
      "learning_rate": 0.00011531755146644135,
      "loss": 1.5606,
      "step": 5100
    },
    {
      "epoch": 0.8022629945389541,
      "grad_norm": 0.1299007385969162,
      "learning_rate": 0.00011444278652391659,
      "loss": 1.5613,
      "step": 5105
    },
    {
      "epoch": 0.8030487565316465,
      "grad_norm": 0.1310184895992279,
      "learning_rate": 0.00011357092310178412,
      "loss": 1.5169,
      "step": 5110
    },
    {
      "epoch": 0.803834518524339,
      "grad_norm": 0.16118641197681427,
      "learning_rate": 0.00011270196776127362,
      "loss": 1.5392,
      "step": 5115
    },
    {
      "epoch": 0.8046202805170314,
      "grad_norm": 0.13983090221881866,
      "learning_rate": 0.00011183592704173029,
      "loss": 1.6329,
      "step": 5120
    },
    {
      "epoch": 0.8054060425097238,
      "grad_norm": 0.14741620421409607,
      "learning_rate": 0.00011097280746056482,
      "loss": 1.5803,
      "step": 5125
    },
    {
      "epoch": 0.8061918045024162,
      "grad_norm": 0.1304011195898056,
      "learning_rate": 0.00011011261551320528,
      "loss": 1.5256,
      "step": 5130
    },
    {
      "epoch": 0.8069775664951087,
      "grad_norm": 0.14894512295722961,
      "learning_rate": 0.00010925535767304751,
      "loss": 1.5956,
      "step": 5135
    },
    {
      "epoch": 0.8077633284878011,
      "grad_norm": 0.11353661864995956,
      "learning_rate": 0.00010840104039140681,
      "loss": 1.6066,
      "step": 5140
    },
    {
      "epoch": 0.8085490904804935,
      "grad_norm": 0.13712039589881897,
      "learning_rate": 0.0001075496700974688,
      "loss": 1.5676,
      "step": 5145
    },
    {
      "epoch": 0.8093348524731858,
      "grad_norm": 0.12836557626724243,
      "learning_rate": 0.00010670125319824203,
      "loss": 1.535,
      "step": 5150
    },
    {
      "epoch": 0.8101206144658782,
      "grad_norm": 0.12600398063659668,
      "learning_rate": 0.00010585579607850903,
      "loss": 1.6436,
      "step": 5155
    },
    {
      "epoch": 0.8109063764585707,
      "grad_norm": 0.11498034000396729,
      "learning_rate": 0.00010501330510077812,
      "loss": 1.5237,
      "step": 5160
    },
    {
      "epoch": 0.8116921384512631,
      "grad_norm": 0.1327584683895111,
      "learning_rate": 0.0001041737866052363,
      "loss": 1.4921,
      "step": 5165
    },
    {
      "epoch": 0.8124779004439555,
      "grad_norm": 0.1261293739080429,
      "learning_rate": 0.0001033372469097007,
      "loss": 1.4974,
      "step": 5170
    },
    {
      "epoch": 0.8132636624366479,
      "grad_norm": 0.1574053317308426,
      "learning_rate": 0.00010250369230957163,
      "loss": 1.5147,
      "step": 5175
    },
    {
      "epoch": 0.8140494244293404,
      "grad_norm": 0.13025632500648499,
      "learning_rate": 0.00010167312907778514,
      "loss": 1.5261,
      "step": 5180
    },
    {
      "epoch": 0.8148351864220328,
      "grad_norm": 0.13743074238300323,
      "learning_rate": 0.00010084556346476526,
      "loss": 1.546,
      "step": 5185
    },
    {
      "epoch": 0.8156209484147252,
      "grad_norm": 0.14771868288516998,
      "learning_rate": 0.0001000210016983777,
      "loss": 1.6203,
      "step": 5190
    },
    {
      "epoch": 0.8164067104074176,
      "grad_norm": 0.12153328210115433,
      "learning_rate": 9.919944998388236e-05,
      "loss": 1.492,
      "step": 5195
    },
    {
      "epoch": 0.8171924724001101,
      "grad_norm": 0.12602929770946503,
      "learning_rate": 9.83809145038872e-05,
      "loss": 1.5288,
      "step": 5200
    },
    {
      "epoch": 0.8179782343928024,
      "grad_norm": 0.11900759488344193,
      "learning_rate": 9.756540141830134e-05,
      "loss": 1.498,
      "step": 5205
    },
    {
      "epoch": 0.8187639963854948,
      "grad_norm": 0.1492881178855896,
      "learning_rate": 9.675291686428883e-05,
      "loss": 1.5234,
      "step": 5210
    },
    {
      "epoch": 0.8195497583781872,
      "grad_norm": 0.1102166399359703,
      "learning_rate": 9.59434669562222e-05,
      "loss": 1.5694,
      "step": 5215
    },
    {
      "epoch": 0.8203355203708796,
      "grad_norm": 0.12211363017559052,
      "learning_rate": 9.513705778563692e-05,
      "loss": 1.5243,
      "step": 5220
    },
    {
      "epoch": 0.8211212823635721,
      "grad_norm": 0.12994244694709778,
      "learning_rate": 9.433369542118537e-05,
      "loss": 1.5714,
      "step": 5225
    },
    {
      "epoch": 0.8219070443562645,
      "grad_norm": 0.126256063580513,
      "learning_rate": 9.353338590859078e-05,
      "loss": 1.5297,
      "step": 5230
    },
    {
      "epoch": 0.8226928063489569,
      "grad_norm": 0.15575683116912842,
      "learning_rate": 9.273613527060254e-05,
      "loss": 1.5624,
      "step": 5235
    },
    {
      "epoch": 0.8234785683416493,
      "grad_norm": 0.14137138426303864,
      "learning_rate": 9.194194950694984e-05,
      "loss": 1.5444,
      "step": 5240
    },
    {
      "epoch": 0.8242643303343418,
      "grad_norm": 0.15421825647354126,
      "learning_rate": 9.115083459429751e-05,
      "loss": 1.6643,
      "step": 5245
    },
    {
      "epoch": 0.8250500923270342,
      "grad_norm": 0.13579821586608887,
      "learning_rate": 9.03627964862005e-05,
      "loss": 1.5539,
      "step": 5250
    },
    {
      "epoch": 0.8258358543197265,
      "grad_norm": 0.10899634659290314,
      "learning_rate": 8.957784111305927e-05,
      "loss": 1.4881,
      "step": 5255
    },
    {
      "epoch": 0.8266216163124189,
      "grad_norm": 0.126531720161438,
      "learning_rate": 8.879597438207481e-05,
      "loss": 1.5383,
      "step": 5260
    },
    {
      "epoch": 0.8274073783051114,
      "grad_norm": 0.18092769384384155,
      "learning_rate": 8.801720217720488e-05,
      "loss": 1.5975,
      "step": 5265
    },
    {
      "epoch": 0.8281931402978038,
      "grad_norm": 0.16945770382881165,
      "learning_rate": 8.724153035911875e-05,
      "loss": 1.559,
      "step": 5270
    },
    {
      "epoch": 0.8289789022904962,
      "grad_norm": 0.1362890601158142,
      "learning_rate": 8.646896476515414e-05,
      "loss": 1.5776,
      "step": 5275
    },
    {
      "epoch": 0.8297646642831886,
      "grad_norm": 0.1530720740556717,
      "learning_rate": 8.569951120927272e-05,
      "loss": 1.5631,
      "step": 5280
    },
    {
      "epoch": 0.830550426275881,
      "grad_norm": 0.14527654647827148,
      "learning_rate": 8.493317548201607e-05,
      "loss": 1.5862,
      "step": 5285
    },
    {
      "epoch": 0.8313361882685735,
      "grad_norm": 0.12368170917034149,
      "learning_rate": 8.416996335046295e-05,
      "loss": 1.5584,
      "step": 5290
    },
    {
      "epoch": 0.8321219502612659,
      "grad_norm": 0.12269091606140137,
      "learning_rate": 8.340988055818522e-05,
      "loss": 1.5286,
      "step": 5295
    },
    {
      "epoch": 0.8329077122539583,
      "grad_norm": 0.11459631472826004,
      "learning_rate": 8.265293282520492e-05,
      "loss": 1.589,
      "step": 5300
    },
    {
      "epoch": 0.8336934742466506,
      "grad_norm": 0.17323504388332367,
      "learning_rate": 8.189912584795072e-05,
      "loss": 1.5524,
      "step": 5305
    },
    {
      "epoch": 0.8344792362393431,
      "grad_norm": 0.12731066346168518,
      "learning_rate": 8.11484652992161e-05,
      "loss": 1.4795,
      "step": 5310
    },
    {
      "epoch": 0.8352649982320355,
      "grad_norm": 0.12014425545930862,
      "learning_rate": 8.040095682811539e-05,
      "loss": 1.5776,
      "step": 5315
    },
    {
      "epoch": 0.8360507602247279,
      "grad_norm": 0.13413500785827637,
      "learning_rate": 7.965660606004232e-05,
      "loss": 1.5533,
      "step": 5320
    },
    {
      "epoch": 0.8368365222174203,
      "grad_norm": 0.11268212646245956,
      "learning_rate": 7.891541859662715e-05,
      "loss": 1.5351,
      "step": 5325
    },
    {
      "epoch": 0.8376222842101128,
      "grad_norm": 0.1282140612602234,
      "learning_rate": 7.81774000156944e-05,
      "loss": 1.5849,
      "step": 5330
    },
    {
      "epoch": 0.8384080462028052,
      "grad_norm": 0.14709092676639557,
      "learning_rate": 7.74425558712214e-05,
      "loss": 1.4882,
      "step": 5335
    },
    {
      "epoch": 0.8391938081954976,
      "grad_norm": 0.13272453844547272,
      "learning_rate": 7.671089169329581e-05,
      "loss": 1.5516,
      "step": 5340
    },
    {
      "epoch": 0.83997957018819,
      "grad_norm": 0.14306382834911346,
      "learning_rate": 7.598241298807479e-05,
      "loss": 1.5861,
      "step": 5345
    },
    {
      "epoch": 0.8407653321808825,
      "grad_norm": 0.17192484438419342,
      "learning_rate": 7.525712523774292e-05,
      "loss": 1.5499,
      "step": 5350
    },
    {
      "epoch": 0.8415510941735749,
      "grad_norm": 0.11474210023880005,
      "learning_rate": 7.453503390047106e-05,
      "loss": 1.5965,
      "step": 5355
    },
    {
      "epoch": 0.8423368561662672,
      "grad_norm": 0.14394626021385193,
      "learning_rate": 7.381614441037554e-05,
      "loss": 1.6175,
      "step": 5360
    },
    {
      "epoch": 0.8431226181589596,
      "grad_norm": 0.1367376148700714,
      "learning_rate": 7.310046217747717e-05,
      "loss": 1.5275,
      "step": 5365
    },
    {
      "epoch": 0.843908380151652,
      "grad_norm": 0.12612248957157135,
      "learning_rate": 7.238799258766048e-05,
      "loss": 1.5517,
      "step": 5370
    },
    {
      "epoch": 0.8446941421443445,
      "grad_norm": 0.12763731181621552,
      "learning_rate": 7.167874100263283e-05,
      "loss": 1.6903,
      "step": 5375
    },
    {
      "epoch": 0.8454799041370369,
      "grad_norm": 0.13539037108421326,
      "learning_rate": 7.097271275988482e-05,
      "loss": 1.599,
      "step": 5380
    },
    {
      "epoch": 0.8462656661297293,
      "grad_norm": 0.1278766244649887,
      "learning_rate": 7.026991317264941e-05,
      "loss": 1.5837,
      "step": 5385
    },
    {
      "epoch": 0.8470514281224217,
      "grad_norm": 0.14802652597427368,
      "learning_rate": 6.95703475298624e-05,
      "loss": 1.5944,
      "step": 5390
    },
    {
      "epoch": 0.8478371901151142,
      "grad_norm": 0.14598144590854645,
      "learning_rate": 6.887402109612262e-05,
      "loss": 1.5986,
      "step": 5395
    },
    {
      "epoch": 0.8486229521078066,
      "grad_norm": 0.17217661440372467,
      "learning_rate": 6.818093911165163e-05,
      "loss": 1.6023,
      "step": 5400
    },
    {
      "epoch": 0.849408714100499,
      "grad_norm": 0.11938291043043137,
      "learning_rate": 6.74911067922554e-05,
      "loss": 1.5861,
      "step": 5405
    },
    {
      "epoch": 0.8501944760931913,
      "grad_norm": 0.13088202476501465,
      "learning_rate": 6.680452932928382e-05,
      "loss": 1.5966,
      "step": 5410
    },
    {
      "epoch": 0.8509802380858837,
      "grad_norm": 0.1346057802438736,
      "learning_rate": 6.612121188959319e-05,
      "loss": 1.5539,
      "step": 5415
    },
    {
      "epoch": 0.8517660000785762,
      "grad_norm": 0.18001966178417206,
      "learning_rate": 6.544115961550545e-05,
      "loss": 1.4906,
      "step": 5420
    },
    {
      "epoch": 0.8525517620712686,
      "grad_norm": 0.12191446870565414,
      "learning_rate": 6.476437762477117e-05,
      "loss": 1.5589,
      "step": 5425
    },
    {
      "epoch": 0.853337524063961,
      "grad_norm": 0.1333189755678177,
      "learning_rate": 6.409087101052979e-05,
      "loss": 1.5569,
      "step": 5430
    },
    {
      "epoch": 0.8541232860566534,
      "grad_norm": 0.12484955042600632,
      "learning_rate": 6.342064484127224e-05,
      "loss": 1.5843,
      "step": 5435
    },
    {
      "epoch": 0.8549090480493459,
      "grad_norm": 0.13213855028152466,
      "learning_rate": 6.275370416080223e-05,
      "loss": 1.6526,
      "step": 5440
    },
    {
      "epoch": 0.8556948100420383,
      "grad_norm": 0.1528891623020172,
      "learning_rate": 6.209005398819828e-05,
      "loss": 1.5803,
      "step": 5445
    },
    {
      "epoch": 0.8564805720347307,
      "grad_norm": 0.10992589592933655,
      "learning_rate": 6.142969931777648e-05,
      "loss": 1.5425,
      "step": 5450
    },
    {
      "epoch": 0.8572663340274231,
      "grad_norm": 0.1314469575881958,
      "learning_rate": 6.077264511905195e-05,
      "loss": 1.5244,
      "step": 5455
    },
    {
      "epoch": 0.8580520960201156,
      "grad_norm": 0.1289691925048828,
      "learning_rate": 6.0118896336702814e-05,
      "loss": 1.5982,
      "step": 5460
    },
    {
      "epoch": 0.8588378580128079,
      "grad_norm": 0.13314056396484375,
      "learning_rate": 5.946845789053146e-05,
      "loss": 1.5215,
      "step": 5465
    },
    {
      "epoch": 0.8596236200055003,
      "grad_norm": 0.11456876993179321,
      "learning_rate": 5.882133467542888e-05,
      "loss": 1.5819,
      "step": 5470
    },
    {
      "epoch": 0.8604093819981927,
      "grad_norm": 0.11907348781824112,
      "learning_rate": 5.8177531561336596e-05,
      "loss": 1.5479,
      "step": 5475
    },
    {
      "epoch": 0.8611951439908851,
      "grad_norm": 0.12371833622455597,
      "learning_rate": 5.753705339321108e-05,
      "loss": 1.5858,
      "step": 5480
    },
    {
      "epoch": 0.8619809059835776,
      "grad_norm": 0.1354624330997467,
      "learning_rate": 5.689990499098685e-05,
      "loss": 1.4952,
      "step": 5485
    },
    {
      "epoch": 0.86276666797627,
      "grad_norm": 0.13625556230545044,
      "learning_rate": 5.626609114953973e-05,
      "loss": 1.5004,
      "step": 5490
    },
    {
      "epoch": 0.8635524299689624,
      "grad_norm": 0.14950565993785858,
      "learning_rate": 5.563561663865191e-05,
      "loss": 1.6327,
      "step": 5495
    },
    {
      "epoch": 0.8643381919616548,
      "grad_norm": 0.1316632628440857,
      "learning_rate": 5.5008486202974494e-05,
      "loss": 1.6422,
      "step": 5500
    },
    {
      "epoch": 0.8651239539543473,
      "grad_norm": 0.11560774594545364,
      "learning_rate": 5.4384704561993735e-05,
      "loss": 1.4734,
      "step": 5505
    },
    {
      "epoch": 0.8659097159470397,
      "grad_norm": 0.13010354340076447,
      "learning_rate": 5.376427640999354e-05,
      "loss": 1.6256,
      "step": 5510
    },
    {
      "epoch": 0.866695477939732,
      "grad_norm": 0.13881151378154755,
      "learning_rate": 5.3147206416021385e-05,
      "loss": 1.5763,
      "step": 5515
    },
    {
      "epoch": 0.8674812399324244,
      "grad_norm": 0.11300093680620193,
      "learning_rate": 5.253349922385298e-05,
      "loss": 1.506,
      "step": 5520
    },
    {
      "epoch": 0.8682670019251169,
      "grad_norm": 0.15863415598869324,
      "learning_rate": 5.19231594519568e-05,
      "loss": 1.535,
      "step": 5525
    },
    {
      "epoch": 0.8690527639178093,
      "grad_norm": 0.12128844112157822,
      "learning_rate": 5.13161916934603e-05,
      "loss": 1.5609,
      "step": 5530
    },
    {
      "epoch": 0.8698385259105017,
      "grad_norm": 0.13314326107501984,
      "learning_rate": 5.07126005161142e-05,
      "loss": 1.521,
      "step": 5535
    },
    {
      "epoch": 0.8706242879031941,
      "grad_norm": 0.11809918284416199,
      "learning_rate": 5.011239046225913e-05,
      "loss": 1.5759,
      "step": 5540
    },
    {
      "epoch": 0.8714100498958866,
      "grad_norm": 0.13252602517604828,
      "learning_rate": 4.9515566048790485e-05,
      "loss": 1.588,
      "step": 5545
    },
    {
      "epoch": 0.872195811888579,
      "grad_norm": 0.14337563514709473,
      "learning_rate": 4.8922131767125345e-05,
      "loss": 1.6085,
      "step": 5550
    },
    {
      "epoch": 0.8729815738812714,
      "grad_norm": 0.1566237211227417,
      "learning_rate": 4.833209208316824e-05,
      "loss": 1.5998,
      "step": 5555
    },
    {
      "epoch": 0.8737673358739638,
      "grad_norm": 0.14884966611862183,
      "learning_rate": 4.774545143727732e-05,
      "loss": 1.592,
      "step": 5560
    },
    {
      "epoch": 0.8745530978666561,
      "grad_norm": 0.13440923392772675,
      "learning_rate": 4.716221424423145e-05,
      "loss": 1.535,
      "step": 5565
    },
    {
      "epoch": 0.8753388598593486,
      "grad_norm": 0.14891453087329865,
      "learning_rate": 4.658238489319655e-05,
      "loss": 1.5672,
      "step": 5570
    },
    {
      "epoch": 0.876124621852041,
      "grad_norm": 0.13791170716285706,
      "learning_rate": 4.600596774769283e-05,
      "loss": 1.5814,
      "step": 5575
    },
    {
      "epoch": 0.8769103838447334,
      "grad_norm": 0.13698305189609528,
      "learning_rate": 4.543296714556189e-05,
      "loss": 1.5656,
      "step": 5580
    },
    {
      "epoch": 0.8776961458374258,
      "grad_norm": 0.12566827237606049,
      "learning_rate": 4.4863387398934086e-05,
      "loss": 1.5333,
      "step": 5585
    },
    {
      "epoch": 0.8784819078301183,
      "grad_norm": 0.17175309360027313,
      "learning_rate": 4.429723279419573e-05,
      "loss": 1.5809,
      "step": 5590
    },
    {
      "epoch": 0.8792676698228107,
      "grad_norm": 0.11969798803329468,
      "learning_rate": 4.3734507591957464e-05,
      "loss": 1.5349,
      "step": 5595
    },
    {
      "epoch": 0.8800534318155031,
      "grad_norm": 0.13722212612628937,
      "learning_rate": 4.31752160270219e-05,
      "loss": 1.5832,
      "step": 5600
    },
    {
      "epoch": 0.8808391938081955,
      "grad_norm": 0.14006294310092926,
      "learning_rate": 4.261936230835145e-05,
      "loss": 1.5148,
      "step": 5605
    },
    {
      "epoch": 0.881624955800888,
      "grad_norm": 0.14070965349674225,
      "learning_rate": 4.20669506190372e-05,
      "loss": 1.5549,
      "step": 5610
    },
    {
      "epoch": 0.8824107177935804,
      "grad_norm": 0.14081864058971405,
      "learning_rate": 4.151798511626698e-05,
      "loss": 1.548,
      "step": 5615
    },
    {
      "epoch": 0.8831964797862727,
      "grad_norm": 0.1195095032453537,
      "learning_rate": 4.097246993129428e-05,
      "loss": 1.6033,
      "step": 5620
    },
    {
      "epoch": 0.8839822417789651,
      "grad_norm": 0.1440693438053131,
      "learning_rate": 4.043040916940727e-05,
      "loss": 1.5329,
      "step": 5625
    },
    {
      "epoch": 0.8847680037716575,
      "grad_norm": 0.12744715809822083,
      "learning_rate": 3.989180690989774e-05,
      "loss": 1.6177,
      "step": 5630
    },
    {
      "epoch": 0.88555376576435,
      "grad_norm": 0.17127704620361328,
      "learning_rate": 3.935666720603026e-05,
      "loss": 1.5456,
      "step": 5635
    },
    {
      "epoch": 0.8863395277570424,
      "grad_norm": 0.11687444895505905,
      "learning_rate": 3.8824994085012054e-05,
      "loss": 1.5794,
      "step": 5640
    },
    {
      "epoch": 0.8871252897497348,
      "grad_norm": 0.12178084254264832,
      "learning_rate": 3.829679154796228e-05,
      "loss": 1.554,
      "step": 5645
    },
    {
      "epoch": 0.8879110517424272,
      "grad_norm": 0.13413095474243164,
      "learning_rate": 3.777206356988239e-05,
      "loss": 1.5518,
      "step": 5650
    },
    {
      "epoch": 0.8886968137351197,
      "grad_norm": 0.1272350400686264,
      "learning_rate": 3.725081409962583e-05,
      "loss": 1.5592,
      "step": 5655
    },
    {
      "epoch": 0.8894825757278121,
      "grad_norm": 0.1570257842540741,
      "learning_rate": 3.67330470598683e-05,
      "loss": 1.6286,
      "step": 5660
    },
    {
      "epoch": 0.8902683377205045,
      "grad_norm": 0.12927798926830292,
      "learning_rate": 3.62187663470786e-05,
      "loss": 1.5626,
      "step": 5665
    },
    {
      "epoch": 0.8910540997131968,
      "grad_norm": 0.13783001899719238,
      "learning_rate": 3.570797583148916e-05,
      "loss": 1.5048,
      "step": 5670
    },
    {
      "epoch": 0.8918398617058892,
      "grad_norm": 0.13481375575065613,
      "learning_rate": 3.52006793570665e-05,
      "loss": 1.5288,
      "step": 5675
    },
    {
      "epoch": 0.8926256236985817,
      "grad_norm": 0.15029197931289673,
      "learning_rate": 3.4696880741482974e-05,
      "loss": 1.549,
      "step": 5680
    },
    {
      "epoch": 0.8934113856912741,
      "grad_norm": 0.14143696427345276,
      "learning_rate": 3.419658377608748e-05,
      "loss": 1.5579,
      "step": 5685
    },
    {
      "epoch": 0.8941971476839665,
      "grad_norm": 0.11702083051204681,
      "learning_rate": 3.369979222587727e-05,
      "loss": 1.579,
      "step": 5690
    },
    {
      "epoch": 0.8949829096766589,
      "grad_norm": 0.15968121588230133,
      "learning_rate": 3.320650982946954e-05,
      "loss": 1.6591,
      "step": 5695
    },
    {
      "epoch": 0.8957686716693514,
      "grad_norm": 0.13863512873649597,
      "learning_rate": 3.271674029907334e-05,
      "loss": 1.5756,
      "step": 5700
    },
    {
      "epoch": 0.8965544336620438,
      "grad_norm": 0.11312685161828995,
      "learning_rate": 3.223048732046124e-05,
      "loss": 1.5231,
      "step": 5705
    },
    {
      "epoch": 0.8973401956547362,
      "grad_norm": 0.13332636654376984,
      "learning_rate": 3.1747754552942324e-05,
      "loss": 1.6331,
      "step": 5710
    },
    {
      "epoch": 0.8981259576474286,
      "grad_norm": 0.1321624368429184,
      "learning_rate": 3.126854562933379e-05,
      "loss": 1.5423,
      "step": 5715
    },
    {
      "epoch": 0.898911719640121,
      "grad_norm": 0.12722861766815186,
      "learning_rate": 3.079286415593441e-05,
      "loss": 1.5334,
      "step": 5720
    },
    {
      "epoch": 0.8996974816328134,
      "grad_norm": 0.12315835803747177,
      "learning_rate": 3.0320713712497062e-05,
      "loss": 1.6664,
      "step": 5725
    },
    {
      "epoch": 0.9004832436255058,
      "grad_norm": 0.20989854633808136,
      "learning_rate": 2.98520978522015e-05,
      "loss": 1.6118,
      "step": 5730
    },
    {
      "epoch": 0.9012690056181982,
      "grad_norm": 0.12240595370531082,
      "learning_rate": 2.938702010162797e-05,
      "loss": 1.541,
      "step": 5735
    },
    {
      "epoch": 0.9020547676108907,
      "grad_norm": 0.1253134161233902,
      "learning_rate": 2.8925483960730803e-05,
      "loss": 1.5571,
      "step": 5740
    },
    {
      "epoch": 0.9028405296035831,
      "grad_norm": 0.12602074444293976,
      "learning_rate": 2.84674929028117e-05,
      "loss": 1.6147,
      "step": 5745
    },
    {
      "epoch": 0.9036262915962755,
      "grad_norm": 0.14237385988235474,
      "learning_rate": 2.801305037449353e-05,
      "loss": 1.5694,
      "step": 5750
    },
    {
      "epoch": 0.9044120535889679,
      "grad_norm": 0.13480807840824127,
      "learning_rate": 2.7562159795695097e-05,
      "loss": 1.5912,
      "step": 5755
    },
    {
      "epoch": 0.9051978155816603,
      "grad_norm": 0.13586966693401337,
      "learning_rate": 2.7114824559604513e-05,
      "loss": 1.552,
      "step": 5760
    },
    {
      "epoch": 0.9059835775743528,
      "grad_norm": 0.11657349020242691,
      "learning_rate": 2.6671048032654188e-05,
      "loss": 1.6195,
      "step": 5765
    },
    {
      "epoch": 0.9067693395670452,
      "grad_norm": 0.13299603760242462,
      "learning_rate": 2.623083355449557e-05,
      "loss": 1.525,
      "step": 5770
    },
    {
      "epoch": 0.9075551015597375,
      "grad_norm": 0.15020130574703217,
      "learning_rate": 2.5794184437973433e-05,
      "loss": 1.5987,
      "step": 5775
    },
    {
      "epoch": 0.9083408635524299,
      "grad_norm": 0.12535901367664337,
      "learning_rate": 2.536110396910174e-05,
      "loss": 1.671,
      "step": 5780
    },
    {
      "epoch": 0.9091266255451224,
      "grad_norm": 0.13589003682136536,
      "learning_rate": 2.4931595407038098e-05,
      "loss": 1.6108,
      "step": 5785
    },
    {
      "epoch": 0.9099123875378148,
      "grad_norm": 0.14382244646549225,
      "learning_rate": 2.450566198406018e-05,
      "loss": 1.5919,
      "step": 5790
    },
    {
      "epoch": 0.9106981495305072,
      "grad_norm": 0.12787427008152008,
      "learning_rate": 2.4083306905540335e-05,
      "loss": 1.5341,
      "step": 5795
    },
    {
      "epoch": 0.9114839115231996,
      "grad_norm": 0.12778525054454803,
      "learning_rate": 2.36645333499223e-05,
      "loss": 1.5093,
      "step": 5800
    },
    {
      "epoch": 0.912269673515892,
      "grad_norm": 0.14423970878124237,
      "learning_rate": 2.3249344468696755e-05,
      "loss": 1.5012,
      "step": 5805
    },
    {
      "epoch": 0.9130554355085845,
      "grad_norm": 0.1335286796092987,
      "learning_rate": 2.2837743386378007e-05,
      "loss": 1.529,
      "step": 5810
    },
    {
      "epoch": 0.9138411975012769,
      "grad_norm": 0.12632855772972107,
      "learning_rate": 2.2429733200480308e-05,
      "loss": 1.5543,
      "step": 5815
    },
    {
      "epoch": 0.9146269594939693,
      "grad_norm": 0.11702293157577515,
      "learning_rate": 2.2025316981494347e-05,
      "loss": 1.6215,
      "step": 5820
    },
    {
      "epoch": 0.9154127214866616,
      "grad_norm": 0.13728240132331848,
      "learning_rate": 2.1624497772864514e-05,
      "loss": 1.551,
      "step": 5825
    },
    {
      "epoch": 0.9161984834793541,
      "grad_norm": 0.1302066147327423,
      "learning_rate": 2.1227278590965572e-05,
      "loss": 1.5828,
      "step": 5830
    },
    {
      "epoch": 0.9169842454720465,
      "grad_norm": 0.1424468457698822,
      "learning_rate": 2.0833662425080557e-05,
      "loss": 1.6033,
      "step": 5835
    },
    {
      "epoch": 0.9177700074647389,
      "grad_norm": 0.15224626660346985,
      "learning_rate": 2.0443652237377596e-05,
      "loss": 1.6329,
      "step": 5840
    },
    {
      "epoch": 0.9185557694574313,
      "grad_norm": 0.1158602312207222,
      "learning_rate": 2.0057250962887964e-05,
      "loss": 1.5337,
      "step": 5845
    },
    {
      "epoch": 0.9193415314501238,
      "grad_norm": 0.145168736577034,
      "learning_rate": 1.9674461509484266e-05,
      "loss": 1.5735,
      "step": 5850
    },
    {
      "epoch": 0.9201272934428162,
      "grad_norm": 0.14250454306602478,
      "learning_rate": 1.92952867578578e-05,
      "loss": 1.5675,
      "step": 5855
    },
    {
      "epoch": 0.9209130554355086,
      "grad_norm": 0.1341368407011032,
      "learning_rate": 1.891972956149779e-05,
      "loss": 1.6223,
      "step": 5860
    },
    {
      "epoch": 0.921698817428201,
      "grad_norm": 0.12028190493583679,
      "learning_rate": 1.854779274666918e-05,
      "loss": 1.5513,
      "step": 5865
    },
    {
      "epoch": 0.9224845794208935,
      "grad_norm": 0.13998378813266754,
      "learning_rate": 1.8179479112391704e-05,
      "loss": 1.5942,
      "step": 5870
    },
    {
      "epoch": 0.9232703414135858,
      "grad_norm": 0.1285070776939392,
      "learning_rate": 1.781479143041875e-05,
      "loss": 1.5804,
      "step": 5875
    },
    {
      "epoch": 0.9240561034062782,
      "grad_norm": 0.10047586262226105,
      "learning_rate": 1.7453732445216586e-05,
      "loss": 1.5314,
      "step": 5880
    },
    {
      "epoch": 0.9248418653989706,
      "grad_norm": 0.1180456280708313,
      "learning_rate": 1.7096304873943535e-05,
      "loss": 1.5613,
      "step": 5885
    },
    {
      "epoch": 0.925627627391663,
      "grad_norm": 0.1225186139345169,
      "learning_rate": 1.6742511406429682e-05,
      "loss": 1.5305,
      "step": 5890
    },
    {
      "epoch": 0.9264133893843555,
      "grad_norm": 0.13830195367336273,
      "learning_rate": 1.639235470515654e-05,
      "loss": 1.5563,
      "step": 5895
    },
    {
      "epoch": 0.9271991513770479,
      "grad_norm": 0.17313812673091888,
      "learning_rate": 1.6045837405237075e-05,
      "loss": 1.4795,
      "step": 5900
    },
    {
      "epoch": 0.9279849133697403,
      "grad_norm": 0.12217851728200912,
      "learning_rate": 1.570296211439587e-05,
      "loss": 1.5596,
      "step": 5905
    },
    {
      "epoch": 0.9287706753624327,
      "grad_norm": 0.11904492974281311,
      "learning_rate": 1.5363731412949444e-05,
      "loss": 1.6062,
      "step": 5910
    },
    {
      "epoch": 0.9295564373551252,
      "grad_norm": 0.13460589945316315,
      "learning_rate": 1.5028147853786867e-05,
      "loss": 1.5486,
      "step": 5915
    },
    {
      "epoch": 0.9303421993478176,
      "grad_norm": 0.10764899104833603,
      "learning_rate": 1.4696213962350492e-05,
      "loss": 1.5923,
      "step": 5920
    },
    {
      "epoch": 0.93112796134051,
      "grad_norm": 0.12785036861896515,
      "learning_rate": 1.4367932236617142e-05,
      "loss": 1.6108,
      "step": 5925
    },
    {
      "epoch": 0.9319137233332023,
      "grad_norm": 0.11170931160449982,
      "learning_rate": 1.4043305147079077e-05,
      "loss": 1.4704,
      "step": 5930
    },
    {
      "epoch": 0.9326994853258948,
      "grad_norm": 0.200826033949852,
      "learning_rate": 1.372233513672555e-05,
      "loss": 1.6068,
      "step": 5935
    },
    {
      "epoch": 0.9334852473185872,
      "grad_norm": 0.11707545816898346,
      "learning_rate": 1.3405024621024332e-05,
      "loss": 1.5703,
      "step": 5940
    },
    {
      "epoch": 0.9342710093112796,
      "grad_norm": 0.11969916522502899,
      "learning_rate": 1.3091375987903498e-05,
      "loss": 1.5001,
      "step": 5945
    },
    {
      "epoch": 0.935056771303972,
      "grad_norm": 0.13447532057762146,
      "learning_rate": 1.2781391597733837e-05,
      "loss": 1.5555,
      "step": 5950
    },
    {
      "epoch": 0.9358425332966644,
      "grad_norm": 0.13763001561164856,
      "learning_rate": 1.247507378331042e-05,
      "loss": 1.5385,
      "step": 5955
    },
    {
      "epoch": 0.9366282952893569,
      "grad_norm": 0.13508008420467377,
      "learning_rate": 1.2172424849835662e-05,
      "loss": 1.555,
      "step": 5960
    },
    {
      "epoch": 0.9374140572820493,
      "grad_norm": 0.122609443962574,
      "learning_rate": 1.1873447074901512e-05,
      "loss": 1.5754,
      "step": 5965
    },
    {
      "epoch": 0.9381998192747417,
      "grad_norm": 0.12040087580680847,
      "learning_rate": 1.1578142708472794e-05,
      "loss": 1.4984,
      "step": 5970
    },
    {
      "epoch": 0.9389855812674341,
      "grad_norm": 0.14471302926540375,
      "learning_rate": 1.1286513972869783e-05,
      "loss": 1.5293,
      "step": 5975
    },
    {
      "epoch": 0.9397713432601265,
      "grad_norm": 0.12743833661079407,
      "learning_rate": 1.0998563062751821e-05,
      "loss": 1.553,
      "step": 5980
    },
    {
      "epoch": 0.9405571052528189,
      "grad_norm": 0.12718817591667175,
      "learning_rate": 1.0714292145100557e-05,
      "loss": 1.5326,
      "step": 5985
    },
    {
      "epoch": 0.9413428672455113,
      "grad_norm": 0.1280830055475235,
      "learning_rate": 1.0433703359203906e-05,
      "loss": 1.5474,
      "step": 5990
    },
    {
      "epoch": 0.9421286292382037,
      "grad_norm": 0.13597886264324188,
      "learning_rate": 1.0156798816639723e-05,
      "loss": 1.564,
      "step": 5995
    },
    {
      "epoch": 0.9429143912308962,
      "grad_norm": 0.14008958637714386,
      "learning_rate": 9.883580601260044e-06,
      "loss": 1.6469,
      "step": 6000
    },
    {
      "epoch": 0.9437001532235886,
      "grad_norm": 0.12624593079090118,
      "learning_rate": 9.61405076917532e-06,
      "loss": 1.5435,
      "step": 6005
    },
    {
      "epoch": 0.944485915216281,
      "grad_norm": 0.1182253360748291,
      "learning_rate": 9.348211348739033e-06,
      "loss": 1.6148,
      "step": 6010
    },
    {
      "epoch": 0.9452716772089734,
      "grad_norm": 0.12373318523168564,
      "learning_rate": 9.086064340532274e-06,
      "loss": 1.6464,
      "step": 6015
    },
    {
      "epoch": 0.9460574392016659,
      "grad_norm": 0.1271478533744812,
      "learning_rate": 8.827611717349027e-06,
      "loss": 1.5245,
      "step": 6020
    },
    {
      "epoch": 0.9468432011943583,
      "grad_norm": 0.12978233397006989,
      "learning_rate": 8.572855424180736e-06,
      "loss": 1.5904,
      "step": 6025
    },
    {
      "epoch": 0.9476289631870506,
      "grad_norm": 0.14304792881011963,
      "learning_rate": 8.321797378202378e-06,
      "loss": 1.5159,
      "step": 6030
    },
    {
      "epoch": 0.948414725179743,
      "grad_norm": 0.1286846399307251,
      "learning_rate": 8.074439468757412e-06,
      "loss": 1.5156,
      "step": 6035
    },
    {
      "epoch": 0.9492004871724354,
      "grad_norm": 0.12985752522945404,
      "learning_rate": 7.830783557343901e-06,
      "loss": 1.5533,
      "step": 6040
    },
    {
      "epoch": 0.9499862491651279,
      "grad_norm": 0.11546095460653305,
      "learning_rate": 7.590831477600646e-06,
      "loss": 1.4813,
      "step": 6045
    },
    {
      "epoch": 0.9507720111578203,
      "grad_norm": 0.13258661329746246,
      "learning_rate": 7.354585035292794e-06,
      "loss": 1.6366,
      "step": 6050
    },
    {
      "epoch": 0.9515577731505127,
      "grad_norm": 0.1464798003435135,
      "learning_rate": 7.122046008298966e-06,
      "loss": 1.5343,
      "step": 6055
    },
    {
      "epoch": 0.9523435351432051,
      "grad_norm": 0.11942831426858902,
      "learning_rate": 6.8932161465972694e-06,
      "loss": 1.548,
      "step": 6060
    },
    {
      "epoch": 0.9531292971358976,
      "grad_norm": 0.11840633302927017,
      "learning_rate": 6.668097172252529e-06,
      "loss": 1.63,
      "step": 6065
    },
    {
      "epoch": 0.95391505912859,
      "grad_norm": 0.12097301334142685,
      "learning_rate": 6.446690779403241e-06,
      "loss": 1.5491,
      "step": 6070
    },
    {
      "epoch": 0.9547008211212824,
      "grad_norm": 0.12309488654136658,
      "learning_rate": 6.228998634248695e-06,
      "loss": 1.598,
      "step": 6075
    },
    {
      "epoch": 0.9554865831139748,
      "grad_norm": 0.14806729555130005,
      "learning_rate": 6.015022375036539e-06,
      "loss": 1.5416,
      "step": 6080
    },
    {
      "epoch": 0.9562723451066671,
      "grad_norm": 0.12924478948116302,
      "learning_rate": 5.8047636120504015e-06,
      "loss": 1.5189,
      "step": 6085
    },
    {
      "epoch": 0.9570581070993596,
      "grad_norm": 0.13043110072612762,
      "learning_rate": 5.598223927597901e-06,
      "loss": 1.67,
      "step": 6090
    },
    {
      "epoch": 0.957843869092052,
      "grad_norm": 0.1203450858592987,
      "learning_rate": 5.395404875998488e-06,
      "loss": 1.4862,
      "step": 6095
    },
    {
      "epoch": 0.9586296310847444,
      "grad_norm": 0.1190691590309143,
      "learning_rate": 5.196307983572069e-06,
      "loss": 1.557,
      "step": 6100
    },
    {
      "epoch": 0.9594153930774368,
      "grad_norm": 0.1312546730041504,
      "learning_rate": 5.000934748627117e-06,
      "loss": 1.5992,
      "step": 6105
    },
    {
      "epoch": 0.9602011550701293,
      "grad_norm": 0.12697143852710724,
      "learning_rate": 4.809286641449862e-06,
      "loss": 1.5766,
      "step": 6110
    },
    {
      "epoch": 0.9609869170628217,
      "grad_norm": 0.11688420176506042,
      "learning_rate": 4.621365104292896e-06,
      "loss": 1.5383,
      "step": 6115
    },
    {
      "epoch": 0.9617726790555141,
      "grad_norm": 0.15452660620212555,
      "learning_rate": 4.437171551364416e-06,
      "loss": 1.6561,
      "step": 6120
    },
    {
      "epoch": 0.9625584410482065,
      "grad_norm": 0.14491859078407288,
      "learning_rate": 4.256707368817503e-06,
      "loss": 1.5707,
      "step": 6125
    },
    {
      "epoch": 0.963344203040899,
      "grad_norm": 0.14828036725521088,
      "learning_rate": 4.079973914739965e-06,
      "loss": 1.5733,
      "step": 6130
    },
    {
      "epoch": 0.9641299650335913,
      "grad_norm": 0.11809247732162476,
      "learning_rate": 3.906972519143736e-06,
      "loss": 1.5501,
      "step": 6135
    },
    {
      "epoch": 0.9649157270262837,
      "grad_norm": 0.11582066118717194,
      "learning_rate": 3.737704483955107e-06,
      "loss": 1.5558,
      "step": 6140
    },
    {
      "epoch": 0.9657014890189761,
      "grad_norm": 0.1349029392004013,
      "learning_rate": 3.5721710830049493e-06,
      "loss": 1.5224,
      "step": 6145
    },
    {
      "epoch": 0.9664872510116685,
      "grad_norm": 0.14910638332366943,
      "learning_rate": 3.4103735620189536e-06,
      "loss": 1.5653,
      "step": 6150
    },
    {
      "epoch": 0.967273013004361,
      "grad_norm": 0.11077561974525452,
      "learning_rate": 3.2523131386083537e-06,
      "loss": 1.5118,
      "step": 6155
    },
    {
      "epoch": 0.9680587749970534,
      "grad_norm": 0.12212218344211578,
      "learning_rate": 3.0979910022607693e-06,
      "loss": 1.5892,
      "step": 6160
    },
    {
      "epoch": 0.9688445369897458,
      "grad_norm": 0.1238313764333725,
      "learning_rate": 2.947408314331379e-06,
      "loss": 1.5932,
      "step": 6165
    },
    {
      "epoch": 0.9696302989824382,
      "grad_norm": 0.128588005900383,
      "learning_rate": 2.8005662080339276e-06,
      "loss": 1.5289,
      "step": 6170
    },
    {
      "epoch": 0.9704160609751307,
      "grad_norm": 0.12846864759922028,
      "learning_rate": 2.657465788432234e-06,
      "loss": 1.5869,
      "step": 6175
    },
    {
      "epoch": 0.9712018229678231,
      "grad_norm": 0.13134929537773132,
      "learning_rate": 2.518108132432084e-06,
      "loss": 1.5328,
      "step": 6180
    },
    {
      "epoch": 0.9719875849605154,
      "grad_norm": 0.13864241540431976,
      "learning_rate": 2.382494288773074e-06,
      "loss": 1.6102,
      "step": 6185
    },
    {
      "epoch": 0.9727733469532078,
      "grad_norm": 0.12534551322460175,
      "learning_rate": 2.2506252780205017e-06,
      "loss": 1.5703,
      "step": 6190
    },
    {
      "epoch": 0.9735591089459003,
      "grad_norm": 0.11990083009004593,
      "learning_rate": 2.122502092557876e-06,
      "loss": 1.6265,
      "step": 6195
    },
    {
      "epoch": 0.9743448709385927,
      "grad_norm": 0.11616454273462296,
      "learning_rate": 1.9981256965794226e-06,
      "loss": 1.5141,
      "step": 6200
    },
    {
      "epoch": 0.9751306329312851,
      "grad_norm": 0.12967850267887115,
      "learning_rate": 1.8774970260827528e-06,
      "loss": 1.5729,
      "step": 6205
    },
    {
      "epoch": 0.9759163949239775,
      "grad_norm": 0.13299420475959778,
      "learning_rate": 1.760616988861985e-06,
      "loss": 1.5698,
      "step": 6210
    },
    {
      "epoch": 0.97670215691667,
      "grad_norm": 0.12787404656410217,
      "learning_rate": 1.6474864645008024e-06,
      "loss": 1.5025,
      "step": 6215
    },
    {
      "epoch": 0.9774879189093624,
      "grad_norm": 0.11461758613586426,
      "learning_rate": 1.5381063043656274e-06,
      "loss": 1.5551,
      "step": 6220
    },
    {
      "epoch": 0.9782736809020548,
      "grad_norm": 0.11087030172348022,
      "learning_rate": 1.4324773315996797e-06,
      "loss": 1.5443,
      "step": 6225
    },
    {
      "epoch": 0.9790594428947472,
      "grad_norm": 0.13169339299201965,
      "learning_rate": 1.3306003411163725e-06,
      "loss": 1.5355,
      "step": 6230
    },
    {
      "epoch": 0.9798452048874396,
      "grad_norm": 0.14098110795021057,
      "learning_rate": 1.232476099593538e-06,
      "loss": 1.5563,
      "step": 6235
    },
    {
      "epoch": 0.980630966880132,
      "grad_norm": 0.1428086906671524,
      "learning_rate": 1.1381053454675994e-06,
      "loss": 1.5886,
      "step": 6240
    },
    {
      "epoch": 0.9814167288728244,
      "grad_norm": 0.11641329526901245,
      "learning_rate": 1.0474887889280749e-06,
      "loss": 1.5523,
      "step": 6245
    },
    {
      "epoch": 0.9822024908655168,
      "grad_norm": 0.11133892089128494,
      "learning_rate": 9.606271119119713e-07,
      "loss": 1.5447,
      "step": 6250
    },
    {
      "epoch": 0.9829882528582092,
      "grad_norm": 0.11940602213144302,
      "learning_rate": 8.775209680991214e-07,
      "loss": 1.5727,
      "step": 6255
    },
    {
      "epoch": 0.9837740148509017,
      "grad_norm": 0.12904313206672668,
      "learning_rate": 7.981709829068539e-07,
      "loss": 1.5067,
      "step": 6260
    },
    {
      "epoch": 0.9845597768435941,
      "grad_norm": 0.15031477808952332,
      "learning_rate": 7.225777534854428e-07,
      "loss": 1.549,
      "step": 6265
    },
    {
      "epoch": 0.9853455388362865,
      "grad_norm": 0.12052863091230392,
      "learning_rate": 6.507418487135542e-07,
      "loss": 1.5189,
      "step": 6270
    },
    {
      "epoch": 0.9861313008289789,
      "grad_norm": 0.1250990927219391,
      "learning_rate": 5.826638091941394e-07,
      "loss": 1.585,
      "step": 6275
    },
    {
      "epoch": 0.9869170628216714,
      "grad_norm": 0.14981551468372345,
      "learning_rate": 5.18344147250216e-07,
      "loss": 1.5447,
      "step": 6280
    },
    {
      "epoch": 0.9877028248143638,
      "grad_norm": 0.12979185581207275,
      "learning_rate": 4.577833469208703e-07,
      "loss": 1.6224,
      "step": 6285
    },
    {
      "epoch": 0.9884885868070561,
      "grad_norm": 0.1247519999742508,
      "learning_rate": 4.009818639580387e-07,
      "loss": 1.5269,
      "step": 6290
    },
    {
      "epoch": 0.9892743487997485,
      "grad_norm": 0.11512382328510284,
      "learning_rate": 3.4794012582262115e-07,
      "loss": 1.5008,
      "step": 6295
    },
    {
      "epoch": 0.9900601107924409,
      "grad_norm": 0.13370949029922485,
      "learning_rate": 2.9865853168159483e-07,
      "loss": 1.5376,
      "step": 6300
    },
    {
      "epoch": 0.9908458727851334,
      "grad_norm": 0.14563091099262238,
      "learning_rate": 2.5313745240473915e-07,
      "loss": 1.5473,
      "step": 6305
    },
    {
      "epoch": 0.9916316347778258,
      "grad_norm": 0.12514719367027283,
      "learning_rate": 2.1137723056213753e-07,
      "loss": 1.5579,
      "step": 6310
    },
    {
      "epoch": 0.9924173967705182,
      "grad_norm": 0.14008943736553192,
      "learning_rate": 1.7337818042134635e-07,
      "loss": 1.5233,
      "step": 6315
    },
    {
      "epoch": 0.9932031587632106,
      "grad_norm": 0.11884216964244843,
      "learning_rate": 1.3914058794511907e-07,
      "loss": 1.5463,
      "step": 6320
    },
    {
      "epoch": 0.9939889207559031,
      "grad_norm": 0.13800643384456635,
      "learning_rate": 1.0866471078940787e-07,
      "loss": 1.5987,
      "step": 6325
    },
    {
      "epoch": 0.9947746827485955,
      "grad_norm": 0.11561693996191025,
      "learning_rate": 8.195077830114306e-08,
      "loss": 1.539,
      "step": 6330
    },
    {
      "epoch": 0.9955604447412879,
      "grad_norm": 0.1500178575515747,
      "learning_rate": 5.8998991516678866e-08,
      "loss": 1.5477,
      "step": 6335
    },
    {
      "epoch": 0.9963462067339802,
      "grad_norm": 0.1255454123020172,
      "learning_rate": 3.9809523160183606e-08,
      "loss": 1.5781,
      "step": 6340
    },
    {
      "epoch": 0.9971319687266726,
      "grad_norm": 0.18067841231822968,
      "learning_rate": 2.4382517642640435e-08,
      "loss": 1.4978,
      "step": 6345
    },
    {
      "epoch": 0.9979177307193651,
      "grad_norm": 0.15813903510570526,
      "learning_rate": 1.2718091060182069e-08,
      "loss": 1.5221,
      "step": 6350
    },
    {
      "epoch": 0.9987034927120575,
      "grad_norm": 0.1291486918926239,
      "learning_rate": 4.816331193924217e-09,
      "loss": 1.6245,
      "step": 6355
    },
    {
      "epoch": 0.9994892547047499,
      "grad_norm": 0.11886509507894516,
      "learning_rate": 6.772975085778121e-10,
      "loss": 1.6374,
      "step": 6360
    },
    {
      "epoch": 0.9999607119003654,
      "eval_loss": 1.557084083557129,
      "eval_runtime": 120.519,
      "eval_samples_per_second": 88.376,
      "eval_steps_per_second": 11.052,
      "step": 6363
    },
    {
      "epoch": 0.9999607119003654,
      "step": 6363,
      "total_flos": 2.657246056189788e+17,
      "train_loss": 1.732499259158573,
      "train_runtime": 9418.2953,
      "train_samples_per_second": 21.62,
      "train_steps_per_second": 0.676
    }
  ],
  "logging_steps": 5,
  "max_steps": 6363,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.657246056189788e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}