{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.748031496062993,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07874015748031496,
      "grad_norm": 5.119835376739502,
      "learning_rate": 1.8e-06,
      "loss": 0.8008,
      "step": 10
    },
    {
      "epoch": 0.15748031496062992,
      "grad_norm": 2.7846670150756836,
      "learning_rate": 3.8e-06,
      "loss": 0.6753,
      "step": 20
    },
    {
      "epoch": 0.23622047244094488,
      "grad_norm": 0.9429484605789185,
      "learning_rate": 5.8e-06,
      "loss": 0.3854,
      "step": 30
    },
    {
      "epoch": 0.31496062992125984,
      "grad_norm": 0.4289324879646301,
      "learning_rate": 7.8e-06,
      "loss": 0.2374,
      "step": 40
    },
    {
      "epoch": 0.3937007874015748,
      "grad_norm": 0.7212084531784058,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.1937,
      "step": 50
    },
    {
      "epoch": 0.47244094488188976,
      "grad_norm": 0.3743739426136017,
      "learning_rate": 1.18e-05,
      "loss": 0.169,
      "step": 60
    },
    {
      "epoch": 0.5511811023622047,
      "grad_norm": 0.33906233310699463,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.1513,
      "step": 70
    },
    {
      "epoch": 0.6299212598425197,
      "grad_norm": 0.4938454031944275,
      "learning_rate": 1.58e-05,
      "loss": 0.1333,
      "step": 80
    },
    {
      "epoch": 0.7086614173228346,
      "grad_norm": 0.3791239261627197,
      "learning_rate": 1.78e-05,
      "loss": 0.1209,
      "step": 90
    },
    {
      "epoch": 0.7874015748031497,
      "grad_norm": 0.31115081906318665,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 0.108,
      "step": 100
    },
    {
      "epoch": 0.8661417322834646,
      "grad_norm": 0.3266952633857727,
      "learning_rate": 2.18e-05,
      "loss": 0.105,
      "step": 110
    },
    {
      "epoch": 0.9448818897637795,
      "grad_norm": 0.42559316754341125,
      "learning_rate": 2.38e-05,
      "loss": 0.0931,
      "step": 120
    },
    {
      "epoch": 1.0236220472440944,
      "grad_norm": 0.40235716104507446,
      "learning_rate": 2.58e-05,
      "loss": 0.0867,
      "step": 130
    },
    {
      "epoch": 1.1023622047244095,
      "grad_norm": 0.3455784022808075,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.0843,
      "step": 140
    },
    {
      "epoch": 1.1811023622047245,
      "grad_norm": 0.39895790815353394,
      "learning_rate": 2.98e-05,
      "loss": 0.0803,
      "step": 150
    },
    {
      "epoch": 1.2598425196850394,
      "grad_norm": 0.32412493228912354,
      "learning_rate": 3.18e-05,
      "loss": 0.075,
      "step": 160
    },
    {
      "epoch": 1.3385826771653544,
      "grad_norm": 0.3238581717014313,
      "learning_rate": 3.38e-05,
      "loss": 0.069,
      "step": 170
    },
    {
      "epoch": 1.4173228346456692,
      "grad_norm": 0.35713112354278564,
      "learning_rate": 3.58e-05,
      "loss": 0.0632,
      "step": 180
    },
    {
      "epoch": 1.4960629921259843,
      "grad_norm": 0.31078988313674927,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 0.0606,
      "step": 190
    },
    {
      "epoch": 1.574803149606299,
      "grad_norm": 0.23168951272964478,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 0.0566,
      "step": 200
    },
    {
      "epoch": 1.6535433070866141,
      "grad_norm": 0.2528112828731537,
      "learning_rate": 4.18e-05,
      "loss": 0.0553,
      "step": 210
    },
    {
      "epoch": 1.7322834645669292,
      "grad_norm": 0.31132972240448,
      "learning_rate": 4.38e-05,
      "loss": 0.0516,
      "step": 220
    },
    {
      "epoch": 1.811023622047244,
      "grad_norm": 0.3506482243537903,
      "learning_rate": 4.58e-05,
      "loss": 0.051,
      "step": 230
    },
    {
      "epoch": 1.889763779527559,
      "grad_norm": 0.30420321226119995,
      "learning_rate": 4.78e-05,
      "loss": 0.0498,
      "step": 240
    },
    {
      "epoch": 1.968503937007874,
      "grad_norm": 0.27608105540275574,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 0.0425,
      "step": 250
    },
    {
      "epoch": 2.047244094488189,
      "grad_norm": 0.24153359234333038,
      "learning_rate": 5.1800000000000005e-05,
      "loss": 0.0425,
      "step": 260
    },
    {
      "epoch": 2.1259842519685037,
      "grad_norm": 0.2684983015060425,
      "learning_rate": 5.380000000000001e-05,
      "loss": 0.0397,
      "step": 270
    },
    {
      "epoch": 2.204724409448819,
      "grad_norm": 0.2812291979789734,
      "learning_rate": 5.580000000000001e-05,
      "loss": 0.0347,
      "step": 280
    },
    {
      "epoch": 2.283464566929134,
      "grad_norm": 0.272079199552536,
      "learning_rate": 5.7799999999999995e-05,
      "loss": 0.0339,
      "step": 290
    },
    {
      "epoch": 2.362204724409449,
      "grad_norm": 0.30601683259010315,
      "learning_rate": 5.9800000000000003e-05,
      "loss": 0.0359,
      "step": 300
    },
    {
      "epoch": 2.440944881889764,
      "grad_norm": 0.3129172921180725,
      "learning_rate": 6.18e-05,
      "loss": 0.0351,
      "step": 310
    },
    {
      "epoch": 2.5196850393700787,
      "grad_norm": 0.27252131700515747,
      "learning_rate": 6.38e-05,
      "loss": 0.0287,
      "step": 320
    },
    {
      "epoch": 2.5984251968503935,
      "grad_norm": 0.2653070390224457,
      "learning_rate": 6.58e-05,
      "loss": 0.0313,
      "step": 330
    },
    {
      "epoch": 2.677165354330709,
      "grad_norm": 0.35808777809143066,
      "learning_rate": 6.780000000000001e-05,
      "loss": 0.0356,
      "step": 340
    },
    {
      "epoch": 2.7559055118110236,
      "grad_norm": 0.26742085814476013,
      "learning_rate": 6.98e-05,
      "loss": 0.0299,
      "step": 350
    },
    {
      "epoch": 2.8346456692913384,
      "grad_norm": 0.4106348156929016,
      "learning_rate": 7.18e-05,
      "loss": 0.0324,
      "step": 360
    },
    {
      "epoch": 2.9133858267716537,
      "grad_norm": 0.213535338640213,
      "learning_rate": 7.38e-05,
      "loss": 0.0273,
      "step": 370
    },
    {
      "epoch": 2.9921259842519685,
      "grad_norm": 0.26808497309684753,
      "learning_rate": 7.58e-05,
      "loss": 0.0254,
      "step": 380
    },
    {
      "epoch": 3.0708661417322833,
      "grad_norm": 0.18177832663059235,
      "learning_rate": 7.780000000000001e-05,
      "loss": 0.0261,
      "step": 390
    },
    {
      "epoch": 3.1496062992125986,
      "grad_norm": 0.2706851065158844,
      "learning_rate": 7.98e-05,
      "loss": 0.0268,
      "step": 400
    },
    {
      "epoch": 3.2283464566929134,
      "grad_norm": 0.29524528980255127,
      "learning_rate": 8.18e-05,
      "loss": 0.0285,
      "step": 410
    },
    {
      "epoch": 3.3070866141732282,
      "grad_norm": 0.21399272978305817,
      "learning_rate": 8.38e-05,
      "loss": 0.0252,
      "step": 420
    },
    {
      "epoch": 3.3858267716535435,
      "grad_norm": 0.290097177028656,
      "learning_rate": 8.58e-05,
      "loss": 0.0279,
      "step": 430
    },
    {
      "epoch": 3.4645669291338583,
      "grad_norm": 0.2789689004421234,
      "learning_rate": 8.78e-05,
      "loss": 0.0236,
      "step": 440
    },
    {
      "epoch": 3.543307086614173,
      "grad_norm": 0.3307545781135559,
      "learning_rate": 8.98e-05,
      "loss": 0.0256,
      "step": 450
    },
    {
      "epoch": 3.622047244094488,
      "grad_norm": 0.2919306457042694,
      "learning_rate": 9.180000000000001e-05,
      "loss": 0.0227,
      "step": 460
    },
    {
      "epoch": 3.7007874015748032,
      "grad_norm": 0.27534034848213196,
      "learning_rate": 9.38e-05,
      "loss": 0.0219,
      "step": 470
    },
    {
      "epoch": 3.779527559055118,
      "grad_norm": 0.26348116993904114,
      "learning_rate": 9.58e-05,
      "loss": 0.0242,
      "step": 480
    },
    {
      "epoch": 3.8582677165354333,
      "grad_norm": 0.29468125104904175,
      "learning_rate": 9.78e-05,
      "loss": 0.0224,
      "step": 490
    },
    {
      "epoch": 3.937007874015748,
      "grad_norm": 0.20534993708133698,
      "learning_rate": 9.98e-05,
      "loss": 0.0235,
      "step": 500
    },
    {
      "epoch": 4.015748031496063,
      "grad_norm": 0.2911393642425537,
      "learning_rate": 9.9999778549206e-05,
      "loss": 0.0206,
      "step": 510
    },
    {
      "epoch": 4.094488188976378,
      "grad_norm": 0.2478438913822174,
      "learning_rate": 9.999901304280685e-05,
      "loss": 0.0219,
      "step": 520
    },
    {
      "epoch": 4.173228346456693,
      "grad_norm": 0.32605063915252686,
      "learning_rate": 9.999770075521164e-05,
      "loss": 0.0232,
      "step": 530
    },
    {
      "epoch": 4.251968503937007,
      "grad_norm": 0.22585000097751617,
      "learning_rate": 9.99958417007713e-05,
      "loss": 0.0228,
      "step": 540
    },
    {
      "epoch": 4.330708661417323,
      "grad_norm": 0.3227289319038391,
      "learning_rate": 9.999343589981615e-05,
      "loss": 0.018,
      "step": 550
    },
    {
      "epoch": 4.409448818897638,
      "grad_norm": 0.2523372769355774,
      "learning_rate": 9.999048337865568e-05,
      "loss": 0.0215,
      "step": 560
    },
    {
      "epoch": 4.488188976377953,
      "grad_norm": 0.34720173478126526,
      "learning_rate": 9.998698416957815e-05,
      "loss": 0.0242,
      "step": 570
    },
    {
      "epoch": 4.566929133858268,
      "grad_norm": 0.24005654454231262,
      "learning_rate": 9.998293831085037e-05,
      "loss": 0.0213,
      "step": 580
    },
    {
      "epoch": 4.645669291338582,
      "grad_norm": 0.28940242528915405,
      "learning_rate": 9.997834584671719e-05,
      "loss": 0.0204,
      "step": 590
    },
    {
      "epoch": 4.724409448818898,
      "grad_norm": 0.2654191255569458,
      "learning_rate": 9.997320682740107e-05,
      "loss": 0.0217,
      "step": 600
    },
    {
      "epoch": 4.803149606299213,
      "grad_norm": 0.2912241816520691,
      "learning_rate": 9.996752130910149e-05,
      "loss": 0.0197,
      "step": 610
    },
    {
      "epoch": 4.881889763779528,
      "grad_norm": 0.23718924820423126,
      "learning_rate": 9.99612893539944e-05,
      "loss": 0.0209,
      "step": 620
    },
    {
      "epoch": 4.960629921259843,
      "grad_norm": 0.2647818326950073,
      "learning_rate": 9.995451103023144e-05,
      "loss": 0.0222,
      "step": 630
    },
    {
      "epoch": 5.039370078740157,
      "grad_norm": 0.35324886441230774,
      "learning_rate": 9.994718641193928e-05,
      "loss": 0.0224,
      "step": 640
    },
    {
      "epoch": 5.118110236220472,
      "grad_norm": 0.2671961188316345,
      "learning_rate": 9.993931557921874e-05,
      "loss": 0.0219,
      "step": 650
    },
    {
      "epoch": 5.196850393700787,
      "grad_norm": 0.2596529722213745,
      "learning_rate": 9.993089861814402e-05,
      "loss": 0.0203,
      "step": 660
    },
    {
      "epoch": 5.275590551181103,
      "grad_norm": 0.25885483622550964,
      "learning_rate": 9.992193562076166e-05,
      "loss": 0.0188,
      "step": 670
    },
    {
      "epoch": 5.354330708661418,
      "grad_norm": 0.24976016581058502,
      "learning_rate": 9.991242668508954e-05,
      "loss": 0.0175,
      "step": 680
    },
    {
      "epoch": 5.433070866141732,
      "grad_norm": 0.24121227860450745,
      "learning_rate": 9.990237191511587e-05,
      "loss": 0.0158,
      "step": 690
    },
    {
      "epoch": 5.511811023622047,
      "grad_norm": 0.22227917611598969,
      "learning_rate": 9.989177142079802e-05,
      "loss": 0.0177,
      "step": 700
    },
    {
      "epoch": 5.590551181102362,
      "grad_norm": 0.231464222073555,
      "learning_rate": 9.988062531806126e-05,
      "loss": 0.0183,
      "step": 710
    },
    {
      "epoch": 5.669291338582677,
      "grad_norm": 0.16609017550945282,
      "learning_rate": 9.986893372879762e-05,
      "loss": 0.018,
      "step": 720
    },
    {
      "epoch": 5.748031496062993,
      "grad_norm": 0.19624024629592896,
      "learning_rate": 9.985669678086443e-05,
      "loss": 0.018,
      "step": 730
    },
    {
      "epoch": 5.826771653543307,
      "grad_norm": 0.22255055606365204,
      "learning_rate": 9.984391460808298e-05,
      "loss": 0.0199,
      "step": 740
    },
    {
      "epoch": 5.905511811023622,
      "grad_norm": 0.22765639424324036,
      "learning_rate": 9.983058735023709e-05,
      "loss": 0.0191,
      "step": 750
    },
    {
      "epoch": 5.984251968503937,
      "grad_norm": 0.23915418982505798,
      "learning_rate": 9.98167151530715e-05,
      "loss": 0.0178,
      "step": 760
    },
    {
      "epoch": 6.062992125984252,
      "grad_norm": 0.2489311248064041,
      "learning_rate": 9.980229816829034e-05,
      "loss": 0.0202,
      "step": 770
    },
    {
      "epoch": 6.141732283464567,
      "grad_norm": 0.22865547239780426,
      "learning_rate": 9.978733655355544e-05,
      "loss": 0.0187,
      "step": 780
    },
    {
      "epoch": 6.2204724409448815,
      "grad_norm": 0.19393905997276306,
      "learning_rate": 9.977183047248464e-05,
      "loss": 0.0168,
      "step": 790
    },
    {
      "epoch": 6.299212598425197,
      "grad_norm": 0.20525363087654114,
      "learning_rate": 9.975578009464992e-05,
      "loss": 0.018,
      "step": 800
    },
    {
      "epoch": 6.377952755905512,
      "grad_norm": 0.2537108063697815,
      "learning_rate": 9.97391855955757e-05,
      "loss": 0.0143,
      "step": 810
    },
    {
      "epoch": 6.456692913385827,
      "grad_norm": 0.2665018141269684,
      "learning_rate": 9.972204715673669e-05,
      "loss": 0.0165,
      "step": 820
    },
    {
      "epoch": 6.535433070866142,
      "grad_norm": 0.18383699655532837,
      "learning_rate": 9.970436496555617e-05,
      "loss": 0.0164,
      "step": 830
    },
    {
      "epoch": 6.6141732283464565,
      "grad_norm": 0.3430931270122528,
      "learning_rate": 9.968613921540373e-05,
      "loss": 0.0176,
      "step": 840
    },
    {
      "epoch": 6.692913385826771,
      "grad_norm": 0.2601425349712372,
      "learning_rate": 9.966737010559326e-05,
      "loss": 0.0175,
      "step": 850
    },
    {
      "epoch": 6.771653543307087,
      "grad_norm": 0.19988982379436493,
      "learning_rate": 9.964805784138072e-05,
      "loss": 0.0172,
      "step": 860
    },
    {
      "epoch": 6.850393700787402,
      "grad_norm": 0.18660953640937805,
      "learning_rate": 9.962820263396195e-05,
      "loss": 0.0158,
      "step": 870
    },
    {
      "epoch": 6.929133858267717,
      "grad_norm": 0.22756962478160858,
      "learning_rate": 9.960780470047033e-05,
      "loss": 0.0185,
      "step": 880
    },
    {
      "epoch": 7.0078740157480315,
      "grad_norm": 0.14548353850841522,
      "learning_rate": 9.958686426397437e-05,
      "loss": 0.0164,
      "step": 890
    },
    {
      "epoch": 7.086614173228346,
      "grad_norm": 0.20737145841121674,
      "learning_rate": 9.956538155347534e-05,
      "loss": 0.0182,
      "step": 900
    },
    {
      "epoch": 7.165354330708661,
      "grad_norm": 0.20689648389816284,
      "learning_rate": 9.95433568039047e-05,
      "loss": 0.0145,
      "step": 910
    },
    {
      "epoch": 7.244094488188976,
      "grad_norm": 0.26220783591270447,
      "learning_rate": 9.952079025612162e-05,
      "loss": 0.0145,
      "step": 920
    },
    {
      "epoch": 7.322834645669292,
      "grad_norm": 0.23523452877998352,
      "learning_rate": 9.949768215691022e-05,
      "loss": 0.0168,
      "step": 930
    },
    {
      "epoch": 7.4015748031496065,
      "grad_norm": 0.207063227891922,
      "learning_rate": 9.9474032758977e-05,
      "loss": 0.0154,
      "step": 940
    },
    {
      "epoch": 7.480314960629921,
      "grad_norm": 0.2092580646276474,
      "learning_rate": 9.944984232094794e-05,
      "loss": 0.0169,
      "step": 950
    },
    {
      "epoch": 7.559055118110236,
      "grad_norm": 0.1808154582977295,
      "learning_rate": 9.942511110736584e-05,
      "loss": 0.0157,
      "step": 960
    },
    {
      "epoch": 7.637795275590551,
      "grad_norm": 0.2190985083580017,
      "learning_rate": 9.939983938868726e-05,
      "loss": 0.0155,
      "step": 970
    },
    {
      "epoch": 7.716535433070866,
      "grad_norm": 0.1607908308506012,
      "learning_rate": 9.93740274412797e-05,
      "loss": 0.0136,
      "step": 980
    },
    {
      "epoch": 7.7952755905511815,
      "grad_norm": 0.20882774889469147,
      "learning_rate": 9.934767554741846e-05,
      "loss": 0.0192,
      "step": 990
    },
    {
      "epoch": 7.874015748031496,
      "grad_norm": 0.18141894042491913,
      "learning_rate": 9.932078399528361e-05,
      "loss": 0.0134,
      "step": 1000
    },
    {
      "epoch": 7.952755905511811,
      "grad_norm": 0.1842644363641739,
      "learning_rate": 9.929335307895689e-05,
      "loss": 0.0145,
      "step": 1010
    },
    {
      "epoch": 8.031496062992126,
      "grad_norm": 0.19102592766284943,
      "learning_rate": 9.926538309841839e-05,
      "loss": 0.0179,
      "step": 1020
    },
    {
      "epoch": 8.11023622047244,
      "grad_norm": 0.2554001212120056,
      "learning_rate": 9.923687435954334e-05,
      "loss": 0.0145,
      "step": 1030
    },
    {
      "epoch": 8.188976377952756,
      "grad_norm": 0.2188219279050827,
      "learning_rate": 9.920782717409873e-05,
      "loss": 0.0133,
      "step": 1040
    },
    {
      "epoch": 8.26771653543307,
      "grad_norm": 0.19668325781822205,
      "learning_rate": 9.917824185973994e-05,
      "loss": 0.013,
      "step": 1050
    },
    {
      "epoch": 8.346456692913385,
      "grad_norm": 0.19224300980567932,
      "learning_rate": 9.914811874000723e-05,
      "loss": 0.012,
      "step": 1060
    },
    {
      "epoch": 8.4251968503937,
      "grad_norm": 0.2617517113685608,
      "learning_rate": 9.911745814432218e-05,
      "loss": 0.0144,
      "step": 1070
    },
    {
      "epoch": 8.503937007874015,
      "grad_norm": 0.340850293636322,
      "learning_rate": 9.90862604079842e-05,
      "loss": 0.0163,
      "step": 1080
    },
    {
      "epoch": 8.582677165354331,
      "grad_norm": 0.24036389589309692,
      "learning_rate": 9.90545258721667e-05,
      "loss": 0.0143,
      "step": 1090
    },
    {
      "epoch": 8.661417322834646,
      "grad_norm": 0.2523621916770935,
      "learning_rate": 9.90222548839135e-05,
      "loss": 0.0137,
      "step": 1100
    },
    {
      "epoch": 8.740157480314961,
      "grad_norm": 0.25303855538368225,
      "learning_rate": 9.898944779613495e-05,
      "loss": 0.0124,
      "step": 1110
    },
    {
      "epoch": 8.818897637795276,
      "grad_norm": 0.2672367990016937,
      "learning_rate": 9.89561049676041e-05,
      "loss": 0.0135,
      "step": 1120
    },
    {
      "epoch": 8.89763779527559,
      "grad_norm": 0.22292408347129822,
      "learning_rate": 9.89222267629528e-05,
      "loss": 0.0155,
      "step": 1130
    },
    {
      "epoch": 8.976377952755906,
      "grad_norm": 0.2113981992006302,
      "learning_rate": 9.888781355266763e-05,
      "loss": 0.0139,
      "step": 1140
    },
    {
      "epoch": 9.05511811023622,
      "grad_norm": 0.16752807796001434,
      "learning_rate": 9.885286571308598e-05,
      "loss": 0.0124,
      "step": 1150
    },
    {
      "epoch": 9.133858267716535,
      "grad_norm": 0.1773703545331955,
      "learning_rate": 9.881738362639182e-05,
      "loss": 0.015,
      "step": 1160
    },
    {
      "epoch": 9.21259842519685,
      "grad_norm": 0.26974138617515564,
      "learning_rate": 9.878136768061154e-05,
      "loss": 0.0162,
      "step": 1170
    },
    {
      "epoch": 9.291338582677165,
      "grad_norm": 0.2184063196182251,
      "learning_rate": 9.874481826960979e-05,
      "loss": 0.0148,
      "step": 1180
    },
    {
      "epoch": 9.37007874015748,
      "grad_norm": 0.1977306753396988,
      "learning_rate": 9.870773579308503e-05,
      "loss": 0.0123,
      "step": 1190
    },
    {
      "epoch": 9.448818897637794,
      "grad_norm": 0.1981269121170044,
      "learning_rate": 9.867012065656533e-05,
      "loss": 0.0152,
      "step": 1200
    },
    {
      "epoch": 9.527559055118111,
      "grad_norm": 0.17817805707454681,
      "learning_rate": 9.863197327140376e-05,
      "loss": 0.0123,
      "step": 1210
    },
    {
      "epoch": 9.606299212598426,
      "grad_norm": 0.23420843482017517,
      "learning_rate": 9.859329405477403e-05,
      "loss": 0.0129,
      "step": 1220
    },
    {
      "epoch": 9.68503937007874,
      "grad_norm": 0.25216200947761536,
      "learning_rate": 9.855408342966585e-05,
      "loss": 0.0138,
      "step": 1230
    },
    {
      "epoch": 9.763779527559056,
      "grad_norm": 0.1990588754415512,
      "learning_rate": 9.851434182488033e-05,
      "loss": 0.0129,
      "step": 1240
    },
    {
      "epoch": 9.84251968503937,
      "grad_norm": 0.27837619185447693,
      "learning_rate": 9.84740696750253e-05,
      "loss": 0.0124,
      "step": 1250
    },
    {
      "epoch": 9.921259842519685,
      "grad_norm": 0.21090054512023926,
      "learning_rate": 9.843326742051055e-05,
      "loss": 0.013,
      "step": 1260
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.19581645727157593,
      "learning_rate": 9.839193550754297e-05,
      "loss": 0.0126,
      "step": 1270
    },
    {
      "epoch": 10.078740157480315,
      "grad_norm": 0.21251627802848816,
      "learning_rate": 9.835007438812177e-05,
      "loss": 0.0148,
      "step": 1280
    },
    {
      "epoch": 10.15748031496063,
      "grad_norm": 0.18511821329593658,
      "learning_rate": 9.830768452003341e-05,
      "loss": 0.0133,
      "step": 1290
    },
    {
      "epoch": 10.236220472440944,
      "grad_norm": 0.18811464309692383,
      "learning_rate": 9.826476636684671e-05,
      "loss": 0.0126,
      "step": 1300
    },
    {
      "epoch": 10.31496062992126,
      "grad_norm": 0.18782231211662292,
      "learning_rate": 9.822132039790773e-05,
      "loss": 0.0117,
      "step": 1310
    },
    {
      "epoch": 10.393700787401574,
      "grad_norm": 0.16824057698249817,
      "learning_rate": 9.817734708833461e-05,
      "loss": 0.0106,
      "step": 1320
    },
    {
      "epoch": 10.472440944881889,
      "grad_norm": 0.1814710795879364,
      "learning_rate": 9.813284691901243e-05,
      "loss": 0.0162,
      "step": 1330
    },
    {
      "epoch": 10.551181102362206,
      "grad_norm": 0.2217687964439392,
      "learning_rate": 9.808782037658792e-05,
      "loss": 0.0155,
      "step": 1340
    },
    {
      "epoch": 10.62992125984252,
      "grad_norm": 0.19781896471977234,
      "learning_rate": 9.804226795346411e-05,
      "loss": 0.0133,
      "step": 1350
    },
    {
      "epoch": 10.708661417322835,
      "grad_norm": 0.24714171886444092,
      "learning_rate": 9.799619014779503e-05,
      "loss": 0.0129,
      "step": 1360
    },
    {
      "epoch": 10.78740157480315,
      "grad_norm": 0.16805458068847656,
      "learning_rate": 9.794958746348013e-05,
      "loss": 0.0125,
      "step": 1370
    },
    {
      "epoch": 10.866141732283465,
      "grad_norm": 0.18694327771663666,
      "learning_rate": 9.790246041015896e-05,
      "loss": 0.0112,
      "step": 1380
    },
    {
      "epoch": 10.94488188976378,
      "grad_norm": 0.21768535673618317,
      "learning_rate": 9.785480950320538e-05,
      "loss": 0.0121,
      "step": 1390
    },
    {
      "epoch": 11.023622047244094,
      "grad_norm": 0.16912485659122467,
      "learning_rate": 9.78066352637221e-05,
      "loss": 0.0109,
      "step": 1400
    },
    {
      "epoch": 11.10236220472441,
      "grad_norm": 0.15913233160972595,
      "learning_rate": 9.775793821853488e-05,
      "loss": 0.0115,
      "step": 1410
    },
    {
      "epoch": 11.181102362204724,
      "grad_norm": 0.15250848233699799,
      "learning_rate": 9.77087189001868e-05,
      "loss": 0.0123,
      "step": 1420
    },
    {
      "epoch": 11.259842519685039,
      "grad_norm": 0.17317131161689758,
      "learning_rate": 9.765897784693243e-05,
      "loss": 0.0117,
      "step": 1430
    },
    {
      "epoch": 11.338582677165354,
      "grad_norm": 0.23304998874664307,
      "learning_rate": 9.760871560273197e-05,
      "loss": 0.0107,
      "step": 1440
    },
    {
      "epoch": 11.417322834645669,
      "grad_norm": 0.2260117381811142,
      "learning_rate": 9.755793271724526e-05,
      "loss": 0.0113,
      "step": 1450
    },
    {
      "epoch": 11.496062992125985,
      "grad_norm": 0.20854035019874573,
      "learning_rate": 9.750662974582584e-05,
      "loss": 0.0156,
      "step": 1460
    },
    {
      "epoch": 11.5748031496063,
      "grad_norm": 0.18729598820209503,
      "learning_rate": 9.745480724951473e-05,
      "loss": 0.0115,
      "step": 1470
    },
    {
      "epoch": 11.653543307086615,
      "grad_norm": 0.1489574909210205,
      "learning_rate": 9.740246579503447e-05,
      "loss": 0.0122,
      "step": 1480
    },
    {
      "epoch": 11.73228346456693,
      "grad_norm": 0.16865724325180054,
      "learning_rate": 9.734960595478284e-05,
      "loss": 0.0121,
      "step": 1490
    },
    {
      "epoch": 11.811023622047244,
      "grad_norm": 0.1705121397972107,
      "learning_rate": 9.729622830682657e-05,
      "loss": 0.0117,
      "step": 1500
    },
    {
      "epoch": 11.88976377952756,
      "grad_norm": 0.12779462337493896,
      "learning_rate": 9.724233343489504e-05,
      "loss": 0.013,
      "step": 1510
    },
    {
      "epoch": 11.968503937007874,
      "grad_norm": 0.21109400689601898,
      "learning_rate": 9.718792192837396e-05,
      "loss": 0.0105,
      "step": 1520
    },
    {
      "epoch": 12.047244094488189,
      "grad_norm": 0.17350123822689056,
      "learning_rate": 9.713299438229886e-05,
      "loss": 0.0129,
      "step": 1530
    },
    {
      "epoch": 12.125984251968504,
      "grad_norm": 0.19555015861988068,
      "learning_rate": 9.707755139734855e-05,
      "loss": 0.0131,
      "step": 1540
    },
    {
      "epoch": 12.204724409448819,
      "grad_norm": 0.22949132323265076,
      "learning_rate": 9.702159357983866e-05,
      "loss": 0.0122,
      "step": 1550
    },
    {
      "epoch": 12.283464566929133,
      "grad_norm": 0.21299389004707336,
      "learning_rate": 9.696512154171492e-05,
      "loss": 0.013,
      "step": 1560
    },
    {
      "epoch": 12.362204724409448,
      "grad_norm": 0.2029636800289154,
      "learning_rate": 9.690813590054645e-05,
      "loss": 0.0127,
      "step": 1570
    },
    {
      "epoch": 12.440944881889763,
      "grad_norm": 0.2509428858757019,
      "learning_rate": 9.685063727951914e-05,
      "loss": 0.0115,
      "step": 1580
    },
    {
      "epoch": 12.519685039370078,
      "grad_norm": 0.17952832579612732,
      "learning_rate": 9.679262630742865e-05,
      "loss": 0.0123,
      "step": 1590
    },
    {
      "epoch": 12.598425196850394,
      "grad_norm": 0.17356553673744202,
      "learning_rate": 9.673410361867373e-05,
      "loss": 0.0133,
      "step": 1600
    },
    {
      "epoch": 12.67716535433071,
      "grad_norm": 0.2649160921573639,
      "learning_rate": 9.667506985324909e-05,
      "loss": 0.0116,
      "step": 1610
    },
    {
      "epoch": 12.755905511811024,
      "grad_norm": 0.209790900349617,
      "learning_rate": 9.661552565673855e-05,
      "loss": 0.0104,
      "step": 1620
    },
    {
      "epoch": 12.834645669291339,
      "grad_norm": 0.21641805768013,
      "learning_rate": 9.655547168030789e-05,
      "loss": 0.0129,
      "step": 1630
    },
    {
      "epoch": 12.913385826771654,
      "grad_norm": 0.2454116940498352,
      "learning_rate": 9.649490858069777e-05,
      "loss": 0.0104,
      "step": 1640
    },
    {
      "epoch": 12.992125984251969,
      "grad_norm": 0.21532991528511047,
      "learning_rate": 9.643383702021658e-05,
      "loss": 0.0111,
      "step": 1650
    },
    {
      "epoch": 13.070866141732283,
      "grad_norm": 0.22105859220027924,
      "learning_rate": 9.637225766673307e-05,
      "loss": 0.0098,
      "step": 1660
    },
    {
      "epoch": 13.149606299212598,
      "grad_norm": 0.16493800282478333,
      "learning_rate": 9.631017119366922e-05,
      "loss": 0.0107,
      "step": 1670
    },
    {
      "epoch": 13.228346456692913,
      "grad_norm": 0.18128368258476257,
      "learning_rate": 9.624757827999273e-05,
      "loss": 0.0117,
      "step": 1680
    },
    {
      "epoch": 13.307086614173228,
      "grad_norm": 0.2412339746952057,
      "learning_rate": 9.618447961020971e-05,
      "loss": 0.015,
      "step": 1690
    },
    {
      "epoch": 13.385826771653543,
      "grad_norm": 0.23829780519008636,
      "learning_rate": 9.612087587435707e-05,
      "loss": 0.0146,
      "step": 1700
    },
    {
      "epoch": 13.464566929133857,
      "grad_norm": 0.17435379326343536,
      "learning_rate": 9.605676776799508e-05,
      "loss": 0.012,
      "step": 1710
    },
    {
      "epoch": 13.543307086614174,
      "grad_norm": 0.26677370071411133,
      "learning_rate": 9.599215599219973e-05,
      "loss": 0.0119,
      "step": 1720
    },
    {
      "epoch": 13.622047244094489,
      "grad_norm": 0.17352107167243958,
      "learning_rate": 9.592704125355505e-05,
      "loss": 0.0119,
      "step": 1730
    },
    {
      "epoch": 13.700787401574804,
      "grad_norm": 0.1817910224199295,
      "learning_rate": 9.586142426414538e-05,
      "loss": 0.011,
      "step": 1740
    },
    {
      "epoch": 13.779527559055119,
      "grad_norm": 0.26779067516326904,
      "learning_rate": 9.57953057415476e-05,
      "loss": 0.0137,
      "step": 1750
    },
    {
      "epoch": 13.858267716535433,
      "grad_norm": 0.16992807388305664,
      "learning_rate": 9.572868640882328e-05,
      "loss": 0.0116,
      "step": 1760
    },
    {
      "epoch": 13.937007874015748,
      "grad_norm": 0.2475721836090088,
      "learning_rate": 9.56615669945108e-05,
      "loss": 0.013,
      "step": 1770
    },
    {
      "epoch": 14.015748031496063,
      "grad_norm": 0.30210572481155396,
      "learning_rate": 9.55939482326173e-05,
      "loss": 0.0124,
      "step": 1780
    },
    {
      "epoch": 14.094488188976378,
      "grad_norm": 0.19526968896389008,
      "learning_rate": 9.552583086261069e-05,
      "loss": 0.0135,
      "step": 1790
    },
    {
      "epoch": 14.173228346456693,
      "grad_norm": 0.1772489845752716,
      "learning_rate": 9.545721562941168e-05,
      "loss": 0.0119,
      "step": 1800
    },
    {
      "epoch": 14.251968503937007,
      "grad_norm": 0.20985430479049683,
      "learning_rate": 9.538810328338543e-05,
      "loss": 0.0106,
      "step": 1810
    },
    {
      "epoch": 14.330708661417322,
      "grad_norm": 0.2288864552974701,
      "learning_rate": 9.531849458033349e-05,
      "loss": 0.0121,
      "step": 1820
    },
    {
      "epoch": 14.409448818897637,
      "grad_norm": 0.14826878905296326,
      "learning_rate": 9.524839028148547e-05,
      "loss": 0.0109,
      "step": 1830
    },
    {
      "epoch": 14.488188976377952,
      "grad_norm": 0.24729447066783905,
      "learning_rate": 9.517779115349077e-05,
      "loss": 0.0122,
      "step": 1840
    },
    {
      "epoch": 14.566929133858268,
      "grad_norm": 0.23712359368801117,
      "learning_rate": 9.510669796841014e-05,
      "loss": 0.012,
      "step": 1850
    },
    {
      "epoch": 14.645669291338583,
      "grad_norm": 0.24393972754478455,
      "learning_rate": 9.503511150370727e-05,
      "loss": 0.014,
      "step": 1860
    },
    {
      "epoch": 14.724409448818898,
      "grad_norm": 0.16620883345603943,
      "learning_rate": 9.496303254224024e-05,
      "loss": 0.0124,
      "step": 1870
    },
    {
      "epoch": 14.803149606299213,
      "grad_norm": 0.19335606694221497,
      "learning_rate": 9.489046187225306e-05,
      "loss": 0.0116,
      "step": 1880
    },
    {
      "epoch": 14.881889763779528,
      "grad_norm": 0.17978379130363464,
      "learning_rate": 9.481740028736692e-05,
      "loss": 0.0127,
      "step": 1890
    },
    {
      "epoch": 14.960629921259843,
      "grad_norm": 0.15655072033405304,
      "learning_rate": 9.474384858657164e-05,
      "loss": 0.0112,
      "step": 1900
    },
    {
      "epoch": 15.039370078740157,
      "grad_norm": 0.13158245384693146,
      "learning_rate": 9.466980757421679e-05,
      "loss": 0.011,
      "step": 1910
    },
    {
      "epoch": 15.118110236220472,
      "grad_norm": 0.21858836710453033,
      "learning_rate": 9.459527806000305e-05,
      "loss": 0.0116,
      "step": 1920
    },
    {
      "epoch": 15.196850393700787,
      "grad_norm": 0.19522692263126373,
      "learning_rate": 9.452026085897325e-05,
      "loss": 0.0118,
      "step": 1930
    },
    {
      "epoch": 15.275590551181102,
      "grad_norm": 0.20890626311302185,
      "learning_rate": 9.444475679150348e-05,
      "loss": 0.0111,
      "step": 1940
    },
    {
      "epoch": 15.354330708661417,
      "grad_norm": 0.20746995508670807,
      "learning_rate": 9.436876668329411e-05,
      "loss": 0.0107,
      "step": 1950
    },
    {
      "epoch": 15.433070866141732,
      "grad_norm": 0.18878526985645294,
      "learning_rate": 9.429229136536079e-05,
      "loss": 0.0105,
      "step": 1960
    },
    {
      "epoch": 15.511811023622048,
      "grad_norm": 0.18786223232746124,
      "learning_rate": 9.421533167402534e-05,
      "loss": 0.0112,
      "step": 1970
    },
    {
      "epoch": 15.590551181102363,
      "grad_norm": 0.12698164582252502,
      "learning_rate": 9.413788845090666e-05,
      "loss": 0.011,
      "step": 1980
    },
    {
      "epoch": 15.669291338582678,
      "grad_norm": 0.22439169883728027,
      "learning_rate": 9.405996254291136e-05,
      "loss": 0.0113,
      "step": 1990
    },
    {
      "epoch": 15.748031496062993,
      "grad_norm": 0.16835476458072662,
      "learning_rate": 9.398155480222474e-05,
      "loss": 0.0111,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 79,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}