{
  "best_metric": 0.030188467353582382,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold3/checkpoint-1950",
  "epoch": 4.9976479443033215,
  "eval_steps": 50,
  "global_step": 3320,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015053156458744943,
      "grad_norm": 1.822564721107483,
      "learning_rate": 3.0120481927710846e-06,
      "loss": 1.7049,
      "step": 10
    },
    {
      "epoch": 0.030106312917489886,
      "grad_norm": 2.226895570755005,
      "learning_rate": 6.024096385542169e-06,
      "loss": 1.6672,
      "step": 20
    },
    {
      "epoch": 0.04515946937623483,
      "grad_norm": 2.538222551345825,
      "learning_rate": 9.036144578313253e-06,
      "loss": 1.615,
      "step": 30
    },
    {
      "epoch": 0.06021262583497977,
      "grad_norm": 3.4118905067443848,
      "learning_rate": 1.2048192771084338e-05,
      "loss": 1.4092,
      "step": 40
    },
    {
      "epoch": 0.07526578229372471,
      "grad_norm": 0.845978856086731,
      "learning_rate": 1.5060240963855424e-05,
      "loss": 0.8176,
      "step": 50
    },
    {
      "epoch": 0.07526578229372471,
      "eval_loss": 0.6253208518028259,
      "eval_runtime": 178.2787,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 50
    },
    {
      "epoch": 0.09031893875246966,
      "grad_norm": 0.6405627727508545,
      "learning_rate": 1.8072289156626505e-05,
      "loss": 0.5568,
      "step": 60
    },
    {
      "epoch": 0.1053720952112146,
      "grad_norm": 0.9309800267219543,
      "learning_rate": 2.1084337349397593e-05,
      "loss": 0.4914,
      "step": 70
    },
    {
      "epoch": 0.12042525166995954,
      "grad_norm": 0.7481735348701477,
      "learning_rate": 2.4096385542168677e-05,
      "loss": 0.3461,
      "step": 80
    },
    {
      "epoch": 0.1354784081287045,
      "grad_norm": 0.840279757976532,
      "learning_rate": 2.7108433734939758e-05,
      "loss": 0.173,
      "step": 90
    },
    {
      "epoch": 0.15053156458744943,
      "grad_norm": 0.6460654735565186,
      "learning_rate": 3.012048192771085e-05,
      "loss": 0.149,
      "step": 100
    },
    {
      "epoch": 0.15053156458744943,
      "eval_loss": 0.12084457278251648,
      "eval_runtime": 178.2793,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 100
    },
    {
      "epoch": 0.16558472104619437,
      "grad_norm": 0.8036267757415771,
      "learning_rate": 3.313253012048193e-05,
      "loss": 0.1049,
      "step": 110
    },
    {
      "epoch": 0.18063787750493931,
      "grad_norm": 0.4858762323856354,
      "learning_rate": 3.614457831325301e-05,
      "loss": 0.1118,
      "step": 120
    },
    {
      "epoch": 0.19569103396368426,
      "grad_norm": 0.5866274833679199,
      "learning_rate": 3.91566265060241e-05,
      "loss": 0.091,
      "step": 130
    },
    {
      "epoch": 0.2107441904224292,
      "grad_norm": 0.6353292465209961,
      "learning_rate": 4.2168674698795186e-05,
      "loss": 0.0815,
      "step": 140
    },
    {
      "epoch": 0.22579734688117414,
      "grad_norm": 0.8361664414405823,
      "learning_rate": 4.5180722891566266e-05,
      "loss": 0.0858,
      "step": 150
    },
    {
      "epoch": 0.22579734688117414,
      "eval_loss": 0.0719056949019432,
      "eval_runtime": 178.4648,
      "eval_samples_per_second": 6.618,
      "eval_steps_per_second": 6.618,
      "step": 150
    },
    {
      "epoch": 0.2408505033399191,
      "grad_norm": 0.5231442451477051,
      "learning_rate": 4.8192771084337354e-05,
      "loss": 0.076,
      "step": 160
    },
    {
      "epoch": 0.25590365979866403,
      "grad_norm": 0.7356946468353271,
      "learning_rate": 5.120481927710844e-05,
      "loss": 0.0844,
      "step": 170
    },
    {
      "epoch": 0.270956816257409,
      "grad_norm": 0.7905665636062622,
      "learning_rate": 5.4216867469879516e-05,
      "loss": 0.0734,
      "step": 180
    },
    {
      "epoch": 0.2860099727161539,
      "grad_norm": 0.9920386075973511,
      "learning_rate": 5.72289156626506e-05,
      "loss": 0.0687,
      "step": 190
    },
    {
      "epoch": 0.30106312917489886,
      "grad_norm": 0.8998428583145142,
      "learning_rate": 6.02409638554217e-05,
      "loss": 0.0728,
      "step": 200
    },
    {
      "epoch": 0.30106312917489886,
      "eval_loss": 0.056333187967538834,
      "eval_runtime": 178.4366,
      "eval_samples_per_second": 6.619,
      "eval_steps_per_second": 6.619,
      "step": 200
    },
    {
      "epoch": 0.3161162856336438,
      "grad_norm": 0.517510175704956,
      "learning_rate": 6.325301204819278e-05,
      "loss": 0.0601,
      "step": 210
    },
    {
      "epoch": 0.33116944209238874,
      "grad_norm": 0.5527799725532532,
      "learning_rate": 6.626506024096386e-05,
      "loss": 0.0636,
      "step": 220
    },
    {
      "epoch": 0.3462225985511337,
      "grad_norm": 0.5699729919433594,
      "learning_rate": 6.927710843373494e-05,
      "loss": 0.0742,
      "step": 230
    },
    {
      "epoch": 0.36127575500987863,
      "grad_norm": 0.4265763461589813,
      "learning_rate": 7.228915662650602e-05,
      "loss": 0.0485,
      "step": 240
    },
    {
      "epoch": 0.3763289114686236,
      "grad_norm": 0.8758322596549988,
      "learning_rate": 7.530120481927712e-05,
      "loss": 0.0574,
      "step": 250
    },
    {
      "epoch": 0.3763289114686236,
      "eval_loss": 0.04959677904844284,
      "eval_runtime": 178.4826,
      "eval_samples_per_second": 6.617,
      "eval_steps_per_second": 6.617,
      "step": 250
    },
    {
      "epoch": 0.3913820679273685,
      "grad_norm": 0.4359073340892792,
      "learning_rate": 7.83132530120482e-05,
      "loss": 0.0393,
      "step": 260
    },
    {
      "epoch": 0.40643522438611346,
      "grad_norm": 0.4522893726825714,
      "learning_rate": 8.132530120481928e-05,
      "loss": 0.0538,
      "step": 270
    },
    {
      "epoch": 0.4214883808448584,
      "grad_norm": 0.531626284122467,
      "learning_rate": 8.433734939759037e-05,
      "loss": 0.0481,
      "step": 280
    },
    {
      "epoch": 0.43654153730360334,
      "grad_norm": 0.26284611225128174,
      "learning_rate": 8.734939759036145e-05,
      "loss": 0.049,
      "step": 290
    },
    {
      "epoch": 0.4515946937623483,
      "grad_norm": 0.639433741569519,
      "learning_rate": 9.036144578313253e-05,
      "loss": 0.0603,
      "step": 300
    },
    {
      "epoch": 0.4515946937623483,
      "eval_loss": 0.05344177782535553,
      "eval_runtime": 178.3037,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 300
    },
    {
      "epoch": 0.46664785022109323,
      "grad_norm": 0.23814865946769714,
      "learning_rate": 9.337349397590361e-05,
      "loss": 0.0443,
      "step": 310
    },
    {
      "epoch": 0.4817010066798382,
      "grad_norm": 0.8485608100891113,
      "learning_rate": 9.638554216867471e-05,
      "loss": 0.0691,
      "step": 320
    },
    {
      "epoch": 0.4967541631385831,
      "grad_norm": 0.7342369556427002,
      "learning_rate": 9.939759036144579e-05,
      "loss": 0.0535,
      "step": 330
    },
    {
      "epoch": 0.5118073195973281,
      "grad_norm": 0.5011388063430786,
      "learning_rate": 9.999823129264712e-05,
      "loss": 0.038,
      "step": 340
    },
    {
      "epoch": 0.526860476056073,
      "grad_norm": 0.2543235421180725,
      "learning_rate": 9.999104613348688e-05,
      "loss": 0.0432,
      "step": 350
    },
    {
      "epoch": 0.526860476056073,
      "eval_loss": 0.043458491563797,
      "eval_runtime": 178.3108,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 350
    },
    {
      "epoch": 0.541913632514818,
      "grad_norm": 0.19088613986968994,
      "learning_rate": 9.997833477197385e-05,
      "loss": 0.054,
      "step": 360
    },
    {
      "epoch": 0.5569667889735629,
      "grad_norm": 0.4376489818096161,
      "learning_rate": 9.996009861327077e-05,
      "loss": 0.043,
      "step": 370
    },
    {
      "epoch": 0.5720199454323078,
      "grad_norm": 0.33153238892555237,
      "learning_rate": 9.993633967327269e-05,
      "loss": 0.0445,
      "step": 380
    },
    {
      "epoch": 0.5870731018910528,
      "grad_norm": 0.5583683252334595,
      "learning_rate": 9.990706057838416e-05,
      "loss": 0.04,
      "step": 390
    },
    {
      "epoch": 0.6021262583497977,
      "grad_norm": 0.4868285655975342,
      "learning_rate": 9.987226456522884e-05,
      "loss": 0.0506,
      "step": 400
    },
    {
      "epoch": 0.6021262583497977,
      "eval_loss": 0.04465359449386597,
      "eval_runtime": 178.2721,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 400
    },
    {
      "epoch": 0.6171794148085427,
      "grad_norm": 0.4056116044521332,
      "learning_rate": 9.983195548029173e-05,
      "loss": 0.0334,
      "step": 410
    },
    {
      "epoch": 0.6322325712672876,
      "grad_norm": 0.5368084907531738,
      "learning_rate": 9.9786137779494e-05,
      "loss": 0.0537,
      "step": 420
    },
    {
      "epoch": 0.6472857277260325,
      "grad_norm": 0.2836042046546936,
      "learning_rate": 9.973481652770038e-05,
      "loss": 0.0403,
      "step": 430
    },
    {
      "epoch": 0.6623388841847775,
      "grad_norm": 0.2168349325656891,
      "learning_rate": 9.967799739815925e-05,
      "loss": 0.0322,
      "step": 440
    },
    {
      "epoch": 0.6773920406435224,
      "grad_norm": 0.5306774973869324,
      "learning_rate": 9.961568667187556e-05,
      "loss": 0.0418,
      "step": 450
    },
    {
      "epoch": 0.6773920406435224,
      "eval_loss": 0.042699169367551804,
      "eval_runtime": 178.1761,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 450
    },
    {
      "epoch": 0.6924451971022674,
      "grad_norm": 0.5590943098068237,
      "learning_rate": 9.954789123691642e-05,
      "loss": 0.0423,
      "step": 460
    },
    {
      "epoch": 0.7074983535610123,
      "grad_norm": 0.46472442150115967,
      "learning_rate": 9.947461858764978e-05,
      "loss": 0.0383,
      "step": 470
    },
    {
      "epoch": 0.7225515100197573,
      "grad_norm": 0.307923823595047,
      "learning_rate": 9.939587682391586e-05,
      "loss": 0.0392,
      "step": 480
    },
    {
      "epoch": 0.7376046664785022,
      "grad_norm": 0.3933861255645752,
      "learning_rate": 9.931167465013182e-05,
      "loss": 0.0519,
      "step": 490
    },
    {
      "epoch": 0.7526578229372471,
      "grad_norm": 0.31609928607940674,
      "learning_rate": 9.922202137432955e-05,
      "loss": 0.0319,
      "step": 500
    },
    {
      "epoch": 0.7526578229372471,
      "eval_loss": 0.03804313763976097,
      "eval_runtime": 178.3743,
      "eval_samples_per_second": 6.621,
      "eval_steps_per_second": 6.621,
      "step": 500
    },
    {
      "epoch": 0.7677109793959921,
      "grad_norm": 0.2750873863697052,
      "learning_rate": 9.912692690712665e-05,
      "loss": 0.0356,
      "step": 510
    },
    {
      "epoch": 0.782764135854737,
      "grad_norm": 0.4082382023334503,
      "learning_rate": 9.902640176063103e-05,
      "loss": 0.0426,
      "step": 520
    },
    {
      "epoch": 0.797817292313482,
      "grad_norm": 0.2213507443666458,
      "learning_rate": 9.892045704727864e-05,
      "loss": 0.0579,
      "step": 530
    },
    {
      "epoch": 0.8128704487722269,
      "grad_norm": 0.24513815343379974,
      "learning_rate": 9.880910447860527e-05,
      "loss": 0.0512,
      "step": 540
    },
    {
      "epoch": 0.8279236052309719,
      "grad_norm": 0.33264800906181335,
      "learning_rate": 9.869235636395177e-05,
      "loss": 0.0409,
      "step": 550
    },
    {
      "epoch": 0.8279236052309719,
      "eval_loss": 0.03934917971491814,
      "eval_runtime": 178.3094,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 550
    },
    {
      "epoch": 0.8429767616897168,
      "grad_norm": 0.4701898694038391,
      "learning_rate": 9.857022560910338e-05,
      "loss": 0.0513,
      "step": 560
    },
    {
      "epoch": 0.8580299181484617,
      "grad_norm": 0.36034145951271057,
      "learning_rate": 9.844272571486311e-05,
      "loss": 0.0375,
      "step": 570
    },
    {
      "epoch": 0.8730830746072067,
      "grad_norm": 0.1605140119791031,
      "learning_rate": 9.830987077555924e-05,
      "loss": 0.0471,
      "step": 580
    },
    {
      "epoch": 0.8881362310659516,
      "grad_norm": 0.3974078893661499,
      "learning_rate": 9.817167547748729e-05,
      "loss": 0.0588,
      "step": 590
    },
    {
      "epoch": 0.9031893875246966,
      "grad_norm": 0.21184875071048737,
      "learning_rate": 9.802815509728662e-05,
      "loss": 0.0289,
      "step": 600
    },
    {
      "epoch": 0.9031893875246966,
      "eval_loss": 0.0372873954474926,
      "eval_runtime": 178.4025,
      "eval_samples_per_second": 6.62,
      "eval_steps_per_second": 6.62,
      "step": 600
    },
    {
      "epoch": 0.9182425439834415,
      "grad_norm": 0.3172430992126465,
      "learning_rate": 9.787932550025158e-05,
      "loss": 0.0351,
      "step": 610
    },
    {
      "epoch": 0.9332957004421865,
      "grad_norm": 0.25452977418899536,
      "learning_rate": 9.772520313857775e-05,
      "loss": 0.0354,
      "step": 620
    },
    {
      "epoch": 0.9483488569009314,
      "grad_norm": 0.31180253624916077,
      "learning_rate": 9.756580504954334e-05,
      "loss": 0.0408,
      "step": 630
    },
    {
      "epoch": 0.9634020133596763,
      "grad_norm": 0.2877283990383148,
      "learning_rate": 9.740114885362562e-05,
      "loss": 0.0357,
      "step": 640
    },
    {
      "epoch": 0.9784551698184213,
      "grad_norm": 0.31497249007225037,
      "learning_rate": 9.723125275255325e-05,
      "loss": 0.0301,
      "step": 650
    },
    {
      "epoch": 0.9784551698184213,
      "eval_loss": 0.03777872771024704,
      "eval_runtime": 178.3928,
      "eval_samples_per_second": 6.62,
      "eval_steps_per_second": 6.62,
      "step": 650
    },
    {
      "epoch": 0.9935083262771662,
      "grad_norm": 0.18348178267478943,
      "learning_rate": 9.705613552729415e-05,
      "loss": 0.0259,
      "step": 660
    },
    {
      "epoch": 1.0085614827359113,
      "grad_norm": 0.15709549188613892,
      "learning_rate": 9.68758165359794e-05,
      "loss": 0.0335,
      "step": 670
    },
    {
      "epoch": 1.0236146391946561,
      "grad_norm": 0.25934895873069763,
      "learning_rate": 9.669031571176322e-05,
      "loss": 0.0353,
      "step": 680
    },
    {
      "epoch": 1.0386677956534012,
      "grad_norm": 0.21780580282211304,
      "learning_rate": 9.64996535606196e-05,
      "loss": 0.0318,
      "step": 690
    },
    {
      "epoch": 1.053720952112146,
      "grad_norm": 0.14777664840221405,
      "learning_rate": 9.630385115907545e-05,
      "loss": 0.0289,
      "step": 700
    },
    {
      "epoch": 1.053720952112146,
      "eval_loss": 0.03666107356548309,
      "eval_runtime": 178.3948,
      "eval_samples_per_second": 6.62,
      "eval_steps_per_second": 6.62,
      "step": 700
    },
    {
      "epoch": 1.068774108570891,
      "grad_norm": 0.09680185467004776,
      "learning_rate": 9.610293015188067e-05,
      "loss": 0.0305,
      "step": 710
    },
    {
      "epoch": 1.083827265029636,
      "grad_norm": 0.21849073469638824,
      "learning_rate": 9.589691274961556e-05,
      "loss": 0.0241,
      "step": 720
    },
    {
      "epoch": 1.098880421488381,
      "grad_norm": 0.23137938976287842,
      "learning_rate": 9.568582172623544e-05,
      "loss": 0.0367,
      "step": 730
    },
    {
      "epoch": 1.1139335779471258,
      "grad_norm": 0.38276636600494385,
      "learning_rate": 9.546968041655326e-05,
      "loss": 0.0302,
      "step": 740
    },
    {
      "epoch": 1.1289867344058706,
      "grad_norm": 0.17088480293750763,
      "learning_rate": 9.524851271366001e-05,
      "loss": 0.0283,
      "step": 750
    },
    {
      "epoch": 1.1289867344058706,
      "eval_loss": 0.03535303846001625,
      "eval_runtime": 178.355,
      "eval_samples_per_second": 6.622,
      "eval_steps_per_second": 6.622,
      "step": 750
    },
    {
      "epoch": 1.1440398908646157,
      "grad_norm": 0.236710324883461,
      "learning_rate": 9.502234306628355e-05,
      "loss": 0.0295,
      "step": 760
    },
    {
      "epoch": 1.1590930473233607,
      "grad_norm": 0.23621885478496552,
      "learning_rate": 9.47911964760858e-05,
      "loss": 0.0272,
      "step": 770
    },
    {
      "epoch": 1.1741462037821055,
      "grad_norm": 0.41253966093063354,
      "learning_rate": 9.455509849489915e-05,
      "loss": 0.0254,
      "step": 780
    },
    {
      "epoch": 1.1891993602408506,
      "grad_norm": 0.2661738991737366,
      "learning_rate": 9.431407522190175e-05,
      "loss": 0.0321,
      "step": 790
    },
    {
      "epoch": 1.2042525166995954,
      "grad_norm": 0.26856687664985657,
      "learning_rate": 9.406815330073244e-05,
      "loss": 0.0338,
      "step": 800
    },
    {
      "epoch": 1.2042525166995954,
      "eval_loss": 0.03521593660116196,
      "eval_runtime": 178.2893,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 800
    },
    {
      "epoch": 1.2193056731583405,
      "grad_norm": 0.15742696821689606,
      "learning_rate": 9.381735991654546e-05,
      "loss": 0.03,
      "step": 810
    },
    {
      "epoch": 1.2343588296170853,
      "grad_norm": 0.38911351561546326,
      "learning_rate": 9.356172279300528e-05,
      "loss": 0.0352,
      "step": 820
    },
    {
      "epoch": 1.2494119860758304,
      "grad_norm": 0.12868359684944153,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.0296,
      "step": 830
    },
    {
      "epoch": 1.2644651425345752,
      "grad_norm": 0.1884261518716812,
      "learning_rate": 9.303603089662716e-05,
      "loss": 0.0368,
      "step": 840
    },
    {
      "epoch": 1.2795182989933203,
      "grad_norm": 0.14614073932170868,
      "learning_rate": 9.276603423579164e-05,
      "loss": 0.0286,
      "step": 850
    },
    {
      "epoch": 1.2795182989933203,
      "eval_loss": 0.035069651901721954,
      "eval_runtime": 178.3662,
      "eval_samples_per_second": 6.621,
      "eval_steps_per_second": 6.621,
      "step": 850
    },
    {
      "epoch": 1.294571455452065,
      "grad_norm": 0.33718934655189514,
      "learning_rate": 9.249131005318387e-05,
      "loss": 0.0412,
      "step": 860
    },
    {
      "epoch": 1.30962461191081,
      "grad_norm": 0.1532343029975891,
      "learning_rate": 9.221188871787075e-05,
      "loss": 0.0327,
      "step": 870
    },
    {
      "epoch": 1.324677768369555,
      "grad_norm": 0.16133852303028107,
      "learning_rate": 9.192780111816047e-05,
      "loss": 0.0297,
      "step": 880
    },
    {
      "epoch": 1.3397309248283,
      "grad_norm": 0.19326288998126984,
      "learning_rate": 9.163907865818806e-05,
      "loss": 0.0304,
      "step": 890
    },
    {
      "epoch": 1.3547840812870449,
      "grad_norm": 0.2810550332069397,
      "learning_rate": 9.134575325444376e-05,
      "loss": 0.0396,
      "step": 900
    },
    {
      "epoch": 1.3547840812870449,
      "eval_loss": 0.0357198566198349,
      "eval_runtime": 178.4266,
      "eval_samples_per_second": 6.619,
      "eval_steps_per_second": 6.619,
      "step": 900
    },
    {
      "epoch": 1.3698372377457897,
      "grad_norm": 0.24054096639156342,
      "learning_rate": 9.104785733224496e-05,
      "loss": 0.0297,
      "step": 910
    },
    {
      "epoch": 1.3848903942045347,
      "grad_norm": 0.1945008933544159,
      "learning_rate": 9.07454238221517e-05,
      "loss": 0.0297,
      "step": 920
    },
    {
      "epoch": 1.3999435506632798,
      "grad_norm": 0.22676967084407806,
      "learning_rate": 9.043848615632642e-05,
      "loss": 0.0258,
      "step": 930
    },
    {
      "epoch": 1.4149967071220246,
      "grad_norm": 0.12921719253063202,
      "learning_rate": 9.012707826483823e-05,
      "loss": 0.0331,
      "step": 940
    },
    {
      "epoch": 1.4300498635807695,
      "grad_norm": 0.2984769940376282,
      "learning_rate": 8.98112345719122e-05,
      "loss": 0.0286,
      "step": 950
    },
    {
      "epoch": 1.4300498635807695,
      "eval_loss": 0.0343579463660717,
      "eval_runtime": 178.3834,
      "eval_samples_per_second": 6.621,
      "eval_steps_per_second": 6.621,
      "step": 950
    },
    {
      "epoch": 1.4451030200395145,
      "grad_norm": 0.5328255891799927,
      "learning_rate": 8.949098999212391e-05,
      "loss": 0.0325,
      "step": 960
    },
    {
      "epoch": 1.4601561764982596,
      "grad_norm": 0.22663827240467072,
      "learning_rate": 8.916637992653991e-05,
      "loss": 0.025,
      "step": 970
    },
    {
      "epoch": 1.4752093329570044,
      "grad_norm": 0.31026363372802734,
      "learning_rate": 8.883744025880428e-05,
      "loss": 0.0322,
      "step": 980
    },
    {
      "epoch": 1.4902624894157492,
      "grad_norm": 0.1781584918498993,
      "learning_rate": 8.850420735117202e-05,
      "loss": 0.0239,
      "step": 990
    },
    {
      "epoch": 1.5053156458744943,
      "grad_norm": 0.25284937024116516,
      "learning_rate": 8.816671804048933e-05,
      "loss": 0.0276,
      "step": 1000
    },
    {
      "epoch": 1.5053156458744943,
      "eval_loss": 0.03388373181223869,
      "eval_runtime": 178.2776,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 1000
    },
    {
      "epoch": 1.5203688023332393,
      "grad_norm": 0.18928562104701996,
      "learning_rate": 8.782500963412156e-05,
      "loss": 0.0317,
      "step": 1010
    },
    {
      "epoch": 1.5354219587919842,
      "grad_norm": 0.24282586574554443,
      "learning_rate": 8.747911990582912e-05,
      "loss": 0.0249,
      "step": 1020
    },
    {
      "epoch": 1.550475115250729,
      "grad_norm": 0.1898348331451416,
      "learning_rate": 8.712908709159183e-05,
      "loss": 0.025,
      "step": 1030
    },
    {
      "epoch": 1.565528271709474,
      "grad_norm": 0.46804437041282654,
      "learning_rate": 8.677494988538211e-05,
      "loss": 0.034,
      "step": 1040
    },
    {
      "epoch": 1.5805814281682191,
      "grad_norm": 0.21851754188537598,
      "learning_rate": 8.641674743488769e-05,
      "loss": 0.0269,
      "step": 1050
    },
    {
      "epoch": 1.5805814281682191,
      "eval_loss": 0.03368181735277176,
      "eval_runtime": 178.3246,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 1050
    },
    {
      "epoch": 1.595634584626964,
      "grad_norm": 0.3493017256259918,
      "learning_rate": 8.605451933718397e-05,
      "loss": 0.028,
      "step": 1060
    },
    {
      "epoch": 1.6106877410857088,
      "grad_norm": 0.3040076792240143,
      "learning_rate": 8.568830563435694e-05,
      "loss": 0.0347,
      "step": 1070
    },
    {
      "epoch": 1.6257408975444538,
      "grad_norm": 0.2517918646335602,
      "learning_rate": 8.531814680907664e-05,
      "loss": 0.0212,
      "step": 1080
    },
    {
      "epoch": 1.6407940540031989,
      "grad_norm": 0.28657108545303345,
      "learning_rate": 8.494408378012209e-05,
      "loss": 0.0253,
      "step": 1090
    },
    {
      "epoch": 1.6558472104619437,
      "grad_norm": 0.12799641489982605,
      "learning_rate": 8.456615789785804e-05,
      "loss": 0.0202,
      "step": 1100
    },
    {
      "epoch": 1.6558472104619437,
      "eval_loss": 0.033622123301029205,
      "eval_runtime": 178.3151,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 1100
    },
    {
      "epoch": 1.6709003669206886,
      "grad_norm": 0.220402330160141,
      "learning_rate": 8.418441093966385e-05,
      "loss": 0.0368,
      "step": 1110
    },
    {
      "epoch": 1.6859535233794336,
      "grad_norm": 0.2065258026123047,
      "learning_rate": 8.379888510531535e-05,
      "loss": 0.0253,
      "step": 1120
    },
    {
      "epoch": 1.7010066798381787,
      "grad_norm": 0.2068171501159668,
      "learning_rate": 8.340962301231981e-05,
      "loss": 0.0214,
      "step": 1130
    },
    {
      "epoch": 1.7160598362969235,
      "grad_norm": 0.12344514578580856,
      "learning_rate": 8.301666769120488e-05,
      "loss": 0.0291,
      "step": 1140
    },
    {
      "epoch": 1.7311129927556683,
      "grad_norm": 0.11504557728767395,
      "learning_rate": 8.262006258076187e-05,
      "loss": 0.0364,
      "step": 1150
    },
    {
      "epoch": 1.7311129927556683,
      "eval_loss": 0.03163878992199898,
      "eval_runtime": 178.2999,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 1150
    },
    {
      "epoch": 1.7461661492144134,
      "grad_norm": 0.2136172354221344,
      "learning_rate": 8.221985152324385e-05,
      "loss": 0.0284,
      "step": 1160
    },
    {
      "epoch": 1.7612193056731584,
      "grad_norm": 0.3486640453338623,
      "learning_rate": 8.18160787595191e-05,
      "loss": 0.0373,
      "step": 1170
    },
    {
      "epoch": 1.7762724621319033,
      "grad_norm": 0.2501416802406311,
      "learning_rate": 8.14087889241806e-05,
      "loss": 0.0265,
      "step": 1180
    },
    {
      "epoch": 1.791325618590648,
      "grad_norm": 0.21038861572742462,
      "learning_rate": 8.099802704061195e-05,
      "loss": 0.0337,
      "step": 1190
    },
    {
      "epoch": 1.8063787750493931,
      "grad_norm": 0.3405999541282654,
      "learning_rate": 8.058383851601027e-05,
      "loss": 0.0265,
      "step": 1200
    },
    {
      "epoch": 1.8063787750493931,
      "eval_loss": 0.032966770231723785,
      "eval_runtime": 178.2353,
      "eval_samples_per_second": 6.626,
      "eval_steps_per_second": 6.626,
      "step": 1200
    },
    {
      "epoch": 1.8214319315081382,
      "grad_norm": 0.5199443101882935,
      "learning_rate": 8.01662691363668e-05,
      "loss": 0.029,
      "step": 1210
    },
    {
      "epoch": 1.836485087966883,
      "grad_norm": 0.35030198097229004,
      "learning_rate": 7.974536506140547e-05,
      "loss": 0.0333,
      "step": 1220
    },
    {
      "epoch": 1.8515382444256279,
      "grad_norm": 0.27158239483833313,
      "learning_rate": 7.932117281948021e-05,
      "loss": 0.0294,
      "step": 1230
    },
    {
      "epoch": 1.866591400884373,
      "grad_norm": 0.18654458224773407,
      "learning_rate": 7.889373930243164e-05,
      "loss": 0.032,
      "step": 1240
    },
    {
      "epoch": 1.881644557343118,
      "grad_norm": 0.13273607194423676,
      "learning_rate": 7.846311176040331e-05,
      "loss": 0.0291,
      "step": 1250
    },
    {
      "epoch": 1.881644557343118,
      "eval_loss": 0.03071535751223564,
      "eval_runtime": 178.2583,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 1250
    },
    {
      "epoch": 1.8966977138018628,
      "grad_norm": 0.1273525208234787,
      "learning_rate": 7.802933779661859e-05,
      "loss": 0.0361,
      "step": 1260
    },
    {
      "epoch": 1.9117508702606076,
      "grad_norm": 0.09461209923028946,
      "learning_rate": 7.759246536211844e-05,
      "loss": 0.0251,
      "step": 1270
    },
    {
      "epoch": 1.9268040267193527,
      "grad_norm": 0.10459864139556885,
      "learning_rate": 7.715254275046062e-05,
      "loss": 0.0308,
      "step": 1280
    },
    {
      "epoch": 1.9418571831780977,
      "grad_norm": 0.21300184726715088,
      "learning_rate": 7.670961859238124e-05,
      "loss": 0.0272,
      "step": 1290
    },
    {
      "epoch": 1.9569103396368426,
      "grad_norm": 0.47472086548805237,
      "learning_rate": 7.626374185041886e-05,
      "loss": 0.0244,
      "step": 1300
    },
    {
      "epoch": 1.9569103396368426,
      "eval_loss": 0.031091291457414627,
      "eval_runtime": 178.3155,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 1300
    },
    {
      "epoch": 1.9719634960955874,
      "grad_norm": 0.29737257957458496,
      "learning_rate": 7.581496181350203e-05,
      "loss": 0.0255,
      "step": 1310
    },
    {
      "epoch": 1.9870166525543325,
      "grad_norm": 0.18417982757091522,
      "learning_rate": 7.536332809150067e-05,
      "loss": 0.0258,
      "step": 1320
    },
    {
      "epoch": 2.0020698090130775,
      "grad_norm": 0.27828750014305115,
      "learning_rate": 7.490889060974201e-05,
      "loss": 0.0272,
      "step": 1330
    },
    {
      "epoch": 2.0171229654718226,
      "grad_norm": 0.194544717669487,
      "learning_rate": 7.445169960349167e-05,
      "loss": 0.015,
      "step": 1340
    },
    {
      "epoch": 2.032176121930567,
      "grad_norm": 0.21474245190620422,
      "learning_rate": 7.399180561240044e-05,
      "loss": 0.0197,
      "step": 1350
    },
    {
      "epoch": 2.032176121930567,
      "eval_loss": 0.031300000846385956,
      "eval_runtime": 178.2318,
      "eval_samples_per_second": 6.626,
      "eval_steps_per_second": 6.626,
      "step": 1350
    },
    {
      "epoch": 2.0472292783893122,
      "grad_norm": 0.10497045516967773,
      "learning_rate": 7.352925947491746e-05,
      "loss": 0.0118,
      "step": 1360
    },
    {
      "epoch": 2.0622824348480573,
      "grad_norm": 0.32971543073654175,
      "learning_rate": 7.306411232267029e-05,
      "loss": 0.0199,
      "step": 1370
    },
    {
      "epoch": 2.0773355913068023,
      "grad_norm": 0.20883502066135406,
      "learning_rate": 7.259641557481269e-05,
      "loss": 0.0221,
      "step": 1380
    },
    {
      "epoch": 2.092388747765547,
      "grad_norm": 0.12358664721250534,
      "learning_rate": 7.212622093234049e-05,
      "loss": 0.0153,
      "step": 1390
    },
    {
      "epoch": 2.107441904224292,
      "grad_norm": 0.2762002646923065,
      "learning_rate": 7.165358037237643e-05,
      "loss": 0.0223,
      "step": 1400
    },
    {
      "epoch": 2.107441904224292,
      "eval_loss": 0.03041531704366207,
      "eval_runtime": 178.1406,
      "eval_samples_per_second": 6.63,
      "eval_steps_per_second": 6.63,
      "step": 1400
    },
    {
      "epoch": 2.122495060683037,
      "grad_norm": 0.21662060916423798,
      "learning_rate": 7.117854614242434e-05,
      "loss": 0.0208,
      "step": 1410
    },
    {
      "epoch": 2.137548217141782,
      "grad_norm": 0.2806037366390228,
      "learning_rate": 7.070117075459352e-05,
      "loss": 0.0203,
      "step": 1420
    },
    {
      "epoch": 2.1526013736005267,
      "grad_norm": 0.17611578106880188,
      "learning_rate": 7.022150697979384e-05,
      "loss": 0.0226,
      "step": 1430
    },
    {
      "epoch": 2.167654530059272,
      "grad_norm": 0.22745737433433533,
      "learning_rate": 6.973960784190237e-05,
      "loss": 0.023,
      "step": 1440
    },
    {
      "epoch": 2.182707686518017,
      "grad_norm": 0.2118409425020218,
      "learning_rate": 6.925552661190166e-05,
      "loss": 0.0197,
      "step": 1450
    },
    {
      "epoch": 2.182707686518017,
      "eval_loss": 0.03106178157031536,
      "eval_runtime": 178.1235,
      "eval_samples_per_second": 6.63,
      "eval_steps_per_second": 6.63,
      "step": 1450
    },
    {
      "epoch": 2.197760842976762,
      "grad_norm": 0.1830424964427948,
      "learning_rate": 6.876931680199121e-05,
      "loss": 0.0156,
      "step": 1460
    },
    {
      "epoch": 2.2128139994355065,
      "grad_norm": 0.10300840437412262,
      "learning_rate": 6.828103215967186e-05,
      "loss": 0.0203,
      "step": 1470
    },
    {
      "epoch": 2.2278671558942515,
      "grad_norm": 0.17760959267616272,
      "learning_rate": 6.779072666180446e-05,
      "loss": 0.0234,
      "step": 1480
    },
    {
      "epoch": 2.2429203123529966,
      "grad_norm": 0.05315260961651802,
      "learning_rate": 6.729845450864294e-05,
      "loss": 0.0159,
      "step": 1490
    },
    {
      "epoch": 2.257973468811741,
      "grad_norm": 0.21539124846458435,
      "learning_rate": 6.680427011784292e-05,
      "loss": 0.0204,
      "step": 1500
    },
    {
      "epoch": 2.257973468811741,
      "eval_loss": 0.0323634073138237,
      "eval_runtime": 178.1357,
      "eval_samples_per_second": 6.63,
      "eval_steps_per_second": 6.63,
      "step": 1500
    },
    {
      "epoch": 2.2730266252704863,
      "grad_norm": 0.24310541152954102,
      "learning_rate": 6.630822811844604e-05,
      "loss": 0.017,
      "step": 1510
    },
    {
      "epoch": 2.2880797817292313,
      "grad_norm": 0.1696529984474182,
      "learning_rate": 6.58103833448412e-05,
      "loss": 0.0222,
      "step": 1520
    },
    {
      "epoch": 2.3031329381879764,
      "grad_norm": 0.3020496666431427,
      "learning_rate": 6.531079083070288e-05,
      "loss": 0.0234,
      "step": 1530
    },
    {
      "epoch": 2.3181860946467214,
      "grad_norm": 0.12914994359016418,
      "learning_rate": 6.480950580290752e-05,
      "loss": 0.017,
      "step": 1540
    },
    {
      "epoch": 2.333239251105466,
      "grad_norm": 0.18893304467201233,
      "learning_rate": 6.430658367542843e-05,
      "loss": 0.0276,
      "step": 1550
    },
    {
      "epoch": 2.333239251105466,
      "eval_loss": 0.031088994815945625,
      "eval_runtime": 178.1568,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 6.629,
      "step": 1550
    },
    {
      "epoch": 2.348292407564211,
      "grad_norm": 0.08729825168848038,
      "learning_rate": 6.380208004321036e-05,
      "loss": 0.0247,
      "step": 1560
    },
    {
      "epoch": 2.363345564022956,
      "grad_norm": 0.12183941900730133,
      "learning_rate": 6.32960506760236e-05,
      "loss": 0.0181,
      "step": 1570
    },
    {
      "epoch": 2.378398720481701,
      "grad_norm": 0.2509489953517914,
      "learning_rate": 6.278855151229901e-05,
      "loss": 0.0158,
      "step": 1580
    },
    {
      "epoch": 2.393451876940446,
      "grad_norm": 0.0914265587925911,
      "learning_rate": 6.227963865294444e-05,
      "loss": 0.0168,
      "step": 1590
    },
    {
      "epoch": 2.408505033399191,
      "grad_norm": 0.11629947274923325,
      "learning_rate": 6.176936835514312e-05,
      "loss": 0.0146,
      "step": 1600
    },
    {
      "epoch": 2.408505033399191,
      "eval_loss": 0.031647298485040665,
      "eval_runtime": 178.063,
      "eval_samples_per_second": 6.632,
      "eval_steps_per_second": 6.632,
      "step": 1600
    },
    {
      "epoch": 2.423558189857936,
      "grad_norm": 0.2622014582157135,
      "learning_rate": 6.125779702613471e-05,
      "loss": 0.0176,
      "step": 1610
    },
    {
      "epoch": 2.438611346316681,
      "grad_norm": 0.177884042263031,
      "learning_rate": 6.074498121697983e-05,
      "loss": 0.0207,
      "step": 1620
    },
    {
      "epoch": 2.4536645027754256,
      "grad_norm": 0.20466935634613037,
      "learning_rate": 6.023097761630879e-05,
      "loss": 0.0254,
      "step": 1630
    },
    {
      "epoch": 2.4687176592341706,
      "grad_norm": 0.1373012661933899,
      "learning_rate": 5.971584304405489e-05,
      "loss": 0.0164,
      "step": 1640
    },
    {
      "epoch": 2.4837708156929157,
      "grad_norm": 0.23096352815628052,
      "learning_rate": 5.919963444517338e-05,
      "loss": 0.0165,
      "step": 1650
    },
    {
      "epoch": 2.4837708156929157,
      "eval_loss": 0.03288305178284645,
      "eval_runtime": 178.0988,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 6.631,
      "step": 1650
    },
    {
      "epoch": 2.4988239721516607,
      "grad_norm": 0.2017652541399002,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0181,
      "step": 1660
    },
    {
      "epoch": 2.5138771286104054,
      "grad_norm": 0.07514988631010056,
      "learning_rate": 5.816422353467562e-05,
      "loss": 0.0148,
      "step": 1670
    },
    {
      "epoch": 2.5289302850691504,
      "grad_norm": 0.11077038943767548,
      "learning_rate": 5.7645135681360496e-05,
      "loss": 0.0183,
      "step": 1680
    },
    {
      "epoch": 2.5439834415278955,
      "grad_norm": 0.24594064056873322,
      "learning_rate": 5.7125202705367234e-05,
      "loss": 0.0228,
      "step": 1690
    },
    {
      "epoch": 2.5590365979866405,
      "grad_norm": 0.2879762649536133,
      "learning_rate": 5.660448208208513e-05,
      "loss": 0.0242,
      "step": 1700
    },
    {
      "epoch": 2.5590365979866405,
      "eval_loss": 0.03144305199384689,
      "eval_runtime": 178.0564,
      "eval_samples_per_second": 6.633,
      "eval_steps_per_second": 6.633,
      "step": 1700
    },
    {
      "epoch": 2.574089754445385,
      "grad_norm": 0.060580622404813766,
      "learning_rate": 5.608303137397294e-05,
      "loss": 0.0156,
      "step": 1710
    },
    {
      "epoch": 2.58914291090413,
      "grad_norm": 0.13122287392616272,
      "learning_rate": 5.5560908224195886e-05,
      "loss": 0.0218,
      "step": 1720
    },
    {
      "epoch": 2.6041960673628752,
      "grad_norm": 0.2391199916601181,
      "learning_rate": 5.503817035025342e-05,
      "loss": 0.0187,
      "step": 1730
    },
    {
      "epoch": 2.61924922382162,
      "grad_norm": 0.2503703236579895,
      "learning_rate": 5.4514875537598985e-05,
      "loss": 0.0143,
      "step": 1740
    },
    {
      "epoch": 2.634302380280365,
      "grad_norm": 0.2618468105792999,
      "learning_rate": 5.399108163325217e-05,
      "loss": 0.0273,
      "step": 1750
    },
    {
      "epoch": 2.634302380280365,
      "eval_loss": 0.032075975090265274,
      "eval_runtime": 178.055,
      "eval_samples_per_second": 6.633,
      "eval_steps_per_second": 6.633,
      "step": 1750
    },
    {
      "epoch": 2.64935553673911,
      "grad_norm": 0.15629079937934875,
      "learning_rate": 5.346684653940408e-05,
      "loss": 0.0217,
      "step": 1760
    },
    {
      "epoch": 2.664408693197855,
      "grad_norm": 0.09079688042402267,
      "learning_rate": 5.294222820701661e-05,
      "loss": 0.0204,
      "step": 1770
    },
    {
      "epoch": 2.6794618496566,
      "grad_norm": 0.29432734847068787,
      "learning_rate": 5.24172846294163e-05,
      "loss": 0.0148,
      "step": 1780
    },
    {
      "epoch": 2.6945150061153447,
      "grad_norm": 0.17906472086906433,
      "learning_rate": 5.1892073835883524e-05,
      "loss": 0.0189,
      "step": 1790
    },
    {
      "epoch": 2.7095681625740897,
      "grad_norm": 0.37694233655929565,
      "learning_rate": 5.136665388523778e-05,
      "loss": 0.0204,
      "step": 1800
    },
    {
      "epoch": 2.7095681625740897,
      "eval_loss": 0.03231995552778244,
      "eval_runtime": 177.9391,
      "eval_samples_per_second": 6.637,
      "eval_steps_per_second": 6.637,
      "step": 1800
    },
    {
      "epoch": 2.7246213190328348,
      "grad_norm": 0.15220922231674194,
      "learning_rate": 5.0841082859419585e-05,
      "loss": 0.0243,
      "step": 1810
    },
    {
      "epoch": 2.7396744754915794,
      "grad_norm": 0.1869482398033142,
      "learning_rate": 5.031541885706987e-05,
      "loss": 0.0215,
      "step": 1820
    },
    {
      "epoch": 2.7547276319503244,
      "grad_norm": 0.12930846214294434,
      "learning_rate": 4.9789719987107545e-05,
      "loss": 0.0223,
      "step": 1830
    },
    {
      "epoch": 2.7697807884090695,
      "grad_norm": 0.4493294060230255,
      "learning_rate": 4.926404436230596e-05,
      "loss": 0.0201,
      "step": 1840
    },
    {
      "epoch": 2.7848339448678145,
      "grad_norm": 0.3657067120075226,
      "learning_rate": 4.8738450092868785e-05,
      "loss": 0.021,
      "step": 1850
    },
    {
      "epoch": 2.7848339448678145,
      "eval_loss": 0.03066716156899929,
      "eval_runtime": 177.9619,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 6.636,
      "step": 1850
    },
    {
      "epoch": 2.7998871013265596,
      "grad_norm": 0.35673412680625916,
      "learning_rate": 4.8212995280006426e-05,
      "loss": 0.0186,
      "step": 1860
    },
    {
      "epoch": 2.814940257785304,
      "grad_norm": 0.20455271005630493,
      "learning_rate": 4.76877380095132e-05,
      "loss": 0.0183,
      "step": 1870
    },
    {
      "epoch": 2.8299934142440493,
      "grad_norm": 0.25493428111076355,
      "learning_rate": 4.7162736345346303e-05,
      "loss": 0.0149,
      "step": 1880
    },
    {
      "epoch": 2.8450465707027943,
      "grad_norm": 0.25855717062950134,
      "learning_rate": 4.663804832320726e-05,
      "loss": 0.0134,
      "step": 1890
    },
    {
      "epoch": 2.860099727161539,
      "grad_norm": 0.16750505566596985,
      "learning_rate": 4.6113731944126406e-05,
      "loss": 0.0195,
      "step": 1900
    },
    {
      "epoch": 2.860099727161539,
      "eval_loss": 0.033523257821798325,
      "eval_runtime": 178.3199,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 1900
    },
    {
      "epoch": 2.875152883620284,
      "grad_norm": 0.17557500302791595,
      "learning_rate": 4.558984516805118e-05,
      "loss": 0.0177,
      "step": 1910
    },
    {
      "epoch": 2.890206040079029,
      "grad_norm": 0.32916489243507385,
      "learning_rate": 4.5066445907439104e-05,
      "loss": 0.0132,
      "step": 1920
    },
    {
      "epoch": 2.905259196537774,
      "grad_norm": 0.3346385359764099,
      "learning_rate": 4.454359202085582e-05,
      "loss": 0.0233,
      "step": 1930
    },
    {
      "epoch": 2.920312352996519,
      "grad_norm": 0.5354165434837341,
      "learning_rate": 4.402134130657925e-05,
      "loss": 0.0163,
      "step": 1940
    },
    {
      "epoch": 2.9353655094552638,
      "grad_norm": 0.24359606206417084,
      "learning_rate": 4.349975149621039e-05,
      "loss": 0.0221,
      "step": 1950
    },
    {
      "epoch": 2.9353655094552638,
      "eval_loss": 0.030188467353582382,
      "eval_runtime": 177.9643,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 6.636,
      "step": 1950
    },
    {
      "epoch": 2.950418665914009,
      "grad_norm": 0.11338020116090775,
      "learning_rate": 4.297888024829126e-05,
      "loss": 0.0163,
      "step": 1960
    },
    {
      "epoch": 2.965471822372754,
      "grad_norm": 0.1252482682466507,
      "learning_rate": 4.2458785141931314e-05,
      "loss": 0.0181,
      "step": 1970
    },
    {
      "epoch": 2.9805249788314985,
      "grad_norm": 0.0980340763926506,
      "learning_rate": 4.1939523670442316e-05,
      "loss": 0.0171,
      "step": 1980
    },
    {
      "epoch": 2.9955781352902435,
      "grad_norm": 0.1608283817768097,
      "learning_rate": 4.14211532349828e-05,
      "loss": 0.0177,
      "step": 1990
    },
    {
      "epoch": 3.0106312917489886,
      "grad_norm": 0.07570704817771912,
      "learning_rate": 4.090373113821281e-05,
      "loss": 0.0129,
      "step": 2000
    },
    {
      "epoch": 3.0106312917489886,
      "eval_loss": 0.03050847351551056,
      "eval_runtime": 177.9373,
      "eval_samples_per_second": 6.637,
      "eval_steps_per_second": 6.637,
      "step": 2000
    },
    {
      "epoch": 3.0256844482077336,
      "grad_norm": 0.15597139298915863,
      "learning_rate": 4.0387314577959315e-05,
      "loss": 0.0113,
      "step": 2010
    },
    {
      "epoch": 3.0407376046664787,
      "grad_norm": 0.055645011365413666,
      "learning_rate": 3.987196064089346e-05,
      "loss": 0.0091,
      "step": 2020
    },
    {
      "epoch": 3.0557907611252233,
      "grad_norm": 0.10545477271080017,
      "learning_rate": 3.935772629621995e-05,
      "loss": 0.0079,
      "step": 2030
    },
    {
      "epoch": 3.0708439175839684,
      "grad_norm": 0.14946284890174866,
      "learning_rate": 3.8844668389379396e-05,
      "loss": 0.0102,
      "step": 2040
    },
    {
      "epoch": 3.0858970740427134,
      "grad_norm": 0.2742091119289398,
      "learning_rate": 3.833284363576447e-05,
      "loss": 0.0127,
      "step": 2050
    },
    {
      "epoch": 3.0858970740427134,
      "eval_loss": 0.03302207961678505,
      "eval_runtime": 178.0745,
      "eval_samples_per_second": 6.632,
      "eval_steps_per_second": 6.632,
      "step": 2050
    },
    {
      "epoch": 3.1009502305014585,
      "grad_norm": 0.19018790125846863,
      "learning_rate": 3.7822308614450406e-05,
      "loss": 0.0096,
      "step": 2060
    },
    {
      "epoch": 3.116003386960203,
      "grad_norm": 0.04711037501692772,
      "learning_rate": 3.7313119761940375e-05,
      "loss": 0.0142,
      "step": 2070
    },
    {
      "epoch": 3.131056543418948,
      "grad_norm": 0.22224950790405273,
      "learning_rate": 3.680533336592694e-05,
      "loss": 0.0091,
      "step": 2080
    },
    {
      "epoch": 3.146109699877693,
      "grad_norm": 0.09899937361478806,
      "learning_rate": 3.62990055590697e-05,
      "loss": 0.0101,
      "step": 2090
    },
    {
      "epoch": 3.1611628563364382,
      "grad_norm": 0.09017643332481384,
      "learning_rate": 3.579419231279023e-05,
      "loss": 0.0122,
      "step": 2100
    },
    {
      "epoch": 3.1611628563364382,
      "eval_loss": 0.03305284306406975,
      "eval_runtime": 178.3274,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 2100
    },
    {
      "epoch": 3.176216012795183,
      "grad_norm": 0.12764662504196167,
      "learning_rate": 3.529094943108475e-05,
      "loss": 0.0085,
      "step": 2110
    },
    {
      "epoch": 3.191269169253928,
      "grad_norm": 0.12719255685806274,
      "learning_rate": 3.478933254435534e-05,
      "loss": 0.0144,
      "step": 2120
    },
    {
      "epoch": 3.206322325712673,
      "grad_norm": 0.164920374751091,
      "learning_rate": 3.4289397103260346e-05,
      "loss": 0.0125,
      "step": 2130
    },
    {
      "epoch": 3.221375482171418,
      "grad_norm": 0.22153641283512115,
      "learning_rate": 3.3791198372584664e-05,
      "loss": 0.0129,
      "step": 2140
    },
    {
      "epoch": 3.2364286386301626,
      "grad_norm": 0.11729541420936584,
      "learning_rate": 3.329479142513051e-05,
      "loss": 0.0129,
      "step": 2150
    },
    {
      "epoch": 3.2364286386301626,
      "eval_loss": 0.03324244171380997,
      "eval_runtime": 178.3386,
      "eval_samples_per_second": 6.622,
      "eval_steps_per_second": 6.622,
      "step": 2150
    },
    {
      "epoch": 3.2514817950889077,
      "grad_norm": 0.344887912273407,
      "learning_rate": 3.280023113562957e-05,
      "loss": 0.0136,
      "step": 2160
    },
    {
      "epoch": 3.2665349515476527,
      "grad_norm": 0.1558815985918045,
      "learning_rate": 3.230757217467677e-05,
      "loss": 0.0171,
      "step": 2170
    },
    {
      "epoch": 3.2815881080063978,
      "grad_norm": 0.19974495470523834,
      "learning_rate": 3.1816869002686936e-05,
      "loss": 0.0108,
      "step": 2180
    },
    {
      "epoch": 3.2966412644651424,
      "grad_norm": 0.13594548404216766,
      "learning_rate": 3.1328175863874464e-05,
      "loss": 0.0086,
      "step": 2190
    },
    {
      "epoch": 3.3116944209238874,
      "grad_norm": 0.21001030504703522,
      "learning_rate": 3.084154678025692e-05,
      "loss": 0.0125,
      "step": 2200
    },
    {
      "epoch": 3.3116944209238874,
      "eval_loss": 0.03326999396085739,
      "eval_runtime": 178.6788,
      "eval_samples_per_second": 6.61,
      "eval_steps_per_second": 6.61,
      "step": 2200
    },
    {
      "epoch": 3.3267475773826325,
      "grad_norm": 0.2704913914203644,
      "learning_rate": 3.035703554568331e-05,
      "loss": 0.0126,
      "step": 2210
    },
    {
      "epoch": 3.3418007338413775,
      "grad_norm": 0.1079394668340683,
      "learning_rate": 2.9874695719887464e-05,
      "loss": 0.0109,
      "step": 2220
    },
    {
      "epoch": 3.356853890300122,
      "grad_norm": 0.3655363917350769,
      "learning_rate": 2.9394580622567312e-05,
      "loss": 0.0076,
      "step": 2230
    },
    {
      "epoch": 3.371907046758867,
      "grad_norm": 0.09557830542325974,
      "learning_rate": 2.8916743327490803e-05,
      "loss": 0.0116,
      "step": 2240
    },
    {
      "epoch": 3.3869602032176123,
      "grad_norm": 0.22106143832206726,
      "learning_rate": 2.8441236656628828e-05,
      "loss": 0.0132,
      "step": 2250
    },
    {
      "epoch": 3.3869602032176123,
      "eval_loss": 0.03426986560225487,
      "eval_runtime": 179.2113,
      "eval_samples_per_second": 6.59,
      "eval_steps_per_second": 6.59,
      "step": 2250
    },
    {
      "epoch": 3.4020133596763573,
      "grad_norm": 0.1936042606830597,
      "learning_rate": 2.79681131743161e-05,
      "loss": 0.0064,
      "step": 2260
    },
    {
      "epoch": 3.417066516135102,
      "grad_norm": 0.18822342157363892,
      "learning_rate": 2.7497425181440607e-05,
      "loss": 0.0131,
      "step": 2270
    },
    {
      "epoch": 3.432119672593847,
      "grad_norm": 0.13977766036987305,
      "learning_rate": 2.702922470966187e-05,
      "loss": 0.0067,
      "step": 2280
    },
    {
      "epoch": 3.447172829052592,
      "grad_norm": 0.1867881566286087,
      "learning_rate": 2.6563563515659306e-05,
      "loss": 0.0104,
      "step": 2290
    },
    {
      "epoch": 3.462225985511337,
      "grad_norm": 0.035174448043107986,
      "learning_rate": 2.6100493075410848e-05,
      "loss": 0.0084,
      "step": 2300
    },
    {
      "epoch": 3.462225985511337,
      "eval_loss": 0.03515166416764259,
      "eval_runtime": 179.2383,
      "eval_samples_per_second": 6.589,
      "eval_steps_per_second": 6.589,
      "step": 2300
    },
    {
      "epoch": 3.4772791419700817,
      "grad_norm": 0.3701817989349365,
      "learning_rate": 2.5640064578502497e-05,
      "loss": 0.0153,
      "step": 2310
    },
    {
      "epoch": 3.4923322984288268,
      "grad_norm": 0.25794854760169983,
      "learning_rate": 2.5182328922469723e-05,
      "loss": 0.0108,
      "step": 2320
    },
    {
      "epoch": 3.507385454887572,
      "grad_norm": 0.29802244901657104,
      "learning_rate": 2.4727336707170973e-05,
      "loss": 0.0102,
      "step": 2330
    },
    {
      "epoch": 3.522438611346317,
      "grad_norm": 0.10990750789642334,
      "learning_rate": 2.427513822919424e-05,
      "loss": 0.0115,
      "step": 2340
    },
    {
      "epoch": 3.5374917678050615,
      "grad_norm": 0.282993346452713,
      "learning_rate": 2.3825783476297087e-05,
      "loss": 0.0086,
      "step": 2350
    },
    {
      "epoch": 3.5374917678050615,
      "eval_loss": 0.03473927080631256,
      "eval_runtime": 179.3483,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 2350
    },
    {
      "epoch": 3.5525449242638065,
      "grad_norm": 0.27041804790496826,
      "learning_rate": 2.337932212188073e-05,
      "loss": 0.0109,
      "step": 2360
    },
    {
      "epoch": 3.5675980807225516,
      "grad_norm": 0.2581341862678528,
      "learning_rate": 2.2935803519499e-05,
      "loss": 0.0122,
      "step": 2370
    },
    {
      "epoch": 3.582651237181296,
      "grad_norm": 0.13770440220832825,
      "learning_rate": 2.2495276697402662e-05,
      "loss": 0.0083,
      "step": 2380
    },
    {
      "epoch": 3.5977043936400412,
      "grad_norm": 0.13981565833091736,
      "learning_rate": 2.2057790353119535e-05,
      "loss": 0.0105,
      "step": 2390
    },
    {
      "epoch": 3.6127575500987863,
      "grad_norm": 0.27494239807128906,
      "learning_rate": 2.1623392848071354e-05,
      "loss": 0.0112,
      "step": 2400
    },
    {
      "epoch": 3.6127575500987863,
      "eval_loss": 0.03513345867395401,
      "eval_runtime": 179.5449,
      "eval_samples_per_second": 6.578,
      "eval_steps_per_second": 6.578,
      "step": 2400
    },
    {
      "epoch": 3.6278107065575314,
      "grad_norm": 0.18222074210643768,
      "learning_rate": 2.1192132202227677e-05,
      "loss": 0.0103,
      "step": 2410
    },
    {
      "epoch": 3.6428638630162764,
      "grad_norm": 0.27348509430885315,
      "learning_rate": 2.0764056088797645e-05,
      "loss": 0.0096,
      "step": 2420
    },
    {
      "epoch": 3.657917019475021,
      "grad_norm": 0.39703041315078735,
      "learning_rate": 2.0339211828959904e-05,
      "loss": 0.0098,
      "step": 2430
    },
    {
      "epoch": 3.672970175933766,
      "grad_norm": 0.21773886680603027,
      "learning_rate": 1.9917646386631577e-05,
      "loss": 0.0094,
      "step": 2440
    },
    {
      "epoch": 3.688023332392511,
      "grad_norm": 0.16702532768249512,
      "learning_rate": 1.949940636327671e-05,
      "loss": 0.012,
      "step": 2450
    },
    {
      "epoch": 3.688023332392511,
      "eval_loss": 0.03648751974105835,
      "eval_runtime": 179.7341,
      "eval_samples_per_second": 6.571,
      "eval_steps_per_second": 6.571,
      "step": 2450
    },
    {
      "epoch": 3.7030764888512557,
      "grad_norm": 0.32012760639190674,
      "learning_rate": 1.9084537992754792e-05,
      "loss": 0.01,
      "step": 2460
    },
    {
      "epoch": 3.718129645310001,
      "grad_norm": 0.24596737325191498,
      "learning_rate": 1.8673087136209803e-05,
      "loss": 0.0076,
      "step": 2470
    },
    {
      "epoch": 3.733182801768746,
      "grad_norm": 0.07540785521268845,
      "learning_rate": 1.8265099277000614e-05,
      "loss": 0.0067,
      "step": 2480
    },
    {
      "epoch": 3.748235958227491,
      "grad_norm": 0.2560862600803375,
      "learning_rate": 1.7860619515673033e-05,
      "loss": 0.0107,
      "step": 2490
    },
    {
      "epoch": 3.763289114686236,
      "grad_norm": 0.09346114844083786,
      "learning_rate": 1.7459692564974316e-05,
      "loss": 0.011,
      "step": 2500
    },
    {
      "epoch": 3.763289114686236,
      "eval_loss": 0.03612372279167175,
      "eval_runtime": 179.868,
      "eval_samples_per_second": 6.566,
      "eval_steps_per_second": 6.566,
      "step": 2500
    },
    {
      "epoch": 3.7783422711449806,
      "grad_norm": 0.07731559872627258,
      "learning_rate": 1.7062362744910322e-05,
      "loss": 0.0059,
      "step": 2510
    },
    {
      "epoch": 3.7933954276037256,
      "grad_norm": 0.35534602403640747,
      "learning_rate": 1.6668673977846254e-05,
      "loss": 0.0122,
      "step": 2520
    },
    {
      "epoch": 3.8084485840624707,
      "grad_norm": 0.4245568513870239,
      "learning_rate": 1.6278669783651395e-05,
      "loss": 0.0089,
      "step": 2530
    },
    {
      "epoch": 3.8235017405212153,
      "grad_norm": 0.0772927924990654,
      "learning_rate": 1.589239327488812e-05,
      "loss": 0.0051,
      "step": 2540
    },
    {
      "epoch": 3.8385548969799603,
      "grad_norm": 0.1755327433347702,
      "learning_rate": 1.5509887152046137e-05,
      "loss": 0.0079,
      "step": 2550
    },
    {
      "epoch": 3.8385548969799603,
      "eval_loss": 0.03893940895795822,
      "eval_runtime": 180.0653,
      "eval_samples_per_second": 6.559,
      "eval_steps_per_second": 6.559,
      "step": 2550
    },
    {
      "epoch": 3.8536080534387054,
      "grad_norm": 0.23755531013011932,
      "learning_rate": 1.5131193698822232e-05,
      "loss": 0.0137,
      "step": 2560
    },
    {
      "epoch": 3.8686612098974504,
      "grad_norm": 0.21640031039714813,
      "learning_rate": 1.4756354777446001e-05,
      "loss": 0.0089,
      "step": 2570
    },
    {
      "epoch": 3.8837143663561955,
      "grad_norm": 0.3311944305896759,
      "learning_rate": 1.4385411824052342e-05,
      "loss": 0.0093,
      "step": 2580
    },
    {
      "epoch": 3.89876752281494,
      "grad_norm": 0.2809070646762848,
      "learning_rate": 1.4018405844100812e-05,
      "loss": 0.0123,
      "step": 2590
    },
    {
      "epoch": 3.913820679273685,
      "grad_norm": 0.19817812740802765,
      "learning_rate": 1.3655377407842812e-05,
      "loss": 0.0089,
      "step": 2600
    },
    {
      "epoch": 3.913820679273685,
      "eval_loss": 0.03675345331430435,
      "eval_runtime": 180.0963,
      "eval_samples_per_second": 6.558,
      "eval_steps_per_second": 6.558,
      "step": 2600
    },
    {
      "epoch": 3.92887383573243,
      "grad_norm": 0.15485577285289764,
      "learning_rate": 1.3296366645836822e-05,
      "loss": 0.0074,
      "step": 2610
    },
    {
      "epoch": 3.943926992191175,
      "grad_norm": 0.2657122313976288,
      "learning_rate": 1.2941413244512113e-05,
      "loss": 0.0128,
      "step": 2620
    },
    {
      "epoch": 3.95898014864992,
      "grad_norm": 0.3065131902694702,
      "learning_rate": 1.2590556441781725e-05,
      "loss": 0.0077,
      "step": 2630
    },
    {
      "epoch": 3.974033305108665,
      "grad_norm": 0.487378865480423,
      "learning_rate": 1.2243835022705003e-05,
      "loss": 0.0115,
      "step": 2640
    },
    {
      "epoch": 3.98908646156741,
      "grad_norm": 0.25239965319633484,
      "learning_rate": 1.1901287315199977e-05,
      "loss": 0.009,
      "step": 2650
    },
    {
      "epoch": 3.98908646156741,
      "eval_loss": 0.03644752502441406,
      "eval_runtime": 180.2417,
      "eval_samples_per_second": 6.552,
      "eval_steps_per_second": 6.552,
      "step": 2650
    },
    {
      "epoch": 4.004139618026155,
      "grad_norm": 0.24973592162132263,
      "learning_rate": 1.1562951185806676e-05,
      "loss": 0.0078,
      "step": 2660
    },
    {
      "epoch": 4.0191927744849,
      "grad_norm": 0.11293565481901169,
      "learning_rate": 1.1228864035501069e-05,
      "loss": 0.0064,
      "step": 2670
    },
    {
      "epoch": 4.034245930943645,
      "grad_norm": 0.24446244537830353,
      "learning_rate": 1.0899062795560573e-05,
      "loss": 0.0042,
      "step": 2680
    },
    {
      "epoch": 4.04929908740239,
      "grad_norm": 0.29575204849243164,
      "learning_rate": 1.0573583923481711e-05,
      "loss": 0.0087,
      "step": 2690
    },
    {
      "epoch": 4.064352243861134,
      "grad_norm": 0.18624891340732574,
      "learning_rate": 1.0252463398949792e-05,
      "loss": 0.0086,
      "step": 2700
    },
    {
      "epoch": 4.064352243861134,
      "eval_loss": 0.037868741899728775,
      "eval_runtime": 180.3651,
      "eval_samples_per_second": 6.548,
      "eval_steps_per_second": 6.548,
      "step": 2700
    },
    {
      "epoch": 4.07940540031988,
      "grad_norm": 0.10995467007160187,
      "learning_rate": 9.935736719861622e-06,
      "loss": 0.0053,
      "step": 2710
    },
    {
      "epoch": 4.0944585567786245,
      "grad_norm": 0.09725473076105118,
      "learning_rate": 9.62343889840151e-06,
      "loss": 0.0051,
      "step": 2720
    },
    {
      "epoch": 4.109511713237369,
      "grad_norm": 0.036554962396621704,
      "learning_rate": 9.315604457170768e-06,
      "loss": 0.0033,
      "step": 2730
    },
    {
      "epoch": 4.124564869696115,
      "grad_norm": 0.10706653445959091,
      "learning_rate": 9.012267425371513e-06,
      "loss": 0.0053,
      "step": 2740
    },
    {
      "epoch": 4.139618026154859,
      "grad_norm": 0.11202232539653778,
      "learning_rate": 8.71346133504498e-06,
      "loss": 0.005,
      "step": 2750
    },
    {
      "epoch": 4.139618026154859,
      "eval_loss": 0.03884652629494667,
      "eval_runtime": 180.4257,
      "eval_samples_per_second": 6.546,
      "eval_steps_per_second": 6.546,
      "step": 2750
    },
    {
      "epoch": 4.154671182613605,
      "grad_norm": 0.05641934648156166,
      "learning_rate": 8.419219217364654e-06,
      "loss": 0.0037,
      "step": 2760
    },
    {
      "epoch": 4.169724339072349,
      "grad_norm": 0.17391756176948547,
      "learning_rate": 8.129573598984997e-06,
      "loss": 0.0054,
      "step": 2770
    },
    {
      "epoch": 4.184777495531094,
      "grad_norm": 0.08039785176515579,
      "learning_rate": 7.844556498445788e-06,
      "loss": 0.0046,
      "step": 2780
    },
    {
      "epoch": 4.199830651989839,
      "grad_norm": 0.24674569070339203,
      "learning_rate": 7.564199422632579e-06,
      "loss": 0.0067,
      "step": 2790
    },
    {
      "epoch": 4.214883808448584,
      "grad_norm": 0.05541248247027397,
      "learning_rate": 7.288533363293959e-06,
      "loss": 0.0051,
      "step": 2800
    },
    {
      "epoch": 4.214883808448584,
      "eval_loss": 0.0397312231361866,
      "eval_runtime": 180.5412,
      "eval_samples_per_second": 6.541,
      "eval_steps_per_second": 6.541,
      "step": 2800
    },
    {
      "epoch": 4.229936964907329,
      "grad_norm": 0.30977103114128113,
      "learning_rate": 7.017588793615498e-06,
      "loss": 0.004,
      "step": 2810
    },
    {
      "epoch": 4.244990121366074,
      "grad_norm": 0.145084947347641,
      "learning_rate": 6.751395664851135e-06,
      "loss": 0.0038,
      "step": 2820
    },
    {
      "epoch": 4.260043277824819,
      "grad_norm": 0.1771937906742096,
      "learning_rate": 6.489983403012312e-06,
      "loss": 0.0047,
      "step": 2830
    },
    {
      "epoch": 4.275096434283564,
      "grad_norm": 0.30625805258750916,
      "learning_rate": 6.233380905615049e-06,
      "loss": 0.0042,
      "step": 2840
    },
    {
      "epoch": 4.290149590742309,
      "grad_norm": 0.15609732270240784,
      "learning_rate": 5.981616538485496e-06,
      "loss": 0.0055,
      "step": 2850
    },
    {
      "epoch": 4.290149590742309,
      "eval_loss": 0.04038526862859726,
      "eval_runtime": 180.0932,
      "eval_samples_per_second": 6.558,
      "eval_steps_per_second": 6.558,
      "step": 2850
    },
    {
      "epoch": 4.3052027472010534,
      "grad_norm": 0.150795578956604,
      "learning_rate": 5.73471813262435e-06,
      "loss": 0.0035,
      "step": 2860
    },
    {
      "epoch": 4.320255903659799,
      "grad_norm": 0.26626893877983093,
      "learning_rate": 5.4927129811301715e-06,
      "loss": 0.0035,
      "step": 2870
    },
    {
      "epoch": 4.335309060118544,
      "grad_norm": 0.042985837906599045,
      "learning_rate": 5.255627836182453e-06,
      "loss": 0.0024,
      "step": 2880
    },
    {
      "epoch": 4.350362216577288,
      "grad_norm": 0.2361251413822174,
      "learning_rate": 5.0234889060842176e-06,
      "loss": 0.0056,
      "step": 2890
    },
    {
      "epoch": 4.365415373036034,
      "grad_norm": 0.12534025311470032,
      "learning_rate": 4.796321852364877e-06,
      "loss": 0.0083,
      "step": 2900
    },
    {
      "epoch": 4.365415373036034,
      "eval_loss": 0.04092538729310036,
      "eval_runtime": 180.6762,
      "eval_samples_per_second": 6.537,
      "eval_steps_per_second": 6.537,
      "step": 2900
    },
    {
      "epoch": 4.380468529494778,
      "grad_norm": 0.10099391639232635,
      "learning_rate": 4.5741517869435706e-06,
      "loss": 0.0033,
      "step": 2910
    },
    {
      "epoch": 4.395521685953524,
      "grad_norm": 0.037613652646541595,
      "learning_rate": 4.357003269353105e-06,
      "loss": 0.0035,
      "step": 2920
    },
    {
      "epoch": 4.410574842412268,
      "grad_norm": 0.10834306478500366,
      "learning_rate": 4.144900304025101e-06,
      "loss": 0.0057,
      "step": 2930
    },
    {
      "epoch": 4.425627998871013,
      "grad_norm": 0.4370887577533722,
      "learning_rate": 3.937866337636459e-06,
      "loss": 0.0077,
      "step": 2940
    },
    {
      "epoch": 4.4406811553297585,
      "grad_norm": 0.2118665874004364,
      "learning_rate": 3.7359242565174423e-06,
      "loss": 0.0055,
      "step": 2950
    },
    {
      "epoch": 4.4406811553297585,
      "eval_loss": 0.04117078334093094,
      "eval_runtime": 180.7007,
      "eval_samples_per_second": 6.536,
      "eval_steps_per_second": 6.536,
      "step": 2950
    },
    {
      "epoch": 4.455734311788503,
      "grad_norm": 0.11663982272148132,
      "learning_rate": 3.539096384121743e-06,
      "loss": 0.005,
      "step": 2960
    },
    {
      "epoch": 4.470787468247248,
      "grad_norm": 0.1609174907207489,
      "learning_rate": 3.34740447855878e-06,
      "loss": 0.0025,
      "step": 2970
    },
    {
      "epoch": 4.485840624705993,
      "grad_norm": 0.2631778419017792,
      "learning_rate": 3.160869730188465e-06,
      "loss": 0.0046,
      "step": 2980
    },
    {
      "epoch": 4.500893781164738,
      "grad_norm": 0.05932645872235298,
      "learning_rate": 2.9795127592787186e-06,
      "loss": 0.0054,
      "step": 2990
    },
    {
      "epoch": 4.515946937623482,
      "grad_norm": 0.5128821134567261,
      "learning_rate": 2.803353613726056e-06,
      "loss": 0.0056,
      "step": 3000
    },
    {
      "epoch": 4.515946937623482,
      "eval_loss": 0.041392985731363297,
      "eval_runtime": 180.7978,
      "eval_samples_per_second": 6.532,
      "eval_steps_per_second": 6.532,
      "step": 3000
    },
    {
      "epoch": 4.531000094082228,
      "grad_norm": 0.12355174124240875,
      "learning_rate": 2.6324117668393877e-06,
      "loss": 0.0044,
      "step": 3010
    },
    {
      "epoch": 4.5460532505409725,
      "grad_norm": 0.32771605253219604,
      "learning_rate": 2.466706115187406e-06,
      "loss": 0.0029,
      "step": 3020
    },
    {
      "epoch": 4.561106406999718,
      "grad_norm": 0.15814469754695892,
      "learning_rate": 2.3062549765096364e-06,
      "loss": 0.0028,
      "step": 3030
    },
    {
      "epoch": 4.576159563458463,
      "grad_norm": 0.20907792448997498,
      "learning_rate": 2.1510760876915505e-06,
      "loss": 0.0071,
      "step": 3040
    },
    {
      "epoch": 4.591212719917207,
      "grad_norm": 0.3278658986091614,
      "learning_rate": 2.0011866028038617e-06,
      "loss": 0.01,
      "step": 3050
    },
    {
      "epoch": 4.591212719917207,
      "eval_loss": 0.04168102145195007,
      "eval_runtime": 180.8128,
      "eval_samples_per_second": 6.532,
      "eval_steps_per_second": 6.532,
      "step": 3050
    },
    {
      "epoch": 4.606265876375953,
      "grad_norm": 0.22570809721946716,
      "learning_rate": 1.8566030912062549e-06,
      "loss": 0.0058,
      "step": 3060
    },
    {
      "epoch": 4.621319032834697,
      "grad_norm": 0.134377583861351,
      "learning_rate": 1.717341535715733e-06,
      "loss": 0.0034,
      "step": 3070
    },
    {
      "epoch": 4.636372189293443,
      "grad_norm": 0.11749345809221268,
      "learning_rate": 1.5834173308397982e-06,
      "loss": 0.0045,
      "step": 3080
    },
    {
      "epoch": 4.6514253457521875,
      "grad_norm": 0.0782150849699974,
      "learning_rate": 1.4548452810747403e-06,
      "loss": 0.0064,
      "step": 3090
    },
    {
      "epoch": 4.666478502210932,
      "grad_norm": 0.13278253376483917,
      "learning_rate": 1.33163959926903e-06,
      "loss": 0.0041,
      "step": 3100
    },
    {
      "epoch": 4.666478502210932,
      "eval_loss": 0.041555680334568024,
      "eval_runtime": 180.8052,
      "eval_samples_per_second": 6.532,
      "eval_steps_per_second": 6.532,
      "step": 3100
    },
    {
      "epoch": 4.681531658669678,
      "grad_norm": 0.06097458675503731,
      "learning_rate": 1.2138139050522023e-06,
      "loss": 0.0046,
      "step": 3110
    },
    {
      "epoch": 4.696584815128422,
      "grad_norm": 0.0928138941526413,
      "learning_rate": 1.101381223329301e-06,
      "loss": 0.0042,
      "step": 3120
    },
    {
      "epoch": 4.711637971587168,
      "grad_norm": 0.11098574846982956,
      "learning_rate": 9.943539828410342e-07,
      "loss": 0.0041,
      "step": 3130
    },
    {
      "epoch": 4.726691128045912,
      "grad_norm": 0.02670971117913723,
      "learning_rate": 8.927440147898702e-07,
      "loss": 0.0053,
      "step": 3140
    },
    {
      "epoch": 4.741744284504657,
      "grad_norm": 0.16498462855815887,
      "learning_rate": 7.96562551532154e-07,
      "loss": 0.0046,
      "step": 3150
    },
    {
      "epoch": 4.741744284504657,
      "eval_loss": 0.041612498462200165,
      "eval_runtime": 180.8943,
      "eval_samples_per_second": 6.529,
      "eval_steps_per_second": 6.529,
      "step": 3150
    },
    {
      "epoch": 4.756797440963402,
      "grad_norm": 0.02326858416199684,
      "learning_rate": 7.05820225336451e-07,
      "loss": 0.0034,
      "step": 3160
    },
    {
      "epoch": 4.771850597422147,
      "grad_norm": 0.03126712515950203,
      "learning_rate": 6.20527067208232e-07,
      "loss": 0.0028,
      "step": 3170
    },
    {
      "epoch": 4.786903753880892,
      "grad_norm": 0.09830054640769958,
      "learning_rate": 5.406925057809653e-07,
      "loss": 0.0042,
      "step": 3180
    },
    {
      "epoch": 4.801956910339637,
      "grad_norm": 0.10579635947942734,
      "learning_rate": 4.6632536627386756e-07,
      "loss": 0.0054,
      "step": 3190
    },
    {
      "epoch": 4.817010066798382,
      "grad_norm": 0.15730389952659607,
      "learning_rate": 3.974338695163393e-07,
      "loss": 0.0029,
      "step": 3200
    },
    {
      "epoch": 4.817010066798382,
      "eval_loss": 0.04158816114068031,
      "eval_runtime": 180.8757,
      "eval_samples_per_second": 6.529,
      "eval_steps_per_second": 6.529,
      "step": 3200
    },
    {
      "epoch": 4.832063223257126,
      "grad_norm": 0.15249422192573547,
      "learning_rate": 3.3402563103916984e-07,
      "loss": 0.003,
      "step": 3210
    },
    {
      "epoch": 4.847116379715872,
      "grad_norm": 0.06642192602157593,
      "learning_rate": 2.7610766023271615e-07,
      "loss": 0.0032,
      "step": 3220
    },
    {
      "epoch": 4.8621695361746164,
      "grad_norm": 0.17876122891902924,
      "learning_rate": 2.2368635957205618e-07,
      "loss": 0.0049,
      "step": 3230
    },
    {
      "epoch": 4.877222692633362,
      "grad_norm": 0.2662655711174011,
      "learning_rate": 1.7676752390920482e-07,
      "loss": 0.0067,
      "step": 3240
    },
    {
      "epoch": 4.8922758490921066,
      "grad_norm": 0.3073679208755493,
      "learning_rate": 1.3535633983257078e-07,
      "loss": 0.0071,
      "step": 3250
    },
    {
      "epoch": 4.8922758490921066,
      "eval_loss": 0.04167798534035683,
      "eval_runtime": 180.9807,
      "eval_samples_per_second": 6.526,
      "eval_steps_per_second": 6.526,
      "step": 3250
    },
    {
      "epoch": 4.907329005550851,
      "grad_norm": 0.07359470427036285,
      "learning_rate": 9.945738509358205e-08,
      "loss": 0.0026,
      "step": 3260
    },
    {
      "epoch": 4.922382162009597,
      "grad_norm": 0.0783749371767044,
      "learning_rate": 6.907462810065158e-08,
      "loss": 0.0037,
      "step": 3270
    },
    {
      "epoch": 4.937435318468341,
      "grad_norm": 0.2075500339269638,
      "learning_rate": 4.4211427480500554e-08,
      "loss": 0.0079,
      "step": 3280
    },
    {
      "epoch": 4.952488474927087,
      "grad_norm": 0.035787180066108704,
      "learning_rate": 2.4870531706872034e-08,
      "loss": 0.0042,
      "step": 3290
    },
    {
      "epoch": 4.967541631385831,
      "grad_norm": 0.28970831632614136,
      "learning_rate": 1.105407879670728e-08,
      "loss": 0.0039,
      "step": 3300
    },
    {
      "epoch": 4.967541631385831,
      "eval_loss": 0.041562460362911224,
      "eval_runtime": 180.9472,
      "eval_samples_per_second": 6.527,
      "eval_steps_per_second": 6.527,
      "step": 3300
    },
    {
      "epoch": 4.982594787844576,
      "grad_norm": 0.1511860340833664,
      "learning_rate": 2.763596073807051e-09,
      "loss": 0.0041,
      "step": 3310
    },
    {
      "epoch": 4.9976479443033215,
      "grad_norm": 0.369547039270401,
      "learning_rate": 0.0,
      "loss": 0.0045,
      "step": 3320
    },
    {
      "epoch": 4.9976479443033215,
      "step": 3320,
      "total_flos": 8.981559806648648e+17,
      "train_loss": 0.048681382634620886,
      "train_runtime": 37834.6736,
      "train_samples_per_second": 1.405,
      "train_steps_per_second": 0.088
    }
  ],
  "logging_steps": 10,
  "max_steps": 3320,
  "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": 8.981559806648648e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}