{
  "best_metric": 0.03400569409132004,
  "best_model_checkpoint": "saves/psy-course/Llama-3.1-8B-Instruct/train/fold2/checkpoint-1300",
  "epoch": 5.0,
  "eval_steps": 50,
  "global_step": 3225,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015503875968992248,
      "grad_norm": 4.586867332458496,
      "learning_rate": 3.0959752321981426e-06,
      "loss": 1.585,
      "step": 10
    },
    {
      "epoch": 0.031007751937984496,
      "grad_norm": 4.746698379516602,
      "learning_rate": 6.191950464396285e-06,
      "loss": 1.523,
      "step": 20
    },
    {
      "epoch": 0.046511627906976744,
      "grad_norm": 5.548886775970459,
      "learning_rate": 9.287925696594429e-06,
      "loss": 1.318,
      "step": 30
    },
    {
      "epoch": 0.06201550387596899,
      "grad_norm": 2.0926177501678467,
      "learning_rate": 1.238390092879257e-05,
      "loss": 0.8377,
      "step": 40
    },
    {
      "epoch": 0.07751937984496124,
      "grad_norm": 1.533339500427246,
      "learning_rate": 1.5479876160990712e-05,
      "loss": 0.5122,
      "step": 50
    },
    {
      "epoch": 0.07751937984496124,
      "eval_loss": 0.41176876425743103,
      "eval_runtime": 173.3032,
      "eval_samples_per_second": 6.618,
      "eval_steps_per_second": 6.618,
      "step": 50
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 0.9713454246520996,
      "learning_rate": 1.8575851393188857e-05,
      "loss": 0.3858,
      "step": 60
    },
    {
      "epoch": 0.10852713178294573,
      "grad_norm": 1.3564386367797852,
      "learning_rate": 2.1671826625387e-05,
      "loss": 0.2398,
      "step": 70
    },
    {
      "epoch": 0.12403100775193798,
      "grad_norm": 1.194110631942749,
      "learning_rate": 2.476780185758514e-05,
      "loss": 0.1607,
      "step": 80
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 1.0269737243652344,
      "learning_rate": 2.7863777089783283e-05,
      "loss": 0.1322,
      "step": 90
    },
    {
      "epoch": 0.15503875968992248,
      "grad_norm": 0.9308303594589233,
      "learning_rate": 3.0959752321981425e-05,
      "loss": 0.0913,
      "step": 100
    },
    {
      "epoch": 0.15503875968992248,
      "eval_loss": 0.08312761038541794,
      "eval_runtime": 172.5746,
      "eval_samples_per_second": 6.646,
      "eval_steps_per_second": 6.646,
      "step": 100
    },
    {
      "epoch": 0.17054263565891473,
      "grad_norm": 1.6514289379119873,
      "learning_rate": 3.4055727554179566e-05,
      "loss": 0.0996,
      "step": 110
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 1.6203035116195679,
      "learning_rate": 3.7151702786377715e-05,
      "loss": 0.0914,
      "step": 120
    },
    {
      "epoch": 0.20155038759689922,
      "grad_norm": 0.9771865606307983,
      "learning_rate": 4.024767801857585e-05,
      "loss": 0.0855,
      "step": 130
    },
    {
      "epoch": 0.21705426356589147,
      "grad_norm": 0.6767793297767639,
      "learning_rate": 4.3343653250774e-05,
      "loss": 0.0888,
      "step": 140
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 0.7697650790214539,
      "learning_rate": 4.6439628482972134e-05,
      "loss": 0.0638,
      "step": 150
    },
    {
      "epoch": 0.23255813953488372,
      "eval_loss": 0.06616330146789551,
      "eval_runtime": 172.3668,
      "eval_samples_per_second": 6.654,
      "eval_steps_per_second": 6.654,
      "step": 150
    },
    {
      "epoch": 0.24806201550387597,
      "grad_norm": 0.8017315864562988,
      "learning_rate": 4.953560371517028e-05,
      "loss": 0.0788,
      "step": 160
    },
    {
      "epoch": 0.26356589147286824,
      "grad_norm": 1.0115169286727905,
      "learning_rate": 5.2631578947368424e-05,
      "loss": 0.073,
      "step": 170
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.4792479872703552,
      "learning_rate": 5.5727554179566566e-05,
      "loss": 0.0644,
      "step": 180
    },
    {
      "epoch": 0.29457364341085274,
      "grad_norm": 2.296128273010254,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.0679,
      "step": 190
    },
    {
      "epoch": 0.31007751937984496,
      "grad_norm": 1.2886589765548706,
      "learning_rate": 6.191950464396285e-05,
      "loss": 0.0589,
      "step": 200
    },
    {
      "epoch": 0.31007751937984496,
      "eval_loss": 0.05583063140511513,
      "eval_runtime": 172.0552,
      "eval_samples_per_second": 6.666,
      "eval_steps_per_second": 6.666,
      "step": 200
    },
    {
      "epoch": 0.32558139534883723,
      "grad_norm": 0.5177517533302307,
      "learning_rate": 6.501547987616098e-05,
      "loss": 0.0585,
      "step": 210
    },
    {
      "epoch": 0.34108527131782945,
      "grad_norm": 0.7501789331436157,
      "learning_rate": 6.811145510835913e-05,
      "loss": 0.0556,
      "step": 220
    },
    {
      "epoch": 0.35658914728682173,
      "grad_norm": 0.5279349088668823,
      "learning_rate": 7.120743034055728e-05,
      "loss": 0.0586,
      "step": 230
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.6271268725395203,
      "learning_rate": 7.430340557275543e-05,
      "loss": 0.0776,
      "step": 240
    },
    {
      "epoch": 0.3875968992248062,
      "grad_norm": 0.8573686480522156,
      "learning_rate": 7.739938080495357e-05,
      "loss": 0.0628,
      "step": 250
    },
    {
      "epoch": 0.3875968992248062,
      "eval_loss": 0.0528356209397316,
      "eval_runtime": 171.8837,
      "eval_samples_per_second": 6.673,
      "eval_steps_per_second": 6.673,
      "step": 250
    },
    {
      "epoch": 0.40310077519379844,
      "grad_norm": 0.38616809248924255,
      "learning_rate": 8.04953560371517e-05,
      "loss": 0.0509,
      "step": 260
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 0.4749118983745575,
      "learning_rate": 8.359133126934985e-05,
      "loss": 0.0489,
      "step": 270
    },
    {
      "epoch": 0.43410852713178294,
      "grad_norm": 0.6327615976333618,
      "learning_rate": 8.6687306501548e-05,
      "loss": 0.0845,
      "step": 280
    },
    {
      "epoch": 0.4496124031007752,
      "grad_norm": 0.9490478038787842,
      "learning_rate": 8.978328173374613e-05,
      "loss": 0.0613,
      "step": 290
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.5700337886810303,
      "learning_rate": 9.287925696594427e-05,
      "loss": 0.0482,
      "step": 300
    },
    {
      "epoch": 0.46511627906976744,
      "eval_loss": 0.04843495413661003,
      "eval_runtime": 171.945,
      "eval_samples_per_second": 6.671,
      "eval_steps_per_second": 6.671,
      "step": 300
    },
    {
      "epoch": 0.4806201550387597,
      "grad_norm": 0.6147510409355164,
      "learning_rate": 9.597523219814242e-05,
      "loss": 0.0594,
      "step": 310
    },
    {
      "epoch": 0.49612403100775193,
      "grad_norm": 0.5589321851730347,
      "learning_rate": 9.907120743034056e-05,
      "loss": 0.044,
      "step": 320
    },
    {
      "epoch": 0.5116279069767442,
      "grad_norm": 0.3106387257575989,
      "learning_rate": 9.999856438185238e-05,
      "loss": 0.0505,
      "step": 330
    },
    {
      "epoch": 0.5271317829457365,
      "grad_norm": 0.4362027943134308,
      "learning_rate": 9.999153298122152e-05,
      "loss": 0.0536,
      "step": 340
    },
    {
      "epoch": 0.5426356589147286,
      "grad_norm": 0.2365163117647171,
      "learning_rate": 9.997864293614054e-05,
      "loss": 0.0429,
      "step": 350
    },
    {
      "epoch": 0.5426356589147286,
      "eval_loss": 0.04433372616767883,
      "eval_runtime": 171.9903,
      "eval_samples_per_second": 6.669,
      "eval_steps_per_second": 6.669,
      "step": 350
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.47546619176864624,
      "learning_rate": 9.995989575722902e-05,
      "loss": 0.0432,
      "step": 360
    },
    {
      "epoch": 0.5736434108527132,
      "grad_norm": 0.3624158203601837,
      "learning_rate": 9.993529364152018e-05,
      "loss": 0.0612,
      "step": 370
    },
    {
      "epoch": 0.5891472868217055,
      "grad_norm": 0.48345044255256653,
      "learning_rate": 9.990483947220319e-05,
      "loss": 0.0611,
      "step": 380
    },
    {
      "epoch": 0.6046511627906976,
      "grad_norm": 0.272274374961853,
      "learning_rate": 9.986853681828546e-05,
      "loss": 0.0574,
      "step": 390
    },
    {
      "epoch": 0.6201550387596899,
      "grad_norm": 0.7242560982704163,
      "learning_rate": 9.982638993417425e-05,
      "loss": 0.0526,
      "step": 400
    },
    {
      "epoch": 0.6201550387596899,
      "eval_loss": 0.04319106042385101,
      "eval_runtime": 171.8171,
      "eval_samples_per_second": 6.676,
      "eval_steps_per_second": 6.676,
      "step": 400
    },
    {
      "epoch": 0.6356589147286822,
      "grad_norm": 0.44166314601898193,
      "learning_rate": 9.977840375917817e-05,
      "loss": 0.0545,
      "step": 410
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.3073597252368927,
      "learning_rate": 9.972458391692827e-05,
      "loss": 0.0529,
      "step": 420
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.4054962396621704,
      "learning_rate": 9.966493671471904e-05,
      "loss": 0.0484,
      "step": 430
    },
    {
      "epoch": 0.6821705426356589,
      "grad_norm": 0.38832733035087585,
      "learning_rate": 9.959946914276922e-05,
      "loss": 0.0607,
      "step": 440
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 0.5320560932159424,
      "learning_rate": 9.952818887340257e-05,
      "loss": 0.0446,
      "step": 450
    },
    {
      "epoch": 0.6976744186046512,
      "eval_loss": 0.03905312716960907,
      "eval_runtime": 171.6179,
      "eval_samples_per_second": 6.683,
      "eval_steps_per_second": 6.683,
      "step": 450
    },
    {
      "epoch": 0.7131782945736435,
      "grad_norm": 0.2739388048648834,
      "learning_rate": 9.945110426014878e-05,
      "loss": 0.0507,
      "step": 460
    },
    {
      "epoch": 0.7286821705426356,
      "grad_norm": 0.5490703582763672,
      "learning_rate": 9.936822433676444e-05,
      "loss": 0.0424,
      "step": 470
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.2568321228027344,
      "learning_rate": 9.927955881617444e-05,
      "loss": 0.047,
      "step": 480
    },
    {
      "epoch": 0.7596899224806202,
      "grad_norm": 0.5469834804534912,
      "learning_rate": 9.918511808933358e-05,
      "loss": 0.0443,
      "step": 490
    },
    {
      "epoch": 0.7751937984496124,
      "grad_norm": 0.3519169092178345,
      "learning_rate": 9.908491322400885e-05,
      "loss": 0.0503,
      "step": 500
    },
    {
      "epoch": 0.7751937984496124,
      "eval_loss": 0.0377386212348938,
      "eval_runtime": 171.3737,
      "eval_samples_per_second": 6.693,
      "eval_steps_per_second": 6.693,
      "step": 500
    },
    {
      "epoch": 0.7906976744186046,
      "grad_norm": 0.18305093050003052,
      "learning_rate": 9.897895596348247e-05,
      "loss": 0.042,
      "step": 510
    },
    {
      "epoch": 0.8062015503875969,
      "grad_norm": 0.34450390934944153,
      "learning_rate": 9.886725872517552e-05,
      "loss": 0.037,
      "step": 520
    },
    {
      "epoch": 0.8217054263565892,
      "grad_norm": 0.3624585270881653,
      "learning_rate": 9.874983459919277e-05,
      "loss": 0.0471,
      "step": 530
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.3540893495082855,
      "learning_rate": 9.862669734678867e-05,
      "loss": 0.0391,
      "step": 540
    },
    {
      "epoch": 0.8527131782945736,
      "grad_norm": 0.26056310534477234,
      "learning_rate": 9.849786139875452e-05,
      "loss": 0.048,
      "step": 550
    },
    {
      "epoch": 0.8527131782945736,
      "eval_loss": 0.03856755420565605,
      "eval_runtime": 171.2806,
      "eval_samples_per_second": 6.697,
      "eval_steps_per_second": 6.697,
      "step": 550
    },
    {
      "epoch": 0.8682170542635659,
      "grad_norm": 0.4234136939048767,
      "learning_rate": 9.836334185372738e-05,
      "loss": 0.048,
      "step": 560
    },
    {
      "epoch": 0.8837209302325582,
      "grad_norm": 0.3422122299671173,
      "learning_rate": 9.822315447642056e-05,
      "loss": 0.0432,
      "step": 570
    },
    {
      "epoch": 0.8992248062015504,
      "grad_norm": 0.4335425794124603,
      "learning_rate": 9.807731569577615e-05,
      "loss": 0.0337,
      "step": 580
    },
    {
      "epoch": 0.9147286821705426,
      "grad_norm": 0.6463870406150818,
      "learning_rate": 9.792584260303964e-05,
      "loss": 0.0379,
      "step": 590
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.4832146167755127,
      "learning_rate": 9.776875294975698e-05,
      "loss": 0.0636,
      "step": 600
    },
    {
      "epoch": 0.9302325581395349,
      "eval_loss": 0.04392372816801071,
      "eval_runtime": 171.0631,
      "eval_samples_per_second": 6.705,
      "eval_steps_per_second": 6.705,
      "step": 600
    },
    {
      "epoch": 0.9457364341085271,
      "grad_norm": 0.42290669679641724,
      "learning_rate": 9.760606514569416e-05,
      "loss": 0.043,
      "step": 610
    },
    {
      "epoch": 0.9612403100775194,
      "grad_norm": 0.2062637358903885,
      "learning_rate": 9.743779825667984e-05,
      "loss": 0.039,
      "step": 620
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 0.2891997694969177,
      "learning_rate": 9.726397200237085e-05,
      "loss": 0.0424,
      "step": 630
    },
    {
      "epoch": 0.9922480620155039,
      "grad_norm": 0.4399895966053009,
      "learning_rate": 9.708460675394126e-05,
      "loss": 0.0506,
      "step": 640
    },
    {
      "epoch": 1.0077519379844961,
      "grad_norm": 0.20108899474143982,
      "learning_rate": 9.689972353169499e-05,
      "loss": 0.0333,
      "step": 650
    },
    {
      "epoch": 1.0077519379844961,
      "eval_loss": 0.03613699600100517,
      "eval_runtime": 171.0726,
      "eval_samples_per_second": 6.705,
      "eval_steps_per_second": 6.705,
      "step": 650
    },
    {
      "epoch": 1.0232558139534884,
      "grad_norm": 0.40628916025161743,
      "learning_rate": 9.67093440026024e-05,
      "loss": 0.0322,
      "step": 660
    },
    {
      "epoch": 1.0387596899224807,
      "grad_norm": 0.23710253834724426,
      "learning_rate": 9.651349047776112e-05,
      "loss": 0.0473,
      "step": 670
    },
    {
      "epoch": 1.054263565891473,
      "grad_norm": 0.4060989022254944,
      "learning_rate": 9.631218590978126e-05,
      "loss": 0.0257,
      "step": 680
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 0.29535090923309326,
      "learning_rate": 9.610545389009562e-05,
      "loss": 0.0339,
      "step": 690
    },
    {
      "epoch": 1.0852713178294573,
      "grad_norm": 0.336176335811615,
      "learning_rate": 9.589331864619486e-05,
      "loss": 0.0319,
      "step": 700
    },
    {
      "epoch": 1.0852713178294573,
      "eval_loss": 0.038464464247226715,
      "eval_runtime": 170.8431,
      "eval_samples_per_second": 6.714,
      "eval_steps_per_second": 6.714,
      "step": 700
    },
    {
      "epoch": 1.1007751937984496,
      "grad_norm": 0.35942918062210083,
      "learning_rate": 9.567580503878833e-05,
      "loss": 0.0285,
      "step": 710
    },
    {
      "epoch": 1.1162790697674418,
      "grad_norm": 0.30677011609077454,
      "learning_rate": 9.545293855889043e-05,
      "loss": 0.0368,
      "step": 720
    },
    {
      "epoch": 1.1317829457364341,
      "grad_norm": 0.16257011890411377,
      "learning_rate": 9.522474532483337e-05,
      "loss": 0.0351,
      "step": 730
    },
    {
      "epoch": 1.1472868217054264,
      "grad_norm": 0.3570394515991211,
      "learning_rate": 9.499125207920622e-05,
      "loss": 0.0477,
      "step": 740
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.145313560962677,
      "learning_rate": 9.475248618572096e-05,
      "loss": 0.033,
      "step": 750
    },
    {
      "epoch": 1.1627906976744187,
      "eval_loss": 0.03572777658700943,
      "eval_runtime": 170.7702,
      "eval_samples_per_second": 6.717,
      "eval_steps_per_second": 6.717,
      "step": 750
    },
    {
      "epoch": 1.178294573643411,
      "grad_norm": 0.43808048963546753,
      "learning_rate": 9.45084756260055e-05,
      "loss": 0.0392,
      "step": 760
    },
    {
      "epoch": 1.193798449612403,
      "grad_norm": 0.1340464949607849,
      "learning_rate": 9.425924899632458e-05,
      "loss": 0.0344,
      "step": 770
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 0.14380404353141785,
      "learning_rate": 9.400483550422845e-05,
      "loss": 0.0295,
      "step": 780
    },
    {
      "epoch": 1.2248062015503876,
      "grad_norm": 0.28932222723960876,
      "learning_rate": 9.374526496512992e-05,
      "loss": 0.0363,
      "step": 790
    },
    {
      "epoch": 1.2403100775193798,
      "grad_norm": 0.353335440158844,
      "learning_rate": 9.348056779881025e-05,
      "loss": 0.0242,
      "step": 800
    },
    {
      "epoch": 1.2403100775193798,
      "eval_loss": 0.037058863788843155,
      "eval_runtime": 170.7322,
      "eval_samples_per_second": 6.718,
      "eval_steps_per_second": 6.718,
      "step": 800
    },
    {
      "epoch": 1.255813953488372,
      "grad_norm": 0.43459367752075195,
      "learning_rate": 9.321077502585417e-05,
      "loss": 0.0315,
      "step": 810
    },
    {
      "epoch": 1.2713178294573644,
      "grad_norm": 0.3110826313495636,
      "learning_rate": 9.293591826401451e-05,
      "loss": 0.0342,
      "step": 820
    },
    {
      "epoch": 1.2868217054263567,
      "grad_norm": 0.38052013516426086,
      "learning_rate": 9.265602972450673e-05,
      "loss": 0.0376,
      "step": 830
    },
    {
      "epoch": 1.302325581395349,
      "grad_norm": 0.25867798924446106,
      "learning_rate": 9.237114220823413e-05,
      "loss": 0.0366,
      "step": 840
    },
    {
      "epoch": 1.3178294573643412,
      "grad_norm": 0.1531626284122467,
      "learning_rate": 9.208128910194377e-05,
      "loss": 0.0363,
      "step": 850
    },
    {
      "epoch": 1.3178294573643412,
      "eval_loss": 0.03434592857956886,
      "eval_runtime": 170.8504,
      "eval_samples_per_second": 6.713,
      "eval_steps_per_second": 6.713,
      "step": 850
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.33476677536964417,
      "learning_rate": 9.17865043743138e-05,
      "loss": 0.0377,
      "step": 860
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 0.14141975343227386,
      "learning_rate": 9.148682257197245e-05,
      "loss": 0.0294,
      "step": 870
    },
    {
      "epoch": 1.3643410852713178,
      "grad_norm": 0.42156022787094116,
      "learning_rate": 9.118227881544967e-05,
      "loss": 0.0469,
      "step": 880
    },
    {
      "epoch": 1.37984496124031,
      "grad_norm": 0.3194252550601959,
      "learning_rate": 9.087290879506104e-05,
      "loss": 0.0285,
      "step": 890
    },
    {
      "epoch": 1.3953488372093024,
      "grad_norm": 0.14607588946819305,
      "learning_rate": 9.055874876672519e-05,
      "loss": 0.0419,
      "step": 900
    },
    {
      "epoch": 1.3953488372093024,
      "eval_loss": 0.036030739545822144,
      "eval_runtime": 171.2298,
      "eval_samples_per_second": 6.699,
      "eval_steps_per_second": 6.699,
      "step": 900
    },
    {
      "epoch": 1.4108527131782946,
      "grad_norm": 0.21363942325115204,
      "learning_rate": 9.023983554771492e-05,
      "loss": 0.0311,
      "step": 910
    },
    {
      "epoch": 1.4263565891472867,
      "grad_norm": 0.23408649861812592,
      "learning_rate": 8.991620651234242e-05,
      "loss": 0.0285,
      "step": 920
    },
    {
      "epoch": 1.441860465116279,
      "grad_norm": 0.3802158236503601,
      "learning_rate": 8.958789958757928e-05,
      "loss": 0.0331,
      "step": 930
    },
    {
      "epoch": 1.4573643410852712,
      "grad_norm": 0.20695924758911133,
      "learning_rate": 8.925495324861177e-05,
      "loss": 0.034,
      "step": 940
    },
    {
      "epoch": 1.4728682170542635,
      "grad_norm": 0.429266095161438,
      "learning_rate": 8.89174065143318e-05,
      "loss": 0.0444,
      "step": 950
    },
    {
      "epoch": 1.4728682170542635,
      "eval_loss": 0.03494586795568466,
      "eval_runtime": 171.5354,
      "eval_samples_per_second": 6.687,
      "eval_steps_per_second": 6.687,
      "step": 950
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 0.3035537898540497,
      "learning_rate": 8.857529894276412e-05,
      "loss": 0.036,
      "step": 960
    },
    {
      "epoch": 1.503875968992248,
      "grad_norm": 0.23642411828041077,
      "learning_rate": 8.822867062643059e-05,
      "loss": 0.0325,
      "step": 970
    },
    {
      "epoch": 1.5193798449612403,
      "grad_norm": 0.22683683037757874,
      "learning_rate": 8.78775621876514e-05,
      "loss": 0.0375,
      "step": 980
    },
    {
      "epoch": 1.5348837209302326,
      "grad_norm": 0.3267425298690796,
      "learning_rate": 8.752201477378459e-05,
      "loss": 0.0266,
      "step": 990
    },
    {
      "epoch": 1.550387596899225,
      "grad_norm": 0.1990777552127838,
      "learning_rate": 8.716207005240382e-05,
      "loss": 0.0297,
      "step": 1000
    },
    {
      "epoch": 1.550387596899225,
      "eval_loss": 0.03623834624886513,
      "eval_runtime": 171.9583,
      "eval_samples_per_second": 6.67,
      "eval_steps_per_second": 6.67,
      "step": 1000
    },
    {
      "epoch": 1.5658914728682172,
      "grad_norm": 0.45734983682632446,
      "learning_rate": 8.679777020641525e-05,
      "loss": 0.0423,
      "step": 1010
    },
    {
      "epoch": 1.5813953488372094,
      "grad_norm": 0.3885669410228729,
      "learning_rate": 8.6429157929114e-05,
      "loss": 0.04,
      "step": 1020
    },
    {
      "epoch": 1.5968992248062015,
      "grad_norm": 0.3537205755710602,
      "learning_rate": 8.60562764191808e-05,
      "loss": 0.0331,
      "step": 1030
    },
    {
      "epoch": 1.6124031007751938,
      "grad_norm": 0.2690904140472412,
      "learning_rate": 8.567916937561944e-05,
      "loss": 0.0342,
      "step": 1040
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 0.12849485874176025,
      "learning_rate": 8.529788099263554e-05,
      "loss": 0.0348,
      "step": 1050
    },
    {
      "epoch": 1.627906976744186,
      "eval_loss": 0.03483985364437103,
      "eval_runtime": 172.4575,
      "eval_samples_per_second": 6.651,
      "eval_steps_per_second": 6.651,
      "step": 1050
    },
    {
      "epoch": 1.6434108527131783,
      "grad_norm": 0.16510334610939026,
      "learning_rate": 8.49124559544573e-05,
      "loss": 0.0289,
      "step": 1060
    },
    {
      "epoch": 1.6589147286821704,
      "grad_norm": 0.3688533902168274,
      "learning_rate": 8.452293943009889e-05,
      "loss": 0.0289,
      "step": 1070
    },
    {
      "epoch": 1.6744186046511627,
      "grad_norm": 0.3377443253993988,
      "learning_rate": 8.412937706806692e-05,
      "loss": 0.0368,
      "step": 1080
    },
    {
      "epoch": 1.689922480620155,
      "grad_norm": 0.24271418154239655,
      "learning_rate": 8.373181499101077e-05,
      "loss": 0.0328,
      "step": 1090
    },
    {
      "epoch": 1.7054263565891472,
      "grad_norm": 0.6041512489318848,
      "learning_rate": 8.333029979031737e-05,
      "loss": 0.0271,
      "step": 1100
    },
    {
      "epoch": 1.7054263565891472,
      "eval_loss": 0.03539975360035896,
      "eval_runtime": 172.8867,
      "eval_samples_per_second": 6.634,
      "eval_steps_per_second": 6.634,
      "step": 1100
    },
    {
      "epoch": 1.7209302325581395,
      "grad_norm": 0.1947883814573288,
      "learning_rate": 8.292487852065104e-05,
      "loss": 0.0488,
      "step": 1110
    },
    {
      "epoch": 1.7364341085271318,
      "grad_norm": 0.5195293426513672,
      "learning_rate": 8.251559869443897e-05,
      "loss": 0.0341,
      "step": 1120
    },
    {
      "epoch": 1.751937984496124,
      "grad_norm": 0.3926396369934082,
      "learning_rate": 8.210250827630313e-05,
      "loss": 0.0327,
      "step": 1130
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 0.31271877884864807,
      "learning_rate": 8.168565567743924e-05,
      "loss": 0.0261,
      "step": 1140
    },
    {
      "epoch": 1.7829457364341086,
      "grad_norm": 0.6073179841041565,
      "learning_rate": 8.126508974994321e-05,
      "loss": 0.0362,
      "step": 1150
    },
    {
      "epoch": 1.7829457364341086,
      "eval_loss": 0.036610543727874756,
      "eval_runtime": 173.3111,
      "eval_samples_per_second": 6.618,
      "eval_steps_per_second": 6.618,
      "step": 1150
    },
    {
      "epoch": 1.7984496124031009,
      "grad_norm": 0.29566118121147156,
      "learning_rate": 8.084085978108611e-05,
      "loss": 0.02,
      "step": 1160
    },
    {
      "epoch": 1.8139534883720931,
      "grad_norm": 0.2208748310804367,
      "learning_rate": 8.041301548753804e-05,
      "loss": 0.0274,
      "step": 1170
    },
    {
      "epoch": 1.8294573643410854,
      "grad_norm": 0.33662283420562744,
      "learning_rate": 7.998160700954165e-05,
      "loss": 0.026,
      "step": 1180
    },
    {
      "epoch": 1.8449612403100775,
      "grad_norm": 0.3825443387031555,
      "learning_rate": 7.954668490503621e-05,
      "loss": 0.0326,
      "step": 1190
    },
    {
      "epoch": 1.8604651162790697,
      "grad_norm": 0.4023016691207886,
      "learning_rate": 7.910830014373237e-05,
      "loss": 0.034,
      "step": 1200
    },
    {
      "epoch": 1.8604651162790697,
      "eval_loss": 0.03443226218223572,
      "eval_runtime": 173.7001,
      "eval_samples_per_second": 6.603,
      "eval_steps_per_second": 6.603,
      "step": 1200
    },
    {
      "epoch": 1.875968992248062,
      "grad_norm": 0.21011577546596527,
      "learning_rate": 7.866650410113907e-05,
      "loss": 0.0314,
      "step": 1210
    },
    {
      "epoch": 1.8914728682170543,
      "grad_norm": 0.4147494435310364,
      "learning_rate": 7.822134855254263e-05,
      "loss": 0.0386,
      "step": 1220
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 0.2679975926876068,
      "learning_rate": 7.777288566693905e-05,
      "loss": 0.0347,
      "step": 1230
    },
    {
      "epoch": 1.9224806201550386,
      "grad_norm": 0.30710622668266296,
      "learning_rate": 7.732116800092018e-05,
      "loss": 0.0309,
      "step": 1240
    },
    {
      "epoch": 1.937984496124031,
      "grad_norm": 0.2684403359889984,
      "learning_rate": 7.686624849251453e-05,
      "loss": 0.039,
      "step": 1250
    },
    {
      "epoch": 1.937984496124031,
      "eval_loss": 0.03436611220240593,
      "eval_runtime": 173.9789,
      "eval_samples_per_second": 6.593,
      "eval_steps_per_second": 6.593,
      "step": 1250
    },
    {
      "epoch": 1.9534883720930232,
      "grad_norm": 0.381550133228302,
      "learning_rate": 7.640818045498324e-05,
      "loss": 0.0471,
      "step": 1260
    },
    {
      "epoch": 1.9689922480620154,
      "grad_norm": 0.12695617973804474,
      "learning_rate": 7.59470175705722e-05,
      "loss": 0.0362,
      "step": 1270
    },
    {
      "epoch": 1.9844961240310077,
      "grad_norm": 0.2954389452934265,
      "learning_rate": 7.548281388422088e-05,
      "loss": 0.0395,
      "step": 1280
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.21959000825881958,
      "learning_rate": 7.50156237972286e-05,
      "loss": 0.0334,
      "step": 1290
    },
    {
      "epoch": 2.0155038759689923,
      "grad_norm": 0.22099755704402924,
      "learning_rate": 7.454550206087921e-05,
      "loss": 0.0248,
      "step": 1300
    },
    {
      "epoch": 2.0155038759689923,
      "eval_loss": 0.03400569409132004,
      "eval_runtime": 174.5025,
      "eval_samples_per_second": 6.573,
      "eval_steps_per_second": 6.573,
      "step": 1300
    },
    {
      "epoch": 2.0310077519379846,
      "grad_norm": 0.25023770332336426,
      "learning_rate": 7.407250377002451e-05,
      "loss": 0.0241,
      "step": 1310
    },
    {
      "epoch": 2.046511627906977,
      "grad_norm": 0.21477766335010529,
      "learning_rate": 7.359668435662757e-05,
      "loss": 0.0223,
      "step": 1320
    },
    {
      "epoch": 2.062015503875969,
      "grad_norm": 0.2844666540622711,
      "learning_rate": 7.311809958326659e-05,
      "loss": 0.0168,
      "step": 1330
    },
    {
      "epoch": 2.0775193798449614,
      "grad_norm": 0.33521831035614014,
      "learning_rate": 7.263680553659983e-05,
      "loss": 0.0216,
      "step": 1340
    },
    {
      "epoch": 2.0930232558139537,
      "grad_norm": 0.3837973177433014,
      "learning_rate": 7.21528586207927e-05,
      "loss": 0.0209,
      "step": 1350
    },
    {
      "epoch": 2.0930232558139537,
      "eval_loss": 0.03692101314663887,
      "eval_runtime": 174.8765,
      "eval_samples_per_second": 6.559,
      "eval_steps_per_second": 6.559,
      "step": 1350
    },
    {
      "epoch": 2.108527131782946,
      "grad_norm": 0.6733439564704895,
      "learning_rate": 7.166631555090759e-05,
      "loss": 0.0269,
      "step": 1360
    },
    {
      "epoch": 2.124031007751938,
      "grad_norm": 0.2908807098865509,
      "learning_rate": 7.117723334625731e-05,
      "loss": 0.0219,
      "step": 1370
    },
    {
      "epoch": 2.13953488372093,
      "grad_norm": 0.3064257502555847,
      "learning_rate": 7.068566932372278e-05,
      "loss": 0.0239,
      "step": 1380
    },
    {
      "epoch": 2.1550387596899223,
      "grad_norm": 0.4648115336894989,
      "learning_rate": 7.019168109103599e-05,
      "loss": 0.0199,
      "step": 1390
    },
    {
      "epoch": 2.1705426356589146,
      "grad_norm": 0.24724262952804565,
      "learning_rate": 6.96953265400287e-05,
      "loss": 0.0211,
      "step": 1400
    },
    {
      "epoch": 2.1705426356589146,
      "eval_loss": 0.03518850728869438,
      "eval_runtime": 174.9572,
      "eval_samples_per_second": 6.556,
      "eval_steps_per_second": 6.556,
      "step": 1400
    },
    {
      "epoch": 2.186046511627907,
      "grad_norm": 0.25550487637519836,
      "learning_rate": 6.9196663839848e-05,
      "loss": 0.0195,
      "step": 1410
    },
    {
      "epoch": 2.201550387596899,
      "grad_norm": 0.37793949246406555,
      "learning_rate": 6.869575143013934e-05,
      "loss": 0.0201,
      "step": 1420
    },
    {
      "epoch": 2.2170542635658914,
      "grad_norm": 0.4046792685985565,
      "learning_rate": 6.819264801419771e-05,
      "loss": 0.0314,
      "step": 1430
    },
    {
      "epoch": 2.2325581395348837,
      "grad_norm": 0.19130779802799225,
      "learning_rate": 6.768741255208813e-05,
      "loss": 0.0203,
      "step": 1440
    },
    {
      "epoch": 2.248062015503876,
      "grad_norm": 0.3072698414325714,
      "learning_rate": 6.718010425373599e-05,
      "loss": 0.0178,
      "step": 1450
    },
    {
      "epoch": 2.248062015503876,
      "eval_loss": 0.0378645583987236,
      "eval_runtime": 175.3079,
      "eval_samples_per_second": 6.543,
      "eval_steps_per_second": 6.543,
      "step": 1450
    },
    {
      "epoch": 2.2635658914728682,
      "grad_norm": 0.518846333026886,
      "learning_rate": 6.667078257198796e-05,
      "loss": 0.0215,
      "step": 1460
    },
    {
      "epoch": 2.2790697674418605,
      "grad_norm": 0.31176698207855225,
      "learning_rate": 6.615950719564466e-05,
      "loss": 0.016,
      "step": 1470
    },
    {
      "epoch": 2.294573643410853,
      "grad_norm": 0.3617725372314453,
      "learning_rate": 6.56463380424655e-05,
      "loss": 0.0331,
      "step": 1480
    },
    {
      "epoch": 2.310077519379845,
      "grad_norm": 0.19033777713775635,
      "learning_rate": 6.51313352521468e-05,
      "loss": 0.0198,
      "step": 1490
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 0.3381364047527313,
      "learning_rate": 6.46145591792738e-05,
      "loss": 0.026,
      "step": 1500
    },
    {
      "epoch": 2.3255813953488373,
      "eval_loss": 0.03550190106034279,
      "eval_runtime": 175.3413,
      "eval_samples_per_second": 6.542,
      "eval_steps_per_second": 6.542,
      "step": 1500
    },
    {
      "epoch": 2.3410852713178296,
      "grad_norm": 0.11263620853424072,
      "learning_rate": 6.409607038624759e-05,
      "loss": 0.0168,
      "step": 1510
    },
    {
      "epoch": 2.356589147286822,
      "grad_norm": 0.24013790488243103,
      "learning_rate": 6.357592963618761e-05,
      "loss": 0.0303,
      "step": 1520
    },
    {
      "epoch": 2.3720930232558137,
      "grad_norm": 0.19617025554180145,
      "learning_rate": 6.305419788581067e-05,
      "loss": 0.0218,
      "step": 1530
    },
    {
      "epoch": 2.387596899224806,
      "grad_norm": 0.19198082387447357,
      "learning_rate": 6.253093627828724e-05,
      "loss": 0.0192,
      "step": 1540
    },
    {
      "epoch": 2.4031007751937983,
      "grad_norm": 0.32724863290786743,
      "learning_rate": 6.200620613607597e-05,
      "loss": 0.0166,
      "step": 1550
    },
    {
      "epoch": 2.4031007751937983,
      "eval_loss": 0.03754343092441559,
      "eval_runtime": 175.4702,
      "eval_samples_per_second": 6.537,
      "eval_steps_per_second": 6.537,
      "step": 1550
    },
    {
      "epoch": 2.4186046511627906,
      "grad_norm": 0.369870662689209,
      "learning_rate": 6.148006895373706e-05,
      "loss": 0.0226,
      "step": 1560
    },
    {
      "epoch": 2.434108527131783,
      "grad_norm": 0.2815375030040741,
      "learning_rate": 6.0952586390725644e-05,
      "loss": 0.0177,
      "step": 1570
    },
    {
      "epoch": 2.449612403100775,
      "grad_norm": 0.3072023391723633,
      "learning_rate": 6.042382026416563e-05,
      "loss": 0.0208,
      "step": 1580
    },
    {
      "epoch": 2.4651162790697674,
      "grad_norm": 0.2327968031167984,
      "learning_rate": 5.989383254160529e-05,
      "loss": 0.0233,
      "step": 1590
    },
    {
      "epoch": 2.4806201550387597,
      "grad_norm": 0.49106815457344055,
      "learning_rate": 5.936268533375506e-05,
      "loss": 0.0237,
      "step": 1600
    },
    {
      "epoch": 2.4806201550387597,
      "eval_loss": 0.03549407050013542,
      "eval_runtime": 175.559,
      "eval_samples_per_second": 6.533,
      "eval_steps_per_second": 6.533,
      "step": 1600
    },
    {
      "epoch": 2.496124031007752,
      "grad_norm": 0.662108838558197,
      "learning_rate": 5.883044088720865e-05,
      "loss": 0.027,
      "step": 1610
    },
    {
      "epoch": 2.511627906976744,
      "grad_norm": 0.1898617446422577,
      "learning_rate": 5.829716157714814e-05,
      "loss": 0.02,
      "step": 1620
    },
    {
      "epoch": 2.5271317829457365,
      "grad_norm": 0.3116423189640045,
      "learning_rate": 5.77629099000341e-05,
      "loss": 0.029,
      "step": 1630
    },
    {
      "epoch": 2.5426356589147288,
      "grad_norm": 0.32322463393211365,
      "learning_rate": 5.7227748466281486e-05,
      "loss": 0.0277,
      "step": 1640
    },
    {
      "epoch": 2.558139534883721,
      "grad_norm": 0.42573103308677673,
      "learning_rate": 5.669173999292208e-05,
      "loss": 0.037,
      "step": 1650
    },
    {
      "epoch": 2.558139534883721,
      "eval_loss": 0.03468121960759163,
      "eval_runtime": 175.6858,
      "eval_samples_per_second": 6.529,
      "eval_steps_per_second": 6.529,
      "step": 1650
    },
    {
      "epoch": 2.5736434108527133,
      "grad_norm": 0.5278474688529968,
      "learning_rate": 5.615494729625458e-05,
      "loss": 0.0225,
      "step": 1660
    },
    {
      "epoch": 2.5891472868217056,
      "grad_norm": 0.3459230661392212,
      "learning_rate": 5.561743328448296e-05,
      "loss": 0.0268,
      "step": 1670
    },
    {
      "epoch": 2.604651162790698,
      "grad_norm": 0.1962781399488449,
      "learning_rate": 5.5079260950344035e-05,
      "loss": 0.0197,
      "step": 1680
    },
    {
      "epoch": 2.62015503875969,
      "grad_norm": 0.19597306847572327,
      "learning_rate": 5.454049336372531e-05,
      "loss": 0.0215,
      "step": 1690
    },
    {
      "epoch": 2.6356589147286824,
      "grad_norm": 0.39598584175109863,
      "learning_rate": 5.4001193664273454e-05,
      "loss": 0.0161,
      "step": 1700
    },
    {
      "epoch": 2.6356589147286824,
      "eval_loss": 0.03870782628655434,
      "eval_runtime": 175.9342,
      "eval_samples_per_second": 6.519,
      "eval_steps_per_second": 6.519,
      "step": 1700
    },
    {
      "epoch": 2.6511627906976747,
      "grad_norm": 0.19037537276744843,
      "learning_rate": 5.346142505399495e-05,
      "loss": 0.0189,
      "step": 1710
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.3089650273323059,
      "learning_rate": 5.292125078984925e-05,
      "loss": 0.0191,
      "step": 1720
    },
    {
      "epoch": 2.682170542635659,
      "grad_norm": 0.12442419677972794,
      "learning_rate": 5.2380734176335425e-05,
      "loss": 0.0182,
      "step": 1730
    },
    {
      "epoch": 2.697674418604651,
      "grad_norm": 0.21611304581165314,
      "learning_rate": 5.183993855807343e-05,
      "loss": 0.0222,
      "step": 1740
    },
    {
      "epoch": 2.7131782945736433,
      "grad_norm": 0.0941254049539566,
      "learning_rate": 5.1298927312380586e-05,
      "loss": 0.0174,
      "step": 1750
    },
    {
      "epoch": 2.7131782945736433,
      "eval_loss": 0.03829415142536163,
      "eval_runtime": 176.3284,
      "eval_samples_per_second": 6.505,
      "eval_steps_per_second": 6.505,
      "step": 1750
    },
    {
      "epoch": 2.7286821705426356,
      "grad_norm": 0.2699877917766571,
      "learning_rate": 5.075776384184411e-05,
      "loss": 0.0298,
      "step": 1760
    },
    {
      "epoch": 2.744186046511628,
      "grad_norm": 0.2365487664937973,
      "learning_rate": 5.021651156689094e-05,
      "loss": 0.0294,
      "step": 1770
    },
    {
      "epoch": 2.75968992248062,
      "grad_norm": 0.26761698722839355,
      "learning_rate": 4.967523391835521e-05,
      "loss": 0.0264,
      "step": 1780
    },
    {
      "epoch": 2.7751937984496124,
      "grad_norm": 0.4799739122390747,
      "learning_rate": 4.9133994330044644e-05,
      "loss": 0.0245,
      "step": 1790
    },
    {
      "epoch": 2.7906976744186047,
      "grad_norm": 0.4386084973812103,
      "learning_rate": 4.85928562313066e-05,
      "loss": 0.0222,
      "step": 1800
    },
    {
      "epoch": 2.7906976744186047,
      "eval_loss": 0.0363197959959507,
      "eval_runtime": 176.3046,
      "eval_samples_per_second": 6.506,
      "eval_steps_per_second": 6.506,
      "step": 1800
    },
    {
      "epoch": 2.806201550387597,
      "grad_norm": 0.40001216530799866,
      "learning_rate": 4.8051883039594616e-05,
      "loss": 0.0167,
      "step": 1810
    },
    {
      "epoch": 2.8217054263565893,
      "grad_norm": 0.33203211426734924,
      "learning_rate": 4.751113815303624e-05,
      "loss": 0.0263,
      "step": 1820
    },
    {
      "epoch": 2.8372093023255816,
      "grad_norm": 0.6366134285926819,
      "learning_rate": 4.697068494300343e-05,
      "loss": 0.0197,
      "step": 1830
    },
    {
      "epoch": 2.8527131782945734,
      "grad_norm": 0.321431428194046,
      "learning_rate": 4.6430586746685724e-05,
      "loss": 0.024,
      "step": 1840
    },
    {
      "epoch": 2.8682170542635657,
      "grad_norm": 0.30680930614471436,
      "learning_rate": 4.589090685966758e-05,
      "loss": 0.0217,
      "step": 1850
    },
    {
      "epoch": 2.8682170542635657,
      "eval_loss": 0.03761348873376846,
      "eval_runtime": 176.3984,
      "eval_samples_per_second": 6.502,
      "eval_steps_per_second": 6.502,
      "step": 1850
    },
    {
      "epoch": 2.883720930232558,
      "grad_norm": 0.30222904682159424,
      "learning_rate": 4.535170852851073e-05,
      "loss": 0.0327,
      "step": 1860
    },
    {
      "epoch": 2.89922480620155,
      "grad_norm": 0.5061004757881165,
      "learning_rate": 4.481305494334201e-05,
      "loss": 0.0146,
      "step": 1870
    },
    {
      "epoch": 2.9147286821705425,
      "grad_norm": 0.1436406373977661,
      "learning_rate": 4.427500923044801e-05,
      "loss": 0.0211,
      "step": 1880
    },
    {
      "epoch": 2.9302325581395348,
      "grad_norm": 0.32733285427093506,
      "learning_rate": 4.373763444487705e-05,
      "loss": 0.0186,
      "step": 1890
    },
    {
      "epoch": 2.945736434108527,
      "grad_norm": 0.1981613039970398,
      "learning_rate": 4.3200993563049725e-05,
      "loss": 0.0198,
      "step": 1900
    },
    {
      "epoch": 2.945736434108527,
      "eval_loss": 0.0361945740878582,
      "eval_runtime": 176.4884,
      "eval_samples_per_second": 6.499,
      "eval_steps_per_second": 6.499,
      "step": 1900
    },
    {
      "epoch": 2.9612403100775193,
      "grad_norm": 0.22807447612285614,
      "learning_rate": 4.266514947537839e-05,
      "loss": 0.0195,
      "step": 1910
    },
    {
      "epoch": 2.9767441860465116,
      "grad_norm": 0.30690813064575195,
      "learning_rate": 4.2130164978896916e-05,
      "loss": 0.0236,
      "step": 1920
    },
    {
      "epoch": 2.992248062015504,
      "grad_norm": 0.02489878423511982,
      "learning_rate": 4.159610276990137e-05,
      "loss": 0.0148,
      "step": 1930
    },
    {
      "epoch": 3.007751937984496,
      "grad_norm": 0.15840907394886017,
      "learning_rate": 4.106302543660235e-05,
      "loss": 0.0154,
      "step": 1940
    },
    {
      "epoch": 3.0232558139534884,
      "grad_norm": 0.14962005615234375,
      "learning_rate": 4.053099545179028e-05,
      "loss": 0.0105,
      "step": 1950
    },
    {
      "epoch": 3.0232558139534884,
      "eval_loss": 0.03878239914774895,
      "eval_runtime": 176.4369,
      "eval_samples_per_second": 6.501,
      "eval_steps_per_second": 6.501,
      "step": 1950
    },
    {
      "epoch": 3.0387596899224807,
      "grad_norm": 0.21889622509479523,
      "learning_rate": 4.0000075165513845e-05,
      "loss": 0.0141,
      "step": 1960
    },
    {
      "epoch": 3.054263565891473,
      "grad_norm": 0.1822437047958374,
      "learning_rate": 3.9470326797773216e-05,
      "loss": 0.0084,
      "step": 1970
    },
    {
      "epoch": 3.0697674418604652,
      "grad_norm": 0.4426482319831848,
      "learning_rate": 3.8941812431228166e-05,
      "loss": 0.0113,
      "step": 1980
    },
    {
      "epoch": 3.0852713178294575,
      "grad_norm": 0.27718520164489746,
      "learning_rate": 3.8414594003922515e-05,
      "loss": 0.0156,
      "step": 1990
    },
    {
      "epoch": 3.10077519379845,
      "grad_norm": 0.08083080500364304,
      "learning_rate": 3.788873330202544e-05,
      "loss": 0.0097,
      "step": 2000
    },
    {
      "epoch": 3.10077519379845,
      "eval_loss": 0.043953340500593185,
      "eval_runtime": 176.3739,
      "eval_samples_per_second": 6.503,
      "eval_steps_per_second": 6.503,
      "step": 2000
    },
    {
      "epoch": 3.116279069767442,
      "grad_norm": 0.519051730632782,
      "learning_rate": 3.736429195259051e-05,
      "loss": 0.0057,
      "step": 2010
    },
    {
      "epoch": 3.1317829457364343,
      "grad_norm": 0.356781005859375,
      "learning_rate": 3.684133141633358e-05,
      "loss": 0.0105,
      "step": 2020
    },
    {
      "epoch": 3.147286821705426,
      "grad_norm": 0.13736553490161896,
      "learning_rate": 3.6319912980429846e-05,
      "loss": 0.0109,
      "step": 2030
    },
    {
      "epoch": 3.1627906976744184,
      "grad_norm": 0.1859646737575531,
      "learning_rate": 3.580009775133168e-05,
      "loss": 0.009,
      "step": 2040
    },
    {
      "epoch": 3.1782945736434107,
      "grad_norm": 0.12270783632993698,
      "learning_rate": 3.528194664760714e-05,
      "loss": 0.008,
      "step": 2050
    },
    {
      "epoch": 3.1782945736434107,
      "eval_loss": 0.04818466678261757,
      "eval_runtime": 176.2871,
      "eval_samples_per_second": 6.506,
      "eval_steps_per_second": 6.506,
      "step": 2050
    },
    {
      "epoch": 3.193798449612403,
      "grad_norm": 0.20273350179195404,
      "learning_rate": 3.476552039280096e-05,
      "loss": 0.0104,
      "step": 2060
    },
    {
      "epoch": 3.2093023255813953,
      "grad_norm": 0.1353200376033783,
      "learning_rate": 3.42508795083181e-05,
      "loss": 0.0106,
      "step": 2070
    },
    {
      "epoch": 3.2248062015503876,
      "grad_norm": 0.3596494197845459,
      "learning_rate": 3.373808430633106e-05,
      "loss": 0.01,
      "step": 2080
    },
    {
      "epoch": 3.24031007751938,
      "grad_norm": 0.6581428647041321,
      "learning_rate": 3.32271948827118e-05,
      "loss": 0.0105,
      "step": 2090
    },
    {
      "epoch": 3.255813953488372,
      "grad_norm": 0.2087380737066269,
      "learning_rate": 3.2718271109988863e-05,
      "loss": 0.0114,
      "step": 2100
    },
    {
      "epoch": 3.255813953488372,
      "eval_loss": 0.043491754680871964,
      "eval_runtime": 176.0628,
      "eval_samples_per_second": 6.515,
      "eval_steps_per_second": 6.515,
      "step": 2100
    },
    {
      "epoch": 3.2713178294573644,
      "grad_norm": 0.3167926073074341,
      "learning_rate": 3.2211372630330835e-05,
      "loss": 0.0093,
      "step": 2110
    },
    {
      "epoch": 3.2868217054263567,
      "grad_norm": 0.3298277258872986,
      "learning_rate": 3.170655884855661e-05,
      "loss": 0.0162,
      "step": 2120
    },
    {
      "epoch": 3.302325581395349,
      "grad_norm": 0.2147381752729416,
      "learning_rate": 3.120388892517368e-05,
      "loss": 0.0182,
      "step": 2130
    },
    {
      "epoch": 3.317829457364341,
      "grad_norm": 0.4693920314311981,
      "learning_rate": 3.070342176944494e-05,
      "loss": 0.0147,
      "step": 2140
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.10257305949926376,
      "learning_rate": 3.0205216032484805e-05,
      "loss": 0.0112,
      "step": 2150
    },
    {
      "epoch": 3.3333333333333335,
      "eval_loss": 0.040094662457704544,
      "eval_runtime": 176.3863,
      "eval_samples_per_second": 6.503,
      "eval_steps_per_second": 6.503,
      "step": 2150
    },
    {
      "epoch": 3.3488372093023258,
      "grad_norm": 0.17424806952476501,
      "learning_rate": 2.970933010038599e-05,
      "loss": 0.0153,
      "step": 2160
    },
    {
      "epoch": 3.3643410852713176,
      "grad_norm": 0.2841956317424774,
      "learning_rate": 2.921582208737681e-05,
      "loss": 0.0101,
      "step": 2170
    },
    {
      "epoch": 3.37984496124031,
      "grad_norm": 0.20709052681922913,
      "learning_rate": 2.872474982901081e-05,
      "loss": 0.0121,
      "step": 2180
    },
    {
      "epoch": 3.395348837209302,
      "grad_norm": 0.20538711547851562,
      "learning_rate": 2.8236170875388744e-05,
      "loss": 0.0139,
      "step": 2190
    },
    {
      "epoch": 3.4108527131782944,
      "grad_norm": 0.4432328939437866,
      "learning_rate": 2.775014248441422e-05,
      "loss": 0.0079,
      "step": 2200
    },
    {
      "epoch": 3.4108527131782944,
      "eval_loss": 0.04478052258491516,
      "eval_runtime": 176.5264,
      "eval_samples_per_second": 6.498,
      "eval_steps_per_second": 6.498,
      "step": 2200
    },
    {
      "epoch": 3.4263565891472867,
      "grad_norm": 0.1635671854019165,
      "learning_rate": 2.726672161508341e-05,
      "loss": 0.01,
      "step": 2210
    },
    {
      "epoch": 3.441860465116279,
      "grad_norm": 0.39635413885116577,
      "learning_rate": 2.678596492080984e-05,
      "loss": 0.0087,
      "step": 2220
    },
    {
      "epoch": 3.4573643410852712,
      "grad_norm": 0.4210352599620819,
      "learning_rate": 2.630792874278516e-05,
      "loss": 0.0081,
      "step": 2230
    },
    {
      "epoch": 3.4728682170542635,
      "grad_norm": 0.5623159408569336,
      "learning_rate": 2.583266910337624e-05,
      "loss": 0.0084,
      "step": 2240
    },
    {
      "epoch": 3.488372093023256,
      "grad_norm": 0.9574199318885803,
      "learning_rate": 2.5360241699559816e-05,
      "loss": 0.0133,
      "step": 2250
    },
    {
      "epoch": 3.488372093023256,
      "eval_loss": 0.04834475368261337,
      "eval_runtime": 176.6597,
      "eval_samples_per_second": 6.493,
      "eval_steps_per_second": 6.493,
      "step": 2250
    },
    {
      "epoch": 3.503875968992248,
      "grad_norm": 0.19124262034893036,
      "learning_rate": 2.4890701896395146e-05,
      "loss": 0.0129,
      "step": 2260
    },
    {
      "epoch": 3.5193798449612403,
      "grad_norm": 0.27771103382110596,
      "learning_rate": 2.4424104720535735e-05,
      "loss": 0.0132,
      "step": 2270
    },
    {
      "epoch": 3.5348837209302326,
      "grad_norm": 0.3095415532588959,
      "learning_rate": 2.3960504853780462e-05,
      "loss": 0.0107,
      "step": 2280
    },
    {
      "epoch": 3.550387596899225,
      "grad_norm": 0.11446672677993774,
      "learning_rate": 2.349995662666547e-05,
      "loss": 0.0061,
      "step": 2290
    },
    {
      "epoch": 3.565891472868217,
      "grad_norm": 0.244533509016037,
      "learning_rate": 2.3042514012096843e-05,
      "loss": 0.0128,
      "step": 2300
    },
    {
      "epoch": 3.565891472868217,
      "eval_loss": 0.04713751748204231,
      "eval_runtime": 176.5166,
      "eval_samples_per_second": 6.498,
      "eval_steps_per_second": 6.498,
      "step": 2300
    },
    {
      "epoch": 3.5813953488372094,
      "grad_norm": 0.5505859851837158,
      "learning_rate": 2.2588230619025407e-05,
      "loss": 0.0182,
      "step": 2310
    },
    {
      "epoch": 3.5968992248062017,
      "grad_norm": 0.2529643177986145,
      "learning_rate": 2.213715968616425e-05,
      "loss": 0.0122,
      "step": 2320
    },
    {
      "epoch": 3.612403100775194,
      "grad_norm": 0.16935287415981293,
      "learning_rate": 2.16893540757494e-05,
      "loss": 0.0094,
      "step": 2330
    },
    {
      "epoch": 3.6279069767441863,
      "grad_norm": 0.3461015820503235,
      "learning_rate": 2.1244866267344866e-05,
      "loss": 0.0094,
      "step": 2340
    },
    {
      "epoch": 3.6434108527131785,
      "grad_norm": 0.4279143214225769,
      "learning_rate": 2.080374835169235e-05,
      "loss": 0.012,
      "step": 2350
    },
    {
      "epoch": 3.6434108527131785,
      "eval_loss": 0.04670505225658417,
      "eval_runtime": 176.4379,
      "eval_samples_per_second": 6.501,
      "eval_steps_per_second": 6.501,
      "step": 2350
    },
    {
      "epoch": 3.6589147286821704,
      "grad_norm": 0.43128836154937744,
      "learning_rate": 2.0366052024606612e-05,
      "loss": 0.0116,
      "step": 2360
    },
    {
      "epoch": 3.6744186046511627,
      "grad_norm": 0.36545148491859436,
      "learning_rate": 1.9931828580917107e-05,
      "loss": 0.0097,
      "step": 2370
    },
    {
      "epoch": 3.689922480620155,
      "grad_norm": 0.4567084312438965,
      "learning_rate": 1.9501128908456523e-05,
      "loss": 0.0107,
      "step": 2380
    },
    {
      "epoch": 3.705426356589147,
      "grad_norm": 0.22181279957294464,
      "learning_rate": 1.9074003482097258e-05,
      "loss": 0.0083,
      "step": 2390
    },
    {
      "epoch": 3.7209302325581395,
      "grad_norm": 0.6979092359542847,
      "learning_rate": 1.8650502357835925e-05,
      "loss": 0.0143,
      "step": 2400
    },
    {
      "epoch": 3.7209302325581395,
      "eval_loss": 0.04637313261628151,
      "eval_runtime": 176.2066,
      "eval_samples_per_second": 6.509,
      "eval_steps_per_second": 6.509,
      "step": 2400
    },
    {
      "epoch": 3.7364341085271318,
      "grad_norm": 0.3935830891132355,
      "learning_rate": 1.82306751669274e-05,
      "loss": 0.0055,
      "step": 2410
    },
    {
      "epoch": 3.751937984496124,
      "grad_norm": 0.8239232301712036,
      "learning_rate": 1.7814571110068135e-05,
      "loss": 0.0141,
      "step": 2420
    },
    {
      "epoch": 3.7674418604651163,
      "grad_norm": 0.2028406262397766,
      "learning_rate": 1.740223895163039e-05,
      "loss": 0.0095,
      "step": 2430
    },
    {
      "epoch": 3.7829457364341086,
      "grad_norm": 0.11796276271343231,
      "learning_rate": 1.6993727013947336e-05,
      "loss": 0.0121,
      "step": 2440
    },
    {
      "epoch": 3.798449612403101,
      "grad_norm": 0.46211379766464233,
      "learning_rate": 1.6589083171649977e-05,
      "loss": 0.0063,
      "step": 2450
    },
    {
      "epoch": 3.798449612403101,
      "eval_loss": 0.048657044768333435,
      "eval_runtime": 176.2691,
      "eval_samples_per_second": 6.507,
      "eval_steps_per_second": 6.507,
      "step": 2450
    },
    {
      "epoch": 3.813953488372093,
      "grad_norm": 0.41259831190109253,
      "learning_rate": 1.6188354846056698e-05,
      "loss": 0.0102,
      "step": 2460
    },
    {
      "epoch": 3.8294573643410854,
      "grad_norm": 0.2538934648036957,
      "learning_rate": 1.579158899961575e-05,
      "loss": 0.0067,
      "step": 2470
    },
    {
      "epoch": 3.8449612403100772,
      "grad_norm": 0.7502545714378357,
      "learning_rate": 1.5398832130401637e-05,
      "loss": 0.0109,
      "step": 2480
    },
    {
      "epoch": 3.8604651162790695,
      "grad_norm": 0.22138559818267822,
      "learning_rate": 1.5010130266665807e-05,
      "loss": 0.012,
      "step": 2490
    },
    {
      "epoch": 3.875968992248062,
      "grad_norm": 0.4992906153202057,
      "learning_rate": 1.4625528961442591e-05,
      "loss": 0.0098,
      "step": 2500
    },
    {
      "epoch": 3.875968992248062,
      "eval_loss": 0.0469583123922348,
      "eval_runtime": 176.489,
      "eval_samples_per_second": 6.499,
      "eval_steps_per_second": 6.499,
      "step": 2500
    },
    {
      "epoch": 3.891472868217054,
      "grad_norm": 0.332923948764801,
      "learning_rate": 1.4245073287210608e-05,
      "loss": 0.0111,
      "step": 2510
    },
    {
      "epoch": 3.9069767441860463,
      "grad_norm": 0.5175971388816833,
      "learning_rate": 1.3868807830610653e-05,
      "loss": 0.0165,
      "step": 2520
    },
    {
      "epoch": 3.9224806201550386,
      "grad_norm": 0.21356546878814697,
      "learning_rate": 1.3496776687220514e-05,
      "loss": 0.0158,
      "step": 2530
    },
    {
      "epoch": 3.937984496124031,
      "grad_norm": 0.2587060034275055,
      "learning_rate": 1.3129023456387151e-05,
      "loss": 0.0049,
      "step": 2540
    },
    {
      "epoch": 3.953488372093023,
      "grad_norm": 0.22560270130634308,
      "learning_rate": 1.2765591236117324e-05,
      "loss": 0.0088,
      "step": 2550
    },
    {
      "epoch": 3.953488372093023,
      "eval_loss": 0.046130869537591934,
      "eval_runtime": 175.5058,
      "eval_samples_per_second": 6.535,
      "eval_steps_per_second": 6.535,
      "step": 2550
    },
    {
      "epoch": 3.9689922480620154,
      "grad_norm": 0.27367764711380005,
      "learning_rate": 1.240652261802669e-05,
      "loss": 0.0097,
      "step": 2560
    },
    {
      "epoch": 3.9844961240310077,
      "grad_norm": 0.47304585576057434,
      "learning_rate": 1.205185968234847e-05,
      "loss": 0.0143,
      "step": 2570
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7547179460525513,
      "learning_rate": 1.1701643993001916e-05,
      "loss": 0.0122,
      "step": 2580
    },
    {
      "epoch": 4.015503875968992,
      "grad_norm": 0.09287036210298538,
      "learning_rate": 1.1355916592721316e-05,
      "loss": 0.0043,
      "step": 2590
    },
    {
      "epoch": 4.0310077519379846,
      "grad_norm": 0.20291487872600555,
      "learning_rate": 1.1014717998246099e-05,
      "loss": 0.0057,
      "step": 2600
    },
    {
      "epoch": 4.0310077519379846,
      "eval_loss": 0.046690039336681366,
      "eval_runtime": 176.7343,
      "eval_samples_per_second": 6.49,
      "eval_steps_per_second": 6.49,
      "step": 2600
    },
    {
      "epoch": 4.046511627906977,
      "grad_norm": 0.11436055600643158,
      "learning_rate": 1.0678088195572517e-05,
      "loss": 0.0049,
      "step": 2610
    },
    {
      "epoch": 4.062015503875969,
      "grad_norm": 0.12677472829818726,
      "learning_rate": 1.0346066635267676e-05,
      "loss": 0.0038,
      "step": 2620
    },
    {
      "epoch": 4.077519379844961,
      "grad_norm": 0.0559503436088562,
      "learning_rate": 1.001869222784611e-05,
      "loss": 0.0064,
      "step": 2630
    },
    {
      "epoch": 4.093023255813954,
      "grad_norm": 0.20214763283729553,
      "learning_rate": 9.696003339209819e-06,
      "loss": 0.0081,
      "step": 2640
    },
    {
      "epoch": 4.108527131782946,
      "grad_norm": 0.6340795755386353,
      "learning_rate": 9.378037786152055e-06,
      "loss": 0.006,
      "step": 2650
    },
    {
      "epoch": 4.108527131782946,
      "eval_loss": 0.0505899041891098,
      "eval_runtime": 177.0154,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 2650
    },
    {
      "epoch": 4.124031007751938,
      "grad_norm": 0.2011650651693344,
      "learning_rate": 9.064832831925441e-06,
      "loss": 0.0036,
      "step": 2660
    },
    {
      "epoch": 4.1395348837209305,
      "grad_norm": 0.054464973509311676,
      "learning_rate": 8.756425181875028e-06,
      "loss": 0.0029,
      "step": 2670
    },
    {
      "epoch": 4.155038759689923,
      "grad_norm": 0.2652602195739746,
      "learning_rate": 8.452850979136617e-06,
      "loss": 0.004,
      "step": 2680
    },
    {
      "epoch": 4.170542635658915,
      "grad_norm": 0.04914356395602226,
      "learning_rate": 8.15414580040117e-06,
      "loss": 0.0018,
      "step": 2690
    },
    {
      "epoch": 4.186046511627907,
      "grad_norm": 0.16105438768863678,
      "learning_rate": 7.860344651745366e-06,
      "loss": 0.0058,
      "step": 2700
    },
    {
      "epoch": 4.186046511627907,
      "eval_loss": 0.0564822219312191,
      "eval_runtime": 177.0095,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 2700
    },
    {
      "epoch": 4.2015503875969,
      "grad_norm": 0.0615265853703022,
      "learning_rate": 7.571481964529226e-06,
      "loss": 0.004,
      "step": 2710
    },
    {
      "epoch": 4.217054263565892,
      "grad_norm": 0.23362399637699127,
      "learning_rate": 7.287591591360926e-06,
      "loss": 0.0021,
      "step": 2720
    },
    {
      "epoch": 4.232558139534884,
      "grad_norm": 0.14505736529827118,
      "learning_rate": 7.008706802129606e-06,
      "loss": 0.0066,
      "step": 2730
    },
    {
      "epoch": 4.248062015503876,
      "grad_norm": 0.0062782918103039265,
      "learning_rate": 6.734860280106292e-06,
      "loss": 0.0022,
      "step": 2740
    },
    {
      "epoch": 4.263565891472869,
      "grad_norm": 0.3620900809764862,
      "learning_rate": 6.466084118113735e-06,
      "loss": 0.0057,
      "step": 2750
    },
    {
      "epoch": 4.263565891472869,
      "eval_loss": 0.059196386486291885,
      "eval_runtime": 177.1851,
      "eval_samples_per_second": 6.473,
      "eval_steps_per_second": 6.473,
      "step": 2750
    },
    {
      "epoch": 4.27906976744186,
      "grad_norm": 0.2628999352455139,
      "learning_rate": 6.202409814765325e-06,
      "loss": 0.0048,
      "step": 2760
    },
    {
      "epoch": 4.294573643410852,
      "grad_norm": 0.21285632252693176,
      "learning_rate": 5.9438682707736725e-06,
      "loss": 0.0068,
      "step": 2770
    },
    {
      "epoch": 4.310077519379845,
      "grad_norm": 0.27160316705703735,
      "learning_rate": 5.690489785329301e-06,
      "loss": 0.0053,
      "step": 2780
    },
    {
      "epoch": 4.325581395348837,
      "grad_norm": 0.010958620347082615,
      "learning_rate": 5.442304052549752e-06,
      "loss": 0.0039,
      "step": 2790
    },
    {
      "epoch": 4.341085271317829,
      "grad_norm": 0.10000675171613693,
      "learning_rate": 5.199340157999733e-06,
      "loss": 0.0043,
      "step": 2800
    },
    {
      "epoch": 4.341085271317829,
      "eval_loss": 0.05947728455066681,
      "eval_runtime": 177.3633,
      "eval_samples_per_second": 6.467,
      "eval_steps_per_second": 6.467,
      "step": 2800
    },
    {
      "epoch": 4.3565891472868215,
      "grad_norm": 0.08185016363859177,
      "learning_rate": 4.961626575282396e-06,
      "loss": 0.0026,
      "step": 2810
    },
    {
      "epoch": 4.372093023255814,
      "grad_norm": 0.3243257403373718,
      "learning_rate": 4.7291911627025235e-06,
      "loss": 0.0051,
      "step": 2820
    },
    {
      "epoch": 4.387596899224806,
      "grad_norm": 0.22814767062664032,
      "learning_rate": 4.502061160001725e-06,
      "loss": 0.002,
      "step": 2830
    },
    {
      "epoch": 4.403100775193798,
      "grad_norm": 0.04655148833990097,
      "learning_rate": 4.280263185166078e-06,
      "loss": 0.0045,
      "step": 2840
    },
    {
      "epoch": 4.4186046511627906,
      "grad_norm": 0.3352436423301697,
      "learning_rate": 4.063823231306757e-06,
      "loss": 0.0043,
      "step": 2850
    },
    {
      "epoch": 4.4186046511627906,
      "eval_loss": 0.061167895793914795,
      "eval_runtime": 177.4729,
      "eval_samples_per_second": 6.463,
      "eval_steps_per_second": 6.463,
      "step": 2850
    },
    {
      "epoch": 4.434108527131783,
      "grad_norm": 0.03668573871254921,
      "learning_rate": 3.8527666636137885e-06,
      "loss": 0.0038,
      "step": 2860
    },
    {
      "epoch": 4.449612403100775,
      "grad_norm": 0.07714797556400299,
      "learning_rate": 3.6471182163834825e-06,
      "loss": 0.0047,
      "step": 2870
    },
    {
      "epoch": 4.465116279069767,
      "grad_norm": 0.25801295042037964,
      "learning_rate": 3.4469019901197054e-06,
      "loss": 0.0027,
      "step": 2880
    },
    {
      "epoch": 4.48062015503876,
      "grad_norm": 0.06319399923086166,
      "learning_rate": 3.252141448709495e-06,
      "loss": 0.0033,
      "step": 2890
    },
    {
      "epoch": 4.496124031007752,
      "grad_norm": 0.1602359414100647,
      "learning_rate": 3.0628594166732693e-06,
      "loss": 0.0055,
      "step": 2900
    },
    {
      "epoch": 4.496124031007752,
      "eval_loss": 0.061658285558223724,
      "eval_runtime": 177.7423,
      "eval_samples_per_second": 6.453,
      "eval_steps_per_second": 6.453,
      "step": 2900
    },
    {
      "epoch": 4.511627906976744,
      "grad_norm": 0.04129059985280037,
      "learning_rate": 2.8790780764899384e-06,
      "loss": 0.0056,
      "step": 2910
    },
    {
      "epoch": 4.5271317829457365,
      "grad_norm": 0.07882144302129745,
      "learning_rate": 2.700818965997315e-06,
      "loss": 0.0089,
      "step": 2920
    },
    {
      "epoch": 4.542635658914729,
      "grad_norm": 0.019726574420928955,
      "learning_rate": 2.528102975867991e-06,
      "loss": 0.002,
      "step": 2930
    },
    {
      "epoch": 4.558139534883721,
      "grad_norm": 0.036272820085287094,
      "learning_rate": 2.3609503471611284e-06,
      "loss": 0.0025,
      "step": 2940
    },
    {
      "epoch": 4.573643410852713,
      "grad_norm": 0.2767897844314575,
      "learning_rate": 2.1993806689503738e-06,
      "loss": 0.0029,
      "step": 2950
    },
    {
      "epoch": 4.573643410852713,
      "eval_loss": 0.06082376092672348,
      "eval_runtime": 177.687,
      "eval_samples_per_second": 6.455,
      "eval_steps_per_second": 6.455,
      "step": 2950
    },
    {
      "epoch": 4.589147286821706,
      "grad_norm": 0.02878037840127945,
      "learning_rate": 2.0434128760281167e-06,
      "loss": 0.0019,
      "step": 2960
    },
    {
      "epoch": 4.604651162790698,
      "grad_norm": 0.09522771835327148,
      "learning_rate": 1.893065246686504e-06,
      "loss": 0.002,
      "step": 2970
    },
    {
      "epoch": 4.62015503875969,
      "grad_norm": 0.007287740241736174,
      "learning_rate": 1.74835540057533e-06,
      "loss": 0.005,
      "step": 2980
    },
    {
      "epoch": 4.635658914728682,
      "grad_norm": 0.1458829939365387,
      "learning_rate": 1.6093002966371617e-06,
      "loss": 0.0017,
      "step": 2990
    },
    {
      "epoch": 4.651162790697675,
      "grad_norm": 0.18788699805736542,
      "learning_rate": 1.4759162311198783e-06,
      "loss": 0.0029,
      "step": 3000
    },
    {
      "epoch": 4.651162790697675,
      "eval_loss": 0.06145617738366127,
      "eval_runtime": 177.5386,
      "eval_samples_per_second": 6.461,
      "eval_steps_per_second": 6.461,
      "step": 3000
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.06372984498739243,
      "learning_rate": 1.348218835666859e-06,
      "loss": 0.0037,
      "step": 3010
    },
    {
      "epoch": 4.682170542635659,
      "grad_norm": 0.02428070269525051,
      "learning_rate": 1.2262230754850445e-06,
      "loss": 0.0032,
      "step": 3020
    },
    {
      "epoch": 4.6976744186046515,
      "grad_norm": 0.06479329615831375,
      "learning_rate": 1.109943247591172e-06,
      "loss": 0.0059,
      "step": 3030
    },
    {
      "epoch": 4.713178294573644,
      "grad_norm": 0.18652325868606567,
      "learning_rate": 9.993929791362323e-07,
      "loss": 0.0041,
      "step": 3040
    },
    {
      "epoch": 4.728682170542635,
      "grad_norm": 0.1359170526266098,
      "learning_rate": 8.945852258084863e-07,
      "loss": 0.0039,
      "step": 3050
    },
    {
      "epoch": 4.728682170542635,
      "eval_loss": 0.061951201409101486,
      "eval_runtime": 177.3573,
      "eval_samples_per_second": 6.467,
      "eval_steps_per_second": 6.467,
      "step": 3050
    },
    {
      "epoch": 4.7441860465116275,
      "grad_norm": 0.5252036452293396,
      "learning_rate": 7.955322703151358e-07,
      "loss": 0.008,
      "step": 3060
    },
    {
      "epoch": 4.75968992248062,
      "grad_norm": 0.3898888826370239,
      "learning_rate": 7.022457209428901e-07,
      "loss": 0.0068,
      "step": 3070
    },
    {
      "epoch": 4.775193798449612,
      "grad_norm": 0.22847306728363037,
      "learning_rate": 6.147365101975666e-07,
      "loss": 0.003,
      "step": 3080
    },
    {
      "epoch": 4.790697674418604,
      "grad_norm": 0.12396875023841858,
      "learning_rate": 5.33014893522854e-07,
      "loss": 0.0046,
      "step": 3090
    },
    {
      "epoch": 4.8062015503875966,
      "grad_norm": 0.335860013961792,
      "learning_rate": 4.57090448098485e-07,
      "loss": 0.0034,
      "step": 3100
    },
    {
      "epoch": 4.8062015503875966,
      "eval_loss": 0.06197270005941391,
      "eval_runtime": 177.4972,
      "eval_samples_per_second": 6.462,
      "eval_steps_per_second": 6.462,
      "step": 3100
    },
    {
      "epoch": 4.821705426356589,
      "grad_norm": 0.04163197800517082,
      "learning_rate": 3.8697207171781714e-07,
      "loss": 0.0044,
      "step": 3110
    },
    {
      "epoch": 4.837209302325581,
      "grad_norm": 0.05860529839992523,
      "learning_rate": 3.2266798174512837e-07,
      "loss": 0.0052,
      "step": 3120
    },
    {
      "epoch": 4.852713178294573,
      "grad_norm": 0.08096914738416672,
      "learning_rate": 2.6418571415255387e-07,
      "loss": 0.0043,
      "step": 3130
    },
    {
      "epoch": 4.868217054263566,
      "grad_norm": 0.07322604209184647,
      "learning_rate": 2.1153212263695378e-07,
      "loss": 0.0051,
      "step": 3140
    },
    {
      "epoch": 4.883720930232558,
      "grad_norm": 0.3177713453769684,
      "learning_rate": 1.6471337781669982e-07,
      "loss": 0.0061,
      "step": 3150
    },
    {
      "epoch": 4.883720930232558,
      "eval_loss": 0.06184118613600731,
      "eval_runtime": 177.919,
      "eval_samples_per_second": 6.447,
      "eval_steps_per_second": 6.447,
      "step": 3150
    },
    {
      "epoch": 4.89922480620155,
      "grad_norm": 0.1601889431476593,
      "learning_rate": 1.237349665085097e-07,
      "loss": 0.0066,
      "step": 3160
    },
    {
      "epoch": 4.9147286821705425,
      "grad_norm": 0.1714223474264145,
      "learning_rate": 8.86016910844667e-08,
      "loss": 0.0065,
      "step": 3170
    },
    {
      "epoch": 4.930232558139535,
      "grad_norm": 0.9038301706314087,
      "learning_rate": 5.9317668909192323e-08,
      "loss": 0.006,
      "step": 3180
    },
    {
      "epoch": 4.945736434108527,
      "grad_norm": 0.08108743280172348,
      "learning_rate": 3.588633185730994e-08,
      "loss": 0.0082,
      "step": 3190
    },
    {
      "epoch": 4.961240310077519,
      "grad_norm": 0.058073218911886215,
      "learning_rate": 1.8310425911294283e-08,
      "loss": 0.0045,
      "step": 3200
    },
    {
      "epoch": 4.961240310077519,
      "eval_loss": 0.061884406954050064,
      "eval_runtime": 177.8676,
      "eval_samples_per_second": 6.449,
      "eval_steps_per_second": 6.449,
      "step": 3200
    },
    {
      "epoch": 4.976744186046512,
      "grad_norm": 0.48708316683769226,
      "learning_rate": 6.592010839612251e-09,
      "loss": 0.0074,
      "step": 3210
    },
    {
      "epoch": 4.992248062015504,
      "grad_norm": 0.044914498925209045,
      "learning_rate": 7.324599553770739e-10,
      "loss": 0.0016,
      "step": 3220
    },
    {
      "epoch": 5.0,
      "step": 3225,
      "total_flos": 8.096268812599296e+17,
      "train_loss": 0.04488553278003783,
      "train_runtime": 38904.1603,
      "train_samples_per_second": 1.326,
      "train_steps_per_second": 0.083
    }
  ],
  "logging_steps": 10,
  "max_steps": 3225,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.096268812599296e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}