{
  "best_metric": 0.036843445152044296,
  "best_model_checkpoint": "saves/psy-course/Llama-3.1-8B-Instruct/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": 4.304145336151123,
      "learning_rate": 3.003003003003003e-06,
      "loss": 1.5464,
      "step": 10
    },
    {
      "epoch": 0.03004694835680751,
      "grad_norm": 5.03688907623291,
      "learning_rate": 6.006006006006006e-06,
      "loss": 1.597,
      "step": 20
    },
    {
      "epoch": 0.04507042253521127,
      "grad_norm": 5.86588191986084,
      "learning_rate": 9.00900900900901e-06,
      "loss": 1.4497,
      "step": 30
    },
    {
      "epoch": 0.06009389671361502,
      "grad_norm": 2.099303722381592,
      "learning_rate": 1.2012012012012012e-05,
      "loss": 0.9009,
      "step": 40
    },
    {
      "epoch": 0.07511737089201878,
      "grad_norm": 1.6121171712875366,
      "learning_rate": 1.5015015015015016e-05,
      "loss": 0.6045,
      "step": 50
    },
    {
      "epoch": 0.07511737089201878,
      "eval_loss": 0.41910141706466675,
      "eval_runtime": 156.3169,
      "eval_samples_per_second": 7.574,
      "eval_steps_per_second": 7.574,
      "step": 50
    },
    {
      "epoch": 0.09014084507042254,
      "grad_norm": 1.1837172508239746,
      "learning_rate": 1.801801801801802e-05,
      "loss": 0.402,
      "step": 60
    },
    {
      "epoch": 0.10516431924882629,
      "grad_norm": 0.9849412441253662,
      "learning_rate": 2.102102102102102e-05,
      "loss": 0.2368,
      "step": 70
    },
    {
      "epoch": 0.12018779342723004,
      "grad_norm": 0.8895949721336365,
      "learning_rate": 2.4024024024024024e-05,
      "loss": 0.1741,
      "step": 80
    },
    {
      "epoch": 0.1352112676056338,
      "grad_norm": 0.9555579423904419,
      "learning_rate": 2.702702702702703e-05,
      "loss": 0.1089,
      "step": 90
    },
    {
      "epoch": 0.15023474178403756,
      "grad_norm": 0.7877600193023682,
      "learning_rate": 3.0030030030030033e-05,
      "loss": 0.118,
      "step": 100
    },
    {
      "epoch": 0.15023474178403756,
      "eval_loss": 0.09762203693389893,
      "eval_runtime": 156.4651,
      "eval_samples_per_second": 7.567,
      "eval_steps_per_second": 7.567,
      "step": 100
    },
    {
      "epoch": 0.1652582159624413,
      "grad_norm": 1.2528424263000488,
      "learning_rate": 3.3033033033033035e-05,
      "loss": 0.1111,
      "step": 110
    },
    {
      "epoch": 0.18028169014084508,
      "grad_norm": 0.836103618144989,
      "learning_rate": 3.603603603603604e-05,
      "loss": 0.0988,
      "step": 120
    },
    {
      "epoch": 0.19530516431924883,
      "grad_norm": 1.0173139572143555,
      "learning_rate": 3.903903903903904e-05,
      "loss": 0.0874,
      "step": 130
    },
    {
      "epoch": 0.21032863849765257,
      "grad_norm": 0.4798429012298584,
      "learning_rate": 4.204204204204204e-05,
      "loss": 0.0904,
      "step": 140
    },
    {
      "epoch": 0.22535211267605634,
      "grad_norm": 0.6175504922866821,
      "learning_rate": 4.5045045045045046e-05,
      "loss": 0.075,
      "step": 150
    },
    {
      "epoch": 0.22535211267605634,
      "eval_loss": 0.09080640226602554,
      "eval_runtime": 156.3504,
      "eval_samples_per_second": 7.573,
      "eval_steps_per_second": 7.573,
      "step": 150
    },
    {
      "epoch": 0.2403755868544601,
      "grad_norm": 0.8490262031555176,
      "learning_rate": 4.804804804804805e-05,
      "loss": 0.0712,
      "step": 160
    },
    {
      "epoch": 0.25539906103286386,
      "grad_norm": 1.1651703119277954,
      "learning_rate": 5.105105105105106e-05,
      "loss": 0.0496,
      "step": 170
    },
    {
      "epoch": 0.2704225352112676,
      "grad_norm": 0.9137387275695801,
      "learning_rate": 5.405405405405406e-05,
      "loss": 0.0628,
      "step": 180
    },
    {
      "epoch": 0.28544600938967135,
      "grad_norm": 0.6661922931671143,
      "learning_rate": 5.705705705705706e-05,
      "loss": 0.0582,
      "step": 190
    },
    {
      "epoch": 0.3004694835680751,
      "grad_norm": 0.47193077206611633,
      "learning_rate": 6.0060060060060066e-05,
      "loss": 0.0475,
      "step": 200
    },
    {
      "epoch": 0.3004694835680751,
      "eval_loss": 0.05551699176430702,
      "eval_runtime": 156.3208,
      "eval_samples_per_second": 7.574,
      "eval_steps_per_second": 7.574,
      "step": 200
    },
    {
      "epoch": 0.3154929577464789,
      "grad_norm": 0.5054419636726379,
      "learning_rate": 6.306306306306306e-05,
      "loss": 0.0529,
      "step": 210
    },
    {
      "epoch": 0.3305164319248826,
      "grad_norm": 0.4073943793773651,
      "learning_rate": 6.606606606606607e-05,
      "loss": 0.0564,
      "step": 220
    },
    {
      "epoch": 0.3455399061032864,
      "grad_norm": 0.5814553499221802,
      "learning_rate": 6.906906906906907e-05,
      "loss": 0.0473,
      "step": 230
    },
    {
      "epoch": 0.36056338028169016,
      "grad_norm": 0.730743944644928,
      "learning_rate": 7.207207207207208e-05,
      "loss": 0.0651,
      "step": 240
    },
    {
      "epoch": 0.3755868544600939,
      "grad_norm": 0.6487244963645935,
      "learning_rate": 7.507507507507507e-05,
      "loss": 0.0612,
      "step": 250
    },
    {
      "epoch": 0.3755868544600939,
      "eval_loss": 0.05300073325634003,
      "eval_runtime": 156.331,
      "eval_samples_per_second": 7.574,
      "eval_steps_per_second": 7.574,
      "step": 250
    },
    {
      "epoch": 0.39061032863849765,
      "grad_norm": 0.3846875727176666,
      "learning_rate": 7.807807807807808e-05,
      "loss": 0.0529,
      "step": 260
    },
    {
      "epoch": 0.4056338028169014,
      "grad_norm": 0.8112801909446716,
      "learning_rate": 8.108108108108109e-05,
      "loss": 0.0609,
      "step": 270
    },
    {
      "epoch": 0.42065727699530514,
      "grad_norm": 0.22141359746456146,
      "learning_rate": 8.408408408408409e-05,
      "loss": 0.0577,
      "step": 280
    },
    {
      "epoch": 0.4356807511737089,
      "grad_norm": 1.3037300109863281,
      "learning_rate": 8.70870870870871e-05,
      "loss": 0.0572,
      "step": 290
    },
    {
      "epoch": 0.4507042253521127,
      "grad_norm": 1.0784958600997925,
      "learning_rate": 9.009009009009009e-05,
      "loss": 0.043,
      "step": 300
    },
    {
      "epoch": 0.4507042253521127,
      "eval_loss": 0.05256583169102669,
      "eval_runtime": 156.2659,
      "eval_samples_per_second": 7.577,
      "eval_steps_per_second": 7.577,
      "step": 300
    },
    {
      "epoch": 0.46572769953051646,
      "grad_norm": 0.4775850176811218,
      "learning_rate": 9.30930930930931e-05,
      "loss": 0.0439,
      "step": 310
    },
    {
      "epoch": 0.4807511737089202,
      "grad_norm": 0.4602857232093811,
      "learning_rate": 9.60960960960961e-05,
      "loss": 0.0597,
      "step": 320
    },
    {
      "epoch": 0.49577464788732395,
      "grad_norm": 0.6467078924179077,
      "learning_rate": 9.90990990990991e-05,
      "loss": 0.0553,
      "step": 330
    },
    {
      "epoch": 0.5107981220657277,
      "grad_norm": 0.5899861454963684,
      "learning_rate": 9.999864944989638e-05,
      "loss": 0.0614,
      "step": 340
    },
    {
      "epoch": 0.5258215962441315,
      "grad_norm": 0.3560892939567566,
      "learning_rate": 9.999203468625017e-05,
      "loss": 0.0497,
      "step": 350
    },
    {
      "epoch": 0.5258215962441315,
      "eval_loss": 0.05027875676751137,
      "eval_runtime": 156.4383,
      "eval_samples_per_second": 7.568,
      "eval_steps_per_second": 7.568,
      "step": 350
    },
    {
      "epoch": 0.5408450704225352,
      "grad_norm": 1.0397191047668457,
      "learning_rate": 9.997990837719421e-05,
      "loss": 0.0614,
      "step": 360
    },
    {
      "epoch": 0.5558685446009389,
      "grad_norm": 0.3409489095211029,
      "learning_rate": 9.996227185963554e-05,
      "loss": 0.0446,
      "step": 370
    },
    {
      "epoch": 0.5708920187793427,
      "grad_norm": 0.6183903217315674,
      "learning_rate": 9.993912707797329e-05,
      "loss": 0.0513,
      "step": 380
    },
    {
      "epoch": 0.5859154929577465,
      "grad_norm": 0.36344006657600403,
      "learning_rate": 9.99104765838842e-05,
      "loss": 0.0433,
      "step": 390
    },
    {
      "epoch": 0.6009389671361502,
      "grad_norm": 0.2808699309825897,
      "learning_rate": 9.987632353604151e-05,
      "loss": 0.049,
      "step": 400
    },
    {
      "epoch": 0.6009389671361502,
      "eval_loss": 0.04805321246385574,
      "eval_runtime": 156.401,
      "eval_samples_per_second": 7.57,
      "eval_steps_per_second": 7.57,
      "step": 400
    },
    {
      "epoch": 0.615962441314554,
      "grad_norm": 0.41244831681251526,
      "learning_rate": 9.98366716997665e-05,
      "loss": 0.0469,
      "step": 410
    },
    {
      "epoch": 0.6309859154929578,
      "grad_norm": 0.4227665960788727,
      "learning_rate": 9.979152544661354e-05,
      "loss": 0.0437,
      "step": 420
    },
    {
      "epoch": 0.6460093896713615,
      "grad_norm": 0.5011832118034363,
      "learning_rate": 9.974088975388802e-05,
      "loss": 0.0479,
      "step": 430
    },
    {
      "epoch": 0.6610328638497652,
      "grad_norm": 0.47126173973083496,
      "learning_rate": 9.968477020409766e-05,
      "loss": 0.0452,
      "step": 440
    },
    {
      "epoch": 0.676056338028169,
      "grad_norm": 0.4116508364677429,
      "learning_rate": 9.962317298433705e-05,
      "loss": 0.0541,
      "step": 450
    },
    {
      "epoch": 0.676056338028169,
      "eval_loss": 0.047716788947582245,
      "eval_runtime": 156.3995,
      "eval_samples_per_second": 7.57,
      "eval_steps_per_second": 7.57,
      "step": 450
    },
    {
      "epoch": 0.6910798122065728,
      "grad_norm": 0.2776816785335541,
      "learning_rate": 9.955610488560551e-05,
      "loss": 0.0589,
      "step": 460
    },
    {
      "epoch": 0.7061032863849765,
      "grad_norm": 0.5058294534683228,
      "learning_rate": 9.948357330205842e-05,
      "loss": 0.0583,
      "step": 470
    },
    {
      "epoch": 0.7211267605633803,
      "grad_norm": 0.22229774296283722,
      "learning_rate": 9.940558623019201e-05,
      "loss": 0.0329,
      "step": 480
    },
    {
      "epoch": 0.7361502347417841,
      "grad_norm": 0.33905524015426636,
      "learning_rate": 9.932215226796172e-05,
      "loss": 0.0501,
      "step": 490
    },
    {
      "epoch": 0.7511737089201878,
      "grad_norm": 0.2562945783138275,
      "learning_rate": 9.923328061383435e-05,
      "loss": 0.0494,
      "step": 500
    },
    {
      "epoch": 0.7511737089201878,
      "eval_loss": 0.043219637125730515,
      "eval_runtime": 156.322,
      "eval_samples_per_second": 7.574,
      "eval_steps_per_second": 7.574,
      "step": 500
    },
    {
      "epoch": 0.7661971830985915,
      "grad_norm": 0.18596145510673523,
      "learning_rate": 9.913898106577393e-05,
      "loss": 0.0445,
      "step": 510
    },
    {
      "epoch": 0.7812206572769953,
      "grad_norm": 0.30711135268211365,
      "learning_rate": 9.903926402016153e-05,
      "loss": 0.0389,
      "step": 520
    },
    {
      "epoch": 0.7962441314553991,
      "grad_norm": 0.21911293268203735,
      "learning_rate": 9.893414047064897e-05,
      "loss": 0.0436,
      "step": 530
    },
    {
      "epoch": 0.8112676056338028,
      "grad_norm": 0.49014413356781006,
      "learning_rate": 9.88236220069469e-05,
      "loss": 0.043,
      "step": 540
    },
    {
      "epoch": 0.8262910798122066,
      "grad_norm": 0.33901888132095337,
      "learning_rate": 9.870772081354705e-05,
      "loss": 0.0556,
      "step": 550
    },
    {
      "epoch": 0.8262910798122066,
      "eval_loss": 0.04144579544663429,
      "eval_runtime": 156.3463,
      "eval_samples_per_second": 7.573,
      "eval_steps_per_second": 7.573,
      "step": 550
    },
    {
      "epoch": 0.8413145539906103,
      "grad_norm": 0.3239082098007202,
      "learning_rate": 9.858644966837878e-05,
      "loss": 0.0463,
      "step": 560
    },
    {
      "epoch": 0.856338028169014,
      "grad_norm": 0.16826266050338745,
      "learning_rate": 9.845982194140051e-05,
      "loss": 0.0456,
      "step": 570
    },
    {
      "epoch": 0.8713615023474178,
      "grad_norm": 0.28091156482696533,
      "learning_rate": 9.832785159312559e-05,
      "loss": 0.042,
      "step": 580
    },
    {
      "epoch": 0.8863849765258216,
      "grad_norm": 0.29264500737190247,
      "learning_rate": 9.819055317308317e-05,
      "loss": 0.04,
      "step": 590
    },
    {
      "epoch": 0.9014084507042254,
      "grad_norm": 0.30719348788261414,
      "learning_rate": 9.804794181821422e-05,
      "loss": 0.0398,
      "step": 600
    },
    {
      "epoch": 0.9014084507042254,
      "eval_loss": 0.04154442995786667,
      "eval_runtime": 156.3701,
      "eval_samples_per_second": 7.572,
      "eval_steps_per_second": 7.572,
      "step": 600
    },
    {
      "epoch": 0.9164319248826291,
      "grad_norm": 0.18495255708694458,
      "learning_rate": 9.790003325120261e-05,
      "loss": 0.0403,
      "step": 610
    },
    {
      "epoch": 0.9314553990610329,
      "grad_norm": 0.514930009841919,
      "learning_rate": 9.774684377874178e-05,
      "loss": 0.0535,
      "step": 620
    },
    {
      "epoch": 0.9464788732394366,
      "grad_norm": 0.2771974503993988,
      "learning_rate": 9.758839028973692e-05,
      "loss": 0.0358,
      "step": 630
    },
    {
      "epoch": 0.9615023474178404,
      "grad_norm": 0.2064761519432068,
      "learning_rate": 9.742469025344298e-05,
      "loss": 0.043,
      "step": 640
    },
    {
      "epoch": 0.9765258215962441,
      "grad_norm": 0.22615325450897217,
      "learning_rate": 9.725576171753874e-05,
      "loss": 0.0535,
      "step": 650
    },
    {
      "epoch": 0.9765258215962441,
      "eval_loss": 0.04291123151779175,
      "eval_runtime": 156.3355,
      "eval_samples_per_second": 7.573,
      "eval_steps_per_second": 7.573,
      "step": 650
    },
    {
      "epoch": 0.9915492957746479,
      "grad_norm": 0.9398748874664307,
      "learning_rate": 9.708162330613708e-05,
      "loss": 0.0441,
      "step": 660
    },
    {
      "epoch": 1.0065727699530516,
      "grad_norm": 0.2263706773519516,
      "learning_rate": 9.690229421773167e-05,
      "loss": 0.0439,
      "step": 670
    },
    {
      "epoch": 1.0215962441314554,
      "grad_norm": 0.3939422070980072,
      "learning_rate": 9.67177942230804e-05,
      "loss": 0.0402,
      "step": 680
    },
    {
      "epoch": 1.036619718309859,
      "grad_norm": 0.4399869740009308,
      "learning_rate": 9.652814366302568e-05,
      "loss": 0.0412,
      "step": 690
    },
    {
      "epoch": 1.051643192488263,
      "grad_norm": 0.29108133912086487,
      "learning_rate": 9.633336344625185e-05,
      "loss": 0.0383,
      "step": 700
    },
    {
      "epoch": 1.051643192488263,
      "eval_loss": 0.04375332221388817,
      "eval_runtime": 156.304,
      "eval_samples_per_second": 7.575,
      "eval_steps_per_second": 7.575,
      "step": 700
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.41116732358932495,
      "learning_rate": 9.61334750469801e-05,
      "loss": 0.0297,
      "step": 710
    },
    {
      "epoch": 1.0816901408450703,
      "grad_norm": 0.49311184883117676,
      "learning_rate": 9.592850050260089e-05,
      "loss": 0.0377,
      "step": 720
    },
    {
      "epoch": 1.0967136150234742,
      "grad_norm": 0.20614486932754517,
      "learning_rate": 9.571846241124446e-05,
      "loss": 0.0329,
      "step": 730
    },
    {
      "epoch": 1.1117370892018779,
      "grad_norm": 0.1830943077802658,
      "learning_rate": 9.55033839292893e-05,
      "loss": 0.0273,
      "step": 740
    },
    {
      "epoch": 1.1267605633802817,
      "grad_norm": 0.49418705701828003,
      "learning_rate": 9.52832887688093e-05,
      "loss": 0.0317,
      "step": 750
    },
    {
      "epoch": 1.1267605633802817,
      "eval_loss": 0.04088291898369789,
      "eval_runtime": 156.2721,
      "eval_samples_per_second": 7.577,
      "eval_steps_per_second": 7.577,
      "step": 750
    },
    {
      "epoch": 1.1417840375586854,
      "grad_norm": 0.43101999163627625,
      "learning_rate": 9.50582011949595e-05,
      "loss": 0.035,
      "step": 760
    },
    {
      "epoch": 1.1568075117370893,
      "grad_norm": 0.19687387347221375,
      "learning_rate": 9.482814602330084e-05,
      "loss": 0.038,
      "step": 770
    },
    {
      "epoch": 1.171830985915493,
      "grad_norm": 0.31862860918045044,
      "learning_rate": 9.459314861706435e-05,
      "loss": 0.0402,
      "step": 780
    },
    {
      "epoch": 1.1868544600938966,
      "grad_norm": 0.18375325202941895,
      "learning_rate": 9.435323488435488e-05,
      "loss": 0.0313,
      "step": 790
    },
    {
      "epoch": 1.2018779342723005,
      "grad_norm": 0.4344165027141571,
      "learning_rate": 9.410843127529473e-05,
      "loss": 0.0311,
      "step": 800
    },
    {
      "epoch": 1.2018779342723005,
      "eval_loss": 0.04096408188343048,
      "eval_runtime": 156.2708,
      "eval_samples_per_second": 7.577,
      "eval_steps_per_second": 7.577,
      "step": 800
    },
    {
      "epoch": 1.2169014084507042,
      "grad_norm": 0.48818346858024597,
      "learning_rate": 9.385876477910765e-05,
      "loss": 0.0344,
      "step": 810
    },
    {
      "epoch": 1.231924882629108,
      "grad_norm": 0.2184937298297882,
      "learning_rate": 9.360426292114314e-05,
      "loss": 0.0343,
      "step": 820
    },
    {
      "epoch": 1.2469483568075117,
      "grad_norm": 0.3244076371192932,
      "learning_rate": 9.334495375984212e-05,
      "loss": 0.0294,
      "step": 830
    },
    {
      "epoch": 1.2619718309859156,
      "grad_norm": 0.3577475845813751,
      "learning_rate": 9.30808658836432e-05,
      "loss": 0.0337,
      "step": 840
    },
    {
      "epoch": 1.2769953051643192,
      "grad_norm": 0.5191630721092224,
      "learning_rate": 9.281202840783108e-05,
      "loss": 0.0351,
      "step": 850
    },
    {
      "epoch": 1.2769953051643192,
      "eval_loss": 0.03982950374484062,
      "eval_runtime": 156.3321,
      "eval_samples_per_second": 7.574,
      "eval_steps_per_second": 7.574,
      "step": 850
    },
    {
      "epoch": 1.292018779342723,
      "grad_norm": 0.21748529374599457,
      "learning_rate": 9.253847097132655e-05,
      "loss": 0.0313,
      "step": 860
    },
    {
      "epoch": 1.3070422535211268,
      "grad_norm": 0.18870031833648682,
      "learning_rate": 9.226022373341882e-05,
      "loss": 0.0323,
      "step": 870
    },
    {
      "epoch": 1.3220657276995305,
      "grad_norm": 0.24485838413238525,
      "learning_rate": 9.19773173704406e-05,
      "loss": 0.0299,
      "step": 880
    },
    {
      "epoch": 1.3370892018779343,
      "grad_norm": 0.6192132830619812,
      "learning_rate": 9.168978307238594e-05,
      "loss": 0.037,
      "step": 890
    },
    {
      "epoch": 1.352112676056338,
      "grad_norm": 0.1277685910463333,
      "learning_rate": 9.13976525394717e-05,
      "loss": 0.0309,
      "step": 900
    },
    {
      "epoch": 1.352112676056338,
      "eval_loss": 0.037233345210552216,
      "eval_runtime": 156.3404,
      "eval_samples_per_second": 7.573,
      "eval_steps_per_second": 7.573,
      "step": 900
    },
    {
      "epoch": 1.3671361502347419,
      "grad_norm": 0.17440356314182281,
      "learning_rate": 9.110095797864263e-05,
      "loss": 0.0264,
      "step": 910
    },
    {
      "epoch": 1.3821596244131455,
      "grad_norm": 0.15580697357654572,
      "learning_rate": 9.079973210002051e-05,
      "loss": 0.0371,
      "step": 920
    },
    {
      "epoch": 1.3971830985915492,
      "grad_norm": 0.2136790156364441,
      "learning_rate": 9.049400811329807e-05,
      "loss": 0.0389,
      "step": 930
    },
    {
      "epoch": 1.412206572769953,
      "grad_norm": 0.32828402519226074,
      "learning_rate": 9.01838197240775e-05,
      "loss": 0.0296,
      "step": 940
    },
    {
      "epoch": 1.4272300469483568,
      "grad_norm": 0.44386857748031616,
      "learning_rate": 8.986920113015461e-05,
      "loss": 0.0373,
      "step": 950
    },
    {
      "epoch": 1.4272300469483568,
      "eval_loss": 0.03834960237145424,
      "eval_runtime": 156.2587,
      "eval_samples_per_second": 7.577,
      "eval_steps_per_second": 7.577,
      "step": 950
    },
    {
      "epoch": 1.4422535211267606,
      "grad_norm": 0.20402352511882782,
      "learning_rate": 8.955018701774846e-05,
      "loss": 0.0244,
      "step": 960
    },
    {
      "epoch": 1.4572769953051643,
      "grad_norm": 0.3417670428752899,
      "learning_rate": 8.922681255767731e-05,
      "loss": 0.0413,
      "step": 970
    },
    {
      "epoch": 1.4723004694835682,
      "grad_norm": 0.21933668851852417,
      "learning_rate": 8.889911340148112e-05,
      "loss": 0.0359,
      "step": 980
    },
    {
      "epoch": 1.4873239436619718,
      "grad_norm": 0.1883007138967514,
      "learning_rate": 8.856712567749095e-05,
      "loss": 0.0336,
      "step": 990
    },
    {
      "epoch": 1.5023474178403755,
      "grad_norm": 0.31359952688217163,
      "learning_rate": 8.82308859868459e-05,
      "loss": 0.0295,
      "step": 1000
    },
    {
      "epoch": 1.5023474178403755,
      "eval_loss": 0.039744190871715546,
      "eval_runtime": 156.3721,
      "eval_samples_per_second": 7.572,
      "eval_steps_per_second": 7.572,
      "step": 1000
    },
    {
      "epoch": 1.5173708920187794,
      "grad_norm": 0.3432142734527588,
      "learning_rate": 8.789043139945795e-05,
      "loss": 0.0335,
      "step": 1010
    },
    {
      "epoch": 1.532394366197183,
      "grad_norm": 0.16293199360370636,
      "learning_rate": 8.754579944992491e-05,
      "loss": 0.0391,
      "step": 1020
    },
    {
      "epoch": 1.5474178403755867,
      "grad_norm": 0.20295502245426178,
      "learning_rate": 8.719702813339248e-05,
      "loss": 0.035,
      "step": 1030
    },
    {
      "epoch": 1.5624413145539906,
      "grad_norm": 0.22714973986148834,
      "learning_rate": 8.684415590136518e-05,
      "loss": 0.0322,
      "step": 1040
    },
    {
      "epoch": 1.5774647887323945,
      "grad_norm": 0.15775541961193085,
      "learning_rate": 8.648722165746722e-05,
      "loss": 0.0349,
      "step": 1050
    },
    {
      "epoch": 1.5774647887323945,
      "eval_loss": 0.04082873836159706,
      "eval_runtime": 156.2306,
      "eval_samples_per_second": 7.579,
      "eval_steps_per_second": 7.579,
      "step": 1050
    },
    {
      "epoch": 1.5924882629107981,
      "grad_norm": 0.18249914050102234,
      "learning_rate": 8.61262647531534e-05,
      "loss": 0.0299,
      "step": 1060
    },
    {
      "epoch": 1.6075117370892018,
      "grad_norm": 0.2968941330909729,
      "learning_rate": 8.576132498337068e-05,
      "loss": 0.0372,
      "step": 1070
    },
    {
      "epoch": 1.6225352112676057,
      "grad_norm": 0.416565477848053,
      "learning_rate": 8.539244258217088e-05,
      "loss": 0.0332,
      "step": 1080
    },
    {
      "epoch": 1.6375586854460094,
      "grad_norm": 0.23764647543430328,
      "learning_rate": 8.501965821827485e-05,
      "loss": 0.0365,
      "step": 1090
    },
    {
      "epoch": 1.652582159624413,
      "grad_norm": 0.20299109816551208,
      "learning_rate": 8.464301299058892e-05,
      "loss": 0.026,
      "step": 1100
    },
    {
      "epoch": 1.652582159624413,
      "eval_loss": 0.04096033051609993,
      "eval_runtime": 156.3005,
      "eval_samples_per_second": 7.575,
      "eval_steps_per_second": 7.575,
      "step": 1100
    },
    {
      "epoch": 1.667605633802817,
      "grad_norm": 0.2365420162677765,
      "learning_rate": 8.426254842367374e-05,
      "loss": 0.0279,
      "step": 1110
    },
    {
      "epoch": 1.6826291079812208,
      "grad_norm": 0.26829490065574646,
      "learning_rate": 8.387830646316623e-05,
      "loss": 0.0388,
      "step": 1120
    },
    {
      "epoch": 1.6976525821596244,
      "grad_norm": 0.39060696959495544,
      "learning_rate": 8.349032947115525e-05,
      "loss": 0.028,
      "step": 1130
    },
    {
      "epoch": 1.712676056338028,
      "grad_norm": 0.3235212564468384,
      "learning_rate": 8.309866022151107e-05,
      "loss": 0.0468,
      "step": 1140
    },
    {
      "epoch": 1.727699530516432,
      "grad_norm": 0.32556065917015076,
      "learning_rate": 8.270334189516983e-05,
      "loss": 0.0305,
      "step": 1150
    },
    {
      "epoch": 1.727699530516432,
      "eval_loss": 0.039029572159051895,
      "eval_runtime": 156.2235,
      "eval_samples_per_second": 7.579,
      "eval_steps_per_second": 7.579,
      "step": 1150
    },
    {
      "epoch": 1.7427230046948357,
      "grad_norm": 0.15808342397212982,
      "learning_rate": 8.230441807537277e-05,
      "loss": 0.0466,
      "step": 1160
    },
    {
      "epoch": 1.7577464788732393,
      "grad_norm": 0.15356343984603882,
      "learning_rate": 8.190193274286122e-05,
      "loss": 0.0284,
      "step": 1170
    },
    {
      "epoch": 1.7727699530516432,
      "grad_norm": 0.26693618297576904,
      "learning_rate": 8.149593027102789e-05,
      "loss": 0.0347,
      "step": 1180
    },
    {
      "epoch": 1.787793427230047,
      "grad_norm": 0.3144373595714569,
      "learning_rate": 8.108645542102469e-05,
      "loss": 0.0277,
      "step": 1190
    },
    {
      "epoch": 1.8028169014084507,
      "grad_norm": 0.26021772623062134,
      "learning_rate": 8.067355333682798e-05,
      "loss": 0.0332,
      "step": 1200
    },
    {
      "epoch": 1.8028169014084507,
      "eval_loss": 0.03854435682296753,
      "eval_runtime": 156.2279,
      "eval_samples_per_second": 7.579,
      "eval_steps_per_second": 7.579,
      "step": 1200
    },
    {
      "epoch": 1.8178403755868544,
      "grad_norm": 0.4863680303096771,
      "learning_rate": 8.025726954026138e-05,
      "loss": 0.0242,
      "step": 1210
    },
    {
      "epoch": 1.8328638497652583,
      "grad_norm": 0.5616618394851685,
      "learning_rate": 7.983764992597716e-05,
      "loss": 0.034,
      "step": 1220
    },
    {
      "epoch": 1.847887323943662,
      "grad_norm": 0.4086838662624359,
      "learning_rate": 7.94147407563964e-05,
      "loss": 0.0314,
      "step": 1230
    },
    {
      "epoch": 1.8629107981220656,
      "grad_norm": 0.2752532362937927,
      "learning_rate": 7.89885886566086e-05,
      "loss": 0.0348,
      "step": 1240
    },
    {
      "epoch": 1.8779342723004695,
      "grad_norm": 0.508071780204773,
      "learning_rate": 7.855924060923141e-05,
      "loss": 0.0304,
      "step": 1250
    },
    {
      "epoch": 1.8779342723004695,
      "eval_loss": 0.03792616352438927,
      "eval_runtime": 156.2438,
      "eval_samples_per_second": 7.578,
      "eval_steps_per_second": 7.578,
      "step": 1250
    },
    {
      "epoch": 1.8929577464788734,
      "grad_norm": 0.2151941955089569,
      "learning_rate": 7.812674394923077e-05,
      "loss": 0.0364,
      "step": 1260
    },
    {
      "epoch": 1.907981220657277,
      "grad_norm": 0.2420804798603058,
      "learning_rate": 7.769114635870231e-05,
      "loss": 0.0357,
      "step": 1270
    },
    {
      "epoch": 1.9230046948356807,
      "grad_norm": 0.37158623337745667,
      "learning_rate": 7.725249586161463e-05,
      "loss": 0.0317,
      "step": 1280
    },
    {
      "epoch": 1.9380281690140846,
      "grad_norm": 0.19411316514015198,
      "learning_rate": 7.68108408185145e-05,
      "loss": 0.0276,
      "step": 1290
    },
    {
      "epoch": 1.9530516431924883,
      "grad_norm": 0.2072768658399582,
      "learning_rate": 7.636622992119536e-05,
      "loss": 0.0497,
      "step": 1300
    },
    {
      "epoch": 1.9530516431924883,
      "eval_loss": 0.039196692407131195,
      "eval_runtime": 156.2151,
      "eval_samples_per_second": 7.579,
      "eval_steps_per_second": 7.579,
      "step": 1300
    },
    {
      "epoch": 1.968075117370892,
      "grad_norm": 0.30601534247398376,
      "learning_rate": 7.591871218732902e-05,
      "loss": 0.036,
      "step": 1310
    },
    {
      "epoch": 1.9830985915492958,
      "grad_norm": 0.2544531226158142,
      "learning_rate": 7.54683369550616e-05,
      "loss": 0.0413,
      "step": 1320
    },
    {
      "epoch": 1.9981220657276997,
      "grad_norm": 0.18643656373023987,
      "learning_rate": 7.501515387757404e-05,
      "loss": 0.0295,
      "step": 1330
    },
    {
      "epoch": 2.013145539906103,
      "grad_norm": 0.2908141016960144,
      "learning_rate": 7.455921291760796e-05,
      "loss": 0.0365,
      "step": 1340
    },
    {
      "epoch": 2.028169014084507,
      "grad_norm": 0.08652251958847046,
      "learning_rate": 7.410056434195725e-05,
      "loss": 0.0183,
      "step": 1350
    },
    {
      "epoch": 2.028169014084507,
      "eval_loss": 0.04124961793422699,
      "eval_runtime": 156.2014,
      "eval_samples_per_second": 7.58,
      "eval_steps_per_second": 7.58,
      "step": 1350
    },
    {
      "epoch": 2.043192488262911,
      "grad_norm": 0.231898233294487,
      "learning_rate": 7.363925871592629e-05,
      "loss": 0.0264,
      "step": 1360
    },
    {
      "epoch": 2.0582159624413148,
      "grad_norm": 0.15739409625530243,
      "learning_rate": 7.317534689775528e-05,
      "loss": 0.0239,
      "step": 1370
    },
    {
      "epoch": 2.073239436619718,
      "grad_norm": 0.16826321184635162,
      "learning_rate": 7.270888003301304e-05,
      "loss": 0.0175,
      "step": 1380
    },
    {
      "epoch": 2.088262910798122,
      "grad_norm": 0.5138995051383972,
      "learning_rate": 7.22399095489584e-05,
      "loss": 0.022,
      "step": 1390
    },
    {
      "epoch": 2.103286384976526,
      "grad_norm": 0.1792786568403244,
      "learning_rate": 7.176848714887042e-05,
      "loss": 0.0269,
      "step": 1400
    },
    {
      "epoch": 2.103286384976526,
      "eval_loss": 0.038102250546216965,
      "eval_runtime": 156.1612,
      "eval_samples_per_second": 7.582,
      "eval_steps_per_second": 7.582,
      "step": 1400
    },
    {
      "epoch": 2.1183098591549294,
      "grad_norm": 0.21858704090118408,
      "learning_rate": 7.129466480634806e-05,
      "loss": 0.0251,
      "step": 1410
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 0.2904298007488251,
      "learning_rate": 7.081849475958042e-05,
      "loss": 0.0208,
      "step": 1420
    },
    {
      "epoch": 2.148356807511737,
      "grad_norm": 0.4409976899623871,
      "learning_rate": 7.034002950558723e-05,
      "loss": 0.0243,
      "step": 1430
    },
    {
      "epoch": 2.1633802816901406,
      "grad_norm": 0.19713151454925537,
      "learning_rate": 6.985932179443144e-05,
      "loss": 0.0253,
      "step": 1440
    },
    {
      "epoch": 2.1784037558685445,
      "grad_norm": 0.24242573976516724,
      "learning_rate": 6.937642462340342e-05,
      "loss": 0.0299,
      "step": 1450
    },
    {
      "epoch": 2.1784037558685445,
      "eval_loss": 0.0379047654569149,
      "eval_runtime": 156.2912,
      "eval_samples_per_second": 7.576,
      "eval_steps_per_second": 7.576,
      "step": 1450
    },
    {
      "epoch": 2.1934272300469484,
      "grad_norm": 0.19243745505809784,
      "learning_rate": 6.889139123117817e-05,
      "loss": 0.0184,
      "step": 1460
    },
    {
      "epoch": 2.2084507042253523,
      "grad_norm": 0.2577214241027832,
      "learning_rate": 6.840427509194575e-05,
      "loss": 0.0249,
      "step": 1470
    },
    {
      "epoch": 2.2234741784037557,
      "grad_norm": 0.2908743917942047,
      "learning_rate": 6.791512990951597e-05,
      "loss": 0.0165,
      "step": 1480
    },
    {
      "epoch": 2.2384976525821596,
      "grad_norm": 0.6111800670623779,
      "learning_rate": 6.74240096113975e-05,
      "loss": 0.0262,
      "step": 1490
    },
    {
      "epoch": 2.2535211267605635,
      "grad_norm": 0.1613549441099167,
      "learning_rate": 6.693096834285256e-05,
      "loss": 0.0129,
      "step": 1500
    },
    {
      "epoch": 2.2535211267605635,
      "eval_loss": 0.04337477684020996,
      "eval_runtime": 156.2934,
      "eval_samples_per_second": 7.575,
      "eval_steps_per_second": 7.575,
      "step": 1500
    },
    {
      "epoch": 2.2685446009389674,
      "grad_norm": 0.40178021788597107,
      "learning_rate": 6.643606046092732e-05,
      "loss": 0.0238,
      "step": 1510
    },
    {
      "epoch": 2.283568075117371,
      "grad_norm": 0.18915718793869019,
      "learning_rate": 6.593934052845929e-05,
      "loss": 0.0257,
      "step": 1520
    },
    {
      "epoch": 2.2985915492957747,
      "grad_norm": 0.35349389910697937,
      "learning_rate": 6.544086330806181e-05,
      "loss": 0.0268,
      "step": 1530
    },
    {
      "epoch": 2.3136150234741786,
      "grad_norm": 0.15677450597286224,
      "learning_rate": 6.494068375608646e-05,
      "loss": 0.0245,
      "step": 1540
    },
    {
      "epoch": 2.328638497652582,
      "grad_norm": 0.2012639194726944,
      "learning_rate": 6.443885701656432e-05,
      "loss": 0.0203,
      "step": 1550
    },
    {
      "epoch": 2.328638497652582,
      "eval_loss": 0.03834431990981102,
      "eval_runtime": 156.2857,
      "eval_samples_per_second": 7.576,
      "eval_steps_per_second": 7.576,
      "step": 1550
    },
    {
      "epoch": 2.343661971830986,
      "grad_norm": 0.28499725461006165,
      "learning_rate": 6.393543841512632e-05,
      "loss": 0.0187,
      "step": 1560
    },
    {
      "epoch": 2.35868544600939,
      "grad_norm": 0.27369531989097595,
      "learning_rate": 6.343048345290386e-05,
      "loss": 0.0143,
      "step": 1570
    },
    {
      "epoch": 2.3737089201877932,
      "grad_norm": 0.31129664182662964,
      "learning_rate": 6.292404780040961e-05,
      "loss": 0.0306,
      "step": 1580
    },
    {
      "epoch": 2.388732394366197,
      "grad_norm": 0.43185606598854065,
      "learning_rate": 6.241618729140018e-05,
      "loss": 0.0198,
      "step": 1590
    },
    {
      "epoch": 2.403755868544601,
      "grad_norm": 0.32471033930778503,
      "learning_rate": 6.190695791672042e-05,
      "loss": 0.0221,
      "step": 1600
    },
    {
      "epoch": 2.403755868544601,
      "eval_loss": 0.03971394523978233,
      "eval_runtime": 156.3498,
      "eval_samples_per_second": 7.573,
      "eval_steps_per_second": 7.573,
      "step": 1600
    },
    {
      "epoch": 2.418779342723005,
      "grad_norm": 0.13160324096679688,
      "learning_rate": 6.139641581813052e-05,
      "loss": 0.0194,
      "step": 1610
    },
    {
      "epoch": 2.4338028169014083,
      "grad_norm": 0.1444709450006485,
      "learning_rate": 6.088461728211642e-05,
      "loss": 0.0206,
      "step": 1620
    },
    {
      "epoch": 2.448826291079812,
      "grad_norm": 0.3128149211406708,
      "learning_rate": 6.0371618733684474e-05,
      "loss": 0.0151,
      "step": 1630
    },
    {
      "epoch": 2.463849765258216,
      "grad_norm": 0.1301877200603485,
      "learning_rate": 5.9857476730140485e-05,
      "loss": 0.0214,
      "step": 1640
    },
    {
      "epoch": 2.4788732394366195,
      "grad_norm": 0.2600902020931244,
      "learning_rate": 5.9342247954854466e-05,
      "loss": 0.023,
      "step": 1650
    },
    {
      "epoch": 2.4788732394366195,
      "eval_loss": 0.041283123195171356,
      "eval_runtime": 156.3863,
      "eval_samples_per_second": 7.571,
      "eval_steps_per_second": 7.571,
      "step": 1650
    },
    {
      "epoch": 2.4938967136150234,
      "grad_norm": 0.20478318631649017,
      "learning_rate": 5.8825989211011335e-05,
      "loss": 0.0242,
      "step": 1660
    },
    {
      "epoch": 2.5089201877934273,
      "grad_norm": 0.212309792637825,
      "learning_rate": 5.830875741534852e-05,
      "loss": 0.0187,
      "step": 1670
    },
    {
      "epoch": 2.523943661971831,
      "grad_norm": 0.2569638192653656,
      "learning_rate": 5.7790609591880826e-05,
      "loss": 0.0259,
      "step": 1680
    },
    {
      "epoch": 2.5389671361502346,
      "grad_norm": 0.305275022983551,
      "learning_rate": 5.727160286561386e-05,
      "loss": 0.0232,
      "step": 1690
    },
    {
      "epoch": 2.5539906103286385,
      "grad_norm": 0.22980214655399323,
      "learning_rate": 5.675179445624581e-05,
      "loss": 0.0257,
      "step": 1700
    },
    {
      "epoch": 2.5539906103286385,
      "eval_loss": 0.04087158292531967,
      "eval_runtime": 156.4861,
      "eval_samples_per_second": 7.566,
      "eval_steps_per_second": 7.566,
      "step": 1700
    },
    {
      "epoch": 2.5690140845070424,
      "grad_norm": 0.13337288796901703,
      "learning_rate": 5.62312416718593e-05,
      "loss": 0.018,
      "step": 1710
    },
    {
      "epoch": 2.584037558685446,
      "grad_norm": 0.06781081855297089,
      "learning_rate": 5.5710001902603116e-05,
      "loss": 0.0238,
      "step": 1720
    },
    {
      "epoch": 2.5990610328638497,
      "grad_norm": 0.15210840106010437,
      "learning_rate": 5.5188132614365094e-05,
      "loss": 0.0213,
      "step": 1730
    },
    {
      "epoch": 2.6140845070422536,
      "grad_norm": 0.18319003283977509,
      "learning_rate": 5.4665691342436565e-05,
      "loss": 0.0214,
      "step": 1740
    },
    {
      "epoch": 2.629107981220657,
      "grad_norm": 0.382030189037323,
      "learning_rate": 5.414273568516919e-05,
      "loss": 0.0241,
      "step": 1750
    },
    {
      "epoch": 2.629107981220657,
      "eval_loss": 0.03998855501413345,
      "eval_runtime": 156.5463,
      "eval_samples_per_second": 7.563,
      "eval_steps_per_second": 7.563,
      "step": 1750
    },
    {
      "epoch": 2.644131455399061,
      "grad_norm": 0.26402246952056885,
      "learning_rate": 5.361932329762481e-05,
      "loss": 0.0201,
      "step": 1760
    },
    {
      "epoch": 2.659154929577465,
      "grad_norm": 0.1534145027399063,
      "learning_rate": 5.309551188521914e-05,
      "loss": 0.0228,
      "step": 1770
    },
    {
      "epoch": 2.6741784037558687,
      "grad_norm": 0.4438458979129791,
      "learning_rate": 5.2571359197359704e-05,
      "loss": 0.0162,
      "step": 1780
    },
    {
      "epoch": 2.6892018779342726,
      "grad_norm": 0.21595679223537445,
      "learning_rate": 5.2046923021079175e-05,
      "loss": 0.0173,
      "step": 1790
    },
    {
      "epoch": 2.704225352112676,
      "grad_norm": 0.4059389531612396,
      "learning_rate": 5.1522261174664346e-05,
      "loss": 0.0258,
      "step": 1800
    },
    {
      "epoch": 2.704225352112676,
      "eval_loss": 0.040539611130952835,
      "eval_runtime": 156.6429,
      "eval_samples_per_second": 7.559,
      "eval_steps_per_second": 7.559,
      "step": 1800
    },
    {
      "epoch": 2.71924882629108,
      "grad_norm": 0.8589577078819275,
      "learning_rate": 5.0997431501281835e-05,
      "loss": 0.035,
      "step": 1810
    },
    {
      "epoch": 2.7342723004694838,
      "grad_norm": 0.2879532277584076,
      "learning_rate": 5.0472491862600915e-05,
      "loss": 0.0223,
      "step": 1820
    },
    {
      "epoch": 2.749295774647887,
      "grad_norm": 0.3457850217819214,
      "learning_rate": 4.994750013241435e-05,
      "loss": 0.0264,
      "step": 1830
    },
    {
      "epoch": 2.764319248826291,
      "grad_norm": 0.29999199509620667,
      "learning_rate": 4.9422514190257974e-05,
      "loss": 0.0242,
      "step": 1840
    },
    {
      "epoch": 2.779342723004695,
      "grad_norm": 0.09458961337804794,
      "learning_rate": 4.88975919150294e-05,
      "loss": 0.021,
      "step": 1850
    },
    {
      "epoch": 2.779342723004695,
      "eval_loss": 0.03918106481432915,
      "eval_runtime": 156.9396,
      "eval_samples_per_second": 7.544,
      "eval_steps_per_second": 7.544,
      "step": 1850
    },
    {
      "epoch": 2.7943661971830984,
      "grad_norm": 0.2146388590335846,
      "learning_rate": 4.83727911786071e-05,
      "loss": 0.0257,
      "step": 1860
    },
    {
      "epoch": 2.8093896713615023,
      "grad_norm": 0.3507349193096161,
      "learning_rate": 4.7848169839470145e-05,
      "loss": 0.0237,
      "step": 1870
    },
    {
      "epoch": 2.824413145539906,
      "grad_norm": 0.14694459736347198,
      "learning_rate": 4.7323785736319244e-05,
      "loss": 0.0271,
      "step": 1880
    },
    {
      "epoch": 2.8394366197183096,
      "grad_norm": 0.16568858921527863,
      "learning_rate": 4.679969668170024e-05,
      "loss": 0.0234,
      "step": 1890
    },
    {
      "epoch": 2.8544600938967135,
      "grad_norm": 0.1532352864742279,
      "learning_rate": 4.627596045563031e-05,
      "loss": 0.0253,
      "step": 1900
    },
    {
      "epoch": 2.8544600938967135,
      "eval_loss": 0.036843445152044296,
      "eval_runtime": 157.0438,
      "eval_samples_per_second": 7.539,
      "eval_steps_per_second": 7.539,
      "step": 1900
    },
    {
      "epoch": 2.8694835680751174,
      "grad_norm": 0.22552986443042755,
      "learning_rate": 4.575263479922783e-05,
      "loss": 0.0156,
      "step": 1910
    },
    {
      "epoch": 2.8845070422535213,
      "grad_norm": 0.1421150416135788,
      "learning_rate": 4.522977740834651e-05,
      "loss": 0.0142,
      "step": 1920
    },
    {
      "epoch": 2.8995305164319247,
      "grad_norm": 0.2992912232875824,
      "learning_rate": 4.4707445927214456e-05,
      "loss": 0.0169,
      "step": 1930
    },
    {
      "epoch": 2.9145539906103286,
      "grad_norm": 0.1591469943523407,
      "learning_rate": 4.4185697942079115e-05,
      "loss": 0.0255,
      "step": 1940
    },
    {
      "epoch": 2.9295774647887325,
      "grad_norm": 0.13977603614330292,
      "learning_rate": 4.366459097485832e-05,
      "loss": 0.018,
      "step": 1950
    },
    {
      "epoch": 2.9295774647887325,
      "eval_loss": 0.03921648487448692,
      "eval_runtime": 157.0461,
      "eval_samples_per_second": 7.539,
      "eval_steps_per_second": 7.539,
      "step": 1950
    },
    {
      "epoch": 2.9446009389671364,
      "grad_norm": 0.3342364728450775,
      "learning_rate": 4.314418247679866e-05,
      "loss": 0.0263,
      "step": 1960
    },
    {
      "epoch": 2.95962441314554,
      "grad_norm": 0.24284552037715912,
      "learning_rate": 4.26245298221416e-05,
      "loss": 0.014,
      "step": 1970
    },
    {
      "epoch": 2.9746478873239437,
      "grad_norm": 0.15542685985565186,
      "learning_rate": 4.2105690301798014e-05,
      "loss": 0.0176,
      "step": 1980
    },
    {
      "epoch": 2.9896713615023476,
      "grad_norm": 0.24745222926139832,
      "learning_rate": 4.158772111703194e-05,
      "loss": 0.0177,
      "step": 1990
    },
    {
      "epoch": 3.004694835680751,
      "grad_norm": 0.07768896967172623,
      "learning_rate": 4.107067937315429e-05,
      "loss": 0.0112,
      "step": 2000
    },
    {
      "epoch": 3.004694835680751,
      "eval_loss": 0.04066740348935127,
      "eval_runtime": 157.2285,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 2000
    },
    {
      "epoch": 3.019718309859155,
      "grad_norm": 0.1813950091600418,
      "learning_rate": 4.055462207322698e-05,
      "loss": 0.0121,
      "step": 2010
    },
    {
      "epoch": 3.034741784037559,
      "grad_norm": 0.22787263989448547,
      "learning_rate": 4.003960611177855e-05,
      "loss": 0.0118,
      "step": 2020
    },
    {
      "epoch": 3.0497652582159622,
      "grad_norm": 0.19959565997123718,
      "learning_rate": 3.952568826853152e-05,
      "loss": 0.0097,
      "step": 2030
    },
    {
      "epoch": 3.064788732394366,
      "grad_norm": 0.0704595223069191,
      "learning_rate": 3.901292520214256e-05,
      "loss": 0.0079,
      "step": 2040
    },
    {
      "epoch": 3.07981220657277,
      "grad_norm": 0.4440910816192627,
      "learning_rate": 3.850137344395598e-05,
      "loss": 0.0082,
      "step": 2050
    },
    {
      "epoch": 3.07981220657277,
      "eval_loss": 0.047751668840646744,
      "eval_runtime": 157.4059,
      "eval_samples_per_second": 7.522,
      "eval_steps_per_second": 7.522,
      "step": 2050
    },
    {
      "epoch": 3.094835680751174,
      "grad_norm": 0.4817383885383606,
      "learning_rate": 3.799108939177118e-05,
      "loss": 0.011,
      "step": 2060
    },
    {
      "epoch": 3.1098591549295773,
      "grad_norm": 0.15736213326454163,
      "learning_rate": 3.7482129303624934e-05,
      "loss": 0.0098,
      "step": 2070
    },
    {
      "epoch": 3.124882629107981,
      "grad_norm": 0.3768656551837921,
      "learning_rate": 3.697454929158901e-05,
      "loss": 0.0123,
      "step": 2080
    },
    {
      "epoch": 3.139906103286385,
      "grad_norm": 0.37794822454452515,
      "learning_rate": 3.6468405315583854e-05,
      "loss": 0.0155,
      "step": 2090
    },
    {
      "epoch": 3.1549295774647885,
      "grad_norm": 0.47674834728240967,
      "learning_rate": 3.59637531772092e-05,
      "loss": 0.0127,
      "step": 2100
    },
    {
      "epoch": 3.1549295774647885,
      "eval_loss": 0.046006690710783005,
      "eval_runtime": 157.5212,
      "eval_samples_per_second": 7.516,
      "eval_steps_per_second": 7.516,
      "step": 2100
    },
    {
      "epoch": 3.1699530516431924,
      "grad_norm": 0.2169833481311798,
      "learning_rate": 3.546064851359192e-05,
      "loss": 0.0048,
      "step": 2110
    },
    {
      "epoch": 3.1849765258215963,
      "grad_norm": 0.12608371675014496,
      "learning_rate": 3.495914679125212e-05,
      "loss": 0.0125,
      "step": 2120
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.21189413964748383,
      "learning_rate": 3.445930329998819e-05,
      "loss": 0.0102,
      "step": 2130
    },
    {
      "epoch": 3.2150234741784036,
      "grad_norm": 0.29917919635772705,
      "learning_rate": 3.396117314678097e-05,
      "loss": 0.0173,
      "step": 2140
    },
    {
      "epoch": 3.2300469483568075,
      "grad_norm": 0.29451966285705566,
      "learning_rate": 3.3464811249718474e-05,
      "loss": 0.0112,
      "step": 2150
    },
    {
      "epoch": 3.2300469483568075,
      "eval_loss": 0.04488084837794304,
      "eval_runtime": 157.4858,
      "eval_samples_per_second": 7.518,
      "eval_steps_per_second": 7.518,
      "step": 2150
    },
    {
      "epoch": 3.2450704225352114,
      "grad_norm": 0.1548345535993576,
      "learning_rate": 3.297027233194114e-05,
      "loss": 0.0128,
      "step": 2160
    },
    {
      "epoch": 3.260093896713615,
      "grad_norm": 0.08690305054187775,
      "learning_rate": 3.2477610915608704e-05,
      "loss": 0.01,
      "step": 2170
    },
    {
      "epoch": 3.2751173708920187,
      "grad_norm": 0.26383867859840393,
      "learning_rate": 3.1986881315889315e-05,
      "loss": 0.014,
      "step": 2180
    },
    {
      "epoch": 3.2901408450704226,
      "grad_norm": 0.4667436480522156,
      "learning_rate": 3.149813763497124e-05,
      "loss": 0.0174,
      "step": 2190
    },
    {
      "epoch": 3.3051643192488265,
      "grad_norm": 0.3968455493450165,
      "learning_rate": 3.101143375609818e-05,
      "loss": 0.0145,
      "step": 2200
    },
    {
      "epoch": 3.3051643192488265,
      "eval_loss": 0.04522354528307915,
      "eval_runtime": 157.475,
      "eval_samples_per_second": 7.519,
      "eval_steps_per_second": 7.519,
      "step": 2200
    },
    {
      "epoch": 3.32018779342723,
      "grad_norm": 0.2882903814315796,
      "learning_rate": 3.0526823337628915e-05,
      "loss": 0.0143,
      "step": 2210
    },
    {
      "epoch": 3.335211267605634,
      "grad_norm": 0.2159668207168579,
      "learning_rate": 3.004435980712129e-05,
      "loss": 0.0094,
      "step": 2220
    },
    {
      "epoch": 3.3502347417840377,
      "grad_norm": 0.6046177744865417,
      "learning_rate": 2.9564096355442116e-05,
      "loss": 0.0133,
      "step": 2230
    },
    {
      "epoch": 3.365258215962441,
      "grad_norm": 0.30788397789001465,
      "learning_rate": 2.9086085930902824e-05,
      "loss": 0.0143,
      "step": 2240
    },
    {
      "epoch": 3.380281690140845,
      "grad_norm": 0.3591701090335846,
      "learning_rate": 2.8610381233422058e-05,
      "loss": 0.0086,
      "step": 2250
    },
    {
      "epoch": 3.380281690140845,
      "eval_loss": 0.04514958709478378,
      "eval_runtime": 157.5469,
      "eval_samples_per_second": 7.515,
      "eval_steps_per_second": 7.515,
      "step": 2250
    },
    {
      "epoch": 3.395305164319249,
      "grad_norm": 0.18443742394447327,
      "learning_rate": 2.8137034708715592e-05,
      "loss": 0.0106,
      "step": 2260
    },
    {
      "epoch": 3.4103286384976528,
      "grad_norm": 0.02844012901186943,
      "learning_rate": 2.7666098542514273e-05,
      "loss": 0.0066,
      "step": 2270
    },
    {
      "epoch": 3.425352112676056,
      "grad_norm": 0.5778037309646606,
      "learning_rate": 2.719762465481055e-05,
      "loss": 0.0117,
      "step": 2280
    },
    {
      "epoch": 3.44037558685446,
      "grad_norm": 0.17093347012996674,
      "learning_rate": 2.6731664694134473e-05,
      "loss": 0.0133,
      "step": 2290
    },
    {
      "epoch": 3.455399061032864,
      "grad_norm": 0.5244073271751404,
      "learning_rate": 2.6268270031859476e-05,
      "loss": 0.0161,
      "step": 2300
    },
    {
      "epoch": 3.455399061032864,
      "eval_loss": 0.04896237701177597,
      "eval_runtime": 157.594,
      "eval_samples_per_second": 7.513,
      "eval_steps_per_second": 7.513,
      "step": 2300
    },
    {
      "epoch": 3.4704225352112674,
      "grad_norm": 0.5648406147956848,
      "learning_rate": 2.580749175653877e-05,
      "loss": 0.0108,
      "step": 2310
    },
    {
      "epoch": 3.4854460093896713,
      "grad_norm": 0.5017344355583191,
      "learning_rate": 2.5349380668272905e-05,
      "loss": 0.0124,
      "step": 2320
    },
    {
      "epoch": 3.500469483568075,
      "grad_norm": 0.1720559000968933,
      "learning_rate": 2.489398727310908e-05,
      "loss": 0.013,
      "step": 2330
    },
    {
      "epoch": 3.5154929577464786,
      "grad_norm": 0.34457623958587646,
      "learning_rate": 2.4441361777473066e-05,
      "loss": 0.0107,
      "step": 2340
    },
    {
      "epoch": 3.5305164319248825,
      "grad_norm": 0.8733149766921997,
      "learning_rate": 2.3991554082633912e-05,
      "loss": 0.0134,
      "step": 2350
    },
    {
      "epoch": 3.5305164319248825,
      "eval_loss": 0.04652281105518341,
      "eval_runtime": 157.7619,
      "eval_samples_per_second": 7.505,
      "eval_steps_per_second": 7.505,
      "step": 2350
    },
    {
      "epoch": 3.5455399061032864,
      "grad_norm": 0.13055922091007233,
      "learning_rate": 2.354461377920239e-05,
      "loss": 0.0088,
      "step": 2360
    },
    {
      "epoch": 3.5605633802816903,
      "grad_norm": 0.20548763871192932,
      "learning_rate": 2.3100590141663807e-05,
      "loss": 0.0115,
      "step": 2370
    },
    {
      "epoch": 3.575586854460094,
      "grad_norm": 0.3360619843006134,
      "learning_rate": 2.265953212294551e-05,
      "loss": 0.0182,
      "step": 2380
    },
    {
      "epoch": 3.5906103286384976,
      "grad_norm": 0.6770099401473999,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.0164,
      "step": 2390
    },
    {
      "epoch": 3.6056338028169015,
      "grad_norm": 0.31347426772117615,
      "learning_rate": 2.1786507113543457e-05,
      "loss": 0.009,
      "step": 2400
    },
    {
      "epoch": 3.6056338028169015,
      "eval_loss": 0.04605977609753609,
      "eval_runtime": 157.7592,
      "eval_samples_per_second": 7.505,
      "eval_steps_per_second": 7.505,
      "step": 2400
    },
    {
      "epoch": 3.6206572769953054,
      "grad_norm": 0.13689975440502167,
      "learning_rate": 2.1354636372532523e-05,
      "loss": 0.0176,
      "step": 2410
    },
    {
      "epoch": 3.635680751173709,
      "grad_norm": 0.2913866937160492,
      "learning_rate": 2.092592373907617e-05,
      "loss": 0.0101,
      "step": 2420
    },
    {
      "epoch": 3.6507042253521127,
      "grad_norm": 0.37444183230400085,
      "learning_rate": 2.0500416478086932e-05,
      "loss": 0.0075,
      "step": 2430
    },
    {
      "epoch": 3.6657276995305166,
      "grad_norm": 0.42321744561195374,
      "learning_rate": 2.0078161501089954e-05,
      "loss": 0.0117,
      "step": 2440
    },
    {
      "epoch": 3.68075117370892,
      "grad_norm": 0.15104345977306366,
      "learning_rate": 1.9659205361050982e-05,
      "loss": 0.0056,
      "step": 2450
    },
    {
      "epoch": 3.68075117370892,
      "eval_loss": 0.04802750423550606,
      "eval_runtime": 157.7669,
      "eval_samples_per_second": 7.505,
      "eval_steps_per_second": 7.505,
      "step": 2450
    },
    {
      "epoch": 3.695774647887324,
      "grad_norm": 0.15667098760604858,
      "learning_rate": 1.924359424724408e-05,
      "loss": 0.0154,
      "step": 2460
    },
    {
      "epoch": 3.710798122065728,
      "grad_norm": 0.49284836649894714,
      "learning_rate": 1.8831373980159296e-05,
      "loss": 0.0145,
      "step": 2470
    },
    {
      "epoch": 3.7258215962441312,
      "grad_norm": 0.22378303110599518,
      "learning_rate": 1.8422590006450947e-05,
      "loss": 0.0063,
      "step": 2480
    },
    {
      "epoch": 3.740845070422535,
      "grad_norm": 0.20172594487667084,
      "learning_rate": 1.801728739392731e-05,
      "loss": 0.0089,
      "step": 2490
    },
    {
      "epoch": 3.755868544600939,
      "grad_norm": 0.2898995280265808,
      "learning_rate": 1.7615510826581904e-05,
      "loss": 0.0069,
      "step": 2500
    },
    {
      "epoch": 3.755868544600939,
      "eval_loss": 0.04896940663456917,
      "eval_runtime": 157.7817,
      "eval_samples_per_second": 7.504,
      "eval_steps_per_second": 7.504,
      "step": 2500
    },
    {
      "epoch": 3.770892018779343,
      "grad_norm": 0.32213857769966125,
      "learning_rate": 1.7217304599667146e-05,
      "loss": 0.0164,
      "step": 2510
    },
    {
      "epoch": 3.7859154929577463,
      "grad_norm": 0.06365455687046051,
      "learning_rate": 1.6822712614810893e-05,
      "loss": 0.01,
      "step": 2520
    },
    {
      "epoch": 3.80093896713615,
      "grad_norm": 0.2050347775220871,
      "learning_rate": 1.643177837517631e-05,
      "loss": 0.0117,
      "step": 2530
    },
    {
      "epoch": 3.815962441314554,
      "grad_norm": 0.40109461545944214,
      "learning_rate": 1.6044544980665767e-05,
      "loss": 0.0108,
      "step": 2540
    },
    {
      "epoch": 3.830985915492958,
      "grad_norm": 0.5502837300300598,
      "learning_rate": 1.5661055123169126e-05,
      "loss": 0.007,
      "step": 2550
    },
    {
      "epoch": 3.830985915492958,
      "eval_loss": 0.04891159385442734,
      "eval_runtime": 157.7489,
      "eval_samples_per_second": 7.506,
      "eval_steps_per_second": 7.506,
      "step": 2550
    },
    {
      "epoch": 3.8460093896713614,
      "grad_norm": 0.17208245396614075,
      "learning_rate": 1.5281351081856974e-05,
      "loss": 0.0079,
      "step": 2560
    },
    {
      "epoch": 3.8610328638497653,
      "grad_norm": 0.2847214639186859,
      "learning_rate": 1.4905474718519491e-05,
      "loss": 0.0075,
      "step": 2570
    },
    {
      "epoch": 3.876056338028169,
      "grad_norm": 0.36303362250328064,
      "learning_rate": 1.453346747295119e-05,
      "loss": 0.0106,
      "step": 2580
    },
    {
      "epoch": 3.8910798122065726,
      "grad_norm": 0.09206058084964752,
      "learning_rate": 1.4165370358382274e-05,
      "loss": 0.0117,
      "step": 2590
    },
    {
      "epoch": 3.9061032863849765,
      "grad_norm": 0.7500032782554626,
      "learning_rate": 1.3801223956956994e-05,
      "loss": 0.0147,
      "step": 2600
    },
    {
      "epoch": 3.9061032863849765,
      "eval_loss": 0.048022232949733734,
      "eval_runtime": 157.7412,
      "eval_samples_per_second": 7.506,
      "eval_steps_per_second": 7.506,
      "step": 2600
    },
    {
      "epoch": 3.9211267605633804,
      "grad_norm": 0.37353694438934326,
      "learning_rate": 1.344106841525946e-05,
      "loss": 0.0101,
      "step": 2610
    },
    {
      "epoch": 3.936150234741784,
      "grad_norm": 0.3442070484161377,
      "learning_rate": 1.3084943439887659e-05,
      "loss": 0.0111,
      "step": 2620
    },
    {
      "epoch": 3.9511737089201877,
      "grad_norm": 0.17266932129859924,
      "learning_rate": 1.273288829307579e-05,
      "loss": 0.0069,
      "step": 2630
    },
    {
      "epoch": 3.9661971830985916,
      "grad_norm": 0.09121805429458618,
      "learning_rate": 1.2384941788365622e-05,
      "loss": 0.0073,
      "step": 2640
    },
    {
      "epoch": 3.981220657276995,
      "grad_norm": 0.37061765789985657,
      "learning_rate": 1.2041142286327477e-05,
      "loss": 0.0123,
      "step": 2650
    },
    {
      "epoch": 3.981220657276995,
      "eval_loss": 0.0474095456302166,
      "eval_runtime": 157.8962,
      "eval_samples_per_second": 7.499,
      "eval_steps_per_second": 7.499,
      "step": 2650
    },
    {
      "epoch": 3.996244131455399,
      "grad_norm": 0.28421324491500854,
      "learning_rate": 1.170152769033095e-05,
      "loss": 0.0058,
      "step": 2660
    },
    {
      "epoch": 4.011267605633803,
      "grad_norm": 0.1725592464208603,
      "learning_rate": 1.1366135442366127e-05,
      "loss": 0.0066,
      "step": 2670
    },
    {
      "epoch": 4.026291079812206,
      "grad_norm": 0.144903302192688,
      "learning_rate": 1.103500251891571e-05,
      "loss": 0.0061,
      "step": 2680
    },
    {
      "epoch": 4.041314553990611,
      "grad_norm": 0.09540867805480957,
      "learning_rate": 1.0708165426878325e-05,
      "loss": 0.0042,
      "step": 2690
    },
    {
      "epoch": 4.056338028169014,
      "grad_norm": 0.04649324715137482,
      "learning_rate": 1.0385660199543812e-05,
      "loss": 0.0053,
      "step": 2700
    },
    {
      "epoch": 4.056338028169014,
      "eval_loss": 0.05291583761572838,
      "eval_runtime": 157.9865,
      "eval_samples_per_second": 7.494,
      "eval_steps_per_second": 7.494,
      "step": 2700
    },
    {
      "epoch": 4.0713615023474174,
      "grad_norm": 0.020398231223225594,
      "learning_rate": 1.0067522392620537e-05,
      "loss": 0.0037,
      "step": 2710
    },
    {
      "epoch": 4.086384976525822,
      "grad_norm": 0.25772586464881897,
      "learning_rate": 9.753787080315385e-06,
      "loss": 0.0084,
      "step": 2720
    },
    {
      "epoch": 4.101408450704225,
      "grad_norm": 0.0540861040353775,
      "learning_rate": 9.444488851467042e-06,
      "loss": 0.0051,
      "step": 2730
    },
    {
      "epoch": 4.1164319248826295,
      "grad_norm": 0.19499164819717407,
      "learning_rate": 9.139661805732435e-06,
      "loss": 0.0043,
      "step": 2740
    },
    {
      "epoch": 4.131455399061033,
      "grad_norm": 0.46163612604141235,
      "learning_rate": 8.839339549827397e-06,
      "loss": 0.0064,
      "step": 2750
    },
    {
      "epoch": 4.131455399061033,
      "eval_loss": 0.05712565779685974,
      "eval_runtime": 157.9487,
      "eval_samples_per_second": 7.496,
      "eval_steps_per_second": 7.496,
      "step": 2750
    },
    {
      "epoch": 4.146478873239436,
      "grad_norm": 0.07052065432071686,
      "learning_rate": 8.543555193821634e-06,
      "loss": 0.0045,
      "step": 2760
    },
    {
      "epoch": 4.161502347417841,
      "grad_norm": 0.3251402974128723,
      "learning_rate": 8.252341347488251e-06,
      "loss": 0.0032,
      "step": 2770
    },
    {
      "epoch": 4.176525821596244,
      "grad_norm": 0.21811746060848236,
      "learning_rate": 7.965730116708681e-06,
      "loss": 0.005,
      "step": 2780
    },
    {
      "epoch": 4.191549295774648,
      "grad_norm": 0.08314239978790283,
      "learning_rate": 7.68375309993304e-06,
      "loss": 0.0025,
      "step": 2790
    },
    {
      "epoch": 4.206572769953052,
      "grad_norm": 0.07649940252304077,
      "learning_rate": 7.406441384696372e-06,
      "loss": 0.0064,
      "step": 2800
    },
    {
      "epoch": 4.206572769953052,
      "eval_loss": 0.05896175280213356,
      "eval_runtime": 157.6893,
      "eval_samples_per_second": 7.508,
      "eval_steps_per_second": 7.508,
      "step": 2800
    },
    {
      "epoch": 4.221596244131455,
      "grad_norm": 0.14924947917461395,
      "learning_rate": 7.133825544191464e-06,
      "loss": 0.0088,
      "step": 2810
    },
    {
      "epoch": 4.236619718309859,
      "grad_norm": 0.07150545716285706,
      "learning_rate": 6.865935633897996e-06,
      "loss": 0.0019,
      "step": 2820
    },
    {
      "epoch": 4.251643192488263,
      "grad_norm": 0.13238117098808289,
      "learning_rate": 6.602801188269081e-06,
      "loss": 0.0057,
      "step": 2830
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 0.182158961892128,
      "learning_rate": 6.344451217475183e-06,
      "loss": 0.0029,
      "step": 2840
    },
    {
      "epoch": 4.28169014084507,
      "grad_norm": 0.12668107450008392,
      "learning_rate": 6.090914204205655e-06,
      "loss": 0.0038,
      "step": 2850
    },
    {
      "epoch": 4.28169014084507,
      "eval_loss": 0.060705944895744324,
      "eval_runtime": 157.8239,
      "eval_samples_per_second": 7.502,
      "eval_steps_per_second": 7.502,
      "step": 2850
    },
    {
      "epoch": 4.296713615023474,
      "grad_norm": 0.22536417841911316,
      "learning_rate": 5.842218100528679e-06,
      "loss": 0.0029,
      "step": 2860
    },
    {
      "epoch": 4.311737089201878,
      "grad_norm": 0.2726947069168091,
      "learning_rate": 5.598390324809555e-06,
      "loss": 0.0041,
      "step": 2870
    },
    {
      "epoch": 4.326760563380281,
      "grad_norm": 0.3938266634941101,
      "learning_rate": 5.359457758687841e-06,
      "loss": 0.0048,
      "step": 2880
    },
    {
      "epoch": 4.341784037558686,
      "grad_norm": 0.29681265354156494,
      "learning_rate": 5.125446744113743e-06,
      "loss": 0.0088,
      "step": 2890
    },
    {
      "epoch": 4.356807511737089,
      "grad_norm": 0.08293790370225906,
      "learning_rate": 4.896383080443934e-06,
      "loss": 0.0011,
      "step": 2900
    },
    {
      "epoch": 4.356807511737089,
      "eval_loss": 0.06272595375776291,
      "eval_runtime": 157.9854,
      "eval_samples_per_second": 7.494,
      "eval_steps_per_second": 7.494,
      "step": 2900
    },
    {
      "epoch": 4.371830985915493,
      "grad_norm": 0.15698398649692535,
      "learning_rate": 4.672292021597174e-06,
      "loss": 0.0053,
      "step": 2910
    },
    {
      "epoch": 4.386854460093897,
      "grad_norm": 0.25638025999069214,
      "learning_rate": 4.4531982732702145e-06,
      "loss": 0.0055,
      "step": 2920
    },
    {
      "epoch": 4.4018779342723,
      "grad_norm": 0.1589283049106598,
      "learning_rate": 4.239125990213883e-06,
      "loss": 0.0045,
      "step": 2930
    },
    {
      "epoch": 4.416901408450705,
      "grad_norm": 0.06673490256071091,
      "learning_rate": 4.030098773570174e-06,
      "loss": 0.001,
      "step": 2940
    },
    {
      "epoch": 4.431924882629108,
      "grad_norm": 0.12453540414571762,
      "learning_rate": 3.826139668270234e-06,
      "loss": 0.0034,
      "step": 2950
    },
    {
      "epoch": 4.431924882629108,
      "eval_loss": 0.06274248659610748,
      "eval_runtime": 157.8646,
      "eval_samples_per_second": 7.5,
      "eval_steps_per_second": 7.5,
      "step": 2950
    },
    {
      "epoch": 4.446948356807511,
      "grad_norm": 0.18250524997711182,
      "learning_rate": 3.6272711604936504e-06,
      "loss": 0.0074,
      "step": 2960
    },
    {
      "epoch": 4.461971830985916,
      "grad_norm": 0.026577528566122055,
      "learning_rate": 3.433515175189428e-06,
      "loss": 0.0033,
      "step": 2970
    },
    {
      "epoch": 4.476995305164319,
      "grad_norm": 0.03135993331670761,
      "learning_rate": 3.2448930736588e-06,
      "loss": 0.0093,
      "step": 2980
    },
    {
      "epoch": 4.492018779342723,
      "grad_norm": 0.054898951202631,
      "learning_rate": 3.061425651200117e-06,
      "loss": 0.0031,
      "step": 2990
    },
    {
      "epoch": 4.507042253521127,
      "grad_norm": 0.12721896171569824,
      "learning_rate": 2.883133134816296e-06,
      "loss": 0.0063,
      "step": 3000
    },
    {
      "epoch": 4.507042253521127,
      "eval_loss": 0.062312256544828415,
      "eval_runtime": 157.8769,
      "eval_samples_per_second": 7.5,
      "eval_steps_per_second": 7.5,
      "step": 3000
    },
    {
      "epoch": 4.52206572769953,
      "grad_norm": 0.050971172749996185,
      "learning_rate": 2.7100351809847326e-06,
      "loss": 0.0026,
      "step": 3010
    },
    {
      "epoch": 4.537089201877935,
      "grad_norm": 0.06001908332109451,
      "learning_rate": 2.542150873490251e-06,
      "loss": 0.0026,
      "step": 3020
    },
    {
      "epoch": 4.552112676056338,
      "grad_norm": 0.1522800624370575,
      "learning_rate": 2.3794987213211383e-06,
      "loss": 0.0039,
      "step": 3030
    },
    {
      "epoch": 4.567136150234742,
      "grad_norm": 0.19140265882015228,
      "learning_rate": 2.222096656628547e-06,
      "loss": 0.0041,
      "step": 3040
    },
    {
      "epoch": 4.582159624413146,
      "grad_norm": 0.031790439039468765,
      "learning_rate": 2.0699620327495174e-06,
      "loss": 0.0033,
      "step": 3050
    },
    {
      "epoch": 4.582159624413146,
      "eval_loss": 0.06346932053565979,
      "eval_runtime": 157.947,
      "eval_samples_per_second": 7.496,
      "eval_steps_per_second": 7.496,
      "step": 3050
    },
    {
      "epoch": 4.597183098591549,
      "grad_norm": 0.011058680713176727,
      "learning_rate": 1.9231116222937996e-06,
      "loss": 0.0014,
      "step": 3060
    },
    {
      "epoch": 4.612206572769953,
      "grad_norm": 0.051989588886499405,
      "learning_rate": 1.7815616152946523e-06,
      "loss": 0.0022,
      "step": 3070
    },
    {
      "epoch": 4.627230046948357,
      "grad_norm": 0.28735625743865967,
      "learning_rate": 1.6453276174240195e-06,
      "loss": 0.0016,
      "step": 3080
    },
    {
      "epoch": 4.642253521126761,
      "grad_norm": 0.7765727043151855,
      "learning_rate": 1.5144246482719114e-06,
      "loss": 0.0032,
      "step": 3090
    },
    {
      "epoch": 4.657276995305164,
      "grad_norm": 0.16394619643688202,
      "learning_rate": 1.3888671396905805e-06,
      "loss": 0.0048,
      "step": 3100
    },
    {
      "epoch": 4.657276995305164,
      "eval_loss": 0.06427957862615585,
      "eval_runtime": 157.8435,
      "eval_samples_per_second": 7.501,
      "eval_steps_per_second": 7.501,
      "step": 3100
    },
    {
      "epoch": 4.672300469483568,
      "grad_norm": 0.07136521488428116,
      "learning_rate": 1.2686689342034431e-06,
      "loss": 0.0075,
      "step": 3110
    },
    {
      "epoch": 4.687323943661972,
      "grad_norm": 0.06406821310520172,
      "learning_rate": 1.1538432834789227e-06,
      "loss": 0.0058,
      "step": 3120
    },
    {
      "epoch": 4.702347417840375,
      "grad_norm": 0.42130792140960693,
      "learning_rate": 1.044402846869491e-06,
      "loss": 0.0041,
      "step": 3130
    },
    {
      "epoch": 4.71737089201878,
      "grad_norm": 0.6147840023040771,
      "learning_rate": 9.403596900160073e-07,
      "loss": 0.003,
      "step": 3140
    },
    {
      "epoch": 4.732394366197183,
      "grad_norm": 0.4716566205024719,
      "learning_rate": 8.417252835174749e-07,
      "loss": 0.0044,
      "step": 3150
    },
    {
      "epoch": 4.732394366197183,
      "eval_loss": 0.06363623589277267,
      "eval_runtime": 157.8646,
      "eval_samples_per_second": 7.5,
      "eval_steps_per_second": 7.5,
      "step": 3150
    },
    {
      "epoch": 4.7474178403755865,
      "grad_norm": 0.2807024419307709,
      "learning_rate": 7.48510501666455e-07,
      "loss": 0.0079,
      "step": 3160
    },
    {
      "epoch": 4.762441314553991,
      "grad_norm": 0.3779400587081909,
      "learning_rate": 6.607256212501578e-07,
      "loss": 0.0041,
      "step": 3170
    },
    {
      "epoch": 4.777464788732394,
      "grad_norm": 0.3377034068107605,
      "learning_rate": 5.783803204174654e-07,
      "loss": 0.005,
      "step": 3180
    },
    {
      "epoch": 4.792488262910798,
      "grad_norm": 0.15139789879322052,
      "learning_rate": 5.014836776119358e-07,
      "loss": 0.0027,
      "step": 3190
    },
    {
      "epoch": 4.807511737089202,
      "grad_norm": 0.15307891368865967,
      "learning_rate": 4.300441705708924e-07,
      "loss": 0.0031,
      "step": 3200
    },
    {
      "epoch": 4.807511737089202,
      "eval_loss": 0.06359302997589111,
      "eval_runtime": 157.9251,
      "eval_samples_per_second": 7.497,
      "eval_steps_per_second": 7.497,
      "step": 3200
    },
    {
      "epoch": 4.822535211267605,
      "grad_norm": 0.28907057642936707,
      "learning_rate": 3.6406967539078796e-07,
      "loss": 0.0047,
      "step": 3210
    },
    {
      "epoch": 4.83755868544601,
      "grad_norm": 0.10809265822172165,
      "learning_rate": 3.0356746565887715e-07,
      "loss": 0.004,
      "step": 3220
    },
    {
      "epoch": 4.852582159624413,
      "grad_norm": 0.07856517285108566,
      "learning_rate": 2.485442116513026e-07,
      "loss": 0.0031,
      "step": 3230
    },
    {
      "epoch": 4.867605633802817,
      "grad_norm": 0.5275461673736572,
      "learning_rate": 1.9900597959770507e-07,
      "loss": 0.0087,
      "step": 3240
    },
    {
      "epoch": 4.882629107981221,
      "grad_norm": 0.014800317585468292,
      "learning_rate": 1.5495823101245866e-07,
      "loss": 0.0022,
      "step": 3250
    },
    {
      "epoch": 4.882629107981221,
      "eval_loss": 0.06390237808227539,
      "eval_runtime": 157.8776,
      "eval_samples_per_second": 7.499,
      "eval_steps_per_second": 7.499,
      "step": 3250
    },
    {
      "epoch": 4.897652582159624,
      "grad_norm": 0.04556208476424217,
      "learning_rate": 1.164058220925135e-07,
      "loss": 0.006,
      "step": 3260
    },
    {
      "epoch": 4.912676056338028,
      "grad_norm": 0.20600998401641846,
      "learning_rate": 8.335300318201844e-08,
      "loss": 0.0067,
      "step": 3270
    },
    {
      "epoch": 4.927699530516432,
      "grad_norm": 0.8991699814796448,
      "learning_rate": 5.5803418303745917e-08,
      "loss": 0.008,
      "step": 3280
    },
    {
      "epoch": 4.942723004694836,
      "grad_norm": 0.015831077471375465,
      "learning_rate": 3.3760104757313284e-08,
      "loss": 0.0015,
      "step": 3290
    },
    {
      "epoch": 4.957746478873239,
      "grad_norm": 0.1406923532485962,
      "learning_rate": 1.7225492784345156e-08,
      "loss": 0.0078,
      "step": 3300
    },
    {
      "epoch": 4.957746478873239,
      "eval_loss": 0.06379479169845581,
      "eval_runtime": 157.8707,
      "eval_samples_per_second": 7.5,
      "eval_steps_per_second": 7.5,
      "step": 3300
    },
    {
      "epoch": 4.972769953051643,
      "grad_norm": 0.062291864305734634,
      "learning_rate": 6.201405300532148e-09,
      "loss": 0.006,
      "step": 3310
    },
    {
      "epoch": 4.987793427230047,
      "grad_norm": 2.030796527862549,
      "learning_rate": 6.890576946805282e-10,
      "loss": 0.0032,
      "step": 3320
    },
    {
      "epoch": 4.995305164319249,
      "step": 3325,
      "total_flos": 8.345060490986127e+17,
      "train_loss": 0.04504659976081965,
      "train_runtime": 35784.009,
      "train_samples_per_second": 1.488,
      "train_steps_per_second": 0.093
    }
  ],
  "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": 8.345060490986127e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}