{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 14950,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006688963210702341,
      "grad_norm": 5.731072425842285,
      "learning_rate": 9.993311036789297e-05,
      "loss": 0.6793,
      "step": 10
    },
    {
      "epoch": 0.013377926421404682,
      "grad_norm": 9.623015403747559,
      "learning_rate": 9.986622073578595e-05,
      "loss": 0.1186,
      "step": 20
    },
    {
      "epoch": 0.020066889632107024,
      "grad_norm": 13.354948043823242,
      "learning_rate": 9.979933110367894e-05,
      "loss": 0.1486,
      "step": 30
    },
    {
      "epoch": 0.026755852842809364,
      "grad_norm": 5.970050811767578,
      "learning_rate": 9.973244147157192e-05,
      "loss": 0.1379,
      "step": 40
    },
    {
      "epoch": 0.033444816053511704,
      "grad_norm": 6.1483354568481445,
      "learning_rate": 9.966555183946489e-05,
      "loss": 0.0901,
      "step": 50
    },
    {
      "epoch": 0.04013377926421405,
      "grad_norm": 1.7410144805908203,
      "learning_rate": 9.959866220735787e-05,
      "loss": 0.1011,
      "step": 60
    },
    {
      "epoch": 0.046822742474916385,
      "grad_norm": 10.210768699645996,
      "learning_rate": 9.953177257525084e-05,
      "loss": 0.1327,
      "step": 70
    },
    {
      "epoch": 0.05351170568561873,
      "grad_norm": 8.604742050170898,
      "learning_rate": 9.946488294314382e-05,
      "loss": 0.1222,
      "step": 80
    },
    {
      "epoch": 0.06020066889632107,
      "grad_norm": 8.2460355758667,
      "learning_rate": 9.93979933110368e-05,
      "loss": 0.12,
      "step": 90
    },
    {
      "epoch": 0.06688963210702341,
      "grad_norm": 4.913175582885742,
      "learning_rate": 9.933110367892977e-05,
      "loss": 0.1165,
      "step": 100
    },
    {
      "epoch": 0.07357859531772576,
      "grad_norm": 3.4663991928100586,
      "learning_rate": 9.926421404682275e-05,
      "loss": 0.1905,
      "step": 110
    },
    {
      "epoch": 0.0802675585284281,
      "grad_norm": 3.857680082321167,
      "learning_rate": 9.919732441471572e-05,
      "loss": 0.0881,
      "step": 120
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 10.378622055053711,
      "learning_rate": 9.91304347826087e-05,
      "loss": 0.0997,
      "step": 130
    },
    {
      "epoch": 0.09364548494983277,
      "grad_norm": 1.7943992614746094,
      "learning_rate": 9.906354515050168e-05,
      "loss": 0.0938,
      "step": 140
    },
    {
      "epoch": 0.10033444816053512,
      "grad_norm": 3.0998451709747314,
      "learning_rate": 9.899665551839465e-05,
      "loss": 0.0681,
      "step": 150
    },
    {
      "epoch": 0.10702341137123746,
      "grad_norm": 5.554123878479004,
      "learning_rate": 9.892976588628763e-05,
      "loss": 0.0726,
      "step": 160
    },
    {
      "epoch": 0.11371237458193979,
      "grad_norm": 1.8740187883377075,
      "learning_rate": 9.88628762541806e-05,
      "loss": 0.0887,
      "step": 170
    },
    {
      "epoch": 0.12040133779264214,
      "grad_norm": 9.424790382385254,
      "learning_rate": 9.879598662207358e-05,
      "loss": 0.1184,
      "step": 180
    },
    {
      "epoch": 0.12709030100334448,
      "grad_norm": 0.5133495330810547,
      "learning_rate": 9.872909698996656e-05,
      "loss": 0.0907,
      "step": 190
    },
    {
      "epoch": 0.13377926421404682,
      "grad_norm": 12.215278625488281,
      "learning_rate": 9.866220735785953e-05,
      "loss": 0.138,
      "step": 200
    },
    {
      "epoch": 0.14046822742474915,
      "grad_norm": 14.582442283630371,
      "learning_rate": 9.859531772575251e-05,
      "loss": 0.1946,
      "step": 210
    },
    {
      "epoch": 0.14715719063545152,
      "grad_norm": 2.9713079929351807,
      "learning_rate": 9.852842809364548e-05,
      "loss": 0.1018,
      "step": 220
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 2.692634344100952,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.1059,
      "step": 230
    },
    {
      "epoch": 0.1605351170568562,
      "grad_norm": 3.160918712615967,
      "learning_rate": 9.839464882943145e-05,
      "loss": 0.1153,
      "step": 240
    },
    {
      "epoch": 0.16722408026755853,
      "grad_norm": 4.2426557540893555,
      "learning_rate": 9.832775919732441e-05,
      "loss": 0.0746,
      "step": 250
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 0.8495791554450989,
      "learning_rate": 9.82608695652174e-05,
      "loss": 0.087,
      "step": 260
    },
    {
      "epoch": 0.1806020066889632,
      "grad_norm": 4.695148468017578,
      "learning_rate": 9.819397993311036e-05,
      "loss": 0.0933,
      "step": 270
    },
    {
      "epoch": 0.18729096989966554,
      "grad_norm": 1.0231961011886597,
      "learning_rate": 9.812709030100335e-05,
      "loss": 0.0762,
      "step": 280
    },
    {
      "epoch": 0.1939799331103679,
      "grad_norm": 0.4477454423904419,
      "learning_rate": 9.806020066889633e-05,
      "loss": 0.0817,
      "step": 290
    },
    {
      "epoch": 0.20066889632107024,
      "grad_norm": 3.472343683242798,
      "learning_rate": 9.799331103678931e-05,
      "loss": 0.0788,
      "step": 300
    },
    {
      "epoch": 0.20735785953177258,
      "grad_norm": 2.307664632797241,
      "learning_rate": 9.792642140468228e-05,
      "loss": 0.0777,
      "step": 310
    },
    {
      "epoch": 0.2140468227424749,
      "grad_norm": 2.007056951522827,
      "learning_rate": 9.785953177257525e-05,
      "loss": 0.0708,
      "step": 320
    },
    {
      "epoch": 0.22073578595317725,
      "grad_norm": 8.99911880493164,
      "learning_rate": 9.779264214046823e-05,
      "loss": 0.07,
      "step": 330
    },
    {
      "epoch": 0.22742474916387959,
      "grad_norm": 0.8066081404685974,
      "learning_rate": 9.772575250836121e-05,
      "loss": 0.0807,
      "step": 340
    },
    {
      "epoch": 0.23411371237458195,
      "grad_norm": 9.597452163696289,
      "learning_rate": 9.765886287625419e-05,
      "loss": 0.0992,
      "step": 350
    },
    {
      "epoch": 0.2408026755852843,
      "grad_norm": 0.36006277799606323,
      "learning_rate": 9.759197324414716e-05,
      "loss": 0.0892,
      "step": 360
    },
    {
      "epoch": 0.24749163879598662,
      "grad_norm": 6.198542594909668,
      "learning_rate": 9.752508361204014e-05,
      "loss": 0.0894,
      "step": 370
    },
    {
      "epoch": 0.25418060200668896,
      "grad_norm": 1.0550386905670166,
      "learning_rate": 9.745819397993311e-05,
      "loss": 0.0836,
      "step": 380
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 3.5416440963745117,
      "learning_rate": 9.739130434782609e-05,
      "loss": 0.0833,
      "step": 390
    },
    {
      "epoch": 0.26755852842809363,
      "grad_norm": 0.4253082573413849,
      "learning_rate": 9.732441471571907e-05,
      "loss": 0.0899,
      "step": 400
    },
    {
      "epoch": 0.27424749163879597,
      "grad_norm": 0.3868394196033478,
      "learning_rate": 9.725752508361204e-05,
      "loss": 0.0776,
      "step": 410
    },
    {
      "epoch": 0.2809364548494983,
      "grad_norm": 3.237781286239624,
      "learning_rate": 9.719063545150502e-05,
      "loss": 0.0927,
      "step": 420
    },
    {
      "epoch": 0.28762541806020064,
      "grad_norm": 2.2645530700683594,
      "learning_rate": 9.712374581939799e-05,
      "loss": 0.1582,
      "step": 430
    },
    {
      "epoch": 0.29431438127090304,
      "grad_norm": 4.430079460144043,
      "learning_rate": 9.705685618729097e-05,
      "loss": 0.0908,
      "step": 440
    },
    {
      "epoch": 0.3010033444816054,
      "grad_norm": 1.9928741455078125,
      "learning_rate": 9.698996655518396e-05,
      "loss": 0.0951,
      "step": 450
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 1.939744472503662,
      "learning_rate": 9.692307692307692e-05,
      "loss": 0.0727,
      "step": 460
    },
    {
      "epoch": 0.31438127090301005,
      "grad_norm": 8.799491882324219,
      "learning_rate": 9.68561872909699e-05,
      "loss": 0.1026,
      "step": 470
    },
    {
      "epoch": 0.3210702341137124,
      "grad_norm": 2.6402878761291504,
      "learning_rate": 9.678929765886287e-05,
      "loss": 0.1168,
      "step": 480
    },
    {
      "epoch": 0.3277591973244147,
      "grad_norm": 0.3376615345478058,
      "learning_rate": 9.672240802675586e-05,
      "loss": 0.0732,
      "step": 490
    },
    {
      "epoch": 0.33444816053511706,
      "grad_norm": 7.645970821380615,
      "learning_rate": 9.665551839464884e-05,
      "loss": 0.0762,
      "step": 500
    },
    {
      "epoch": 0.3411371237458194,
      "grad_norm": 0.24924737215042114,
      "learning_rate": 9.65886287625418e-05,
      "loss": 0.0831,
      "step": 510
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 3.0840044021606445,
      "learning_rate": 9.652173913043479e-05,
      "loss": 0.0865,
      "step": 520
    },
    {
      "epoch": 0.35451505016722407,
      "grad_norm": 0.9078582525253296,
      "learning_rate": 9.645484949832776e-05,
      "loss": 0.0911,
      "step": 530
    },
    {
      "epoch": 0.3612040133779264,
      "grad_norm": 1.0846281051635742,
      "learning_rate": 9.638795986622075e-05,
      "loss": 0.1001,
      "step": 540
    },
    {
      "epoch": 0.36789297658862874,
      "grad_norm": 1.852506399154663,
      "learning_rate": 9.632107023411372e-05,
      "loss": 0.0888,
      "step": 550
    },
    {
      "epoch": 0.3745819397993311,
      "grad_norm": 2.859666347503662,
      "learning_rate": 9.625418060200669e-05,
      "loss": 0.0911,
      "step": 560
    },
    {
      "epoch": 0.38127090301003347,
      "grad_norm": 1.9279602766036987,
      "learning_rate": 9.618729096989967e-05,
      "loss": 0.0864,
      "step": 570
    },
    {
      "epoch": 0.3879598662207358,
      "grad_norm": 8.074692726135254,
      "learning_rate": 9.612040133779264e-05,
      "loss": 0.0692,
      "step": 580
    },
    {
      "epoch": 0.39464882943143814,
      "grad_norm": 0.68867427110672,
      "learning_rate": 9.605351170568563e-05,
      "loss": 0.0787,
      "step": 590
    },
    {
      "epoch": 0.4013377926421405,
      "grad_norm": 5.155614376068115,
      "learning_rate": 9.59866220735786e-05,
      "loss": 0.0814,
      "step": 600
    },
    {
      "epoch": 0.4080267558528428,
      "grad_norm": 4.687347412109375,
      "learning_rate": 9.591973244147158e-05,
      "loss": 0.0845,
      "step": 610
    },
    {
      "epoch": 0.41471571906354515,
      "grad_norm": 6.637145042419434,
      "learning_rate": 9.585284280936455e-05,
      "loss": 0.0934,
      "step": 620
    },
    {
      "epoch": 0.4214046822742475,
      "grad_norm": 1.045323133468628,
      "learning_rate": 9.578595317725752e-05,
      "loss": 0.0718,
      "step": 630
    },
    {
      "epoch": 0.4280936454849498,
      "grad_norm": 7.606454849243164,
      "learning_rate": 9.571906354515052e-05,
      "loss": 0.0983,
      "step": 640
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 4.973568916320801,
      "learning_rate": 9.565217391304348e-05,
      "loss": 0.0979,
      "step": 650
    },
    {
      "epoch": 0.4414715719063545,
      "grad_norm": 4.984686851501465,
      "learning_rate": 9.558528428093647e-05,
      "loss": 0.0808,
      "step": 660
    },
    {
      "epoch": 0.44816053511705684,
      "grad_norm": 5.981488227844238,
      "learning_rate": 9.551839464882943e-05,
      "loss": 0.089,
      "step": 670
    },
    {
      "epoch": 0.45484949832775917,
      "grad_norm": 1.167374610900879,
      "learning_rate": 9.545150501672242e-05,
      "loss": 0.0829,
      "step": 680
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 0.4526713192462921,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0691,
      "step": 690
    },
    {
      "epoch": 0.4682274247491639,
      "grad_norm": 8.227622032165527,
      "learning_rate": 9.531772575250837e-05,
      "loss": 0.0949,
      "step": 700
    },
    {
      "epoch": 0.47491638795986624,
      "grad_norm": 0.20492926239967346,
      "learning_rate": 9.525083612040135e-05,
      "loss": 0.0778,
      "step": 710
    },
    {
      "epoch": 0.4816053511705686,
      "grad_norm": 3.324284315109253,
      "learning_rate": 9.518394648829432e-05,
      "loss": 0.0944,
      "step": 720
    },
    {
      "epoch": 0.4882943143812709,
      "grad_norm": 5.781021595001221,
      "learning_rate": 9.51170568561873e-05,
      "loss": 0.0906,
      "step": 730
    },
    {
      "epoch": 0.49498327759197325,
      "grad_norm": 0.3161633014678955,
      "learning_rate": 9.505016722408028e-05,
      "loss": 0.0878,
      "step": 740
    },
    {
      "epoch": 0.5016722408026756,
      "grad_norm": 0.9887397289276123,
      "learning_rate": 9.498327759197325e-05,
      "loss": 0.0766,
      "step": 750
    },
    {
      "epoch": 0.5083612040133779,
      "grad_norm": 0.4013742506504059,
      "learning_rate": 9.491638795986623e-05,
      "loss": 0.0828,
      "step": 760
    },
    {
      "epoch": 0.5150501672240803,
      "grad_norm": 4.095464706420898,
      "learning_rate": 9.48494983277592e-05,
      "loss": 0.0941,
      "step": 770
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 10.108964920043945,
      "learning_rate": 9.478260869565218e-05,
      "loss": 0.0914,
      "step": 780
    },
    {
      "epoch": 0.5284280936454849,
      "grad_norm": 7.85491943359375,
      "learning_rate": 9.471571906354515e-05,
      "loss": 0.0985,
      "step": 790
    },
    {
      "epoch": 0.5351170568561873,
      "grad_norm": 0.8438981771469116,
      "learning_rate": 9.464882943143813e-05,
      "loss": 0.0743,
      "step": 800
    },
    {
      "epoch": 0.5418060200668896,
      "grad_norm": 1.7622973918914795,
      "learning_rate": 9.458193979933111e-05,
      "loss": 0.095,
      "step": 810
    },
    {
      "epoch": 0.5484949832775919,
      "grad_norm": 3.505446672439575,
      "learning_rate": 9.451505016722408e-05,
      "loss": 0.0811,
      "step": 820
    },
    {
      "epoch": 0.5551839464882943,
      "grad_norm": 0.6634843349456787,
      "learning_rate": 9.444816053511706e-05,
      "loss": 0.0917,
      "step": 830
    },
    {
      "epoch": 0.5618729096989966,
      "grad_norm": 4.3299880027771,
      "learning_rate": 9.438127090301003e-05,
      "loss": 0.077,
      "step": 840
    },
    {
      "epoch": 0.568561872909699,
      "grad_norm": 3.079010009765625,
      "learning_rate": 9.431438127090302e-05,
      "loss": 0.0775,
      "step": 850
    },
    {
      "epoch": 0.5752508361204013,
      "grad_norm": 3.117349863052368,
      "learning_rate": 9.424749163879599e-05,
      "loss": 0.0581,
      "step": 860
    },
    {
      "epoch": 0.5819397993311036,
      "grad_norm": 7.100534915924072,
      "learning_rate": 9.418060200668896e-05,
      "loss": 0.0963,
      "step": 870
    },
    {
      "epoch": 0.5886287625418061,
      "grad_norm": 2.5692174434661865,
      "learning_rate": 9.411371237458194e-05,
      "loss": 0.0813,
      "step": 880
    },
    {
      "epoch": 0.5953177257525084,
      "grad_norm": 5.7321271896362305,
      "learning_rate": 9.404682274247491e-05,
      "loss": 0.0824,
      "step": 890
    },
    {
      "epoch": 0.6020066889632107,
      "grad_norm": 0.9095240831375122,
      "learning_rate": 9.39799331103679e-05,
      "loss": 0.0656,
      "step": 900
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 2.5167980194091797,
      "learning_rate": 9.391304347826087e-05,
      "loss": 0.0745,
      "step": 910
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.2390034794807434,
      "learning_rate": 9.384615384615386e-05,
      "loss": 0.0896,
      "step": 920
    },
    {
      "epoch": 0.6220735785953178,
      "grad_norm": 2.4616341590881348,
      "learning_rate": 9.377926421404682e-05,
      "loss": 0.0833,
      "step": 930
    },
    {
      "epoch": 0.6287625418060201,
      "grad_norm": 10.120244026184082,
      "learning_rate": 9.371237458193979e-05,
      "loss": 0.0887,
      "step": 940
    },
    {
      "epoch": 0.6354515050167224,
      "grad_norm": 2.8883373737335205,
      "learning_rate": 9.364548494983279e-05,
      "loss": 0.1058,
      "step": 950
    },
    {
      "epoch": 0.6421404682274248,
      "grad_norm": 7.05233097076416,
      "learning_rate": 9.357859531772576e-05,
      "loss": 0.087,
      "step": 960
    },
    {
      "epoch": 0.6488294314381271,
      "grad_norm": 8.142528533935547,
      "learning_rate": 9.351170568561874e-05,
      "loss": 0.1047,
      "step": 970
    },
    {
      "epoch": 0.6555183946488294,
      "grad_norm": 6.927243232727051,
      "learning_rate": 9.344481605351171e-05,
      "loss": 0.088,
      "step": 980
    },
    {
      "epoch": 0.6622073578595318,
      "grad_norm": 6.964981555938721,
      "learning_rate": 9.337792642140469e-05,
      "loss": 0.0822,
      "step": 990
    },
    {
      "epoch": 0.6688963210702341,
      "grad_norm": 0.3616563081741333,
      "learning_rate": 9.331103678929767e-05,
      "loss": 0.0849,
      "step": 1000
    },
    {
      "epoch": 0.6755852842809364,
      "grad_norm": 0.9265075922012329,
      "learning_rate": 9.324414715719064e-05,
      "loss": 0.0885,
      "step": 1010
    },
    {
      "epoch": 0.6822742474916388,
      "grad_norm": 2.175069570541382,
      "learning_rate": 9.317725752508362e-05,
      "loss": 0.0609,
      "step": 1020
    },
    {
      "epoch": 0.6889632107023411,
      "grad_norm": 0.4830019474029541,
      "learning_rate": 9.311036789297659e-05,
      "loss": 0.069,
      "step": 1030
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 10.464764595031738,
      "learning_rate": 9.304347826086957e-05,
      "loss": 0.0956,
      "step": 1040
    },
    {
      "epoch": 0.7023411371237458,
      "grad_norm": 4.139832973480225,
      "learning_rate": 9.297658862876255e-05,
      "loss": 0.0925,
      "step": 1050
    },
    {
      "epoch": 0.7090301003344481,
      "grad_norm": 2.826423406600952,
      "learning_rate": 9.290969899665552e-05,
      "loss": 0.0898,
      "step": 1060
    },
    {
      "epoch": 0.7157190635451505,
      "grad_norm": 5.344399452209473,
      "learning_rate": 9.28428093645485e-05,
      "loss": 0.0696,
      "step": 1070
    },
    {
      "epoch": 0.7224080267558528,
      "grad_norm": 3.084770441055298,
      "learning_rate": 9.277591973244147e-05,
      "loss": 0.0643,
      "step": 1080
    },
    {
      "epoch": 0.7290969899665551,
      "grad_norm": 0.4706350862979889,
      "learning_rate": 9.270903010033445e-05,
      "loss": 0.0819,
      "step": 1090
    },
    {
      "epoch": 0.7357859531772575,
      "grad_norm": 3.9289710521698,
      "learning_rate": 9.264214046822743e-05,
      "loss": 0.1029,
      "step": 1100
    },
    {
      "epoch": 0.7424749163879598,
      "grad_norm": 0.404439240694046,
      "learning_rate": 9.25752508361204e-05,
      "loss": 0.0877,
      "step": 1110
    },
    {
      "epoch": 0.7491638795986622,
      "grad_norm": 7.014498710632324,
      "learning_rate": 9.250836120401338e-05,
      "loss": 0.0883,
      "step": 1120
    },
    {
      "epoch": 0.7558528428093646,
      "grad_norm": 5.416506290435791,
      "learning_rate": 9.244147157190635e-05,
      "loss": 0.1351,
      "step": 1130
    },
    {
      "epoch": 0.7625418060200669,
      "grad_norm": 5.6372904777526855,
      "learning_rate": 9.237458193979933e-05,
      "loss": 0.0914,
      "step": 1140
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 2.7677652835845947,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0809,
      "step": 1150
    },
    {
      "epoch": 0.7759197324414716,
      "grad_norm": 5.408329010009766,
      "learning_rate": 9.22408026755853e-05,
      "loss": 0.0739,
      "step": 1160
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 0.7951474189758301,
      "learning_rate": 9.217391304347827e-05,
      "loss": 0.0703,
      "step": 1170
    },
    {
      "epoch": 0.7892976588628763,
      "grad_norm": 8.31872844696045,
      "learning_rate": 9.210702341137123e-05,
      "loss": 0.0777,
      "step": 1180
    },
    {
      "epoch": 0.7959866220735786,
      "grad_norm": 1.7713472843170166,
      "learning_rate": 9.204013377926422e-05,
      "loss": 0.0626,
      "step": 1190
    },
    {
      "epoch": 0.802675585284281,
      "grad_norm": 5.379355430603027,
      "learning_rate": 9.19732441471572e-05,
      "loss": 0.0763,
      "step": 1200
    },
    {
      "epoch": 0.8093645484949833,
      "grad_norm": 6.51313591003418,
      "learning_rate": 9.190635451505018e-05,
      "loss": 0.0869,
      "step": 1210
    },
    {
      "epoch": 0.8160535117056856,
      "grad_norm": 3.0633490085601807,
      "learning_rate": 9.183946488294315e-05,
      "loss": 0.0807,
      "step": 1220
    },
    {
      "epoch": 0.822742474916388,
      "grad_norm": 1.5585048198699951,
      "learning_rate": 9.177257525083613e-05,
      "loss": 0.0714,
      "step": 1230
    },
    {
      "epoch": 0.8294314381270903,
      "grad_norm": 1.0530630350112915,
      "learning_rate": 9.17056856187291e-05,
      "loss": 0.0777,
      "step": 1240
    },
    {
      "epoch": 0.8361204013377926,
      "grad_norm": 4.880084991455078,
      "learning_rate": 9.163879598662207e-05,
      "loss": 0.062,
      "step": 1250
    },
    {
      "epoch": 0.842809364548495,
      "grad_norm": 4.956668853759766,
      "learning_rate": 9.157190635451506e-05,
      "loss": 0.0936,
      "step": 1260
    },
    {
      "epoch": 0.8494983277591973,
      "grad_norm": 4.809473037719727,
      "learning_rate": 9.150501672240803e-05,
      "loss": 0.0944,
      "step": 1270
    },
    {
      "epoch": 0.8561872909698997,
      "grad_norm": 4.575435638427734,
      "learning_rate": 9.143812709030101e-05,
      "loss": 0.0795,
      "step": 1280
    },
    {
      "epoch": 0.862876254180602,
      "grad_norm": 5.542465686798096,
      "learning_rate": 9.137123745819398e-05,
      "loss": 0.0902,
      "step": 1290
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 2.882326364517212,
      "learning_rate": 9.130434782608696e-05,
      "loss": 0.077,
      "step": 1300
    },
    {
      "epoch": 0.8762541806020067,
      "grad_norm": 5.928070545196533,
      "learning_rate": 9.123745819397994e-05,
      "loss": 0.0959,
      "step": 1310
    },
    {
      "epoch": 0.882943143812709,
      "grad_norm": 5.219929218292236,
      "learning_rate": 9.117056856187291e-05,
      "loss": 0.1166,
      "step": 1320
    },
    {
      "epoch": 0.8896321070234113,
      "grad_norm": 2.3765580654144287,
      "learning_rate": 9.11036789297659e-05,
      "loss": 0.0778,
      "step": 1330
    },
    {
      "epoch": 0.8963210702341137,
      "grad_norm": 5.088015079498291,
      "learning_rate": 9.103678929765886e-05,
      "loss": 0.0715,
      "step": 1340
    },
    {
      "epoch": 0.903010033444816,
      "grad_norm": 2.053117513656616,
      "learning_rate": 9.096989966555184e-05,
      "loss": 0.0676,
      "step": 1350
    },
    {
      "epoch": 0.9096989966555183,
      "grad_norm": 4.23502254486084,
      "learning_rate": 9.090301003344483e-05,
      "loss": 0.0972,
      "step": 1360
    },
    {
      "epoch": 0.9163879598662207,
      "grad_norm": 1.1127852201461792,
      "learning_rate": 9.08361204013378e-05,
      "loss": 0.0889,
      "step": 1370
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 2.8048794269561768,
      "learning_rate": 9.076923076923078e-05,
      "loss": 0.0753,
      "step": 1380
    },
    {
      "epoch": 0.9297658862876255,
      "grad_norm": 2.873419761657715,
      "learning_rate": 9.070234113712374e-05,
      "loss": 0.0755,
      "step": 1390
    },
    {
      "epoch": 0.9364548494983278,
      "grad_norm": 1.651400089263916,
      "learning_rate": 9.063545150501673e-05,
      "loss": 0.0861,
      "step": 1400
    },
    {
      "epoch": 0.9431438127090301,
      "grad_norm": 7.603207111358643,
      "learning_rate": 9.056856187290971e-05,
      "loss": 0.0839,
      "step": 1410
    },
    {
      "epoch": 0.9498327759197325,
      "grad_norm": 3.371145009994507,
      "learning_rate": 9.050167224080268e-05,
      "loss": 0.0841,
      "step": 1420
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 2.119283437728882,
      "learning_rate": 9.043478260869566e-05,
      "loss": 0.0654,
      "step": 1430
    },
    {
      "epoch": 0.9632107023411371,
      "grad_norm": 3.469703435897827,
      "learning_rate": 9.036789297658863e-05,
      "loss": 0.0876,
      "step": 1440
    },
    {
      "epoch": 0.9698996655518395,
      "grad_norm": 4.874169826507568,
      "learning_rate": 9.030100334448161e-05,
      "loss": 0.0776,
      "step": 1450
    },
    {
      "epoch": 0.9765886287625418,
      "grad_norm": 7.210989952087402,
      "learning_rate": 9.023411371237459e-05,
      "loss": 0.0905,
      "step": 1460
    },
    {
      "epoch": 0.9832775919732442,
      "grad_norm": 2.3983662128448486,
      "learning_rate": 9.016722408026757e-05,
      "loss": 0.0847,
      "step": 1470
    },
    {
      "epoch": 0.9899665551839465,
      "grad_norm": 5.223353862762451,
      "learning_rate": 9.010033444816054e-05,
      "loss": 0.0877,
      "step": 1480
    },
    {
      "epoch": 0.9966555183946488,
      "grad_norm": 0.9329972863197327,
      "learning_rate": 9.003344481605351e-05,
      "loss": 0.0816,
      "step": 1490
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.0947796180844307,
      "eval_mse": 0.0947796180844307,
      "eval_runtime": 227.7814,
      "eval_samples_per_second": 13.127,
      "eval_steps_per_second": 1.642,
      "step": 1495
    },
    {
      "epoch": 1.0033444816053512,
      "grad_norm": 9.278254508972168,
      "learning_rate": 8.996655518394649e-05,
      "loss": 0.0897,
      "step": 1500
    },
    {
      "epoch": 1.0100334448160535,
      "grad_norm": 3.634450912475586,
      "learning_rate": 8.989966555183947e-05,
      "loss": 0.0784,
      "step": 1510
    },
    {
      "epoch": 1.0167224080267558,
      "grad_norm": 1.6217931509017944,
      "learning_rate": 8.983277591973245e-05,
      "loss": 0.086,
      "step": 1520
    },
    {
      "epoch": 1.0234113712374582,
      "grad_norm": 3.606285333633423,
      "learning_rate": 8.976588628762542e-05,
      "loss": 0.0768,
      "step": 1530
    },
    {
      "epoch": 1.0301003344481605,
      "grad_norm": 1.5076968669891357,
      "learning_rate": 8.96989966555184e-05,
      "loss": 0.0859,
      "step": 1540
    },
    {
      "epoch": 1.0367892976588629,
      "grad_norm": 6.945788383483887,
      "learning_rate": 8.963210702341137e-05,
      "loss": 0.0792,
      "step": 1550
    },
    {
      "epoch": 1.0434782608695652,
      "grad_norm": 0.16002212464809418,
      "learning_rate": 8.956521739130435e-05,
      "loss": 0.0613,
      "step": 1560
    },
    {
      "epoch": 1.0501672240802675,
      "grad_norm": 0.5019150972366333,
      "learning_rate": 8.949832775919734e-05,
      "loss": 0.0794,
      "step": 1570
    },
    {
      "epoch": 1.0568561872909699,
      "grad_norm": 3.1986889839172363,
      "learning_rate": 8.94314381270903e-05,
      "loss": 0.0671,
      "step": 1580
    },
    {
      "epoch": 1.0635451505016722,
      "grad_norm": 0.7359128594398499,
      "learning_rate": 8.936454849498329e-05,
      "loss": 0.0867,
      "step": 1590
    },
    {
      "epoch": 1.0702341137123745,
      "grad_norm": 3.642479419708252,
      "learning_rate": 8.929765886287625e-05,
      "loss": 0.0934,
      "step": 1600
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 2.9425642490386963,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.1037,
      "step": 1610
    },
    {
      "epoch": 1.0836120401337792,
      "grad_norm": 2.242279052734375,
      "learning_rate": 8.916387959866222e-05,
      "loss": 0.0793,
      "step": 1620
    },
    {
      "epoch": 1.0903010033444815,
      "grad_norm": 3.2683675289154053,
      "learning_rate": 8.909698996655519e-05,
      "loss": 0.0895,
      "step": 1630
    },
    {
      "epoch": 1.0969899665551839,
      "grad_norm": 0.4252338707447052,
      "learning_rate": 8.903010033444817e-05,
      "loss": 0.0727,
      "step": 1640
    },
    {
      "epoch": 1.1036789297658862,
      "grad_norm": 5.4478373527526855,
      "learning_rate": 8.896321070234114e-05,
      "loss": 0.0795,
      "step": 1650
    },
    {
      "epoch": 1.1103678929765886,
      "grad_norm": 3.975977659225464,
      "learning_rate": 8.889632107023412e-05,
      "loss": 0.0705,
      "step": 1660
    },
    {
      "epoch": 1.117056856187291,
      "grad_norm": 4.362052917480469,
      "learning_rate": 8.88294314381271e-05,
      "loss": 0.0787,
      "step": 1670
    },
    {
      "epoch": 1.1237458193979932,
      "grad_norm": 3.969586133956909,
      "learning_rate": 8.876254180602007e-05,
      "loss": 0.0942,
      "step": 1680
    },
    {
      "epoch": 1.1304347826086956,
      "grad_norm": 1.0709367990493774,
      "learning_rate": 8.869565217391305e-05,
      "loss": 0.0834,
      "step": 1690
    },
    {
      "epoch": 1.137123745819398,
      "grad_norm": 1.190765619277954,
      "learning_rate": 8.862876254180602e-05,
      "loss": 0.0827,
      "step": 1700
    },
    {
      "epoch": 1.1438127090301002,
      "grad_norm": 7.36898136138916,
      "learning_rate": 8.856187290969901e-05,
      "loss": 0.1021,
      "step": 1710
    },
    {
      "epoch": 1.1505016722408028,
      "grad_norm": 3.5522820949554443,
      "learning_rate": 8.849498327759198e-05,
      "loss": 0.0856,
      "step": 1720
    },
    {
      "epoch": 1.1571906354515051,
      "grad_norm": 5.038424015045166,
      "learning_rate": 8.842809364548495e-05,
      "loss": 0.0845,
      "step": 1730
    },
    {
      "epoch": 1.1638795986622075,
      "grad_norm": 3.5751819610595703,
      "learning_rate": 8.836120401337793e-05,
      "loss": 0.077,
      "step": 1740
    },
    {
      "epoch": 1.1705685618729098,
      "grad_norm": 5.335000991821289,
      "learning_rate": 8.82943143812709e-05,
      "loss": 0.0886,
      "step": 1750
    },
    {
      "epoch": 1.1772575250836121,
      "grad_norm": 3.961160898208618,
      "learning_rate": 8.822742474916388e-05,
      "loss": 0.0715,
      "step": 1760
    },
    {
      "epoch": 1.1839464882943145,
      "grad_norm": 2.157200574874878,
      "learning_rate": 8.816053511705686e-05,
      "loss": 0.0882,
      "step": 1770
    },
    {
      "epoch": 1.1906354515050168,
      "grad_norm": 6.394442081451416,
      "learning_rate": 8.809364548494984e-05,
      "loss": 0.0808,
      "step": 1780
    },
    {
      "epoch": 1.1973244147157192,
      "grad_norm": 1.3427674770355225,
      "learning_rate": 8.802675585284281e-05,
      "loss": 0.0606,
      "step": 1790
    },
    {
      "epoch": 1.2040133779264215,
      "grad_norm": 0.8644570708274841,
      "learning_rate": 8.795986622073578e-05,
      "loss": 0.0782,
      "step": 1800
    },
    {
      "epoch": 1.2107023411371238,
      "grad_norm": 1.9095417261123657,
      "learning_rate": 8.789297658862876e-05,
      "loss": 0.0843,
      "step": 1810
    },
    {
      "epoch": 1.2173913043478262,
      "grad_norm": 5.084800720214844,
      "learning_rate": 8.782608695652174e-05,
      "loss": 0.1011,
      "step": 1820
    },
    {
      "epoch": 1.2240802675585285,
      "grad_norm": 1.1336281299591064,
      "learning_rate": 8.775919732441473e-05,
      "loss": 0.1059,
      "step": 1830
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 2.755981683731079,
      "learning_rate": 8.76923076923077e-05,
      "loss": 0.0657,
      "step": 1840
    },
    {
      "epoch": 1.2374581939799332,
      "grad_norm": 1.1654160022735596,
      "learning_rate": 8.762541806020068e-05,
      "loss": 0.079,
      "step": 1850
    },
    {
      "epoch": 1.2441471571906355,
      "grad_norm": 2.7007975578308105,
      "learning_rate": 8.755852842809365e-05,
      "loss": 0.0932,
      "step": 1860
    },
    {
      "epoch": 1.2508361204013378,
      "grad_norm": 4.843238830566406,
      "learning_rate": 8.749163879598663e-05,
      "loss": 0.071,
      "step": 1870
    },
    {
      "epoch": 1.2575250836120402,
      "grad_norm": 5.805212020874023,
      "learning_rate": 8.742474916387961e-05,
      "loss": 0.0814,
      "step": 1880
    },
    {
      "epoch": 1.2642140468227425,
      "grad_norm": 3.9875166416168213,
      "learning_rate": 8.735785953177258e-05,
      "loss": 0.0848,
      "step": 1890
    },
    {
      "epoch": 1.2709030100334449,
      "grad_norm": 0.26979923248291016,
      "learning_rate": 8.729096989966556e-05,
      "loss": 0.0797,
      "step": 1900
    },
    {
      "epoch": 1.2775919732441472,
      "grad_norm": 4.210855007171631,
      "learning_rate": 8.722408026755853e-05,
      "loss": 0.074,
      "step": 1910
    },
    {
      "epoch": 1.2842809364548495,
      "grad_norm": 6.115736484527588,
      "learning_rate": 8.715719063545151e-05,
      "loss": 0.0752,
      "step": 1920
    },
    {
      "epoch": 1.2909698996655519,
      "grad_norm": 0.48328760266304016,
      "learning_rate": 8.709030100334449e-05,
      "loss": 0.0744,
      "step": 1930
    },
    {
      "epoch": 1.2976588628762542,
      "grad_norm": 0.3940099775791168,
      "learning_rate": 8.702341137123746e-05,
      "loss": 0.0846,
      "step": 1940
    },
    {
      "epoch": 1.3043478260869565,
      "grad_norm": 6.541823863983154,
      "learning_rate": 8.695652173913044e-05,
      "loss": 0.0767,
      "step": 1950
    },
    {
      "epoch": 1.3110367892976589,
      "grad_norm": 7.311049461364746,
      "learning_rate": 8.688963210702341e-05,
      "loss": 0.0826,
      "step": 1960
    },
    {
      "epoch": 1.3177257525083612,
      "grad_norm": 1.2747488021850586,
      "learning_rate": 8.682274247491639e-05,
      "loss": 0.0677,
      "step": 1970
    },
    {
      "epoch": 1.3244147157190636,
      "grad_norm": 9.177070617675781,
      "learning_rate": 8.675585284280937e-05,
      "loss": 0.0675,
      "step": 1980
    },
    {
      "epoch": 1.3311036789297659,
      "grad_norm": 0.9875199198722839,
      "learning_rate": 8.668896321070234e-05,
      "loss": 0.072,
      "step": 1990
    },
    {
      "epoch": 1.3377926421404682,
      "grad_norm": 5.233514308929443,
      "learning_rate": 8.662207357859532e-05,
      "loss": 0.0734,
      "step": 2000
    },
    {
      "epoch": 1.3444816053511706,
      "grad_norm": 3.531949996948242,
      "learning_rate": 8.655518394648829e-05,
      "loss": 0.0979,
      "step": 2010
    },
    {
      "epoch": 1.351170568561873,
      "grad_norm": 1.9295690059661865,
      "learning_rate": 8.648829431438129e-05,
      "loss": 0.0838,
      "step": 2020
    },
    {
      "epoch": 1.3578595317725752,
      "grad_norm": 5.854907989501953,
      "learning_rate": 8.642140468227425e-05,
      "loss": 0.1134,
      "step": 2030
    },
    {
      "epoch": 1.3645484949832776,
      "grad_norm": 2.4425508975982666,
      "learning_rate": 8.635451505016722e-05,
      "loss": 0.0727,
      "step": 2040
    },
    {
      "epoch": 1.37123745819398,
      "grad_norm": 5.542223930358887,
      "learning_rate": 8.62876254180602e-05,
      "loss": 0.0816,
      "step": 2050
    },
    {
      "epoch": 1.3779264214046822,
      "grad_norm": 1.6907216310501099,
      "learning_rate": 8.622073578595317e-05,
      "loss": 0.0907,
      "step": 2060
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 4.605374813079834,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.081,
      "step": 2070
    },
    {
      "epoch": 1.391304347826087,
      "grad_norm": 4.201385498046875,
      "learning_rate": 8.608695652173914e-05,
      "loss": 0.1161,
      "step": 2080
    },
    {
      "epoch": 1.3979933110367893,
      "grad_norm": 2.6294329166412354,
      "learning_rate": 8.602006688963212e-05,
      "loss": 0.0685,
      "step": 2090
    },
    {
      "epoch": 1.4046822742474916,
      "grad_norm": 0.2493625432252884,
      "learning_rate": 8.595317725752509e-05,
      "loss": 0.0683,
      "step": 2100
    },
    {
      "epoch": 1.411371237458194,
      "grad_norm": 3.8422601222991943,
      "learning_rate": 8.588628762541805e-05,
      "loss": 0.094,
      "step": 2110
    },
    {
      "epoch": 1.4180602006688963,
      "grad_norm": 1.6194677352905273,
      "learning_rate": 8.581939799331105e-05,
      "loss": 0.078,
      "step": 2120
    },
    {
      "epoch": 1.4247491638795986,
      "grad_norm": 3.9038643836975098,
      "learning_rate": 8.575250836120402e-05,
      "loss": 0.0675,
      "step": 2130
    },
    {
      "epoch": 1.431438127090301,
      "grad_norm": 6.922024250030518,
      "learning_rate": 8.5685618729097e-05,
      "loss": 0.082,
      "step": 2140
    },
    {
      "epoch": 1.4381270903010033,
      "grad_norm": 1.8225903511047363,
      "learning_rate": 8.561872909698997e-05,
      "loss": 0.0935,
      "step": 2150
    },
    {
      "epoch": 1.4448160535117056,
      "grad_norm": 9.204913139343262,
      "learning_rate": 8.555183946488295e-05,
      "loss": 0.0878,
      "step": 2160
    },
    {
      "epoch": 1.451505016722408,
      "grad_norm": 3.7089452743530273,
      "learning_rate": 8.548494983277593e-05,
      "loss": 0.0776,
      "step": 2170
    },
    {
      "epoch": 1.4581939799331103,
      "grad_norm": 4.970576763153076,
      "learning_rate": 8.54180602006689e-05,
      "loss": 0.0956,
      "step": 2180
    },
    {
      "epoch": 1.4648829431438126,
      "grad_norm": 4.007099151611328,
      "learning_rate": 8.535117056856188e-05,
      "loss": 0.0937,
      "step": 2190
    },
    {
      "epoch": 1.471571906354515,
      "grad_norm": 4.514796733856201,
      "learning_rate": 8.528428093645485e-05,
      "loss": 0.0773,
      "step": 2200
    },
    {
      "epoch": 1.4782608695652173,
      "grad_norm": 2.7079756259918213,
      "learning_rate": 8.521739130434783e-05,
      "loss": 0.0848,
      "step": 2210
    },
    {
      "epoch": 1.4849498327759196,
      "grad_norm": 0.419909805059433,
      "learning_rate": 8.515050167224081e-05,
      "loss": 0.0627,
      "step": 2220
    },
    {
      "epoch": 1.491638795986622,
      "grad_norm": 0.8812950849533081,
      "learning_rate": 8.508361204013378e-05,
      "loss": 0.0624,
      "step": 2230
    },
    {
      "epoch": 1.4983277591973243,
      "grad_norm": 2.6033947467803955,
      "learning_rate": 8.501672240802676e-05,
      "loss": 0.0989,
      "step": 2240
    },
    {
      "epoch": 1.5050167224080266,
      "grad_norm": 5.632299423217773,
      "learning_rate": 8.494983277591973e-05,
      "loss": 0.0938,
      "step": 2250
    },
    {
      "epoch": 1.511705685618729,
      "grad_norm": 2.372530221939087,
      "learning_rate": 8.488294314381271e-05,
      "loss": 0.0877,
      "step": 2260
    },
    {
      "epoch": 1.5183946488294313,
      "grad_norm": 3.246079921722412,
      "learning_rate": 8.481605351170568e-05,
      "loss": 0.0826,
      "step": 2270
    },
    {
      "epoch": 1.5250836120401337,
      "grad_norm": 3.317453384399414,
      "learning_rate": 8.474916387959866e-05,
      "loss": 0.0775,
      "step": 2280
    },
    {
      "epoch": 1.531772575250836,
      "grad_norm": 4.335958003997803,
      "learning_rate": 8.468227424749165e-05,
      "loss": 0.0666,
      "step": 2290
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 3.737240791320801,
      "learning_rate": 8.461538461538461e-05,
      "loss": 0.0647,
      "step": 2300
    },
    {
      "epoch": 1.5451505016722407,
      "grad_norm": 1.582314372062683,
      "learning_rate": 8.45484949832776e-05,
      "loss": 0.0834,
      "step": 2310
    },
    {
      "epoch": 1.551839464882943,
      "grad_norm": 3.502819061279297,
      "learning_rate": 8.448160535117056e-05,
      "loss": 0.0905,
      "step": 2320
    },
    {
      "epoch": 1.5585284280936453,
      "grad_norm": 0.8796928524971008,
      "learning_rate": 8.441471571906356e-05,
      "loss": 0.0752,
      "step": 2330
    },
    {
      "epoch": 1.5652173913043477,
      "grad_norm": 3.127408742904663,
      "learning_rate": 8.434782608695653e-05,
      "loss": 0.0596,
      "step": 2340
    },
    {
      "epoch": 1.57190635451505,
      "grad_norm": 2.658625602722168,
      "learning_rate": 8.42809364548495e-05,
      "loss": 0.091,
      "step": 2350
    },
    {
      "epoch": 1.5785953177257523,
      "grad_norm": 1.1064062118530273,
      "learning_rate": 8.421404682274248e-05,
      "loss": 0.0784,
      "step": 2360
    },
    {
      "epoch": 1.585284280936455,
      "grad_norm": 5.8419389724731445,
      "learning_rate": 8.414715719063545e-05,
      "loss": 0.085,
      "step": 2370
    },
    {
      "epoch": 1.5919732441471572,
      "grad_norm": 3.167968273162842,
      "learning_rate": 8.408026755852844e-05,
      "loss": 0.1401,
      "step": 2380
    },
    {
      "epoch": 1.5986622073578596,
      "grad_norm": 6.581013202667236,
      "learning_rate": 8.401337792642141e-05,
      "loss": 0.0865,
      "step": 2390
    },
    {
      "epoch": 1.605351170568562,
      "grad_norm": 0.5135999321937561,
      "learning_rate": 8.394648829431439e-05,
      "loss": 0.0883,
      "step": 2400
    },
    {
      "epoch": 1.6120401337792643,
      "grad_norm": 1.0057544708251953,
      "learning_rate": 8.387959866220736e-05,
      "loss": 0.1091,
      "step": 2410
    },
    {
      "epoch": 1.6187290969899666,
      "grad_norm": 1.3300502300262451,
      "learning_rate": 8.381270903010033e-05,
      "loss": 0.0762,
      "step": 2420
    },
    {
      "epoch": 1.625418060200669,
      "grad_norm": 10.380025863647461,
      "learning_rate": 8.374581939799332e-05,
      "loss": 0.0745,
      "step": 2430
    },
    {
      "epoch": 1.6321070234113713,
      "grad_norm": 0.41489875316619873,
      "learning_rate": 8.367892976588629e-05,
      "loss": 0.0659,
      "step": 2440
    },
    {
      "epoch": 1.6387959866220736,
      "grad_norm": 2.31946063041687,
      "learning_rate": 8.361204013377927e-05,
      "loss": 0.069,
      "step": 2450
    },
    {
      "epoch": 1.645484949832776,
      "grad_norm": 3.089249849319458,
      "learning_rate": 8.354515050167224e-05,
      "loss": 0.0932,
      "step": 2460
    },
    {
      "epoch": 1.6521739130434783,
      "grad_norm": 1.3205960988998413,
      "learning_rate": 8.347826086956521e-05,
      "loss": 0.0795,
      "step": 2470
    },
    {
      "epoch": 1.6588628762541806,
      "grad_norm": 3.7268786430358887,
      "learning_rate": 8.34113712374582e-05,
      "loss": 0.0877,
      "step": 2480
    },
    {
      "epoch": 1.665551839464883,
      "grad_norm": 2.373908758163452,
      "learning_rate": 8.334448160535117e-05,
      "loss": 0.0831,
      "step": 2490
    },
    {
      "epoch": 1.6722408026755853,
      "grad_norm": 0.46990734338760376,
      "learning_rate": 8.327759197324416e-05,
      "loss": 0.0741,
      "step": 2500
    },
    {
      "epoch": 1.6789297658862876,
      "grad_norm": 1.2849098443984985,
      "learning_rate": 8.321070234113712e-05,
      "loss": 0.0728,
      "step": 2510
    },
    {
      "epoch": 1.68561872909699,
      "grad_norm": 8.4157133102417,
      "learning_rate": 8.31438127090301e-05,
      "loss": 0.07,
      "step": 2520
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 2.4645774364471436,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0717,
      "step": 2530
    },
    {
      "epoch": 1.6989966555183946,
      "grad_norm": 1.3757083415985107,
      "learning_rate": 8.301003344481606e-05,
      "loss": 0.0652,
      "step": 2540
    },
    {
      "epoch": 1.705685618729097,
      "grad_norm": 6.6431050300598145,
      "learning_rate": 8.294314381270904e-05,
      "loss": 0.0796,
      "step": 2550
    },
    {
      "epoch": 1.7123745819397993,
      "grad_norm": 0.2695080637931824,
      "learning_rate": 8.2876254180602e-05,
      "loss": 0.0724,
      "step": 2560
    },
    {
      "epoch": 1.7190635451505016,
      "grad_norm": 4.07316255569458,
      "learning_rate": 8.280936454849499e-05,
      "loss": 0.0898,
      "step": 2570
    },
    {
      "epoch": 1.725752508361204,
      "grad_norm": 0.32654353976249695,
      "learning_rate": 8.274247491638797e-05,
      "loss": 0.0838,
      "step": 2580
    },
    {
      "epoch": 1.7324414715719063,
      "grad_norm": 4.416987419128418,
      "learning_rate": 8.267558528428094e-05,
      "loss": 0.063,
      "step": 2590
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 3.2613632678985596,
      "learning_rate": 8.260869565217392e-05,
      "loss": 0.0756,
      "step": 2600
    },
    {
      "epoch": 1.745819397993311,
      "grad_norm": 1.9059773683547974,
      "learning_rate": 8.254180602006689e-05,
      "loss": 0.0794,
      "step": 2610
    },
    {
      "epoch": 1.7525083612040135,
      "grad_norm": 4.8873443603515625,
      "learning_rate": 8.247491638795987e-05,
      "loss": 0.0819,
      "step": 2620
    },
    {
      "epoch": 1.7591973244147159,
      "grad_norm": 1.6874901056289673,
      "learning_rate": 8.240802675585285e-05,
      "loss": 0.0772,
      "step": 2630
    },
    {
      "epoch": 1.7658862876254182,
      "grad_norm": 4.475655555725098,
      "learning_rate": 8.234113712374582e-05,
      "loss": 0.0745,
      "step": 2640
    },
    {
      "epoch": 1.7725752508361206,
      "grad_norm": 1.547790288925171,
      "learning_rate": 8.22742474916388e-05,
      "loss": 0.0936,
      "step": 2650
    },
    {
      "epoch": 1.779264214046823,
      "grad_norm": 2.782203197479248,
      "learning_rate": 8.220735785953177e-05,
      "loss": 0.082,
      "step": 2660
    },
    {
      "epoch": 1.7859531772575252,
      "grad_norm": 0.9756410121917725,
      "learning_rate": 8.214046822742475e-05,
      "loss": 0.08,
      "step": 2670
    },
    {
      "epoch": 1.7926421404682276,
      "grad_norm": 4.201378345489502,
      "learning_rate": 8.207357859531773e-05,
      "loss": 0.0856,
      "step": 2680
    },
    {
      "epoch": 1.79933110367893,
      "grad_norm": 11.08785629272461,
      "learning_rate": 8.200668896321071e-05,
      "loss": 0.0786,
      "step": 2690
    },
    {
      "epoch": 1.8060200668896322,
      "grad_norm": 2.3140785694122314,
      "learning_rate": 8.193979933110368e-05,
      "loss": 0.0986,
      "step": 2700
    },
    {
      "epoch": 1.8127090301003346,
      "grad_norm": 4.409708499908447,
      "learning_rate": 8.187290969899665e-05,
      "loss": 0.0768,
      "step": 2710
    },
    {
      "epoch": 1.819397993311037,
      "grad_norm": 5.257035732269287,
      "learning_rate": 8.180602006688963e-05,
      "loss": 0.0616,
      "step": 2720
    },
    {
      "epoch": 1.8260869565217392,
      "grad_norm": 2.065892457962036,
      "learning_rate": 8.173913043478262e-05,
      "loss": 0.086,
      "step": 2730
    },
    {
      "epoch": 1.8327759197324416,
      "grad_norm": 1.7501225471496582,
      "learning_rate": 8.16722408026756e-05,
      "loss": 0.0744,
      "step": 2740
    },
    {
      "epoch": 1.839464882943144,
      "grad_norm": 7.9652557373046875,
      "learning_rate": 8.160535117056857e-05,
      "loss": 0.0991,
      "step": 2750
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 4.254938125610352,
      "learning_rate": 8.153846153846155e-05,
      "loss": 0.0789,
      "step": 2760
    },
    {
      "epoch": 1.8528428093645486,
      "grad_norm": 0.2958781123161316,
      "learning_rate": 8.147157190635452e-05,
      "loss": 0.0733,
      "step": 2770
    },
    {
      "epoch": 1.859531772575251,
      "grad_norm": 0.6094931364059448,
      "learning_rate": 8.140468227424748e-05,
      "loss": 0.079,
      "step": 2780
    },
    {
      "epoch": 1.8662207357859533,
      "grad_norm": 4.592105388641357,
      "learning_rate": 8.133779264214048e-05,
      "loss": 0.0717,
      "step": 2790
    },
    {
      "epoch": 1.8729096989966556,
      "grad_norm": 0.26445138454437256,
      "learning_rate": 8.127090301003345e-05,
      "loss": 0.0809,
      "step": 2800
    },
    {
      "epoch": 1.879598662207358,
      "grad_norm": 7.2576799392700195,
      "learning_rate": 8.120401337792643e-05,
      "loss": 0.064,
      "step": 2810
    },
    {
      "epoch": 1.8862876254180603,
      "grad_norm": 3.6506426334381104,
      "learning_rate": 8.11371237458194e-05,
      "loss": 0.0766,
      "step": 2820
    },
    {
      "epoch": 1.8929765886287626,
      "grad_norm": 1.1458144187927246,
      "learning_rate": 8.107023411371238e-05,
      "loss": 0.0811,
      "step": 2830
    },
    {
      "epoch": 1.899665551839465,
      "grad_norm": 2.733182191848755,
      "learning_rate": 8.100334448160536e-05,
      "loss": 0.0685,
      "step": 2840
    },
    {
      "epoch": 1.9063545150501673,
      "grad_norm": 2.7656662464141846,
      "learning_rate": 8.093645484949833e-05,
      "loss": 0.0772,
      "step": 2850
    },
    {
      "epoch": 1.9130434782608696,
      "grad_norm": 1.3046529293060303,
      "learning_rate": 8.086956521739131e-05,
      "loss": 0.0654,
      "step": 2860
    },
    {
      "epoch": 1.919732441471572,
      "grad_norm": 0.3139551281929016,
      "learning_rate": 8.080267558528428e-05,
      "loss": 0.0864,
      "step": 2870
    },
    {
      "epoch": 1.9264214046822743,
      "grad_norm": 0.244270458817482,
      "learning_rate": 8.073578595317726e-05,
      "loss": 0.0698,
      "step": 2880
    },
    {
      "epoch": 1.9331103678929766,
      "grad_norm": 1.8901079893112183,
      "learning_rate": 8.066889632107024e-05,
      "loss": 0.0912,
      "step": 2890
    },
    {
      "epoch": 1.939799331103679,
      "grad_norm": 3.75931715965271,
      "learning_rate": 8.060200668896321e-05,
      "loss": 0.0708,
      "step": 2900
    },
    {
      "epoch": 1.9464882943143813,
      "grad_norm": 3.791531801223755,
      "learning_rate": 8.053511705685619e-05,
      "loss": 0.0808,
      "step": 2910
    },
    {
      "epoch": 1.9531772575250836,
      "grad_norm": 9.437219619750977,
      "learning_rate": 8.046822742474916e-05,
      "loss": 0.1003,
      "step": 2920
    },
    {
      "epoch": 1.959866220735786,
      "grad_norm": 0.38832196593284607,
      "learning_rate": 8.040133779264214e-05,
      "loss": 0.0823,
      "step": 2930
    },
    {
      "epoch": 1.9665551839464883,
      "grad_norm": 4.084001541137695,
      "learning_rate": 8.033444816053512e-05,
      "loss": 0.0799,
      "step": 2940
    },
    {
      "epoch": 1.9732441471571907,
      "grad_norm": 6.1558837890625,
      "learning_rate": 8.026755852842809e-05,
      "loss": 0.0936,
      "step": 2950
    },
    {
      "epoch": 1.979933110367893,
      "grad_norm": 8.414342880249023,
      "learning_rate": 8.020066889632107e-05,
      "loss": 0.0786,
      "step": 2960
    },
    {
      "epoch": 1.9866220735785953,
      "grad_norm": 2.3542683124542236,
      "learning_rate": 8.013377926421404e-05,
      "loss": 0.0725,
      "step": 2970
    },
    {
      "epoch": 1.9933110367892977,
      "grad_norm": 0.4237585961818695,
      "learning_rate": 8.006688963210702e-05,
      "loss": 0.0882,
      "step": 2980
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.528468132019043,
      "learning_rate": 8e-05,
      "loss": 0.0833,
      "step": 2990
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.0793251246213913,
      "eval_mse": 0.0793251246213913,
      "eval_runtime": 230.277,
      "eval_samples_per_second": 12.984,
      "eval_steps_per_second": 1.624,
      "step": 2990
    },
    {
      "epoch": 2.0066889632107023,
      "grad_norm": 0.42202022671699524,
      "learning_rate": 7.993311036789299e-05,
      "loss": 0.0608,
      "step": 3000
    },
    {
      "epoch": 2.0133779264214047,
      "grad_norm": 2.761110544204712,
      "learning_rate": 7.986622073578596e-05,
      "loss": 0.0771,
      "step": 3010
    },
    {
      "epoch": 2.020066889632107,
      "grad_norm": 0.31655821204185486,
      "learning_rate": 7.979933110367892e-05,
      "loss": 0.0668,
      "step": 3020
    },
    {
      "epoch": 2.0267558528428093,
      "grad_norm": 0.8060646653175354,
      "learning_rate": 7.97324414715719e-05,
      "loss": 0.1023,
      "step": 3030
    },
    {
      "epoch": 2.0334448160535117,
      "grad_norm": 4.09045934677124,
      "learning_rate": 7.966555183946489e-05,
      "loss": 0.0823,
      "step": 3040
    },
    {
      "epoch": 2.040133779264214,
      "grad_norm": 3.433368682861328,
      "learning_rate": 7.959866220735787e-05,
      "loss": 0.0775,
      "step": 3050
    },
    {
      "epoch": 2.0468227424749164,
      "grad_norm": 2.6346817016601562,
      "learning_rate": 7.953177257525084e-05,
      "loss": 0.0725,
      "step": 3060
    },
    {
      "epoch": 2.0535117056856187,
      "grad_norm": 4.599377632141113,
      "learning_rate": 7.946488294314382e-05,
      "loss": 0.0823,
      "step": 3070
    },
    {
      "epoch": 2.060200668896321,
      "grad_norm": 4.843750476837158,
      "learning_rate": 7.939799331103679e-05,
      "loss": 0.0682,
      "step": 3080
    },
    {
      "epoch": 2.0668896321070234,
      "grad_norm": 3.9601993560791016,
      "learning_rate": 7.933110367892977e-05,
      "loss": 0.0775,
      "step": 3090
    },
    {
      "epoch": 2.0735785953177257,
      "grad_norm": 0.437264084815979,
      "learning_rate": 7.926421404682275e-05,
      "loss": 0.0603,
      "step": 3100
    },
    {
      "epoch": 2.080267558528428,
      "grad_norm": 2.8297762870788574,
      "learning_rate": 7.919732441471572e-05,
      "loss": 0.0815,
      "step": 3110
    },
    {
      "epoch": 2.0869565217391304,
      "grad_norm": 1.021713137626648,
      "learning_rate": 7.91304347826087e-05,
      "loss": 0.0575,
      "step": 3120
    },
    {
      "epoch": 2.0936454849498327,
      "grad_norm": 0.2781499922275543,
      "learning_rate": 7.906354515050167e-05,
      "loss": 0.0761,
      "step": 3130
    },
    {
      "epoch": 2.100334448160535,
      "grad_norm": 1.3915780782699585,
      "learning_rate": 7.899665551839465e-05,
      "loss": 0.0688,
      "step": 3140
    },
    {
      "epoch": 2.1070234113712374,
      "grad_norm": 1.546858787536621,
      "learning_rate": 7.892976588628763e-05,
      "loss": 0.0762,
      "step": 3150
    },
    {
      "epoch": 2.1137123745819397,
      "grad_norm": 3.6546237468719482,
      "learning_rate": 7.88628762541806e-05,
      "loss": 0.0569,
      "step": 3160
    },
    {
      "epoch": 2.120401337792642,
      "grad_norm": 3.8396148681640625,
      "learning_rate": 7.879598662207358e-05,
      "loss": 0.0783,
      "step": 3170
    },
    {
      "epoch": 2.1270903010033444,
      "grad_norm": 1.614401936531067,
      "learning_rate": 7.872909698996655e-05,
      "loss": 0.0782,
      "step": 3180
    },
    {
      "epoch": 2.1337792642140467,
      "grad_norm": 1.955979347229004,
      "learning_rate": 7.866220735785953e-05,
      "loss": 0.0659,
      "step": 3190
    },
    {
      "epoch": 2.140468227424749,
      "grad_norm": 10.376070976257324,
      "learning_rate": 7.859531772575252e-05,
      "loss": 0.0948,
      "step": 3200
    },
    {
      "epoch": 2.1471571906354514,
      "grad_norm": 0.34439998865127563,
      "learning_rate": 7.852842809364548e-05,
      "loss": 0.1018,
      "step": 3210
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 5.3788862228393555,
      "learning_rate": 7.846153846153847e-05,
      "loss": 0.0891,
      "step": 3220
    },
    {
      "epoch": 2.160535117056856,
      "grad_norm": 1.8117215633392334,
      "learning_rate": 7.839464882943143e-05,
      "loss": 0.0854,
      "step": 3230
    },
    {
      "epoch": 2.1672240802675584,
      "grad_norm": 2.610339403152466,
      "learning_rate": 7.832775919732442e-05,
      "loss": 0.0753,
      "step": 3240
    },
    {
      "epoch": 2.1739130434782608,
      "grad_norm": 1.4385907649993896,
      "learning_rate": 7.82608695652174e-05,
      "loss": 0.0785,
      "step": 3250
    },
    {
      "epoch": 2.180602006688963,
      "grad_norm": 6.955935001373291,
      "learning_rate": 7.819397993311037e-05,
      "loss": 0.074,
      "step": 3260
    },
    {
      "epoch": 2.1872909698996654,
      "grad_norm": 1.2500567436218262,
      "learning_rate": 7.812709030100335e-05,
      "loss": 0.0788,
      "step": 3270
    },
    {
      "epoch": 2.1939799331103678,
      "grad_norm": 3.9600324630737305,
      "learning_rate": 7.806020066889632e-05,
      "loss": 0.0749,
      "step": 3280
    },
    {
      "epoch": 2.20066889632107,
      "grad_norm": 7.548783302307129,
      "learning_rate": 7.79933110367893e-05,
      "loss": 0.0903,
      "step": 3290
    },
    {
      "epoch": 2.2073578595317724,
      "grad_norm": 0.5083264708518982,
      "learning_rate": 7.792642140468228e-05,
      "loss": 0.07,
      "step": 3300
    },
    {
      "epoch": 2.2140468227424748,
      "grad_norm": 0.17018166184425354,
      "learning_rate": 7.785953177257526e-05,
      "loss": 0.0867,
      "step": 3310
    },
    {
      "epoch": 2.220735785953177,
      "grad_norm": 1.3731433153152466,
      "learning_rate": 7.779264214046823e-05,
      "loss": 0.0653,
      "step": 3320
    },
    {
      "epoch": 2.2274247491638794,
      "grad_norm": 1.8665434122085571,
      "learning_rate": 7.77257525083612e-05,
      "loss": 0.0951,
      "step": 3330
    },
    {
      "epoch": 2.234113712374582,
      "grad_norm": 4.514686107635498,
      "learning_rate": 7.765886287625418e-05,
      "loss": 0.0513,
      "step": 3340
    },
    {
      "epoch": 2.240802675585284,
      "grad_norm": 0.6165162324905396,
      "learning_rate": 7.759197324414716e-05,
      "loss": 0.0676,
      "step": 3350
    },
    {
      "epoch": 2.2474916387959865,
      "grad_norm": 1.839833378791809,
      "learning_rate": 7.752508361204014e-05,
      "loss": 0.1025,
      "step": 3360
    },
    {
      "epoch": 2.254180602006689,
      "grad_norm": 2.2654104232788086,
      "learning_rate": 7.745819397993311e-05,
      "loss": 0.0821,
      "step": 3370
    },
    {
      "epoch": 2.260869565217391,
      "grad_norm": 6.743686199188232,
      "learning_rate": 7.73913043478261e-05,
      "loss": 0.0741,
      "step": 3380
    },
    {
      "epoch": 2.2675585284280935,
      "grad_norm": 3.1228573322296143,
      "learning_rate": 7.732441471571906e-05,
      "loss": 0.0797,
      "step": 3390
    },
    {
      "epoch": 2.274247491638796,
      "grad_norm": 0.12537412345409393,
      "learning_rate": 7.725752508361204e-05,
      "loss": 0.0856,
      "step": 3400
    },
    {
      "epoch": 2.280936454849498,
      "grad_norm": 1.829524278640747,
      "learning_rate": 7.719063545150503e-05,
      "loss": 0.084,
      "step": 3410
    },
    {
      "epoch": 2.2876254180602005,
      "grad_norm": 3.228325366973877,
      "learning_rate": 7.7123745819398e-05,
      "loss": 0.0894,
      "step": 3420
    },
    {
      "epoch": 2.294314381270903,
      "grad_norm": 4.834877014160156,
      "learning_rate": 7.705685618729098e-05,
      "loss": 0.0778,
      "step": 3430
    },
    {
      "epoch": 2.3010033444816056,
      "grad_norm": 4.732447624206543,
      "learning_rate": 7.698996655518394e-05,
      "loss": 0.1107,
      "step": 3440
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 2.6647109985351562,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0858,
      "step": 3450
    },
    {
      "epoch": 2.3143812709030103,
      "grad_norm": 4.5095415115356445,
      "learning_rate": 7.685618729096991e-05,
      "loss": 0.0825,
      "step": 3460
    },
    {
      "epoch": 2.321070234113712,
      "grad_norm": 1.0812541246414185,
      "learning_rate": 7.678929765886288e-05,
      "loss": 0.081,
      "step": 3470
    },
    {
      "epoch": 2.327759197324415,
      "grad_norm": 7.102626800537109,
      "learning_rate": 7.672240802675586e-05,
      "loss": 0.0928,
      "step": 3480
    },
    {
      "epoch": 2.334448160535117,
      "grad_norm": 4.619367599487305,
      "learning_rate": 7.665551839464883e-05,
      "loss": 0.0918,
      "step": 3490
    },
    {
      "epoch": 2.3411371237458196,
      "grad_norm": 3.246306896209717,
      "learning_rate": 7.658862876254181e-05,
      "loss": 0.079,
      "step": 3500
    },
    {
      "epoch": 2.3478260869565215,
      "grad_norm": 4.399368762969971,
      "learning_rate": 7.652173913043479e-05,
      "loss": 0.0943,
      "step": 3510
    },
    {
      "epoch": 2.3545150501672243,
      "grad_norm": 1.4770549535751343,
      "learning_rate": 7.645484949832776e-05,
      "loss": 0.0861,
      "step": 3520
    },
    {
      "epoch": 2.361204013377926,
      "grad_norm": 3.9449102878570557,
      "learning_rate": 7.638795986622074e-05,
      "loss": 0.082,
      "step": 3530
    },
    {
      "epoch": 2.367892976588629,
      "grad_norm": 3.8862507343292236,
      "learning_rate": 7.632107023411371e-05,
      "loss": 0.0703,
      "step": 3540
    },
    {
      "epoch": 2.374581939799331,
      "grad_norm": 3.4087064266204834,
      "learning_rate": 7.62541806020067e-05,
      "loss": 0.0915,
      "step": 3550
    },
    {
      "epoch": 2.3812709030100336,
      "grad_norm": 0.5102267265319824,
      "learning_rate": 7.618729096989967e-05,
      "loss": 0.094,
      "step": 3560
    },
    {
      "epoch": 2.387959866220736,
      "grad_norm": 1.3995593786239624,
      "learning_rate": 7.612040133779264e-05,
      "loss": 0.0895,
      "step": 3570
    },
    {
      "epoch": 2.3946488294314383,
      "grad_norm": 1.626395583152771,
      "learning_rate": 7.605351170568562e-05,
      "loss": 0.0999,
      "step": 3580
    },
    {
      "epoch": 2.4013377926421406,
      "grad_norm": 1.7516860961914062,
      "learning_rate": 7.598662207357859e-05,
      "loss": 0.0853,
      "step": 3590
    },
    {
      "epoch": 2.408026755852843,
      "grad_norm": 5.477409362792969,
      "learning_rate": 7.591973244147159e-05,
      "loss": 0.0775,
      "step": 3600
    },
    {
      "epoch": 2.4147157190635453,
      "grad_norm": 4.511748313903809,
      "learning_rate": 7.585284280936455e-05,
      "loss": 0.0806,
      "step": 3610
    },
    {
      "epoch": 2.4214046822742477,
      "grad_norm": 3.8276188373565674,
      "learning_rate": 7.578595317725754e-05,
      "loss": 0.0785,
      "step": 3620
    },
    {
      "epoch": 2.42809364548495,
      "grad_norm": 3.8297007083892822,
      "learning_rate": 7.57190635451505e-05,
      "loss": 0.0809,
      "step": 3630
    },
    {
      "epoch": 2.4347826086956523,
      "grad_norm": 3.4355835914611816,
      "learning_rate": 7.565217391304347e-05,
      "loss": 0.0806,
      "step": 3640
    },
    {
      "epoch": 2.4414715719063547,
      "grad_norm": 5.955551624298096,
      "learning_rate": 7.558528428093647e-05,
      "loss": 0.076,
      "step": 3650
    },
    {
      "epoch": 2.448160535117057,
      "grad_norm": 0.3564576804637909,
      "learning_rate": 7.551839464882944e-05,
      "loss": 0.0885,
      "step": 3660
    },
    {
      "epoch": 2.4548494983277593,
      "grad_norm": 0.8594129681587219,
      "learning_rate": 7.545150501672242e-05,
      "loss": 0.0723,
      "step": 3670
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 5.289486408233643,
      "learning_rate": 7.538461538461539e-05,
      "loss": 0.1042,
      "step": 3680
    },
    {
      "epoch": 2.468227424749164,
      "grad_norm": 3.3944950103759766,
      "learning_rate": 7.531772575250837e-05,
      "loss": 0.0716,
      "step": 3690
    },
    {
      "epoch": 2.4749163879598663,
      "grad_norm": 2.454094171524048,
      "learning_rate": 7.525083612040135e-05,
      "loss": 0.0709,
      "step": 3700
    },
    {
      "epoch": 2.4816053511705687,
      "grad_norm": 0.41697606444358826,
      "learning_rate": 7.518394648829432e-05,
      "loss": 0.101,
      "step": 3710
    },
    {
      "epoch": 2.488294314381271,
      "grad_norm": 3.9693078994750977,
      "learning_rate": 7.51170568561873e-05,
      "loss": 0.0779,
      "step": 3720
    },
    {
      "epoch": 2.4949832775919734,
      "grad_norm": 4.420345306396484,
      "learning_rate": 7.505016722408027e-05,
      "loss": 0.0738,
      "step": 3730
    },
    {
      "epoch": 2.5016722408026757,
      "grad_norm": 1.1933796405792236,
      "learning_rate": 7.498327759197325e-05,
      "loss": 0.085,
      "step": 3740
    },
    {
      "epoch": 2.508361204013378,
      "grad_norm": 0.7892850637435913,
      "learning_rate": 7.491638795986622e-05,
      "loss": 0.0715,
      "step": 3750
    },
    {
      "epoch": 2.5150501672240804,
      "grad_norm": 5.415370941162109,
      "learning_rate": 7.48494983277592e-05,
      "loss": 0.0833,
      "step": 3760
    },
    {
      "epoch": 2.5217391304347827,
      "grad_norm": 2.976743698120117,
      "learning_rate": 7.478260869565218e-05,
      "loss": 0.0747,
      "step": 3770
    },
    {
      "epoch": 2.528428093645485,
      "grad_norm": 4.22029972076416,
      "learning_rate": 7.471571906354515e-05,
      "loss": 0.0747,
      "step": 3780
    },
    {
      "epoch": 2.5351170568561874,
      "grad_norm": 0.27683568000793457,
      "learning_rate": 7.464882943143813e-05,
      "loss": 0.0635,
      "step": 3790
    },
    {
      "epoch": 2.5418060200668897,
      "grad_norm": 3.5258395671844482,
      "learning_rate": 7.45819397993311e-05,
      "loss": 0.0847,
      "step": 3800
    },
    {
      "epoch": 2.548494983277592,
      "grad_norm": 1.4285045862197876,
      "learning_rate": 7.451505016722408e-05,
      "loss": 0.0561,
      "step": 3810
    },
    {
      "epoch": 2.5551839464882944,
      "grad_norm": 3.0891616344451904,
      "learning_rate": 7.444816053511706e-05,
      "loss": 0.0738,
      "step": 3820
    },
    {
      "epoch": 2.5618729096989967,
      "grad_norm": 2.099480628967285,
      "learning_rate": 7.438127090301003e-05,
      "loss": 0.1008,
      "step": 3830
    },
    {
      "epoch": 2.568561872909699,
      "grad_norm": 0.7076812982559204,
      "learning_rate": 7.431438127090301e-05,
      "loss": 0.0709,
      "step": 3840
    },
    {
      "epoch": 2.5752508361204014,
      "grad_norm": 4.342235088348389,
      "learning_rate": 7.424749163879598e-05,
      "loss": 0.0722,
      "step": 3850
    },
    {
      "epoch": 2.5819397993311037,
      "grad_norm": 2.153355360031128,
      "learning_rate": 7.418060200668898e-05,
      "loss": 0.0706,
      "step": 3860
    },
    {
      "epoch": 2.588628762541806,
      "grad_norm": 2.8034729957580566,
      "learning_rate": 7.411371237458194e-05,
      "loss": 0.0801,
      "step": 3870
    },
    {
      "epoch": 2.5953177257525084,
      "grad_norm": 2.267157554626465,
      "learning_rate": 7.404682274247491e-05,
      "loss": 0.0922,
      "step": 3880
    },
    {
      "epoch": 2.6020066889632107,
      "grad_norm": 2.9926929473876953,
      "learning_rate": 7.39799331103679e-05,
      "loss": 0.0845,
      "step": 3890
    },
    {
      "epoch": 2.608695652173913,
      "grad_norm": 3.7677009105682373,
      "learning_rate": 7.391304347826086e-05,
      "loss": 0.0769,
      "step": 3900
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 6.213174819946289,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.092,
      "step": 3910
    },
    {
      "epoch": 2.6220735785953178,
      "grad_norm": 0.17510397732257843,
      "learning_rate": 7.377926421404683e-05,
      "loss": 0.0758,
      "step": 3920
    },
    {
      "epoch": 2.62876254180602,
      "grad_norm": 0.8683602213859558,
      "learning_rate": 7.371237458193981e-05,
      "loss": 0.0888,
      "step": 3930
    },
    {
      "epoch": 2.6354515050167224,
      "grad_norm": 2.76550030708313,
      "learning_rate": 7.364548494983278e-05,
      "loss": 0.0666,
      "step": 3940
    },
    {
      "epoch": 2.6421404682274248,
      "grad_norm": 0.7504422068595886,
      "learning_rate": 7.357859531772575e-05,
      "loss": 0.0674,
      "step": 3950
    },
    {
      "epoch": 2.648829431438127,
      "grad_norm": 1.3935351371765137,
      "learning_rate": 7.351170568561874e-05,
      "loss": 0.0858,
      "step": 3960
    },
    {
      "epoch": 2.6555183946488294,
      "grad_norm": 1.4868053197860718,
      "learning_rate": 7.344481605351171e-05,
      "loss": 0.0853,
      "step": 3970
    },
    {
      "epoch": 2.6622073578595318,
      "grad_norm": 4.424736976623535,
      "learning_rate": 7.337792642140469e-05,
      "loss": 0.0801,
      "step": 3980
    },
    {
      "epoch": 2.668896321070234,
      "grad_norm": 2.0431973934173584,
      "learning_rate": 7.331103678929766e-05,
      "loss": 0.0769,
      "step": 3990
    },
    {
      "epoch": 2.6755852842809364,
      "grad_norm": 1.8978101015090942,
      "learning_rate": 7.324414715719064e-05,
      "loss": 0.0883,
      "step": 4000
    },
    {
      "epoch": 2.682274247491639,
      "grad_norm": 3.53125262260437,
      "learning_rate": 7.317725752508362e-05,
      "loss": 0.0925,
      "step": 4010
    },
    {
      "epoch": 2.688963210702341,
      "grad_norm": 1.823233962059021,
      "learning_rate": 7.311036789297659e-05,
      "loss": 0.0686,
      "step": 4020
    },
    {
      "epoch": 2.6956521739130435,
      "grad_norm": 0.9586164951324463,
      "learning_rate": 7.304347826086957e-05,
      "loss": 0.0779,
      "step": 4030
    },
    {
      "epoch": 2.702341137123746,
      "grad_norm": 3.304990768432617,
      "learning_rate": 7.297658862876254e-05,
      "loss": 0.0716,
      "step": 4040
    },
    {
      "epoch": 2.709030100334448,
      "grad_norm": 5.012158393859863,
      "learning_rate": 7.290969899665552e-05,
      "loss": 0.0733,
      "step": 4050
    },
    {
      "epoch": 2.7157190635451505,
      "grad_norm": 4.342820167541504,
      "learning_rate": 7.28428093645485e-05,
      "loss": 0.0782,
      "step": 4060
    },
    {
      "epoch": 2.722408026755853,
      "grad_norm": 4.751095294952393,
      "learning_rate": 7.277591973244147e-05,
      "loss": 0.0687,
      "step": 4070
    },
    {
      "epoch": 2.729096989966555,
      "grad_norm": 1.3369343280792236,
      "learning_rate": 7.270903010033445e-05,
      "loss": 0.0693,
      "step": 4080
    },
    {
      "epoch": 2.7357859531772575,
      "grad_norm": 2.1024954319000244,
      "learning_rate": 7.264214046822742e-05,
      "loss": 0.0776,
      "step": 4090
    },
    {
      "epoch": 2.74247491638796,
      "grad_norm": 8.521135330200195,
      "learning_rate": 7.25752508361204e-05,
      "loss": 0.0761,
      "step": 4100
    },
    {
      "epoch": 2.749163879598662,
      "grad_norm": 1.152764916419983,
      "learning_rate": 7.250836120401339e-05,
      "loss": 0.0959,
      "step": 4110
    },
    {
      "epoch": 2.7558528428093645,
      "grad_norm": 2.0863466262817383,
      "learning_rate": 7.244147157190635e-05,
      "loss": 0.067,
      "step": 4120
    },
    {
      "epoch": 2.762541806020067,
      "grad_norm": 0.3307766318321228,
      "learning_rate": 7.237458193979934e-05,
      "loss": 0.0771,
      "step": 4130
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 1.6041626930236816,
      "learning_rate": 7.23076923076923e-05,
      "loss": 0.0711,
      "step": 4140
    },
    {
      "epoch": 2.7759197324414715,
      "grad_norm": 0.37076133489608765,
      "learning_rate": 7.224080267558529e-05,
      "loss": 0.0692,
      "step": 4150
    },
    {
      "epoch": 2.782608695652174,
      "grad_norm": 1.4220796823501587,
      "learning_rate": 7.217391304347827e-05,
      "loss": 0.0867,
      "step": 4160
    },
    {
      "epoch": 2.789297658862876,
      "grad_norm": 6.7247090339660645,
      "learning_rate": 7.210702341137125e-05,
      "loss": 0.0932,
      "step": 4170
    },
    {
      "epoch": 2.7959866220735785,
      "grad_norm": 0.6634153127670288,
      "learning_rate": 7.204013377926422e-05,
      "loss": 0.077,
      "step": 4180
    },
    {
      "epoch": 2.802675585284281,
      "grad_norm": 1.5781530141830444,
      "learning_rate": 7.197324414715719e-05,
      "loss": 0.0726,
      "step": 4190
    },
    {
      "epoch": 2.809364548494983,
      "grad_norm": 1.701782464981079,
      "learning_rate": 7.190635451505017e-05,
      "loss": 0.0616,
      "step": 4200
    },
    {
      "epoch": 2.8160535117056855,
      "grad_norm": 2.859126329421997,
      "learning_rate": 7.183946488294315e-05,
      "loss": 0.088,
      "step": 4210
    },
    {
      "epoch": 2.822742474916388,
      "grad_norm": 1.8869216442108154,
      "learning_rate": 7.177257525083613e-05,
      "loss": 0.0766,
      "step": 4220
    },
    {
      "epoch": 2.82943143812709,
      "grad_norm": 0.26754462718963623,
      "learning_rate": 7.17056856187291e-05,
      "loss": 0.0771,
      "step": 4230
    },
    {
      "epoch": 2.8361204013377925,
      "grad_norm": 0.4016328454017639,
      "learning_rate": 7.163879598662208e-05,
      "loss": 0.0887,
      "step": 4240
    },
    {
      "epoch": 2.842809364548495,
      "grad_norm": 3.2005503177642822,
      "learning_rate": 7.157190635451505e-05,
      "loss": 0.0742,
      "step": 4250
    },
    {
      "epoch": 2.849498327759197,
      "grad_norm": 3.5342941284179688,
      "learning_rate": 7.150501672240802e-05,
      "loss": 0.084,
      "step": 4260
    },
    {
      "epoch": 2.8561872909698995,
      "grad_norm": 1.4917818307876587,
      "learning_rate": 7.143812709030101e-05,
      "loss": 0.0885,
      "step": 4270
    },
    {
      "epoch": 2.862876254180602,
      "grad_norm": 4.428409099578857,
      "learning_rate": 7.137123745819398e-05,
      "loss": 0.0918,
      "step": 4280
    },
    {
      "epoch": 2.869565217391304,
      "grad_norm": 1.8964093923568726,
      "learning_rate": 7.130434782608696e-05,
      "loss": 0.0719,
      "step": 4290
    },
    {
      "epoch": 2.8762541806020065,
      "grad_norm": 0.6107605695724487,
      "learning_rate": 7.123745819397993e-05,
      "loss": 0.0815,
      "step": 4300
    },
    {
      "epoch": 2.882943143812709,
      "grad_norm": 0.24433688819408417,
      "learning_rate": 7.117056856187291e-05,
      "loss": 0.0829,
      "step": 4310
    },
    {
      "epoch": 2.8896321070234112,
      "grad_norm": 7.567927360534668,
      "learning_rate": 7.11036789297659e-05,
      "loss": 0.0786,
      "step": 4320
    },
    {
      "epoch": 2.8963210702341136,
      "grad_norm": 2.2083115577697754,
      "learning_rate": 7.103678929765886e-05,
      "loss": 0.0766,
      "step": 4330
    },
    {
      "epoch": 2.903010033444816,
      "grad_norm": 0.35634785890579224,
      "learning_rate": 7.096989966555185e-05,
      "loss": 0.0723,
      "step": 4340
    },
    {
      "epoch": 2.9096989966555182,
      "grad_norm": 0.25455230474472046,
      "learning_rate": 7.090301003344481e-05,
      "loss": 0.0866,
      "step": 4350
    },
    {
      "epoch": 2.9163879598662206,
      "grad_norm": 3.7779791355133057,
      "learning_rate": 7.08361204013378e-05,
      "loss": 0.0655,
      "step": 4360
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 8.164266586303711,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0778,
      "step": 4370
    },
    {
      "epoch": 2.9297658862876252,
      "grad_norm": 1.9915484189987183,
      "learning_rate": 7.070234113712375e-05,
      "loss": 0.0705,
      "step": 4380
    },
    {
      "epoch": 2.936454849498328,
      "grad_norm": 0.46880605816841125,
      "learning_rate": 7.063545150501673e-05,
      "loss": 0.0749,
      "step": 4390
    },
    {
      "epoch": 2.94314381270903,
      "grad_norm": 0.22903496026992798,
      "learning_rate": 7.05685618729097e-05,
      "loss": 0.0599,
      "step": 4400
    },
    {
      "epoch": 2.9498327759197327,
      "grad_norm": 0.978259265422821,
      "learning_rate": 7.050167224080268e-05,
      "loss": 0.0745,
      "step": 4410
    },
    {
      "epoch": 2.9565217391304346,
      "grad_norm": 2.0547633171081543,
      "learning_rate": 7.043478260869566e-05,
      "loss": 0.0722,
      "step": 4420
    },
    {
      "epoch": 2.9632107023411374,
      "grad_norm": 3.652096748352051,
      "learning_rate": 7.036789297658863e-05,
      "loss": 0.0871,
      "step": 4430
    },
    {
      "epoch": 2.9698996655518393,
      "grad_norm": 3.154127836227417,
      "learning_rate": 7.030100334448161e-05,
      "loss": 0.0756,
      "step": 4440
    },
    {
      "epoch": 2.976588628762542,
      "grad_norm": 0.7841221690177917,
      "learning_rate": 7.023411371237458e-05,
      "loss": 0.0825,
      "step": 4450
    },
    {
      "epoch": 2.983277591973244,
      "grad_norm": 0.5690768957138062,
      "learning_rate": 7.016722408026756e-05,
      "loss": 0.0615,
      "step": 4460
    },
    {
      "epoch": 2.9899665551839467,
      "grad_norm": 4.73695182800293,
      "learning_rate": 7.010033444816054e-05,
      "loss": 0.0763,
      "step": 4470
    },
    {
      "epoch": 2.9966555183946486,
      "grad_norm": 3.4800970554351807,
      "learning_rate": 7.003344481605352e-05,
      "loss": 0.074,
      "step": 4480
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.08599434792995453,
      "eval_mse": 0.08599434792995453,
      "eval_runtime": 220.1132,
      "eval_samples_per_second": 13.584,
      "eval_steps_per_second": 1.699,
      "step": 4485
    },
    {
      "epoch": 3.0033444816053514,
      "grad_norm": 0.4188951551914215,
      "learning_rate": 6.996655518394649e-05,
      "loss": 0.0636,
      "step": 4490
    },
    {
      "epoch": 3.0100334448160537,
      "grad_norm": 0.22950677573680878,
      "learning_rate": 6.989966555183946e-05,
      "loss": 0.0741,
      "step": 4500
    },
    {
      "epoch": 3.016722408026756,
      "grad_norm": 5.341685771942139,
      "learning_rate": 6.983277591973244e-05,
      "loss": 0.0804,
      "step": 4510
    },
    {
      "epoch": 3.0234113712374584,
      "grad_norm": 3.4862024784088135,
      "learning_rate": 6.976588628762542e-05,
      "loss": 0.0872,
      "step": 4520
    },
    {
      "epoch": 3.0301003344481607,
      "grad_norm": 0.4761682450771332,
      "learning_rate": 6.96989966555184e-05,
      "loss": 0.0795,
      "step": 4530
    },
    {
      "epoch": 3.036789297658863,
      "grad_norm": 0.49775540828704834,
      "learning_rate": 6.963210702341137e-05,
      "loss": 0.0818,
      "step": 4540
    },
    {
      "epoch": 3.0434782608695654,
      "grad_norm": 2.6934797763824463,
      "learning_rate": 6.956521739130436e-05,
      "loss": 0.0588,
      "step": 4550
    },
    {
      "epoch": 3.0501672240802677,
      "grad_norm": 1.939240574836731,
      "learning_rate": 6.949832775919732e-05,
      "loss": 0.0835,
      "step": 4560
    },
    {
      "epoch": 3.05685618729097,
      "grad_norm": 0.32804200053215027,
      "learning_rate": 6.94314381270903e-05,
      "loss": 0.0583,
      "step": 4570
    },
    {
      "epoch": 3.0635451505016724,
      "grad_norm": 3.560800075531006,
      "learning_rate": 6.936454849498329e-05,
      "loss": 0.0757,
      "step": 4580
    },
    {
      "epoch": 3.0702341137123748,
      "grad_norm": 5.4448065757751465,
      "learning_rate": 6.929765886287626e-05,
      "loss": 0.0695,
      "step": 4590
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 3.402082920074463,
      "learning_rate": 6.923076923076924e-05,
      "loss": 0.0689,
      "step": 4600
    },
    {
      "epoch": 3.0836120401337794,
      "grad_norm": 1.682234764099121,
      "learning_rate": 6.91638795986622e-05,
      "loss": 0.069,
      "step": 4610
    },
    {
      "epoch": 3.0903010033444818,
      "grad_norm": 2.9638731479644775,
      "learning_rate": 6.909698996655519e-05,
      "loss": 0.0621,
      "step": 4620
    },
    {
      "epoch": 3.096989966555184,
      "grad_norm": 2.818448543548584,
      "learning_rate": 6.903010033444817e-05,
      "loss": 0.0723,
      "step": 4630
    },
    {
      "epoch": 3.1036789297658864,
      "grad_norm": 2.1265201568603516,
      "learning_rate": 6.896321070234114e-05,
      "loss": 0.1035,
      "step": 4640
    },
    {
      "epoch": 3.1103678929765888,
      "grad_norm": 2.025054931640625,
      "learning_rate": 6.889632107023412e-05,
      "loss": 0.0769,
      "step": 4650
    },
    {
      "epoch": 3.117056856187291,
      "grad_norm": 3.5865674018859863,
      "learning_rate": 6.882943143812709e-05,
      "loss": 0.0784,
      "step": 4660
    },
    {
      "epoch": 3.1237458193979935,
      "grad_norm": 4.239811420440674,
      "learning_rate": 6.876254180602007e-05,
      "loss": 0.0805,
      "step": 4670
    },
    {
      "epoch": 3.130434782608696,
      "grad_norm": 8.931589126586914,
      "learning_rate": 6.869565217391305e-05,
      "loss": 0.0751,
      "step": 4680
    },
    {
      "epoch": 3.137123745819398,
      "grad_norm": 6.283422470092773,
      "learning_rate": 6.862876254180602e-05,
      "loss": 0.0885,
      "step": 4690
    },
    {
      "epoch": 3.1438127090301005,
      "grad_norm": 6.964988708496094,
      "learning_rate": 6.8561872909699e-05,
      "loss": 0.0881,
      "step": 4700
    },
    {
      "epoch": 3.150501672240803,
      "grad_norm": 2.8225507736206055,
      "learning_rate": 6.849498327759197e-05,
      "loss": 0.0722,
      "step": 4710
    },
    {
      "epoch": 3.157190635451505,
      "grad_norm": 0.5449556708335876,
      "learning_rate": 6.842809364548496e-05,
      "loss": 0.0861,
      "step": 4720
    },
    {
      "epoch": 3.1638795986622075,
      "grad_norm": 4.718214511871338,
      "learning_rate": 6.836120401337793e-05,
      "loss": 0.085,
      "step": 4730
    },
    {
      "epoch": 3.17056856187291,
      "grad_norm": 3.2649168968200684,
      "learning_rate": 6.82943143812709e-05,
      "loss": 0.0658,
      "step": 4740
    },
    {
      "epoch": 3.177257525083612,
      "grad_norm": 1.1325840950012207,
      "learning_rate": 6.822742474916388e-05,
      "loss": 0.0711,
      "step": 4750
    },
    {
      "epoch": 3.1839464882943145,
      "grad_norm": 1.378689169883728,
      "learning_rate": 6.816053511705685e-05,
      "loss": 0.0538,
      "step": 4760
    },
    {
      "epoch": 3.190635451505017,
      "grad_norm": 2.0602452754974365,
      "learning_rate": 6.809364548494983e-05,
      "loss": 0.0746,
      "step": 4770
    },
    {
      "epoch": 3.197324414715719,
      "grad_norm": 5.678889274597168,
      "learning_rate": 6.802675585284281e-05,
      "loss": 0.0891,
      "step": 4780
    },
    {
      "epoch": 3.2040133779264215,
      "grad_norm": 1.952165126800537,
      "learning_rate": 6.79598662207358e-05,
      "loss": 0.0844,
      "step": 4790
    },
    {
      "epoch": 3.210702341137124,
      "grad_norm": 2.186636447906494,
      "learning_rate": 6.789297658862876e-05,
      "loss": 0.063,
      "step": 4800
    },
    {
      "epoch": 3.217391304347826,
      "grad_norm": 0.32168132066726685,
      "learning_rate": 6.782608695652173e-05,
      "loss": 0.0873,
      "step": 4810
    },
    {
      "epoch": 3.2240802675585285,
      "grad_norm": 4.04616641998291,
      "learning_rate": 6.775919732441471e-05,
      "loss": 0.0696,
      "step": 4820
    },
    {
      "epoch": 3.230769230769231,
      "grad_norm": 1.4718605279922485,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0719,
      "step": 4830
    },
    {
      "epoch": 3.237458193979933,
      "grad_norm": 3.9757933616638184,
      "learning_rate": 6.762541806020068e-05,
      "loss": 0.0888,
      "step": 4840
    },
    {
      "epoch": 3.2441471571906355,
      "grad_norm": 0.3360963463783264,
      "learning_rate": 6.755852842809365e-05,
      "loss": 0.0789,
      "step": 4850
    },
    {
      "epoch": 3.250836120401338,
      "grad_norm": 3.2653157711029053,
      "learning_rate": 6.749163879598663e-05,
      "loss": 0.0757,
      "step": 4860
    },
    {
      "epoch": 3.25752508361204,
      "grad_norm": 1.1410871744155884,
      "learning_rate": 6.74247491638796e-05,
      "loss": 0.0675,
      "step": 4870
    },
    {
      "epoch": 3.2642140468227425,
      "grad_norm": 2.160099506378174,
      "learning_rate": 6.735785953177258e-05,
      "loss": 0.0696,
      "step": 4880
    },
    {
      "epoch": 3.270903010033445,
      "grad_norm": 3.953361749649048,
      "learning_rate": 6.729096989966556e-05,
      "loss": 0.0761,
      "step": 4890
    },
    {
      "epoch": 3.277591973244147,
      "grad_norm": 2.248314380645752,
      "learning_rate": 6.722408026755853e-05,
      "loss": 0.0819,
      "step": 4900
    },
    {
      "epoch": 3.2842809364548495,
      "grad_norm": 1.036260724067688,
      "learning_rate": 6.715719063545151e-05,
      "loss": 0.0853,
      "step": 4910
    },
    {
      "epoch": 3.290969899665552,
      "grad_norm": 1.3433880805969238,
      "learning_rate": 6.709030100334448e-05,
      "loss": 0.0715,
      "step": 4920
    },
    {
      "epoch": 3.297658862876254,
      "grad_norm": 2.211595296859741,
      "learning_rate": 6.702341137123746e-05,
      "loss": 0.0541,
      "step": 4930
    },
    {
      "epoch": 3.3043478260869565,
      "grad_norm": 6.310031414031982,
      "learning_rate": 6.695652173913044e-05,
      "loss": 0.0765,
      "step": 4940
    },
    {
      "epoch": 3.311036789297659,
      "grad_norm": 1.9298906326293945,
      "learning_rate": 6.688963210702341e-05,
      "loss": 0.0766,
      "step": 4950
    },
    {
      "epoch": 3.317725752508361,
      "grad_norm": 0.2267463505268097,
      "learning_rate": 6.682274247491639e-05,
      "loss": 0.0627,
      "step": 4960
    },
    {
      "epoch": 3.3244147157190636,
      "grad_norm": 0.6416743993759155,
      "learning_rate": 6.675585284280936e-05,
      "loss": 0.0775,
      "step": 4970
    },
    {
      "epoch": 3.331103678929766,
      "grad_norm": 3.208157539367676,
      "learning_rate": 6.668896321070234e-05,
      "loss": 0.0783,
      "step": 4980
    },
    {
      "epoch": 3.3377926421404682,
      "grad_norm": 1.661583423614502,
      "learning_rate": 6.662207357859532e-05,
      "loss": 0.1092,
      "step": 4990
    },
    {
      "epoch": 3.3444816053511706,
      "grad_norm": 0.8186647295951843,
      "learning_rate": 6.655518394648829e-05,
      "loss": 0.0831,
      "step": 5000
    },
    {
      "epoch": 3.351170568561873,
      "grad_norm": 7.43472957611084,
      "learning_rate": 6.648829431438127e-05,
      "loss": 0.0948,
      "step": 5010
    },
    {
      "epoch": 3.3578595317725752,
      "grad_norm": 1.8813217878341675,
      "learning_rate": 6.642140468227424e-05,
      "loss": 0.076,
      "step": 5020
    },
    {
      "epoch": 3.3645484949832776,
      "grad_norm": 3.208296060562134,
      "learning_rate": 6.635451505016724e-05,
      "loss": 0.0695,
      "step": 5030
    },
    {
      "epoch": 3.37123745819398,
      "grad_norm": 1.160740613937378,
      "learning_rate": 6.62876254180602e-05,
      "loss": 0.0713,
      "step": 5040
    },
    {
      "epoch": 3.3779264214046822,
      "grad_norm": 4.097474575042725,
      "learning_rate": 6.622073578595317e-05,
      "loss": 0.0714,
      "step": 5050
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 1.559029221534729,
      "learning_rate": 6.615384615384616e-05,
      "loss": 0.0682,
      "step": 5060
    },
    {
      "epoch": 3.391304347826087,
      "grad_norm": 0.24774087965488434,
      "learning_rate": 6.608695652173912e-05,
      "loss": 0.0797,
      "step": 5070
    },
    {
      "epoch": 3.3979933110367893,
      "grad_norm": 3.935304880142212,
      "learning_rate": 6.602006688963212e-05,
      "loss": 0.0659,
      "step": 5080
    },
    {
      "epoch": 3.4046822742474916,
      "grad_norm": 1.9889538288116455,
      "learning_rate": 6.595317725752509e-05,
      "loss": 0.0628,
      "step": 5090
    },
    {
      "epoch": 3.411371237458194,
      "grad_norm": 6.523388385772705,
      "learning_rate": 6.588628762541807e-05,
      "loss": 0.0844,
      "step": 5100
    },
    {
      "epoch": 3.4180602006688963,
      "grad_norm": 0.7219478487968445,
      "learning_rate": 6.581939799331104e-05,
      "loss": 0.057,
      "step": 5110
    },
    {
      "epoch": 3.4247491638795986,
      "grad_norm": 0.7636262774467468,
      "learning_rate": 6.5752508361204e-05,
      "loss": 0.0707,
      "step": 5120
    },
    {
      "epoch": 3.431438127090301,
      "grad_norm": 1.4830290079116821,
      "learning_rate": 6.5685618729097e-05,
      "loss": 0.0799,
      "step": 5130
    },
    {
      "epoch": 3.4381270903010033,
      "grad_norm": 3.003439426422119,
      "learning_rate": 6.561872909698997e-05,
      "loss": 0.086,
      "step": 5140
    },
    {
      "epoch": 3.4448160535117056,
      "grad_norm": 0.34735190868377686,
      "learning_rate": 6.555183946488295e-05,
      "loss": 0.0755,
      "step": 5150
    },
    {
      "epoch": 3.451505016722408,
      "grad_norm": 0.4009545147418976,
      "learning_rate": 6.548494983277592e-05,
      "loss": 0.0835,
      "step": 5160
    },
    {
      "epoch": 3.4581939799331103,
      "grad_norm": 0.18325521051883698,
      "learning_rate": 6.54180602006689e-05,
      "loss": 0.0741,
      "step": 5170
    },
    {
      "epoch": 3.4648829431438126,
      "grad_norm": 0.6062169671058655,
      "learning_rate": 6.535117056856188e-05,
      "loss": 0.0679,
      "step": 5180
    },
    {
      "epoch": 3.471571906354515,
      "grad_norm": 0.6367049217224121,
      "learning_rate": 6.528428093645485e-05,
      "loss": 0.0865,
      "step": 5190
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 3.2684335708618164,
      "learning_rate": 6.521739130434783e-05,
      "loss": 0.0671,
      "step": 5200
    },
    {
      "epoch": 3.4849498327759196,
      "grad_norm": 1.9086042642593384,
      "learning_rate": 6.51505016722408e-05,
      "loss": 0.0683,
      "step": 5210
    },
    {
      "epoch": 3.491638795986622,
      "grad_norm": 3.634255886077881,
      "learning_rate": 6.508361204013378e-05,
      "loss": 0.0746,
      "step": 5220
    },
    {
      "epoch": 3.4983277591973243,
      "grad_norm": 0.94216388463974,
      "learning_rate": 6.501672240802677e-05,
      "loss": 0.0774,
      "step": 5230
    },
    {
      "epoch": 3.5050167224080266,
      "grad_norm": 2.3354969024658203,
      "learning_rate": 6.494983277591973e-05,
      "loss": 0.069,
      "step": 5240
    },
    {
      "epoch": 3.511705685618729,
      "grad_norm": 1.3407578468322754,
      "learning_rate": 6.488294314381272e-05,
      "loss": 0.0785,
      "step": 5250
    },
    {
      "epoch": 3.5183946488294313,
      "grad_norm": 1.516933560371399,
      "learning_rate": 6.481605351170568e-05,
      "loss": 0.0837,
      "step": 5260
    },
    {
      "epoch": 3.5250836120401337,
      "grad_norm": 4.001212120056152,
      "learning_rate": 6.474916387959867e-05,
      "loss": 0.0774,
      "step": 5270
    },
    {
      "epoch": 3.531772575250836,
      "grad_norm": 6.833039283752441,
      "learning_rate": 6.468227424749163e-05,
      "loss": 0.0778,
      "step": 5280
    },
    {
      "epoch": 3.5384615384615383,
      "grad_norm": 0.9666281342506409,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0628,
      "step": 5290
    },
    {
      "epoch": 3.5451505016722407,
      "grad_norm": 6.497124195098877,
      "learning_rate": 6.45484949832776e-05,
      "loss": 0.077,
      "step": 5300
    },
    {
      "epoch": 3.551839464882943,
      "grad_norm": 1.4599709510803223,
      "learning_rate": 6.448160535117057e-05,
      "loss": 0.0714,
      "step": 5310
    },
    {
      "epoch": 3.5585284280936453,
      "grad_norm": 3.878899097442627,
      "learning_rate": 6.441471571906355e-05,
      "loss": 0.0643,
      "step": 5320
    },
    {
      "epoch": 3.5652173913043477,
      "grad_norm": 0.6432303786277771,
      "learning_rate": 6.434782608695652e-05,
      "loss": 0.0701,
      "step": 5330
    },
    {
      "epoch": 3.57190635451505,
      "grad_norm": 1.6723068952560425,
      "learning_rate": 6.428093645484951e-05,
      "loss": 0.0794,
      "step": 5340
    },
    {
      "epoch": 3.5785953177257523,
      "grad_norm": 6.770298957824707,
      "learning_rate": 6.421404682274248e-05,
      "loss": 0.0909,
      "step": 5350
    },
    {
      "epoch": 3.585284280936455,
      "grad_norm": 1.5146554708480835,
      "learning_rate": 6.414715719063545e-05,
      "loss": 0.0828,
      "step": 5360
    },
    {
      "epoch": 3.591973244147157,
      "grad_norm": 0.2019585371017456,
      "learning_rate": 6.408026755852843e-05,
      "loss": 0.0681,
      "step": 5370
    },
    {
      "epoch": 3.59866220735786,
      "grad_norm": 0.3240724503993988,
      "learning_rate": 6.40133779264214e-05,
      "loss": 0.0846,
      "step": 5380
    },
    {
      "epoch": 3.6053511705685617,
      "grad_norm": 0.20901960134506226,
      "learning_rate": 6.39464882943144e-05,
      "loss": 0.0724,
      "step": 5390
    },
    {
      "epoch": 3.6120401337792645,
      "grad_norm": 2.741830587387085,
      "learning_rate": 6.387959866220736e-05,
      "loss": 0.0717,
      "step": 5400
    },
    {
      "epoch": 3.6187290969899664,
      "grad_norm": 3.1877925395965576,
      "learning_rate": 6.381270903010034e-05,
      "loss": 0.0693,
      "step": 5410
    },
    {
      "epoch": 3.625418060200669,
      "grad_norm": 1.9705578088760376,
      "learning_rate": 6.374581939799331e-05,
      "loss": 0.0648,
      "step": 5420
    },
    {
      "epoch": 3.632107023411371,
      "grad_norm": 0.9284788370132446,
      "learning_rate": 6.367892976588628e-05,
      "loss": 0.064,
      "step": 5430
    },
    {
      "epoch": 3.638795986622074,
      "grad_norm": 0.4431762993335724,
      "learning_rate": 6.361204013377928e-05,
      "loss": 0.0783,
      "step": 5440
    },
    {
      "epoch": 3.6454849498327757,
      "grad_norm": 2.4773430824279785,
      "learning_rate": 6.354515050167224e-05,
      "loss": 0.0885,
      "step": 5450
    },
    {
      "epoch": 3.6521739130434785,
      "grad_norm": 1.8923101425170898,
      "learning_rate": 6.347826086956523e-05,
      "loss": 0.082,
      "step": 5460
    },
    {
      "epoch": 3.6588628762541804,
      "grad_norm": 4.674037933349609,
      "learning_rate": 6.34113712374582e-05,
      "loss": 0.0631,
      "step": 5470
    },
    {
      "epoch": 3.665551839464883,
      "grad_norm": 0.2656029462814331,
      "learning_rate": 6.334448160535118e-05,
      "loss": 0.0736,
      "step": 5480
    },
    {
      "epoch": 3.672240802675585,
      "grad_norm": 4.043028354644775,
      "learning_rate": 6.327759197324416e-05,
      "loss": 0.0697,
      "step": 5490
    },
    {
      "epoch": 3.678929765886288,
      "grad_norm": 0.7581077218055725,
      "learning_rate": 6.321070234113713e-05,
      "loss": 0.0742,
      "step": 5500
    },
    {
      "epoch": 3.6856187290969897,
      "grad_norm": 0.6322870254516602,
      "learning_rate": 6.314381270903011e-05,
      "loss": 0.0704,
      "step": 5510
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 3.356288433074951,
      "learning_rate": 6.307692307692308e-05,
      "loss": 0.0842,
      "step": 5520
    },
    {
      "epoch": 3.6989966555183944,
      "grad_norm": 1.0668654441833496,
      "learning_rate": 6.301003344481606e-05,
      "loss": 0.0595,
      "step": 5530
    },
    {
      "epoch": 3.705685618729097,
      "grad_norm": 1.2336496114730835,
      "learning_rate": 6.294314381270904e-05,
      "loss": 0.0731,
      "step": 5540
    },
    {
      "epoch": 3.712374581939799,
      "grad_norm": 2.575582504272461,
      "learning_rate": 6.287625418060201e-05,
      "loss": 0.0688,
      "step": 5550
    },
    {
      "epoch": 3.719063545150502,
      "grad_norm": 2.5532429218292236,
      "learning_rate": 6.280936454849499e-05,
      "loss": 0.0786,
      "step": 5560
    },
    {
      "epoch": 3.7257525083612038,
      "grad_norm": 0.3744869828224182,
      "learning_rate": 6.274247491638796e-05,
      "loss": 0.081,
      "step": 5570
    },
    {
      "epoch": 3.7324414715719065,
      "grad_norm": 1.8331739902496338,
      "learning_rate": 6.267558528428094e-05,
      "loss": 0.0729,
      "step": 5580
    },
    {
      "epoch": 3.7391304347826084,
      "grad_norm": 3.997131586074829,
      "learning_rate": 6.260869565217392e-05,
      "loss": 0.0744,
      "step": 5590
    },
    {
      "epoch": 3.745819397993311,
      "grad_norm": 0.29028135538101196,
      "learning_rate": 6.254180602006689e-05,
      "loss": 0.0736,
      "step": 5600
    },
    {
      "epoch": 3.7525083612040135,
      "grad_norm": 1.71115243434906,
      "learning_rate": 6.247491638795987e-05,
      "loss": 0.0649,
      "step": 5610
    },
    {
      "epoch": 3.759197324414716,
      "grad_norm": 0.6625193953514099,
      "learning_rate": 6.240802675585284e-05,
      "loss": 0.0844,
      "step": 5620
    },
    {
      "epoch": 3.765886287625418,
      "grad_norm": 1.3801952600479126,
      "learning_rate": 6.234113712374582e-05,
      "loss": 0.0715,
      "step": 5630
    },
    {
      "epoch": 3.7725752508361206,
      "grad_norm": 0.7554713487625122,
      "learning_rate": 6.22742474916388e-05,
      "loss": 0.0667,
      "step": 5640
    },
    {
      "epoch": 3.779264214046823,
      "grad_norm": 1.6294810771942139,
      "learning_rate": 6.220735785953178e-05,
      "loss": 0.0836,
      "step": 5650
    },
    {
      "epoch": 3.7859531772575252,
      "grad_norm": 1.9828605651855469,
      "learning_rate": 6.214046822742475e-05,
      "loss": 0.096,
      "step": 5660
    },
    {
      "epoch": 3.7926421404682276,
      "grad_norm": 3.5717391967773438,
      "learning_rate": 6.207357859531772e-05,
      "loss": 0.0784,
      "step": 5670
    },
    {
      "epoch": 3.79933110367893,
      "grad_norm": 1.6683013439178467,
      "learning_rate": 6.20066889632107e-05,
      "loss": 0.0767,
      "step": 5680
    },
    {
      "epoch": 3.8060200668896322,
      "grad_norm": 0.3348728120326996,
      "learning_rate": 6.193979933110368e-05,
      "loss": 0.0639,
      "step": 5690
    },
    {
      "epoch": 3.8127090301003346,
      "grad_norm": 0.779590904712677,
      "learning_rate": 6.187290969899667e-05,
      "loss": 0.075,
      "step": 5700
    },
    {
      "epoch": 3.819397993311037,
      "grad_norm": 0.2324654906988144,
      "learning_rate": 6.180602006688964e-05,
      "loss": 0.0701,
      "step": 5710
    },
    {
      "epoch": 3.8260869565217392,
      "grad_norm": 0.3135102689266205,
      "learning_rate": 6.173913043478262e-05,
      "loss": 0.0659,
      "step": 5720
    },
    {
      "epoch": 3.8327759197324416,
      "grad_norm": 1.7761188745498657,
      "learning_rate": 6.167224080267559e-05,
      "loss": 0.0852,
      "step": 5730
    },
    {
      "epoch": 3.839464882943144,
      "grad_norm": 6.760908603668213,
      "learning_rate": 6.160535117056855e-05,
      "loss": 0.0892,
      "step": 5740
    },
    {
      "epoch": 3.8461538461538463,
      "grad_norm": 0.9731970429420471,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.079,
      "step": 5750
    },
    {
      "epoch": 3.8528428093645486,
      "grad_norm": 2.6022255420684814,
      "learning_rate": 6.147157190635452e-05,
      "loss": 0.0601,
      "step": 5760
    },
    {
      "epoch": 3.859531772575251,
      "grad_norm": 4.185039520263672,
      "learning_rate": 6.14046822742475e-05,
      "loss": 0.0726,
      "step": 5770
    },
    {
      "epoch": 3.8662207357859533,
      "grad_norm": 0.7964357137680054,
      "learning_rate": 6.133779264214047e-05,
      "loss": 0.0745,
      "step": 5780
    },
    {
      "epoch": 3.8729096989966556,
      "grad_norm": 0.5310606956481934,
      "learning_rate": 6.127090301003345e-05,
      "loss": 0.0725,
      "step": 5790
    },
    {
      "epoch": 3.879598662207358,
      "grad_norm": 1.5212914943695068,
      "learning_rate": 6.120401337792643e-05,
      "loss": 0.0598,
      "step": 5800
    },
    {
      "epoch": 3.8862876254180603,
      "grad_norm": 3.6158175468444824,
      "learning_rate": 6.11371237458194e-05,
      "loss": 0.0717,
      "step": 5810
    },
    {
      "epoch": 3.8929765886287626,
      "grad_norm": 0.4467984139919281,
      "learning_rate": 6.107023411371238e-05,
      "loss": 0.0625,
      "step": 5820
    },
    {
      "epoch": 3.899665551839465,
      "grad_norm": 3.0646955966949463,
      "learning_rate": 6.100334448160535e-05,
      "loss": 0.0607,
      "step": 5830
    },
    {
      "epoch": 3.9063545150501673,
      "grad_norm": 2.663433313369751,
      "learning_rate": 6.0936454849498324e-05,
      "loss": 0.0595,
      "step": 5840
    },
    {
      "epoch": 3.9130434782608696,
      "grad_norm": 3.1213862895965576,
      "learning_rate": 6.086956521739131e-05,
      "loss": 0.0649,
      "step": 5850
    },
    {
      "epoch": 3.919732441471572,
      "grad_norm": 2.0309112071990967,
      "learning_rate": 6.080267558528429e-05,
      "loss": 0.0699,
      "step": 5860
    },
    {
      "epoch": 3.9264214046822743,
      "grad_norm": 2.9740498065948486,
      "learning_rate": 6.073578595317726e-05,
      "loss": 0.063,
      "step": 5870
    },
    {
      "epoch": 3.9331103678929766,
      "grad_norm": 3.238435745239258,
      "learning_rate": 6.066889632107024e-05,
      "loss": 0.0731,
      "step": 5880
    },
    {
      "epoch": 3.939799331103679,
      "grad_norm": 1.1853530406951904,
      "learning_rate": 6.0602006688963206e-05,
      "loss": 0.0932,
      "step": 5890
    },
    {
      "epoch": 3.9464882943143813,
      "grad_norm": 0.3700926601886749,
      "learning_rate": 6.0535117056856194e-05,
      "loss": 0.0819,
      "step": 5900
    },
    {
      "epoch": 3.9531772575250836,
      "grad_norm": 2.7900564670562744,
      "learning_rate": 6.046822742474917e-05,
      "loss": 0.0664,
      "step": 5910
    },
    {
      "epoch": 3.959866220735786,
      "grad_norm": 0.8620715141296387,
      "learning_rate": 6.0401337792642145e-05,
      "loss": 0.0774,
      "step": 5920
    },
    {
      "epoch": 3.9665551839464883,
      "grad_norm": 1.8979065418243408,
      "learning_rate": 6.033444816053512e-05,
      "loss": 0.0733,
      "step": 5930
    },
    {
      "epoch": 3.9732441471571907,
      "grad_norm": 3.0854146480560303,
      "learning_rate": 6.0267558528428095e-05,
      "loss": 0.0993,
      "step": 5940
    },
    {
      "epoch": 3.979933110367893,
      "grad_norm": 0.19406002759933472,
      "learning_rate": 6.0200668896321076e-05,
      "loss": 0.0686,
      "step": 5950
    },
    {
      "epoch": 3.9866220735785953,
      "grad_norm": 2.1533548831939697,
      "learning_rate": 6.013377926421405e-05,
      "loss": 0.0719,
      "step": 5960
    },
    {
      "epoch": 3.9933110367892977,
      "grad_norm": 1.405422329902649,
      "learning_rate": 6.0066889632107026e-05,
      "loss": 0.0624,
      "step": 5970
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.511916160583496,
      "learning_rate": 6e-05,
      "loss": 0.0779,
      "step": 5980
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.07555365562438965,
      "eval_mse": 0.07555365562438965,
      "eval_runtime": 220.2714,
      "eval_samples_per_second": 13.574,
      "eval_steps_per_second": 1.698,
      "step": 5980
    },
    {
      "epoch": 4.006688963210703,
      "grad_norm": 0.7200626134872437,
      "learning_rate": 5.9933110367892977e-05,
      "loss": 0.0659,
      "step": 5990
    },
    {
      "epoch": 4.013377926421405,
      "grad_norm": 7.074632167816162,
      "learning_rate": 5.986622073578596e-05,
      "loss": 0.0949,
      "step": 6000
    },
    {
      "epoch": 4.0200668896321075,
      "grad_norm": 1.3575361967086792,
      "learning_rate": 5.9799331103678933e-05,
      "loss": 0.0656,
      "step": 6010
    },
    {
      "epoch": 4.026755852842809,
      "grad_norm": 3.287247657775879,
      "learning_rate": 5.973244147157191e-05,
      "loss": 0.0626,
      "step": 6020
    },
    {
      "epoch": 4.033444816053512,
      "grad_norm": 3.9007692337036133,
      "learning_rate": 5.9665551839464883e-05,
      "loss": 0.0531,
      "step": 6030
    },
    {
      "epoch": 4.040133779264214,
      "grad_norm": 0.33864256739616394,
      "learning_rate": 5.959866220735786e-05,
      "loss": 0.0533,
      "step": 6040
    },
    {
      "epoch": 4.046822742474917,
      "grad_norm": 1.0063756704330444,
      "learning_rate": 5.953177257525085e-05,
      "loss": 0.076,
      "step": 6050
    },
    {
      "epoch": 4.053511705685619,
      "grad_norm": 1.4410839080810547,
      "learning_rate": 5.9464882943143815e-05,
      "loss": 0.0621,
      "step": 6060
    },
    {
      "epoch": 4.0602006688963215,
      "grad_norm": 2.621447801589966,
      "learning_rate": 5.939799331103679e-05,
      "loss": 0.0666,
      "step": 6070
    },
    {
      "epoch": 4.066889632107023,
      "grad_norm": 2.981430768966675,
      "learning_rate": 5.9331103678929765e-05,
      "loss": 0.0608,
      "step": 6080
    },
    {
      "epoch": 4.073578595317726,
      "grad_norm": 1.6998625993728638,
      "learning_rate": 5.926421404682274e-05,
      "loss": 0.0733,
      "step": 6090
    },
    {
      "epoch": 4.080267558528428,
      "grad_norm": 0.7376275062561035,
      "learning_rate": 5.919732441471573e-05,
      "loss": 0.0707,
      "step": 6100
    },
    {
      "epoch": 4.086956521739131,
      "grad_norm": 0.9681304693222046,
      "learning_rate": 5.9130434782608704e-05,
      "loss": 0.0633,
      "step": 6110
    },
    {
      "epoch": 4.093645484949833,
      "grad_norm": 1.3118129968643188,
      "learning_rate": 5.906354515050168e-05,
      "loss": 0.0586,
      "step": 6120
    },
    {
      "epoch": 4.1003344481605355,
      "grad_norm": 4.774895191192627,
      "learning_rate": 5.899665551839465e-05,
      "loss": 0.067,
      "step": 6130
    },
    {
      "epoch": 4.107023411371237,
      "grad_norm": 0.9589107632637024,
      "learning_rate": 5.892976588628762e-05,
      "loss": 0.0631,
      "step": 6140
    },
    {
      "epoch": 4.11371237458194,
      "grad_norm": 2.461914300918579,
      "learning_rate": 5.886287625418061e-05,
      "loss": 0.0714,
      "step": 6150
    },
    {
      "epoch": 4.120401337792642,
      "grad_norm": 2.1283793449401855,
      "learning_rate": 5.8795986622073586e-05,
      "loss": 0.0841,
      "step": 6160
    },
    {
      "epoch": 4.127090301003345,
      "grad_norm": 0.8651756048202515,
      "learning_rate": 5.872909698996656e-05,
      "loss": 0.0668,
      "step": 6170
    },
    {
      "epoch": 4.133779264214047,
      "grad_norm": 2.440626859664917,
      "learning_rate": 5.8662207357859536e-05,
      "loss": 0.0763,
      "step": 6180
    },
    {
      "epoch": 4.1404682274247495,
      "grad_norm": 3.1031172275543213,
      "learning_rate": 5.859531772575251e-05,
      "loss": 0.0702,
      "step": 6190
    },
    {
      "epoch": 4.147157190635451,
      "grad_norm": 3.748271942138672,
      "learning_rate": 5.852842809364549e-05,
      "loss": 0.0716,
      "step": 6200
    },
    {
      "epoch": 4.153846153846154,
      "grad_norm": 1.2731395959854126,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0847,
      "step": 6210
    },
    {
      "epoch": 4.160535117056856,
      "grad_norm": 0.48583781719207764,
      "learning_rate": 5.839464882943144e-05,
      "loss": 0.0604,
      "step": 6220
    },
    {
      "epoch": 4.167224080267559,
      "grad_norm": 0.9449681639671326,
      "learning_rate": 5.832775919732442e-05,
      "loss": 0.075,
      "step": 6230
    },
    {
      "epoch": 4.173913043478261,
      "grad_norm": 1.5034922361373901,
      "learning_rate": 5.826086956521739e-05,
      "loss": 0.0795,
      "step": 6240
    },
    {
      "epoch": 4.1806020066889635,
      "grad_norm": 6.206334114074707,
      "learning_rate": 5.819397993311037e-05,
      "loss": 0.0714,
      "step": 6250
    },
    {
      "epoch": 4.187290969899665,
      "grad_norm": 1.9744735956192017,
      "learning_rate": 5.812709030100335e-05,
      "loss": 0.0671,
      "step": 6260
    },
    {
      "epoch": 4.193979933110368,
      "grad_norm": 1.4443360567092896,
      "learning_rate": 5.8060200668896325e-05,
      "loss": 0.0628,
      "step": 6270
    },
    {
      "epoch": 4.20066889632107,
      "grad_norm": 3.5589351654052734,
      "learning_rate": 5.79933110367893e-05,
      "loss": 0.0724,
      "step": 6280
    },
    {
      "epoch": 4.207357859531773,
      "grad_norm": 1.0574309825897217,
      "learning_rate": 5.7926421404682275e-05,
      "loss": 0.0595,
      "step": 6290
    },
    {
      "epoch": 4.214046822742475,
      "grad_norm": 0.8373066186904907,
      "learning_rate": 5.785953177257525e-05,
      "loss": 0.0755,
      "step": 6300
    },
    {
      "epoch": 4.2207357859531776,
      "grad_norm": 0.5989928841590881,
      "learning_rate": 5.779264214046823e-05,
      "loss": 0.0708,
      "step": 6310
    },
    {
      "epoch": 4.2274247491638794,
      "grad_norm": 3.6913678646087646,
      "learning_rate": 5.772575250836121e-05,
      "loss": 0.0731,
      "step": 6320
    },
    {
      "epoch": 4.234113712374582,
      "grad_norm": 6.884210586547852,
      "learning_rate": 5.765886287625418e-05,
      "loss": 0.0782,
      "step": 6330
    },
    {
      "epoch": 4.240802675585284,
      "grad_norm": 2.3835830688476562,
      "learning_rate": 5.759197324414716e-05,
      "loss": 0.0708,
      "step": 6340
    },
    {
      "epoch": 4.247491638795987,
      "grad_norm": 1.463895559310913,
      "learning_rate": 5.752508361204013e-05,
      "loss": 0.0583,
      "step": 6350
    },
    {
      "epoch": 4.254180602006689,
      "grad_norm": 3.8699686527252197,
      "learning_rate": 5.7458193979933114e-05,
      "loss": 0.0588,
      "step": 6360
    },
    {
      "epoch": 4.260869565217392,
      "grad_norm": 0.9512161016464233,
      "learning_rate": 5.739130434782609e-05,
      "loss": 0.0647,
      "step": 6370
    },
    {
      "epoch": 4.2675585284280935,
      "grad_norm": 0.32455405592918396,
      "learning_rate": 5.7324414715719064e-05,
      "loss": 0.063,
      "step": 6380
    },
    {
      "epoch": 4.274247491638796,
      "grad_norm": 0.3892381191253662,
      "learning_rate": 5.725752508361204e-05,
      "loss": 0.0945,
      "step": 6390
    },
    {
      "epoch": 4.280936454849498,
      "grad_norm": 0.3295302093029022,
      "learning_rate": 5.7190635451505014e-05,
      "loss": 0.0554,
      "step": 6400
    },
    {
      "epoch": 4.287625418060201,
      "grad_norm": 0.6870085000991821,
      "learning_rate": 5.7123745819398e-05,
      "loss": 0.0813,
      "step": 6410
    },
    {
      "epoch": 4.294314381270903,
      "grad_norm": 0.5889727473258972,
      "learning_rate": 5.705685618729098e-05,
      "loss": 0.0788,
      "step": 6420
    },
    {
      "epoch": 4.301003344481606,
      "grad_norm": 2.1253087520599365,
      "learning_rate": 5.6989966555183946e-05,
      "loss": 0.071,
      "step": 6430
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 2.5222198963165283,
      "learning_rate": 5.692307692307692e-05,
      "loss": 0.0553,
      "step": 6440
    },
    {
      "epoch": 4.31438127090301,
      "grad_norm": 3.0310568809509277,
      "learning_rate": 5.6856187290969896e-05,
      "loss": 0.0863,
      "step": 6450
    },
    {
      "epoch": 4.321070234113712,
      "grad_norm": 5.671077251434326,
      "learning_rate": 5.6789297658862884e-05,
      "loss": 0.0625,
      "step": 6460
    },
    {
      "epoch": 4.327759197324415,
      "grad_norm": 2.766726493835449,
      "learning_rate": 5.672240802675586e-05,
      "loss": 0.0711,
      "step": 6470
    },
    {
      "epoch": 4.334448160535117,
      "grad_norm": 0.4096842110157013,
      "learning_rate": 5.6655518394648834e-05,
      "loss": 0.0671,
      "step": 6480
    },
    {
      "epoch": 4.34113712374582,
      "grad_norm": 8.858095169067383,
      "learning_rate": 5.658862876254181e-05,
      "loss": 0.0789,
      "step": 6490
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 0.9603244066238403,
      "learning_rate": 5.652173913043478e-05,
      "loss": 0.0638,
      "step": 6500
    },
    {
      "epoch": 4.354515050167224,
      "grad_norm": 0.19360791146755219,
      "learning_rate": 5.6454849498327766e-05,
      "loss": 0.0669,
      "step": 6510
    },
    {
      "epoch": 4.361204013377926,
      "grad_norm": 0.7423932552337646,
      "learning_rate": 5.638795986622074e-05,
      "loss": 0.0712,
      "step": 6520
    },
    {
      "epoch": 4.367892976588629,
      "grad_norm": 0.4607205092906952,
      "learning_rate": 5.6321070234113716e-05,
      "loss": 0.0818,
      "step": 6530
    },
    {
      "epoch": 4.374581939799331,
      "grad_norm": 0.6850534677505493,
      "learning_rate": 5.625418060200669e-05,
      "loss": 0.0709,
      "step": 6540
    },
    {
      "epoch": 4.381270903010034,
      "grad_norm": 1.2787178754806519,
      "learning_rate": 5.6187290969899666e-05,
      "loss": 0.09,
      "step": 6550
    },
    {
      "epoch": 4.3879598662207355,
      "grad_norm": 0.7124513387680054,
      "learning_rate": 5.612040133779265e-05,
      "loss": 0.0683,
      "step": 6560
    },
    {
      "epoch": 4.394648829431438,
      "grad_norm": 1.9447962045669556,
      "learning_rate": 5.605351170568562e-05,
      "loss": 0.0674,
      "step": 6570
    },
    {
      "epoch": 4.40133779264214,
      "grad_norm": 3.2783830165863037,
      "learning_rate": 5.59866220735786e-05,
      "loss": 0.0754,
      "step": 6580
    },
    {
      "epoch": 4.408026755852843,
      "grad_norm": 1.072465419769287,
      "learning_rate": 5.591973244147157e-05,
      "loss": 0.0757,
      "step": 6590
    },
    {
      "epoch": 4.414715719063545,
      "grad_norm": 4.134634971618652,
      "learning_rate": 5.585284280936455e-05,
      "loss": 0.0573,
      "step": 6600
    },
    {
      "epoch": 4.421404682274248,
      "grad_norm": 6.002267837524414,
      "learning_rate": 5.578595317725753e-05,
      "loss": 0.074,
      "step": 6610
    },
    {
      "epoch": 4.4280936454849495,
      "grad_norm": 1.9408398866653442,
      "learning_rate": 5.5719063545150505e-05,
      "loss": 0.0671,
      "step": 6620
    },
    {
      "epoch": 4.434782608695652,
      "grad_norm": 1.891671895980835,
      "learning_rate": 5.565217391304348e-05,
      "loss": 0.0655,
      "step": 6630
    },
    {
      "epoch": 4.441471571906354,
      "grad_norm": 5.870092391967773,
      "learning_rate": 5.5585284280936455e-05,
      "loss": 0.069,
      "step": 6640
    },
    {
      "epoch": 4.448160535117057,
      "grad_norm": 5.0089921951293945,
      "learning_rate": 5.551839464882943e-05,
      "loss": 0.0658,
      "step": 6650
    },
    {
      "epoch": 4.454849498327759,
      "grad_norm": 1.1499357223510742,
      "learning_rate": 5.545150501672242e-05,
      "loss": 0.0855,
      "step": 6660
    },
    {
      "epoch": 4.461538461538462,
      "grad_norm": 2.1092236042022705,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0826,
      "step": 6670
    },
    {
      "epoch": 4.468227424749164,
      "grad_norm": 2.6977195739746094,
      "learning_rate": 5.531772575250836e-05,
      "loss": 0.0622,
      "step": 6680
    },
    {
      "epoch": 4.474916387959866,
      "grad_norm": 2.9836738109588623,
      "learning_rate": 5.525083612040134e-05,
      "loss": 0.0659,
      "step": 6690
    },
    {
      "epoch": 4.481605351170568,
      "grad_norm": 0.8441120982170105,
      "learning_rate": 5.518394648829431e-05,
      "loss": 0.0663,
      "step": 6700
    },
    {
      "epoch": 4.488294314381271,
      "grad_norm": 1.7105242013931274,
      "learning_rate": 5.51170568561873e-05,
      "loss": 0.0787,
      "step": 6710
    },
    {
      "epoch": 4.494983277591973,
      "grad_norm": 3.3104045391082764,
      "learning_rate": 5.5050167224080276e-05,
      "loss": 0.072,
      "step": 6720
    },
    {
      "epoch": 4.501672240802676,
      "grad_norm": 0.2517462372779846,
      "learning_rate": 5.498327759197325e-05,
      "loss": 0.0592,
      "step": 6730
    },
    {
      "epoch": 4.508361204013378,
      "grad_norm": 3.130553722381592,
      "learning_rate": 5.491638795986622e-05,
      "loss": 0.0597,
      "step": 6740
    },
    {
      "epoch": 4.51505016722408,
      "grad_norm": 2.540755033493042,
      "learning_rate": 5.4849498327759194e-05,
      "loss": 0.0527,
      "step": 6750
    },
    {
      "epoch": 4.521739130434782,
      "grad_norm": 4.94587516784668,
      "learning_rate": 5.478260869565217e-05,
      "loss": 0.0755,
      "step": 6760
    },
    {
      "epoch": 4.528428093645485,
      "grad_norm": 5.80070161819458,
      "learning_rate": 5.471571906354516e-05,
      "loss": 0.0653,
      "step": 6770
    },
    {
      "epoch": 4.535117056856187,
      "grad_norm": 1.3168888092041016,
      "learning_rate": 5.464882943143813e-05,
      "loss": 0.074,
      "step": 6780
    },
    {
      "epoch": 4.54180602006689,
      "grad_norm": 1.5611118078231812,
      "learning_rate": 5.458193979933111e-05,
      "loss": 0.0709,
      "step": 6790
    },
    {
      "epoch": 4.548494983277592,
      "grad_norm": 2.30637264251709,
      "learning_rate": 5.451505016722408e-05,
      "loss": 0.0769,
      "step": 6800
    },
    {
      "epoch": 4.555183946488294,
      "grad_norm": 2.6152946949005127,
      "learning_rate": 5.444816053511705e-05,
      "loss": 0.079,
      "step": 6810
    },
    {
      "epoch": 4.561872909698996,
      "grad_norm": 0.3139130473136902,
      "learning_rate": 5.438127090301004e-05,
      "loss": 0.055,
      "step": 6820
    },
    {
      "epoch": 4.568561872909699,
      "grad_norm": 0.623741626739502,
      "learning_rate": 5.4314381270903015e-05,
      "loss": 0.0619,
      "step": 6830
    },
    {
      "epoch": 4.575250836120401,
      "grad_norm": 1.389875054359436,
      "learning_rate": 5.424749163879599e-05,
      "loss": 0.0686,
      "step": 6840
    },
    {
      "epoch": 4.581939799331104,
      "grad_norm": 0.4257790744304657,
      "learning_rate": 5.4180602006688965e-05,
      "loss": 0.0872,
      "step": 6850
    },
    {
      "epoch": 4.588628762541806,
      "grad_norm": 0.7867361307144165,
      "learning_rate": 5.411371237458194e-05,
      "loss": 0.0746,
      "step": 6860
    },
    {
      "epoch": 4.595317725752508,
      "grad_norm": 1.4453585147857666,
      "learning_rate": 5.404682274247492e-05,
      "loss": 0.0605,
      "step": 6870
    },
    {
      "epoch": 4.602006688963211,
      "grad_norm": 1.342901349067688,
      "learning_rate": 5.3979933110367897e-05,
      "loss": 0.0675,
      "step": 6880
    },
    {
      "epoch": 4.608695652173913,
      "grad_norm": 2.2974514961242676,
      "learning_rate": 5.391304347826087e-05,
      "loss": 0.0748,
      "step": 6890
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 6.246169567108154,
      "learning_rate": 5.384615384615385e-05,
      "loss": 0.0907,
      "step": 6900
    },
    {
      "epoch": 4.622073578595318,
      "grad_norm": 0.721906304359436,
      "learning_rate": 5.377926421404682e-05,
      "loss": 0.0707,
      "step": 6910
    },
    {
      "epoch": 4.6287625418060205,
      "grad_norm": 0.5669277310371399,
      "learning_rate": 5.3712374581939803e-05,
      "loss": 0.0728,
      "step": 6920
    },
    {
      "epoch": 4.635451505016722,
      "grad_norm": 6.285121917724609,
      "learning_rate": 5.364548494983278e-05,
      "loss": 0.0923,
      "step": 6930
    },
    {
      "epoch": 4.642140468227424,
      "grad_norm": 1.2872087955474854,
      "learning_rate": 5.3578595317725754e-05,
      "loss": 0.0735,
      "step": 6940
    },
    {
      "epoch": 4.648829431438127,
      "grad_norm": 3.112480878829956,
      "learning_rate": 5.351170568561873e-05,
      "loss": 0.0638,
      "step": 6950
    },
    {
      "epoch": 4.65551839464883,
      "grad_norm": 3.5378661155700684,
      "learning_rate": 5.3444816053511704e-05,
      "loss": 0.0825,
      "step": 6960
    },
    {
      "epoch": 4.662207357859532,
      "grad_norm": 0.5296181440353394,
      "learning_rate": 5.337792642140469e-05,
      "loss": 0.0756,
      "step": 6970
    },
    {
      "epoch": 4.668896321070234,
      "grad_norm": 5.031908988952637,
      "learning_rate": 5.331103678929766e-05,
      "loss": 0.077,
      "step": 6980
    },
    {
      "epoch": 4.6755852842809364,
      "grad_norm": 0.9596044421195984,
      "learning_rate": 5.3244147157190635e-05,
      "loss": 0.0684,
      "step": 6990
    },
    {
      "epoch": 4.682274247491639,
      "grad_norm": 1.9678120613098145,
      "learning_rate": 5.317725752508361e-05,
      "loss": 0.0816,
      "step": 7000
    },
    {
      "epoch": 4.688963210702341,
      "grad_norm": 2.694352149963379,
      "learning_rate": 5.3110367892976586e-05,
      "loss": 0.0808,
      "step": 7010
    },
    {
      "epoch": 4.695652173913043,
      "grad_norm": 2.666954517364502,
      "learning_rate": 5.3043478260869574e-05,
      "loss": 0.0788,
      "step": 7020
    },
    {
      "epoch": 4.702341137123746,
      "grad_norm": 0.9346604347229004,
      "learning_rate": 5.297658862876255e-05,
      "loss": 0.0742,
      "step": 7030
    },
    {
      "epoch": 4.709030100334449,
      "grad_norm": 1.9348862171173096,
      "learning_rate": 5.2909698996655524e-05,
      "loss": 0.0748,
      "step": 7040
    },
    {
      "epoch": 4.7157190635451505,
      "grad_norm": 0.41695094108581543,
      "learning_rate": 5.284280936454849e-05,
      "loss": 0.0708,
      "step": 7050
    },
    {
      "epoch": 4.722408026755852,
      "grad_norm": 1.0469141006469727,
      "learning_rate": 5.277591973244147e-05,
      "loss": 0.0875,
      "step": 7060
    },
    {
      "epoch": 4.729096989966555,
      "grad_norm": 1.3788455724716187,
      "learning_rate": 5.2709030100334456e-05,
      "loss": 0.0542,
      "step": 7070
    },
    {
      "epoch": 4.735785953177258,
      "grad_norm": 1.5151710510253906,
      "learning_rate": 5.264214046822743e-05,
      "loss": 0.0576,
      "step": 7080
    },
    {
      "epoch": 4.74247491638796,
      "grad_norm": 5.557274341583252,
      "learning_rate": 5.2575250836120406e-05,
      "loss": 0.0848,
      "step": 7090
    },
    {
      "epoch": 4.749163879598662,
      "grad_norm": 3.972078561782837,
      "learning_rate": 5.250836120401338e-05,
      "loss": 0.0757,
      "step": 7100
    },
    {
      "epoch": 4.7558528428093645,
      "grad_norm": 1.610681176185608,
      "learning_rate": 5.2441471571906356e-05,
      "loss": 0.0678,
      "step": 7110
    },
    {
      "epoch": 4.762541806020067,
      "grad_norm": 3.7423255443573,
      "learning_rate": 5.237458193979934e-05,
      "loss": 0.0767,
      "step": 7120
    },
    {
      "epoch": 4.769230769230769,
      "grad_norm": 1.9917365312576294,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0757,
      "step": 7130
    },
    {
      "epoch": 4.775919732441472,
      "grad_norm": 0.4224523901939392,
      "learning_rate": 5.224080267558529e-05,
      "loss": 0.0613,
      "step": 7140
    },
    {
      "epoch": 4.782608695652174,
      "grad_norm": 2.9698522090911865,
      "learning_rate": 5.217391304347826e-05,
      "loss": 0.0731,
      "step": 7150
    },
    {
      "epoch": 4.789297658862877,
      "grad_norm": 1.6401015520095825,
      "learning_rate": 5.210702341137124e-05,
      "loss": 0.0685,
      "step": 7160
    },
    {
      "epoch": 4.7959866220735785,
      "grad_norm": 1.937436580657959,
      "learning_rate": 5.204013377926422e-05,
      "loss": 0.0662,
      "step": 7170
    },
    {
      "epoch": 4.802675585284281,
      "grad_norm": 2.4356794357299805,
      "learning_rate": 5.1973244147157195e-05,
      "loss": 0.0585,
      "step": 7180
    },
    {
      "epoch": 4.809364548494983,
      "grad_norm": 1.4742059707641602,
      "learning_rate": 5.190635451505017e-05,
      "loss": 0.0684,
      "step": 7190
    },
    {
      "epoch": 4.816053511705686,
      "grad_norm": 4.7270965576171875,
      "learning_rate": 5.1839464882943145e-05,
      "loss": 0.0754,
      "step": 7200
    },
    {
      "epoch": 4.822742474916388,
      "grad_norm": 2.557396173477173,
      "learning_rate": 5.177257525083612e-05,
      "loss": 0.0626,
      "step": 7210
    },
    {
      "epoch": 4.829431438127091,
      "grad_norm": 3.0992376804351807,
      "learning_rate": 5.17056856187291e-05,
      "loss": 0.0718,
      "step": 7220
    },
    {
      "epoch": 4.8361204013377925,
      "grad_norm": 3.5227317810058594,
      "learning_rate": 5.163879598662208e-05,
      "loss": 0.0696,
      "step": 7230
    },
    {
      "epoch": 4.842809364548495,
      "grad_norm": 0.42812949419021606,
      "learning_rate": 5.157190635451505e-05,
      "loss": 0.0622,
      "step": 7240
    },
    {
      "epoch": 4.849498327759197,
      "grad_norm": 0.9504969120025635,
      "learning_rate": 5.150501672240803e-05,
      "loss": 0.0587,
      "step": 7250
    },
    {
      "epoch": 4.8561872909699,
      "grad_norm": 1.9355140924453735,
      "learning_rate": 5.1438127090301e-05,
      "loss": 0.0721,
      "step": 7260
    },
    {
      "epoch": 4.862876254180602,
      "grad_norm": 1.7390964031219482,
      "learning_rate": 5.137123745819398e-05,
      "loss": 0.0733,
      "step": 7270
    },
    {
      "epoch": 4.869565217391305,
      "grad_norm": 0.4878229796886444,
      "learning_rate": 5.1304347826086966e-05,
      "loss": 0.0678,
      "step": 7280
    },
    {
      "epoch": 4.8762541806020065,
      "grad_norm": 0.5825105905532837,
      "learning_rate": 5.1237458193979934e-05,
      "loss": 0.0633,
      "step": 7290
    },
    {
      "epoch": 4.882943143812709,
      "grad_norm": 1.2219446897506714,
      "learning_rate": 5.117056856187291e-05,
      "loss": 0.074,
      "step": 7300
    },
    {
      "epoch": 4.889632107023411,
      "grad_norm": 7.823890209197998,
      "learning_rate": 5.1103678929765884e-05,
      "loss": 0.0663,
      "step": 7310
    },
    {
      "epoch": 4.896321070234114,
      "grad_norm": 1.276883602142334,
      "learning_rate": 5.103678929765886e-05,
      "loss": 0.068,
      "step": 7320
    },
    {
      "epoch": 4.903010033444816,
      "grad_norm": 0.39954355359077454,
      "learning_rate": 5.096989966555185e-05,
      "loss": 0.0672,
      "step": 7330
    },
    {
      "epoch": 4.909698996655519,
      "grad_norm": 1.9998856782913208,
      "learning_rate": 5.090301003344482e-05,
      "loss": 0.0663,
      "step": 7340
    },
    {
      "epoch": 4.916387959866221,
      "grad_norm": 3.3592422008514404,
      "learning_rate": 5.08361204013378e-05,
      "loss": 0.0717,
      "step": 7350
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 2.421261787414551,
      "learning_rate": 5.0769230769230766e-05,
      "loss": 0.0778,
      "step": 7360
    },
    {
      "epoch": 4.929765886287625,
      "grad_norm": 3.719966411590576,
      "learning_rate": 5.070234113712374e-05,
      "loss": 0.067,
      "step": 7370
    },
    {
      "epoch": 4.936454849498328,
      "grad_norm": 0.6203343868255615,
      "learning_rate": 5.063545150501673e-05,
      "loss": 0.0643,
      "step": 7380
    },
    {
      "epoch": 4.94314381270903,
      "grad_norm": 0.692283570766449,
      "learning_rate": 5.0568561872909704e-05,
      "loss": 0.0684,
      "step": 7390
    },
    {
      "epoch": 4.949832775919733,
      "grad_norm": 1.4749492406845093,
      "learning_rate": 5.050167224080268e-05,
      "loss": 0.0746,
      "step": 7400
    },
    {
      "epoch": 4.956521739130435,
      "grad_norm": 2.1798877716064453,
      "learning_rate": 5.0434782608695655e-05,
      "loss": 0.0655,
      "step": 7410
    },
    {
      "epoch": 4.963210702341137,
      "grad_norm": 2.3222920894622803,
      "learning_rate": 5.036789297658863e-05,
      "loss": 0.0658,
      "step": 7420
    },
    {
      "epoch": 4.969899665551839,
      "grad_norm": 3.0843615531921387,
      "learning_rate": 5.030100334448161e-05,
      "loss": 0.0573,
      "step": 7430
    },
    {
      "epoch": 4.976588628762542,
      "grad_norm": 2.0864250659942627,
      "learning_rate": 5.0234113712374586e-05,
      "loss": 0.0607,
      "step": 7440
    },
    {
      "epoch": 4.983277591973244,
      "grad_norm": 2.6795477867126465,
      "learning_rate": 5.016722408026756e-05,
      "loss": 0.0631,
      "step": 7450
    },
    {
      "epoch": 4.989966555183947,
      "grad_norm": 0.6230664849281311,
      "learning_rate": 5.0100334448160536e-05,
      "loss": 0.066,
      "step": 7460
    },
    {
      "epoch": 4.996655518394649,
      "grad_norm": 2.3801419734954834,
      "learning_rate": 5.003344481605351e-05,
      "loss": 0.0841,
      "step": 7470
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.07865732908248901,
      "eval_mse": 0.07865732908248901,
      "eval_runtime": 221.7747,
      "eval_samples_per_second": 13.482,
      "eval_steps_per_second": 1.686,
      "step": 7475
    },
    {
      "epoch": 5.003344481605351,
      "grad_norm": 3.762542486190796,
      "learning_rate": 4.9966555183946487e-05,
      "loss": 0.062,
      "step": 7480
    },
    {
      "epoch": 5.010033444816053,
      "grad_norm": 1.8806613683700562,
      "learning_rate": 4.989966555183947e-05,
      "loss": 0.0603,
      "step": 7490
    },
    {
      "epoch": 5.016722408026756,
      "grad_norm": 5.983439922332764,
      "learning_rate": 4.983277591973244e-05,
      "loss": 0.0627,
      "step": 7500
    },
    {
      "epoch": 5.023411371237458,
      "grad_norm": 1.454197645187378,
      "learning_rate": 4.976588628762542e-05,
      "loss": 0.0464,
      "step": 7510
    },
    {
      "epoch": 5.030100334448161,
      "grad_norm": 3.129765510559082,
      "learning_rate": 4.96989966555184e-05,
      "loss": 0.0504,
      "step": 7520
    },
    {
      "epoch": 5.036789297658863,
      "grad_norm": 1.3404700756072998,
      "learning_rate": 4.9632107023411375e-05,
      "loss": 0.0671,
      "step": 7530
    },
    {
      "epoch": 5.043478260869565,
      "grad_norm": 4.357332706451416,
      "learning_rate": 4.956521739130435e-05,
      "loss": 0.0741,
      "step": 7540
    },
    {
      "epoch": 5.050167224080267,
      "grad_norm": 1.8956897258758545,
      "learning_rate": 4.9498327759197325e-05,
      "loss": 0.0676,
      "step": 7550
    },
    {
      "epoch": 5.05685618729097,
      "grad_norm": 5.321130275726318,
      "learning_rate": 4.94314381270903e-05,
      "loss": 0.0732,
      "step": 7560
    },
    {
      "epoch": 5.063545150501672,
      "grad_norm": 0.6879299283027649,
      "learning_rate": 4.936454849498328e-05,
      "loss": 0.0569,
      "step": 7570
    },
    {
      "epoch": 5.070234113712375,
      "grad_norm": 2.887319326400757,
      "learning_rate": 4.929765886287626e-05,
      "loss": 0.0625,
      "step": 7580
    },
    {
      "epoch": 5.076923076923077,
      "grad_norm": 5.584492206573486,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0664,
      "step": 7590
    },
    {
      "epoch": 5.083612040133779,
      "grad_norm": 7.167271137237549,
      "learning_rate": 4.916387959866221e-05,
      "loss": 0.0637,
      "step": 7600
    },
    {
      "epoch": 5.090301003344481,
      "grad_norm": 1.2168267965316772,
      "learning_rate": 4.909698996655518e-05,
      "loss": 0.0751,
      "step": 7610
    },
    {
      "epoch": 5.096989966555184,
      "grad_norm": 3.569265127182007,
      "learning_rate": 4.9030100334448164e-05,
      "loss": 0.0605,
      "step": 7620
    },
    {
      "epoch": 5.103678929765886,
      "grad_norm": 1.1153556108474731,
      "learning_rate": 4.896321070234114e-05,
      "loss": 0.0662,
      "step": 7630
    },
    {
      "epoch": 5.110367892976589,
      "grad_norm": 0.45676007866859436,
      "learning_rate": 4.8896321070234114e-05,
      "loss": 0.0531,
      "step": 7640
    },
    {
      "epoch": 5.117056856187291,
      "grad_norm": 3.829061269760132,
      "learning_rate": 4.8829431438127096e-05,
      "loss": 0.0575,
      "step": 7650
    },
    {
      "epoch": 5.1237458193979935,
      "grad_norm": 1.676592469215393,
      "learning_rate": 4.876254180602007e-05,
      "loss": 0.06,
      "step": 7660
    },
    {
      "epoch": 5.130434782608695,
      "grad_norm": 0.6828649640083313,
      "learning_rate": 4.8695652173913046e-05,
      "loss": 0.0578,
      "step": 7670
    },
    {
      "epoch": 5.137123745819398,
      "grad_norm": 3.1590874195098877,
      "learning_rate": 4.862876254180602e-05,
      "loss": 0.0775,
      "step": 7680
    },
    {
      "epoch": 5.1438127090301,
      "grad_norm": 0.5831063985824585,
      "learning_rate": 4.8561872909698996e-05,
      "loss": 0.0808,
      "step": 7690
    },
    {
      "epoch": 5.150501672240803,
      "grad_norm": 0.17127808928489685,
      "learning_rate": 4.849498327759198e-05,
      "loss": 0.0548,
      "step": 7700
    },
    {
      "epoch": 5.157190635451505,
      "grad_norm": 3.0591585636138916,
      "learning_rate": 4.842809364548495e-05,
      "loss": 0.0789,
      "step": 7710
    },
    {
      "epoch": 5.1638795986622075,
      "grad_norm": 0.9182135462760925,
      "learning_rate": 4.836120401337793e-05,
      "loss": 0.0649,
      "step": 7720
    },
    {
      "epoch": 5.170568561872909,
      "grad_norm": 0.8472630977630615,
      "learning_rate": 4.82943143812709e-05,
      "loss": 0.0585,
      "step": 7730
    },
    {
      "epoch": 5.177257525083612,
      "grad_norm": 2.960695266723633,
      "learning_rate": 4.822742474916388e-05,
      "loss": 0.0596,
      "step": 7740
    },
    {
      "epoch": 5.183946488294314,
      "grad_norm": 1.086195468902588,
      "learning_rate": 4.816053511705686e-05,
      "loss": 0.0599,
      "step": 7750
    },
    {
      "epoch": 5.190635451505017,
      "grad_norm": 3.7133307456970215,
      "learning_rate": 4.8093645484949835e-05,
      "loss": 0.05,
      "step": 7760
    },
    {
      "epoch": 5.197324414715719,
      "grad_norm": 1.9042307138442993,
      "learning_rate": 4.802675585284282e-05,
      "loss": 0.0653,
      "step": 7770
    },
    {
      "epoch": 5.2040133779264215,
      "grad_norm": 2.424346685409546,
      "learning_rate": 4.795986622073579e-05,
      "loss": 0.0531,
      "step": 7780
    },
    {
      "epoch": 5.210702341137123,
      "grad_norm": 0.8446156978607178,
      "learning_rate": 4.789297658862876e-05,
      "loss": 0.0542,
      "step": 7790
    },
    {
      "epoch": 5.217391304347826,
      "grad_norm": 5.944769859313965,
      "learning_rate": 4.782608695652174e-05,
      "loss": 0.0634,
      "step": 7800
    },
    {
      "epoch": 5.224080267558528,
      "grad_norm": 0.392313688993454,
      "learning_rate": 4.775919732441472e-05,
      "loss": 0.0522,
      "step": 7810
    },
    {
      "epoch": 5.230769230769231,
      "grad_norm": 3.464770793914795,
      "learning_rate": 4.76923076923077e-05,
      "loss": 0.0695,
      "step": 7820
    },
    {
      "epoch": 5.237458193979933,
      "grad_norm": 5.194563865661621,
      "learning_rate": 4.7625418060200674e-05,
      "loss": 0.047,
      "step": 7830
    },
    {
      "epoch": 5.2441471571906355,
      "grad_norm": 3.7293553352355957,
      "learning_rate": 4.755852842809365e-05,
      "loss": 0.0679,
      "step": 7840
    },
    {
      "epoch": 5.250836120401337,
      "grad_norm": 1.6341503858566284,
      "learning_rate": 4.7491638795986624e-05,
      "loss": 0.069,
      "step": 7850
    },
    {
      "epoch": 5.25752508361204,
      "grad_norm": 2.9764022827148438,
      "learning_rate": 4.74247491638796e-05,
      "loss": 0.056,
      "step": 7860
    },
    {
      "epoch": 5.264214046822742,
      "grad_norm": 0.9984621405601501,
      "learning_rate": 4.7357859531772574e-05,
      "loss": 0.0619,
      "step": 7870
    },
    {
      "epoch": 5.270903010033445,
      "grad_norm": 1.5726069211959839,
      "learning_rate": 4.7290969899665556e-05,
      "loss": 0.0554,
      "step": 7880
    },
    {
      "epoch": 5.277591973244147,
      "grad_norm": 2.6059932708740234,
      "learning_rate": 4.722408026755853e-05,
      "loss": 0.0486,
      "step": 7890
    },
    {
      "epoch": 5.2842809364548495,
      "grad_norm": 1.7960805892944336,
      "learning_rate": 4.715719063545151e-05,
      "loss": 0.0602,
      "step": 7900
    },
    {
      "epoch": 5.290969899665551,
      "grad_norm": 1.2459561824798584,
      "learning_rate": 4.709030100334448e-05,
      "loss": 0.0685,
      "step": 7910
    },
    {
      "epoch": 5.297658862876254,
      "grad_norm": 0.4798709750175476,
      "learning_rate": 4.7023411371237456e-05,
      "loss": 0.0456,
      "step": 7920
    },
    {
      "epoch": 5.304347826086957,
      "grad_norm": 1.6899269819259644,
      "learning_rate": 4.695652173913044e-05,
      "loss": 0.0856,
      "step": 7930
    },
    {
      "epoch": 5.311036789297659,
      "grad_norm": 4.1698899269104,
      "learning_rate": 4.688963210702341e-05,
      "loss": 0.071,
      "step": 7940
    },
    {
      "epoch": 5.317725752508361,
      "grad_norm": 3.4093782901763916,
      "learning_rate": 4.6822742474916394e-05,
      "loss": 0.0696,
      "step": 7950
    },
    {
      "epoch": 5.3244147157190636,
      "grad_norm": 5.086987018585205,
      "learning_rate": 4.675585284280937e-05,
      "loss": 0.0815,
      "step": 7960
    },
    {
      "epoch": 5.331103678929766,
      "grad_norm": 0.9084048867225647,
      "learning_rate": 4.6688963210702344e-05,
      "loss": 0.0613,
      "step": 7970
    },
    {
      "epoch": 5.337792642140468,
      "grad_norm": 1.4398747682571411,
      "learning_rate": 4.662207357859532e-05,
      "loss": 0.0737,
      "step": 7980
    },
    {
      "epoch": 5.34448160535117,
      "grad_norm": 1.1616789102554321,
      "learning_rate": 4.6555183946488294e-05,
      "loss": 0.0661,
      "step": 7990
    },
    {
      "epoch": 5.351170568561873,
      "grad_norm": 1.2757660150527954,
      "learning_rate": 4.6488294314381276e-05,
      "loss": 0.0581,
      "step": 8000
    },
    {
      "epoch": 5.357859531772576,
      "grad_norm": 0.2481812834739685,
      "learning_rate": 4.642140468227425e-05,
      "loss": 0.0695,
      "step": 8010
    },
    {
      "epoch": 5.364548494983278,
      "grad_norm": 0.8452910780906677,
      "learning_rate": 4.6354515050167226e-05,
      "loss": 0.058,
      "step": 8020
    },
    {
      "epoch": 5.3712374581939795,
      "grad_norm": 5.494908332824707,
      "learning_rate": 4.62876254180602e-05,
      "loss": 0.0755,
      "step": 8030
    },
    {
      "epoch": 5.377926421404682,
      "grad_norm": 3.610271692276001,
      "learning_rate": 4.6220735785953176e-05,
      "loss": 0.0704,
      "step": 8040
    },
    {
      "epoch": 5.384615384615385,
      "grad_norm": 0.33079513907432556,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.059,
      "step": 8050
    },
    {
      "epoch": 5.391304347826087,
      "grad_norm": 3.8127245903015137,
      "learning_rate": 4.608695652173913e-05,
      "loss": 0.0665,
      "step": 8060
    },
    {
      "epoch": 5.39799331103679,
      "grad_norm": 1.6623046398162842,
      "learning_rate": 4.602006688963211e-05,
      "loss": 0.0913,
      "step": 8070
    },
    {
      "epoch": 5.404682274247492,
      "grad_norm": 2.1431431770324707,
      "learning_rate": 4.595317725752509e-05,
      "loss": 0.0615,
      "step": 8080
    },
    {
      "epoch": 5.411371237458194,
      "grad_norm": 3.0918350219726562,
      "learning_rate": 4.5886287625418065e-05,
      "loss": 0.0647,
      "step": 8090
    },
    {
      "epoch": 5.418060200668896,
      "grad_norm": 0.3633984625339508,
      "learning_rate": 4.581939799331103e-05,
      "loss": 0.0587,
      "step": 8100
    },
    {
      "epoch": 5.424749163879599,
      "grad_norm": 0.8349478244781494,
      "learning_rate": 4.5752508361204015e-05,
      "loss": 0.0743,
      "step": 8110
    },
    {
      "epoch": 5.431438127090301,
      "grad_norm": 0.17242014408111572,
      "learning_rate": 4.568561872909699e-05,
      "loss": 0.0713,
      "step": 8120
    },
    {
      "epoch": 5.438127090301004,
      "grad_norm": 1.0146863460540771,
      "learning_rate": 4.561872909698997e-05,
      "loss": 0.0519,
      "step": 8130
    },
    {
      "epoch": 5.444816053511706,
      "grad_norm": 1.4058053493499756,
      "learning_rate": 4.555183946488295e-05,
      "loss": 0.057,
      "step": 8140
    },
    {
      "epoch": 5.451505016722408,
      "grad_norm": 2.4681694507598877,
      "learning_rate": 4.548494983277592e-05,
      "loss": 0.0603,
      "step": 8150
    },
    {
      "epoch": 5.45819397993311,
      "grad_norm": 0.7661647796630859,
      "learning_rate": 4.54180602006689e-05,
      "loss": 0.055,
      "step": 8160
    },
    {
      "epoch": 5.464882943143813,
      "grad_norm": 1.6556464433670044,
      "learning_rate": 4.535117056856187e-05,
      "loss": 0.061,
      "step": 8170
    },
    {
      "epoch": 5.471571906354515,
      "grad_norm": 1.1254479885101318,
      "learning_rate": 4.5284280936454854e-05,
      "loss": 0.0895,
      "step": 8180
    },
    {
      "epoch": 5.478260869565218,
      "grad_norm": 3.0038845539093018,
      "learning_rate": 4.521739130434783e-05,
      "loss": 0.0593,
      "step": 8190
    },
    {
      "epoch": 5.48494983277592,
      "grad_norm": 3.134263753890991,
      "learning_rate": 4.5150501672240804e-05,
      "loss": 0.0626,
      "step": 8200
    },
    {
      "epoch": 5.491638795986622,
      "grad_norm": 0.7873281836509705,
      "learning_rate": 4.5083612040133786e-05,
      "loss": 0.0601,
      "step": 8210
    },
    {
      "epoch": 5.498327759197324,
      "grad_norm": 2.841003894805908,
      "learning_rate": 4.5016722408026754e-05,
      "loss": 0.0562,
      "step": 8220
    },
    {
      "epoch": 5.505016722408027,
      "grad_norm": 3.942023992538452,
      "learning_rate": 4.4949832775919736e-05,
      "loss": 0.0816,
      "step": 8230
    },
    {
      "epoch": 5.511705685618729,
      "grad_norm": 0.525076687335968,
      "learning_rate": 4.488294314381271e-05,
      "loss": 0.0645,
      "step": 8240
    },
    {
      "epoch": 5.518394648829432,
      "grad_norm": 0.6077231764793396,
      "learning_rate": 4.4816053511705686e-05,
      "loss": 0.069,
      "step": 8250
    },
    {
      "epoch": 5.525083612040134,
      "grad_norm": 2.8950376510620117,
      "learning_rate": 4.474916387959867e-05,
      "loss": 0.0577,
      "step": 8260
    },
    {
      "epoch": 5.531772575250836,
      "grad_norm": 1.7112760543823242,
      "learning_rate": 4.468227424749164e-05,
      "loss": 0.0787,
      "step": 8270
    },
    {
      "epoch": 5.538461538461538,
      "grad_norm": 0.3955395221710205,
      "learning_rate": 4.461538461538462e-05,
      "loss": 0.0578,
      "step": 8280
    },
    {
      "epoch": 5.545150501672241,
      "grad_norm": 1.8074870109558105,
      "learning_rate": 4.454849498327759e-05,
      "loss": 0.0642,
      "step": 8290
    },
    {
      "epoch": 5.551839464882943,
      "grad_norm": 1.9293107986450195,
      "learning_rate": 4.448160535117057e-05,
      "loss": 0.0624,
      "step": 8300
    },
    {
      "epoch": 5.558528428093646,
      "grad_norm": 1.021624207496643,
      "learning_rate": 4.441471571906355e-05,
      "loss": 0.078,
      "step": 8310
    },
    {
      "epoch": 5.565217391304348,
      "grad_norm": 1.299368977546692,
      "learning_rate": 4.4347826086956525e-05,
      "loss": 0.07,
      "step": 8320
    },
    {
      "epoch": 5.5719063545150505,
      "grad_norm": 0.6169732809066772,
      "learning_rate": 4.4280936454849506e-05,
      "loss": 0.0629,
      "step": 8330
    },
    {
      "epoch": 5.578595317725752,
      "grad_norm": 1.1304540634155273,
      "learning_rate": 4.4214046822742475e-05,
      "loss": 0.052,
      "step": 8340
    },
    {
      "epoch": 5.585284280936455,
      "grad_norm": 2.0306649208068848,
      "learning_rate": 4.414715719063545e-05,
      "loss": 0.0772,
      "step": 8350
    },
    {
      "epoch": 5.591973244147157,
      "grad_norm": 2.0321860313415527,
      "learning_rate": 4.408026755852843e-05,
      "loss": 0.0586,
      "step": 8360
    },
    {
      "epoch": 5.59866220735786,
      "grad_norm": 2.1765122413635254,
      "learning_rate": 4.4013377926421407e-05,
      "loss": 0.0619,
      "step": 8370
    },
    {
      "epoch": 5.605351170568562,
      "grad_norm": 2.641141414642334,
      "learning_rate": 4.394648829431438e-05,
      "loss": 0.0693,
      "step": 8380
    },
    {
      "epoch": 5.6120401337792645,
      "grad_norm": 4.06684684753418,
      "learning_rate": 4.3879598662207363e-05,
      "loss": 0.076,
      "step": 8390
    },
    {
      "epoch": 5.618729096989966,
      "grad_norm": 1.448151707649231,
      "learning_rate": 4.381270903010034e-05,
      "loss": 0.0795,
      "step": 8400
    },
    {
      "epoch": 5.625418060200669,
      "grad_norm": 2.3511595726013184,
      "learning_rate": 4.3745819397993313e-05,
      "loss": 0.05,
      "step": 8410
    },
    {
      "epoch": 5.632107023411371,
      "grad_norm": 3.952510356903076,
      "learning_rate": 4.367892976588629e-05,
      "loss": 0.0676,
      "step": 8420
    },
    {
      "epoch": 5.638795986622074,
      "grad_norm": 2.635272979736328,
      "learning_rate": 4.3612040133779264e-05,
      "loss": 0.0663,
      "step": 8430
    },
    {
      "epoch": 5.645484949832776,
      "grad_norm": 2.334019184112549,
      "learning_rate": 4.3545150501672245e-05,
      "loss": 0.0704,
      "step": 8440
    },
    {
      "epoch": 5.6521739130434785,
      "grad_norm": 3.036574602127075,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.0623,
      "step": 8450
    },
    {
      "epoch": 5.65886287625418,
      "grad_norm": 3.2615692615509033,
      "learning_rate": 4.3411371237458195e-05,
      "loss": 0.0603,
      "step": 8460
    },
    {
      "epoch": 5.665551839464883,
      "grad_norm": 0.9925899505615234,
      "learning_rate": 4.334448160535117e-05,
      "loss": 0.0697,
      "step": 8470
    },
    {
      "epoch": 5.672240802675585,
      "grad_norm": 5.593775272369385,
      "learning_rate": 4.3277591973244145e-05,
      "loss": 0.056,
      "step": 8480
    },
    {
      "epoch": 5.678929765886288,
      "grad_norm": 2.722613573074341,
      "learning_rate": 4.321070234113713e-05,
      "loss": 0.0671,
      "step": 8490
    },
    {
      "epoch": 5.68561872909699,
      "grad_norm": 6.139028072357178,
      "learning_rate": 4.31438127090301e-05,
      "loss": 0.0705,
      "step": 8500
    },
    {
      "epoch": 5.6923076923076925,
      "grad_norm": 1.3252087831497192,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0599,
      "step": 8510
    },
    {
      "epoch": 5.698996655518394,
      "grad_norm": 5.210198402404785,
      "learning_rate": 4.301003344481606e-05,
      "loss": 0.0635,
      "step": 8520
    },
    {
      "epoch": 5.705685618729097,
      "grad_norm": 0.547988772392273,
      "learning_rate": 4.294314381270903e-05,
      "loss": 0.0511,
      "step": 8530
    },
    {
      "epoch": 5.712374581939799,
      "grad_norm": 2.0190672874450684,
      "learning_rate": 4.287625418060201e-05,
      "loss": 0.0677,
      "step": 8540
    },
    {
      "epoch": 5.719063545150502,
      "grad_norm": 1.9402360916137695,
      "learning_rate": 4.2809364548494984e-05,
      "loss": 0.0714,
      "step": 8550
    },
    {
      "epoch": 5.725752508361204,
      "grad_norm": 2.5025672912597656,
      "learning_rate": 4.2742474916387966e-05,
      "loss": 0.0701,
      "step": 8560
    },
    {
      "epoch": 5.7324414715719065,
      "grad_norm": 5.094533443450928,
      "learning_rate": 4.267558528428094e-05,
      "loss": 0.0584,
      "step": 8570
    },
    {
      "epoch": 5.739130434782608,
      "grad_norm": 3.5709147453308105,
      "learning_rate": 4.2608695652173916e-05,
      "loss": 0.0646,
      "step": 8580
    },
    {
      "epoch": 5.745819397993311,
      "grad_norm": 0.7281627655029297,
      "learning_rate": 4.254180602006689e-05,
      "loss": 0.0684,
      "step": 8590
    },
    {
      "epoch": 5.752508361204013,
      "grad_norm": 0.7778987288475037,
      "learning_rate": 4.2474916387959866e-05,
      "loss": 0.059,
      "step": 8600
    },
    {
      "epoch": 5.759197324414716,
      "grad_norm": 0.7435514330863953,
      "learning_rate": 4.240802675585284e-05,
      "loss": 0.0686,
      "step": 8610
    },
    {
      "epoch": 5.765886287625418,
      "grad_norm": 5.178036212921143,
      "learning_rate": 4.234113712374582e-05,
      "loss": 0.0583,
      "step": 8620
    },
    {
      "epoch": 5.7725752508361206,
      "grad_norm": 1.6079000234603882,
      "learning_rate": 4.22742474916388e-05,
      "loss": 0.0469,
      "step": 8630
    },
    {
      "epoch": 5.7792642140468224,
      "grad_norm": 0.29372864961624146,
      "learning_rate": 4.220735785953178e-05,
      "loss": 0.0573,
      "step": 8640
    },
    {
      "epoch": 5.785953177257525,
      "grad_norm": 1.2265444993972778,
      "learning_rate": 4.214046822742475e-05,
      "loss": 0.0613,
      "step": 8650
    },
    {
      "epoch": 5.792642140468227,
      "grad_norm": 4.454895496368408,
      "learning_rate": 4.207357859531772e-05,
      "loss": 0.0676,
      "step": 8660
    },
    {
      "epoch": 5.79933110367893,
      "grad_norm": 1.5004006624221802,
      "learning_rate": 4.2006688963210705e-05,
      "loss": 0.0631,
      "step": 8670
    },
    {
      "epoch": 5.806020066889632,
      "grad_norm": 3.0780856609344482,
      "learning_rate": 4.193979933110368e-05,
      "loss": 0.0586,
      "step": 8680
    },
    {
      "epoch": 5.812709030100335,
      "grad_norm": 0.17715483903884888,
      "learning_rate": 4.187290969899666e-05,
      "loss": 0.0616,
      "step": 8690
    },
    {
      "epoch": 5.8193979933110365,
      "grad_norm": 0.9322307109832764,
      "learning_rate": 4.180602006688964e-05,
      "loss": 0.0733,
      "step": 8700
    },
    {
      "epoch": 5.826086956521739,
      "grad_norm": 1.79049551486969,
      "learning_rate": 4.1739130434782605e-05,
      "loss": 0.0733,
      "step": 8710
    },
    {
      "epoch": 5.832775919732441,
      "grad_norm": 3.6763522624969482,
      "learning_rate": 4.167224080267559e-05,
      "loss": 0.0568,
      "step": 8720
    },
    {
      "epoch": 5.839464882943144,
      "grad_norm": 0.3415290117263794,
      "learning_rate": 4.160535117056856e-05,
      "loss": 0.0544,
      "step": 8730
    },
    {
      "epoch": 5.846153846153846,
      "grad_norm": 0.7277694940567017,
      "learning_rate": 4.1538461538461544e-05,
      "loss": 0.0623,
      "step": 8740
    },
    {
      "epoch": 5.852842809364549,
      "grad_norm": 2.3001549243927,
      "learning_rate": 4.147157190635452e-05,
      "loss": 0.0552,
      "step": 8750
    },
    {
      "epoch": 5.8595317725752505,
      "grad_norm": 1.6906051635742188,
      "learning_rate": 4.1404682274247494e-05,
      "loss": 0.0733,
      "step": 8760
    },
    {
      "epoch": 5.866220735785953,
      "grad_norm": 1.930910348892212,
      "learning_rate": 4.133779264214047e-05,
      "loss": 0.0523,
      "step": 8770
    },
    {
      "epoch": 5.872909698996655,
      "grad_norm": 0.4897262454032898,
      "learning_rate": 4.1270903010033444e-05,
      "loss": 0.0646,
      "step": 8780
    },
    {
      "epoch": 5.879598662207358,
      "grad_norm": 1.7537932395935059,
      "learning_rate": 4.1204013377926426e-05,
      "loss": 0.0591,
      "step": 8790
    },
    {
      "epoch": 5.88628762541806,
      "grad_norm": 0.8843321800231934,
      "learning_rate": 4.11371237458194e-05,
      "loss": 0.0763,
      "step": 8800
    },
    {
      "epoch": 5.892976588628763,
      "grad_norm": 3.215991735458374,
      "learning_rate": 4.1070234113712376e-05,
      "loss": 0.048,
      "step": 8810
    },
    {
      "epoch": 5.8996655518394645,
      "grad_norm": 1.1658564805984497,
      "learning_rate": 4.100334448160536e-05,
      "loss": 0.0551,
      "step": 8820
    },
    {
      "epoch": 5.906354515050167,
      "grad_norm": 7.578488826751709,
      "learning_rate": 4.0936454849498326e-05,
      "loss": 0.0715,
      "step": 8830
    },
    {
      "epoch": 5.913043478260869,
      "grad_norm": 1.8251421451568604,
      "learning_rate": 4.086956521739131e-05,
      "loss": 0.0519,
      "step": 8840
    },
    {
      "epoch": 5.919732441471572,
      "grad_norm": 1.7481938600540161,
      "learning_rate": 4.080267558528428e-05,
      "loss": 0.0492,
      "step": 8850
    },
    {
      "epoch": 5.926421404682275,
      "grad_norm": 0.3837336003780365,
      "learning_rate": 4.073578595317726e-05,
      "loss": 0.0636,
      "step": 8860
    },
    {
      "epoch": 5.933110367892977,
      "grad_norm": 1.3640086650848389,
      "learning_rate": 4.066889632107024e-05,
      "loss": 0.0443,
      "step": 8870
    },
    {
      "epoch": 5.9397993311036785,
      "grad_norm": 1.6527479887008667,
      "learning_rate": 4.0602006688963214e-05,
      "loss": 0.064,
      "step": 8880
    },
    {
      "epoch": 5.946488294314381,
      "grad_norm": 0.7127817869186401,
      "learning_rate": 4.053511705685619e-05,
      "loss": 0.079,
      "step": 8890
    },
    {
      "epoch": 5.953177257525084,
      "grad_norm": 0.5994347929954529,
      "learning_rate": 4.0468227424749165e-05,
      "loss": 0.0585,
      "step": 8900
    },
    {
      "epoch": 5.959866220735786,
      "grad_norm": 2.5335919857025146,
      "learning_rate": 4.040133779264214e-05,
      "loss": 0.0588,
      "step": 8910
    },
    {
      "epoch": 5.966555183946488,
      "grad_norm": 2.706871747970581,
      "learning_rate": 4.033444816053512e-05,
      "loss": 0.0672,
      "step": 8920
    },
    {
      "epoch": 5.973244147157191,
      "grad_norm": 5.2530107498168945,
      "learning_rate": 4.0267558528428096e-05,
      "loss": 0.0659,
      "step": 8930
    },
    {
      "epoch": 5.979933110367893,
      "grad_norm": 1.829155445098877,
      "learning_rate": 4.020066889632107e-05,
      "loss": 0.07,
      "step": 8940
    },
    {
      "epoch": 5.986622073578595,
      "grad_norm": 1.090369462966919,
      "learning_rate": 4.0133779264214046e-05,
      "loss": 0.0542,
      "step": 8950
    },
    {
      "epoch": 5.993311036789297,
      "grad_norm": 2.2025110721588135,
      "learning_rate": 4.006688963210702e-05,
      "loss": 0.0644,
      "step": 8960
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.740294337272644,
      "learning_rate": 4e-05,
      "loss": 0.0677,
      "step": 8970
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.07782479375600815,
      "eval_mse": 0.07782479375600815,
      "eval_runtime": 221.7684,
      "eval_samples_per_second": 13.483,
      "eval_steps_per_second": 1.686,
      "step": 8970
    },
    {
      "epoch": 6.006688963210703,
      "grad_norm": 1.5802152156829834,
      "learning_rate": 3.993311036789298e-05,
      "loss": 0.0508,
      "step": 8980
    },
    {
      "epoch": 6.013377926421405,
      "grad_norm": 3.263096332550049,
      "learning_rate": 3.986622073578595e-05,
      "loss": 0.0465,
      "step": 8990
    },
    {
      "epoch": 6.0200668896321075,
      "grad_norm": 3.2240211963653564,
      "learning_rate": 3.9799331103678935e-05,
      "loss": 0.0614,
      "step": 9000
    },
    {
      "epoch": 6.026755852842809,
      "grad_norm": 1.102103352546692,
      "learning_rate": 3.973244147157191e-05,
      "loss": 0.0621,
      "step": 9010
    },
    {
      "epoch": 6.033444816053512,
      "grad_norm": 2.7606148719787598,
      "learning_rate": 3.9665551839464885e-05,
      "loss": 0.0633,
      "step": 9020
    },
    {
      "epoch": 6.040133779264214,
      "grad_norm": 0.45631539821624756,
      "learning_rate": 3.959866220735786e-05,
      "loss": 0.0574,
      "step": 9030
    },
    {
      "epoch": 6.046822742474917,
      "grad_norm": 6.115387439727783,
      "learning_rate": 3.9531772575250835e-05,
      "loss": 0.0597,
      "step": 9040
    },
    {
      "epoch": 6.053511705685619,
      "grad_norm": 4.481198787689209,
      "learning_rate": 3.946488294314382e-05,
      "loss": 0.0632,
      "step": 9050
    },
    {
      "epoch": 6.0602006688963215,
      "grad_norm": 0.34394732117652893,
      "learning_rate": 3.939799331103679e-05,
      "loss": 0.0436,
      "step": 9060
    },
    {
      "epoch": 6.066889632107023,
      "grad_norm": 2.672247886657715,
      "learning_rate": 3.933110367892977e-05,
      "loss": 0.0429,
      "step": 9070
    },
    {
      "epoch": 6.073578595317726,
      "grad_norm": 5.2561869621276855,
      "learning_rate": 3.926421404682274e-05,
      "loss": 0.0498,
      "step": 9080
    },
    {
      "epoch": 6.080267558528428,
      "grad_norm": 1.9382920265197754,
      "learning_rate": 3.919732441471572e-05,
      "loss": 0.0578,
      "step": 9090
    },
    {
      "epoch": 6.086956521739131,
      "grad_norm": 1.8460969924926758,
      "learning_rate": 3.91304347826087e-05,
      "loss": 0.0486,
      "step": 9100
    },
    {
      "epoch": 6.093645484949833,
      "grad_norm": 1.0311009883880615,
      "learning_rate": 3.9063545150501674e-05,
      "loss": 0.0343,
      "step": 9110
    },
    {
      "epoch": 6.1003344481605355,
      "grad_norm": 0.8744246959686279,
      "learning_rate": 3.899665551839465e-05,
      "loss": 0.0577,
      "step": 9120
    },
    {
      "epoch": 6.107023411371237,
      "grad_norm": 1.2633397579193115,
      "learning_rate": 3.892976588628763e-05,
      "loss": 0.0464,
      "step": 9130
    },
    {
      "epoch": 6.11371237458194,
      "grad_norm": 0.67624431848526,
      "learning_rate": 3.88628762541806e-05,
      "loss": 0.0373,
      "step": 9140
    },
    {
      "epoch": 6.120401337792642,
      "grad_norm": 3.1662580966949463,
      "learning_rate": 3.879598662207358e-05,
      "loss": 0.0545,
      "step": 9150
    },
    {
      "epoch": 6.127090301003345,
      "grad_norm": 0.6137005686759949,
      "learning_rate": 3.8729096989966556e-05,
      "loss": 0.0501,
      "step": 9160
    },
    {
      "epoch": 6.133779264214047,
      "grad_norm": 1.7040530443191528,
      "learning_rate": 3.866220735785953e-05,
      "loss": 0.0545,
      "step": 9170
    },
    {
      "epoch": 6.1404682274247495,
      "grad_norm": 2.090524196624756,
      "learning_rate": 3.859531772575251e-05,
      "loss": 0.0441,
      "step": 9180
    },
    {
      "epoch": 6.147157190635451,
      "grad_norm": 2.0296332836151123,
      "learning_rate": 3.852842809364549e-05,
      "loss": 0.0662,
      "step": 9190
    },
    {
      "epoch": 6.153846153846154,
      "grad_norm": 0.7337088584899902,
      "learning_rate": 3.846153846153846e-05,
      "loss": 0.0434,
      "step": 9200
    },
    {
      "epoch": 6.160535117056856,
      "grad_norm": 0.595910370349884,
      "learning_rate": 3.839464882943144e-05,
      "loss": 0.0655,
      "step": 9210
    },
    {
      "epoch": 6.167224080267559,
      "grad_norm": 1.5148868560791016,
      "learning_rate": 3.832775919732441e-05,
      "loss": 0.0555,
      "step": 9220
    },
    {
      "epoch": 6.173913043478261,
      "grad_norm": 1.5558432340621948,
      "learning_rate": 3.8260869565217395e-05,
      "loss": 0.0531,
      "step": 9230
    },
    {
      "epoch": 6.1806020066889635,
      "grad_norm": 3.5125138759613037,
      "learning_rate": 3.819397993311037e-05,
      "loss": 0.046,
      "step": 9240
    },
    {
      "epoch": 6.187290969899665,
      "grad_norm": 2.2275428771972656,
      "learning_rate": 3.812709030100335e-05,
      "loss": 0.0516,
      "step": 9250
    },
    {
      "epoch": 6.193979933110368,
      "grad_norm": 2.8403425216674805,
      "learning_rate": 3.806020066889632e-05,
      "loss": 0.066,
      "step": 9260
    },
    {
      "epoch": 6.20066889632107,
      "grad_norm": 2.0408856868743896,
      "learning_rate": 3.7993311036789295e-05,
      "loss": 0.053,
      "step": 9270
    },
    {
      "epoch": 6.207357859531773,
      "grad_norm": 1.795935869216919,
      "learning_rate": 3.792642140468228e-05,
      "loss": 0.0492,
      "step": 9280
    },
    {
      "epoch": 6.214046822742475,
      "grad_norm": 2.151693105697632,
      "learning_rate": 3.785953177257525e-05,
      "loss": 0.043,
      "step": 9290
    },
    {
      "epoch": 6.2207357859531776,
      "grad_norm": 2.5936758518218994,
      "learning_rate": 3.7792642140468233e-05,
      "loss": 0.0528,
      "step": 9300
    },
    {
      "epoch": 6.2274247491638794,
      "grad_norm": 1.5297855138778687,
      "learning_rate": 3.772575250836121e-05,
      "loss": 0.0465,
      "step": 9310
    },
    {
      "epoch": 6.234113712374582,
      "grad_norm": 3.3159210681915283,
      "learning_rate": 3.7658862876254184e-05,
      "loss": 0.0476,
      "step": 9320
    },
    {
      "epoch": 6.240802675585284,
      "grad_norm": 2.997654676437378,
      "learning_rate": 3.759197324414716e-05,
      "loss": 0.0573,
      "step": 9330
    },
    {
      "epoch": 6.247491638795987,
      "grad_norm": 0.6075377464294434,
      "learning_rate": 3.7525083612040134e-05,
      "loss": 0.0588,
      "step": 9340
    },
    {
      "epoch": 6.254180602006689,
      "grad_norm": 2.5842342376708984,
      "learning_rate": 3.745819397993311e-05,
      "loss": 0.0637,
      "step": 9350
    },
    {
      "epoch": 6.260869565217392,
      "grad_norm": 4.031673431396484,
      "learning_rate": 3.739130434782609e-05,
      "loss": 0.0525,
      "step": 9360
    },
    {
      "epoch": 6.2675585284280935,
      "grad_norm": 2.174769401550293,
      "learning_rate": 3.7324414715719065e-05,
      "loss": 0.0368,
      "step": 9370
    },
    {
      "epoch": 6.274247491638796,
      "grad_norm": 3.260265588760376,
      "learning_rate": 3.725752508361204e-05,
      "loss": 0.0483,
      "step": 9380
    },
    {
      "epoch": 6.280936454849498,
      "grad_norm": 1.3505406379699707,
      "learning_rate": 3.7190635451505016e-05,
      "loss": 0.0521,
      "step": 9390
    },
    {
      "epoch": 6.287625418060201,
      "grad_norm": 1.0265694856643677,
      "learning_rate": 3.712374581939799e-05,
      "loss": 0.0726,
      "step": 9400
    },
    {
      "epoch": 6.294314381270903,
      "grad_norm": 0.7887771129608154,
      "learning_rate": 3.705685618729097e-05,
      "loss": 0.0679,
      "step": 9410
    },
    {
      "epoch": 6.301003344481606,
      "grad_norm": 1.9342752695083618,
      "learning_rate": 3.698996655518395e-05,
      "loss": 0.0562,
      "step": 9420
    },
    {
      "epoch": 6.3076923076923075,
      "grad_norm": 0.5005589723587036,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0629,
      "step": 9430
    },
    {
      "epoch": 6.31438127090301,
      "grad_norm": 0.8035380244255066,
      "learning_rate": 3.6856187290969904e-05,
      "loss": 0.0565,
      "step": 9440
    },
    {
      "epoch": 6.321070234113712,
      "grad_norm": 2.018810272216797,
      "learning_rate": 3.678929765886287e-05,
      "loss": 0.064,
      "step": 9450
    },
    {
      "epoch": 6.327759197324415,
      "grad_norm": 0.6776765584945679,
      "learning_rate": 3.6722408026755854e-05,
      "loss": 0.0479,
      "step": 9460
    },
    {
      "epoch": 6.334448160535117,
      "grad_norm": 2.903470754623413,
      "learning_rate": 3.665551839464883e-05,
      "loss": 0.0626,
      "step": 9470
    },
    {
      "epoch": 6.34113712374582,
      "grad_norm": 1.5163041353225708,
      "learning_rate": 3.658862876254181e-05,
      "loss": 0.0557,
      "step": 9480
    },
    {
      "epoch": 6.3478260869565215,
      "grad_norm": 2.5668625831604004,
      "learning_rate": 3.6521739130434786e-05,
      "loss": 0.0504,
      "step": 9490
    },
    {
      "epoch": 6.354515050167224,
      "grad_norm": 1.1242213249206543,
      "learning_rate": 3.645484949832776e-05,
      "loss": 0.0471,
      "step": 9500
    },
    {
      "epoch": 6.361204013377926,
      "grad_norm": 0.8804264068603516,
      "learning_rate": 3.6387959866220736e-05,
      "loss": 0.0535,
      "step": 9510
    },
    {
      "epoch": 6.367892976588629,
      "grad_norm": 0.8456764221191406,
      "learning_rate": 3.632107023411371e-05,
      "loss": 0.0598,
      "step": 9520
    },
    {
      "epoch": 6.374581939799331,
      "grad_norm": 1.6663057804107666,
      "learning_rate": 3.625418060200669e-05,
      "loss": 0.0537,
      "step": 9530
    },
    {
      "epoch": 6.381270903010034,
      "grad_norm": 0.485980361700058,
      "learning_rate": 3.618729096989967e-05,
      "loss": 0.0498,
      "step": 9540
    },
    {
      "epoch": 6.3879598662207355,
      "grad_norm": 2.3799917697906494,
      "learning_rate": 3.612040133779264e-05,
      "loss": 0.0568,
      "step": 9550
    },
    {
      "epoch": 6.394648829431438,
      "grad_norm": 1.0425069332122803,
      "learning_rate": 3.6053511705685625e-05,
      "loss": 0.0441,
      "step": 9560
    },
    {
      "epoch": 6.40133779264214,
      "grad_norm": 3.5270988941192627,
      "learning_rate": 3.598662207357859e-05,
      "loss": 0.0502,
      "step": 9570
    },
    {
      "epoch": 6.408026755852843,
      "grad_norm": 6.216033458709717,
      "learning_rate": 3.5919732441471575e-05,
      "loss": 0.052,
      "step": 9580
    },
    {
      "epoch": 6.414715719063545,
      "grad_norm": 2.594477653503418,
      "learning_rate": 3.585284280936455e-05,
      "loss": 0.0553,
      "step": 9590
    },
    {
      "epoch": 6.421404682274248,
      "grad_norm": 3.850069999694824,
      "learning_rate": 3.5785953177257525e-05,
      "loss": 0.0531,
      "step": 9600
    },
    {
      "epoch": 6.4280936454849495,
      "grad_norm": 2.2683510780334473,
      "learning_rate": 3.571906354515051e-05,
      "loss": 0.0514,
      "step": 9610
    },
    {
      "epoch": 6.434782608695652,
      "grad_norm": 2.9711992740631104,
      "learning_rate": 3.565217391304348e-05,
      "loss": 0.0505,
      "step": 9620
    },
    {
      "epoch": 6.441471571906354,
      "grad_norm": 3.0900185108184814,
      "learning_rate": 3.558528428093646e-05,
      "loss": 0.0473,
      "step": 9630
    },
    {
      "epoch": 6.448160535117057,
      "grad_norm": 2.725559711456299,
      "learning_rate": 3.551839464882943e-05,
      "loss": 0.0599,
      "step": 9640
    },
    {
      "epoch": 6.454849498327759,
      "grad_norm": 0.8586891293525696,
      "learning_rate": 3.545150501672241e-05,
      "loss": 0.0574,
      "step": 9650
    },
    {
      "epoch": 6.461538461538462,
      "grad_norm": 0.4257071912288666,
      "learning_rate": 3.538461538461539e-05,
      "loss": 0.0521,
      "step": 9660
    },
    {
      "epoch": 6.468227424749164,
      "grad_norm": 2.654646635055542,
      "learning_rate": 3.5317725752508364e-05,
      "loss": 0.0488,
      "step": 9670
    },
    {
      "epoch": 6.474916387959866,
      "grad_norm": 1.2186239957809448,
      "learning_rate": 3.525083612040134e-05,
      "loss": 0.0558,
      "step": 9680
    },
    {
      "epoch": 6.481605351170568,
      "grad_norm": 0.5296560525894165,
      "learning_rate": 3.5183946488294314e-05,
      "loss": 0.0508,
      "step": 9690
    },
    {
      "epoch": 6.488294314381271,
      "grad_norm": 4.062180995941162,
      "learning_rate": 3.511705685618729e-05,
      "loss": 0.0634,
      "step": 9700
    },
    {
      "epoch": 6.494983277591973,
      "grad_norm": 3.340278148651123,
      "learning_rate": 3.505016722408027e-05,
      "loss": 0.0442,
      "step": 9710
    },
    {
      "epoch": 6.501672240802676,
      "grad_norm": 1.239185094833374,
      "learning_rate": 3.4983277591973246e-05,
      "loss": 0.0546,
      "step": 9720
    },
    {
      "epoch": 6.508361204013378,
      "grad_norm": 0.5998998880386353,
      "learning_rate": 3.491638795986622e-05,
      "loss": 0.0544,
      "step": 9730
    },
    {
      "epoch": 6.51505016722408,
      "grad_norm": 1.277424931526184,
      "learning_rate": 3.48494983277592e-05,
      "loss": 0.0596,
      "step": 9740
    },
    {
      "epoch": 6.521739130434782,
      "grad_norm": 0.9304616451263428,
      "learning_rate": 3.478260869565218e-05,
      "loss": 0.0641,
      "step": 9750
    },
    {
      "epoch": 6.528428093645485,
      "grad_norm": 0.8859513998031616,
      "learning_rate": 3.471571906354515e-05,
      "loss": 0.0566,
      "step": 9760
    },
    {
      "epoch": 6.535117056856187,
      "grad_norm": 2.800158977508545,
      "learning_rate": 3.464882943143813e-05,
      "loss": 0.0525,
      "step": 9770
    },
    {
      "epoch": 6.54180602006689,
      "grad_norm": 1.4394855499267578,
      "learning_rate": 3.45819397993311e-05,
      "loss": 0.0488,
      "step": 9780
    },
    {
      "epoch": 6.548494983277592,
      "grad_norm": 0.5787505507469177,
      "learning_rate": 3.4515050167224085e-05,
      "loss": 0.0487,
      "step": 9790
    },
    {
      "epoch": 6.555183946488294,
      "grad_norm": 1.568294644355774,
      "learning_rate": 3.444816053511706e-05,
      "loss": 0.06,
      "step": 9800
    },
    {
      "epoch": 6.561872909698996,
      "grad_norm": 2.2750236988067627,
      "learning_rate": 3.4381270903010035e-05,
      "loss": 0.0546,
      "step": 9810
    },
    {
      "epoch": 6.568561872909699,
      "grad_norm": 6.165664196014404,
      "learning_rate": 3.431438127090301e-05,
      "loss": 0.0459,
      "step": 9820
    },
    {
      "epoch": 6.575250836120401,
      "grad_norm": 0.3739396631717682,
      "learning_rate": 3.4247491638795985e-05,
      "loss": 0.0477,
      "step": 9830
    },
    {
      "epoch": 6.581939799331104,
      "grad_norm": 1.5687717199325562,
      "learning_rate": 3.4180602006688966e-05,
      "loss": 0.0552,
      "step": 9840
    },
    {
      "epoch": 6.588628762541806,
      "grad_norm": 4.11144495010376,
      "learning_rate": 3.411371237458194e-05,
      "loss": 0.0594,
      "step": 9850
    },
    {
      "epoch": 6.595317725752508,
      "grad_norm": 0.7815970182418823,
      "learning_rate": 3.4046822742474917e-05,
      "loss": 0.0615,
      "step": 9860
    },
    {
      "epoch": 6.602006688963211,
      "grad_norm": 1.0173360109329224,
      "learning_rate": 3.39799331103679e-05,
      "loss": 0.0693,
      "step": 9870
    },
    {
      "epoch": 6.608695652173913,
      "grad_norm": 2.378054618835449,
      "learning_rate": 3.3913043478260867e-05,
      "loss": 0.0643,
      "step": 9880
    },
    {
      "epoch": 6.615384615384615,
      "grad_norm": 2.3293285369873047,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0445,
      "step": 9890
    },
    {
      "epoch": 6.622073578595318,
      "grad_norm": 4.405140399932861,
      "learning_rate": 3.3779264214046823e-05,
      "loss": 0.0604,
      "step": 9900
    },
    {
      "epoch": 6.6287625418060205,
      "grad_norm": 0.9924785494804382,
      "learning_rate": 3.37123745819398e-05,
      "loss": 0.0399,
      "step": 9910
    },
    {
      "epoch": 6.635451505016722,
      "grad_norm": 0.6336735486984253,
      "learning_rate": 3.364548494983278e-05,
      "loss": 0.0485,
      "step": 9920
    },
    {
      "epoch": 6.642140468227424,
      "grad_norm": 3.0844953060150146,
      "learning_rate": 3.3578595317725755e-05,
      "loss": 0.0421,
      "step": 9930
    },
    {
      "epoch": 6.648829431438127,
      "grad_norm": 0.759816586971283,
      "learning_rate": 3.351170568561873e-05,
      "loss": 0.0551,
      "step": 9940
    },
    {
      "epoch": 6.65551839464883,
      "grad_norm": 2.6838886737823486,
      "learning_rate": 3.3444816053511705e-05,
      "loss": 0.051,
      "step": 9950
    },
    {
      "epoch": 6.662207357859532,
      "grad_norm": 1.9430067539215088,
      "learning_rate": 3.337792642140468e-05,
      "loss": 0.0404,
      "step": 9960
    },
    {
      "epoch": 6.668896321070234,
      "grad_norm": 0.7252715229988098,
      "learning_rate": 3.331103678929766e-05,
      "loss": 0.0556,
      "step": 9970
    },
    {
      "epoch": 6.6755852842809364,
      "grad_norm": 1.3465676307678223,
      "learning_rate": 3.324414715719064e-05,
      "loss": 0.0617,
      "step": 9980
    },
    {
      "epoch": 6.682274247491639,
      "grad_norm": 1.4243336915969849,
      "learning_rate": 3.317725752508362e-05,
      "loss": 0.0561,
      "step": 9990
    },
    {
      "epoch": 6.688963210702341,
      "grad_norm": 0.5687856078147888,
      "learning_rate": 3.311036789297659e-05,
      "loss": 0.0505,
      "step": 10000
    },
    {
      "epoch": 6.695652173913043,
      "grad_norm": 1.7072571516036987,
      "learning_rate": 3.304347826086956e-05,
      "loss": 0.0536,
      "step": 10010
    },
    {
      "epoch": 6.702341137123746,
      "grad_norm": 1.2271828651428223,
      "learning_rate": 3.2976588628762544e-05,
      "loss": 0.0457,
      "step": 10020
    },
    {
      "epoch": 6.709030100334449,
      "grad_norm": 2.639836311340332,
      "learning_rate": 3.290969899665552e-05,
      "loss": 0.0657,
      "step": 10030
    },
    {
      "epoch": 6.7157190635451505,
      "grad_norm": 0.9130740761756897,
      "learning_rate": 3.28428093645485e-05,
      "loss": 0.0394,
      "step": 10040
    },
    {
      "epoch": 6.722408026755852,
      "grad_norm": 2.828885555267334,
      "learning_rate": 3.2775919732441476e-05,
      "loss": 0.052,
      "step": 10050
    },
    {
      "epoch": 6.729096989966555,
      "grad_norm": 6.914234638214111,
      "learning_rate": 3.270903010033445e-05,
      "loss": 0.0664,
      "step": 10060
    },
    {
      "epoch": 6.735785953177258,
      "grad_norm": 3.1282074451446533,
      "learning_rate": 3.2642140468227426e-05,
      "loss": 0.0642,
      "step": 10070
    },
    {
      "epoch": 6.74247491638796,
      "grad_norm": 1.06096351146698,
      "learning_rate": 3.25752508361204e-05,
      "loss": 0.0625,
      "step": 10080
    },
    {
      "epoch": 6.749163879598662,
      "grad_norm": 0.637898325920105,
      "learning_rate": 3.250836120401338e-05,
      "loss": 0.0521,
      "step": 10090
    },
    {
      "epoch": 6.7558528428093645,
      "grad_norm": 1.6803759336471558,
      "learning_rate": 3.244147157190636e-05,
      "loss": 0.0593,
      "step": 10100
    },
    {
      "epoch": 6.762541806020067,
      "grad_norm": 1.0126163959503174,
      "learning_rate": 3.237458193979933e-05,
      "loss": 0.0509,
      "step": 10110
    },
    {
      "epoch": 6.769230769230769,
      "grad_norm": 1.9007771015167236,
      "learning_rate": 3.230769230769231e-05,
      "loss": 0.051,
      "step": 10120
    },
    {
      "epoch": 6.775919732441472,
      "grad_norm": 1.0006183385849,
      "learning_rate": 3.224080267558528e-05,
      "loss": 0.0452,
      "step": 10130
    },
    {
      "epoch": 6.782608695652174,
      "grad_norm": 1.4270817041397095,
      "learning_rate": 3.217391304347826e-05,
      "loss": 0.0585,
      "step": 10140
    },
    {
      "epoch": 6.789297658862877,
      "grad_norm": 1.6937892436981201,
      "learning_rate": 3.210702341137124e-05,
      "loss": 0.0662,
      "step": 10150
    },
    {
      "epoch": 6.7959866220735785,
      "grad_norm": 1.8574788570404053,
      "learning_rate": 3.2040133779264215e-05,
      "loss": 0.0686,
      "step": 10160
    },
    {
      "epoch": 6.802675585284281,
      "grad_norm": 3.8167994022369385,
      "learning_rate": 3.19732441471572e-05,
      "loss": 0.0611,
      "step": 10170
    },
    {
      "epoch": 6.809364548494983,
      "grad_norm": 0.6884555816650391,
      "learning_rate": 3.190635451505017e-05,
      "loss": 0.0609,
      "step": 10180
    },
    {
      "epoch": 6.816053511705686,
      "grad_norm": 2.10984206199646,
      "learning_rate": 3.183946488294314e-05,
      "loss": 0.0542,
      "step": 10190
    },
    {
      "epoch": 6.822742474916388,
      "grad_norm": 1.2238664627075195,
      "learning_rate": 3.177257525083612e-05,
      "loss": 0.0501,
      "step": 10200
    },
    {
      "epoch": 6.829431438127091,
      "grad_norm": 1.3676847219467163,
      "learning_rate": 3.17056856187291e-05,
      "loss": 0.0494,
      "step": 10210
    },
    {
      "epoch": 6.8361204013377925,
      "grad_norm": 1.1930272579193115,
      "learning_rate": 3.163879598662208e-05,
      "loss": 0.0693,
      "step": 10220
    },
    {
      "epoch": 6.842809364548495,
      "grad_norm": 2.3715999126434326,
      "learning_rate": 3.1571906354515054e-05,
      "loss": 0.0476,
      "step": 10230
    },
    {
      "epoch": 6.849498327759197,
      "grad_norm": 4.478176593780518,
      "learning_rate": 3.150501672240803e-05,
      "loss": 0.0526,
      "step": 10240
    },
    {
      "epoch": 6.8561872909699,
      "grad_norm": 3.742927074432373,
      "learning_rate": 3.1438127090301004e-05,
      "loss": 0.0522,
      "step": 10250
    },
    {
      "epoch": 6.862876254180602,
      "grad_norm": 0.9782775044441223,
      "learning_rate": 3.137123745819398e-05,
      "loss": 0.0553,
      "step": 10260
    },
    {
      "epoch": 6.869565217391305,
      "grad_norm": 1.5323350429534912,
      "learning_rate": 3.130434782608696e-05,
      "loss": 0.0704,
      "step": 10270
    },
    {
      "epoch": 6.8762541806020065,
      "grad_norm": 3.9432942867279053,
      "learning_rate": 3.1237458193979936e-05,
      "loss": 0.0506,
      "step": 10280
    },
    {
      "epoch": 6.882943143812709,
      "grad_norm": 5.13473653793335,
      "learning_rate": 3.117056856187291e-05,
      "loss": 0.0399,
      "step": 10290
    },
    {
      "epoch": 6.889632107023411,
      "grad_norm": 0.9223735928535461,
      "learning_rate": 3.110367892976589e-05,
      "loss": 0.0368,
      "step": 10300
    },
    {
      "epoch": 6.896321070234114,
      "grad_norm": 2.1676201820373535,
      "learning_rate": 3.103678929765886e-05,
      "loss": 0.0462,
      "step": 10310
    },
    {
      "epoch": 6.903010033444816,
      "grad_norm": 2.290289878845215,
      "learning_rate": 3.096989966555184e-05,
      "loss": 0.0502,
      "step": 10320
    },
    {
      "epoch": 6.909698996655519,
      "grad_norm": 2.9174978733062744,
      "learning_rate": 3.090301003344482e-05,
      "loss": 0.0475,
      "step": 10330
    },
    {
      "epoch": 6.916387959866221,
      "grad_norm": 2.0384767055511475,
      "learning_rate": 3.083612040133779e-05,
      "loss": 0.0651,
      "step": 10340
    },
    {
      "epoch": 6.923076923076923,
      "grad_norm": 2.338029384613037,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.053,
      "step": 10350
    },
    {
      "epoch": 6.929765886287625,
      "grad_norm": 1.1559053659439087,
      "learning_rate": 3.070234113712375e-05,
      "loss": 0.0635,
      "step": 10360
    },
    {
      "epoch": 6.936454849498328,
      "grad_norm": 5.035431861877441,
      "learning_rate": 3.0635451505016724e-05,
      "loss": 0.0644,
      "step": 10370
    },
    {
      "epoch": 6.94314381270903,
      "grad_norm": 1.137031078338623,
      "learning_rate": 3.05685618729097e-05,
      "loss": 0.0476,
      "step": 10380
    },
    {
      "epoch": 6.949832775919733,
      "grad_norm": 2.5256316661834717,
      "learning_rate": 3.0501672240802674e-05,
      "loss": 0.0671,
      "step": 10390
    },
    {
      "epoch": 6.956521739130435,
      "grad_norm": 1.1378544569015503,
      "learning_rate": 3.0434782608695656e-05,
      "loss": 0.0548,
      "step": 10400
    },
    {
      "epoch": 6.963210702341137,
      "grad_norm": 0.49568507075309753,
      "learning_rate": 3.036789297658863e-05,
      "loss": 0.0563,
      "step": 10410
    },
    {
      "epoch": 6.969899665551839,
      "grad_norm": 2.079551935195923,
      "learning_rate": 3.0301003344481603e-05,
      "loss": 0.0547,
      "step": 10420
    },
    {
      "epoch": 6.976588628762542,
      "grad_norm": 1.71834135055542,
      "learning_rate": 3.0234113712374585e-05,
      "loss": 0.0644,
      "step": 10430
    },
    {
      "epoch": 6.983277591973244,
      "grad_norm": 1.868596076965332,
      "learning_rate": 3.016722408026756e-05,
      "loss": 0.0585,
      "step": 10440
    },
    {
      "epoch": 6.989966555183947,
      "grad_norm": 2.3574764728546143,
      "learning_rate": 3.0100334448160538e-05,
      "loss": 0.051,
      "step": 10450
    },
    {
      "epoch": 6.996655518394649,
      "grad_norm": 3.6523358821868896,
      "learning_rate": 3.0033444816053513e-05,
      "loss": 0.0553,
      "step": 10460
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.08115968108177185,
      "eval_mse": 0.08115968108177185,
      "eval_runtime": 226.6126,
      "eval_samples_per_second": 13.194,
      "eval_steps_per_second": 1.65,
      "step": 10465
    },
    {
      "epoch": 7.003344481605351,
      "grad_norm": 1.6401000022888184,
      "learning_rate": 2.9966555183946488e-05,
      "loss": 0.0587,
      "step": 10470
    },
    {
      "epoch": 7.010033444816053,
      "grad_norm": 0.4730532467365265,
      "learning_rate": 2.9899665551839467e-05,
      "loss": 0.0344,
      "step": 10480
    },
    {
      "epoch": 7.016722408026756,
      "grad_norm": 3.996577501296997,
      "learning_rate": 2.9832775919732442e-05,
      "loss": 0.0429,
      "step": 10490
    },
    {
      "epoch": 7.023411371237458,
      "grad_norm": 2.086496591567993,
      "learning_rate": 2.9765886287625424e-05,
      "loss": 0.0382,
      "step": 10500
    },
    {
      "epoch": 7.030100334448161,
      "grad_norm": 0.7093859910964966,
      "learning_rate": 2.9698996655518395e-05,
      "loss": 0.0539,
      "step": 10510
    },
    {
      "epoch": 7.036789297658863,
      "grad_norm": 6.086569786071777,
      "learning_rate": 2.963210702341137e-05,
      "loss": 0.0519,
      "step": 10520
    },
    {
      "epoch": 7.043478260869565,
      "grad_norm": 2.9343221187591553,
      "learning_rate": 2.9565217391304352e-05,
      "loss": 0.0402,
      "step": 10530
    },
    {
      "epoch": 7.050167224080267,
      "grad_norm": 1.3587912321090698,
      "learning_rate": 2.9498327759197324e-05,
      "loss": 0.0328,
      "step": 10540
    },
    {
      "epoch": 7.05685618729097,
      "grad_norm": 0.7912111878395081,
      "learning_rate": 2.9431438127090305e-05,
      "loss": 0.0397,
      "step": 10550
    },
    {
      "epoch": 7.063545150501672,
      "grad_norm": 1.5741112232208252,
      "learning_rate": 2.936454849498328e-05,
      "loss": 0.0342,
      "step": 10560
    },
    {
      "epoch": 7.070234113712375,
      "grad_norm": 0.40685224533081055,
      "learning_rate": 2.9297658862876256e-05,
      "loss": 0.0286,
      "step": 10570
    },
    {
      "epoch": 7.076923076923077,
      "grad_norm": 0.5212937593460083,
      "learning_rate": 2.9230769230769234e-05,
      "loss": 0.0405,
      "step": 10580
    },
    {
      "epoch": 7.083612040133779,
      "grad_norm": 1.1463117599487305,
      "learning_rate": 2.916387959866221e-05,
      "loss": 0.0367,
      "step": 10590
    },
    {
      "epoch": 7.090301003344481,
      "grad_norm": 0.6308630108833313,
      "learning_rate": 2.9096989966555184e-05,
      "loss": 0.034,
      "step": 10600
    },
    {
      "epoch": 7.096989966555184,
      "grad_norm": 0.5892065763473511,
      "learning_rate": 2.9030100334448162e-05,
      "loss": 0.0349,
      "step": 10610
    },
    {
      "epoch": 7.103678929765886,
      "grad_norm": 2.7921645641326904,
      "learning_rate": 2.8963210702341137e-05,
      "loss": 0.0467,
      "step": 10620
    },
    {
      "epoch": 7.110367892976589,
      "grad_norm": 0.6055333018302917,
      "learning_rate": 2.8896321070234116e-05,
      "loss": 0.0439,
      "step": 10630
    },
    {
      "epoch": 7.117056856187291,
      "grad_norm": 0.6409103274345398,
      "learning_rate": 2.882943143812709e-05,
      "loss": 0.0423,
      "step": 10640
    },
    {
      "epoch": 7.1237458193979935,
      "grad_norm": 2.2501044273376465,
      "learning_rate": 2.8762541806020066e-05,
      "loss": 0.0344,
      "step": 10650
    },
    {
      "epoch": 7.130434782608695,
      "grad_norm": 0.6633714437484741,
      "learning_rate": 2.8695652173913044e-05,
      "loss": 0.0447,
      "step": 10660
    },
    {
      "epoch": 7.137123745819398,
      "grad_norm": 4.393712043762207,
      "learning_rate": 2.862876254180602e-05,
      "loss": 0.0338,
      "step": 10670
    },
    {
      "epoch": 7.1438127090301,
      "grad_norm": 2.0332045555114746,
      "learning_rate": 2.8561872909699e-05,
      "loss": 0.0417,
      "step": 10680
    },
    {
      "epoch": 7.150501672240803,
      "grad_norm": 2.332609176635742,
      "learning_rate": 2.8494983277591973e-05,
      "loss": 0.0423,
      "step": 10690
    },
    {
      "epoch": 7.157190635451505,
      "grad_norm": 1.8318698406219482,
      "learning_rate": 2.8428093645484948e-05,
      "loss": 0.0381,
      "step": 10700
    },
    {
      "epoch": 7.1638795986622075,
      "grad_norm": 1.953809380531311,
      "learning_rate": 2.836120401337793e-05,
      "loss": 0.0498,
      "step": 10710
    },
    {
      "epoch": 7.170568561872909,
      "grad_norm": 3.4478530883789062,
      "learning_rate": 2.8294314381270905e-05,
      "loss": 0.0338,
      "step": 10720
    },
    {
      "epoch": 7.177257525083612,
      "grad_norm": 1.5791703462600708,
      "learning_rate": 2.8227424749163883e-05,
      "loss": 0.0459,
      "step": 10730
    },
    {
      "epoch": 7.183946488294314,
      "grad_norm": 1.9636425971984863,
      "learning_rate": 2.8160535117056858e-05,
      "loss": 0.0445,
      "step": 10740
    },
    {
      "epoch": 7.190635451505017,
      "grad_norm": 2.8458187580108643,
      "learning_rate": 2.8093645484949833e-05,
      "loss": 0.0386,
      "step": 10750
    },
    {
      "epoch": 7.197324414715719,
      "grad_norm": 1.9556761980056763,
      "learning_rate": 2.802675585284281e-05,
      "loss": 0.0501,
      "step": 10760
    },
    {
      "epoch": 7.2040133779264215,
      "grad_norm": 0.8367358446121216,
      "learning_rate": 2.7959866220735787e-05,
      "loss": 0.0484,
      "step": 10770
    },
    {
      "epoch": 7.210702341137123,
      "grad_norm": 0.5826350450515747,
      "learning_rate": 2.7892976588628765e-05,
      "loss": 0.0354,
      "step": 10780
    },
    {
      "epoch": 7.217391304347826,
      "grad_norm": 1.0834014415740967,
      "learning_rate": 2.782608695652174e-05,
      "loss": 0.0401,
      "step": 10790
    },
    {
      "epoch": 7.224080267558528,
      "grad_norm": 2.8577046394348145,
      "learning_rate": 2.7759197324414715e-05,
      "loss": 0.0389,
      "step": 10800
    },
    {
      "epoch": 7.230769230769231,
      "grad_norm": 1.558532476425171,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.0465,
      "step": 10810
    },
    {
      "epoch": 7.237458193979933,
      "grad_norm": 1.4853328466415405,
      "learning_rate": 2.762541806020067e-05,
      "loss": 0.0425,
      "step": 10820
    },
    {
      "epoch": 7.2441471571906355,
      "grad_norm": 2.165236473083496,
      "learning_rate": 2.755852842809365e-05,
      "loss": 0.0415,
      "step": 10830
    },
    {
      "epoch": 7.250836120401337,
      "grad_norm": 0.9280331134796143,
      "learning_rate": 2.7491638795986625e-05,
      "loss": 0.0412,
      "step": 10840
    },
    {
      "epoch": 7.25752508361204,
      "grad_norm": 3.1878480911254883,
      "learning_rate": 2.7424749163879597e-05,
      "loss": 0.047,
      "step": 10850
    },
    {
      "epoch": 7.264214046822742,
      "grad_norm": 5.7657623291015625,
      "learning_rate": 2.735785953177258e-05,
      "loss": 0.0489,
      "step": 10860
    },
    {
      "epoch": 7.270903010033445,
      "grad_norm": 2.0762507915496826,
      "learning_rate": 2.7290969899665554e-05,
      "loss": 0.041,
      "step": 10870
    },
    {
      "epoch": 7.277591973244147,
      "grad_norm": 2.477379560470581,
      "learning_rate": 2.7224080267558526e-05,
      "loss": 0.0513,
      "step": 10880
    },
    {
      "epoch": 7.2842809364548495,
      "grad_norm": 1.3350303173065186,
      "learning_rate": 2.7157190635451507e-05,
      "loss": 0.048,
      "step": 10890
    },
    {
      "epoch": 7.290969899665551,
      "grad_norm": 2.863292694091797,
      "learning_rate": 2.7090301003344482e-05,
      "loss": 0.0492,
      "step": 10900
    },
    {
      "epoch": 7.297658862876254,
      "grad_norm": 4.157398700714111,
      "learning_rate": 2.702341137123746e-05,
      "loss": 0.0367,
      "step": 10910
    },
    {
      "epoch": 7.304347826086957,
      "grad_norm": 0.9202874302864075,
      "learning_rate": 2.6956521739130436e-05,
      "loss": 0.0305,
      "step": 10920
    },
    {
      "epoch": 7.311036789297659,
      "grad_norm": 0.9513516426086426,
      "learning_rate": 2.688963210702341e-05,
      "loss": 0.0409,
      "step": 10930
    },
    {
      "epoch": 7.317725752508361,
      "grad_norm": 2.04758620262146,
      "learning_rate": 2.682274247491639e-05,
      "loss": 0.0405,
      "step": 10940
    },
    {
      "epoch": 7.3244147157190636,
      "grad_norm": 4.638932228088379,
      "learning_rate": 2.6755852842809364e-05,
      "loss": 0.0473,
      "step": 10950
    },
    {
      "epoch": 7.331103678929766,
      "grad_norm": 1.4520858526229858,
      "learning_rate": 2.6688963210702346e-05,
      "loss": 0.0381,
      "step": 10960
    },
    {
      "epoch": 7.337792642140468,
      "grad_norm": 2.646759510040283,
      "learning_rate": 2.6622073578595318e-05,
      "loss": 0.0427,
      "step": 10970
    },
    {
      "epoch": 7.34448160535117,
      "grad_norm": 2.3960533142089844,
      "learning_rate": 2.6555183946488293e-05,
      "loss": 0.0464,
      "step": 10980
    },
    {
      "epoch": 7.351170568561873,
      "grad_norm": 1.6093047857284546,
      "learning_rate": 2.6488294314381275e-05,
      "loss": 0.0469,
      "step": 10990
    },
    {
      "epoch": 7.357859531772576,
      "grad_norm": 0.6260418891906738,
      "learning_rate": 2.6421404682274246e-05,
      "loss": 0.0397,
      "step": 11000
    },
    {
      "epoch": 7.364548494983278,
      "grad_norm": 4.12513542175293,
      "learning_rate": 2.6354515050167228e-05,
      "loss": 0.0452,
      "step": 11010
    },
    {
      "epoch": 7.3712374581939795,
      "grad_norm": 2.9965901374816895,
      "learning_rate": 2.6287625418060203e-05,
      "loss": 0.0348,
      "step": 11020
    },
    {
      "epoch": 7.377926421404682,
      "grad_norm": 1.5032777786254883,
      "learning_rate": 2.6220735785953178e-05,
      "loss": 0.0425,
      "step": 11030
    },
    {
      "epoch": 7.384615384615385,
      "grad_norm": 0.8793359398841858,
      "learning_rate": 2.6153846153846157e-05,
      "loss": 0.0469,
      "step": 11040
    },
    {
      "epoch": 7.391304347826087,
      "grad_norm": 2.3625988960266113,
      "learning_rate": 2.608695652173913e-05,
      "loss": 0.0365,
      "step": 11050
    },
    {
      "epoch": 7.39799331103679,
      "grad_norm": 2.840656280517578,
      "learning_rate": 2.602006688963211e-05,
      "loss": 0.0482,
      "step": 11060
    },
    {
      "epoch": 7.404682274247492,
      "grad_norm": 1.521253228187561,
      "learning_rate": 2.5953177257525085e-05,
      "loss": 0.0488,
      "step": 11070
    },
    {
      "epoch": 7.411371237458194,
      "grad_norm": 0.7154616117477417,
      "learning_rate": 2.588628762541806e-05,
      "loss": 0.0462,
      "step": 11080
    },
    {
      "epoch": 7.418060200668896,
      "grad_norm": 3.5693202018737793,
      "learning_rate": 2.581939799331104e-05,
      "loss": 0.0355,
      "step": 11090
    },
    {
      "epoch": 7.424749163879599,
      "grad_norm": 1.9669532775878906,
      "learning_rate": 2.5752508361204013e-05,
      "loss": 0.0358,
      "step": 11100
    },
    {
      "epoch": 7.431438127090301,
      "grad_norm": 0.8268588185310364,
      "learning_rate": 2.568561872909699e-05,
      "loss": 0.042,
      "step": 11110
    },
    {
      "epoch": 7.438127090301004,
      "grad_norm": 2.0618767738342285,
      "learning_rate": 2.5618729096989967e-05,
      "loss": 0.0458,
      "step": 11120
    },
    {
      "epoch": 7.444816053511706,
      "grad_norm": 3.76814341545105,
      "learning_rate": 2.5551839464882942e-05,
      "loss": 0.0428,
      "step": 11130
    },
    {
      "epoch": 7.451505016722408,
      "grad_norm": 1.0186822414398193,
      "learning_rate": 2.5484949832775924e-05,
      "loss": 0.0393,
      "step": 11140
    },
    {
      "epoch": 7.45819397993311,
      "grad_norm": 2.9750678539276123,
      "learning_rate": 2.54180602006689e-05,
      "loss": 0.0386,
      "step": 11150
    },
    {
      "epoch": 7.464882943143813,
      "grad_norm": 4.203641891479492,
      "learning_rate": 2.535117056856187e-05,
      "loss": 0.0434,
      "step": 11160
    },
    {
      "epoch": 7.471571906354515,
      "grad_norm": 1.066295862197876,
      "learning_rate": 2.5284280936454852e-05,
      "loss": 0.0439,
      "step": 11170
    },
    {
      "epoch": 7.478260869565218,
      "grad_norm": 1.0334880352020264,
      "learning_rate": 2.5217391304347827e-05,
      "loss": 0.0429,
      "step": 11180
    },
    {
      "epoch": 7.48494983277592,
      "grad_norm": 3.2917025089263916,
      "learning_rate": 2.5150501672240806e-05,
      "loss": 0.0475,
      "step": 11190
    },
    {
      "epoch": 7.491638795986622,
      "grad_norm": 3.197359800338745,
      "learning_rate": 2.508361204013378e-05,
      "loss": 0.0458,
      "step": 11200
    },
    {
      "epoch": 7.498327759197324,
      "grad_norm": 1.312531590461731,
      "learning_rate": 2.5016722408026756e-05,
      "loss": 0.0425,
      "step": 11210
    },
    {
      "epoch": 7.505016722408027,
      "grad_norm": 1.0706628561019897,
      "learning_rate": 2.4949832775919734e-05,
      "loss": 0.0351,
      "step": 11220
    },
    {
      "epoch": 7.511705685618729,
      "grad_norm": 2.260401725769043,
      "learning_rate": 2.488294314381271e-05,
      "loss": 0.0469,
      "step": 11230
    },
    {
      "epoch": 7.518394648829432,
      "grad_norm": 2.654482841491699,
      "learning_rate": 2.4816053511705688e-05,
      "loss": 0.0431,
      "step": 11240
    },
    {
      "epoch": 7.525083612040134,
      "grad_norm": 2.9351720809936523,
      "learning_rate": 2.4749163879598663e-05,
      "loss": 0.0343,
      "step": 11250
    },
    {
      "epoch": 7.531772575250836,
      "grad_norm": 3.859576463699341,
      "learning_rate": 2.468227424749164e-05,
      "loss": 0.0336,
      "step": 11260
    },
    {
      "epoch": 7.538461538461538,
      "grad_norm": 1.0542091131210327,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.0326,
      "step": 11270
    },
    {
      "epoch": 7.545150501672241,
      "grad_norm": 1.7328006029129028,
      "learning_rate": 2.454849498327759e-05,
      "loss": 0.0413,
      "step": 11280
    },
    {
      "epoch": 7.551839464882943,
      "grad_norm": 1.8054059743881226,
      "learning_rate": 2.448160535117057e-05,
      "loss": 0.0407,
      "step": 11290
    },
    {
      "epoch": 7.558528428093646,
      "grad_norm": 1.1257389783859253,
      "learning_rate": 2.4414715719063548e-05,
      "loss": 0.0427,
      "step": 11300
    },
    {
      "epoch": 7.565217391304348,
      "grad_norm": 1.9701489210128784,
      "learning_rate": 2.4347826086956523e-05,
      "loss": 0.0439,
      "step": 11310
    },
    {
      "epoch": 7.5719063545150505,
      "grad_norm": 2.4522111415863037,
      "learning_rate": 2.4280936454849498e-05,
      "loss": 0.0284,
      "step": 11320
    },
    {
      "epoch": 7.578595317725752,
      "grad_norm": 1.3907763957977295,
      "learning_rate": 2.4214046822742476e-05,
      "loss": 0.0301,
      "step": 11330
    },
    {
      "epoch": 7.585284280936455,
      "grad_norm": 0.5545759201049805,
      "learning_rate": 2.414715719063545e-05,
      "loss": 0.0344,
      "step": 11340
    },
    {
      "epoch": 7.591973244147157,
      "grad_norm": 1.364746332168579,
      "learning_rate": 2.408026755852843e-05,
      "loss": 0.0485,
      "step": 11350
    },
    {
      "epoch": 7.59866220735786,
      "grad_norm": 1.384533405303955,
      "learning_rate": 2.401337792642141e-05,
      "loss": 0.0527,
      "step": 11360
    },
    {
      "epoch": 7.605351170568562,
      "grad_norm": 1.4667154550552368,
      "learning_rate": 2.394648829431438e-05,
      "loss": 0.0462,
      "step": 11370
    },
    {
      "epoch": 7.6120401337792645,
      "grad_norm": 2.1494758129119873,
      "learning_rate": 2.387959866220736e-05,
      "loss": 0.0363,
      "step": 11380
    },
    {
      "epoch": 7.618729096989966,
      "grad_norm": 1.8320368528366089,
      "learning_rate": 2.3812709030100337e-05,
      "loss": 0.0415,
      "step": 11390
    },
    {
      "epoch": 7.625418060200669,
      "grad_norm": 1.5719565153121948,
      "learning_rate": 2.3745819397993312e-05,
      "loss": 0.0454,
      "step": 11400
    },
    {
      "epoch": 7.632107023411371,
      "grad_norm": 0.6953816413879395,
      "learning_rate": 2.3678929765886287e-05,
      "loss": 0.0366,
      "step": 11410
    },
    {
      "epoch": 7.638795986622074,
      "grad_norm": 0.7792525887489319,
      "learning_rate": 2.3612040133779265e-05,
      "loss": 0.0385,
      "step": 11420
    },
    {
      "epoch": 7.645484949832776,
      "grad_norm": 1.783232569694519,
      "learning_rate": 2.354515050167224e-05,
      "loss": 0.0362,
      "step": 11430
    },
    {
      "epoch": 7.6521739130434785,
      "grad_norm": 0.43922555446624756,
      "learning_rate": 2.347826086956522e-05,
      "loss": 0.0529,
      "step": 11440
    },
    {
      "epoch": 7.65886287625418,
      "grad_norm": 2.1042943000793457,
      "learning_rate": 2.3411371237458197e-05,
      "loss": 0.0408,
      "step": 11450
    },
    {
      "epoch": 7.665551839464883,
      "grad_norm": 1.4735808372497559,
      "learning_rate": 2.3344481605351172e-05,
      "loss": 0.0519,
      "step": 11460
    },
    {
      "epoch": 7.672240802675585,
      "grad_norm": 0.5197001099586487,
      "learning_rate": 2.3277591973244147e-05,
      "loss": 0.0519,
      "step": 11470
    },
    {
      "epoch": 7.678929765886288,
      "grad_norm": 1.7215490341186523,
      "learning_rate": 2.3210702341137126e-05,
      "loss": 0.0451,
      "step": 11480
    },
    {
      "epoch": 7.68561872909699,
      "grad_norm": 0.8570510745048523,
      "learning_rate": 2.31438127090301e-05,
      "loss": 0.033,
      "step": 11490
    },
    {
      "epoch": 7.6923076923076925,
      "grad_norm": 1.0296355485916138,
      "learning_rate": 2.307692307692308e-05,
      "loss": 0.0561,
      "step": 11500
    },
    {
      "epoch": 7.698996655518394,
      "grad_norm": 3.8507938385009766,
      "learning_rate": 2.3010033444816054e-05,
      "loss": 0.0453,
      "step": 11510
    },
    {
      "epoch": 7.705685618729097,
      "grad_norm": 1.8961751461029053,
      "learning_rate": 2.2943143812709033e-05,
      "loss": 0.0492,
      "step": 11520
    },
    {
      "epoch": 7.712374581939799,
      "grad_norm": 2.693507671356201,
      "learning_rate": 2.2876254180602008e-05,
      "loss": 0.0497,
      "step": 11530
    },
    {
      "epoch": 7.719063545150502,
      "grad_norm": 2.789548635482788,
      "learning_rate": 2.2809364548494986e-05,
      "loss": 0.0295,
      "step": 11540
    },
    {
      "epoch": 7.725752508361204,
      "grad_norm": 3.2721359729766846,
      "learning_rate": 2.274247491638796e-05,
      "loss": 0.0397,
      "step": 11550
    },
    {
      "epoch": 7.7324414715719065,
      "grad_norm": 0.6751932501792908,
      "learning_rate": 2.2675585284280936e-05,
      "loss": 0.0418,
      "step": 11560
    },
    {
      "epoch": 7.739130434782608,
      "grad_norm": 2.357602596282959,
      "learning_rate": 2.2608695652173914e-05,
      "loss": 0.0502,
      "step": 11570
    },
    {
      "epoch": 7.745819397993311,
      "grad_norm": 0.9620494246482849,
      "learning_rate": 2.2541806020066893e-05,
      "loss": 0.0487,
      "step": 11580
    },
    {
      "epoch": 7.752508361204013,
      "grad_norm": 2.187662363052368,
      "learning_rate": 2.2474916387959868e-05,
      "loss": 0.0435,
      "step": 11590
    },
    {
      "epoch": 7.759197324414716,
      "grad_norm": 1.5170892477035522,
      "learning_rate": 2.2408026755852843e-05,
      "loss": 0.0431,
      "step": 11600
    },
    {
      "epoch": 7.765886287625418,
      "grad_norm": 1.1562601327896118,
      "learning_rate": 2.234113712374582e-05,
      "loss": 0.0323,
      "step": 11610
    },
    {
      "epoch": 7.7725752508361206,
      "grad_norm": 3.6104776859283447,
      "learning_rate": 2.2274247491638796e-05,
      "loss": 0.042,
      "step": 11620
    },
    {
      "epoch": 7.7792642140468224,
      "grad_norm": 0.7847824692726135,
      "learning_rate": 2.2207357859531775e-05,
      "loss": 0.0399,
      "step": 11630
    },
    {
      "epoch": 7.785953177257525,
      "grad_norm": 1.0568705797195435,
      "learning_rate": 2.2140468227424753e-05,
      "loss": 0.0446,
      "step": 11640
    },
    {
      "epoch": 7.792642140468227,
      "grad_norm": 0.699948787689209,
      "learning_rate": 2.2073578595317725e-05,
      "loss": 0.0347,
      "step": 11650
    },
    {
      "epoch": 7.79933110367893,
      "grad_norm": 1.0860364437103271,
      "learning_rate": 2.2006688963210703e-05,
      "loss": 0.0403,
      "step": 11660
    },
    {
      "epoch": 7.806020066889632,
      "grad_norm": 0.90751713514328,
      "learning_rate": 2.1939799331103682e-05,
      "loss": 0.0413,
      "step": 11670
    },
    {
      "epoch": 7.812709030100335,
      "grad_norm": 2.4008915424346924,
      "learning_rate": 2.1872909698996657e-05,
      "loss": 0.0463,
      "step": 11680
    },
    {
      "epoch": 7.8193979933110365,
      "grad_norm": 2.8335297107696533,
      "learning_rate": 2.1806020066889632e-05,
      "loss": 0.048,
      "step": 11690
    },
    {
      "epoch": 7.826086956521739,
      "grad_norm": 0.7693464756011963,
      "learning_rate": 2.173913043478261e-05,
      "loss": 0.047,
      "step": 11700
    },
    {
      "epoch": 7.832775919732441,
      "grad_norm": 4.794638633728027,
      "learning_rate": 2.1672240802675585e-05,
      "loss": 0.0384,
      "step": 11710
    },
    {
      "epoch": 7.839464882943144,
      "grad_norm": 3.020655632019043,
      "learning_rate": 2.1605351170568564e-05,
      "loss": 0.0468,
      "step": 11720
    },
    {
      "epoch": 7.846153846153846,
      "grad_norm": 1.1203726530075073,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.0489,
      "step": 11730
    },
    {
      "epoch": 7.852842809364549,
      "grad_norm": 0.991735577583313,
      "learning_rate": 2.1471571906354514e-05,
      "loss": 0.0393,
      "step": 11740
    },
    {
      "epoch": 7.8595317725752505,
      "grad_norm": 2.7917697429656982,
      "learning_rate": 2.1404682274247492e-05,
      "loss": 0.0433,
      "step": 11750
    },
    {
      "epoch": 7.866220735785953,
      "grad_norm": 1.8894798755645752,
      "learning_rate": 2.133779264214047e-05,
      "loss": 0.0299,
      "step": 11760
    },
    {
      "epoch": 7.872909698996655,
      "grad_norm": 2.0870378017425537,
      "learning_rate": 2.1270903010033446e-05,
      "loss": 0.0344,
      "step": 11770
    },
    {
      "epoch": 7.879598662207358,
      "grad_norm": 0.9423313140869141,
      "learning_rate": 2.120401337792642e-05,
      "loss": 0.0449,
      "step": 11780
    },
    {
      "epoch": 7.88628762541806,
      "grad_norm": 0.9477387070655823,
      "learning_rate": 2.11371237458194e-05,
      "loss": 0.0361,
      "step": 11790
    },
    {
      "epoch": 7.892976588628763,
      "grad_norm": 0.6451069712638855,
      "learning_rate": 2.1070234113712374e-05,
      "loss": 0.053,
      "step": 11800
    },
    {
      "epoch": 7.8996655518394645,
      "grad_norm": 5.747945308685303,
      "learning_rate": 2.1003344481605352e-05,
      "loss": 0.0429,
      "step": 11810
    },
    {
      "epoch": 7.906354515050167,
      "grad_norm": 0.995347797870636,
      "learning_rate": 2.093645484949833e-05,
      "loss": 0.0343,
      "step": 11820
    },
    {
      "epoch": 7.913043478260869,
      "grad_norm": 2.1813809871673584,
      "learning_rate": 2.0869565217391303e-05,
      "loss": 0.0487,
      "step": 11830
    },
    {
      "epoch": 7.919732441471572,
      "grad_norm": 1.4194324016571045,
      "learning_rate": 2.080267558528428e-05,
      "loss": 0.0435,
      "step": 11840
    },
    {
      "epoch": 7.926421404682275,
      "grad_norm": 0.39779338240623474,
      "learning_rate": 2.073578595317726e-05,
      "loss": 0.0369,
      "step": 11850
    },
    {
      "epoch": 7.933110367892977,
      "grad_norm": 4.078008651733398,
      "learning_rate": 2.0668896321070234e-05,
      "loss": 0.0424,
      "step": 11860
    },
    {
      "epoch": 7.9397993311036785,
      "grad_norm": 1.5877165794372559,
      "learning_rate": 2.0602006688963213e-05,
      "loss": 0.0333,
      "step": 11870
    },
    {
      "epoch": 7.946488294314381,
      "grad_norm": 2.7322449684143066,
      "learning_rate": 2.0535117056856188e-05,
      "loss": 0.0421,
      "step": 11880
    },
    {
      "epoch": 7.953177257525084,
      "grad_norm": 2.233501434326172,
      "learning_rate": 2.0468227424749163e-05,
      "loss": 0.0452,
      "step": 11890
    },
    {
      "epoch": 7.959866220735786,
      "grad_norm": 1.474515676498413,
      "learning_rate": 2.040133779264214e-05,
      "loss": 0.0375,
      "step": 11900
    },
    {
      "epoch": 7.966555183946488,
      "grad_norm": 3.1244900226593018,
      "learning_rate": 2.033444816053512e-05,
      "loss": 0.0429,
      "step": 11910
    },
    {
      "epoch": 7.973244147157191,
      "grad_norm": 2.144104480743408,
      "learning_rate": 2.0267558528428095e-05,
      "loss": 0.0346,
      "step": 11920
    },
    {
      "epoch": 7.979933110367893,
      "grad_norm": 0.5135546922683716,
      "learning_rate": 2.020066889632107e-05,
      "loss": 0.0266,
      "step": 11930
    },
    {
      "epoch": 7.986622073578595,
      "grad_norm": 2.0495548248291016,
      "learning_rate": 2.0133779264214048e-05,
      "loss": 0.038,
      "step": 11940
    },
    {
      "epoch": 7.993311036789297,
      "grad_norm": 0.7567475438117981,
      "learning_rate": 2.0066889632107023e-05,
      "loss": 0.0334,
      "step": 11950
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.2615234851837158,
      "learning_rate": 2e-05,
      "loss": 0.0326,
      "step": 11960
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.08834142982959747,
      "eval_mse": 0.08834142982959747,
      "eval_runtime": 229.8467,
      "eval_samples_per_second": 13.009,
      "eval_steps_per_second": 1.627,
      "step": 11960
    },
    {
      "epoch": 8.006688963210703,
      "grad_norm": 3.9783406257629395,
      "learning_rate": 1.9933110367892977e-05,
      "loss": 0.0337,
      "step": 11970
    },
    {
      "epoch": 8.013377926421406,
      "grad_norm": 1.0547821521759033,
      "learning_rate": 1.9866220735785955e-05,
      "loss": 0.0265,
      "step": 11980
    },
    {
      "epoch": 8.020066889632107,
      "grad_norm": 3.261388063430786,
      "learning_rate": 1.979933110367893e-05,
      "loss": 0.028,
      "step": 11990
    },
    {
      "epoch": 8.02675585284281,
      "grad_norm": 1.504028558731079,
      "learning_rate": 1.973244147157191e-05,
      "loss": 0.029,
      "step": 12000
    },
    {
      "epoch": 8.033444816053512,
      "grad_norm": 3.6684327125549316,
      "learning_rate": 1.9665551839464884e-05,
      "loss": 0.0311,
      "step": 12010
    },
    {
      "epoch": 8.040133779264215,
      "grad_norm": 2.4501540660858154,
      "learning_rate": 1.959866220735786e-05,
      "loss": 0.0313,
      "step": 12020
    },
    {
      "epoch": 8.046822742474916,
      "grad_norm": 3.4440808296203613,
      "learning_rate": 1.9531772575250837e-05,
      "loss": 0.0382,
      "step": 12030
    },
    {
      "epoch": 8.053511705685619,
      "grad_norm": 0.9953986406326294,
      "learning_rate": 1.9464882943143815e-05,
      "loss": 0.0313,
      "step": 12040
    },
    {
      "epoch": 8.060200668896321,
      "grad_norm": 0.80042564868927,
      "learning_rate": 1.939799331103679e-05,
      "loss": 0.0269,
      "step": 12050
    },
    {
      "epoch": 8.066889632107024,
      "grad_norm": 1.4230451583862305,
      "learning_rate": 1.9331103678929765e-05,
      "loss": 0.0236,
      "step": 12060
    },
    {
      "epoch": 8.073578595317725,
      "grad_norm": 2.201822519302368,
      "learning_rate": 1.9264214046822744e-05,
      "loss": 0.0294,
      "step": 12070
    },
    {
      "epoch": 8.080267558528428,
      "grad_norm": 0.8343908786773682,
      "learning_rate": 1.919732441471572e-05,
      "loss": 0.0302,
      "step": 12080
    },
    {
      "epoch": 8.08695652173913,
      "grad_norm": 1.7819195985794067,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.0152,
      "step": 12090
    },
    {
      "epoch": 8.093645484949834,
      "grad_norm": 0.8807457685470581,
      "learning_rate": 1.9063545150501676e-05,
      "loss": 0.0176,
      "step": 12100
    },
    {
      "epoch": 8.100334448160535,
      "grad_norm": 0.37105417251586914,
      "learning_rate": 1.8996655518394647e-05,
      "loss": 0.0311,
      "step": 12110
    },
    {
      "epoch": 8.107023411371237,
      "grad_norm": 3.2414398193359375,
      "learning_rate": 1.8929765886287626e-05,
      "loss": 0.0311,
      "step": 12120
    },
    {
      "epoch": 8.11371237458194,
      "grad_norm": 3.936710834503174,
      "learning_rate": 1.8862876254180604e-05,
      "loss": 0.0267,
      "step": 12130
    },
    {
      "epoch": 8.120401337792643,
      "grad_norm": 1.1247128248214722,
      "learning_rate": 1.879598662207358e-05,
      "loss": 0.033,
      "step": 12140
    },
    {
      "epoch": 8.127090301003344,
      "grad_norm": 3.2280757427215576,
      "learning_rate": 1.8729096989966554e-05,
      "loss": 0.0286,
      "step": 12150
    },
    {
      "epoch": 8.133779264214047,
      "grad_norm": 2.556563138961792,
      "learning_rate": 1.8662207357859533e-05,
      "loss": 0.0294,
      "step": 12160
    },
    {
      "epoch": 8.14046822742475,
      "grad_norm": 0.8224531412124634,
      "learning_rate": 1.8595317725752508e-05,
      "loss": 0.0309,
      "step": 12170
    },
    {
      "epoch": 8.147157190635452,
      "grad_norm": 0.8672958016395569,
      "learning_rate": 1.8528428093645486e-05,
      "loss": 0.0211,
      "step": 12180
    },
    {
      "epoch": 8.153846153846153,
      "grad_norm": 1.4439815282821655,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.0296,
      "step": 12190
    },
    {
      "epoch": 8.160535117056856,
      "grad_norm": 0.7979484796524048,
      "learning_rate": 1.8394648829431436e-05,
      "loss": 0.0373,
      "step": 12200
    },
    {
      "epoch": 8.167224080267559,
      "grad_norm": 5.962237358093262,
      "learning_rate": 1.8327759197324415e-05,
      "loss": 0.0333,
      "step": 12210
    },
    {
      "epoch": 8.173913043478262,
      "grad_norm": 1.737164855003357,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.0225,
      "step": 12220
    },
    {
      "epoch": 8.180602006688963,
      "grad_norm": 2.052433729171753,
      "learning_rate": 1.8193979933110368e-05,
      "loss": 0.0321,
      "step": 12230
    },
    {
      "epoch": 8.187290969899665,
      "grad_norm": 1.0267506837844849,
      "learning_rate": 1.8127090301003347e-05,
      "loss": 0.0341,
      "step": 12240
    },
    {
      "epoch": 8.193979933110368,
      "grad_norm": 0.891290545463562,
      "learning_rate": 1.806020066889632e-05,
      "loss": 0.0251,
      "step": 12250
    },
    {
      "epoch": 8.200668896321071,
      "grad_norm": 1.1818610429763794,
      "learning_rate": 1.7993311036789297e-05,
      "loss": 0.0355,
      "step": 12260
    },
    {
      "epoch": 8.207357859531772,
      "grad_norm": 2.8808891773223877,
      "learning_rate": 1.7926421404682275e-05,
      "loss": 0.0363,
      "step": 12270
    },
    {
      "epoch": 8.214046822742475,
      "grad_norm": 1.9035134315490723,
      "learning_rate": 1.7859531772575253e-05,
      "loss": 0.0243,
      "step": 12280
    },
    {
      "epoch": 8.220735785953178,
      "grad_norm": 1.533340573310852,
      "learning_rate": 1.779264214046823e-05,
      "loss": 0.0434,
      "step": 12290
    },
    {
      "epoch": 8.22742474916388,
      "grad_norm": 1.0126433372497559,
      "learning_rate": 1.7725752508361204e-05,
      "loss": 0.0266,
      "step": 12300
    },
    {
      "epoch": 8.234113712374581,
      "grad_norm": 1.2792842388153076,
      "learning_rate": 1.7658862876254182e-05,
      "loss": 0.0268,
      "step": 12310
    },
    {
      "epoch": 8.240802675585284,
      "grad_norm": 0.51844722032547,
      "learning_rate": 1.7591973244147157e-05,
      "loss": 0.0324,
      "step": 12320
    },
    {
      "epoch": 8.247491638795987,
      "grad_norm": 1.5483566522598267,
      "learning_rate": 1.7525083612040135e-05,
      "loss": 0.0356,
      "step": 12330
    },
    {
      "epoch": 8.25418060200669,
      "grad_norm": 2.015435218811035,
      "learning_rate": 1.745819397993311e-05,
      "loss": 0.0305,
      "step": 12340
    },
    {
      "epoch": 8.26086956521739,
      "grad_norm": 1.5077990293502808,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.0258,
      "step": 12350
    },
    {
      "epoch": 8.267558528428093,
      "grad_norm": 2.064847946166992,
      "learning_rate": 1.7324414715719064e-05,
      "loss": 0.022,
      "step": 12360
    },
    {
      "epoch": 8.274247491638796,
      "grad_norm": 2.8898634910583496,
      "learning_rate": 1.7257525083612042e-05,
      "loss": 0.0386,
      "step": 12370
    },
    {
      "epoch": 8.280936454849499,
      "grad_norm": 1.5126811265945435,
      "learning_rate": 1.7190635451505017e-05,
      "loss": 0.0348,
      "step": 12380
    },
    {
      "epoch": 8.2876254180602,
      "grad_norm": 1.4772700071334839,
      "learning_rate": 1.7123745819397992e-05,
      "loss": 0.0414,
      "step": 12390
    },
    {
      "epoch": 8.294314381270903,
      "grad_norm": 1.3124395608901978,
      "learning_rate": 1.705685618729097e-05,
      "loss": 0.0363,
      "step": 12400
    },
    {
      "epoch": 8.301003344481606,
      "grad_norm": 0.8630233407020569,
      "learning_rate": 1.698996655518395e-05,
      "loss": 0.0281,
      "step": 12410
    },
    {
      "epoch": 8.307692307692308,
      "grad_norm": 1.1018539667129517,
      "learning_rate": 1.6923076923076924e-05,
      "loss": 0.0248,
      "step": 12420
    },
    {
      "epoch": 8.31438127090301,
      "grad_norm": 1.0768871307373047,
      "learning_rate": 1.68561872909699e-05,
      "loss": 0.0292,
      "step": 12430
    },
    {
      "epoch": 8.321070234113712,
      "grad_norm": 1.1558284759521484,
      "learning_rate": 1.6789297658862878e-05,
      "loss": 0.0289,
      "step": 12440
    },
    {
      "epoch": 8.327759197324415,
      "grad_norm": 2.288069248199463,
      "learning_rate": 1.6722408026755853e-05,
      "loss": 0.0237,
      "step": 12450
    },
    {
      "epoch": 8.334448160535118,
      "grad_norm": 0.7263951897621155,
      "learning_rate": 1.665551839464883e-05,
      "loss": 0.033,
      "step": 12460
    },
    {
      "epoch": 8.341137123745819,
      "grad_norm": 4.321417331695557,
      "learning_rate": 1.658862876254181e-05,
      "loss": 0.0406,
      "step": 12470
    },
    {
      "epoch": 8.347826086956522,
      "grad_norm": 0.8176807761192322,
      "learning_rate": 1.652173913043478e-05,
      "loss": 0.0172,
      "step": 12480
    },
    {
      "epoch": 8.354515050167224,
      "grad_norm": 0.8886902928352356,
      "learning_rate": 1.645484949832776e-05,
      "loss": 0.0246,
      "step": 12490
    },
    {
      "epoch": 8.361204013377927,
      "grad_norm": 0.4712463915348053,
      "learning_rate": 1.6387959866220738e-05,
      "loss": 0.0249,
      "step": 12500
    },
    {
      "epoch": 8.367892976588628,
      "grad_norm": 1.5403294563293457,
      "learning_rate": 1.6321070234113713e-05,
      "loss": 0.0295,
      "step": 12510
    },
    {
      "epoch": 8.37458193979933,
      "grad_norm": 0.9423514604568481,
      "learning_rate": 1.625418060200669e-05,
      "loss": 0.0314,
      "step": 12520
    },
    {
      "epoch": 8.381270903010034,
      "grad_norm": 1.1313178539276123,
      "learning_rate": 1.6187290969899666e-05,
      "loss": 0.0281,
      "step": 12530
    },
    {
      "epoch": 8.387959866220736,
      "grad_norm": 2.3116352558135986,
      "learning_rate": 1.612040133779264e-05,
      "loss": 0.027,
      "step": 12540
    },
    {
      "epoch": 8.394648829431437,
      "grad_norm": 0.6153112649917603,
      "learning_rate": 1.605351170568562e-05,
      "loss": 0.0295,
      "step": 12550
    },
    {
      "epoch": 8.40133779264214,
      "grad_norm": 0.900283932685852,
      "learning_rate": 1.59866220735786e-05,
      "loss": 0.0204,
      "step": 12560
    },
    {
      "epoch": 8.408026755852843,
      "grad_norm": 0.7899017333984375,
      "learning_rate": 1.591973244147157e-05,
      "loss": 0.0317,
      "step": 12570
    },
    {
      "epoch": 8.414715719063546,
      "grad_norm": 1.4934643507003784,
      "learning_rate": 1.585284280936455e-05,
      "loss": 0.0272,
      "step": 12580
    },
    {
      "epoch": 8.421404682274247,
      "grad_norm": 3.2169320583343506,
      "learning_rate": 1.5785953177257527e-05,
      "loss": 0.0331,
      "step": 12590
    },
    {
      "epoch": 8.42809364548495,
      "grad_norm": 0.30572786927223206,
      "learning_rate": 1.5719063545150502e-05,
      "loss": 0.0195,
      "step": 12600
    },
    {
      "epoch": 8.434782608695652,
      "grad_norm": 3.049567222595215,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.0368,
      "step": 12610
    },
    {
      "epoch": 8.441471571906355,
      "grad_norm": 0.841004490852356,
      "learning_rate": 1.5585284280936455e-05,
      "loss": 0.0302,
      "step": 12620
    },
    {
      "epoch": 8.448160535117056,
      "grad_norm": 1.2091842889785767,
      "learning_rate": 1.551839464882943e-05,
      "loss": 0.0294,
      "step": 12630
    },
    {
      "epoch": 8.454849498327759,
      "grad_norm": 0.7122257947921753,
      "learning_rate": 1.545150501672241e-05,
      "loss": 0.0175,
      "step": 12640
    },
    {
      "epoch": 8.461538461538462,
      "grad_norm": 2.2872982025146484,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.0301,
      "step": 12650
    },
    {
      "epoch": 8.468227424749164,
      "grad_norm": 2.8640732765197754,
      "learning_rate": 1.5317725752508362e-05,
      "loss": 0.0418,
      "step": 12660
    },
    {
      "epoch": 8.474916387959865,
      "grad_norm": 1.7518962621688843,
      "learning_rate": 1.5250836120401337e-05,
      "loss": 0.0354,
      "step": 12670
    },
    {
      "epoch": 8.481605351170568,
      "grad_norm": 1.138287901878357,
      "learning_rate": 1.5183946488294316e-05,
      "loss": 0.0299,
      "step": 12680
    },
    {
      "epoch": 8.488294314381271,
      "grad_norm": 3.845442533493042,
      "learning_rate": 1.5117056856187292e-05,
      "loss": 0.0221,
      "step": 12690
    },
    {
      "epoch": 8.494983277591974,
      "grad_norm": 0.7240458726882935,
      "learning_rate": 1.5050167224080269e-05,
      "loss": 0.0307,
      "step": 12700
    },
    {
      "epoch": 8.501672240802675,
      "grad_norm": 2.00895094871521,
      "learning_rate": 1.4983277591973244e-05,
      "loss": 0.0242,
      "step": 12710
    },
    {
      "epoch": 8.508361204013378,
      "grad_norm": 0.557636559009552,
      "learning_rate": 1.4916387959866221e-05,
      "loss": 0.0207,
      "step": 12720
    },
    {
      "epoch": 8.51505016722408,
      "grad_norm": 0.8012935519218445,
      "learning_rate": 1.4849498327759198e-05,
      "loss": 0.0241,
      "step": 12730
    },
    {
      "epoch": 8.521739130434783,
      "grad_norm": 0.6977702975273132,
      "learning_rate": 1.4782608695652176e-05,
      "loss": 0.0275,
      "step": 12740
    },
    {
      "epoch": 8.528428093645484,
      "grad_norm": 1.9219177961349487,
      "learning_rate": 1.4715719063545153e-05,
      "loss": 0.024,
      "step": 12750
    },
    {
      "epoch": 8.535117056856187,
      "grad_norm": 1.120877981185913,
      "learning_rate": 1.4648829431438128e-05,
      "loss": 0.0268,
      "step": 12760
    },
    {
      "epoch": 8.54180602006689,
      "grad_norm": 2.783856153488159,
      "learning_rate": 1.4581939799331104e-05,
      "loss": 0.0362,
      "step": 12770
    },
    {
      "epoch": 8.548494983277592,
      "grad_norm": 1.8937408924102783,
      "learning_rate": 1.4515050167224081e-05,
      "loss": 0.0329,
      "step": 12780
    },
    {
      "epoch": 8.555183946488294,
      "grad_norm": 1.783046007156372,
      "learning_rate": 1.4448160535117058e-05,
      "loss": 0.0372,
      "step": 12790
    },
    {
      "epoch": 8.561872909698996,
      "grad_norm": 1.2558856010437012,
      "learning_rate": 1.4381270903010033e-05,
      "loss": 0.0229,
      "step": 12800
    },
    {
      "epoch": 8.568561872909699,
      "grad_norm": 1.831705927848816,
      "learning_rate": 1.431438127090301e-05,
      "loss": 0.0224,
      "step": 12810
    },
    {
      "epoch": 8.575250836120402,
      "grad_norm": 2.1247737407684326,
      "learning_rate": 1.4247491638795986e-05,
      "loss": 0.0274,
      "step": 12820
    },
    {
      "epoch": 8.581939799331103,
      "grad_norm": 1.0554836988449097,
      "learning_rate": 1.4180602006688965e-05,
      "loss": 0.0335,
      "step": 12830
    },
    {
      "epoch": 8.588628762541806,
      "grad_norm": 1.7146085500717163,
      "learning_rate": 1.4113712374581942e-05,
      "loss": 0.0323,
      "step": 12840
    },
    {
      "epoch": 8.595317725752508,
      "grad_norm": 1.0601108074188232,
      "learning_rate": 1.4046822742474917e-05,
      "loss": 0.0253,
      "step": 12850
    },
    {
      "epoch": 8.602006688963211,
      "grad_norm": 0.7944843173027039,
      "learning_rate": 1.3979933110367893e-05,
      "loss": 0.0309,
      "step": 12860
    },
    {
      "epoch": 8.608695652173914,
      "grad_norm": 1.291366457939148,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.0222,
      "step": 12870
    },
    {
      "epoch": 8.615384615384615,
      "grad_norm": 2.837893009185791,
      "learning_rate": 1.3846153846153847e-05,
      "loss": 0.0344,
      "step": 12880
    },
    {
      "epoch": 8.622073578595318,
      "grad_norm": 2.289451837539673,
      "learning_rate": 1.3779264214046825e-05,
      "loss": 0.0201,
      "step": 12890
    },
    {
      "epoch": 8.62876254180602,
      "grad_norm": 1.1213974952697754,
      "learning_rate": 1.3712374581939799e-05,
      "loss": 0.0193,
      "step": 12900
    },
    {
      "epoch": 8.635451505016722,
      "grad_norm": 2.0777931213378906,
      "learning_rate": 1.3645484949832777e-05,
      "loss": 0.0217,
      "step": 12910
    },
    {
      "epoch": 8.642140468227424,
      "grad_norm": 1.627186894416809,
      "learning_rate": 1.3578595317725754e-05,
      "loss": 0.0298,
      "step": 12920
    },
    {
      "epoch": 8.648829431438127,
      "grad_norm": 1.691390037536621,
      "learning_rate": 1.351170568561873e-05,
      "loss": 0.0371,
      "step": 12930
    },
    {
      "epoch": 8.65551839464883,
      "grad_norm": 4.63961935043335,
      "learning_rate": 1.3444816053511705e-05,
      "loss": 0.0411,
      "step": 12940
    },
    {
      "epoch": 8.662207357859533,
      "grad_norm": 0.9765155911445618,
      "learning_rate": 1.3377926421404682e-05,
      "loss": 0.0354,
      "step": 12950
    },
    {
      "epoch": 8.668896321070234,
      "grad_norm": 1.394568920135498,
      "learning_rate": 1.3311036789297659e-05,
      "loss": 0.0327,
      "step": 12960
    },
    {
      "epoch": 8.675585284280936,
      "grad_norm": 0.6211327314376831,
      "learning_rate": 1.3244147157190637e-05,
      "loss": 0.0254,
      "step": 12970
    },
    {
      "epoch": 8.68227424749164,
      "grad_norm": 3.3042550086975098,
      "learning_rate": 1.3177257525083614e-05,
      "loss": 0.029,
      "step": 12980
    },
    {
      "epoch": 8.68896321070234,
      "grad_norm": 1.2794384956359863,
      "learning_rate": 1.3110367892976589e-05,
      "loss": 0.0408,
      "step": 12990
    },
    {
      "epoch": 8.695652173913043,
      "grad_norm": 1.3925954103469849,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.0363,
      "step": 13000
    },
    {
      "epoch": 8.702341137123746,
      "grad_norm": 2.9687554836273193,
      "learning_rate": 1.2976588628762542e-05,
      "loss": 0.0245,
      "step": 13010
    },
    {
      "epoch": 8.709030100334449,
      "grad_norm": 2.5190224647521973,
      "learning_rate": 1.290969899665552e-05,
      "loss": 0.0344,
      "step": 13020
    },
    {
      "epoch": 8.715719063545151,
      "grad_norm": 0.9365471005439758,
      "learning_rate": 1.2842809364548494e-05,
      "loss": 0.0263,
      "step": 13030
    },
    {
      "epoch": 8.722408026755852,
      "grad_norm": 1.4775091409683228,
      "learning_rate": 1.2775919732441471e-05,
      "loss": 0.0256,
      "step": 13040
    },
    {
      "epoch": 8.729096989966555,
      "grad_norm": 3.659803867340088,
      "learning_rate": 1.270903010033445e-05,
      "loss": 0.0328,
      "step": 13050
    },
    {
      "epoch": 8.735785953177258,
      "grad_norm": 1.5067414045333862,
      "learning_rate": 1.2642140468227426e-05,
      "loss": 0.0303,
      "step": 13060
    },
    {
      "epoch": 8.742474916387959,
      "grad_norm": 1.6895188093185425,
      "learning_rate": 1.2575250836120403e-05,
      "loss": 0.0248,
      "step": 13070
    },
    {
      "epoch": 8.749163879598662,
      "grad_norm": 2.7330615520477295,
      "learning_rate": 1.2508361204013378e-05,
      "loss": 0.0314,
      "step": 13080
    },
    {
      "epoch": 8.755852842809364,
      "grad_norm": 2.122676134109497,
      "learning_rate": 1.2441471571906355e-05,
      "loss": 0.0284,
      "step": 13090
    },
    {
      "epoch": 8.762541806020067,
      "grad_norm": 1.499622106552124,
      "learning_rate": 1.2374581939799331e-05,
      "loss": 0.0288,
      "step": 13100
    },
    {
      "epoch": 8.76923076923077,
      "grad_norm": 3.3748297691345215,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.0284,
      "step": 13110
    },
    {
      "epoch": 8.775919732441471,
      "grad_norm": 0.9968297481536865,
      "learning_rate": 1.2240802675585285e-05,
      "loss": 0.0279,
      "step": 13120
    },
    {
      "epoch": 8.782608695652174,
      "grad_norm": 0.6592386960983276,
      "learning_rate": 1.2173913043478261e-05,
      "loss": 0.0316,
      "step": 13130
    },
    {
      "epoch": 8.789297658862877,
      "grad_norm": 1.2029937505722046,
      "learning_rate": 1.2107023411371238e-05,
      "loss": 0.0244,
      "step": 13140
    },
    {
      "epoch": 8.79598662207358,
      "grad_norm": 0.8343057632446289,
      "learning_rate": 1.2040133779264215e-05,
      "loss": 0.0242,
      "step": 13150
    },
    {
      "epoch": 8.80267558528428,
      "grad_norm": 1.8353803157806396,
      "learning_rate": 1.197324414715719e-05,
      "loss": 0.0219,
      "step": 13160
    },
    {
      "epoch": 8.809364548494983,
      "grad_norm": 1.6544325351715088,
      "learning_rate": 1.1906354515050168e-05,
      "loss": 0.0272,
      "step": 13170
    },
    {
      "epoch": 8.816053511705686,
      "grad_norm": 0.7823971509933472,
      "learning_rate": 1.1839464882943143e-05,
      "loss": 0.0275,
      "step": 13180
    },
    {
      "epoch": 8.822742474916389,
      "grad_norm": 0.49826306104660034,
      "learning_rate": 1.177257525083612e-05,
      "loss": 0.0217,
      "step": 13190
    },
    {
      "epoch": 8.82943143812709,
      "grad_norm": 1.0228309631347656,
      "learning_rate": 1.1705685618729099e-05,
      "loss": 0.0241,
      "step": 13200
    },
    {
      "epoch": 8.836120401337793,
      "grad_norm": 2.224672317504883,
      "learning_rate": 1.1638795986622074e-05,
      "loss": 0.0317,
      "step": 13210
    },
    {
      "epoch": 8.842809364548495,
      "grad_norm": 2.9379260540008545,
      "learning_rate": 1.157190635451505e-05,
      "loss": 0.0318,
      "step": 13220
    },
    {
      "epoch": 8.849498327759198,
      "grad_norm": 2.134108066558838,
      "learning_rate": 1.1505016722408027e-05,
      "loss": 0.0195,
      "step": 13230
    },
    {
      "epoch": 8.856187290969899,
      "grad_norm": 3.203746795654297,
      "learning_rate": 1.1438127090301004e-05,
      "loss": 0.0285,
      "step": 13240
    },
    {
      "epoch": 8.862876254180602,
      "grad_norm": 0.8799943923950195,
      "learning_rate": 1.137123745819398e-05,
      "loss": 0.0369,
      "step": 13250
    },
    {
      "epoch": 8.869565217391305,
      "grad_norm": 0.89390629529953,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.0339,
      "step": 13260
    },
    {
      "epoch": 8.876254180602007,
      "grad_norm": 2.9978513717651367,
      "learning_rate": 1.1237458193979934e-05,
      "loss": 0.0361,
      "step": 13270
    },
    {
      "epoch": 8.882943143812708,
      "grad_norm": 1.2196823358535767,
      "learning_rate": 1.117056856187291e-05,
      "loss": 0.0295,
      "step": 13280
    },
    {
      "epoch": 8.889632107023411,
      "grad_norm": 0.9094672799110413,
      "learning_rate": 1.1103678929765887e-05,
      "loss": 0.0254,
      "step": 13290
    },
    {
      "epoch": 8.896321070234114,
      "grad_norm": 2.8375725746154785,
      "learning_rate": 1.1036789297658862e-05,
      "loss": 0.0246,
      "step": 13300
    },
    {
      "epoch": 8.903010033444817,
      "grad_norm": 2.510275363922119,
      "learning_rate": 1.0969899665551841e-05,
      "loss": 0.0271,
      "step": 13310
    },
    {
      "epoch": 8.909698996655518,
      "grad_norm": 1.341292381286621,
      "learning_rate": 1.0903010033444816e-05,
      "loss": 0.02,
      "step": 13320
    },
    {
      "epoch": 8.91638795986622,
      "grad_norm": 0.7900250554084778,
      "learning_rate": 1.0836120401337793e-05,
      "loss": 0.0255,
      "step": 13330
    },
    {
      "epoch": 8.923076923076923,
      "grad_norm": 2.9481964111328125,
      "learning_rate": 1.0769230769230771e-05,
      "loss": 0.0329,
      "step": 13340
    },
    {
      "epoch": 8.929765886287626,
      "grad_norm": 1.1457481384277344,
      "learning_rate": 1.0702341137123746e-05,
      "loss": 0.0283,
      "step": 13350
    },
    {
      "epoch": 8.936454849498327,
      "grad_norm": 1.039831280708313,
      "learning_rate": 1.0635451505016723e-05,
      "loss": 0.0258,
      "step": 13360
    },
    {
      "epoch": 8.94314381270903,
      "grad_norm": 0.9117240309715271,
      "learning_rate": 1.05685618729097e-05,
      "loss": 0.0353,
      "step": 13370
    },
    {
      "epoch": 8.949832775919733,
      "grad_norm": 0.45773544907569885,
      "learning_rate": 1.0501672240802676e-05,
      "loss": 0.0285,
      "step": 13380
    },
    {
      "epoch": 8.956521739130435,
      "grad_norm": 0.676830530166626,
      "learning_rate": 1.0434782608695651e-05,
      "loss": 0.0208,
      "step": 13390
    },
    {
      "epoch": 8.963210702341136,
      "grad_norm": 1.4347857236862183,
      "learning_rate": 1.036789297658863e-05,
      "loss": 0.0281,
      "step": 13400
    },
    {
      "epoch": 8.96989966555184,
      "grad_norm": 0.6630375981330872,
      "learning_rate": 1.0301003344481606e-05,
      "loss": 0.0158,
      "step": 13410
    },
    {
      "epoch": 8.976588628762542,
      "grad_norm": 0.9883447289466858,
      "learning_rate": 1.0234113712374581e-05,
      "loss": 0.0208,
      "step": 13420
    },
    {
      "epoch": 8.983277591973245,
      "grad_norm": 1.1196333169937134,
      "learning_rate": 1.016722408026756e-05,
      "loss": 0.0248,
      "step": 13430
    },
    {
      "epoch": 8.989966555183946,
      "grad_norm": 1.1069833040237427,
      "learning_rate": 1.0100334448160535e-05,
      "loss": 0.0318,
      "step": 13440
    },
    {
      "epoch": 8.996655518394649,
      "grad_norm": 1.9332762956619263,
      "learning_rate": 1.0033444816053512e-05,
      "loss": 0.0311,
      "step": 13450
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.0904405489563942,
      "eval_mse": 0.09044055640697479,
      "eval_runtime": 224.6401,
      "eval_samples_per_second": 13.31,
      "eval_steps_per_second": 1.665,
      "step": 13455
    },
    {
      "epoch": 9.003344481605351,
      "grad_norm": 0.9701605439186096,
      "learning_rate": 9.966555183946488e-06,
      "loss": 0.0237,
      "step": 13460
    },
    {
      "epoch": 9.010033444816054,
      "grad_norm": 2.653738260269165,
      "learning_rate": 9.899665551839465e-06,
      "loss": 0.0257,
      "step": 13470
    },
    {
      "epoch": 9.016722408026755,
      "grad_norm": 0.45054054260253906,
      "learning_rate": 9.832775919732442e-06,
      "loss": 0.0152,
      "step": 13480
    },
    {
      "epoch": 9.023411371237458,
      "grad_norm": 2.150256395339966,
      "learning_rate": 9.765886287625419e-06,
      "loss": 0.0194,
      "step": 13490
    },
    {
      "epoch": 9.03010033444816,
      "grad_norm": 1.6995000839233398,
      "learning_rate": 9.698996655518395e-06,
      "loss": 0.0217,
      "step": 13500
    },
    {
      "epoch": 9.036789297658864,
      "grad_norm": 0.4809478223323822,
      "learning_rate": 9.632107023411372e-06,
      "loss": 0.0187,
      "step": 13510
    },
    {
      "epoch": 9.043478260869565,
      "grad_norm": 1.1911168098449707,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.0142,
      "step": 13520
    },
    {
      "epoch": 9.050167224080267,
      "grad_norm": 2.6061320304870605,
      "learning_rate": 9.498327759197324e-06,
      "loss": 0.0227,
      "step": 13530
    },
    {
      "epoch": 9.05685618729097,
      "grad_norm": 0.42797496914863586,
      "learning_rate": 9.431438127090302e-06,
      "loss": 0.0195,
      "step": 13540
    },
    {
      "epoch": 9.063545150501673,
      "grad_norm": 1.5085166692733765,
      "learning_rate": 9.364548494983277e-06,
      "loss": 0.0269,
      "step": 13550
    },
    {
      "epoch": 9.070234113712374,
      "grad_norm": 0.48043566942214966,
      "learning_rate": 9.297658862876254e-06,
      "loss": 0.0165,
      "step": 13560
    },
    {
      "epoch": 9.076923076923077,
      "grad_norm": 1.4516146183013916,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.0162,
      "step": 13570
    },
    {
      "epoch": 9.08361204013378,
      "grad_norm": 0.9354628920555115,
      "learning_rate": 9.163879598662207e-06,
      "loss": 0.0185,
      "step": 13580
    },
    {
      "epoch": 9.090301003344482,
      "grad_norm": 0.7186703681945801,
      "learning_rate": 9.096989966555184e-06,
      "loss": 0.0164,
      "step": 13590
    },
    {
      "epoch": 9.096989966555183,
      "grad_norm": 0.677481472492218,
      "learning_rate": 9.03010033444816e-06,
      "loss": 0.0128,
      "step": 13600
    },
    {
      "epoch": 9.103678929765886,
      "grad_norm": 1.1981468200683594,
      "learning_rate": 8.963210702341138e-06,
      "loss": 0.0168,
      "step": 13610
    },
    {
      "epoch": 9.110367892976589,
      "grad_norm": 1.0445246696472168,
      "learning_rate": 8.896321070234114e-06,
      "loss": 0.0154,
      "step": 13620
    },
    {
      "epoch": 9.117056856187292,
      "grad_norm": 2.4534506797790527,
      "learning_rate": 8.829431438127091e-06,
      "loss": 0.0179,
      "step": 13630
    },
    {
      "epoch": 9.123745819397993,
      "grad_norm": 1.3992637395858765,
      "learning_rate": 8.762541806020068e-06,
      "loss": 0.0122,
      "step": 13640
    },
    {
      "epoch": 9.130434782608695,
      "grad_norm": 0.882490873336792,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.0178,
      "step": 13650
    },
    {
      "epoch": 9.137123745819398,
      "grad_norm": 2.6310842037200928,
      "learning_rate": 8.628762541806021e-06,
      "loss": 0.0206,
      "step": 13660
    },
    {
      "epoch": 9.143812709030101,
      "grad_norm": 2.426238536834717,
      "learning_rate": 8.561872909698996e-06,
      "loss": 0.0215,
      "step": 13670
    },
    {
      "epoch": 9.150501672240802,
      "grad_norm": 1.248909831047058,
      "learning_rate": 8.494983277591975e-06,
      "loss": 0.017,
      "step": 13680
    },
    {
      "epoch": 9.157190635451505,
      "grad_norm": 0.7255083322525024,
      "learning_rate": 8.42809364548495e-06,
      "loss": 0.0225,
      "step": 13690
    },
    {
      "epoch": 9.163879598662207,
      "grad_norm": 1.6974523067474365,
      "learning_rate": 8.361204013377926e-06,
      "loss": 0.0176,
      "step": 13700
    },
    {
      "epoch": 9.17056856187291,
      "grad_norm": 1.3242517709732056,
      "learning_rate": 8.294314381270905e-06,
      "loss": 0.0248,
      "step": 13710
    },
    {
      "epoch": 9.177257525083611,
      "grad_norm": 1.7846802473068237,
      "learning_rate": 8.22742474916388e-06,
      "loss": 0.0199,
      "step": 13720
    },
    {
      "epoch": 9.183946488294314,
      "grad_norm": 1.2862492799758911,
      "learning_rate": 8.160535117056857e-06,
      "loss": 0.0229,
      "step": 13730
    },
    {
      "epoch": 9.190635451505017,
      "grad_norm": 2.379707098007202,
      "learning_rate": 8.093645484949833e-06,
      "loss": 0.0218,
      "step": 13740
    },
    {
      "epoch": 9.19732441471572,
      "grad_norm": 1.0562094449996948,
      "learning_rate": 8.02675585284281e-06,
      "loss": 0.0169,
      "step": 13750
    },
    {
      "epoch": 9.20401337792642,
      "grad_norm": 2.632425308227539,
      "learning_rate": 7.959866220735785e-06,
      "loss": 0.0288,
      "step": 13760
    },
    {
      "epoch": 9.210702341137123,
      "grad_norm": 0.6736680865287781,
      "learning_rate": 7.892976588628763e-06,
      "loss": 0.0157,
      "step": 13770
    },
    {
      "epoch": 9.217391304347826,
      "grad_norm": 2.769725799560547,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.0193,
      "step": 13780
    },
    {
      "epoch": 9.224080267558529,
      "grad_norm": 1.3798658847808838,
      "learning_rate": 7.759197324414715e-06,
      "loss": 0.0335,
      "step": 13790
    },
    {
      "epoch": 9.23076923076923,
      "grad_norm": 0.8057318329811096,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.0226,
      "step": 13800
    },
    {
      "epoch": 9.237458193979933,
      "grad_norm": 1.60343337059021,
      "learning_rate": 7.625418060200669e-06,
      "loss": 0.0213,
      "step": 13810
    },
    {
      "epoch": 9.244147157190636,
      "grad_norm": 0.9643651247024536,
      "learning_rate": 7.558528428093646e-06,
      "loss": 0.0201,
      "step": 13820
    },
    {
      "epoch": 9.250836120401338,
      "grad_norm": 1.777146577835083,
      "learning_rate": 7.491638795986622e-06,
      "loss": 0.0198,
      "step": 13830
    },
    {
      "epoch": 9.25752508361204,
      "grad_norm": 0.708709180355072,
      "learning_rate": 7.424749163879599e-06,
      "loss": 0.0134,
      "step": 13840
    },
    {
      "epoch": 9.264214046822742,
      "grad_norm": 0.22371377050876617,
      "learning_rate": 7.357859531772576e-06,
      "loss": 0.0186,
      "step": 13850
    },
    {
      "epoch": 9.270903010033445,
      "grad_norm": 3.0788733959198,
      "learning_rate": 7.290969899665552e-06,
      "loss": 0.0178,
      "step": 13860
    },
    {
      "epoch": 9.277591973244148,
      "grad_norm": 1.888277530670166,
      "learning_rate": 7.224080267558529e-06,
      "loss": 0.0184,
      "step": 13870
    },
    {
      "epoch": 9.284280936454849,
      "grad_norm": 1.3582139015197754,
      "learning_rate": 7.157190635451505e-06,
      "loss": 0.026,
      "step": 13880
    },
    {
      "epoch": 9.290969899665551,
      "grad_norm": 1.5523234605789185,
      "learning_rate": 7.090301003344482e-06,
      "loss": 0.0173,
      "step": 13890
    },
    {
      "epoch": 9.297658862876254,
      "grad_norm": 0.42221876978874207,
      "learning_rate": 7.023411371237458e-06,
      "loss": 0.023,
      "step": 13900
    },
    {
      "epoch": 9.304347826086957,
      "grad_norm": 0.5811980962753296,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.0175,
      "step": 13910
    },
    {
      "epoch": 9.31103678929766,
      "grad_norm": 1.5026706457138062,
      "learning_rate": 6.889632107023413e-06,
      "loss": 0.0126,
      "step": 13920
    },
    {
      "epoch": 9.31772575250836,
      "grad_norm": 0.8601313233375549,
      "learning_rate": 6.8227424749163885e-06,
      "loss": 0.0266,
      "step": 13930
    },
    {
      "epoch": 9.324414715719064,
      "grad_norm": 0.26971110701560974,
      "learning_rate": 6.755852842809365e-06,
      "loss": 0.0122,
      "step": 13940
    },
    {
      "epoch": 9.331103678929766,
      "grad_norm": 0.6851016879081726,
      "learning_rate": 6.688963210702341e-06,
      "loss": 0.0228,
      "step": 13950
    },
    {
      "epoch": 9.337792642140467,
      "grad_norm": 1.128944993019104,
      "learning_rate": 6.622073578595319e-06,
      "loss": 0.0166,
      "step": 13960
    },
    {
      "epoch": 9.34448160535117,
      "grad_norm": 0.764406144618988,
      "learning_rate": 6.5551839464882945e-06,
      "loss": 0.0134,
      "step": 13970
    },
    {
      "epoch": 9.351170568561873,
      "grad_norm": 1.1070597171783447,
      "learning_rate": 6.488294314381271e-06,
      "loss": 0.0223,
      "step": 13980
    },
    {
      "epoch": 9.357859531772576,
      "grad_norm": 2.363032102584839,
      "learning_rate": 6.421404682274247e-06,
      "loss": 0.0163,
      "step": 13990
    },
    {
      "epoch": 9.364548494983278,
      "grad_norm": 2.318466901779175,
      "learning_rate": 6.354515050167225e-06,
      "loss": 0.0213,
      "step": 14000
    },
    {
      "epoch": 9.37123745819398,
      "grad_norm": 2.3102550506591797,
      "learning_rate": 6.287625418060201e-06,
      "loss": 0.0194,
      "step": 14010
    },
    {
      "epoch": 9.377926421404682,
      "grad_norm": 0.9690243601799011,
      "learning_rate": 6.220735785953177e-06,
      "loss": 0.0186,
      "step": 14020
    },
    {
      "epoch": 9.384615384615385,
      "grad_norm": 4.312711238861084,
      "learning_rate": 6.153846153846155e-06,
      "loss": 0.021,
      "step": 14030
    },
    {
      "epoch": 9.391304347826088,
      "grad_norm": 1.6641044616699219,
      "learning_rate": 6.086956521739131e-06,
      "loss": 0.0266,
      "step": 14040
    },
    {
      "epoch": 9.397993311036789,
      "grad_norm": 0.8408416509628296,
      "learning_rate": 6.0200668896321075e-06,
      "loss": 0.0162,
      "step": 14050
    },
    {
      "epoch": 9.404682274247492,
      "grad_norm": 0.6300095319747925,
      "learning_rate": 5.953177257525084e-06,
      "loss": 0.0135,
      "step": 14060
    },
    {
      "epoch": 9.411371237458194,
      "grad_norm": 1.318575143814087,
      "learning_rate": 5.88628762541806e-06,
      "loss": 0.0159,
      "step": 14070
    },
    {
      "epoch": 9.418060200668897,
      "grad_norm": 2.5081329345703125,
      "learning_rate": 5.819397993311037e-06,
      "loss": 0.0137,
      "step": 14080
    },
    {
      "epoch": 9.424749163879598,
      "grad_norm": 2.2743217945098877,
      "learning_rate": 5.7525083612040135e-06,
      "loss": 0.0195,
      "step": 14090
    },
    {
      "epoch": 9.431438127090301,
      "grad_norm": 0.530133068561554,
      "learning_rate": 5.68561872909699e-06,
      "loss": 0.0239,
      "step": 14100
    },
    {
      "epoch": 9.438127090301004,
      "grad_norm": 0.6645513772964478,
      "learning_rate": 5.618729096989967e-06,
      "loss": 0.0149,
      "step": 14110
    },
    {
      "epoch": 9.444816053511706,
      "grad_norm": 0.9330313801765442,
      "learning_rate": 5.551839464882944e-06,
      "loss": 0.0256,
      "step": 14120
    },
    {
      "epoch": 9.451505016722408,
      "grad_norm": 0.29505959153175354,
      "learning_rate": 5.4849498327759204e-06,
      "loss": 0.0155,
      "step": 14130
    },
    {
      "epoch": 9.45819397993311,
      "grad_norm": 0.9493227601051331,
      "learning_rate": 5.418060200668896e-06,
      "loss": 0.0184,
      "step": 14140
    },
    {
      "epoch": 9.464882943143813,
      "grad_norm": 2.8090169429779053,
      "learning_rate": 5.351170568561873e-06,
      "loss": 0.0114,
      "step": 14150
    },
    {
      "epoch": 9.471571906354516,
      "grad_norm": 0.5139400362968445,
      "learning_rate": 5.28428093645485e-06,
      "loss": 0.0172,
      "step": 14160
    },
    {
      "epoch": 9.478260869565217,
      "grad_norm": 0.5248110294342041,
      "learning_rate": 5.217391304347826e-06,
      "loss": 0.0149,
      "step": 14170
    },
    {
      "epoch": 9.48494983277592,
      "grad_norm": 1.211573600769043,
      "learning_rate": 5.150501672240803e-06,
      "loss": 0.0204,
      "step": 14180
    },
    {
      "epoch": 9.491638795986622,
      "grad_norm": 1.8907419443130493,
      "learning_rate": 5.08361204013378e-06,
      "loss": 0.0155,
      "step": 14190
    },
    {
      "epoch": 9.498327759197325,
      "grad_norm": 1.7588986158370972,
      "learning_rate": 5.016722408026756e-06,
      "loss": 0.0156,
      "step": 14200
    },
    {
      "epoch": 9.505016722408026,
      "grad_norm": 0.8350905776023865,
      "learning_rate": 4.9498327759197325e-06,
      "loss": 0.0203,
      "step": 14210
    },
    {
      "epoch": 9.511705685618729,
      "grad_norm": 0.8791153430938721,
      "learning_rate": 4.882943143812709e-06,
      "loss": 0.0239,
      "step": 14220
    },
    {
      "epoch": 9.518394648829432,
      "grad_norm": 3.2824668884277344,
      "learning_rate": 4.816053511705686e-06,
      "loss": 0.0278,
      "step": 14230
    },
    {
      "epoch": 9.525083612040135,
      "grad_norm": 0.2794869542121887,
      "learning_rate": 4.749163879598662e-06,
      "loss": 0.0236,
      "step": 14240
    },
    {
      "epoch": 9.531772575250836,
      "grad_norm": 0.37374019622802734,
      "learning_rate": 4.682274247491639e-06,
      "loss": 0.0223,
      "step": 14250
    },
    {
      "epoch": 9.538461538461538,
      "grad_norm": 0.9982376098632812,
      "learning_rate": 4.615384615384616e-06,
      "loss": 0.018,
      "step": 14260
    },
    {
      "epoch": 9.545150501672241,
      "grad_norm": 0.356210321187973,
      "learning_rate": 4.548494983277592e-06,
      "loss": 0.0162,
      "step": 14270
    },
    {
      "epoch": 9.551839464882944,
      "grad_norm": 0.5105440616607666,
      "learning_rate": 4.481605351170569e-06,
      "loss": 0.0252,
      "step": 14280
    },
    {
      "epoch": 9.558528428093645,
      "grad_norm": 2.1598117351531982,
      "learning_rate": 4.4147157190635455e-06,
      "loss": 0.0177,
      "step": 14290
    },
    {
      "epoch": 9.565217391304348,
      "grad_norm": 1.0519219636917114,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.016,
      "step": 14300
    },
    {
      "epoch": 9.57190635451505,
      "grad_norm": 0.4185671806335449,
      "learning_rate": 4.280936454849498e-06,
      "loss": 0.0271,
      "step": 14310
    },
    {
      "epoch": 9.578595317725753,
      "grad_norm": 1.2913966178894043,
      "learning_rate": 4.214046822742475e-06,
      "loss": 0.0227,
      "step": 14320
    },
    {
      "epoch": 9.585284280936454,
      "grad_norm": 2.225336790084839,
      "learning_rate": 4.147157190635452e-06,
      "loss": 0.0174,
      "step": 14330
    },
    {
      "epoch": 9.591973244147157,
      "grad_norm": 1.5562801361083984,
      "learning_rate": 4.080267558528428e-06,
      "loss": 0.0201,
      "step": 14340
    },
    {
      "epoch": 9.59866220735786,
      "grad_norm": 1.298086404800415,
      "learning_rate": 4.013377926421405e-06,
      "loss": 0.0189,
      "step": 14350
    },
    {
      "epoch": 9.605351170568563,
      "grad_norm": 0.8389520049095154,
      "learning_rate": 3.946488294314382e-06,
      "loss": 0.0206,
      "step": 14360
    },
    {
      "epoch": 9.612040133779264,
      "grad_norm": 1.0966200828552246,
      "learning_rate": 3.879598662207358e-06,
      "loss": 0.0198,
      "step": 14370
    },
    {
      "epoch": 9.618729096989966,
      "grad_norm": 1.133457899093628,
      "learning_rate": 3.8127090301003343e-06,
      "loss": 0.0109,
      "step": 14380
    },
    {
      "epoch": 9.62541806020067,
      "grad_norm": 1.0652483701705933,
      "learning_rate": 3.745819397993311e-06,
      "loss": 0.019,
      "step": 14390
    },
    {
      "epoch": 9.632107023411372,
      "grad_norm": 1.9861187934875488,
      "learning_rate": 3.678929765886288e-06,
      "loss": 0.0271,
      "step": 14400
    },
    {
      "epoch": 9.638795986622073,
      "grad_norm": 2.377269744873047,
      "learning_rate": 3.6120401337792645e-06,
      "loss": 0.021,
      "step": 14410
    },
    {
      "epoch": 9.645484949832776,
      "grad_norm": 1.7457759380340576,
      "learning_rate": 3.545150501672241e-06,
      "loss": 0.0234,
      "step": 14420
    },
    {
      "epoch": 9.652173913043478,
      "grad_norm": 4.8636884689331055,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.0197,
      "step": 14430
    },
    {
      "epoch": 9.658862876254181,
      "grad_norm": 0.41581761837005615,
      "learning_rate": 3.4113712374581942e-06,
      "loss": 0.0149,
      "step": 14440
    },
    {
      "epoch": 9.665551839464882,
      "grad_norm": 0.6414255499839783,
      "learning_rate": 3.3444816053511705e-06,
      "loss": 0.0232,
      "step": 14450
    },
    {
      "epoch": 9.672240802675585,
      "grad_norm": 0.9758875966072083,
      "learning_rate": 3.2775919732441473e-06,
      "loss": 0.0246,
      "step": 14460
    },
    {
      "epoch": 9.678929765886288,
      "grad_norm": 1.265362024307251,
      "learning_rate": 3.2107023411371236e-06,
      "loss": 0.0137,
      "step": 14470
    },
    {
      "epoch": 9.68561872909699,
      "grad_norm": 1.263174295425415,
      "learning_rate": 3.1438127090301007e-06,
      "loss": 0.0156,
      "step": 14480
    },
    {
      "epoch": 9.692307692307692,
      "grad_norm": 2.886913776397705,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 0.0175,
      "step": 14490
    },
    {
      "epoch": 9.698996655518394,
      "grad_norm": 0.6035112142562866,
      "learning_rate": 3.0100334448160537e-06,
      "loss": 0.0172,
      "step": 14500
    },
    {
      "epoch": 9.705685618729097,
      "grad_norm": 1.9714527130126953,
      "learning_rate": 2.94314381270903e-06,
      "loss": 0.0191,
      "step": 14510
    },
    {
      "epoch": 9.7123745819398,
      "grad_norm": 0.6139861345291138,
      "learning_rate": 2.8762541806020068e-06,
      "loss": 0.011,
      "step": 14520
    },
    {
      "epoch": 9.719063545150501,
      "grad_norm": 1.1653517484664917,
      "learning_rate": 2.8093645484949835e-06,
      "loss": 0.0269,
      "step": 14530
    },
    {
      "epoch": 9.725752508361204,
      "grad_norm": 1.1893435716629028,
      "learning_rate": 2.7424749163879602e-06,
      "loss": 0.0121,
      "step": 14540
    },
    {
      "epoch": 9.732441471571907,
      "grad_norm": 0.5057249665260315,
      "learning_rate": 2.6755852842809365e-06,
      "loss": 0.03,
      "step": 14550
    },
    {
      "epoch": 9.73913043478261,
      "grad_norm": 0.7675619125366211,
      "learning_rate": 2.608695652173913e-06,
      "loss": 0.0228,
      "step": 14560
    },
    {
      "epoch": 9.74581939799331,
      "grad_norm": 0.9619132876396179,
      "learning_rate": 2.54180602006689e-06,
      "loss": 0.0199,
      "step": 14570
    },
    {
      "epoch": 9.752508361204013,
      "grad_norm": 1.727118968963623,
      "learning_rate": 2.4749163879598663e-06,
      "loss": 0.0255,
      "step": 14580
    },
    {
      "epoch": 9.759197324414716,
      "grad_norm": 0.8463283777236938,
      "learning_rate": 2.408026755852843e-06,
      "loss": 0.0178,
      "step": 14590
    },
    {
      "epoch": 9.765886287625419,
      "grad_norm": 1.2253174781799316,
      "learning_rate": 2.3411371237458193e-06,
      "loss": 0.0201,
      "step": 14600
    },
    {
      "epoch": 9.77257525083612,
      "grad_norm": 0.49370265007019043,
      "learning_rate": 2.274247491638796e-06,
      "loss": 0.0218,
      "step": 14610
    },
    {
      "epoch": 9.779264214046822,
      "grad_norm": 1.9811979532241821,
      "learning_rate": 2.2073578595317727e-06,
      "loss": 0.0242,
      "step": 14620
    },
    {
      "epoch": 9.785953177257525,
      "grad_norm": 0.901732325553894,
      "learning_rate": 2.140468227424749e-06,
      "loss": 0.0163,
      "step": 14630
    },
    {
      "epoch": 9.792642140468228,
      "grad_norm": 1.5021089315414429,
      "learning_rate": 2.073578595317726e-06,
      "loss": 0.0171,
      "step": 14640
    },
    {
      "epoch": 9.799331103678929,
      "grad_norm": 0.9714933037757874,
      "learning_rate": 2.0066889632107025e-06,
      "loss": 0.0175,
      "step": 14650
    },
    {
      "epoch": 9.806020066889632,
      "grad_norm": 2.403262138366699,
      "learning_rate": 1.939799331103679e-06,
      "loss": 0.0261,
      "step": 14660
    },
    {
      "epoch": 9.812709030100335,
      "grad_norm": 0.4679270386695862,
      "learning_rate": 1.8729096989966555e-06,
      "loss": 0.0148,
      "step": 14670
    },
    {
      "epoch": 9.819397993311037,
      "grad_norm": 0.800066351890564,
      "learning_rate": 1.8060200668896322e-06,
      "loss": 0.0122,
      "step": 14680
    },
    {
      "epoch": 9.826086956521738,
      "grad_norm": 1.5774024724960327,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.013,
      "step": 14690
    },
    {
      "epoch": 9.832775919732441,
      "grad_norm": 0.4182083308696747,
      "learning_rate": 1.6722408026755853e-06,
      "loss": 0.017,
      "step": 14700
    },
    {
      "epoch": 9.839464882943144,
      "grad_norm": 2.029318332672119,
      "learning_rate": 1.6053511705685618e-06,
      "loss": 0.0318,
      "step": 14710
    },
    {
      "epoch": 9.846153846153847,
      "grad_norm": 0.7166652083396912,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 0.015,
      "step": 14720
    },
    {
      "epoch": 9.852842809364548,
      "grad_norm": 1.4269553422927856,
      "learning_rate": 1.471571906354515e-06,
      "loss": 0.0156,
      "step": 14730
    },
    {
      "epoch": 9.85953177257525,
      "grad_norm": 2.2864603996276855,
      "learning_rate": 1.4046822742474917e-06,
      "loss": 0.0179,
      "step": 14740
    },
    {
      "epoch": 9.866220735785953,
      "grad_norm": 0.5288352370262146,
      "learning_rate": 1.3377926421404683e-06,
      "loss": 0.0147,
      "step": 14750
    },
    {
      "epoch": 9.872909698996656,
      "grad_norm": 2.4506185054779053,
      "learning_rate": 1.270903010033445e-06,
      "loss": 0.0146,
      "step": 14760
    },
    {
      "epoch": 9.879598662207357,
      "grad_norm": 0.7271729111671448,
      "learning_rate": 1.2040133779264215e-06,
      "loss": 0.0187,
      "step": 14770
    },
    {
      "epoch": 9.88628762541806,
      "grad_norm": 2.37717342376709,
      "learning_rate": 1.137123745819398e-06,
      "loss": 0.0194,
      "step": 14780
    },
    {
      "epoch": 9.892976588628763,
      "grad_norm": 1.2071961164474487,
      "learning_rate": 1.0702341137123745e-06,
      "loss": 0.0124,
      "step": 14790
    },
    {
      "epoch": 9.899665551839465,
      "grad_norm": 0.8646990656852722,
      "learning_rate": 1.0033444816053512e-06,
      "loss": 0.026,
      "step": 14800
    },
    {
      "epoch": 9.906354515050166,
      "grad_norm": 1.959161400794983,
      "learning_rate": 9.364548494983278e-07,
      "loss": 0.0171,
      "step": 14810
    },
    {
      "epoch": 9.91304347826087,
      "grad_norm": 0.8739883303642273,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.0146,
      "step": 14820
    },
    {
      "epoch": 9.919732441471572,
      "grad_norm": 1.5968456268310547,
      "learning_rate": 8.026755852842809e-07,
      "loss": 0.0176,
      "step": 14830
    },
    {
      "epoch": 9.926421404682275,
      "grad_norm": 0.7556308507919312,
      "learning_rate": 7.357859531772575e-07,
      "loss": 0.0199,
      "step": 14840
    },
    {
      "epoch": 9.933110367892976,
      "grad_norm": 1.8411972522735596,
      "learning_rate": 6.688963210702341e-07,
      "loss": 0.0185,
      "step": 14850
    },
    {
      "epoch": 9.939799331103679,
      "grad_norm": 2.0228002071380615,
      "learning_rate": 6.020066889632107e-07,
      "loss": 0.0174,
      "step": 14860
    },
    {
      "epoch": 9.946488294314381,
      "grad_norm": 0.7559150457382202,
      "learning_rate": 5.351170568561873e-07,
      "loss": 0.0239,
      "step": 14870
    },
    {
      "epoch": 9.953177257525084,
      "grad_norm": 1.5705734491348267,
      "learning_rate": 4.682274247491639e-07,
      "loss": 0.0191,
      "step": 14880
    },
    {
      "epoch": 9.959866220735787,
      "grad_norm": 0.48421409726142883,
      "learning_rate": 4.0133779264214045e-07,
      "loss": 0.0183,
      "step": 14890
    },
    {
      "epoch": 9.966555183946488,
      "grad_norm": 1.3886953592300415,
      "learning_rate": 3.3444816053511706e-07,
      "loss": 0.0226,
      "step": 14900
    },
    {
      "epoch": 9.97324414715719,
      "grad_norm": 0.9066689610481262,
      "learning_rate": 2.6755852842809363e-07,
      "loss": 0.0131,
      "step": 14910
    },
    {
      "epoch": 9.979933110367893,
      "grad_norm": 2.4503183364868164,
      "learning_rate": 2.0066889632107022e-07,
      "loss": 0.0182,
      "step": 14920
    },
    {
      "epoch": 9.986622073578594,
      "grad_norm": 1.642620325088501,
      "learning_rate": 1.3377926421404682e-07,
      "loss": 0.0119,
      "step": 14930
    },
    {
      "epoch": 9.993311036789297,
      "grad_norm": 1.9654911756515503,
      "learning_rate": 6.688963210702341e-08,
      "loss": 0.0209,
      "step": 14940
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.5067110061645508,
      "learning_rate": 0.0,
      "loss": 0.0161,
      "step": 14950
    }
  ],
  "logging_steps": 10,
  "max_steps": 14950,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}