{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4711425206124853,
  "eval_steps": 500,
  "global_step": 2800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0016826518593303045,
      "grad_norm": 3.75418950341011,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 0.9983,
      "step": 10
    },
    {
      "epoch": 0.003365303718660609,
      "grad_norm": 4.027030925274863,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.9697,
      "step": 20
    },
    {
      "epoch": 0.005047955577990914,
      "grad_norm": 4.048987349136423,
      "learning_rate": 1.5e-05,
      "loss": 0.9412,
      "step": 30
    },
    {
      "epoch": 0.006730607437321218,
      "grad_norm": 5.720158971431411,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.8783,
      "step": 40
    },
    {
      "epoch": 0.008413259296651522,
      "grad_norm": 4.718965032869529,
      "learning_rate": 2.5e-05,
      "loss": 0.8454,
      "step": 50
    },
    {
      "epoch": 0.010095911155981827,
      "grad_norm": 3.5785181087788835,
      "learning_rate": 3e-05,
      "loss": 0.809,
      "step": 60
    },
    {
      "epoch": 0.011778563015312132,
      "grad_norm": 4.11981684712826,
      "learning_rate": 2.9999786123888308e-05,
      "loss": 0.7556,
      "step": 70
    },
    {
      "epoch": 0.013461214874642436,
      "grad_norm": 6.082559649594005,
      "learning_rate": 2.9999144501652298e-05,
      "loss": 0.7613,
      "step": 80
    },
    {
      "epoch": 0.01514386673397274,
      "grad_norm": 1.957553999291205,
      "learning_rate": 2.9998075151588992e-05,
      "loss": 0.7784,
      "step": 90
    },
    {
      "epoch": 0.016826518593303044,
      "grad_norm": 1.6706087540201593,
      "learning_rate": 2.999657810419285e-05,
      "loss": 0.7658,
      "step": 100
    },
    {
      "epoch": 0.01850917045263335,
      "grad_norm": 2.909734954037323,
      "learning_rate": 2.999465340215489e-05,
      "loss": 0.7331,
      "step": 110
    },
    {
      "epoch": 0.020191822311963654,
      "grad_norm": 1.977272298268717,
      "learning_rate": 2.999230110036149e-05,
      "loss": 0.7507,
      "step": 120
    },
    {
      "epoch": 0.02187447417129396,
      "grad_norm": 1.8089524113272115,
      "learning_rate": 2.99895212658928e-05,
      "loss": 0.7309,
      "step": 130
    },
    {
      "epoch": 0.023557126030624265,
      "grad_norm": 2.134962179309057,
      "learning_rate": 2.9986313978020846e-05,
      "loss": 0.721,
      "step": 140
    },
    {
      "epoch": 0.02523977788995457,
      "grad_norm": 11.10353091330302,
      "learning_rate": 2.9982679328207262e-05,
      "loss": 0.7338,
      "step": 150
    },
    {
      "epoch": 0.02692242974928487,
      "grad_norm": 1.4444344817739057,
      "learning_rate": 2.9978617420100692e-05,
      "loss": 0.7227,
      "step": 160
    },
    {
      "epoch": 0.028605081608615177,
      "grad_norm": 1.453288161439029,
      "learning_rate": 2.9974128369533805e-05,
      "loss": 0.7107,
      "step": 170
    },
    {
      "epoch": 0.03028773346794548,
      "grad_norm": 3.475164856876678,
      "learning_rate": 2.9969212304520034e-05,
      "loss": 0.7303,
      "step": 180
    },
    {
      "epoch": 0.03197038532727579,
      "grad_norm": 1.1636824531496957,
      "learning_rate": 2.9963869365249895e-05,
      "loss": 0.6688,
      "step": 190
    },
    {
      "epoch": 0.03365303718660609,
      "grad_norm": 1.8518695174363622,
      "learning_rate": 2.995809970408699e-05,
      "loss": 0.7003,
      "step": 200
    },
    {
      "epoch": 0.0353356890459364,
      "grad_norm": 4.09791760479377,
      "learning_rate": 2.9951903485563685e-05,
      "loss": 0.7442,
      "step": 210
    },
    {
      "epoch": 0.0370183409052667,
      "grad_norm": 2.4987929291159956,
      "learning_rate": 2.99452808863764e-05,
      "loss": 0.7517,
      "step": 220
    },
    {
      "epoch": 0.03870099276459701,
      "grad_norm": 3.4584802037194087,
      "learning_rate": 2.993823209538056e-05,
      "loss": 0.7537,
      "step": 230
    },
    {
      "epoch": 0.04038364462392731,
      "grad_norm": 2.511130636368107,
      "learning_rate": 2.9930757313585238e-05,
      "loss": 0.7599,
      "step": 240
    },
    {
      "epoch": 0.04206629648325761,
      "grad_norm": 1.7030446444812277,
      "learning_rate": 2.9922856754147406e-05,
      "loss": 0.7126,
      "step": 250
    },
    {
      "epoch": 0.04374894834258792,
      "grad_norm": 4.790377413030976,
      "learning_rate": 2.9914530642365852e-05,
      "loss": 0.72,
      "step": 260
    },
    {
      "epoch": 0.04543160020191822,
      "grad_norm": 2.0321244924961976,
      "learning_rate": 2.990577921567476e-05,
      "loss": 0.6733,
      "step": 270
    },
    {
      "epoch": 0.04711425206124853,
      "grad_norm": 2.310370624749643,
      "learning_rate": 2.989660272363696e-05,
      "loss": 0.7212,
      "step": 280
    },
    {
      "epoch": 0.04879690392057883,
      "grad_norm": 3.451763592410144,
      "learning_rate": 2.988700142793676e-05,
      "loss": 0.7237,
      "step": 290
    },
    {
      "epoch": 0.05047955577990914,
      "grad_norm": 5.317302731978485,
      "learning_rate": 2.9876975602372536e-05,
      "loss": 0.7558,
      "step": 300
    },
    {
      "epoch": 0.05216220763923944,
      "grad_norm": 2.3026448136142914,
      "learning_rate": 2.9866525532848906e-05,
      "loss": 0.6985,
      "step": 310
    },
    {
      "epoch": 0.05384485949856974,
      "grad_norm": 1.8320545447196381,
      "learning_rate": 2.9855651517368567e-05,
      "loss": 0.7227,
      "step": 320
    },
    {
      "epoch": 0.05552751135790005,
      "grad_norm": 1.9908218789466392,
      "learning_rate": 2.9844353866023802e-05,
      "loss": 0.7075,
      "step": 330
    },
    {
      "epoch": 0.05721016321723035,
      "grad_norm": 5.182840115712529,
      "learning_rate": 2.9832632900987642e-05,
      "loss": 0.7207,
      "step": 340
    },
    {
      "epoch": 0.05889281507656066,
      "grad_norm": 1.5483797249278837,
      "learning_rate": 2.982048895650468e-05,
      "loss": 0.7233,
      "step": 350
    },
    {
      "epoch": 0.06057546693589096,
      "grad_norm": 2.3382590504722693,
      "learning_rate": 2.9807922378881537e-05,
      "loss": 0.7002,
      "step": 360
    },
    {
      "epoch": 0.06225811879522127,
      "grad_norm": 3.1859655239636937,
      "learning_rate": 2.979493352647697e-05,
      "loss": 0.7201,
      "step": 370
    },
    {
      "epoch": 0.06394077065455157,
      "grad_norm": 0.9149159742557087,
      "learning_rate": 2.9781522769691686e-05,
      "loss": 0.7136,
      "step": 380
    },
    {
      "epoch": 0.06562342251388188,
      "grad_norm": 10.861566072795899,
      "learning_rate": 2.9767690490957758e-05,
      "loss": 0.7068,
      "step": 390
    },
    {
      "epoch": 0.06730607437321218,
      "grad_norm": 2.8618866775651006,
      "learning_rate": 2.9753437084727713e-05,
      "loss": 0.7239,
      "step": 400
    },
    {
      "epoch": 0.06898872623254249,
      "grad_norm": 2.8726068570785097,
      "learning_rate": 2.9738762957463292e-05,
      "loss": 0.7245,
      "step": 410
    },
    {
      "epoch": 0.0706713780918728,
      "grad_norm": 2.4481298042739112,
      "learning_rate": 2.9723668527623877e-05,
      "loss": 0.7752,
      "step": 420
    },
    {
      "epoch": 0.0723540299512031,
      "grad_norm": 1.8599931346602536,
      "learning_rate": 2.9708154225654526e-05,
      "loss": 0.7323,
      "step": 430
    },
    {
      "epoch": 0.0740366818105334,
      "grad_norm": 1.2855737813743626,
      "learning_rate": 2.9692220493973712e-05,
      "loss": 0.7037,
      "step": 440
    },
    {
      "epoch": 0.0757193336698637,
      "grad_norm": 4.629091463528233,
      "learning_rate": 2.9675867786960718e-05,
      "loss": 0.6867,
      "step": 450
    },
    {
      "epoch": 0.07740198552919401,
      "grad_norm": 6.294427059845777,
      "learning_rate": 2.9659096570942654e-05,
      "loss": 0.7272,
      "step": 460
    },
    {
      "epoch": 0.07908463738852431,
      "grad_norm": 2.4758348810051345,
      "learning_rate": 2.9641907324181194e-05,
      "loss": 0.6779,
      "step": 470
    },
    {
      "epoch": 0.08076728924785462,
      "grad_norm": 1.3455245255212915,
      "learning_rate": 2.96243005368589e-05,
      "loss": 0.7051,
      "step": 480
    },
    {
      "epoch": 0.08244994110718493,
      "grad_norm": 4.796150475871981,
      "learning_rate": 2.960627671106527e-05,
      "loss": 0.7547,
      "step": 490
    },
    {
      "epoch": 0.08413259296651522,
      "grad_norm": 2.684441445075641,
      "learning_rate": 2.9587836360782405e-05,
      "loss": 0.709,
      "step": 500
    },
    {
      "epoch": 0.08581524482584553,
      "grad_norm": 1.3869329152815553,
      "learning_rate": 2.9568980011870357e-05,
      "loss": 0.7073,
      "step": 510
    },
    {
      "epoch": 0.08749789668517584,
      "grad_norm": 2.5576974478207197,
      "learning_rate": 2.954970820205214e-05,
      "loss": 0.6918,
      "step": 520
    },
    {
      "epoch": 0.08918054854450615,
      "grad_norm": 1.1525450967004647,
      "learning_rate": 2.9530021480898393e-05,
      "loss": 0.6698,
      "step": 530
    },
    {
      "epoch": 0.09086320040383644,
      "grad_norm": 2.847083851829901,
      "learning_rate": 2.9509920409811696e-05,
      "loss": 0.671,
      "step": 540
    },
    {
      "epoch": 0.09254585226316675,
      "grad_norm": 2.561042091789346,
      "learning_rate": 2.9489405562010565e-05,
      "loss": 0.75,
      "step": 550
    },
    {
      "epoch": 0.09422850412249706,
      "grad_norm": 4.458337350053255,
      "learning_rate": 2.9468477522513132e-05,
      "loss": 0.7277,
      "step": 560
    },
    {
      "epoch": 0.09591115598182735,
      "grad_norm": 3.114622509219852,
      "learning_rate": 2.9447136888120408e-05,
      "loss": 0.6967,
      "step": 570
    },
    {
      "epoch": 0.09759380784115766,
      "grad_norm": 1.6295210229360877,
      "learning_rate": 2.9425384267399327e-05,
      "loss": 0.6867,
      "step": 580
    },
    {
      "epoch": 0.09927645970048797,
      "grad_norm": 1.7579117810504754,
      "learning_rate": 2.940322028066534e-05,
      "loss": 0.7236,
      "step": 590
    },
    {
      "epoch": 0.10095911155981828,
      "grad_norm": 1.788183804411441,
      "learning_rate": 2.938064555996476e-05,
      "loss": 0.6864,
      "step": 600
    },
    {
      "epoch": 0.10264176341914857,
      "grad_norm": 2.8340511721646373,
      "learning_rate": 2.9357660749056713e-05,
      "loss": 0.6847,
      "step": 610
    },
    {
      "epoch": 0.10432441527847888,
      "grad_norm": 2.5230840193297985,
      "learning_rate": 2.9334266503394803e-05,
      "loss": 0.6889,
      "step": 620
    },
    {
      "epoch": 0.10600706713780919,
      "grad_norm": 7.346086885083334,
      "learning_rate": 2.9310463490108397e-05,
      "loss": 0.7419,
      "step": 630
    },
    {
      "epoch": 0.10768971899713949,
      "grad_norm": 2.356832890545339,
      "learning_rate": 2.928625238798362e-05,
      "loss": 0.7369,
      "step": 640
    },
    {
      "epoch": 0.1093723708564698,
      "grad_norm": 2.4978380391841095,
      "learning_rate": 2.9261633887443993e-05,
      "loss": 0.6948,
      "step": 650
    },
    {
      "epoch": 0.1110550227158001,
      "grad_norm": 3.535487375505793,
      "learning_rate": 2.9236608690530738e-05,
      "loss": 0.7081,
      "step": 660
    },
    {
      "epoch": 0.11273767457513041,
      "grad_norm": 2.522638625540884,
      "learning_rate": 2.921117751088276e-05,
      "loss": 0.7191,
      "step": 670
    },
    {
      "epoch": 0.1144203264344607,
      "grad_norm": 3.055823541699581,
      "learning_rate": 2.91853410737163e-05,
      "loss": 0.74,
      "step": 680
    },
    {
      "epoch": 0.11610297829379101,
      "grad_norm": 3.270117047516123,
      "learning_rate": 2.915910011580426e-05,
      "loss": 0.6829,
      "step": 690
    },
    {
      "epoch": 0.11778563015312132,
      "grad_norm": 2.3219806056695367,
      "learning_rate": 2.9132455385455176e-05,
      "loss": 0.7062,
      "step": 700
    },
    {
      "epoch": 0.11946828201245162,
      "grad_norm": 1.541921603113568,
      "learning_rate": 2.9105407642491895e-05,
      "loss": 0.7217,
      "step": 710
    },
    {
      "epoch": 0.12115093387178193,
      "grad_norm": 1.557595298876376,
      "learning_rate": 2.907795765822989e-05,
      "loss": 0.7083,
      "step": 720
    },
    {
      "epoch": 0.12283358573111224,
      "grad_norm": 2.3829156571868753,
      "learning_rate": 2.9050106215455283e-05,
      "loss": 0.6992,
      "step": 730
    },
    {
      "epoch": 0.12451623759044254,
      "grad_norm": 7.536777098548366,
      "learning_rate": 2.9021854108402516e-05,
      "loss": 0.7248,
      "step": 740
    },
    {
      "epoch": 0.12619888944977284,
      "grad_norm": 1.3408030642895519,
      "learning_rate": 2.8993202142731693e-05,
      "loss": 0.6375,
      "step": 750
    },
    {
      "epoch": 0.12788154130910315,
      "grad_norm": 2.4880776314537254,
      "learning_rate": 2.8964151135505616e-05,
      "loss": 0.7063,
      "step": 760
    },
    {
      "epoch": 0.12956419316843346,
      "grad_norm": 1.5507053769862247,
      "learning_rate": 2.8934701915166477e-05,
      "loss": 0.73,
      "step": 770
    },
    {
      "epoch": 0.13124684502776376,
      "grad_norm": 3.5622930633942564,
      "learning_rate": 2.890485532151225e-05,
      "loss": 0.7521,
      "step": 780
    },
    {
      "epoch": 0.13292949688709407,
      "grad_norm": 4.188153799459233,
      "learning_rate": 2.887461220567271e-05,
      "loss": 0.6841,
      "step": 790
    },
    {
      "epoch": 0.13461214874642435,
      "grad_norm": 2.702901312773331,
      "learning_rate": 2.8843973430085204e-05,
      "loss": 0.694,
      "step": 800
    },
    {
      "epoch": 0.13629480060575466,
      "grad_norm": 3.8663384632605293,
      "learning_rate": 2.8812939868470016e-05,
      "loss": 0.7376,
      "step": 810
    },
    {
      "epoch": 0.13797745246508497,
      "grad_norm": 7.613582881082294,
      "learning_rate": 2.878151240580548e-05,
      "loss": 0.7082,
      "step": 820
    },
    {
      "epoch": 0.13966010432441528,
      "grad_norm": 2.8755666754814015,
      "learning_rate": 2.874969193830274e-05,
      "loss": 0.7486,
      "step": 830
    },
    {
      "epoch": 0.1413427561837456,
      "grad_norm": 2.049640563529798,
      "learning_rate": 2.871747937338016e-05,
      "loss": 0.7375,
      "step": 840
    },
    {
      "epoch": 0.1430254080430759,
      "grad_norm": 3.2253208680917993,
      "learning_rate": 2.8684875629637505e-05,
      "loss": 0.7183,
      "step": 850
    },
    {
      "epoch": 0.1447080599024062,
      "grad_norm": 2.0453993741696306,
      "learning_rate": 2.8651881636829698e-05,
      "loss": 0.6953,
      "step": 860
    },
    {
      "epoch": 0.1463907117617365,
      "grad_norm": 1.3478445170381042,
      "learning_rate": 2.861849833584032e-05,
      "loss": 0.7205,
      "step": 870
    },
    {
      "epoch": 0.1480733636210668,
      "grad_norm": 6.483405424500114,
      "learning_rate": 2.8584726678654787e-05,
      "loss": 0.7331,
      "step": 880
    },
    {
      "epoch": 0.1497560154803971,
      "grad_norm": 1.6912080503281164,
      "learning_rate": 2.85505676283332e-05,
      "loss": 0.6985,
      "step": 890
    },
    {
      "epoch": 0.1514386673397274,
      "grad_norm": 2.089097733011486,
      "learning_rate": 2.851602215898287e-05,
      "loss": 0.7291,
      "step": 900
    },
    {
      "epoch": 0.15312131919905772,
      "grad_norm": 3.3599665631038325,
      "learning_rate": 2.8481091255730552e-05,
      "loss": 0.7125,
      "step": 910
    },
    {
      "epoch": 0.15480397105838803,
      "grad_norm": 5.803874517218743,
      "learning_rate": 2.844577591469435e-05,
      "loss": 0.6614,
      "step": 920
    },
    {
      "epoch": 0.15648662291771834,
      "grad_norm": 4.180624256153927,
      "learning_rate": 2.8410077142955304e-05,
      "loss": 0.6921,
      "step": 930
    },
    {
      "epoch": 0.15816927477704862,
      "grad_norm": 2.51395384445247,
      "learning_rate": 2.8373995958528683e-05,
      "loss": 0.6788,
      "step": 940
    },
    {
      "epoch": 0.15985192663637893,
      "grad_norm": 2.0786229734439,
      "learning_rate": 2.8337533390334942e-05,
      "loss": 0.6324,
      "step": 950
    },
    {
      "epoch": 0.16153457849570924,
      "grad_norm": 2.1798201763285774,
      "learning_rate": 2.8300690478170388e-05,
      "loss": 0.7128,
      "step": 960
    },
    {
      "epoch": 0.16321723035503954,
      "grad_norm": 1.7736042633296192,
      "learning_rate": 2.826346827267753e-05,
      "loss": 0.6854,
      "step": 970
    },
    {
      "epoch": 0.16489988221436985,
      "grad_norm": 3.6499571810784377,
      "learning_rate": 2.8225867835315114e-05,
      "loss": 0.7246,
      "step": 980
    },
    {
      "epoch": 0.16658253407370016,
      "grad_norm": 8.401076529411414,
      "learning_rate": 2.8187890238327842e-05,
      "loss": 0.7166,
      "step": 990
    },
    {
      "epoch": 0.16826518593303044,
      "grad_norm": 1.6815155727131568,
      "learning_rate": 2.814953656471583e-05,
      "loss": 0.6962,
      "step": 1000
    },
    {
      "epoch": 0.16994783779236075,
      "grad_norm": 3.59100648398944,
      "learning_rate": 2.8110807908203682e-05,
      "loss": 0.7271,
      "step": 1010
    },
    {
      "epoch": 0.17163048965169106,
      "grad_norm": 2.9612400836384034,
      "learning_rate": 2.8071705373209328e-05,
      "loss": 0.7048,
      "step": 1020
    },
    {
      "epoch": 0.17331314151102137,
      "grad_norm": 1.6314524411685434,
      "learning_rate": 2.803223007481252e-05,
      "loss": 0.7237,
      "step": 1030
    },
    {
      "epoch": 0.17499579337035168,
      "grad_norm": 4.046292885407821,
      "learning_rate": 2.7992383138723034e-05,
      "loss": 0.7066,
      "step": 1040
    },
    {
      "epoch": 0.17667844522968199,
      "grad_norm": 3.4626891652569665,
      "learning_rate": 2.7952165701248573e-05,
      "loss": 0.7537,
      "step": 1050
    },
    {
      "epoch": 0.1783610970890123,
      "grad_norm": 4.129895397644279,
      "learning_rate": 2.7911578909262353e-05,
      "loss": 0.7348,
      "step": 1060
    },
    {
      "epoch": 0.18004374894834257,
      "grad_norm": 2.1894044487856847,
      "learning_rate": 2.787062392017041e-05,
      "loss": 0.7145,
      "step": 1070
    },
    {
      "epoch": 0.18172640080767288,
      "grad_norm": 2.988495224416439,
      "learning_rate": 2.7829301901878592e-05,
      "loss": 0.7091,
      "step": 1080
    },
    {
      "epoch": 0.1834090526670032,
      "grad_norm": 2.493227176786327,
      "learning_rate": 2.7787614032759243e-05,
      "loss": 0.7427,
      "step": 1090
    },
    {
      "epoch": 0.1850917045263335,
      "grad_norm": 2.9382266505350723,
      "learning_rate": 2.7745561501617605e-05,
      "loss": 0.7081,
      "step": 1100
    },
    {
      "epoch": 0.1867743563856638,
      "grad_norm": 1.9294251174769146,
      "learning_rate": 2.7703145507657923e-05,
      "loss": 0.679,
      "step": 1110
    },
    {
      "epoch": 0.18845700824499412,
      "grad_norm": 7.011830550553666,
      "learning_rate": 2.766036726044926e-05,
      "loss": 0.6962,
      "step": 1120
    },
    {
      "epoch": 0.19013966010432443,
      "grad_norm": 1.8058177496791177,
      "learning_rate": 2.7617227979890957e-05,
      "loss": 0.6953,
      "step": 1130
    },
    {
      "epoch": 0.1918223119636547,
      "grad_norm": 2.2546595962288727,
      "learning_rate": 2.7573728896177897e-05,
      "loss": 0.6853,
      "step": 1140
    },
    {
      "epoch": 0.19350496382298502,
      "grad_norm": 1.7701647300358836,
      "learning_rate": 2.7529871249765397e-05,
      "loss": 0.737,
      "step": 1150
    },
    {
      "epoch": 0.19518761568231532,
      "grad_norm": 3.2767535691041396,
      "learning_rate": 2.7485656291333845e-05,
      "loss": 0.6878,
      "step": 1160
    },
    {
      "epoch": 0.19687026754164563,
      "grad_norm": 1.231100350207441,
      "learning_rate": 2.7441085281753028e-05,
      "loss": 0.7044,
      "step": 1170
    },
    {
      "epoch": 0.19855291940097594,
      "grad_norm": 5.103379397758491,
      "learning_rate": 2.739615949204617e-05,
      "loss": 0.7028,
      "step": 1180
    },
    {
      "epoch": 0.20023557126030625,
      "grad_norm": 1.745258105735824,
      "learning_rate": 2.7350880203353703e-05,
      "loss": 0.7123,
      "step": 1190
    },
    {
      "epoch": 0.20191822311963656,
      "grad_norm": 2.528898960464809,
      "learning_rate": 2.7305248706896722e-05,
      "loss": 0.7242,
      "step": 1200
    },
    {
      "epoch": 0.20360087497896684,
      "grad_norm": 1.329326803950539,
      "learning_rate": 2.7259266303940164e-05,
      "loss": 0.7315,
      "step": 1210
    },
    {
      "epoch": 0.20528352683829715,
      "grad_norm": 3.523954433912976,
      "learning_rate": 2.7212934305755697e-05,
      "loss": 0.7022,
      "step": 1220
    },
    {
      "epoch": 0.20696617869762746,
      "grad_norm": 1.3845861665687345,
      "learning_rate": 2.7166254033584343e-05,
      "loss": 0.6788,
      "step": 1230
    },
    {
      "epoch": 0.20864883055695777,
      "grad_norm": 1.6893702845026013,
      "learning_rate": 2.7119226818598784e-05,
      "loss": 0.7083,
      "step": 1240
    },
    {
      "epoch": 0.21033148241628807,
      "grad_norm": 3.481606379952265,
      "learning_rate": 2.7071854001865402e-05,
      "loss": 0.7104,
      "step": 1250
    },
    {
      "epoch": 0.21201413427561838,
      "grad_norm": 1.3880604016054,
      "learning_rate": 2.702413693430604e-05,
      "loss": 0.7192,
      "step": 1260
    },
    {
      "epoch": 0.2136967861349487,
      "grad_norm": 2.7420634271532625,
      "learning_rate": 2.697607697665948e-05,
      "loss": 0.7329,
      "step": 1270
    },
    {
      "epoch": 0.21537943799427897,
      "grad_norm": 1.3383701328350484,
      "learning_rate": 2.6927675499442648e-05,
      "loss": 0.7523,
      "step": 1280
    },
    {
      "epoch": 0.21706208985360928,
      "grad_norm": 5.63600709352392,
      "learning_rate": 2.68789338829115e-05,
      "loss": 0.6938,
      "step": 1290
    },
    {
      "epoch": 0.2187447417129396,
      "grad_norm": 1.973997298554772,
      "learning_rate": 2.6829853517021698e-05,
      "loss": 0.7024,
      "step": 1300
    },
    {
      "epoch": 0.2204273935722699,
      "grad_norm": 5.331233664305369,
      "learning_rate": 2.6780435801388945e-05,
      "loss": 0.6978,
      "step": 1310
    },
    {
      "epoch": 0.2221100454316002,
      "grad_norm": 14.545018258920948,
      "learning_rate": 2.6730682145249093e-05,
      "loss": 0.7288,
      "step": 1320
    },
    {
      "epoch": 0.22379269729093051,
      "grad_norm": 2.772459303589031,
      "learning_rate": 2.668059396741795e-05,
      "loss": 0.69,
      "step": 1330
    },
    {
      "epoch": 0.22547534915026082,
      "grad_norm": 1.9806140492727284,
      "learning_rate": 2.6630172696250804e-05,
      "loss": 0.7194,
      "step": 1340
    },
    {
      "epoch": 0.2271580010095911,
      "grad_norm": 2.5305067313330305,
      "learning_rate": 2.6579419769601715e-05,
      "loss": 0.7209,
      "step": 1350
    },
    {
      "epoch": 0.2288406528689214,
      "grad_norm": 4.329479239778255,
      "learning_rate": 2.6528336634782493e-05,
      "loss": 0.7263,
      "step": 1360
    },
    {
      "epoch": 0.23052330472825172,
      "grad_norm": 2.4385930080514124,
      "learning_rate": 2.6476924748521443e-05,
      "loss": 0.7169,
      "step": 1370
    },
    {
      "epoch": 0.23220595658758203,
      "grad_norm": 4.486791723774815,
      "learning_rate": 2.6425185576921812e-05,
      "loss": 0.6791,
      "step": 1380
    },
    {
      "epoch": 0.23388860844691234,
      "grad_norm": 2.1648975510177353,
      "learning_rate": 2.637312059541997e-05,
      "loss": 0.722,
      "step": 1390
    },
    {
      "epoch": 0.23557126030624265,
      "grad_norm": 2.497984836932449,
      "learning_rate": 2.632073128874336e-05,
      "loss": 0.737,
      "step": 1400
    },
    {
      "epoch": 0.23725391216557296,
      "grad_norm": 1.6911389710154248,
      "learning_rate": 2.6268019150868144e-05,
      "loss": 0.7027,
      "step": 1410
    },
    {
      "epoch": 0.23893656402490324,
      "grad_norm": 5.094854691429602,
      "learning_rate": 2.62149856849766e-05,
      "loss": 0.7431,
      "step": 1420
    },
    {
      "epoch": 0.24061921588423354,
      "grad_norm": 1.6056704058079299,
      "learning_rate": 2.616163240341426e-05,
      "loss": 0.7215,
      "step": 1430
    },
    {
      "epoch": 0.24230186774356385,
      "grad_norm": 2.0440590394408793,
      "learning_rate": 2.6107960827646774e-05,
      "loss": 0.6864,
      "step": 1440
    },
    {
      "epoch": 0.24398451960289416,
      "grad_norm": 1.4019933491248435,
      "learning_rate": 2.6053972488216538e-05,
      "loss": 0.7007,
      "step": 1450
    },
    {
      "epoch": 0.24566717146222447,
      "grad_norm": 6.4772716175425185,
      "learning_rate": 2.5999668924699035e-05,
      "loss": 0.6963,
      "step": 1460
    },
    {
      "epoch": 0.24734982332155478,
      "grad_norm": 1.235157923543473,
      "learning_rate": 2.5945051685658923e-05,
      "loss": 0.7158,
      "step": 1470
    },
    {
      "epoch": 0.2490324751808851,
      "grad_norm": 1.6576585358395288,
      "learning_rate": 2.5890122328605908e-05,
      "loss": 0.6918,
      "step": 1480
    },
    {
      "epoch": 0.25071512704021537,
      "grad_norm": 2.6005430314710645,
      "learning_rate": 2.5834882419950295e-05,
      "loss": 0.6666,
      "step": 1490
    },
    {
      "epoch": 0.2523977788995457,
      "grad_norm": 3.83061566974576,
      "learning_rate": 2.577933353495833e-05,
      "loss": 0.724,
      "step": 1500
    },
    {
      "epoch": 0.254080430758876,
      "grad_norm": 2.259260300802235,
      "learning_rate": 2.5723477257707293e-05,
      "loss": 0.725,
      "step": 1510
    },
    {
      "epoch": 0.2557630826182063,
      "grad_norm": 3.1023391020410283,
      "learning_rate": 2.566731518104029e-05,
      "loss": 0.709,
      "step": 1520
    },
    {
      "epoch": 0.2574457344775366,
      "grad_norm": 2.375072076607274,
      "learning_rate": 2.5610848906520878e-05,
      "loss": 0.7031,
      "step": 1530
    },
    {
      "epoch": 0.2591283863368669,
      "grad_norm": 1.638162563319741,
      "learning_rate": 2.5554080044387344e-05,
      "loss": 0.7031,
      "step": 1540
    },
    {
      "epoch": 0.2608110381961972,
      "grad_norm": 8.846026339935685,
      "learning_rate": 2.5497010213506825e-05,
      "loss": 0.7119,
      "step": 1550
    },
    {
      "epoch": 0.26249369005552753,
      "grad_norm": 4.589496329936434,
      "learning_rate": 2.5439641041329128e-05,
      "loss": 0.7043,
      "step": 1560
    },
    {
      "epoch": 0.26417634191485784,
      "grad_norm": 0.9945782670551377,
      "learning_rate": 2.5381974163840313e-05,
      "loss": 0.7026,
      "step": 1570
    },
    {
      "epoch": 0.26585899377418815,
      "grad_norm": 2.341138070970226,
      "learning_rate": 2.532401122551605e-05,
      "loss": 0.744,
      "step": 1580
    },
    {
      "epoch": 0.2675416456335184,
      "grad_norm": 3.446122331658564,
      "learning_rate": 2.526575387927473e-05,
      "loss": 0.6861,
      "step": 1590
    },
    {
      "epoch": 0.2692242974928487,
      "grad_norm": 4.165637435951758,
      "learning_rate": 2.52072037864303e-05,
      "loss": 0.7065,
      "step": 1600
    },
    {
      "epoch": 0.270906949352179,
      "grad_norm": 2.156163863520989,
      "learning_rate": 2.5148362616644926e-05,
      "loss": 0.7383,
      "step": 1610
    },
    {
      "epoch": 0.2725896012115093,
      "grad_norm": 1.489844754983356,
      "learning_rate": 2.508923204788135e-05,
      "loss": 0.7061,
      "step": 1620
    },
    {
      "epoch": 0.27427225307083963,
      "grad_norm": 1.330475018363876,
      "learning_rate": 2.5029813766355062e-05,
      "loss": 0.6916,
      "step": 1630
    },
    {
      "epoch": 0.27595490493016994,
      "grad_norm": 1.5166804777641398,
      "learning_rate": 2.4970109466486202e-05,
      "loss": 0.6998,
      "step": 1640
    },
    {
      "epoch": 0.27763755678950025,
      "grad_norm": 1.1006885605957994,
      "learning_rate": 2.491012085085122e-05,
      "loss": 0.7095,
      "step": 1650
    },
    {
      "epoch": 0.27932020864883056,
      "grad_norm": 3.137721907226618,
      "learning_rate": 2.4849849630134384e-05,
      "loss": 0.7204,
      "step": 1660
    },
    {
      "epoch": 0.28100286050816087,
      "grad_norm": 2.0954950376873747,
      "learning_rate": 2.4789297523078924e-05,
      "loss": 0.7149,
      "step": 1670
    },
    {
      "epoch": 0.2826855123674912,
      "grad_norm": 1.59599915913629,
      "learning_rate": 2.4728466256438072e-05,
      "loss": 0.7052,
      "step": 1680
    },
    {
      "epoch": 0.2843681642268215,
      "grad_norm": 8.944835293630385,
      "learning_rate": 2.4667357564925798e-05,
      "loss": 0.7161,
      "step": 1690
    },
    {
      "epoch": 0.2860508160861518,
      "grad_norm": 6.306096521027849,
      "learning_rate": 2.460597319116735e-05,
      "loss": 0.7219,
      "step": 1700
    },
    {
      "epoch": 0.2877334679454821,
      "grad_norm": 8.531648843126508,
      "learning_rate": 2.4544314885649552e-05,
      "loss": 0.7195,
      "step": 1710
    },
    {
      "epoch": 0.2894161198048124,
      "grad_norm": 6.214651933223859,
      "learning_rate": 2.4482384406670883e-05,
      "loss": 0.6836,
      "step": 1720
    },
    {
      "epoch": 0.29109877166414266,
      "grad_norm": 3.495702421141607,
      "learning_rate": 2.4420183520291354e-05,
      "loss": 0.7241,
      "step": 1730
    },
    {
      "epoch": 0.292781423523473,
      "grad_norm": 7.473552612209333,
      "learning_rate": 2.4357714000282127e-05,
      "loss": 0.6664,
      "step": 1740
    },
    {
      "epoch": 0.2944640753828033,
      "grad_norm": 2.528030042904385,
      "learning_rate": 2.4294977628074938e-05,
      "loss": 0.7415,
      "step": 1750
    },
    {
      "epoch": 0.2961467272421336,
      "grad_norm": 15.300967237030209,
      "learning_rate": 2.42319761927113e-05,
      "loss": 0.7336,
      "step": 1760
    },
    {
      "epoch": 0.2978293791014639,
      "grad_norm": 5.5616370399668345,
      "learning_rate": 2.4168711490791484e-05,
      "loss": 0.72,
      "step": 1770
    },
    {
      "epoch": 0.2995120309607942,
      "grad_norm": 3.7113016371508993,
      "learning_rate": 2.4105185326423286e-05,
      "loss": 0.723,
      "step": 1780
    },
    {
      "epoch": 0.3011946828201245,
      "grad_norm": 2.0578017620481397,
      "learning_rate": 2.4041399511170574e-05,
      "loss": 0.7008,
      "step": 1790
    },
    {
      "epoch": 0.3028773346794548,
      "grad_norm": 1.5192292607159725,
      "learning_rate": 2.3977355864001635e-05,
      "loss": 0.7107,
      "step": 1800
    },
    {
      "epoch": 0.30455998653878513,
      "grad_norm": 1.4432235055034852,
      "learning_rate": 2.3913056211237304e-05,
      "loss": 0.7112,
      "step": 1810
    },
    {
      "epoch": 0.30624263839811544,
      "grad_norm": 3.8368406145537577,
      "learning_rate": 2.3848502386498866e-05,
      "loss": 0.6875,
      "step": 1820
    },
    {
      "epoch": 0.30792529025744575,
      "grad_norm": 1.7029662451991225,
      "learning_rate": 2.3783696230655802e-05,
      "loss": 0.6797,
      "step": 1830
    },
    {
      "epoch": 0.30960794211677606,
      "grad_norm": 2.340409810623994,
      "learning_rate": 2.371863959177326e-05,
      "loss": 0.7211,
      "step": 1840
    },
    {
      "epoch": 0.31129059397610637,
      "grad_norm": 8.08876148980577,
      "learning_rate": 2.365333432505937e-05,
      "loss": 0.7208,
      "step": 1850
    },
    {
      "epoch": 0.3129732458354367,
      "grad_norm": 3.415957276427778,
      "learning_rate": 2.3587782292812323e-05,
      "loss": 0.707,
      "step": 1860
    },
    {
      "epoch": 0.31465589769476693,
      "grad_norm": 1.116630095030083,
      "learning_rate": 2.35219853643673e-05,
      "loss": 0.6863,
      "step": 1870
    },
    {
      "epoch": 0.31633854955409724,
      "grad_norm": 1.9365729913801322,
      "learning_rate": 2.3455945416043132e-05,
      "loss": 0.705,
      "step": 1880
    },
    {
      "epoch": 0.31802120141342755,
      "grad_norm": 1.401982720531144,
      "learning_rate": 2.338966433108879e-05,
      "loss": 0.6872,
      "step": 1890
    },
    {
      "epoch": 0.31970385327275785,
      "grad_norm": 1.6958509020945827,
      "learning_rate": 2.3323143999629712e-05,
      "loss": 0.7146,
      "step": 1900
    },
    {
      "epoch": 0.32138650513208816,
      "grad_norm": 1.6193006043035303,
      "learning_rate": 2.3256386318613877e-05,
      "loss": 0.6887,
      "step": 1910
    },
    {
      "epoch": 0.32306915699141847,
      "grad_norm": 12.979287887341894,
      "learning_rate": 2.318939319175771e-05,
      "loss": 0.7063,
      "step": 1920
    },
    {
      "epoch": 0.3247518088507488,
      "grad_norm": 2.053699792676608,
      "learning_rate": 2.3122166529491822e-05,
      "loss": 0.7921,
      "step": 1930
    },
    {
      "epoch": 0.3264344607100791,
      "grad_norm": 2.1805832639041993,
      "learning_rate": 2.3054708248906483e-05,
      "loss": 0.6892,
      "step": 1940
    },
    {
      "epoch": 0.3281171125694094,
      "grad_norm": 2.627137721499779,
      "learning_rate": 2.2987020273696996e-05,
      "loss": 0.6937,
      "step": 1950
    },
    {
      "epoch": 0.3297997644287397,
      "grad_norm": 3.4422222646515284,
      "learning_rate": 2.2919104534108825e-05,
      "loss": 0.7274,
      "step": 1960
    },
    {
      "epoch": 0.33148241628807,
      "grad_norm": 3.942895129812573,
      "learning_rate": 2.2850962966882547e-05,
      "loss": 0.7089,
      "step": 1970
    },
    {
      "epoch": 0.3331650681474003,
      "grad_norm": 2.250499246098397,
      "learning_rate": 2.278259751519861e-05,
      "loss": 0.7288,
      "step": 1980
    },
    {
      "epoch": 0.33484772000673063,
      "grad_norm": 1.9217973398784702,
      "learning_rate": 2.2714010128621957e-05,
      "loss": 0.6971,
      "step": 1990
    },
    {
      "epoch": 0.3365303718660609,
      "grad_norm": 7.4365036240913005,
      "learning_rate": 2.2645202763046385e-05,
      "loss": 0.7027,
      "step": 2000
    },
    {
      "epoch": 0.3382130237253912,
      "grad_norm": 3.6010458486700267,
      "learning_rate": 2.2576177380638808e-05,
      "loss": 0.7135,
      "step": 2010
    },
    {
      "epoch": 0.3398956755847215,
      "grad_norm": 3.6134685076054756,
      "learning_rate": 2.2506935949783277e-05,
      "loss": 0.6703,
      "step": 2020
    },
    {
      "epoch": 0.3415783274440518,
      "grad_norm": 2.047032124552865,
      "learning_rate": 2.243748044502485e-05,
      "loss": 0.7021,
      "step": 2030
    },
    {
      "epoch": 0.3432609793033821,
      "grad_norm": 7.2853023401907135,
      "learning_rate": 2.236781284701332e-05,
      "loss": 0.698,
      "step": 2040
    },
    {
      "epoch": 0.3449436311627124,
      "grad_norm": 10.62863599608186,
      "learning_rate": 2.229793514244666e-05,
      "loss": 0.7083,
      "step": 2050
    },
    {
      "epoch": 0.34662628302204274,
      "grad_norm": 2.7228396995619293,
      "learning_rate": 2.222784932401445e-05,
      "loss": 0.6714,
      "step": 2060
    },
    {
      "epoch": 0.34830893488137304,
      "grad_norm": 8.496361795340652,
      "learning_rate": 2.2157557390341e-05,
      "loss": 0.7215,
      "step": 2070
    },
    {
      "epoch": 0.34999158674070335,
      "grad_norm": 2.7703412032428885,
      "learning_rate": 2.2087061345928375e-05,
      "loss": 0.7355,
      "step": 2080
    },
    {
      "epoch": 0.35167423860003366,
      "grad_norm": 1.615589476440187,
      "learning_rate": 2.2016363201099205e-05,
      "loss": 0.6552,
      "step": 2090
    },
    {
      "epoch": 0.35335689045936397,
      "grad_norm": 2.325164551192993,
      "learning_rate": 2.1945464971939424e-05,
      "loss": 0.7059,
      "step": 2100
    },
    {
      "epoch": 0.3550395423186943,
      "grad_norm": 3.5764361397956113,
      "learning_rate": 2.1874368680240692e-05,
      "loss": 0.7005,
      "step": 2110
    },
    {
      "epoch": 0.3567221941780246,
      "grad_norm": 1.347667616644207,
      "learning_rate": 2.1803076353442806e-05,
      "loss": 0.7102,
      "step": 2120
    },
    {
      "epoch": 0.3584048460373549,
      "grad_norm": 2.0318308073252425,
      "learning_rate": 2.1731590024575848e-05,
      "loss": 0.7173,
      "step": 2130
    },
    {
      "epoch": 0.36008749789668515,
      "grad_norm": 1.3373304099117234,
      "learning_rate": 2.165991173220223e-05,
      "loss": 0.7601,
      "step": 2140
    },
    {
      "epoch": 0.36177014975601546,
      "grad_norm": 1.9787611361857145,
      "learning_rate": 2.158804352035855e-05,
      "loss": 0.721,
      "step": 2150
    },
    {
      "epoch": 0.36345280161534577,
      "grad_norm": 1.0674118498737855,
      "learning_rate": 2.1515987438497295e-05,
      "loss": 0.6998,
      "step": 2160
    },
    {
      "epoch": 0.3651354534746761,
      "grad_norm": 2.0137284693641386,
      "learning_rate": 2.1443745541428416e-05,
      "loss": 0.6765,
      "step": 2170
    },
    {
      "epoch": 0.3668181053340064,
      "grad_norm": 4.298295885455176,
      "learning_rate": 2.137131988926072e-05,
      "loss": 0.7012,
      "step": 2180
    },
    {
      "epoch": 0.3685007571933367,
      "grad_norm": 9.91613007277119,
      "learning_rate": 2.129871254734312e-05,
      "loss": 0.7062,
      "step": 2190
    },
    {
      "epoch": 0.370183409052667,
      "grad_norm": 1.6702084657906056,
      "learning_rate": 2.122592558620575e-05,
      "loss": 0.7319,
      "step": 2200
    },
    {
      "epoch": 0.3718660609119973,
      "grad_norm": 2.2151204900067634,
      "learning_rate": 2.1152961081500906e-05,
      "loss": 0.6853,
      "step": 2210
    },
    {
      "epoch": 0.3735487127713276,
      "grad_norm": 2.06281971620876,
      "learning_rate": 2.1079821113943863e-05,
      "loss": 0.7137,
      "step": 2220
    },
    {
      "epoch": 0.3752313646306579,
      "grad_norm": 2.865770701923554,
      "learning_rate": 2.100650776925353e-05,
      "loss": 0.7066,
      "step": 2230
    },
    {
      "epoch": 0.37691401648998824,
      "grad_norm": 2.36832805891566,
      "learning_rate": 2.0933023138092995e-05,
      "loss": 0.7401,
      "step": 2240
    },
    {
      "epoch": 0.37859666834931854,
      "grad_norm": 1.54287595832039,
      "learning_rate": 2.0859369316009877e-05,
      "loss": 0.7535,
      "step": 2250
    },
    {
      "epoch": 0.38027932020864885,
      "grad_norm": 2.919513911195001,
      "learning_rate": 2.0785548403376592e-05,
      "loss": 0.71,
      "step": 2260
    },
    {
      "epoch": 0.38196197206797916,
      "grad_norm": 1.618528464339543,
      "learning_rate": 2.0711562505330437e-05,
      "loss": 0.7181,
      "step": 2270
    },
    {
      "epoch": 0.3836446239273094,
      "grad_norm": 1.1878290222512224,
      "learning_rate": 2.063741373171357e-05,
      "loss": 0.732,
      "step": 2280
    },
    {
      "epoch": 0.3853272757866397,
      "grad_norm": 2.6515115997158554,
      "learning_rate": 2.0563104197012847e-05,
      "loss": 0.709,
      "step": 2290
    },
    {
      "epoch": 0.38700992764597003,
      "grad_norm": 1.3199743255372487,
      "learning_rate": 2.048863602029951e-05,
      "loss": 0.7276,
      "step": 2300
    },
    {
      "epoch": 0.38869257950530034,
      "grad_norm": 1.9262341939684873,
      "learning_rate": 2.0414011325168777e-05,
      "loss": 0.6964,
      "step": 2310
    },
    {
      "epoch": 0.39037523136463065,
      "grad_norm": 1.3563137443786035,
      "learning_rate": 2.0339232239679252e-05,
      "loss": 0.685,
      "step": 2320
    },
    {
      "epoch": 0.39205788322396096,
      "grad_norm": 1.6219810171886226,
      "learning_rate": 2.026430089629229e-05,
      "loss": 0.6916,
      "step": 2330
    },
    {
      "epoch": 0.39374053508329127,
      "grad_norm": 2.441077712269494,
      "learning_rate": 2.0189219431811123e-05,
      "loss": 0.7447,
      "step": 2340
    },
    {
      "epoch": 0.3954231869426216,
      "grad_norm": 2.1115157955300616,
      "learning_rate": 2.0113989987319988e-05,
      "loss": 0.6802,
      "step": 2350
    },
    {
      "epoch": 0.3971058388019519,
      "grad_norm": 1.4354556930544433,
      "learning_rate": 2.0038614708123023e-05,
      "loss": 0.7004,
      "step": 2360
    },
    {
      "epoch": 0.3987884906612822,
      "grad_norm": 2.4000999827221756,
      "learning_rate": 1.996309574368311e-05,
      "loss": 0.6898,
      "step": 2370
    },
    {
      "epoch": 0.4004711425206125,
      "grad_norm": 1.03142424758583,
      "learning_rate": 1.9887435247560586e-05,
      "loss": 0.7086,
      "step": 2380
    },
    {
      "epoch": 0.4021537943799428,
      "grad_norm": 2.8813939519904914,
      "learning_rate": 1.981163537735181e-05,
      "loss": 0.6954,
      "step": 2390
    },
    {
      "epoch": 0.4038364462392731,
      "grad_norm": 1.1816627689525059,
      "learning_rate": 1.9735698294627644e-05,
      "loss": 0.7142,
      "step": 2400
    },
    {
      "epoch": 0.4055190980986034,
      "grad_norm": 1.3945618099849828,
      "learning_rate": 1.9659626164871828e-05,
      "loss": 0.7097,
      "step": 2410
    },
    {
      "epoch": 0.4072017499579337,
      "grad_norm": 1.9043603505045925,
      "learning_rate": 1.95834211574192e-05,
      "loss": 0.6992,
      "step": 2420
    },
    {
      "epoch": 0.408884401817264,
      "grad_norm": 3.5606637873787648,
      "learning_rate": 1.9507085445393855e-05,
      "loss": 0.7118,
      "step": 2430
    },
    {
      "epoch": 0.4105670536765943,
      "grad_norm": 1.569864161622481,
      "learning_rate": 1.9430621205647156e-05,
      "loss": 0.6971,
      "step": 2440
    },
    {
      "epoch": 0.4122497055359246,
      "grad_norm": 4.4331480365167,
      "learning_rate": 1.935403061869568e-05,
      "loss": 0.6726,
      "step": 2450
    },
    {
      "epoch": 0.4139323573952549,
      "grad_norm": 4.432472927572385,
      "learning_rate": 1.9277315868659017e-05,
      "loss": 0.718,
      "step": 2460
    },
    {
      "epoch": 0.4156150092545852,
      "grad_norm": 1.8000872219993729,
      "learning_rate": 1.920047914319749e-05,
      "loss": 0.7391,
      "step": 2470
    },
    {
      "epoch": 0.41729766111391553,
      "grad_norm": 2.2415704708896294,
      "learning_rate": 1.9123522633449772e-05,
      "loss": 0.7093,
      "step": 2480
    },
    {
      "epoch": 0.41898031297324584,
      "grad_norm": 0.9860844679413583,
      "learning_rate": 1.9046448533970423e-05,
      "loss": 0.6874,
      "step": 2490
    },
    {
      "epoch": 0.42066296483257615,
      "grad_norm": 4.705732364227723,
      "learning_rate": 1.8969259042667255e-05,
      "loss": 0.6766,
      "step": 2500
    },
    {
      "epoch": 0.42234561669190646,
      "grad_norm": 3.3731463958248264,
      "learning_rate": 1.8891956360738706e-05,
      "loss": 0.7155,
      "step": 2510
    },
    {
      "epoch": 0.42402826855123676,
      "grad_norm": 2.2599456108772915,
      "learning_rate": 1.881454269261105e-05,
      "loss": 0.6901,
      "step": 2520
    },
    {
      "epoch": 0.4257109204105671,
      "grad_norm": 1.887949729545168,
      "learning_rate": 1.873702024587553e-05,
      "loss": 0.7528,
      "step": 2530
    },
    {
      "epoch": 0.4273935722698974,
      "grad_norm": 3.5001965876274115,
      "learning_rate": 1.865939123122541e-05,
      "loss": 0.6795,
      "step": 2540
    },
    {
      "epoch": 0.4290762241292277,
      "grad_norm": 3.739512720034628,
      "learning_rate": 1.858165786239293e-05,
      "loss": 0.6902,
      "step": 2550
    },
    {
      "epoch": 0.43075887598855794,
      "grad_norm": 2.835735723710262,
      "learning_rate": 1.850382235608618e-05,
      "loss": 0.7013,
      "step": 2560
    },
    {
      "epoch": 0.43244152784788825,
      "grad_norm": 3.0809412432066594,
      "learning_rate": 1.842588693192589e-05,
      "loss": 0.6916,
      "step": 2570
    },
    {
      "epoch": 0.43412417970721856,
      "grad_norm": 1.9629938051045357,
      "learning_rate": 1.834785381238212e-05,
      "loss": 0.7165,
      "step": 2580
    },
    {
      "epoch": 0.43580683156654887,
      "grad_norm": 2.2447376170957023,
      "learning_rate": 1.8269725222710895e-05,
      "loss": 0.7273,
      "step": 2590
    },
    {
      "epoch": 0.4374894834258792,
      "grad_norm": 2.214057365632279,
      "learning_rate": 1.8191503390890745e-05,
      "loss": 0.7073,
      "step": 2600
    },
    {
      "epoch": 0.4391721352852095,
      "grad_norm": 1.8172031418314936,
      "learning_rate": 1.8113190547559167e-05,
      "loss": 0.715,
      "step": 2610
    },
    {
      "epoch": 0.4408547871445398,
      "grad_norm": 2.835706633111018,
      "learning_rate": 1.8034788925949024e-05,
      "loss": 0.706,
      "step": 2620
    },
    {
      "epoch": 0.4425374390038701,
      "grad_norm": 1.1621082466184123,
      "learning_rate": 1.795630076182484e-05,
      "loss": 0.7084,
      "step": 2630
    },
    {
      "epoch": 0.4442200908632004,
      "grad_norm": 11.176574486878446,
      "learning_rate": 1.7877728293419067e-05,
      "loss": 0.7092,
      "step": 2640
    },
    {
      "epoch": 0.4459027427225307,
      "grad_norm": 10.672692821884855,
      "learning_rate": 1.7799073761368236e-05,
      "loss": 0.7247,
      "step": 2650
    },
    {
      "epoch": 0.44758539458186103,
      "grad_norm": 3.089797606768679,
      "learning_rate": 1.7720339408649084e-05,
      "loss": 0.7212,
      "step": 2660
    },
    {
      "epoch": 0.44926804644119134,
      "grad_norm": 11.51700032068762,
      "learning_rate": 1.7641527480514575e-05,
      "loss": 0.708,
      "step": 2670
    },
    {
      "epoch": 0.45095069830052165,
      "grad_norm": 2.9535720728987007,
      "learning_rate": 1.756264022442987e-05,
      "loss": 0.6986,
      "step": 2680
    },
    {
      "epoch": 0.4526333501598519,
      "grad_norm": 4.148500258838489,
      "learning_rate": 1.7483679890008247e-05,
      "loss": 0.6571,
      "step": 2690
    },
    {
      "epoch": 0.4543160020191822,
      "grad_norm": 1.7950072159263486,
      "learning_rate": 1.740464872894695e-05,
      "loss": 0.6751,
      "step": 2700
    },
    {
      "epoch": 0.4559986538785125,
      "grad_norm": 1.3583282737793725,
      "learning_rate": 1.732554899496297e-05,
      "loss": 0.7705,
      "step": 2710
    },
    {
      "epoch": 0.4576813057378428,
      "grad_norm": 2.184156205512354,
      "learning_rate": 1.7246382943728777e-05,
      "loss": 0.6631,
      "step": 2720
    },
    {
      "epoch": 0.45936395759717313,
      "grad_norm": 1.6297260663890405,
      "learning_rate": 1.7167152832807997e-05,
      "loss": 0.6946,
      "step": 2730
    },
    {
      "epoch": 0.46104660945650344,
      "grad_norm": 1.6037313522955436,
      "learning_rate": 1.7087860921591047e-05,
      "loss": 0.6914,
      "step": 2740
    },
    {
      "epoch": 0.46272926131583375,
      "grad_norm": 1.6197737501652023,
      "learning_rate": 1.7008509471230673e-05,
      "loss": 0.7312,
      "step": 2750
    },
    {
      "epoch": 0.46441191317516406,
      "grad_norm": 2.0167791329962395,
      "learning_rate": 1.69291007445775e-05,
      "loss": 0.6459,
      "step": 2760
    },
    {
      "epoch": 0.46609456503449437,
      "grad_norm": 7.215038584013174,
      "learning_rate": 1.684963700611548e-05,
      "loss": 0.6989,
      "step": 2770
    },
    {
      "epoch": 0.4677772168938247,
      "grad_norm": 3.3650128982999603,
      "learning_rate": 1.677012052189734e-05,
      "loss": 0.7156,
      "step": 2780
    },
    {
      "epoch": 0.469459868753155,
      "grad_norm": 1.6353091631825594,
      "learning_rate": 1.6690553559479946e-05,
      "loss": 0.6916,
      "step": 2790
    },
    {
      "epoch": 0.4711425206124853,
      "grad_norm": 2.314420263644613,
      "learning_rate": 1.6610938387859623e-05,
      "loss": 0.7351,
      "step": 2800
    }
  ],
  "logging_steps": 10,
  "max_steps": 5943,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2756662754476032e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}