{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.90207715133531,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02967359050445104,
      "grad_norm": 1.9033336639404297,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.2317,
      "step": 10
    },
    {
      "epoch": 0.05934718100890208,
      "grad_norm": 1.0404284000396729,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.1977,
      "step": 20
    },
    {
      "epoch": 0.08902077151335312,
      "grad_norm": 0.6579734683036804,
      "learning_rate": 6e-06,
      "loss": 0.1451,
      "step": 30
    },
    {
      "epoch": 0.11869436201780416,
      "grad_norm": 0.33155006170272827,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0959,
      "step": 40
    },
    {
      "epoch": 0.14836795252225518,
      "grad_norm": 0.5317391753196716,
      "learning_rate": 1e-05,
      "loss": 0.0828,
      "step": 50
    },
    {
      "epoch": 0.17804154302670624,
      "grad_norm": 0.45179909467697144,
      "learning_rate": 1.2e-05,
      "loss": 0.0814,
      "step": 60
    },
    {
      "epoch": 0.20771513353115728,
      "grad_norm": 0.2707938849925995,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.0562,
      "step": 70
    },
    {
      "epoch": 0.23738872403560832,
      "grad_norm": 0.22402559220790863,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0594,
      "step": 80
    },
    {
      "epoch": 0.26706231454005935,
      "grad_norm": 0.16533811390399933,
      "learning_rate": 1.8e-05,
      "loss": 0.0529,
      "step": 90
    },
    {
      "epoch": 0.29673590504451036,
      "grad_norm": 0.222530797123909,
      "learning_rate": 2e-05,
      "loss": 0.0522,
      "step": 100
    },
    {
      "epoch": 0.3264094955489614,
      "grad_norm": 0.1894129067659378,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.0514,
      "step": 110
    },
    {
      "epoch": 0.3560830860534125,
      "grad_norm": 0.20559543371200562,
      "learning_rate": 2.4e-05,
      "loss": 0.0462,
      "step": 120
    },
    {
      "epoch": 0.3857566765578635,
      "grad_norm": 0.157830610871315,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.0471,
      "step": 130
    },
    {
      "epoch": 0.41543026706231456,
      "grad_norm": 0.14663924276828766,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.0449,
      "step": 140
    },
    {
      "epoch": 0.44510385756676557,
      "grad_norm": 0.14772620797157288,
      "learning_rate": 3e-05,
      "loss": 0.0424,
      "step": 150
    },
    {
      "epoch": 0.47477744807121663,
      "grad_norm": 0.16058433055877686,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.0424,
      "step": 160
    },
    {
      "epoch": 0.5044510385756676,
      "grad_norm": 0.15857172012329102,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.041,
      "step": 170
    },
    {
      "epoch": 0.5341246290801187,
      "grad_norm": 0.17435680329799652,
      "learning_rate": 3.6e-05,
      "loss": 0.0408,
      "step": 180
    },
    {
      "epoch": 0.5637982195845698,
      "grad_norm": 0.1439993977546692,
      "learning_rate": 3.8e-05,
      "loss": 0.0352,
      "step": 190
    },
    {
      "epoch": 0.5934718100890207,
      "grad_norm": 0.15629075467586517,
      "learning_rate": 4e-05,
      "loss": 0.0383,
      "step": 200
    },
    {
      "epoch": 0.6231454005934718,
      "grad_norm": 0.1610369235277176,
      "learning_rate": 4.2e-05,
      "loss": 0.0392,
      "step": 210
    },
    {
      "epoch": 0.6528189910979229,
      "grad_norm": 0.17589861154556274,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0374,
      "step": 220
    },
    {
      "epoch": 0.6824925816023739,
      "grad_norm": 0.19186066091060638,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.0358,
      "step": 230
    },
    {
      "epoch": 0.712166172106825,
      "grad_norm": 0.1579175740480423,
      "learning_rate": 4.8e-05,
      "loss": 0.0357,
      "step": 240
    },
    {
      "epoch": 0.7418397626112759,
      "grad_norm": 0.17220136523246765,
      "learning_rate": 5e-05,
      "loss": 0.0334,
      "step": 250
    },
    {
      "epoch": 0.771513353115727,
      "grad_norm": 0.18591266870498657,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 0.0315,
      "step": 260
    },
    {
      "epoch": 0.8011869436201781,
      "grad_norm": 0.2341579794883728,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.0375,
      "step": 270
    },
    {
      "epoch": 0.8308605341246291,
      "grad_norm": 0.15227168798446655,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 0.031,
      "step": 280
    },
    {
      "epoch": 0.8605341246290801,
      "grad_norm": 0.1876339167356491,
      "learning_rate": 5.8e-05,
      "loss": 0.0371,
      "step": 290
    },
    {
      "epoch": 0.8902077151335311,
      "grad_norm": 0.1789393573999405,
      "learning_rate": 6e-05,
      "loss": 0.0313,
      "step": 300
    },
    {
      "epoch": 0.9198813056379822,
      "grad_norm": 0.1678636074066162,
      "learning_rate": 6.2e-05,
      "loss": 0.0349,
      "step": 310
    },
    {
      "epoch": 0.9495548961424333,
      "grad_norm": 0.17457032203674316,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.0296,
      "step": 320
    },
    {
      "epoch": 0.9792284866468842,
      "grad_norm": 0.14290577173233032,
      "learning_rate": 6.6e-05,
      "loss": 0.0308,
      "step": 330
    },
    {
      "epoch": 1.0089020771513353,
      "grad_norm": 0.23601128160953522,
      "learning_rate": 6.800000000000001e-05,
      "loss": 0.0298,
      "step": 340
    },
    {
      "epoch": 1.0385756676557865,
      "grad_norm": 0.14039042592048645,
      "learning_rate": 7e-05,
      "loss": 0.0262,
      "step": 350
    },
    {
      "epoch": 1.0682492581602374,
      "grad_norm": 0.1804966777563095,
      "learning_rate": 7.2e-05,
      "loss": 0.0284,
      "step": 360
    },
    {
      "epoch": 1.0979228486646884,
      "grad_norm": 0.22986947000026703,
      "learning_rate": 7.4e-05,
      "loss": 0.0308,
      "step": 370
    },
    {
      "epoch": 1.1275964391691395,
      "grad_norm": 0.20188020169734955,
      "learning_rate": 7.6e-05,
      "loss": 0.0261,
      "step": 380
    },
    {
      "epoch": 1.1572700296735905,
      "grad_norm": 0.14067409932613373,
      "learning_rate": 7.800000000000001e-05,
      "loss": 0.028,
      "step": 390
    },
    {
      "epoch": 1.1869436201780414,
      "grad_norm": 0.16516339778900146,
      "learning_rate": 8e-05,
      "loss": 0.0247,
      "step": 400
    },
    {
      "epoch": 1.2166172106824926,
      "grad_norm": 0.19918474555015564,
      "learning_rate": 8.2e-05,
      "loss": 0.0301,
      "step": 410
    },
    {
      "epoch": 1.2462908011869436,
      "grad_norm": 0.1878385990858078,
      "learning_rate": 8.4e-05,
      "loss": 0.0251,
      "step": 420
    },
    {
      "epoch": 1.2759643916913945,
      "grad_norm": 0.20107118785381317,
      "learning_rate": 8.6e-05,
      "loss": 0.0279,
      "step": 430
    },
    {
      "epoch": 1.3056379821958457,
      "grad_norm": 0.24616649746894836,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.0259,
      "step": 440
    },
    {
      "epoch": 1.3353115727002967,
      "grad_norm": 0.19029636681079865,
      "learning_rate": 9e-05,
      "loss": 0.0262,
      "step": 450
    },
    {
      "epoch": 1.3649851632047478,
      "grad_norm": 0.194508358836174,
      "learning_rate": 9.200000000000001e-05,
      "loss": 0.0275,
      "step": 460
    },
    {
      "epoch": 1.3946587537091988,
      "grad_norm": 0.20826251804828644,
      "learning_rate": 9.4e-05,
      "loss": 0.0289,
      "step": 470
    },
    {
      "epoch": 1.4243323442136497,
      "grad_norm": 0.13222843408584595,
      "learning_rate": 9.6e-05,
      "loss": 0.0249,
      "step": 480
    },
    {
      "epoch": 1.454005934718101,
      "grad_norm": 0.13967235386371613,
      "learning_rate": 9.8e-05,
      "loss": 0.0231,
      "step": 490
    },
    {
      "epoch": 1.4836795252225519,
      "grad_norm": 0.21556402742862701,
      "learning_rate": 0.0001,
      "loss": 0.0232,
      "step": 500
    },
    {
      "epoch": 1.513353115727003,
      "grad_norm": 0.2407234013080597,
      "learning_rate": 9.999972660400536e-05,
      "loss": 0.025,
      "step": 510
    },
    {
      "epoch": 1.543026706231454,
      "grad_norm": 0.1544090360403061,
      "learning_rate": 9.999890641901125e-05,
      "loss": 0.0224,
      "step": 520
    },
    {
      "epoch": 1.572700296735905,
      "grad_norm": 0.1930345594882965,
      "learning_rate": 9.999753945398704e-05,
      "loss": 0.0244,
      "step": 530
    },
    {
      "epoch": 1.6023738872403561,
      "grad_norm": 0.2288358211517334,
      "learning_rate": 9.99956257238817e-05,
      "loss": 0.0223,
      "step": 540
    },
    {
      "epoch": 1.632047477744807,
      "grad_norm": 0.2028588354587555,
      "learning_rate": 9.999316524962345e-05,
      "loss": 0.022,
      "step": 550
    },
    {
      "epoch": 1.6617210682492582,
      "grad_norm": 0.17989283800125122,
      "learning_rate": 9.999015805811965e-05,
      "loss": 0.0201,
      "step": 560
    },
    {
      "epoch": 1.6913946587537092,
      "grad_norm": 0.20576386153697968,
      "learning_rate": 9.998660418225645e-05,
      "loss": 0.0211,
      "step": 570
    },
    {
      "epoch": 1.7210682492581602,
      "grad_norm": 0.1263924539089203,
      "learning_rate": 9.998250366089848e-05,
      "loss": 0.0205,
      "step": 580
    },
    {
      "epoch": 1.7507418397626113,
      "grad_norm": 0.23239193856716156,
      "learning_rate": 9.997785653888835e-05,
      "loss": 0.0223,
      "step": 590
    },
    {
      "epoch": 1.7804154302670623,
      "grad_norm": 0.15964201092720032,
      "learning_rate": 9.997266286704631e-05,
      "loss": 0.0236,
      "step": 600
    },
    {
      "epoch": 1.8100890207715135,
      "grad_norm": 0.1998339742422104,
      "learning_rate": 9.996692270216947e-05,
      "loss": 0.0254,
      "step": 610
    },
    {
      "epoch": 1.8397626112759644,
      "grad_norm": 0.25945162773132324,
      "learning_rate": 9.996063610703137e-05,
      "loss": 0.0207,
      "step": 620
    },
    {
      "epoch": 1.8694362017804154,
      "grad_norm": 0.1687825620174408,
      "learning_rate": 9.995380315038119e-05,
      "loss": 0.0169,
      "step": 630
    },
    {
      "epoch": 1.8991097922848663,
      "grad_norm": 0.13454881310462952,
      "learning_rate": 9.994642390694308e-05,
      "loss": 0.0185,
      "step": 640
    },
    {
      "epoch": 1.9287833827893175,
      "grad_norm": 0.1880808174610138,
      "learning_rate": 9.993849845741524e-05,
      "loss": 0.0199,
      "step": 650
    },
    {
      "epoch": 1.9584569732937687,
      "grad_norm": 0.17725640535354614,
      "learning_rate": 9.993002688846913e-05,
      "loss": 0.019,
      "step": 660
    },
    {
      "epoch": 1.9881305637982196,
      "grad_norm": 0.17680853605270386,
      "learning_rate": 9.992100929274846e-05,
      "loss": 0.0257,
      "step": 670
    },
    {
      "epoch": 2.0178041543026706,
      "grad_norm": 0.18133607506752014,
      "learning_rate": 9.991144576886823e-05,
      "loss": 0.0222,
      "step": 680
    },
    {
      "epoch": 2.0474777448071215,
      "grad_norm": 0.21174193918704987,
      "learning_rate": 9.990133642141359e-05,
      "loss": 0.0192,
      "step": 690
    },
    {
      "epoch": 2.077151335311573,
      "grad_norm": 0.15766288340091705,
      "learning_rate": 9.989068136093873e-05,
      "loss": 0.0199,
      "step": 700
    },
    {
      "epoch": 2.106824925816024,
      "grad_norm": 0.19250448048114777,
      "learning_rate": 9.987948070396571e-05,
      "loss": 0.0252,
      "step": 710
    },
    {
      "epoch": 2.136498516320475,
      "grad_norm": 0.1412709653377533,
      "learning_rate": 9.986773457298311e-05,
      "loss": 0.0186,
      "step": 720
    },
    {
      "epoch": 2.166172106824926,
      "grad_norm": 0.2136259824037552,
      "learning_rate": 9.985544309644475e-05,
      "loss": 0.0204,
      "step": 730
    },
    {
      "epoch": 2.1958456973293767,
      "grad_norm": 0.10815251618623734,
      "learning_rate": 9.984260640876821e-05,
      "loss": 0.0153,
      "step": 740
    },
    {
      "epoch": 2.2255192878338277,
      "grad_norm": 0.14663298428058624,
      "learning_rate": 9.98292246503335e-05,
      "loss": 0.0189,
      "step": 750
    },
    {
      "epoch": 2.255192878338279,
      "grad_norm": 0.18529076874256134,
      "learning_rate": 9.981529796748134e-05,
      "loss": 0.0155,
      "step": 760
    },
    {
      "epoch": 2.28486646884273,
      "grad_norm": 0.2349974811077118,
      "learning_rate": 9.980082651251175e-05,
      "loss": 0.0172,
      "step": 770
    },
    {
      "epoch": 2.314540059347181,
      "grad_norm": 0.10778886079788208,
      "learning_rate": 9.97858104436822e-05,
      "loss": 0.0161,
      "step": 780
    },
    {
      "epoch": 2.344213649851632,
      "grad_norm": 0.15675969421863556,
      "learning_rate": 9.977024992520602e-05,
      "loss": 0.0165,
      "step": 790
    },
    {
      "epoch": 2.373887240356083,
      "grad_norm": 0.23468513786792755,
      "learning_rate": 9.975414512725057e-05,
      "loss": 0.0196,
      "step": 800
    },
    {
      "epoch": 2.4035608308605343,
      "grad_norm": 0.1332869678735733,
      "learning_rate": 9.973749622593534e-05,
      "loss": 0.0193,
      "step": 810
    },
    {
      "epoch": 2.4332344213649852,
      "grad_norm": 0.1406887024641037,
      "learning_rate": 9.972030340333001e-05,
      "loss": 0.0186,
      "step": 820
    },
    {
      "epoch": 2.462908011869436,
      "grad_norm": 0.11544730514287949,
      "learning_rate": 9.970256684745258e-05,
      "loss": 0.0195,
      "step": 830
    },
    {
      "epoch": 2.492581602373887,
      "grad_norm": 0.19476240873336792,
      "learning_rate": 9.968428675226714e-05,
      "loss": 0.0171,
      "step": 840
    },
    {
      "epoch": 2.5222551928783385,
      "grad_norm": 0.22309833765029907,
      "learning_rate": 9.966546331768191e-05,
      "loss": 0.0223,
      "step": 850
    },
    {
      "epoch": 2.551928783382789,
      "grad_norm": 0.2214643657207489,
      "learning_rate": 9.964609674954696e-05,
      "loss": 0.019,
      "step": 860
    },
    {
      "epoch": 2.5816023738872405,
      "grad_norm": 0.2298765480518341,
      "learning_rate": 9.962618725965196e-05,
      "loss": 0.0178,
      "step": 870
    },
    {
      "epoch": 2.6112759643916914,
      "grad_norm": 0.1542595475912094,
      "learning_rate": 9.96057350657239e-05,
      "loss": 0.0152,
      "step": 880
    },
    {
      "epoch": 2.6409495548961424,
      "grad_norm": 0.12994691729545593,
      "learning_rate": 9.95847403914247e-05,
      "loss": 0.0153,
      "step": 890
    },
    {
      "epoch": 2.6706231454005933,
      "grad_norm": 0.1726643294095993,
      "learning_rate": 9.956320346634876e-05,
      "loss": 0.0219,
      "step": 900
    },
    {
      "epoch": 2.7002967359050443,
      "grad_norm": 0.18292242288589478,
      "learning_rate": 9.954112452602045e-05,
      "loss": 0.0137,
      "step": 910
    },
    {
      "epoch": 2.7299703264094957,
      "grad_norm": 0.19749417901039124,
      "learning_rate": 9.95185038118915e-05,
      "loss": 0.0179,
      "step": 920
    },
    {
      "epoch": 2.7596439169139466,
      "grad_norm": 0.17808304727077484,
      "learning_rate": 9.949534157133844e-05,
      "loss": 0.0155,
      "step": 930
    },
    {
      "epoch": 2.7893175074183976,
      "grad_norm": 0.13589969277381897,
      "learning_rate": 9.94716380576598e-05,
      "loss": 0.0143,
      "step": 940
    },
    {
      "epoch": 2.8189910979228485,
      "grad_norm": 0.17047159373760223,
      "learning_rate": 9.944739353007344e-05,
      "loss": 0.0211,
      "step": 950
    },
    {
      "epoch": 2.8486646884272995,
      "grad_norm": 0.15535619854927063,
      "learning_rate": 9.942260825371358e-05,
      "loss": 0.0132,
      "step": 960
    },
    {
      "epoch": 2.878338278931751,
      "grad_norm": 0.1609240472316742,
      "learning_rate": 9.939728249962807e-05,
      "loss": 0.0145,
      "step": 970
    },
    {
      "epoch": 2.908011869436202,
      "grad_norm": 0.20487530529499054,
      "learning_rate": 9.937141654477528e-05,
      "loss": 0.0172,
      "step": 980
    },
    {
      "epoch": 2.9376854599406528,
      "grad_norm": 0.12371553480625153,
      "learning_rate": 9.934501067202117e-05,
      "loss": 0.0191,
      "step": 990
    },
    {
      "epoch": 2.9673590504451037,
      "grad_norm": 0.15513016283512115,
      "learning_rate": 9.931806517013612e-05,
      "loss": 0.0158,
      "step": 1000
    },
    {
      "epoch": 2.9970326409495547,
      "grad_norm": 0.1723584234714508,
      "learning_rate": 9.929058033379181e-05,
      "loss": 0.0139,
      "step": 1010
    },
    {
      "epoch": 3.026706231454006,
      "grad_norm": 0.1482209414243698,
      "learning_rate": 9.926255646355804e-05,
      "loss": 0.0148,
      "step": 1020
    },
    {
      "epoch": 3.056379821958457,
      "grad_norm": 0.15559454262256622,
      "learning_rate": 9.923399386589933e-05,
      "loss": 0.0147,
      "step": 1030
    },
    {
      "epoch": 3.086053412462908,
      "grad_norm": 0.22350917756557465,
      "learning_rate": 9.92048928531717e-05,
      "loss": 0.0157,
      "step": 1040
    },
    {
      "epoch": 3.115727002967359,
      "grad_norm": 0.1581738144159317,
      "learning_rate": 9.917525374361912e-05,
      "loss": 0.0133,
      "step": 1050
    },
    {
      "epoch": 3.14540059347181,
      "grad_norm": 0.20087914168834686,
      "learning_rate": 9.914507686137019e-05,
      "loss": 0.0208,
      "step": 1060
    },
    {
      "epoch": 3.1750741839762613,
      "grad_norm": 0.16003265976905823,
      "learning_rate": 9.911436253643445e-05,
      "loss": 0.0168,
      "step": 1070
    },
    {
      "epoch": 3.2047477744807122,
      "grad_norm": 0.15235169231891632,
      "learning_rate": 9.90831111046988e-05,
      "loss": 0.0168,
      "step": 1080
    },
    {
      "epoch": 3.234421364985163,
      "grad_norm": 0.15660665929317474,
      "learning_rate": 9.905132290792394e-05,
      "loss": 0.0106,
      "step": 1090
    },
    {
      "epoch": 3.264094955489614,
      "grad_norm": 0.16853424906730652,
      "learning_rate": 9.901899829374047e-05,
      "loss": 0.0149,
      "step": 1100
    },
    {
      "epoch": 3.293768545994065,
      "grad_norm": 0.1335846185684204,
      "learning_rate": 9.89861376156452e-05,
      "loss": 0.0168,
      "step": 1110
    },
    {
      "epoch": 3.3234421364985165,
      "grad_norm": 0.20238997042179108,
      "learning_rate": 9.895274123299723e-05,
      "loss": 0.0154,
      "step": 1120
    },
    {
      "epoch": 3.3531157270029674,
      "grad_norm": 0.22216491401195526,
      "learning_rate": 9.891880951101407e-05,
      "loss": 0.019,
      "step": 1130
    },
    {
      "epoch": 3.3827893175074184,
      "grad_norm": 0.2017626017332077,
      "learning_rate": 9.888434282076758e-05,
      "loss": 0.017,
      "step": 1140
    },
    {
      "epoch": 3.4124629080118694,
      "grad_norm": 0.18049117922782898,
      "learning_rate": 9.884934153917997e-05,
      "loss": 0.0163,
      "step": 1150
    },
    {
      "epoch": 3.4421364985163203,
      "grad_norm": 0.28145721554756165,
      "learning_rate": 9.881380604901964e-05,
      "loss": 0.0166,
      "step": 1160
    },
    {
      "epoch": 3.4718100890207717,
      "grad_norm": 0.2356300801038742,
      "learning_rate": 9.877773673889701e-05,
      "loss": 0.0155,
      "step": 1170
    },
    {
      "epoch": 3.5014836795252227,
      "grad_norm": 0.24113395810127258,
      "learning_rate": 9.87411340032603e-05,
      "loss": 0.0156,
      "step": 1180
    },
    {
      "epoch": 3.5311572700296736,
      "grad_norm": 0.18665863573551178,
      "learning_rate": 9.870399824239117e-05,
      "loss": 0.0159,
      "step": 1190
    },
    {
      "epoch": 3.5608308605341246,
      "grad_norm": 0.16171567142009735,
      "learning_rate": 9.86663298624003e-05,
      "loss": 0.0167,
      "step": 1200
    },
    {
      "epoch": 3.5905044510385755,
      "grad_norm": 0.17315839231014252,
      "learning_rate": 9.862812927522309e-05,
      "loss": 0.017,
      "step": 1210
    },
    {
      "epoch": 3.620178041543027,
      "grad_norm": 0.14727933704853058,
      "learning_rate": 9.858939689861506e-05,
      "loss": 0.0132,
      "step": 1220
    },
    {
      "epoch": 3.649851632047478,
      "grad_norm": 0.1552547812461853,
      "learning_rate": 9.855013315614725e-05,
      "loss": 0.0175,
      "step": 1230
    },
    {
      "epoch": 3.679525222551929,
      "grad_norm": 0.1715100109577179,
      "learning_rate": 9.851033847720166e-05,
      "loss": 0.0139,
      "step": 1240
    },
    {
      "epoch": 3.7091988130563798,
      "grad_norm": 0.16414763033390045,
      "learning_rate": 9.847001329696653e-05,
      "loss": 0.0189,
      "step": 1250
    },
    {
      "epoch": 3.7388724035608307,
      "grad_norm": 0.1251063048839569,
      "learning_rate": 9.842915805643155e-05,
      "loss": 0.0145,
      "step": 1260
    },
    {
      "epoch": 3.768545994065282,
      "grad_norm": 0.17011059820652008,
      "learning_rate": 9.838777320238312e-05,
      "loss": 0.0161,
      "step": 1270
    },
    {
      "epoch": 3.798219584569733,
      "grad_norm": 0.14429537951946259,
      "learning_rate": 9.834585918739936e-05,
      "loss": 0.0159,
      "step": 1280
    },
    {
      "epoch": 3.827893175074184,
      "grad_norm": 0.138567715883255,
      "learning_rate": 9.830341646984521e-05,
      "loss": 0.0175,
      "step": 1290
    },
    {
      "epoch": 3.857566765578635,
      "grad_norm": 0.08295896649360657,
      "learning_rate": 9.826044551386744e-05,
      "loss": 0.0145,
      "step": 1300
    },
    {
      "epoch": 3.887240356083086,
      "grad_norm": 0.0911448523402214,
      "learning_rate": 9.821694678938953e-05,
      "loss": 0.0134,
      "step": 1310
    },
    {
      "epoch": 3.9169139465875373,
      "grad_norm": 0.14157798886299133,
      "learning_rate": 9.817292077210659e-05,
      "loss": 0.0158,
      "step": 1320
    },
    {
      "epoch": 3.9465875370919883,
      "grad_norm": 0.17415288090705872,
      "learning_rate": 9.812836794348004e-05,
      "loss": 0.0125,
      "step": 1330
    },
    {
      "epoch": 3.9762611275964392,
      "grad_norm": 0.22007249295711517,
      "learning_rate": 9.808328879073251e-05,
      "loss": 0.0153,
      "step": 1340
    },
    {
      "epoch": 4.005934718100891,
      "grad_norm": 0.144961416721344,
      "learning_rate": 9.803768380684242e-05,
      "loss": 0.0117,
      "step": 1350
    },
    {
      "epoch": 4.035608308605341,
      "grad_norm": 0.14304885268211365,
      "learning_rate": 9.799155349053851e-05,
      "loss": 0.0138,
      "step": 1360
    },
    {
      "epoch": 4.0652818991097925,
      "grad_norm": 0.18843571841716766,
      "learning_rate": 9.794489834629455e-05,
      "loss": 0.0106,
      "step": 1370
    },
    {
      "epoch": 4.094955489614243,
      "grad_norm": 0.16858817636966705,
      "learning_rate": 9.789771888432375e-05,
      "loss": 0.0138,
      "step": 1380
    },
    {
      "epoch": 4.1246290801186944,
      "grad_norm": 0.19177594780921936,
      "learning_rate": 9.785001562057309e-05,
      "loss": 0.0114,
      "step": 1390
    },
    {
      "epoch": 4.154302670623146,
      "grad_norm": 0.20158767700195312,
      "learning_rate": 9.780178907671789e-05,
      "loss": 0.0146,
      "step": 1400
    },
    {
      "epoch": 4.183976261127596,
      "grad_norm": 0.17675232887268066,
      "learning_rate": 9.775303978015585e-05,
      "loss": 0.0116,
      "step": 1410
    },
    {
      "epoch": 4.213649851632048,
      "grad_norm": 0.20077385008335114,
      "learning_rate": 9.77037682640015e-05,
      "loss": 0.0172,
      "step": 1420
    },
    {
      "epoch": 4.243323442136498,
      "grad_norm": 0.17185665667057037,
      "learning_rate": 9.765397506708023e-05,
      "loss": 0.0138,
      "step": 1430
    },
    {
      "epoch": 4.27299703264095,
      "grad_norm": 0.1641971468925476,
      "learning_rate": 9.760366073392246e-05,
      "loss": 0.0145,
      "step": 1440
    },
    {
      "epoch": 4.302670623145401,
      "grad_norm": 0.13757102191448212,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.0158,
      "step": 1450
    },
    {
      "epoch": 4.332344213649852,
      "grad_norm": 0.18012432754039764,
      "learning_rate": 9.750147086550844e-05,
      "loss": 0.0139,
      "step": 1460
    },
    {
      "epoch": 4.362017804154303,
      "grad_norm": 0.1639927178621292,
      "learning_rate": 9.744959644778422e-05,
      "loss": 0.0142,
      "step": 1470
    },
    {
      "epoch": 4.3916913946587535,
      "grad_norm": 0.17392724752426147,
      "learning_rate": 9.739720312887535e-05,
      "loss": 0.018,
      "step": 1480
    },
    {
      "epoch": 4.421364985163205,
      "grad_norm": 0.15744361281394958,
      "learning_rate": 9.734429148174675e-05,
      "loss": 0.0128,
      "step": 1490
    },
    {
      "epoch": 4.451038575667655,
      "grad_norm": 0.12954673171043396,
      "learning_rate": 9.729086208503174e-05,
      "loss": 0.0188,
      "step": 1500
    },
    {
      "epoch": 4.480712166172107,
      "grad_norm": 0.16809141635894775,
      "learning_rate": 9.723691552302562e-05,
      "loss": 0.0144,
      "step": 1510
    },
    {
      "epoch": 4.510385756676558,
      "grad_norm": 0.1523902416229248,
      "learning_rate": 9.718245238567939e-05,
      "loss": 0.0145,
      "step": 1520
    },
    {
      "epoch": 4.540059347181009,
      "grad_norm": 0.26666077971458435,
      "learning_rate": 9.712747326859315e-05,
      "loss": 0.015,
      "step": 1530
    },
    {
      "epoch": 4.56973293768546,
      "grad_norm": 0.193909153342247,
      "learning_rate": 9.707197877300974e-05,
      "loss": 0.0167,
      "step": 1540
    },
    {
      "epoch": 4.5994065281899115,
      "grad_norm": 0.17812030017375946,
      "learning_rate": 9.701596950580806e-05,
      "loss": 0.0157,
      "step": 1550
    },
    {
      "epoch": 4.629080118694362,
      "grad_norm": 0.26170411705970764,
      "learning_rate": 9.695944607949649e-05,
      "loss": 0.0139,
      "step": 1560
    },
    {
      "epoch": 4.658753709198813,
      "grad_norm": 0.14579689502716064,
      "learning_rate": 9.690240911220618e-05,
      "loss": 0.013,
      "step": 1570
    },
    {
      "epoch": 4.688427299703264,
      "grad_norm": 0.17479683458805084,
      "learning_rate": 9.684485922768422e-05,
      "loss": 0.014,
      "step": 1580
    },
    {
      "epoch": 4.718100890207715,
      "grad_norm": 0.15623094141483307,
      "learning_rate": 9.6786797055287e-05,
      "loss": 0.0151,
      "step": 1590
    },
    {
      "epoch": 4.747774480712166,
      "grad_norm": 0.1983657032251358,
      "learning_rate": 9.672822322997305e-05,
      "loss": 0.0138,
      "step": 1600
    },
    {
      "epoch": 4.777448071216617,
      "grad_norm": 0.13670101761817932,
      "learning_rate": 9.66691383922964e-05,
      "loss": 0.0114,
      "step": 1610
    },
    {
      "epoch": 4.807121661721069,
      "grad_norm": 0.13849210739135742,
      "learning_rate": 9.660954318839933e-05,
      "loss": 0.0105,
      "step": 1620
    },
    {
      "epoch": 4.836795252225519,
      "grad_norm": 0.13254909217357635,
      "learning_rate": 9.654943827000548e-05,
      "loss": 0.0113,
      "step": 1630
    },
    {
      "epoch": 4.8664688427299705,
      "grad_norm": 0.12954489886760712,
      "learning_rate": 9.648882429441257e-05,
      "loss": 0.0133,
      "step": 1640
    },
    {
      "epoch": 4.896142433234421,
      "grad_norm": 0.13290920853614807,
      "learning_rate": 9.642770192448536e-05,
      "loss": 0.0156,
      "step": 1650
    },
    {
      "epoch": 4.925816023738872,
      "grad_norm": 0.13683238625526428,
      "learning_rate": 9.636607182864827e-05,
      "loss": 0.0096,
      "step": 1660
    },
    {
      "epoch": 4.955489614243324,
      "grad_norm": 0.07962514460086823,
      "learning_rate": 9.630393468087818e-05,
      "loss": 0.0128,
      "step": 1670
    },
    {
      "epoch": 4.985163204747774,
      "grad_norm": 0.15497808158397675,
      "learning_rate": 9.624129116069694e-05,
      "loss": 0.014,
      "step": 1680
    },
    {
      "epoch": 5.014836795252226,
      "grad_norm": 0.1419367492198944,
      "learning_rate": 9.617814195316411e-05,
      "loss": 0.0132,
      "step": 1690
    },
    {
      "epoch": 5.044510385756676,
      "grad_norm": 0.2201174795627594,
      "learning_rate": 9.611448774886924e-05,
      "loss": 0.0133,
      "step": 1700
    },
    {
      "epoch": 5.074183976261128,
      "grad_norm": 0.16477946937084198,
      "learning_rate": 9.605032924392457e-05,
      "loss": 0.0131,
      "step": 1710
    },
    {
      "epoch": 5.103857566765579,
      "grad_norm": 0.19834354519844055,
      "learning_rate": 9.598566713995718e-05,
      "loss": 0.0154,
      "step": 1720
    },
    {
      "epoch": 5.1335311572700295,
      "grad_norm": 0.22880475223064423,
      "learning_rate": 9.59205021441015e-05,
      "loss": 0.014,
      "step": 1730
    },
    {
      "epoch": 5.163204747774481,
      "grad_norm": 0.16253937780857086,
      "learning_rate": 9.58548349689915e-05,
      "loss": 0.0122,
      "step": 1740
    },
    {
      "epoch": 5.192878338278931,
      "grad_norm": 0.22121521830558777,
      "learning_rate": 9.578866633275288e-05,
      "loss": 0.015,
      "step": 1750
    },
    {
      "epoch": 5.222551928783383,
      "grad_norm": 0.182882621884346,
      "learning_rate": 9.572199695899522e-05,
      "loss": 0.0163,
      "step": 1760
    },
    {
      "epoch": 5.252225519287834,
      "grad_norm": 0.17466451227664948,
      "learning_rate": 9.565482757680415e-05,
      "loss": 0.0145,
      "step": 1770
    },
    {
      "epoch": 5.281899109792285,
      "grad_norm": 0.12321746349334717,
      "learning_rate": 9.558715892073323e-05,
      "loss": 0.0127,
      "step": 1780
    },
    {
      "epoch": 5.311572700296736,
      "grad_norm": 0.1869288980960846,
      "learning_rate": 9.551899173079607e-05,
      "loss": 0.0185,
      "step": 1790
    },
    {
      "epoch": 5.341246290801187,
      "grad_norm": 0.1715238392353058,
      "learning_rate": 9.545032675245813e-05,
      "loss": 0.0136,
      "step": 1800
    },
    {
      "epoch": 5.370919881305638,
      "grad_norm": 0.20112700760364532,
      "learning_rate": 9.538116473662861e-05,
      "loss": 0.013,
      "step": 1810
    },
    {
      "epoch": 5.400593471810089,
      "grad_norm": 0.13722355663776398,
      "learning_rate": 9.531150643965223e-05,
      "loss": 0.0112,
      "step": 1820
    },
    {
      "epoch": 5.43026706231454,
      "grad_norm": 0.2131308764219284,
      "learning_rate": 9.524135262330098e-05,
      "loss": 0.012,
      "step": 1830
    },
    {
      "epoch": 5.459940652818991,
      "grad_norm": 0.13212528824806213,
      "learning_rate": 9.517070405476575e-05,
      "loss": 0.0112,
      "step": 1840
    },
    {
      "epoch": 5.489614243323442,
      "grad_norm": 0.1769159734249115,
      "learning_rate": 9.509956150664796e-05,
      "loss": 0.0122,
      "step": 1850
    },
    {
      "epoch": 5.519287833827893,
      "grad_norm": 0.15795955061912537,
      "learning_rate": 9.502792575695112e-05,
      "loss": 0.0152,
      "step": 1860
    },
    {
      "epoch": 5.548961424332344,
      "grad_norm": 0.1461247205734253,
      "learning_rate": 9.49557975890723e-05,
      "loss": 0.0138,
      "step": 1870
    },
    {
      "epoch": 5.578635014836795,
      "grad_norm": 0.12148546427488327,
      "learning_rate": 9.488317779179361e-05,
      "loss": 0.0105,
      "step": 1880
    },
    {
      "epoch": 5.6083086053412465,
      "grad_norm": 0.17779065668582916,
      "learning_rate": 9.481006715927351e-05,
      "loss": 0.0107,
      "step": 1890
    },
    {
      "epoch": 5.637982195845697,
      "grad_norm": 0.17717322707176208,
      "learning_rate": 9.473646649103818e-05,
      "loss": 0.0158,
      "step": 1900
    },
    {
      "epoch": 5.667655786350148,
      "grad_norm": 0.19820888340473175,
      "learning_rate": 9.46623765919727e-05,
      "loss": 0.0137,
      "step": 1910
    },
    {
      "epoch": 5.697329376854599,
      "grad_norm": 0.15721198916435242,
      "learning_rate": 9.458779827231237e-05,
      "loss": 0.0119,
      "step": 1920
    },
    {
      "epoch": 5.72700296735905,
      "grad_norm": 0.17158320546150208,
      "learning_rate": 9.451273234763371e-05,
      "loss": 0.0133,
      "step": 1930
    },
    {
      "epoch": 5.756676557863502,
      "grad_norm": 0.15745575726032257,
      "learning_rate": 9.443717963884569e-05,
      "loss": 0.0129,
      "step": 1940
    },
    {
      "epoch": 5.786350148367952,
      "grad_norm": 0.17795485258102417,
      "learning_rate": 9.43611409721806e-05,
      "loss": 0.013,
      "step": 1950
    },
    {
      "epoch": 5.816023738872404,
      "grad_norm": 0.18350425362586975,
      "learning_rate": 9.428461717918511e-05,
      "loss": 0.0103,
      "step": 1960
    },
    {
      "epoch": 5.845697329376854,
      "grad_norm": 0.1675357222557068,
      "learning_rate": 9.420760909671118e-05,
      "loss": 0.0149,
      "step": 1970
    },
    {
      "epoch": 5.8753709198813056,
      "grad_norm": 0.1395285427570343,
      "learning_rate": 9.413011756690685e-05,
      "loss": 0.0174,
      "step": 1980
    },
    {
      "epoch": 5.905044510385757,
      "grad_norm": 0.18694210052490234,
      "learning_rate": 9.405214343720707e-05,
      "loss": 0.011,
      "step": 1990
    },
    {
      "epoch": 5.9347181008902075,
      "grad_norm": 0.13509497046470642,
      "learning_rate": 9.397368756032445e-05,
      "loss": 0.0105,
      "step": 2000
    },
    {
      "epoch": 5.964391691394659,
      "grad_norm": 0.1707238107919693,
      "learning_rate": 9.389475079423988e-05,
      "loss": 0.0099,
      "step": 2010
    },
    {
      "epoch": 5.994065281899109,
      "grad_norm": 0.18797723948955536,
      "learning_rate": 9.381533400219318e-05,
      "loss": 0.0144,
      "step": 2020
    },
    {
      "epoch": 6.023738872403561,
      "grad_norm": 0.15263915061950684,
      "learning_rate": 9.373543805267368e-05,
      "loss": 0.0145,
      "step": 2030
    },
    {
      "epoch": 6.053412462908012,
      "grad_norm": 0.10786967724561691,
      "learning_rate": 9.365506381941066e-05,
      "loss": 0.0167,
      "step": 2040
    },
    {
      "epoch": 6.083086053412463,
      "grad_norm": 0.10059456527233124,
      "learning_rate": 9.357421218136386e-05,
      "loss": 0.0127,
      "step": 2050
    },
    {
      "epoch": 6.112759643916914,
      "grad_norm": 0.12447630614042282,
      "learning_rate": 9.349288402271388e-05,
      "loss": 0.0109,
      "step": 2060
    },
    {
      "epoch": 6.142433234421365,
      "grad_norm": 0.14649389684200287,
      "learning_rate": 9.341108023285238e-05,
      "loss": 0.0115,
      "step": 2070
    },
    {
      "epoch": 6.172106824925816,
      "grad_norm": 0.09072308987379074,
      "learning_rate": 9.332880170637252e-05,
      "loss": 0.0097,
      "step": 2080
    },
    {
      "epoch": 6.201780415430267,
      "grad_norm": 0.1700124889612198,
      "learning_rate": 9.32460493430591e-05,
      "loss": 0.0107,
      "step": 2090
    },
    {
      "epoch": 6.231454005934718,
      "grad_norm": 0.13447318971157074,
      "learning_rate": 9.316282404787871e-05,
      "loss": 0.0104,
      "step": 2100
    },
    {
      "epoch": 6.261127596439169,
      "grad_norm": 0.14388670027256012,
      "learning_rate": 9.30791267309698e-05,
      "loss": 0.0101,
      "step": 2110
    },
    {
      "epoch": 6.29080118694362,
      "grad_norm": 0.14514364302158356,
      "learning_rate": 9.299495830763286e-05,
      "loss": 0.0096,
      "step": 2120
    },
    {
      "epoch": 6.320474777448071,
      "grad_norm": 0.09969841688871384,
      "learning_rate": 9.291031969832026e-05,
      "loss": 0.0115,
      "step": 2130
    },
    {
      "epoch": 6.350148367952523,
      "grad_norm": 0.14247213304042816,
      "learning_rate": 9.282521182862629e-05,
      "loss": 0.0115,
      "step": 2140
    },
    {
      "epoch": 6.379821958456973,
      "grad_norm": 0.13603922724723816,
      "learning_rate": 9.273963562927695e-05,
      "loss": 0.0147,
      "step": 2150
    },
    {
      "epoch": 6.4094955489614245,
      "grad_norm": 0.23838986456394196,
      "learning_rate": 9.265359203611987e-05,
      "loss": 0.0115,
      "step": 2160
    },
    {
      "epoch": 6.439169139465875,
      "grad_norm": 0.11037889868021011,
      "learning_rate": 9.256708199011401e-05,
      "loss": 0.0096,
      "step": 2170
    },
    {
      "epoch": 6.468842729970326,
      "grad_norm": 0.15438543260097504,
      "learning_rate": 9.248010643731935e-05,
      "loss": 0.0126,
      "step": 2180
    },
    {
      "epoch": 6.498516320474778,
      "grad_norm": 0.12337090075016022,
      "learning_rate": 9.239266632888659e-05,
      "loss": 0.0093,
      "step": 2190
    },
    {
      "epoch": 6.528189910979228,
      "grad_norm": 0.09916039556264877,
      "learning_rate": 9.230476262104677e-05,
      "loss": 0.0114,
      "step": 2200
    },
    {
      "epoch": 6.55786350148368,
      "grad_norm": 0.1173950806260109,
      "learning_rate": 9.221639627510076e-05,
      "loss": 0.0095,
      "step": 2210
    },
    {
      "epoch": 6.58753709198813,
      "grad_norm": 0.14159461855888367,
      "learning_rate": 9.212756825740873e-05,
      "loss": 0.0096,
      "step": 2220
    },
    {
      "epoch": 6.617210682492582,
      "grad_norm": 0.14629167318344116,
      "learning_rate": 9.20382795393797e-05,
      "loss": 0.0118,
      "step": 2230
    },
    {
      "epoch": 6.646884272997033,
      "grad_norm": 0.11691708117723465,
      "learning_rate": 9.194853109746074e-05,
      "loss": 0.0095,
      "step": 2240
    },
    {
      "epoch": 6.6765578635014835,
      "grad_norm": 0.12816114723682404,
      "learning_rate": 9.185832391312644e-05,
      "loss": 0.0133,
      "step": 2250
    },
    {
      "epoch": 6.706231454005935,
      "grad_norm": 0.11063099652528763,
      "learning_rate": 9.176765897286813e-05,
      "loss": 0.0125,
      "step": 2260
    },
    {
      "epoch": 6.735905044510385,
      "grad_norm": 0.12592542171478271,
      "learning_rate": 9.167653726818305e-05,
      "loss": 0.0097,
      "step": 2270
    },
    {
      "epoch": 6.765578635014837,
      "grad_norm": 0.20816679298877716,
      "learning_rate": 9.158495979556358e-05,
      "loss": 0.0127,
      "step": 2280
    },
    {
      "epoch": 6.795252225519288,
      "grad_norm": 0.13589587807655334,
      "learning_rate": 9.14929275564863e-05,
      "loss": 0.0112,
      "step": 2290
    },
    {
      "epoch": 6.824925816023739,
      "grad_norm": 0.17974646389484406,
      "learning_rate": 9.140044155740101e-05,
      "loss": 0.0088,
      "step": 2300
    },
    {
      "epoch": 6.85459940652819,
      "grad_norm": 0.18915049731731415,
      "learning_rate": 9.130750280971978e-05,
      "loss": 0.0156,
      "step": 2310
    },
    {
      "epoch": 6.884272997032641,
      "grad_norm": 0.14018063247203827,
      "learning_rate": 9.121411232980588e-05,
      "loss": 0.0098,
      "step": 2320
    },
    {
      "epoch": 6.913946587537092,
      "grad_norm": 0.13840338587760925,
      "learning_rate": 9.112027113896262e-05,
      "loss": 0.017,
      "step": 2330
    },
    {
      "epoch": 6.943620178041543,
      "grad_norm": 0.11696403473615646,
      "learning_rate": 9.102598026342222e-05,
      "loss": 0.0099,
      "step": 2340
    },
    {
      "epoch": 6.973293768545994,
      "grad_norm": 0.13574601709842682,
      "learning_rate": 9.093124073433463e-05,
      "loss": 0.014,
      "step": 2350
    },
    {
      "epoch": 7.002967359050445,
      "grad_norm": 0.10100409388542175,
      "learning_rate": 9.083605358775612e-05,
      "loss": 0.0103,
      "step": 2360
    },
    {
      "epoch": 7.032640949554896,
      "grad_norm": 0.16500301659107208,
      "learning_rate": 9.074041986463808e-05,
      "loss": 0.0109,
      "step": 2370
    },
    {
      "epoch": 7.062314540059347,
      "grad_norm": 0.18439586460590363,
      "learning_rate": 9.064434061081562e-05,
      "loss": 0.0095,
      "step": 2380
    },
    {
      "epoch": 7.091988130563799,
      "grad_norm": 0.12944291532039642,
      "learning_rate": 9.0547816876996e-05,
      "loss": 0.0123,
      "step": 2390
    },
    {
      "epoch": 7.121661721068249,
      "grad_norm": 0.1536119282245636,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.0125,
      "step": 2400
    },
    {
      "epoch": 7.1513353115727005,
      "grad_norm": 0.1650673747062683,
      "learning_rate": 9.035344019648702e-05,
      "loss": 0.0092,
      "step": 2410
    },
    {
      "epoch": 7.181008902077151,
      "grad_norm": 0.12277387827634811,
      "learning_rate": 9.025558937546988e-05,
      "loss": 0.0102,
      "step": 2420
    },
    {
      "epoch": 7.210682492581602,
      "grad_norm": 0.15055687725543976,
      "learning_rate": 9.015729832577681e-05,
      "loss": 0.0094,
      "step": 2430
    },
    {
      "epoch": 7.240356083086054,
      "grad_norm": 0.16825971007347107,
      "learning_rate": 9.005856812230304e-05,
      "loss": 0.0116,
      "step": 2440
    },
    {
      "epoch": 7.270029673590504,
      "grad_norm": 0.10691312700510025,
      "learning_rate": 8.995939984474624e-05,
      "loss": 0.0095,
      "step": 2450
    },
    {
      "epoch": 7.299703264094956,
      "grad_norm": 0.14602239429950714,
      "learning_rate": 8.98597945775948e-05,
      "loss": 0.0124,
      "step": 2460
    },
    {
      "epoch": 7.329376854599406,
      "grad_norm": 0.13875631988048553,
      "learning_rate": 8.975975341011596e-05,
      "loss": 0.0106,
      "step": 2470
    },
    {
      "epoch": 7.359050445103858,
      "grad_norm": 0.12208565324544907,
      "learning_rate": 8.965927743634391e-05,
      "loss": 0.0108,
      "step": 2480
    },
    {
      "epoch": 7.388724035608309,
      "grad_norm": 0.11230789124965668,
      "learning_rate": 8.955836775506776e-05,
      "loss": 0.0081,
      "step": 2490
    },
    {
      "epoch": 7.4183976261127595,
      "grad_norm": 0.13064904510974884,
      "learning_rate": 8.945702546981969e-05,
      "loss": 0.0122,
      "step": 2500
    },
    {
      "epoch": 7.448071216617211,
      "grad_norm": 0.16824467480182648,
      "learning_rate": 8.935525168886262e-05,
      "loss": 0.0112,
      "step": 2510
    },
    {
      "epoch": 7.4777448071216615,
      "grad_norm": 0.11342830210924149,
      "learning_rate": 8.92530475251784e-05,
      "loss": 0.0103,
      "step": 2520
    },
    {
      "epoch": 7.507418397626113,
      "grad_norm": 0.15296466648578644,
      "learning_rate": 8.91504140964553e-05,
      "loss": 0.0085,
      "step": 2530
    },
    {
      "epoch": 7.537091988130564,
      "grad_norm": 0.16064001619815826,
      "learning_rate": 8.90473525250761e-05,
      "loss": 0.0114,
      "step": 2540
    },
    {
      "epoch": 7.566765578635015,
      "grad_norm": 0.10076630860567093,
      "learning_rate": 8.894386393810563e-05,
      "loss": 0.0144,
      "step": 2550
    },
    {
      "epoch": 7.596439169139466,
      "grad_norm": 0.15510450303554535,
      "learning_rate": 8.883994946727849e-05,
      "loss": 0.0112,
      "step": 2560
    },
    {
      "epoch": 7.626112759643917,
      "grad_norm": 0.21251456439495087,
      "learning_rate": 8.873561024898668e-05,
      "loss": 0.0106,
      "step": 2570
    },
    {
      "epoch": 7.655786350148368,
      "grad_norm": 0.17526623606681824,
      "learning_rate": 8.863084742426719e-05,
      "loss": 0.0113,
      "step": 2580
    },
    {
      "epoch": 7.6854599406528195,
      "grad_norm": 0.12284035235643387,
      "learning_rate": 8.852566213878947e-05,
      "loss": 0.0114,
      "step": 2590
    },
    {
      "epoch": 7.71513353115727,
      "grad_norm": 0.12916874885559082,
      "learning_rate": 8.842005554284296e-05,
      "loss": 0.0099,
      "step": 2600
    },
    {
      "epoch": 7.744807121661721,
      "grad_norm": 0.17493458092212677,
      "learning_rate": 8.831402879132446e-05,
      "loss": 0.0092,
      "step": 2610
    },
    {
      "epoch": 7.774480712166172,
      "grad_norm": 0.12995202839374542,
      "learning_rate": 8.820758304372557e-05,
      "loss": 0.0104,
      "step": 2620
    },
    {
      "epoch": 7.804154302670623,
      "grad_norm": 0.08063960820436478,
      "learning_rate": 8.810071946411989e-05,
      "loss": 0.0087,
      "step": 2630
    },
    {
      "epoch": 7.833827893175075,
      "grad_norm": 0.10341209173202515,
      "learning_rate": 8.799343922115044e-05,
      "loss": 0.0077,
      "step": 2640
    },
    {
      "epoch": 7.863501483679525,
      "grad_norm": 0.108217254281044,
      "learning_rate": 8.788574348801675e-05,
      "loss": 0.0117,
      "step": 2650
    },
    {
      "epoch": 7.893175074183977,
      "grad_norm": 0.1359342336654663,
      "learning_rate": 8.77776334424621e-05,
      "loss": 0.0157,
      "step": 2660
    },
    {
      "epoch": 7.922848664688427,
      "grad_norm": 0.13467204570770264,
      "learning_rate": 8.766911026676064e-05,
      "loss": 0.011,
      "step": 2670
    },
    {
      "epoch": 7.9525222551928785,
      "grad_norm": 0.1321392059326172,
      "learning_rate": 8.756017514770443e-05,
      "loss": 0.0136,
      "step": 2680
    },
    {
      "epoch": 7.98219584569733,
      "grad_norm": 0.16591744124889374,
      "learning_rate": 8.745082927659047e-05,
      "loss": 0.0093,
      "step": 2690
    },
    {
      "epoch": 8.011869436201781,
      "grad_norm": 0.14482906460762024,
      "learning_rate": 8.73410738492077e-05,
      "loss": 0.012,
      "step": 2700
    },
    {
      "epoch": 8.041543026706231,
      "grad_norm": 0.12772271037101746,
      "learning_rate": 8.723091006582389e-05,
      "loss": 0.0118,
      "step": 2710
    },
    {
      "epoch": 8.071216617210682,
      "grad_norm": 0.12283479422330856,
      "learning_rate": 8.71203391311725e-05,
      "loss": 0.0104,
      "step": 2720
    },
    {
      "epoch": 8.100890207715134,
      "grad_norm": 0.11549960821866989,
      "learning_rate": 8.700936225443959e-05,
      "loss": 0.0101,
      "step": 2730
    },
    {
      "epoch": 8.130563798219585,
      "grad_norm": 0.12340424209833145,
      "learning_rate": 8.689798064925049e-05,
      "loss": 0.0108,
      "step": 2740
    },
    {
      "epoch": 8.160237388724036,
      "grad_norm": 0.13828441500663757,
      "learning_rate": 8.678619553365659e-05,
      "loss": 0.0084,
      "step": 2750
    },
    {
      "epoch": 8.189910979228486,
      "grad_norm": 0.176404669880867,
      "learning_rate": 8.6674008130122e-05,
      "loss": 0.0093,
      "step": 2760
    },
    {
      "epoch": 8.219584569732937,
      "grad_norm": 0.12829335033893585,
      "learning_rate": 8.656141966551019e-05,
      "loss": 0.0095,
      "step": 2770
    },
    {
      "epoch": 8.249258160237389,
      "grad_norm": 0.15259785950183868,
      "learning_rate": 8.644843137107059e-05,
      "loss": 0.01,
      "step": 2780
    },
    {
      "epoch": 8.27893175074184,
      "grad_norm": 0.17509141564369202,
      "learning_rate": 8.633504448242505e-05,
      "loss": 0.0086,
      "step": 2790
    },
    {
      "epoch": 8.308605341246292,
      "grad_norm": 0.18248887360095978,
      "learning_rate": 8.622126023955446e-05,
      "loss": 0.0098,
      "step": 2800
    },
    {
      "epoch": 8.338278931750741,
      "grad_norm": 0.13852570950984955,
      "learning_rate": 8.610707988678503e-05,
      "loss": 0.0109,
      "step": 2810
    },
    {
      "epoch": 8.367952522255193,
      "grad_norm": 0.15752212703227997,
      "learning_rate": 8.599250467277483e-05,
      "loss": 0.0088,
      "step": 2820
    },
    {
      "epoch": 8.397626112759644,
      "grad_norm": 0.08893997222185135,
      "learning_rate": 8.587753585050004e-05,
      "loss": 0.0096,
      "step": 2830
    },
    {
      "epoch": 8.427299703264095,
      "grad_norm": 0.1134849488735199,
      "learning_rate": 8.576217467724128e-05,
      "loss": 0.0105,
      "step": 2840
    },
    {
      "epoch": 8.456973293768545,
      "grad_norm": 0.18662723898887634,
      "learning_rate": 8.564642241456986e-05,
      "loss": 0.0102,
      "step": 2850
    },
    {
      "epoch": 8.486646884272997,
      "grad_norm": 0.07733399420976639,
      "learning_rate": 8.553028032833397e-05,
      "loss": 0.0104,
      "step": 2860
    },
    {
      "epoch": 8.516320474777448,
      "grad_norm": 0.13568611443042755,
      "learning_rate": 8.541374968864487e-05,
      "loss": 0.0086,
      "step": 2870
    },
    {
      "epoch": 8.5459940652819,
      "grad_norm": 0.07520133256912231,
      "learning_rate": 8.529683176986295e-05,
      "loss": 0.0084,
      "step": 2880
    },
    {
      "epoch": 8.57566765578635,
      "grad_norm": 0.06504914909601212,
      "learning_rate": 8.517952785058385e-05,
      "loss": 0.0086,
      "step": 2890
    },
    {
      "epoch": 8.605341246290802,
      "grad_norm": 0.15347328782081604,
      "learning_rate": 8.506183921362443e-05,
      "loss": 0.0097,
      "step": 2900
    },
    {
      "epoch": 8.635014836795252,
      "grad_norm": 0.15778006613254547,
      "learning_rate": 8.494376714600878e-05,
      "loss": 0.0157,
      "step": 2910
    },
    {
      "epoch": 8.664688427299703,
      "grad_norm": 0.15846551954746246,
      "learning_rate": 8.482531293895412e-05,
      "loss": 0.0093,
      "step": 2920
    },
    {
      "epoch": 8.694362017804155,
      "grad_norm": 0.19215027987957,
      "learning_rate": 8.470647788785665e-05,
      "loss": 0.0101,
      "step": 2930
    },
    {
      "epoch": 8.724035608308606,
      "grad_norm": 0.14507389068603516,
      "learning_rate": 8.458726329227747e-05,
      "loss": 0.0102,
      "step": 2940
    },
    {
      "epoch": 8.753709198813056,
      "grad_norm": 0.15245003998279572,
      "learning_rate": 8.44676704559283e-05,
      "loss": 0.009,
      "step": 2950
    },
    {
      "epoch": 8.783382789317507,
      "grad_norm": 0.16085323691368103,
      "learning_rate": 8.434770068665723e-05,
      "loss": 0.0122,
      "step": 2960
    },
    {
      "epoch": 8.813056379821958,
      "grad_norm": 0.12772567570209503,
      "learning_rate": 8.422735529643444e-05,
      "loss": 0.012,
      "step": 2970
    },
    {
      "epoch": 8.84272997032641,
      "grad_norm": 0.12985776364803314,
      "learning_rate": 8.410663560133784e-05,
      "loss": 0.0085,
      "step": 2980
    },
    {
      "epoch": 8.872403560830861,
      "grad_norm": 0.13308796286582947,
      "learning_rate": 8.398554292153866e-05,
      "loss": 0.0076,
      "step": 2990
    },
    {
      "epoch": 8.90207715133531,
      "grad_norm": 0.18778546154499054,
      "learning_rate": 8.386407858128706e-05,
      "loss": 0.0116,
      "step": 3000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.375898115552051e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}