{
  "best_metric": 0.03188331797719002,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold6/checkpoint-1900",
  "epoch": 4.995305164319249,
  "eval_steps": 50,
  "global_step": 3325,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015023474178403756,
      "grad_norm": 1.6643978357315063,
      "learning_rate": 3.003003003003003e-06,
      "loss": 1.6617,
      "step": 10
    },
    {
      "epoch": 0.03004694835680751,
      "grad_norm": 1.970920443534851,
      "learning_rate": 6.006006006006006e-06,
      "loss": 1.7717,
      "step": 20
    },
    {
      "epoch": 0.04507042253521127,
      "grad_norm": 2.2089600563049316,
      "learning_rate": 9.00900900900901e-06,
      "loss": 1.6755,
      "step": 30
    },
    {
      "epoch": 0.06009389671361502,
      "grad_norm": 3.677905797958374,
      "learning_rate": 1.2012012012012012e-05,
      "loss": 1.3133,
      "step": 40
    },
    {
      "epoch": 0.07511737089201878,
      "grad_norm": 1.0802251100540161,
      "learning_rate": 1.5015015015015016e-05,
      "loss": 0.8292,
      "step": 50
    },
    {
      "epoch": 0.07511737089201878,
      "eval_loss": 0.6473308205604553,
      "eval_runtime": 207.0843,
      "eval_samples_per_second": 5.717,
      "eval_steps_per_second": 5.717,
      "step": 50
    },
    {
      "epoch": 0.09014084507042254,
      "grad_norm": 0.8090672492980957,
      "learning_rate": 1.801801801801802e-05,
      "loss": 0.6816,
      "step": 60
    },
    {
      "epoch": 0.10516431924882629,
      "grad_norm": 0.7934019565582275,
      "learning_rate": 2.102102102102102e-05,
      "loss": 0.4595,
      "step": 70
    },
    {
      "epoch": 0.12018779342723004,
      "grad_norm": 0.7469967007637024,
      "learning_rate": 2.4024024024024024e-05,
      "loss": 0.2868,
      "step": 80
    },
    {
      "epoch": 0.1352112676056338,
      "grad_norm": 0.6989622712135315,
      "learning_rate": 2.702702702702703e-05,
      "loss": 0.1691,
      "step": 90
    },
    {
      "epoch": 0.15023474178403756,
      "grad_norm": 0.714993417263031,
      "learning_rate": 3.0030030030030033e-05,
      "loss": 0.1595,
      "step": 100
    },
    {
      "epoch": 0.15023474178403756,
      "eval_loss": 0.11690635979175568,
      "eval_runtime": 208.0685,
      "eval_samples_per_second": 5.69,
      "eval_steps_per_second": 5.69,
      "step": 100
    },
    {
      "epoch": 0.1652582159624413,
      "grad_norm": 0.6915186643600464,
      "learning_rate": 3.3033033033033035e-05,
      "loss": 0.1358,
      "step": 110
    },
    {
      "epoch": 0.18028169014084508,
      "grad_norm": 0.7759954333305359,
      "learning_rate": 3.603603603603604e-05,
      "loss": 0.1063,
      "step": 120
    },
    {
      "epoch": 0.19530516431924883,
      "grad_norm": 0.8066917061805725,
      "learning_rate": 3.903903903903904e-05,
      "loss": 0.0996,
      "step": 130
    },
    {
      "epoch": 0.21032863849765257,
      "grad_norm": 0.4547995924949646,
      "learning_rate": 4.204204204204204e-05,
      "loss": 0.1047,
      "step": 140
    },
    {
      "epoch": 0.22535211267605634,
      "grad_norm": 0.7537536025047302,
      "learning_rate": 4.5045045045045046e-05,
      "loss": 0.0933,
      "step": 150
    },
    {
      "epoch": 0.22535211267605634,
      "eval_loss": 0.07271095365285873,
      "eval_runtime": 207.977,
      "eval_samples_per_second": 5.693,
      "eval_steps_per_second": 5.693,
      "step": 150
    },
    {
      "epoch": 0.2403755868544601,
      "grad_norm": 0.652043879032135,
      "learning_rate": 4.804804804804805e-05,
      "loss": 0.0726,
      "step": 160
    },
    {
      "epoch": 0.25539906103286386,
      "grad_norm": 0.7913267016410828,
      "learning_rate": 5.105105105105106e-05,
      "loss": 0.0503,
      "step": 170
    },
    {
      "epoch": 0.2704225352112676,
      "grad_norm": 0.6111043095588684,
      "learning_rate": 5.405405405405406e-05,
      "loss": 0.0682,
      "step": 180
    },
    {
      "epoch": 0.28544600938967135,
      "grad_norm": 0.6898893713951111,
      "learning_rate": 5.705705705705706e-05,
      "loss": 0.0651,
      "step": 190
    },
    {
      "epoch": 0.3004694835680751,
      "grad_norm": 0.5223749876022339,
      "learning_rate": 6.0060060060060066e-05,
      "loss": 0.0512,
      "step": 200
    },
    {
      "epoch": 0.3004694835680751,
      "eval_loss": 0.05808287858963013,
      "eval_runtime": 207.9535,
      "eval_samples_per_second": 5.694,
      "eval_steps_per_second": 5.694,
      "step": 200
    },
    {
      "epoch": 0.3154929577464789,
      "grad_norm": 0.7966965436935425,
      "learning_rate": 6.306306306306306e-05,
      "loss": 0.0562,
      "step": 210
    },
    {
      "epoch": 0.3305164319248826,
      "grad_norm": 0.4759034514427185,
      "learning_rate": 6.606606606606607e-05,
      "loss": 0.0591,
      "step": 220
    },
    {
      "epoch": 0.3455399061032864,
      "grad_norm": 0.421832412481308,
      "learning_rate": 6.906906906906907e-05,
      "loss": 0.0504,
      "step": 230
    },
    {
      "epoch": 0.36056338028169016,
      "grad_norm": 0.6044343113899231,
      "learning_rate": 7.207207207207208e-05,
      "loss": 0.0609,
      "step": 240
    },
    {
      "epoch": 0.3755868544600939,
      "grad_norm": 0.702263355255127,
      "learning_rate": 7.507507507507507e-05,
      "loss": 0.0619,
      "step": 250
    },
    {
      "epoch": 0.3755868544600939,
      "eval_loss": 0.04737601429224014,
      "eval_runtime": 207.9514,
      "eval_samples_per_second": 5.694,
      "eval_steps_per_second": 5.694,
      "step": 250
    },
    {
      "epoch": 0.39061032863849765,
      "grad_norm": 0.4086180627346039,
      "learning_rate": 7.807807807807808e-05,
      "loss": 0.0473,
      "step": 260
    },
    {
      "epoch": 0.4056338028169014,
      "grad_norm": 0.3927551209926605,
      "learning_rate": 8.108108108108109e-05,
      "loss": 0.0552,
      "step": 270
    },
    {
      "epoch": 0.42065727699530514,
      "grad_norm": 0.26502013206481934,
      "learning_rate": 8.408408408408409e-05,
      "loss": 0.0567,
      "step": 280
    },
    {
      "epoch": 0.4356807511737089,
      "grad_norm": 0.8767386674880981,
      "learning_rate": 8.70870870870871e-05,
      "loss": 0.0572,
      "step": 290
    },
    {
      "epoch": 0.4507042253521127,
      "grad_norm": 0.4874660074710846,
      "learning_rate": 9.009009009009009e-05,
      "loss": 0.0395,
      "step": 300
    },
    {
      "epoch": 0.4507042253521127,
      "eval_loss": 0.04596845433115959,
      "eval_runtime": 207.5439,
      "eval_samples_per_second": 5.705,
      "eval_steps_per_second": 5.705,
      "step": 300
    },
    {
      "epoch": 0.46572769953051646,
      "grad_norm": 0.43807005882263184,
      "learning_rate": 9.30930930930931e-05,
      "loss": 0.0444,
      "step": 310
    },
    {
      "epoch": 0.4807511737089202,
      "grad_norm": 0.7260940074920654,
      "learning_rate": 9.60960960960961e-05,
      "loss": 0.063,
      "step": 320
    },
    {
      "epoch": 0.49577464788732395,
      "grad_norm": 0.40789633989334106,
      "learning_rate": 9.90990990990991e-05,
      "loss": 0.0497,
      "step": 330
    },
    {
      "epoch": 0.5107981220657277,
      "grad_norm": 0.5078976154327393,
      "learning_rate": 9.999864944989638e-05,
      "loss": 0.058,
      "step": 340
    },
    {
      "epoch": 0.5258215962441315,
      "grad_norm": 0.3737334907054901,
      "learning_rate": 9.999203468625017e-05,
      "loss": 0.0476,
      "step": 350
    },
    {
      "epoch": 0.5258215962441315,
      "eval_loss": 0.045401681214571,
      "eval_runtime": 207.0191,
      "eval_samples_per_second": 5.719,
      "eval_steps_per_second": 5.719,
      "step": 350
    },
    {
      "epoch": 0.5408450704225352,
      "grad_norm": 0.5752517580986023,
      "learning_rate": 9.997990837719421e-05,
      "loss": 0.055,
      "step": 360
    },
    {
      "epoch": 0.5558685446009389,
      "grad_norm": 0.28732818365097046,
      "learning_rate": 9.996227185963554e-05,
      "loss": 0.0412,
      "step": 370
    },
    {
      "epoch": 0.5708920187793427,
      "grad_norm": 0.41930803656578064,
      "learning_rate": 9.993912707797329e-05,
      "loss": 0.0448,
      "step": 380
    },
    {
      "epoch": 0.5859154929577465,
      "grad_norm": 0.35693687200546265,
      "learning_rate": 9.99104765838842e-05,
      "loss": 0.0423,
      "step": 390
    },
    {
      "epoch": 0.6009389671361502,
      "grad_norm": 0.11278364062309265,
      "learning_rate": 9.987632353604151e-05,
      "loss": 0.0444,
      "step": 400
    },
    {
      "epoch": 0.6009389671361502,
      "eval_loss": 0.0407099612057209,
      "eval_runtime": 207.012,
      "eval_samples_per_second": 5.719,
      "eval_steps_per_second": 5.719,
      "step": 400
    },
    {
      "epoch": 0.615962441314554,
      "grad_norm": 0.41261032223701477,
      "learning_rate": 9.98366716997665e-05,
      "loss": 0.0478,
      "step": 410
    },
    {
      "epoch": 0.6309859154929578,
      "grad_norm": 0.4349944293498993,
      "learning_rate": 9.979152544661354e-05,
      "loss": 0.0381,
      "step": 420
    },
    {
      "epoch": 0.6460093896713615,
      "grad_norm": 0.609887957572937,
      "learning_rate": 9.974088975388802e-05,
      "loss": 0.0435,
      "step": 430
    },
    {
      "epoch": 0.6610328638497652,
      "grad_norm": 0.37665075063705444,
      "learning_rate": 9.968477020409766e-05,
      "loss": 0.0378,
      "step": 440
    },
    {
      "epoch": 0.676056338028169,
      "grad_norm": 0.30656006932258606,
      "learning_rate": 9.962317298433705e-05,
      "loss": 0.0543,
      "step": 450
    },
    {
      "epoch": 0.676056338028169,
      "eval_loss": 0.042538370937108994,
      "eval_runtime": 207.26,
      "eval_samples_per_second": 5.713,
      "eval_steps_per_second": 5.713,
      "step": 450
    },
    {
      "epoch": 0.6910798122065728,
      "grad_norm": 0.2620968520641327,
      "learning_rate": 9.955610488560551e-05,
      "loss": 0.0529,
      "step": 460
    },
    {
      "epoch": 0.7061032863849765,
      "grad_norm": 0.4962412118911743,
      "learning_rate": 9.948357330205842e-05,
      "loss": 0.0488,
      "step": 470
    },
    {
      "epoch": 0.7211267605633803,
      "grad_norm": 0.11893105506896973,
      "learning_rate": 9.940558623019201e-05,
      "loss": 0.0282,
      "step": 480
    },
    {
      "epoch": 0.7361502347417841,
      "grad_norm": 0.24164587259292603,
      "learning_rate": 9.932215226796172e-05,
      "loss": 0.0377,
      "step": 490
    },
    {
      "epoch": 0.7511737089201878,
      "grad_norm": 0.23037177324295044,
      "learning_rate": 9.923328061383435e-05,
      "loss": 0.0454,
      "step": 500
    },
    {
      "epoch": 0.7511737089201878,
      "eval_loss": 0.0371524840593338,
      "eval_runtime": 207.3218,
      "eval_samples_per_second": 5.711,
      "eval_steps_per_second": 5.711,
      "step": 500
    },
    {
      "epoch": 0.7661971830985915,
      "grad_norm": 0.1442870795726776,
      "learning_rate": 9.913898106577393e-05,
      "loss": 0.0382,
      "step": 510
    },
    {
      "epoch": 0.7812206572769953,
      "grad_norm": 0.2532236576080322,
      "learning_rate": 9.903926402016153e-05,
      "loss": 0.0344,
      "step": 520
    },
    {
      "epoch": 0.7962441314553991,
      "grad_norm": 0.14402006566524506,
      "learning_rate": 9.893414047064897e-05,
      "loss": 0.0398,
      "step": 530
    },
    {
      "epoch": 0.8112676056338028,
      "grad_norm": 0.2404378205537796,
      "learning_rate": 9.88236220069469e-05,
      "loss": 0.0381,
      "step": 540
    },
    {
      "epoch": 0.8262910798122066,
      "grad_norm": 0.27102166414260864,
      "learning_rate": 9.870772081354705e-05,
      "loss": 0.0562,
      "step": 550
    },
    {
      "epoch": 0.8262910798122066,
      "eval_loss": 0.03766689449548721,
      "eval_runtime": 207.5601,
      "eval_samples_per_second": 5.704,
      "eval_steps_per_second": 5.704,
      "step": 550
    },
    {
      "epoch": 0.8413145539906103,
      "grad_norm": 0.2679292857646942,
      "learning_rate": 9.858644966837878e-05,
      "loss": 0.0406,
      "step": 560
    },
    {
      "epoch": 0.856338028169014,
      "grad_norm": 0.0837322548031807,
      "learning_rate": 9.845982194140051e-05,
      "loss": 0.0381,
      "step": 570
    },
    {
      "epoch": 0.8713615023474178,
      "grad_norm": 0.1658477932214737,
      "learning_rate": 9.832785159312559e-05,
      "loss": 0.0404,
      "step": 580
    },
    {
      "epoch": 0.8863849765258216,
      "grad_norm": 0.2504975199699402,
      "learning_rate": 9.819055317308317e-05,
      "loss": 0.0365,
      "step": 590
    },
    {
      "epoch": 0.9014084507042254,
      "grad_norm": 0.24881432950496674,
      "learning_rate": 9.804794181821422e-05,
      "loss": 0.0336,
      "step": 600
    },
    {
      "epoch": 0.9014084507042254,
      "eval_loss": 0.036089226603507996,
      "eval_runtime": 208.6413,
      "eval_samples_per_second": 5.675,
      "eval_steps_per_second": 5.675,
      "step": 600
    },
    {
      "epoch": 0.9164319248826291,
      "grad_norm": 0.20400799810886383,
      "learning_rate": 9.790003325120261e-05,
      "loss": 0.0375,
      "step": 610
    },
    {
      "epoch": 0.9314553990610329,
      "grad_norm": 0.4139688313007355,
      "learning_rate": 9.774684377874178e-05,
      "loss": 0.0466,
      "step": 620
    },
    {
      "epoch": 0.9464788732394366,
      "grad_norm": 0.302133172750473,
      "learning_rate": 9.758839028973692e-05,
      "loss": 0.0308,
      "step": 630
    },
    {
      "epoch": 0.9615023474178404,
      "grad_norm": 0.26886606216430664,
      "learning_rate": 9.742469025344298e-05,
      "loss": 0.0416,
      "step": 640
    },
    {
      "epoch": 0.9765258215962441,
      "grad_norm": 0.22003549337387085,
      "learning_rate": 9.725576171753874e-05,
      "loss": 0.0494,
      "step": 650
    },
    {
      "epoch": 0.9765258215962441,
      "eval_loss": 0.03678149729967117,
      "eval_runtime": 209.7916,
      "eval_samples_per_second": 5.644,
      "eval_steps_per_second": 5.644,
      "step": 650
    },
    {
      "epoch": 0.9915492957746479,
      "grad_norm": 0.686739981174469,
      "learning_rate": 9.708162330613708e-05,
      "loss": 0.0367,
      "step": 660
    },
    {
      "epoch": 1.0065727699530516,
      "grad_norm": 0.1825464963912964,
      "learning_rate": 9.690229421773167e-05,
      "loss": 0.0369,
      "step": 670
    },
    {
      "epoch": 1.0215962441314554,
      "grad_norm": 0.442874014377594,
      "learning_rate": 9.67177942230804e-05,
      "loss": 0.0312,
      "step": 680
    },
    {
      "epoch": 1.036619718309859,
      "grad_norm": 0.45542624592781067,
      "learning_rate": 9.652814366302568e-05,
      "loss": 0.0392,
      "step": 690
    },
    {
      "epoch": 1.051643192488263,
      "grad_norm": 0.16704720258712769,
      "learning_rate": 9.633336344625185e-05,
      "loss": 0.0354,
      "step": 700
    },
    {
      "epoch": 1.051643192488263,
      "eval_loss": 0.0386400930583477,
      "eval_runtime": 210.8244,
      "eval_samples_per_second": 5.616,
      "eval_steps_per_second": 5.616,
      "step": 700
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.2750037610530853,
      "learning_rate": 9.61334750469801e-05,
      "loss": 0.0282,
      "step": 710
    },
    {
      "epoch": 1.0816901408450703,
      "grad_norm": 0.5251227021217346,
      "learning_rate": 9.592850050260089e-05,
      "loss": 0.0349,
      "step": 720
    },
    {
      "epoch": 1.0967136150234742,
      "grad_norm": 0.13910022377967834,
      "learning_rate": 9.571846241124446e-05,
      "loss": 0.0291,
      "step": 730
    },
    {
      "epoch": 1.1117370892018779,
      "grad_norm": 0.17034609615802765,
      "learning_rate": 9.55033839292893e-05,
      "loss": 0.0257,
      "step": 740
    },
    {
      "epoch": 1.1267605633802817,
      "grad_norm": 0.4655674397945404,
      "learning_rate": 9.52832887688093e-05,
      "loss": 0.029,
      "step": 750
    },
    {
      "epoch": 1.1267605633802817,
      "eval_loss": 0.037642158567905426,
      "eval_runtime": 212.6123,
      "eval_samples_per_second": 5.569,
      "eval_steps_per_second": 5.569,
      "step": 750
    },
    {
      "epoch": 1.1417840375586854,
      "grad_norm": 0.33910053968429565,
      "learning_rate": 9.50582011949595e-05,
      "loss": 0.0329,
      "step": 760
    },
    {
      "epoch": 1.1568075117370893,
      "grad_norm": 0.1620519608259201,
      "learning_rate": 9.482814602330084e-05,
      "loss": 0.0328,
      "step": 770
    },
    {
      "epoch": 1.171830985915493,
      "grad_norm": 0.3181590735912323,
      "learning_rate": 9.459314861706435e-05,
      "loss": 0.0363,
      "step": 780
    },
    {
      "epoch": 1.1868544600938966,
      "grad_norm": 0.3142762780189514,
      "learning_rate": 9.435323488435488e-05,
      "loss": 0.0279,
      "step": 790
    },
    {
      "epoch": 1.2018779342723005,
      "grad_norm": 0.34480199217796326,
      "learning_rate": 9.410843127529473e-05,
      "loss": 0.0301,
      "step": 800
    },
    {
      "epoch": 1.2018779342723005,
      "eval_loss": 0.03519855812191963,
      "eval_runtime": 213.9318,
      "eval_samples_per_second": 5.534,
      "eval_steps_per_second": 5.534,
      "step": 800
    },
    {
      "epoch": 1.2169014084507042,
      "grad_norm": 0.3214387893676758,
      "learning_rate": 9.385876477910765e-05,
      "loss": 0.0317,
      "step": 810
    },
    {
      "epoch": 1.231924882629108,
      "grad_norm": 0.20507577061653137,
      "learning_rate": 9.360426292114314e-05,
      "loss": 0.0291,
      "step": 820
    },
    {
      "epoch": 1.2469483568075117,
      "grad_norm": 0.2585867941379547,
      "learning_rate": 9.334495375984212e-05,
      "loss": 0.0279,
      "step": 830
    },
    {
      "epoch": 1.2619718309859156,
      "grad_norm": 0.21939821541309357,
      "learning_rate": 9.30808658836432e-05,
      "loss": 0.0293,
      "step": 840
    },
    {
      "epoch": 1.2769953051643192,
      "grad_norm": 0.42434459924697876,
      "learning_rate": 9.281202840783108e-05,
      "loss": 0.0321,
      "step": 850
    },
    {
      "epoch": 1.2769953051643192,
      "eval_loss": 0.034056928008794785,
      "eval_runtime": 215.0317,
      "eval_samples_per_second": 5.506,
      "eval_steps_per_second": 5.506,
      "step": 850
    },
    {
      "epoch": 1.292018779342723,
      "grad_norm": 0.22999003529548645,
      "learning_rate": 9.253847097132655e-05,
      "loss": 0.0243,
      "step": 860
    },
    {
      "epoch": 1.3070422535211268,
      "grad_norm": 0.1908729076385498,
      "learning_rate": 9.226022373341882e-05,
      "loss": 0.0247,
      "step": 870
    },
    {
      "epoch": 1.3220657276995305,
      "grad_norm": 0.16463080048561096,
      "learning_rate": 9.19773173704406e-05,
      "loss": 0.0245,
      "step": 880
    },
    {
      "epoch": 1.3370892018779343,
      "grad_norm": 0.5442132949829102,
      "learning_rate": 9.168978307238594e-05,
      "loss": 0.032,
      "step": 890
    },
    {
      "epoch": 1.352112676056338,
      "grad_norm": 0.11218319833278656,
      "learning_rate": 9.13976525394717e-05,
      "loss": 0.0271,
      "step": 900
    },
    {
      "epoch": 1.352112676056338,
      "eval_loss": 0.03430046886205673,
      "eval_runtime": 215.5861,
      "eval_samples_per_second": 5.492,
      "eval_steps_per_second": 5.492,
      "step": 900
    },
    {
      "epoch": 1.3671361502347419,
      "grad_norm": 0.13387706875801086,
      "learning_rate": 9.110095797864263e-05,
      "loss": 0.0224,
      "step": 910
    },
    {
      "epoch": 1.3821596244131455,
      "grad_norm": 0.20326955616474152,
      "learning_rate": 9.079973210002051e-05,
      "loss": 0.0318,
      "step": 920
    },
    {
      "epoch": 1.3971830985915492,
      "grad_norm": 0.2666754424571991,
      "learning_rate": 9.049400811329807e-05,
      "loss": 0.0322,
      "step": 930
    },
    {
      "epoch": 1.412206572769953,
      "grad_norm": 0.29467517137527466,
      "learning_rate": 9.01838197240775e-05,
      "loss": 0.0305,
      "step": 940
    },
    {
      "epoch": 1.4272300469483568,
      "grad_norm": 0.27433130145072937,
      "learning_rate": 8.986920113015461e-05,
      "loss": 0.0351,
      "step": 950
    },
    {
      "epoch": 1.4272300469483568,
      "eval_loss": 0.03298981115221977,
      "eval_runtime": 216.0148,
      "eval_samples_per_second": 5.481,
      "eval_steps_per_second": 5.481,
      "step": 950
    },
    {
      "epoch": 1.4422535211267606,
      "grad_norm": 0.23651893436908722,
      "learning_rate": 8.955018701774846e-05,
      "loss": 0.0262,
      "step": 960
    },
    {
      "epoch": 1.4572769953051643,
      "grad_norm": 0.2778637111186981,
      "learning_rate": 8.922681255767731e-05,
      "loss": 0.0347,
      "step": 970
    },
    {
      "epoch": 1.4723004694835682,
      "grad_norm": 0.18652495741844177,
      "learning_rate": 8.889911340148112e-05,
      "loss": 0.0308,
      "step": 980
    },
    {
      "epoch": 1.4873239436619718,
      "grad_norm": 0.17595553398132324,
      "learning_rate": 8.856712567749095e-05,
      "loss": 0.0278,
      "step": 990
    },
    {
      "epoch": 1.5023474178403755,
      "grad_norm": 0.4016423523426056,
      "learning_rate": 8.82308859868459e-05,
      "loss": 0.0244,
      "step": 1000
    },
    {
      "epoch": 1.5023474178403755,
      "eval_loss": 0.033029552549123764,
      "eval_runtime": 217.6657,
      "eval_samples_per_second": 5.44,
      "eval_steps_per_second": 5.44,
      "step": 1000
    },
    {
      "epoch": 1.5173708920187794,
      "grad_norm": 0.2475142925977707,
      "learning_rate": 8.789043139945795e-05,
      "loss": 0.0298,
      "step": 1010
    },
    {
      "epoch": 1.532394366197183,
      "grad_norm": 0.2059515416622162,
      "learning_rate": 8.754579944992491e-05,
      "loss": 0.036,
      "step": 1020
    },
    {
      "epoch": 1.5474178403755867,
      "grad_norm": 0.16493824124336243,
      "learning_rate": 8.719702813339248e-05,
      "loss": 0.0288,
      "step": 1030
    },
    {
      "epoch": 1.5624413145539906,
      "grad_norm": 0.20951515436172485,
      "learning_rate": 8.684415590136518e-05,
      "loss": 0.0277,
      "step": 1040
    },
    {
      "epoch": 1.5774647887323945,
      "grad_norm": 0.0918218195438385,
      "learning_rate": 8.648722165746722e-05,
      "loss": 0.0277,
      "step": 1050
    },
    {
      "epoch": 1.5774647887323945,
      "eval_loss": 0.034053802490234375,
      "eval_runtime": 219.1262,
      "eval_samples_per_second": 5.403,
      "eval_steps_per_second": 5.403,
      "step": 1050
    },
    {
      "epoch": 1.5924882629107981,
      "grad_norm": 0.09690333157777786,
      "learning_rate": 8.61262647531534e-05,
      "loss": 0.0251,
      "step": 1060
    },
    {
      "epoch": 1.6075117370892018,
      "grad_norm": 0.19759806990623474,
      "learning_rate": 8.576132498337068e-05,
      "loss": 0.0362,
      "step": 1070
    },
    {
      "epoch": 1.6225352112676057,
      "grad_norm": 0.2693615257740021,
      "learning_rate": 8.539244258217088e-05,
      "loss": 0.0319,
      "step": 1080
    },
    {
      "epoch": 1.6375586854460094,
      "grad_norm": 0.20964917540550232,
      "learning_rate": 8.501965821827485e-05,
      "loss": 0.032,
      "step": 1090
    },
    {
      "epoch": 1.652582159624413,
      "grad_norm": 0.2406897395849228,
      "learning_rate": 8.464301299058892e-05,
      "loss": 0.0231,
      "step": 1100
    },
    {
      "epoch": 1.652582159624413,
      "eval_loss": 0.034014929085969925,
      "eval_runtime": 219.9896,
      "eval_samples_per_second": 5.382,
      "eval_steps_per_second": 5.382,
      "step": 1100
    },
    {
      "epoch": 1.667605633802817,
      "grad_norm": 0.25754356384277344,
      "learning_rate": 8.426254842367374e-05,
      "loss": 0.0258,
      "step": 1110
    },
    {
      "epoch": 1.6826291079812208,
      "grad_norm": 0.2006315290927887,
      "learning_rate": 8.387830646316623e-05,
      "loss": 0.0334,
      "step": 1120
    },
    {
      "epoch": 1.6976525821596244,
      "grad_norm": 0.24149750173091888,
      "learning_rate": 8.349032947115525e-05,
      "loss": 0.0249,
      "step": 1130
    },
    {
      "epoch": 1.712676056338028,
      "grad_norm": 0.30418938398361206,
      "learning_rate": 8.309866022151107e-05,
      "loss": 0.0382,
      "step": 1140
    },
    {
      "epoch": 1.727699530516432,
      "grad_norm": 0.18970516324043274,
      "learning_rate": 8.270334189516983e-05,
      "loss": 0.0261,
      "step": 1150
    },
    {
      "epoch": 1.727699530516432,
      "eval_loss": 0.03269151970744133,
      "eval_runtime": 220.3747,
      "eval_samples_per_second": 5.373,
      "eval_steps_per_second": 5.373,
      "step": 1150
    },
    {
      "epoch": 1.7427230046948357,
      "grad_norm": 0.11434226483106613,
      "learning_rate": 8.230441807537277e-05,
      "loss": 0.0393,
      "step": 1160
    },
    {
      "epoch": 1.7577464788732393,
      "grad_norm": 0.18536998331546783,
      "learning_rate": 8.190193274286122e-05,
      "loss": 0.026,
      "step": 1170
    },
    {
      "epoch": 1.7727699530516432,
      "grad_norm": 0.1809949427843094,
      "learning_rate": 8.149593027102789e-05,
      "loss": 0.0295,
      "step": 1180
    },
    {
      "epoch": 1.787793427230047,
      "grad_norm": 0.2668783664703369,
      "learning_rate": 8.108645542102469e-05,
      "loss": 0.0255,
      "step": 1190
    },
    {
      "epoch": 1.8028169014084507,
      "grad_norm": 0.19622312486171722,
      "learning_rate": 8.067355333682798e-05,
      "loss": 0.0297,
      "step": 1200
    },
    {
      "epoch": 1.8028169014084507,
      "eval_loss": 0.03482316806912422,
      "eval_runtime": 220.3204,
      "eval_samples_per_second": 5.374,
      "eval_steps_per_second": 5.374,
      "step": 1200
    },
    {
      "epoch": 1.8178403755868544,
      "grad_norm": 0.28984394669532776,
      "learning_rate": 8.025726954026138e-05,
      "loss": 0.0228,
      "step": 1210
    },
    {
      "epoch": 1.8328638497652583,
      "grad_norm": 0.2442556470632553,
      "learning_rate": 7.983764992597716e-05,
      "loss": 0.0272,
      "step": 1220
    },
    {
      "epoch": 1.847887323943662,
      "grad_norm": 0.33816081285476685,
      "learning_rate": 7.94147407563964e-05,
      "loss": 0.0296,
      "step": 1230
    },
    {
      "epoch": 1.8629107981220656,
      "grad_norm": 0.3621574938297272,
      "learning_rate": 7.89885886566086e-05,
      "loss": 0.0333,
      "step": 1240
    },
    {
      "epoch": 1.8779342723004695,
      "grad_norm": 0.27948886156082153,
      "learning_rate": 7.855924060923141e-05,
      "loss": 0.027,
      "step": 1250
    },
    {
      "epoch": 1.8779342723004695,
      "eval_loss": 0.03339933604001999,
      "eval_runtime": 220.8261,
      "eval_samples_per_second": 5.362,
      "eval_steps_per_second": 5.362,
      "step": 1250
    },
    {
      "epoch": 1.8929577464788734,
      "grad_norm": 0.19844475388526917,
      "learning_rate": 7.812674394923077e-05,
      "loss": 0.03,
      "step": 1260
    },
    {
      "epoch": 1.907981220657277,
      "grad_norm": 0.17690996825695038,
      "learning_rate": 7.769114635870231e-05,
      "loss": 0.0275,
      "step": 1270
    },
    {
      "epoch": 1.9230046948356807,
      "grad_norm": 0.26047441363334656,
      "learning_rate": 7.725249586161463e-05,
      "loss": 0.0294,
      "step": 1280
    },
    {
      "epoch": 1.9380281690140846,
      "grad_norm": 0.16973556578159332,
      "learning_rate": 7.68108408185145e-05,
      "loss": 0.031,
      "step": 1290
    },
    {
      "epoch": 1.9530516431924883,
      "grad_norm": 0.18663664162158966,
      "learning_rate": 7.636622992119536e-05,
      "loss": 0.0417,
      "step": 1300
    },
    {
      "epoch": 1.9530516431924883,
      "eval_loss": 0.03484388440847397,
      "eval_runtime": 221.5077,
      "eval_samples_per_second": 5.345,
      "eval_steps_per_second": 5.345,
      "step": 1300
    },
    {
      "epoch": 1.968075117370892,
      "grad_norm": 0.2905583083629608,
      "learning_rate": 7.591871218732902e-05,
      "loss": 0.0329,
      "step": 1310
    },
    {
      "epoch": 1.9830985915492958,
      "grad_norm": 0.18613983690738678,
      "learning_rate": 7.54683369550616e-05,
      "loss": 0.0364,
      "step": 1320
    },
    {
      "epoch": 1.9981220657276997,
      "grad_norm": 0.1276523619890213,
      "learning_rate": 7.501515387757404e-05,
      "loss": 0.0269,
      "step": 1330
    },
    {
      "epoch": 2.013145539906103,
      "grad_norm": 0.22376187145709991,
      "learning_rate": 7.455921291760796e-05,
      "loss": 0.0313,
      "step": 1340
    },
    {
      "epoch": 2.028169014084507,
      "grad_norm": 0.09134189784526825,
      "learning_rate": 7.410056434195725e-05,
      "loss": 0.0173,
      "step": 1350
    },
    {
      "epoch": 2.028169014084507,
      "eval_loss": 0.03278001397848129,
      "eval_runtime": 221.4183,
      "eval_samples_per_second": 5.347,
      "eval_steps_per_second": 5.347,
      "step": 1350
    },
    {
      "epoch": 2.043192488262911,
      "grad_norm": 0.13673311471939087,
      "learning_rate": 7.363925871592629e-05,
      "loss": 0.0205,
      "step": 1360
    },
    {
      "epoch": 2.0582159624413148,
      "grad_norm": 0.161691814661026,
      "learning_rate": 7.317534689775528e-05,
      "loss": 0.0201,
      "step": 1370
    },
    {
      "epoch": 2.073239436619718,
      "grad_norm": 0.1258237212896347,
      "learning_rate": 7.270888003301304e-05,
      "loss": 0.0159,
      "step": 1380
    },
    {
      "epoch": 2.088262910798122,
      "grad_norm": 0.29005277156829834,
      "learning_rate": 7.22399095489584e-05,
      "loss": 0.0222,
      "step": 1390
    },
    {
      "epoch": 2.103286384976526,
      "grad_norm": 0.11337249726057053,
      "learning_rate": 7.176848714887042e-05,
      "loss": 0.0207,
      "step": 1400
    },
    {
      "epoch": 2.103286384976526,
      "eval_loss": 0.03226166218519211,
      "eval_runtime": 221.8527,
      "eval_samples_per_second": 5.337,
      "eval_steps_per_second": 5.337,
      "step": 1400
    },
    {
      "epoch": 2.1183098591549294,
      "grad_norm": 0.16560769081115723,
      "learning_rate": 7.129466480634806e-05,
      "loss": 0.0225,
      "step": 1410
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 0.3505042493343353,
      "learning_rate": 7.081849475958042e-05,
      "loss": 0.0187,
      "step": 1420
    },
    {
      "epoch": 2.148356807511737,
      "grad_norm": 0.16660155355930328,
      "learning_rate": 7.034002950558723e-05,
      "loss": 0.0185,
      "step": 1430
    },
    {
      "epoch": 2.1633802816901406,
      "grad_norm": 0.13852401077747345,
      "learning_rate": 6.985932179443144e-05,
      "loss": 0.0192,
      "step": 1440
    },
    {
      "epoch": 2.1784037558685445,
      "grad_norm": 0.21416670083999634,
      "learning_rate": 6.937642462340342e-05,
      "loss": 0.0223,
      "step": 1450
    },
    {
      "epoch": 2.1784037558685445,
      "eval_loss": 0.03254189342260361,
      "eval_runtime": 222.3283,
      "eval_samples_per_second": 5.325,
      "eval_steps_per_second": 5.325,
      "step": 1450
    },
    {
      "epoch": 2.1934272300469484,
      "grad_norm": 0.1465296745300293,
      "learning_rate": 6.889139123117817e-05,
      "loss": 0.0191,
      "step": 1460
    },
    {
      "epoch": 2.2084507042253523,
      "grad_norm": 0.19770440459251404,
      "learning_rate": 6.840427509194575e-05,
      "loss": 0.0209,
      "step": 1470
    },
    {
      "epoch": 2.2234741784037557,
      "grad_norm": 0.22087684273719788,
      "learning_rate": 6.791512990951597e-05,
      "loss": 0.0149,
      "step": 1480
    },
    {
      "epoch": 2.2384976525821596,
      "grad_norm": 0.34373584389686584,
      "learning_rate": 6.74240096113975e-05,
      "loss": 0.0211,
      "step": 1490
    },
    {
      "epoch": 2.2535211267605635,
      "grad_norm": 0.12940266728401184,
      "learning_rate": 6.693096834285256e-05,
      "loss": 0.0107,
      "step": 1500
    },
    {
      "epoch": 2.2535211267605635,
      "eval_loss": 0.035872478038072586,
      "eval_runtime": 223.0953,
      "eval_samples_per_second": 5.307,
      "eval_steps_per_second": 5.307,
      "step": 1500
    },
    {
      "epoch": 2.2685446009389674,
      "grad_norm": 0.39470982551574707,
      "learning_rate": 6.643606046092732e-05,
      "loss": 0.0226,
      "step": 1510
    },
    {
      "epoch": 2.283568075117371,
      "grad_norm": 0.22834685444831848,
      "learning_rate": 6.593934052845929e-05,
      "loss": 0.0253,
      "step": 1520
    },
    {
      "epoch": 2.2985915492957747,
      "grad_norm": 0.41707947850227356,
      "learning_rate": 6.544086330806181e-05,
      "loss": 0.0205,
      "step": 1530
    },
    {
      "epoch": 2.3136150234741786,
      "grad_norm": 0.1707378327846527,
      "learning_rate": 6.494068375608646e-05,
      "loss": 0.0197,
      "step": 1540
    },
    {
      "epoch": 2.328638497652582,
      "grad_norm": 0.17165277898311615,
      "learning_rate": 6.443885701656432e-05,
      "loss": 0.0182,
      "step": 1550
    },
    {
      "epoch": 2.328638497652582,
      "eval_loss": 0.03318268060684204,
      "eval_runtime": 223.282,
      "eval_samples_per_second": 5.303,
      "eval_steps_per_second": 5.303,
      "step": 1550
    },
    {
      "epoch": 2.343661971830986,
      "grad_norm": 0.14184924960136414,
      "learning_rate": 6.393543841512632e-05,
      "loss": 0.02,
      "step": 1560
    },
    {
      "epoch": 2.35868544600939,
      "grad_norm": 0.31733861565589905,
      "learning_rate": 6.343048345290386e-05,
      "loss": 0.0144,
      "step": 1570
    },
    {
      "epoch": 2.3737089201877932,
      "grad_norm": 0.16232426464557648,
      "learning_rate": 6.292404780040961e-05,
      "loss": 0.023,
      "step": 1580
    },
    {
      "epoch": 2.388732394366197,
      "grad_norm": 0.18042533099651337,
      "learning_rate": 6.241618729140018e-05,
      "loss": 0.0157,
      "step": 1590
    },
    {
      "epoch": 2.403755868544601,
      "grad_norm": 0.4602644741535187,
      "learning_rate": 6.190695791672042e-05,
      "loss": 0.0187,
      "step": 1600
    },
    {
      "epoch": 2.403755868544601,
      "eval_loss": 0.032255951315164566,
      "eval_runtime": 222.5999,
      "eval_samples_per_second": 5.319,
      "eval_steps_per_second": 5.319,
      "step": 1600
    },
    {
      "epoch": 2.418779342723005,
      "grad_norm": 0.0981522649526596,
      "learning_rate": 6.139641581813052e-05,
      "loss": 0.0183,
      "step": 1610
    },
    {
      "epoch": 2.4338028169014083,
      "grad_norm": 0.1687208116054535,
      "learning_rate": 6.088461728211642e-05,
      "loss": 0.0177,
      "step": 1620
    },
    {
      "epoch": 2.448826291079812,
      "grad_norm": 0.19458502531051636,
      "learning_rate": 6.0371618733684474e-05,
      "loss": 0.0117,
      "step": 1630
    },
    {
      "epoch": 2.463849765258216,
      "grad_norm": 0.12449292838573456,
      "learning_rate": 5.9857476730140485e-05,
      "loss": 0.0175,
      "step": 1640
    },
    {
      "epoch": 2.4788732394366195,
      "grad_norm": 0.20214103162288666,
      "learning_rate": 5.9342247954854466e-05,
      "loss": 0.018,
      "step": 1650
    },
    {
      "epoch": 2.4788732394366195,
      "eval_loss": 0.03270378336310387,
      "eval_runtime": 221.3125,
      "eval_samples_per_second": 5.35,
      "eval_steps_per_second": 5.35,
      "step": 1650
    },
    {
      "epoch": 2.4938967136150234,
      "grad_norm": 0.5592941641807556,
      "learning_rate": 5.8825989211011335e-05,
      "loss": 0.022,
      "step": 1660
    },
    {
      "epoch": 2.5089201877934273,
      "grad_norm": 0.20930063724517822,
      "learning_rate": 5.830875741534852e-05,
      "loss": 0.0159,
      "step": 1670
    },
    {
      "epoch": 2.523943661971831,
      "grad_norm": 0.24804644286632538,
      "learning_rate": 5.7790609591880826e-05,
      "loss": 0.0239,
      "step": 1680
    },
    {
      "epoch": 2.5389671361502346,
      "grad_norm": 0.17634353041648865,
      "learning_rate": 5.727160286561386e-05,
      "loss": 0.0224,
      "step": 1690
    },
    {
      "epoch": 2.5539906103286385,
      "grad_norm": 0.23160183429718018,
      "learning_rate": 5.675179445624581e-05,
      "loss": 0.0205,
      "step": 1700
    },
    {
      "epoch": 2.5539906103286385,
      "eval_loss": 0.03496889770030975,
      "eval_runtime": 221.4535,
      "eval_samples_per_second": 5.346,
      "eval_steps_per_second": 5.346,
      "step": 1700
    },
    {
      "epoch": 2.5690140845070424,
      "grad_norm": 0.09332422912120819,
      "learning_rate": 5.62312416718593e-05,
      "loss": 0.0164,
      "step": 1710
    },
    {
      "epoch": 2.584037558685446,
      "grad_norm": 0.09970489144325256,
      "learning_rate": 5.5710001902603116e-05,
      "loss": 0.0257,
      "step": 1720
    },
    {
      "epoch": 2.5990610328638497,
      "grad_norm": 0.17599700391292572,
      "learning_rate": 5.5188132614365094e-05,
      "loss": 0.0198,
      "step": 1730
    },
    {
      "epoch": 2.6140845070422536,
      "grad_norm": 0.11823566257953644,
      "learning_rate": 5.4665691342436565e-05,
      "loss": 0.0152,
      "step": 1740
    },
    {
      "epoch": 2.629107981220657,
      "grad_norm": 0.31226423382759094,
      "learning_rate": 5.414273568516919e-05,
      "loss": 0.0182,
      "step": 1750
    },
    {
      "epoch": 2.629107981220657,
      "eval_loss": 0.032272376120090485,
      "eval_runtime": 221.7599,
      "eval_samples_per_second": 5.339,
      "eval_steps_per_second": 5.339,
      "step": 1750
    },
    {
      "epoch": 2.644131455399061,
      "grad_norm": 0.12536101043224335,
      "learning_rate": 5.361932329762481e-05,
      "loss": 0.0197,
      "step": 1760
    },
    {
      "epoch": 2.659154929577465,
      "grad_norm": 0.11700065433979034,
      "learning_rate": 5.309551188521914e-05,
      "loss": 0.0197,
      "step": 1770
    },
    {
      "epoch": 2.6741784037558687,
      "grad_norm": 0.30704525113105774,
      "learning_rate": 5.2571359197359704e-05,
      "loss": 0.0161,
      "step": 1780
    },
    {
      "epoch": 2.6892018779342726,
      "grad_norm": 0.06998223811388016,
      "learning_rate": 5.2046923021079175e-05,
      "loss": 0.0151,
      "step": 1790
    },
    {
      "epoch": 2.704225352112676,
      "grad_norm": 0.09223470091819763,
      "learning_rate": 5.1522261174664346e-05,
      "loss": 0.0202,
      "step": 1800
    },
    {
      "epoch": 2.704225352112676,
      "eval_loss": 0.032467469573020935,
      "eval_runtime": 221.8494,
      "eval_samples_per_second": 5.337,
      "eval_steps_per_second": 5.337,
      "step": 1800
    },
    {
      "epoch": 2.71924882629108,
      "grad_norm": 0.3536123037338257,
      "learning_rate": 5.0997431501281835e-05,
      "loss": 0.0262,
      "step": 1810
    },
    {
      "epoch": 2.7342723004694838,
      "grad_norm": 0.2920973300933838,
      "learning_rate": 5.0472491862600915e-05,
      "loss": 0.0202,
      "step": 1820
    },
    {
      "epoch": 2.749295774647887,
      "grad_norm": 0.23328785598278046,
      "learning_rate": 4.994750013241435e-05,
      "loss": 0.022,
      "step": 1830
    },
    {
      "epoch": 2.764319248826291,
      "grad_norm": 0.13965994119644165,
      "learning_rate": 4.9422514190257974e-05,
      "loss": 0.0201,
      "step": 1840
    },
    {
      "epoch": 2.779342723004695,
      "grad_norm": 0.12157478928565979,
      "learning_rate": 4.88975919150294e-05,
      "loss": 0.0218,
      "step": 1850
    },
    {
      "epoch": 2.779342723004695,
      "eval_loss": 0.03232384845614433,
      "eval_runtime": 221.306,
      "eval_samples_per_second": 5.35,
      "eval_steps_per_second": 5.35,
      "step": 1850
    },
    {
      "epoch": 2.7943661971830984,
      "grad_norm": 0.17367467284202576,
      "learning_rate": 4.83727911786071e-05,
      "loss": 0.022,
      "step": 1860
    },
    {
      "epoch": 2.8093896713615023,
      "grad_norm": 0.30343097448349,
      "learning_rate": 4.7848169839470145e-05,
      "loss": 0.0224,
      "step": 1870
    },
    {
      "epoch": 2.824413145539906,
      "grad_norm": 0.12888950109481812,
      "learning_rate": 4.7323785736319244e-05,
      "loss": 0.021,
      "step": 1880
    },
    {
      "epoch": 2.8394366197183096,
      "grad_norm": 0.15008972585201263,
      "learning_rate": 4.679969668170024e-05,
      "loss": 0.0232,
      "step": 1890
    },
    {
      "epoch": 2.8544600938967135,
      "grad_norm": 0.10453104227781296,
      "learning_rate": 4.627596045563031e-05,
      "loss": 0.0179,
      "step": 1900
    },
    {
      "epoch": 2.8544600938967135,
      "eval_loss": 0.03188331797719002,
      "eval_runtime": 221.3325,
      "eval_samples_per_second": 5.349,
      "eval_steps_per_second": 5.349,
      "step": 1900
    },
    {
      "epoch": 2.8694835680751174,
      "grad_norm": 0.20146887004375458,
      "learning_rate": 4.575263479922783e-05,
      "loss": 0.0147,
      "step": 1910
    },
    {
      "epoch": 2.8845070422535213,
      "grad_norm": 0.10789839178323746,
      "learning_rate": 4.522977740834651e-05,
      "loss": 0.0118,
      "step": 1920
    },
    {
      "epoch": 2.8995305164319247,
      "grad_norm": 0.1628926396369934,
      "learning_rate": 4.4707445927214456e-05,
      "loss": 0.0126,
      "step": 1930
    },
    {
      "epoch": 2.9145539906103286,
      "grad_norm": 0.14242199063301086,
      "learning_rate": 4.4185697942079115e-05,
      "loss": 0.0199,
      "step": 1940
    },
    {
      "epoch": 2.9295774647887325,
      "grad_norm": 0.15721061825752258,
      "learning_rate": 4.366459097485832e-05,
      "loss": 0.0213,
      "step": 1950
    },
    {
      "epoch": 2.9295774647887325,
      "eval_loss": 0.03299366310238838,
      "eval_runtime": 221.4895,
      "eval_samples_per_second": 5.346,
      "eval_steps_per_second": 5.346,
      "step": 1950
    },
    {
      "epoch": 2.9446009389671364,
      "grad_norm": 0.2670339047908783,
      "learning_rate": 4.314418247679866e-05,
      "loss": 0.0231,
      "step": 1960
    },
    {
      "epoch": 2.95962441314554,
      "grad_norm": 0.18777424097061157,
      "learning_rate": 4.26245298221416e-05,
      "loss": 0.0141,
      "step": 1970
    },
    {
      "epoch": 2.9746478873239437,
      "grad_norm": 0.08959437906742096,
      "learning_rate": 4.2105690301798014e-05,
      "loss": 0.0143,
      "step": 1980
    },
    {
      "epoch": 2.9896713615023476,
      "grad_norm": 0.23774972558021545,
      "learning_rate": 4.158772111703194e-05,
      "loss": 0.0167,
      "step": 1990
    },
    {
      "epoch": 3.004694835680751,
      "grad_norm": 0.025070570409297943,
      "learning_rate": 4.107067937315429e-05,
      "loss": 0.0104,
      "step": 2000
    },
    {
      "epoch": 3.004694835680751,
      "eval_loss": 0.03279737010598183,
      "eval_runtime": 222.0291,
      "eval_samples_per_second": 5.333,
      "eval_steps_per_second": 5.333,
      "step": 2000
    },
    {
      "epoch": 3.019718309859155,
      "grad_norm": 0.12388351559638977,
      "learning_rate": 4.055462207322698e-05,
      "loss": 0.01,
      "step": 2010
    },
    {
      "epoch": 3.034741784037559,
      "grad_norm": 0.024688435718417168,
      "learning_rate": 4.003960611177855e-05,
      "loss": 0.0101,
      "step": 2020
    },
    {
      "epoch": 3.0497652582159622,
      "grad_norm": 0.24434754252433777,
      "learning_rate": 3.952568826853152e-05,
      "loss": 0.0094,
      "step": 2030
    },
    {
      "epoch": 3.064788732394366,
      "grad_norm": 0.05700072646141052,
      "learning_rate": 3.901292520214256e-05,
      "loss": 0.0068,
      "step": 2040
    },
    {
      "epoch": 3.07981220657277,
      "grad_norm": 0.6166403889656067,
      "learning_rate": 3.850137344395598e-05,
      "loss": 0.0097,
      "step": 2050
    },
    {
      "epoch": 3.07981220657277,
      "eval_loss": 0.035948336124420166,
      "eval_runtime": 221.527,
      "eval_samples_per_second": 5.345,
      "eval_steps_per_second": 5.345,
      "step": 2050
    },
    {
      "epoch": 3.094835680751174,
      "grad_norm": 0.2868582308292389,
      "learning_rate": 3.799108939177118e-05,
      "loss": 0.0076,
      "step": 2060
    },
    {
      "epoch": 3.1098591549295773,
      "grad_norm": 0.201212540268898,
      "learning_rate": 3.7482129303624934e-05,
      "loss": 0.0148,
      "step": 2070
    },
    {
      "epoch": 3.124882629107981,
      "grad_norm": 0.512485921382904,
      "learning_rate": 3.697454929158901e-05,
      "loss": 0.0115,
      "step": 2080
    },
    {
      "epoch": 3.139906103286385,
      "grad_norm": 0.23441621661186218,
      "learning_rate": 3.6468405315583854e-05,
      "loss": 0.0147,
      "step": 2090
    },
    {
      "epoch": 3.1549295774647885,
      "grad_norm": 0.20293329656124115,
      "learning_rate": 3.59637531772092e-05,
      "loss": 0.0103,
      "step": 2100
    },
    {
      "epoch": 3.1549295774647885,
      "eval_loss": 0.0363353006541729,
      "eval_runtime": 221.0841,
      "eval_samples_per_second": 5.355,
      "eval_steps_per_second": 5.355,
      "step": 2100
    },
    {
      "epoch": 3.1699530516431924,
      "grad_norm": 0.2901355028152466,
      "learning_rate": 3.546064851359192e-05,
      "loss": 0.0082,
      "step": 2110
    },
    {
      "epoch": 3.1849765258215963,
      "grad_norm": 0.11046724766492844,
      "learning_rate": 3.495914679125212e-05,
      "loss": 0.0097,
      "step": 2120
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.07721813023090363,
      "learning_rate": 3.445930329998819e-05,
      "loss": 0.0103,
      "step": 2130
    },
    {
      "epoch": 3.2150234741784036,
      "grad_norm": 0.26007962226867676,
      "learning_rate": 3.396117314678097e-05,
      "loss": 0.0142,
      "step": 2140
    },
    {
      "epoch": 3.2300469483568075,
      "grad_norm": 0.4068683087825775,
      "learning_rate": 3.3464811249718474e-05,
      "loss": 0.0131,
      "step": 2150
    },
    {
      "epoch": 3.2300469483568075,
      "eval_loss": 0.03587425872683525,
      "eval_runtime": 221.6748,
      "eval_samples_per_second": 5.341,
      "eval_steps_per_second": 5.341,
      "step": 2150
    },
    {
      "epoch": 3.2450704225352114,
      "grad_norm": 0.18533091247081757,
      "learning_rate": 3.297027233194114e-05,
      "loss": 0.0126,
      "step": 2160
    },
    {
      "epoch": 3.260093896713615,
      "grad_norm": 0.10798130184412003,
      "learning_rate": 3.2477610915608704e-05,
      "loss": 0.0081,
      "step": 2170
    },
    {
      "epoch": 3.2751173708920187,
      "grad_norm": 0.19456113874912262,
      "learning_rate": 3.1986881315889315e-05,
      "loss": 0.0091,
      "step": 2180
    },
    {
      "epoch": 3.2901408450704226,
      "grad_norm": 0.26297393441200256,
      "learning_rate": 3.149813763497124e-05,
      "loss": 0.0155,
      "step": 2190
    },
    {
      "epoch": 3.3051643192488265,
      "grad_norm": 0.26686403155326843,
      "learning_rate": 3.101143375609818e-05,
      "loss": 0.0149,
      "step": 2200
    },
    {
      "epoch": 3.3051643192488265,
      "eval_loss": 0.036220405250787735,
      "eval_runtime": 221.9816,
      "eval_samples_per_second": 5.334,
      "eval_steps_per_second": 5.334,
      "step": 2200
    },
    {
      "epoch": 3.32018779342723,
      "grad_norm": 0.3246915638446808,
      "learning_rate": 3.0526823337628915e-05,
      "loss": 0.0129,
      "step": 2210
    },
    {
      "epoch": 3.335211267605634,
      "grad_norm": 0.4109603464603424,
      "learning_rate": 3.004435980712129e-05,
      "loss": 0.0098,
      "step": 2220
    },
    {
      "epoch": 3.3502347417840377,
      "grad_norm": 0.4035777747631073,
      "learning_rate": 2.9564096355442116e-05,
      "loss": 0.0128,
      "step": 2230
    },
    {
      "epoch": 3.365258215962441,
      "grad_norm": 0.1403612494468689,
      "learning_rate": 2.9086085930902824e-05,
      "loss": 0.0137,
      "step": 2240
    },
    {
      "epoch": 3.380281690140845,
      "grad_norm": 0.21675720810890198,
      "learning_rate": 2.8610381233422058e-05,
      "loss": 0.0083,
      "step": 2250
    },
    {
      "epoch": 3.380281690140845,
      "eval_loss": 0.03651123493909836,
      "eval_runtime": 222.0068,
      "eval_samples_per_second": 5.333,
      "eval_steps_per_second": 5.333,
      "step": 2250
    },
    {
      "epoch": 3.395305164319249,
      "grad_norm": 0.30155134201049805,
      "learning_rate": 2.8137034708715592e-05,
      "loss": 0.0098,
      "step": 2260
    },
    {
      "epoch": 3.4103286384976528,
      "grad_norm": 0.05509026348590851,
      "learning_rate": 2.7666098542514273e-05,
      "loss": 0.0096,
      "step": 2270
    },
    {
      "epoch": 3.425352112676056,
      "grad_norm": 0.25339803099632263,
      "learning_rate": 2.719762465481055e-05,
      "loss": 0.0123,
      "step": 2280
    },
    {
      "epoch": 3.44037558685446,
      "grad_norm": 0.052862461656332016,
      "learning_rate": 2.6731664694134473e-05,
      "loss": 0.0118,
      "step": 2290
    },
    {
      "epoch": 3.455399061032864,
      "grad_norm": 0.16749408841133118,
      "learning_rate": 2.6268270031859476e-05,
      "loss": 0.0115,
      "step": 2300
    },
    {
      "epoch": 3.455399061032864,
      "eval_loss": 0.035897351801395416,
      "eval_runtime": 221.6066,
      "eval_samples_per_second": 5.343,
      "eval_steps_per_second": 5.343,
      "step": 2300
    },
    {
      "epoch": 3.4704225352112674,
      "grad_norm": 0.28338125348091125,
      "learning_rate": 2.580749175653877e-05,
      "loss": 0.0097,
      "step": 2310
    },
    {
      "epoch": 3.4854460093896713,
      "grad_norm": 0.36096012592315674,
      "learning_rate": 2.5349380668272905e-05,
      "loss": 0.0122,
      "step": 2320
    },
    {
      "epoch": 3.500469483568075,
      "grad_norm": 0.20076704025268555,
      "learning_rate": 2.489398727310908e-05,
      "loss": 0.0096,
      "step": 2330
    },
    {
      "epoch": 3.5154929577464786,
      "grad_norm": 0.28074732422828674,
      "learning_rate": 2.4441361777473066e-05,
      "loss": 0.0088,
      "step": 2340
    },
    {
      "epoch": 3.5305164319248825,
      "grad_norm": 0.5027658343315125,
      "learning_rate": 2.3991554082633912e-05,
      "loss": 0.0111,
      "step": 2350
    },
    {
      "epoch": 3.5305164319248825,
      "eval_loss": 0.038713712245225906,
      "eval_runtime": 221.2043,
      "eval_samples_per_second": 5.353,
      "eval_steps_per_second": 5.353,
      "step": 2350
    },
    {
      "epoch": 3.5455399061032864,
      "grad_norm": 0.11489449441432953,
      "learning_rate": 2.354461377920239e-05,
      "loss": 0.009,
      "step": 2360
    },
    {
      "epoch": 3.5605633802816903,
      "grad_norm": 0.42795881628990173,
      "learning_rate": 2.3100590141663807e-05,
      "loss": 0.0123,
      "step": 2370
    },
    {
      "epoch": 3.575586854460094,
      "grad_norm": 0.34739652276039124,
      "learning_rate": 2.265953212294551e-05,
      "loss": 0.0143,
      "step": 2380
    },
    {
      "epoch": 3.5906103286384976,
      "grad_norm": 0.5267793536186218,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.015,
      "step": 2390
    },
    {
      "epoch": 3.6056338028169015,
      "grad_norm": 0.2632468342781067,
      "learning_rate": 2.1786507113543457e-05,
      "loss": 0.0094,
      "step": 2400
    },
    {
      "epoch": 3.6056338028169015,
      "eval_loss": 0.03762707859277725,
      "eval_runtime": 222.4043,
      "eval_samples_per_second": 5.324,
      "eval_steps_per_second": 5.324,
      "step": 2400
    },
    {
      "epoch": 3.6206572769953054,
      "grad_norm": 0.15667402744293213,
      "learning_rate": 2.1354636372532523e-05,
      "loss": 0.0149,
      "step": 2410
    },
    {
      "epoch": 3.635680751173709,
      "grad_norm": 0.22469666600227356,
      "learning_rate": 2.092592373907617e-05,
      "loss": 0.0106,
      "step": 2420
    },
    {
      "epoch": 3.6507042253521127,
      "grad_norm": 0.2074466347694397,
      "learning_rate": 2.0500416478086932e-05,
      "loss": 0.0079,
      "step": 2430
    },
    {
      "epoch": 3.6657276995305166,
      "grad_norm": 0.21211783587932587,
      "learning_rate": 2.0078161501089954e-05,
      "loss": 0.0106,
      "step": 2440
    },
    {
      "epoch": 3.68075117370892,
      "grad_norm": 0.13224627077579498,
      "learning_rate": 1.9659205361050982e-05,
      "loss": 0.0051,
      "step": 2450
    },
    {
      "epoch": 3.68075117370892,
      "eval_loss": 0.03760899230837822,
      "eval_runtime": 222.6155,
      "eval_samples_per_second": 5.319,
      "eval_steps_per_second": 5.319,
      "step": 2450
    },
    {
      "epoch": 3.695774647887324,
      "grad_norm": 0.07832111418247223,
      "learning_rate": 1.924359424724408e-05,
      "loss": 0.0143,
      "step": 2460
    },
    {
      "epoch": 3.710798122065728,
      "grad_norm": 0.34279316663742065,
      "learning_rate": 1.8831373980159296e-05,
      "loss": 0.0129,
      "step": 2470
    },
    {
      "epoch": 3.7258215962441312,
      "grad_norm": 0.13393940031528473,
      "learning_rate": 1.8422590006450947e-05,
      "loss": 0.0062,
      "step": 2480
    },
    {
      "epoch": 3.740845070422535,
      "grad_norm": 0.05350172892212868,
      "learning_rate": 1.801728739392731e-05,
      "loss": 0.0087,
      "step": 2490
    },
    {
      "epoch": 3.755868544600939,
      "grad_norm": 0.0817003846168518,
      "learning_rate": 1.7615510826581904e-05,
      "loss": 0.0053,
      "step": 2500
    },
    {
      "epoch": 3.755868544600939,
      "eval_loss": 0.03752273693680763,
      "eval_runtime": 222.372,
      "eval_samples_per_second": 5.324,
      "eval_steps_per_second": 5.324,
      "step": 2500
    },
    {
      "epoch": 3.770892018779343,
      "grad_norm": 0.21435320377349854,
      "learning_rate": 1.7217304599667146e-05,
      "loss": 0.0111,
      "step": 2510
    },
    {
      "epoch": 3.7859154929577463,
      "grad_norm": 0.15997354686260223,
      "learning_rate": 1.6822712614810893e-05,
      "loss": 0.0101,
      "step": 2520
    },
    {
      "epoch": 3.80093896713615,
      "grad_norm": 0.182434543967247,
      "learning_rate": 1.643177837517631e-05,
      "loss": 0.0093,
      "step": 2530
    },
    {
      "epoch": 3.815962441314554,
      "grad_norm": 0.22743846476078033,
      "learning_rate": 1.6044544980665767e-05,
      "loss": 0.0125,
      "step": 2540
    },
    {
      "epoch": 3.830985915492958,
      "grad_norm": 0.33443447947502136,
      "learning_rate": 1.5661055123169126e-05,
      "loss": 0.0078,
      "step": 2550
    },
    {
      "epoch": 3.830985915492958,
      "eval_loss": 0.037725865840911865,
      "eval_runtime": 221.8726,
      "eval_samples_per_second": 5.336,
      "eval_steps_per_second": 5.336,
      "step": 2550
    },
    {
      "epoch": 3.8460093896713614,
      "grad_norm": 0.5252001285552979,
      "learning_rate": 1.5281351081856974e-05,
      "loss": 0.0118,
      "step": 2560
    },
    {
      "epoch": 3.8610328638497653,
      "grad_norm": 0.17551793158054352,
      "learning_rate": 1.4905474718519491e-05,
      "loss": 0.007,
      "step": 2570
    },
    {
      "epoch": 3.876056338028169,
      "grad_norm": 0.1480369120836258,
      "learning_rate": 1.453346747295119e-05,
      "loss": 0.0141,
      "step": 2580
    },
    {
      "epoch": 3.8910798122065726,
      "grad_norm": 0.19130997359752655,
      "learning_rate": 1.4165370358382274e-05,
      "loss": 0.0093,
      "step": 2590
    },
    {
      "epoch": 3.9061032863849765,
      "grad_norm": 0.1769220381975174,
      "learning_rate": 1.3801223956956994e-05,
      "loss": 0.0105,
      "step": 2600
    },
    {
      "epoch": 3.9061032863849765,
      "eval_loss": 0.0372486487030983,
      "eval_runtime": 221.9339,
      "eval_samples_per_second": 5.335,
      "eval_steps_per_second": 5.335,
      "step": 2600
    },
    {
      "epoch": 3.9211267605633804,
      "grad_norm": 0.11678332090377808,
      "learning_rate": 1.344106841525946e-05,
      "loss": 0.0141,
      "step": 2610
    },
    {
      "epoch": 3.936150234741784,
      "grad_norm": 0.24193385243415833,
      "learning_rate": 1.3084943439887659e-05,
      "loss": 0.0085,
      "step": 2620
    },
    {
      "epoch": 3.9511737089201877,
      "grad_norm": 0.10747163742780685,
      "learning_rate": 1.273288829307579e-05,
      "loss": 0.0054,
      "step": 2630
    },
    {
      "epoch": 3.9661971830985916,
      "grad_norm": 0.28762826323509216,
      "learning_rate": 1.2384941788365622e-05,
      "loss": 0.0078,
      "step": 2640
    },
    {
      "epoch": 3.981220657276995,
      "grad_norm": 0.11588622629642487,
      "learning_rate": 1.2041142286327477e-05,
      "loss": 0.0105,
      "step": 2650
    },
    {
      "epoch": 3.981220657276995,
      "eval_loss": 0.03705061599612236,
      "eval_runtime": 222.758,
      "eval_samples_per_second": 5.315,
      "eval_steps_per_second": 5.315,
      "step": 2650
    },
    {
      "epoch": 3.996244131455399,
      "grad_norm": 0.2062222808599472,
      "learning_rate": 1.170152769033095e-05,
      "loss": 0.0085,
      "step": 2660
    },
    {
      "epoch": 4.011267605633803,
      "grad_norm": 0.15399198234081268,
      "learning_rate": 1.1366135442366127e-05,
      "loss": 0.0057,
      "step": 2670
    },
    {
      "epoch": 4.026291079812206,
      "grad_norm": 0.12908077239990234,
      "learning_rate": 1.103500251891571e-05,
      "loss": 0.0074,
      "step": 2680
    },
    {
      "epoch": 4.041314553990611,
      "grad_norm": 0.06553179025650024,
      "learning_rate": 1.0708165426878325e-05,
      "loss": 0.0049,
      "step": 2690
    },
    {
      "epoch": 4.056338028169014,
      "grad_norm": 0.12032724171876907,
      "learning_rate": 1.0385660199543812e-05,
      "loss": 0.0064,
      "step": 2700
    },
    {
      "epoch": 4.056338028169014,
      "eval_loss": 0.038154859095811844,
      "eval_runtime": 220.5195,
      "eval_samples_per_second": 5.369,
      "eval_steps_per_second": 5.369,
      "step": 2700
    },
    {
      "epoch": 4.0713615023474174,
      "grad_norm": 0.06930918991565704,
      "learning_rate": 1.0067522392620537e-05,
      "loss": 0.0052,
      "step": 2710
    },
    {
      "epoch": 4.086384976525822,
      "grad_norm": 0.10133420675992966,
      "learning_rate": 9.753787080315385e-06,
      "loss": 0.0075,
      "step": 2720
    },
    {
      "epoch": 4.101408450704225,
      "grad_norm": 0.01755693554878235,
      "learning_rate": 9.444488851467042e-06,
      "loss": 0.0059,
      "step": 2730
    },
    {
      "epoch": 4.1164319248826295,
      "grad_norm": 0.1588931679725647,
      "learning_rate": 9.139661805732435e-06,
      "loss": 0.0042,
      "step": 2740
    },
    {
      "epoch": 4.131455399061033,
      "grad_norm": 0.39881432056427,
      "learning_rate": 8.839339549827397e-06,
      "loss": 0.0048,
      "step": 2750
    },
    {
      "epoch": 4.131455399061033,
      "eval_loss": 0.03982308879494667,
      "eval_runtime": 221.9797,
      "eval_samples_per_second": 5.334,
      "eval_steps_per_second": 5.334,
      "step": 2750
    },
    {
      "epoch": 4.146478873239436,
      "grad_norm": 0.07863107323646545,
      "learning_rate": 8.543555193821634e-06,
      "loss": 0.0075,
      "step": 2760
    },
    {
      "epoch": 4.161502347417841,
      "grad_norm": 0.25240081548690796,
      "learning_rate": 8.252341347488251e-06,
      "loss": 0.0043,
      "step": 2770
    },
    {
      "epoch": 4.176525821596244,
      "grad_norm": 0.22023944556713104,
      "learning_rate": 7.965730116708681e-06,
      "loss": 0.008,
      "step": 2780
    },
    {
      "epoch": 4.191549295774648,
      "grad_norm": 0.010785081423819065,
      "learning_rate": 7.68375309993304e-06,
      "loss": 0.0031,
      "step": 2790
    },
    {
      "epoch": 4.206572769953052,
      "grad_norm": 0.11607765406370163,
      "learning_rate": 7.406441384696372e-06,
      "loss": 0.0065,
      "step": 2800
    },
    {
      "epoch": 4.206572769953052,
      "eval_loss": 0.04066622629761696,
      "eval_runtime": 221.9129,
      "eval_samples_per_second": 5.335,
      "eval_steps_per_second": 5.335,
      "step": 2800
    },
    {
      "epoch": 4.221596244131455,
      "grad_norm": 0.05429501459002495,
      "learning_rate": 7.133825544191464e-06,
      "loss": 0.0095,
      "step": 2810
    },
    {
      "epoch": 4.236619718309859,
      "grad_norm": 0.09971319139003754,
      "learning_rate": 6.865935633897996e-06,
      "loss": 0.004,
      "step": 2820
    },
    {
      "epoch": 4.251643192488263,
      "grad_norm": 0.09597835689783096,
      "learning_rate": 6.602801188269081e-06,
      "loss": 0.0064,
      "step": 2830
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 0.1952216774225235,
      "learning_rate": 6.344451217475183e-06,
      "loss": 0.0047,
      "step": 2840
    },
    {
      "epoch": 4.28169014084507,
      "grad_norm": 0.0561475045979023,
      "learning_rate": 6.090914204205655e-06,
      "loss": 0.0031,
      "step": 2850
    },
    {
      "epoch": 4.28169014084507,
      "eval_loss": 0.04169252887368202,
      "eval_runtime": 222.1513,
      "eval_samples_per_second": 5.33,
      "eval_steps_per_second": 5.33,
      "step": 2850
    },
    {
      "epoch": 4.296713615023474,
      "grad_norm": 0.2749840021133423,
      "learning_rate": 5.842218100528679e-06,
      "loss": 0.0052,
      "step": 2860
    },
    {
      "epoch": 4.311737089201878,
      "grad_norm": 0.22904397547245026,
      "learning_rate": 5.598390324809555e-06,
      "loss": 0.0046,
      "step": 2870
    },
    {
      "epoch": 4.326760563380281,
      "grad_norm": 0.26261669397354126,
      "learning_rate": 5.359457758687841e-06,
      "loss": 0.0089,
      "step": 2880
    },
    {
      "epoch": 4.341784037558686,
      "grad_norm": 0.21348492801189423,
      "learning_rate": 5.125446744113743e-06,
      "loss": 0.0057,
      "step": 2890
    },
    {
      "epoch": 4.356807511737089,
      "grad_norm": 0.0984061136841774,
      "learning_rate": 4.896383080443934e-06,
      "loss": 0.0028,
      "step": 2900
    },
    {
      "epoch": 4.356807511737089,
      "eval_loss": 0.041960135102272034,
      "eval_runtime": 222.6261,
      "eval_samples_per_second": 5.318,
      "eval_steps_per_second": 5.318,
      "step": 2900
    },
    {
      "epoch": 4.371830985915493,
      "grad_norm": 0.2619991898536682,
      "learning_rate": 4.672292021597174e-06,
      "loss": 0.0056,
      "step": 2910
    },
    {
      "epoch": 4.386854460093897,
      "grad_norm": 0.16088174283504486,
      "learning_rate": 4.4531982732702145e-06,
      "loss": 0.0064,
      "step": 2920
    },
    {
      "epoch": 4.4018779342723,
      "grad_norm": 0.3044176995754242,
      "learning_rate": 4.239125990213883e-06,
      "loss": 0.0039,
      "step": 2930
    },
    {
      "epoch": 4.416901408450705,
      "grad_norm": 0.06394259631633759,
      "learning_rate": 4.030098773570174e-06,
      "loss": 0.0016,
      "step": 2940
    },
    {
      "epoch": 4.431924882629108,
      "grad_norm": 0.15494537353515625,
      "learning_rate": 3.826139668270234e-06,
      "loss": 0.0043,
      "step": 2950
    },
    {
      "epoch": 4.431924882629108,
      "eval_loss": 0.042107198387384415,
      "eval_runtime": 222.671,
      "eval_samples_per_second": 5.317,
      "eval_steps_per_second": 5.317,
      "step": 2950
    },
    {
      "epoch": 4.446948356807511,
      "grad_norm": 0.1741316169500351,
      "learning_rate": 3.6272711604936504e-06,
      "loss": 0.0049,
      "step": 2960
    },
    {
      "epoch": 4.461971830985916,
      "grad_norm": 0.08207106590270996,
      "learning_rate": 3.433515175189428e-06,
      "loss": 0.0053,
      "step": 2970
    },
    {
      "epoch": 4.476995305164319,
      "grad_norm": 0.09723722189664841,
      "learning_rate": 3.2448930736588e-06,
      "loss": 0.0077,
      "step": 2980
    },
    {
      "epoch": 4.492018779342723,
      "grad_norm": 0.03407386317849159,
      "learning_rate": 3.061425651200117e-06,
      "loss": 0.0038,
      "step": 2990
    },
    {
      "epoch": 4.507042253521127,
      "grad_norm": 0.21065545082092285,
      "learning_rate": 2.883133134816296e-06,
      "loss": 0.0048,
      "step": 3000
    },
    {
      "epoch": 4.507042253521127,
      "eval_loss": 0.04240112751722336,
      "eval_runtime": 222.0883,
      "eval_samples_per_second": 5.331,
      "eval_steps_per_second": 5.331,
      "step": 3000
    },
    {
      "epoch": 4.52206572769953,
      "grad_norm": 0.06989271938800812,
      "learning_rate": 2.7100351809847326e-06,
      "loss": 0.0046,
      "step": 3010
    },
    {
      "epoch": 4.537089201877935,
      "grad_norm": 0.07983122766017914,
      "learning_rate": 2.542150873490251e-06,
      "loss": 0.0028,
      "step": 3020
    },
    {
      "epoch": 4.552112676056338,
      "grad_norm": 0.27960672974586487,
      "learning_rate": 2.3794987213211383e-06,
      "loss": 0.0055,
      "step": 3030
    },
    {
      "epoch": 4.567136150234742,
      "grad_norm": 0.2628881633281708,
      "learning_rate": 2.222096656628547e-06,
      "loss": 0.0035,
      "step": 3040
    },
    {
      "epoch": 4.582159624413146,
      "grad_norm": 0.13972419500350952,
      "learning_rate": 2.0699620327495174e-06,
      "loss": 0.0038,
      "step": 3050
    },
    {
      "epoch": 4.582159624413146,
      "eval_loss": 0.04280637204647064,
      "eval_runtime": 222.2913,
      "eval_samples_per_second": 5.326,
      "eval_steps_per_second": 5.326,
      "step": 3050
    },
    {
      "epoch": 4.597183098591549,
      "grad_norm": 0.017372630536556244,
      "learning_rate": 1.9231116222937996e-06,
      "loss": 0.0025,
      "step": 3060
    },
    {
      "epoch": 4.612206572769953,
      "grad_norm": 0.14479027688503265,
      "learning_rate": 1.7815616152946523e-06,
      "loss": 0.0045,
      "step": 3070
    },
    {
      "epoch": 4.627230046948357,
      "grad_norm": 0.03770596161484718,
      "learning_rate": 1.6453276174240195e-06,
      "loss": 0.004,
      "step": 3080
    },
    {
      "epoch": 4.642253521126761,
      "grad_norm": 0.19395627081394196,
      "learning_rate": 1.5144246482719114e-06,
      "loss": 0.0043,
      "step": 3090
    },
    {
      "epoch": 4.657276995305164,
      "grad_norm": 0.05098709091544151,
      "learning_rate": 1.3888671396905805e-06,
      "loss": 0.0041,
      "step": 3100
    },
    {
      "epoch": 4.657276995305164,
      "eval_loss": 0.0430205836892128,
      "eval_runtime": 222.6399,
      "eval_samples_per_second": 5.318,
      "eval_steps_per_second": 5.318,
      "step": 3100
    },
    {
      "epoch": 4.672300469483568,
      "grad_norm": 0.3063965439796448,
      "learning_rate": 1.2686689342034431e-06,
      "loss": 0.0082,
      "step": 3110
    },
    {
      "epoch": 4.687323943661972,
      "grad_norm": 0.21486127376556396,
      "learning_rate": 1.1538432834789227e-06,
      "loss": 0.0054,
      "step": 3120
    },
    {
      "epoch": 4.702347417840375,
      "grad_norm": 0.3423399329185486,
      "learning_rate": 1.044402846869491e-06,
      "loss": 0.0049,
      "step": 3130
    },
    {
      "epoch": 4.71737089201878,
      "grad_norm": 0.15926554799079895,
      "learning_rate": 9.403596900160073e-07,
      "loss": 0.0034,
      "step": 3140
    },
    {
      "epoch": 4.732394366197183,
      "grad_norm": 0.24265912175178528,
      "learning_rate": 8.417252835174749e-07,
      "loss": 0.0066,
      "step": 3150
    },
    {
      "epoch": 4.732394366197183,
      "eval_loss": 0.043101657181978226,
      "eval_runtime": 223.1112,
      "eval_samples_per_second": 5.307,
      "eval_steps_per_second": 5.307,
      "step": 3150
    },
    {
      "epoch": 4.7474178403755865,
      "grad_norm": 0.13980786502361298,
      "learning_rate": 7.48510501666455e-07,
      "loss": 0.0073,
      "step": 3160
    },
    {
      "epoch": 4.762441314553991,
      "grad_norm": 0.14822153747081757,
      "learning_rate": 6.607256212501578e-07,
      "loss": 0.0067,
      "step": 3170
    },
    {
      "epoch": 4.777464788732394,
      "grad_norm": 0.2598550021648407,
      "learning_rate": 5.783803204174654e-07,
      "loss": 0.0054,
      "step": 3180
    },
    {
      "epoch": 4.792488262910798,
      "grad_norm": 0.145918071269989,
      "learning_rate": 5.014836776119358e-07,
      "loss": 0.0051,
      "step": 3190
    },
    {
      "epoch": 4.807511737089202,
      "grad_norm": 0.152046799659729,
      "learning_rate": 4.300441705708924e-07,
      "loss": 0.003,
      "step": 3200
    },
    {
      "epoch": 4.807511737089202,
      "eval_loss": 0.043016187846660614,
      "eval_runtime": 222.4587,
      "eval_samples_per_second": 5.322,
      "eval_steps_per_second": 5.322,
      "step": 3200
    },
    {
      "epoch": 4.822535211267605,
      "grad_norm": 0.1859809011220932,
      "learning_rate": 3.6406967539078796e-07,
      "loss": 0.0042,
      "step": 3210
    },
    {
      "epoch": 4.83755868544601,
      "grad_norm": 0.10558822751045227,
      "learning_rate": 3.0356746565887715e-07,
      "loss": 0.0056,
      "step": 3220
    },
    {
      "epoch": 4.852582159624413,
      "grad_norm": 0.3504423201084137,
      "learning_rate": 2.485442116513026e-07,
      "loss": 0.0048,
      "step": 3230
    },
    {
      "epoch": 4.867605633802817,
      "grad_norm": 0.21484680473804474,
      "learning_rate": 1.9900597959770507e-07,
      "loss": 0.0079,
      "step": 3240
    },
    {
      "epoch": 4.882629107981221,
      "grad_norm": 0.07226985692977905,
      "learning_rate": 1.5495823101245866e-07,
      "loss": 0.0031,
      "step": 3250
    },
    {
      "epoch": 4.882629107981221,
      "eval_loss": 0.042939841747283936,
      "eval_runtime": 222.1162,
      "eval_samples_per_second": 5.331,
      "eval_steps_per_second": 5.331,
      "step": 3250
    },
    {
      "epoch": 4.897652582159624,
      "grad_norm": 0.0429568886756897,
      "learning_rate": 1.164058220925135e-07,
      "loss": 0.0043,
      "step": 3260
    },
    {
      "epoch": 4.912676056338028,
      "grad_norm": 0.3218914568424225,
      "learning_rate": 8.335300318201844e-08,
      "loss": 0.0074,
      "step": 3270
    },
    {
      "epoch": 4.927699530516432,
      "grad_norm": 0.10779738426208496,
      "learning_rate": 5.5803418303745917e-08,
      "loss": 0.005,
      "step": 3280
    },
    {
      "epoch": 4.942723004694836,
      "grad_norm": 0.0708744004368782,
      "learning_rate": 3.3760104757313284e-08,
      "loss": 0.003,
      "step": 3290
    },
    {
      "epoch": 4.957746478873239,
      "grad_norm": 0.19295717775821686,
      "learning_rate": 1.7225492784345156e-08,
      "loss": 0.0046,
      "step": 3300
    },
    {
      "epoch": 4.957746478873239,
      "eval_loss": 0.042976681143045425,
      "eval_runtime": 222.2614,
      "eval_samples_per_second": 5.327,
      "eval_steps_per_second": 5.327,
      "step": 3300
    },
    {
      "epoch": 4.972769953051643,
      "grad_norm": 0.10675567388534546,
      "learning_rate": 6.201405300532148e-09,
      "loss": 0.0058,
      "step": 3310
    },
    {
      "epoch": 4.987793427230047,
      "grad_norm": 0.09640572220087051,
      "learning_rate": 6.890576946805282e-10,
      "loss": 0.0043,
      "step": 3320
    },
    {
      "epoch": 4.995305164319249,
      "step": 3325,
      "total_flos": 8.987524405535048e+17,
      "train_loss": 0.049058449685125426,
      "train_runtime": 45792.1032,
      "train_samples_per_second": 1.163,
      "train_steps_per_second": 0.073
    }
  ],
  "logging_steps": 10,
  "max_steps": 3325,
  "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": 8.987524405535048e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}