{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 11838,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00016894745734076703,
      "grad_norm": 6.621600151062012,
      "learning_rate": 0.0,
      "loss": 10.540443420410156,
      "step": 1
    },
    {
      "epoch": 0.0016894745734076701,
      "grad_norm": 6.755760669708252,
      "learning_rate": 1.3499999999999998e-06,
      "loss": 10.498290167914497,
      "step": 10
    },
    {
      "epoch": 0.0033789491468153403,
      "grad_norm": 5.475676536560059,
      "learning_rate": 2.85e-06,
      "loss": 10.216492462158204,
      "step": 20
    },
    {
      "epoch": 0.00506842372022301,
      "grad_norm": 2.3924124240875244,
      "learning_rate": 4.35e-06,
      "loss": 9.751193237304687,
      "step": 30
    },
    {
      "epoch": 0.0067578982936306806,
      "grad_norm": 2.018623113632202,
      "learning_rate": 5.85e-06,
      "loss": 9.445990753173827,
      "step": 40
    },
    {
      "epoch": 0.00844737286703835,
      "grad_norm": 1.15117609500885,
      "learning_rate": 7.35e-06,
      "loss": 9.269255065917969,
      "step": 50
    },
    {
      "epoch": 0.01013684744044602,
      "grad_norm": 0.9644901752471924,
      "learning_rate": 8.849999999999998e-06,
      "loss": 9.1482421875,
      "step": 60
    },
    {
      "epoch": 0.011826322013853691,
      "grad_norm": 0.9443461894989014,
      "learning_rate": 1.035e-05,
      "loss": 9.042950439453126,
      "step": 70
    },
    {
      "epoch": 0.013515796587261361,
      "grad_norm": 0.8729987144470215,
      "learning_rate": 1.1849999999999998e-05,
      "loss": 8.95867462158203,
      "step": 80
    },
    {
      "epoch": 0.015205271160669031,
      "grad_norm": 0.7886430621147156,
      "learning_rate": 1.3349999999999998e-05,
      "loss": 8.854803466796875,
      "step": 90
    },
    {
      "epoch": 0.0168947457340767,
      "grad_norm": 0.8168472647666931,
      "learning_rate": 1.485e-05,
      "loss": 8.736968231201171,
      "step": 100
    },
    {
      "epoch": 0.018584220307484373,
      "grad_norm": 0.6868988275527954,
      "learning_rate": 1.6349999999999998e-05,
      "loss": 8.659466552734376,
      "step": 110
    },
    {
      "epoch": 0.02027369488089204,
      "grad_norm": 0.6943208575248718,
      "learning_rate": 1.7849999999999997e-05,
      "loss": 8.55049819946289,
      "step": 120
    },
    {
      "epoch": 0.021963169454299714,
      "grad_norm": 0.6718711853027344,
      "learning_rate": 1.935e-05,
      "loss": 8.454410552978516,
      "step": 130
    },
    {
      "epoch": 0.023652644027707382,
      "grad_norm": 0.5750948190689087,
      "learning_rate": 2.085e-05,
      "loss": 8.361714172363282,
      "step": 140
    },
    {
      "epoch": 0.025342118601115054,
      "grad_norm": 0.545462965965271,
      "learning_rate": 2.2349999999999998e-05,
      "loss": 8.286084747314453,
      "step": 150
    },
    {
      "epoch": 0.027031593174522722,
      "grad_norm": 0.6024239659309387,
      "learning_rate": 2.3849999999999997e-05,
      "loss": 8.2337158203125,
      "step": 160
    },
    {
      "epoch": 0.028721067747930394,
      "grad_norm": 0.5649603605270386,
      "learning_rate": 2.535e-05,
      "loss": 8.169093322753906,
      "step": 170
    },
    {
      "epoch": 0.030410542321338063,
      "grad_norm": 0.42989474534988403,
      "learning_rate": 2.6849999999999995e-05,
      "loss": 8.12140121459961,
      "step": 180
    },
    {
      "epoch": 0.03210001689474573,
      "grad_norm": 0.5494393110275269,
      "learning_rate": 2.8349999999999998e-05,
      "loss": 8.079795837402344,
      "step": 190
    },
    {
      "epoch": 0.0337894914681534,
      "grad_norm": 0.48430609703063965,
      "learning_rate": 2.985e-05,
      "loss": 8.047111511230469,
      "step": 200
    },
    {
      "epoch": 0.035478966041561075,
      "grad_norm": 0.45849987864494324,
      "learning_rate": 3.1349999999999996e-05,
      "loss": 7.997683715820313,
      "step": 210
    },
    {
      "epoch": 0.03716844061496875,
      "grad_norm": 0.3945513665676117,
      "learning_rate": 3.285e-05,
      "loss": 7.981051635742188,
      "step": 220
    },
    {
      "epoch": 0.03885791518837641,
      "grad_norm": 0.4015548825263977,
      "learning_rate": 3.435e-05,
      "loss": 7.976935577392578,
      "step": 230
    },
    {
      "epoch": 0.04054738976178408,
      "grad_norm": 0.6867141127586365,
      "learning_rate": 3.585e-05,
      "loss": 7.935280609130859,
      "step": 240
    },
    {
      "epoch": 0.042236864335191755,
      "grad_norm": 0.47820013761520386,
      "learning_rate": 3.735e-05,
      "loss": 7.920246124267578,
      "step": 250
    },
    {
      "epoch": 0.04392633890859943,
      "grad_norm": 0.44240179657936096,
      "learning_rate": 3.8849999999999996e-05,
      "loss": 7.917320251464844,
      "step": 260
    },
    {
      "epoch": 0.0456158134820071,
      "grad_norm": 0.49838986992836,
      "learning_rate": 4.035e-05,
      "loss": 7.892914581298828,
      "step": 270
    },
    {
      "epoch": 0.047305288055414764,
      "grad_norm": 0.41489648818969727,
      "learning_rate": 4.185e-05,
      "loss": 7.873424530029297,
      "step": 280
    },
    {
      "epoch": 0.048994762628822436,
      "grad_norm": 0.3524978756904602,
      "learning_rate": 4.334999999999999e-05,
      "loss": 7.852528381347656,
      "step": 290
    },
    {
      "epoch": 0.05068423720223011,
      "grad_norm": 0.46830496191978455,
      "learning_rate": 4.484999999999999e-05,
      "loss": 7.83319091796875,
      "step": 300
    },
    {
      "epoch": 0.05237371177563778,
      "grad_norm": 0.5290191173553467,
      "learning_rate": 4.6349999999999995e-05,
      "loss": 7.8142448425292965,
      "step": 310
    },
    {
      "epoch": 0.054063186349045445,
      "grad_norm": 0.4697173535823822,
      "learning_rate": 4.785e-05,
      "loss": 7.787006378173828,
      "step": 320
    },
    {
      "epoch": 0.055752660922453116,
      "grad_norm": 0.5864154100418091,
      "learning_rate": 4.935e-05,
      "loss": 7.743091583251953,
      "step": 330
    },
    {
      "epoch": 0.05744213549586079,
      "grad_norm": 0.5467583537101746,
      "learning_rate": 5.0849999999999996e-05,
      "loss": 7.7266998291015625,
      "step": 340
    },
    {
      "epoch": 0.05913161006926846,
      "grad_norm": 0.5317718982696533,
      "learning_rate": 5.234999999999999e-05,
      "loss": 7.715788269042969,
      "step": 350
    },
    {
      "epoch": 0.060821084642676125,
      "grad_norm": 0.45864003896713257,
      "learning_rate": 5.3849999999999994e-05,
      "loss": 7.676011657714843,
      "step": 360
    },
    {
      "epoch": 0.0625105592160838,
      "grad_norm": 0.6899635195732117,
      "learning_rate": 5.535e-05,
      "loss": 7.6587471008300785,
      "step": 370
    },
    {
      "epoch": 0.06420003378949146,
      "grad_norm": 0.4785831868648529,
      "learning_rate": 5.684999999999999e-05,
      "loss": 7.648049163818359,
      "step": 380
    },
    {
      "epoch": 0.06588950836289914,
      "grad_norm": 0.42162397503852844,
      "learning_rate": 5.8349999999999995e-05,
      "loss": 7.612094116210938,
      "step": 390
    },
    {
      "epoch": 0.0675789829363068,
      "grad_norm": 0.6696052551269531,
      "learning_rate": 5.985e-05,
      "loss": 7.605092620849609,
      "step": 400
    },
    {
      "epoch": 0.06926845750971448,
      "grad_norm": 0.5291442275047302,
      "learning_rate": 6.134999999999999e-05,
      "loss": 7.5769294738769535,
      "step": 410
    },
    {
      "epoch": 0.07095793208312215,
      "grad_norm": 0.6115548014640808,
      "learning_rate": 6.285e-05,
      "loss": 7.553981781005859,
      "step": 420
    },
    {
      "epoch": 0.07264740665652981,
      "grad_norm": 0.5771138668060303,
      "learning_rate": 6.434999999999999e-05,
      "loss": 7.541645812988281,
      "step": 430
    },
    {
      "epoch": 0.0743368812299375,
      "grad_norm": 0.647227942943573,
      "learning_rate": 6.584999999999999e-05,
      "loss": 7.522480010986328,
      "step": 440
    },
    {
      "epoch": 0.07602635580334516,
      "grad_norm": 0.6701403856277466,
      "learning_rate": 6.735e-05,
      "loss": 7.49889907836914,
      "step": 450
    },
    {
      "epoch": 0.07771583037675282,
      "grad_norm": 0.844932496547699,
      "learning_rate": 6.884999999999999e-05,
      "loss": 7.476522064208984,
      "step": 460
    },
    {
      "epoch": 0.0794053049501605,
      "grad_norm": 0.5116700530052185,
      "learning_rate": 7.034999999999999e-05,
      "loss": 7.456998443603515,
      "step": 470
    },
    {
      "epoch": 0.08109477952356817,
      "grad_norm": 0.5343000292778015,
      "learning_rate": 7.184999999999998e-05,
      "loss": 7.443318939208984,
      "step": 480
    },
    {
      "epoch": 0.08278425409697585,
      "grad_norm": 0.6147258281707764,
      "learning_rate": 7.335e-05,
      "loss": 7.403359985351562,
      "step": 490
    },
    {
      "epoch": 0.08447372867038351,
      "grad_norm": 0.6813654899597168,
      "learning_rate": 7.484999999999999e-05,
      "loss": 7.403208160400391,
      "step": 500
    },
    {
      "epoch": 0.08447372867038351,
      "eval_loss": 7.390021324157715,
      "eval_runtime": 4.0235,
      "eval_samples_per_second": 248.538,
      "eval_steps_per_second": 5.219,
      "step": 500
    },
    {
      "epoch": 0.08616320324379118,
      "grad_norm": 0.6618097424507141,
      "learning_rate": 7.635e-05,
      "loss": 7.392906188964844,
      "step": 510
    },
    {
      "epoch": 0.08785267781719885,
      "grad_norm": 0.6140709519386292,
      "learning_rate": 7.785e-05,
      "loss": 7.364067840576172,
      "step": 520
    },
    {
      "epoch": 0.08954215239060652,
      "grad_norm": 0.6116703748703003,
      "learning_rate": 7.934999999999999e-05,
      "loss": 7.337810516357422,
      "step": 530
    },
    {
      "epoch": 0.0912316269640142,
      "grad_norm": 0.8000091314315796,
      "learning_rate": 8.085e-05,
      "loss": 7.299466705322265,
      "step": 540
    },
    {
      "epoch": 0.09292110153742186,
      "grad_norm": 0.5890388488769531,
      "learning_rate": 8.235e-05,
      "loss": 7.308570098876953,
      "step": 550
    },
    {
      "epoch": 0.09461057611082953,
      "grad_norm": 1.0396614074707031,
      "learning_rate": 8.385e-05,
      "loss": 7.27392349243164,
      "step": 560
    },
    {
      "epoch": 0.0963000506842372,
      "grad_norm": 0.5742290019989014,
      "learning_rate": 8.534999999999999e-05,
      "loss": 7.271208953857422,
      "step": 570
    },
    {
      "epoch": 0.09798952525764487,
      "grad_norm": 0.684992790222168,
      "learning_rate": 8.684999999999998e-05,
      "loss": 7.2550514221191404,
      "step": 580
    },
    {
      "epoch": 0.09967899983105254,
      "grad_norm": 1.2290043830871582,
      "learning_rate": 8.834999999999999e-05,
      "loss": 7.2304443359375,
      "step": 590
    },
    {
      "epoch": 0.10136847440446022,
      "grad_norm": 0.7645843029022217,
      "learning_rate": 8.984999999999999e-05,
      "loss": 7.205104064941406,
      "step": 600
    },
    {
      "epoch": 0.10305794897786788,
      "grad_norm": 0.730484664440155,
      "learning_rate": 9.134999999999998e-05,
      "loss": 7.210204315185547,
      "step": 610
    },
    {
      "epoch": 0.10474742355127556,
      "grad_norm": 0.7423863410949707,
      "learning_rate": 9.285e-05,
      "loss": 7.166588592529297,
      "step": 620
    },
    {
      "epoch": 0.10643689812468322,
      "grad_norm": 0.888006329536438,
      "learning_rate": 9.434999999999999e-05,
      "loss": 7.162047576904297,
      "step": 630
    },
    {
      "epoch": 0.10812637269809089,
      "grad_norm": 0.9920506477355957,
      "learning_rate": 9.585e-05,
      "loss": 7.145941925048828,
      "step": 640
    },
    {
      "epoch": 0.10981584727149857,
      "grad_norm": 0.8996961712837219,
      "learning_rate": 9.735e-05,
      "loss": 7.110871124267578,
      "step": 650
    },
    {
      "epoch": 0.11150532184490623,
      "grad_norm": 0.7783015370368958,
      "learning_rate": 9.884999999999999e-05,
      "loss": 7.120133972167968,
      "step": 660
    },
    {
      "epoch": 0.1131947964183139,
      "grad_norm": 0.8237811923027039,
      "learning_rate": 0.00010035,
      "loss": 7.082501220703125,
      "step": 670
    },
    {
      "epoch": 0.11488427099172158,
      "grad_norm": 0.8586721420288086,
      "learning_rate": 0.00010185,
      "loss": 7.066880798339843,
      "step": 680
    },
    {
      "epoch": 0.11657374556512924,
      "grad_norm": 0.9714040160179138,
      "learning_rate": 0.00010334999999999998,
      "loss": 7.058338928222656,
      "step": 690
    },
    {
      "epoch": 0.11826322013853692,
      "grad_norm": 0.8379534482955933,
      "learning_rate": 0.00010484999999999999,
      "loss": 7.033222198486328,
      "step": 700
    },
    {
      "epoch": 0.11995269471194459,
      "grad_norm": 1.147356629371643,
      "learning_rate": 0.00010634999999999998,
      "loss": 7.022679138183594,
      "step": 710
    },
    {
      "epoch": 0.12164216928535225,
      "grad_norm": 0.946237325668335,
      "learning_rate": 0.00010784999999999999,
      "loss": 6.991328430175781,
      "step": 720
    },
    {
      "epoch": 0.12333164385875993,
      "grad_norm": 1.0189383029937744,
      "learning_rate": 0.00010934999999999999,
      "loss": 6.9945930480957035,
      "step": 730
    },
    {
      "epoch": 0.1250211184321676,
      "grad_norm": 1.0500218868255615,
      "learning_rate": 0.00011084999999999998,
      "loss": 6.963920593261719,
      "step": 740
    },
    {
      "epoch": 0.12671059300557527,
      "grad_norm": 0.9184631109237671,
      "learning_rate": 0.00011235,
      "loss": 6.966143798828125,
      "step": 750
    },
    {
      "epoch": 0.12840006757898292,
      "grad_norm": 0.7820301651954651,
      "learning_rate": 0.00011384999999999999,
      "loss": 6.938487243652344,
      "step": 760
    },
    {
      "epoch": 0.1300895421523906,
      "grad_norm": 0.861544668674469,
      "learning_rate": 0.00011535,
      "loss": 6.920912170410157,
      "step": 770
    },
    {
      "epoch": 0.13177901672579828,
      "grad_norm": 0.7475805878639221,
      "learning_rate": 0.00011685,
      "loss": 6.9139961242675785,
      "step": 780
    },
    {
      "epoch": 0.13346849129920596,
      "grad_norm": 1.1264002323150635,
      "learning_rate": 0.00011834999999999999,
      "loss": 6.893434143066406,
      "step": 790
    },
    {
      "epoch": 0.1351579658726136,
      "grad_norm": 0.869057834148407,
      "learning_rate": 0.00011985,
      "loss": 6.887288665771484,
      "step": 800
    },
    {
      "epoch": 0.1368474404460213,
      "grad_norm": 0.831230878829956,
      "learning_rate": 0.00012135,
      "loss": 6.863740539550781,
      "step": 810
    },
    {
      "epoch": 0.13853691501942897,
      "grad_norm": 1.068192720413208,
      "learning_rate": 0.00012284999999999998,
      "loss": 6.873618316650391,
      "step": 820
    },
    {
      "epoch": 0.14022638959283662,
      "grad_norm": 0.9752544164657593,
      "learning_rate": 0.00012435,
      "loss": 6.84736328125,
      "step": 830
    },
    {
      "epoch": 0.1419158641662443,
      "grad_norm": 0.9449293613433838,
      "learning_rate": 0.00012585,
      "loss": 6.829524230957031,
      "step": 840
    },
    {
      "epoch": 0.14360533873965198,
      "grad_norm": 1.104444980621338,
      "learning_rate": 0.00012734999999999998,
      "loss": 6.8291679382324215,
      "step": 850
    },
    {
      "epoch": 0.14529481331305963,
      "grad_norm": 0.9382540583610535,
      "learning_rate": 0.00012885,
      "loss": 6.8139289855957035,
      "step": 860
    },
    {
      "epoch": 0.1469842878864673,
      "grad_norm": 0.7313889861106873,
      "learning_rate": 0.00013035,
      "loss": 6.798196411132812,
      "step": 870
    },
    {
      "epoch": 0.148673762459875,
      "grad_norm": 0.7715932130813599,
      "learning_rate": 0.00013184999999999998,
      "loss": 6.805503845214844,
      "step": 880
    },
    {
      "epoch": 0.15036323703328264,
      "grad_norm": 1.0334839820861816,
      "learning_rate": 0.00013335,
      "loss": 6.745892333984375,
      "step": 890
    },
    {
      "epoch": 0.15205271160669032,
      "grad_norm": 1.1189385652542114,
      "learning_rate": 0.00013485,
      "loss": 6.769204711914062,
      "step": 900
    },
    {
      "epoch": 0.153742186180098,
      "grad_norm": 1.289933681488037,
      "learning_rate": 0.00013634999999999998,
      "loss": 6.737556457519531,
      "step": 910
    },
    {
      "epoch": 0.15543166075350565,
      "grad_norm": 1.0107234716415405,
      "learning_rate": 0.00013785,
      "loss": 6.7412353515625,
      "step": 920
    },
    {
      "epoch": 0.15712113532691332,
      "grad_norm": 0.9233148097991943,
      "learning_rate": 0.00013935,
      "loss": 6.707360076904297,
      "step": 930
    },
    {
      "epoch": 0.158810609900321,
      "grad_norm": 0.9500652551651001,
      "learning_rate": 0.00014084999999999998,
      "loss": 6.697336578369141,
      "step": 940
    },
    {
      "epoch": 0.16050008447372868,
      "grad_norm": 1.0929033756256104,
      "learning_rate": 0.00014235,
      "loss": 6.673794555664062,
      "step": 950
    },
    {
      "epoch": 0.16218955904713633,
      "grad_norm": 1.0387179851531982,
      "learning_rate": 0.00014384999999999997,
      "loss": 6.667636871337891,
      "step": 960
    },
    {
      "epoch": 0.163879033620544,
      "grad_norm": 1.1298182010650635,
      "learning_rate": 0.00014534999999999998,
      "loss": 6.645402526855468,
      "step": 970
    },
    {
      "epoch": 0.1655685081939517,
      "grad_norm": 0.9608763456344604,
      "learning_rate": 0.00014685,
      "loss": 6.658983612060547,
      "step": 980
    },
    {
      "epoch": 0.16725798276735934,
      "grad_norm": 1.2303314208984375,
      "learning_rate": 0.00014834999999999997,
      "loss": 6.633333587646485,
      "step": 990
    },
    {
      "epoch": 0.16894745734076702,
      "grad_norm": 0.9978023767471313,
      "learning_rate": 0.00014984999999999998,
      "loss": 6.6367958068847654,
      "step": 1000
    },
    {
      "epoch": 0.16894745734076702,
      "eval_loss": 6.617567539215088,
      "eval_runtime": 3.6651,
      "eval_samples_per_second": 272.845,
      "eval_steps_per_second": 5.73,
      "step": 1000
    },
    {
      "epoch": 0.1706369319141747,
      "grad_norm": 0.9963025450706482,
      "learning_rate": 0.00015134999999999997,
      "loss": 6.613154602050781,
      "step": 1010
    },
    {
      "epoch": 0.17232640648758235,
      "grad_norm": 0.872097909450531,
      "learning_rate": 0.00015284999999999997,
      "loss": 6.613529968261719,
      "step": 1020
    },
    {
      "epoch": 0.17401588106099003,
      "grad_norm": 1.2607650756835938,
      "learning_rate": 0.00015434999999999998,
      "loss": 6.587220001220703,
      "step": 1030
    },
    {
      "epoch": 0.1757053556343977,
      "grad_norm": 1.0194809436798096,
      "learning_rate": 0.00015584999999999997,
      "loss": 6.585498046875,
      "step": 1040
    },
    {
      "epoch": 0.17739483020780536,
      "grad_norm": 0.9153720736503601,
      "learning_rate": 0.00015734999999999998,
      "loss": 6.5845489501953125,
      "step": 1050
    },
    {
      "epoch": 0.17908430478121304,
      "grad_norm": 1.1903005838394165,
      "learning_rate": 0.00015884999999999999,
      "loss": 6.566903686523437,
      "step": 1060
    },
    {
      "epoch": 0.18077377935462072,
      "grad_norm": 0.9262056350708008,
      "learning_rate": 0.00016034999999999997,
      "loss": 6.520059204101562,
      "step": 1070
    },
    {
      "epoch": 0.1824632539280284,
      "grad_norm": 1.0881860256195068,
      "learning_rate": 0.00016184999999999998,
      "loss": 6.543362426757812,
      "step": 1080
    },
    {
      "epoch": 0.18415272850143605,
      "grad_norm": 0.9753679633140564,
      "learning_rate": 0.00016334999999999999,
      "loss": 6.528910064697266,
      "step": 1090
    },
    {
      "epoch": 0.18584220307484373,
      "grad_norm": 1.2809370756149292,
      "learning_rate": 0.00016485,
      "loss": 6.49705810546875,
      "step": 1100
    },
    {
      "epoch": 0.1875316776482514,
      "grad_norm": 1.0647395849227905,
      "learning_rate": 0.00016634999999999998,
      "loss": 6.508152008056641,
      "step": 1110
    },
    {
      "epoch": 0.18922115222165906,
      "grad_norm": 0.9427017569541931,
      "learning_rate": 0.00016785,
      "loss": 6.492857360839844,
      "step": 1120
    },
    {
      "epoch": 0.19091062679506673,
      "grad_norm": 1.1307021379470825,
      "learning_rate": 0.00016935,
      "loss": 6.474656677246093,
      "step": 1130
    },
    {
      "epoch": 0.1926001013684744,
      "grad_norm": 1.182411789894104,
      "learning_rate": 0.00017084999999999998,
      "loss": 6.457868194580078,
      "step": 1140
    },
    {
      "epoch": 0.19428957594188206,
      "grad_norm": 1.1442158222198486,
      "learning_rate": 0.00017235,
      "loss": 6.443910217285156,
      "step": 1150
    },
    {
      "epoch": 0.19597905051528974,
      "grad_norm": 1.2637932300567627,
      "learning_rate": 0.00017385,
      "loss": 6.428031158447266,
      "step": 1160
    },
    {
      "epoch": 0.19766852508869742,
      "grad_norm": 1.334306001663208,
      "learning_rate": 0.00017534999999999998,
      "loss": 6.415740966796875,
      "step": 1170
    },
    {
      "epoch": 0.19935799966210507,
      "grad_norm": 0.882560670375824,
      "learning_rate": 0.00017685,
      "loss": 6.413926696777343,
      "step": 1180
    },
    {
      "epoch": 0.20104747423551275,
      "grad_norm": 0.9657256603240967,
      "learning_rate": 0.00017835,
      "loss": 6.425054931640625,
      "step": 1190
    },
    {
      "epoch": 0.20273694880892043,
      "grad_norm": 1.0196014642715454,
      "learning_rate": 0.00017984999999999998,
      "loss": 6.391595077514649,
      "step": 1200
    },
    {
      "epoch": 0.2044264233823281,
      "grad_norm": 1.297837257385254,
      "learning_rate": 0.00018135,
      "loss": 6.382472991943359,
      "step": 1210
    },
    {
      "epoch": 0.20611589795573576,
      "grad_norm": 1.1288139820098877,
      "learning_rate": 0.00018285,
      "loss": 6.358099746704101,
      "step": 1220
    },
    {
      "epoch": 0.20780537252914344,
      "grad_norm": 0.9396995306015015,
      "learning_rate": 0.00018435,
      "loss": 6.355449676513672,
      "step": 1230
    },
    {
      "epoch": 0.20949484710255112,
      "grad_norm": 1.1936787366867065,
      "learning_rate": 0.00018585,
      "loss": 6.356659698486328,
      "step": 1240
    },
    {
      "epoch": 0.21118432167595877,
      "grad_norm": 0.9550564289093018,
      "learning_rate": 0.00018735,
      "loss": 6.337493515014648,
      "step": 1250
    },
    {
      "epoch": 0.21287379624936645,
      "grad_norm": 1.2012646198272705,
      "learning_rate": 0.00018884999999999996,
      "loss": 6.317781829833985,
      "step": 1260
    },
    {
      "epoch": 0.21456327082277413,
      "grad_norm": 1.0816755294799805,
      "learning_rate": 0.00019034999999999996,
      "loss": 6.316750335693359,
      "step": 1270
    },
    {
      "epoch": 0.21625274539618178,
      "grad_norm": 1.3777987957000732,
      "learning_rate": 0.00019184999999999997,
      "loss": 6.3194934844970705,
      "step": 1280
    },
    {
      "epoch": 0.21794221996958946,
      "grad_norm": 1.187603235244751,
      "learning_rate": 0.00019334999999999998,
      "loss": 6.30432357788086,
      "step": 1290
    },
    {
      "epoch": 0.21963169454299714,
      "grad_norm": 1.0069150924682617,
      "learning_rate": 0.00019484999999999997,
      "loss": 6.2757713317871096,
      "step": 1300
    },
    {
      "epoch": 0.2213211691164048,
      "grad_norm": 1.2410210371017456,
      "learning_rate": 0.00019634999999999998,
      "loss": 6.2698211669921875,
      "step": 1310
    },
    {
      "epoch": 0.22301064368981247,
      "grad_norm": 1.1892989873886108,
      "learning_rate": 0.00019784999999999998,
      "loss": 6.2431591033935545,
      "step": 1320
    },
    {
      "epoch": 0.22470011826322014,
      "grad_norm": 1.1054743528366089,
      "learning_rate": 0.00019934999999999997,
      "loss": 6.26300163269043,
      "step": 1330
    },
    {
      "epoch": 0.2263895928366278,
      "grad_norm": 1.145757794380188,
      "learning_rate": 0.00020084999999999998,
      "loss": 6.226350021362305,
      "step": 1340
    },
    {
      "epoch": 0.22807906741003547,
      "grad_norm": 1.0067166090011597,
      "learning_rate": 0.00020234999999999999,
      "loss": 6.2175750732421875,
      "step": 1350
    },
    {
      "epoch": 0.22976854198344315,
      "grad_norm": 1.5041327476501465,
      "learning_rate": 0.00020384999999999997,
      "loss": 6.191579055786133,
      "step": 1360
    },
    {
      "epoch": 0.23145801655685083,
      "grad_norm": 1.2780109643936157,
      "learning_rate": 0.00020534999999999998,
      "loss": 6.204021835327149,
      "step": 1370
    },
    {
      "epoch": 0.23314749113025848,
      "grad_norm": 1.1531580686569214,
      "learning_rate": 0.00020684999999999999,
      "loss": 6.191404342651367,
      "step": 1380
    },
    {
      "epoch": 0.23483696570366616,
      "grad_norm": 1.056857705116272,
      "learning_rate": 0.00020835,
      "loss": 6.17081298828125,
      "step": 1390
    },
    {
      "epoch": 0.23652644027707384,
      "grad_norm": 1.1238850355148315,
      "learning_rate": 0.00020984999999999998,
      "loss": 6.153195190429687,
      "step": 1400
    },
    {
      "epoch": 0.2382159148504815,
      "grad_norm": 1.2115790843963623,
      "learning_rate": 0.00021135,
      "loss": 6.157797622680664,
      "step": 1410
    },
    {
      "epoch": 0.23990538942388917,
      "grad_norm": 1.1303883790969849,
      "learning_rate": 0.00021285,
      "loss": 6.119416809082031,
      "step": 1420
    },
    {
      "epoch": 0.24159486399729685,
      "grad_norm": 1.2523441314697266,
      "learning_rate": 0.00021434999999999998,
      "loss": 6.133832550048828,
      "step": 1430
    },
    {
      "epoch": 0.2432843385707045,
      "grad_norm": 1.1120916604995728,
      "learning_rate": 0.00021585,
      "loss": 6.122848129272461,
      "step": 1440
    },
    {
      "epoch": 0.24497381314411218,
      "grad_norm": 1.239675521850586,
      "learning_rate": 0.00021735,
      "loss": 6.106191253662109,
      "step": 1450
    },
    {
      "epoch": 0.24666328771751986,
      "grad_norm": 1.1382733583450317,
      "learning_rate": 0.00021884999999999998,
      "loss": 6.0912620544433596,
      "step": 1460
    },
    {
      "epoch": 0.2483527622909275,
      "grad_norm": 1.3199714422225952,
      "learning_rate": 0.00022035,
      "loss": 6.09831428527832,
      "step": 1470
    },
    {
      "epoch": 0.2500422368643352,
      "grad_norm": 1.2705349922180176,
      "learning_rate": 0.00022185,
      "loss": 6.078111267089843,
      "step": 1480
    },
    {
      "epoch": 0.25173171143774287,
      "grad_norm": 1.436306357383728,
      "learning_rate": 0.00022335,
      "loss": 6.058963012695313,
      "step": 1490
    },
    {
      "epoch": 0.25342118601115055,
      "grad_norm": 1.179898977279663,
      "learning_rate": 0.00022485,
      "loss": 6.029299545288086,
      "step": 1500
    },
    {
      "epoch": 0.25342118601115055,
      "eval_loss": 6.033608436584473,
      "eval_runtime": 3.6064,
      "eval_samples_per_second": 277.282,
      "eval_steps_per_second": 5.823,
      "step": 1500
    },
    {
      "epoch": 0.2551106605845582,
      "grad_norm": 1.3389363288879395,
      "learning_rate": 0.00022634999999999997,
      "loss": 6.027260589599609,
      "step": 1510
    },
    {
      "epoch": 0.25680013515796585,
      "grad_norm": 1.2689851522445679,
      "learning_rate": 0.00022784999999999995,
      "loss": 6.00293083190918,
      "step": 1520
    },
    {
      "epoch": 0.2584896097313735,
      "grad_norm": 1.4860210418701172,
      "learning_rate": 0.00022934999999999996,
      "loss": 5.998868942260742,
      "step": 1530
    },
    {
      "epoch": 0.2601790843047812,
      "grad_norm": 1.2490425109863281,
      "learning_rate": 0.00023084999999999997,
      "loss": 5.984478759765625,
      "step": 1540
    },
    {
      "epoch": 0.2618685588781889,
      "grad_norm": 1.5586382150650024,
      "learning_rate": 0.00023234999999999998,
      "loss": 5.9672401428222654,
      "step": 1550
    },
    {
      "epoch": 0.26355803345159656,
      "grad_norm": 1.3526853322982788,
      "learning_rate": 0.00023384999999999997,
      "loss": 5.982438278198242,
      "step": 1560
    },
    {
      "epoch": 0.26524750802500424,
      "grad_norm": 1.3406753540039062,
      "learning_rate": 0.00023534999999999997,
      "loss": 5.938652801513672,
      "step": 1570
    },
    {
      "epoch": 0.2669369825984119,
      "grad_norm": 1.0397038459777832,
      "learning_rate": 0.00023684999999999998,
      "loss": 5.920218658447266,
      "step": 1580
    },
    {
      "epoch": 0.26862645717181954,
      "grad_norm": 1.7000986337661743,
      "learning_rate": 0.00023834999999999997,
      "loss": 5.896316146850586,
      "step": 1590
    },
    {
      "epoch": 0.2703159317452272,
      "grad_norm": 1.1729341745376587,
      "learning_rate": 0.00023984999999999998,
      "loss": 5.8752281188964846,
      "step": 1600
    },
    {
      "epoch": 0.2720054063186349,
      "grad_norm": 1.3115921020507812,
      "learning_rate": 0.00024134999999999998,
      "loss": 5.877028274536133,
      "step": 1610
    },
    {
      "epoch": 0.2736948808920426,
      "grad_norm": 1.5481823682785034,
      "learning_rate": 0.00024284999999999997,
      "loss": 5.863247299194336,
      "step": 1620
    },
    {
      "epoch": 0.27538435546545026,
      "grad_norm": 1.4173649549484253,
      "learning_rate": 0.00024435,
      "loss": 5.848538970947265,
      "step": 1630
    },
    {
      "epoch": 0.27707383003885794,
      "grad_norm": 1.2587963342666626,
      "learning_rate": 0.00024585,
      "loss": 5.841713333129883,
      "step": 1640
    },
    {
      "epoch": 0.27876330461226556,
      "grad_norm": 1.0922702550888062,
      "learning_rate": 0.00024734999999999997,
      "loss": 5.8486980438232425,
      "step": 1650
    },
    {
      "epoch": 0.28045277918567324,
      "grad_norm": 1.6068239212036133,
      "learning_rate": 0.00024885,
      "loss": 5.819171142578125,
      "step": 1660
    },
    {
      "epoch": 0.2821422537590809,
      "grad_norm": 1.5260576009750366,
      "learning_rate": 0.00025035,
      "loss": 5.809968566894531,
      "step": 1670
    },
    {
      "epoch": 0.2838317283324886,
      "grad_norm": 1.2246356010437012,
      "learning_rate": 0.00025184999999999997,
      "loss": 5.788796997070312,
      "step": 1680
    },
    {
      "epoch": 0.2855212029058963,
      "grad_norm": 1.0366030931472778,
      "learning_rate": 0.00025335,
      "loss": 5.78180160522461,
      "step": 1690
    },
    {
      "epoch": 0.28721067747930396,
      "grad_norm": 1.2072358131408691,
      "learning_rate": 0.00025485,
      "loss": 5.770789337158203,
      "step": 1700
    },
    {
      "epoch": 0.28890015205271163,
      "grad_norm": 1.3359684944152832,
      "learning_rate": 0.00025634999999999997,
      "loss": 5.737417221069336,
      "step": 1710
    },
    {
      "epoch": 0.29058962662611926,
      "grad_norm": 1.355406403541565,
      "learning_rate": 0.00025785,
      "loss": 5.725430297851562,
      "step": 1720
    },
    {
      "epoch": 0.29227910119952694,
      "grad_norm": 1.1998307704925537,
      "learning_rate": 0.00025935,
      "loss": 5.723165130615234,
      "step": 1730
    },
    {
      "epoch": 0.2939685757729346,
      "grad_norm": 1.0525386333465576,
      "learning_rate": 0.00026084999999999997,
      "loss": 5.720573043823242,
      "step": 1740
    },
    {
      "epoch": 0.2956580503463423,
      "grad_norm": 1.2880501747131348,
      "learning_rate": 0.00026235,
      "loss": 5.684521102905274,
      "step": 1750
    },
    {
      "epoch": 0.29734752491975,
      "grad_norm": 1.2246838808059692,
      "learning_rate": 0.00026384999999999994,
      "loss": 5.670655059814453,
      "step": 1760
    },
    {
      "epoch": 0.29903699949315765,
      "grad_norm": 1.2167463302612305,
      "learning_rate": 0.00026534999999999997,
      "loss": 5.690992736816407,
      "step": 1770
    },
    {
      "epoch": 0.3007264740665653,
      "grad_norm": 1.2467341423034668,
      "learning_rate": 0.00026684999999999995,
      "loss": 5.694464492797851,
      "step": 1780
    },
    {
      "epoch": 0.30241594863997295,
      "grad_norm": 1.2740100622177124,
      "learning_rate": 0.00026835,
      "loss": 5.679082870483398,
      "step": 1790
    },
    {
      "epoch": 0.30410542321338063,
      "grad_norm": 1.2217073440551758,
      "learning_rate": 0.00026984999999999997,
      "loss": 5.650615692138672,
      "step": 1800
    },
    {
      "epoch": 0.3057948977867883,
      "grad_norm": 1.1172698736190796,
      "learning_rate": 0.00027134999999999995,
      "loss": 5.651753234863281,
      "step": 1810
    },
    {
      "epoch": 0.307484372360196,
      "grad_norm": 1.1706960201263428,
      "learning_rate": 0.00027285,
      "loss": 5.6512096405029295,
      "step": 1820
    },
    {
      "epoch": 0.30917384693360367,
      "grad_norm": 0.91384357213974,
      "learning_rate": 0.00027435,
      "loss": 5.63836784362793,
      "step": 1830
    },
    {
      "epoch": 0.3108633215070113,
      "grad_norm": 1.1929048299789429,
      "learning_rate": 0.00027584999999999996,
      "loss": 5.628775787353516,
      "step": 1840
    },
    {
      "epoch": 0.31255279608041897,
      "grad_norm": 1.023672103881836,
      "learning_rate": 0.00027735,
      "loss": 5.616031265258789,
      "step": 1850
    },
    {
      "epoch": 0.31424227065382665,
      "grad_norm": 1.1450271606445312,
      "learning_rate": 0.00027885,
      "loss": 5.612253952026367,
      "step": 1860
    },
    {
      "epoch": 0.31593174522723433,
      "grad_norm": 1.0316193103790283,
      "learning_rate": 0.00028034999999999996,
      "loss": 5.577928161621093,
      "step": 1870
    },
    {
      "epoch": 0.317621219800642,
      "grad_norm": 1.1516318321228027,
      "learning_rate": 0.00028185,
      "loss": 5.589142227172852,
      "step": 1880
    },
    {
      "epoch": 0.3193106943740497,
      "grad_norm": 1.426249384880066,
      "learning_rate": 0.00028335,
      "loss": 5.594329071044922,
      "step": 1890
    },
    {
      "epoch": 0.32100016894745736,
      "grad_norm": 1.0666186809539795,
      "learning_rate": 0.00028484999999999996,
      "loss": 5.582658386230468,
      "step": 1900
    },
    {
      "epoch": 0.322689643520865,
      "grad_norm": 0.8879145979881287,
      "learning_rate": 0.00028635,
      "loss": 5.542075347900391,
      "step": 1910
    },
    {
      "epoch": 0.32437911809427267,
      "grad_norm": 1.2985228300094604,
      "learning_rate": 0.00028785,
      "loss": 5.572188949584961,
      "step": 1920
    },
    {
      "epoch": 0.32606859266768035,
      "grad_norm": 1.1801198720932007,
      "learning_rate": 0.00028934999999999996,
      "loss": 5.531465530395508,
      "step": 1930
    },
    {
      "epoch": 0.327758067241088,
      "grad_norm": 1.3345341682434082,
      "learning_rate": 0.00029085,
      "loss": 5.5121315002441404,
      "step": 1940
    },
    {
      "epoch": 0.3294475418144957,
      "grad_norm": 0.9832890629768372,
      "learning_rate": 0.00029235,
      "loss": 5.515644073486328,
      "step": 1950
    },
    {
      "epoch": 0.3311370163879034,
      "grad_norm": 1.379388689994812,
      "learning_rate": 0.00029384999999999996,
      "loss": 5.5223854064941404,
      "step": 1960
    },
    {
      "epoch": 0.332826490961311,
      "grad_norm": 1.0441769361495972,
      "learning_rate": 0.00029535,
      "loss": 5.502047729492188,
      "step": 1970
    },
    {
      "epoch": 0.3345159655347187,
      "grad_norm": 1.0386887788772583,
      "learning_rate": 0.00029685,
      "loss": 5.521197128295898,
      "step": 1980
    },
    {
      "epoch": 0.33620544010812636,
      "grad_norm": 0.8223176598548889,
      "learning_rate": 0.00029835,
      "loss": 5.479276275634765,
      "step": 1990
    },
    {
      "epoch": 0.33789491468153404,
      "grad_norm": 1.2531520128250122,
      "learning_rate": 0.00029985,
      "loss": 5.487053298950196,
      "step": 2000
    },
    {
      "epoch": 0.33789491468153404,
      "eval_loss": 5.460203170776367,
      "eval_runtime": 3.9099,
      "eval_samples_per_second": 255.761,
      "eval_steps_per_second": 5.371,
      "step": 2000
    },
    {
      "epoch": 0.3395843892549417,
      "grad_norm": 1.0625675916671753,
      "learning_rate": 0.0002999993805131495,
      "loss": 5.482983016967774,
      "step": 2010
    },
    {
      "epoch": 0.3412738638283494,
      "grad_norm": 0.9310702681541443,
      "learning_rate": 0.00029999723908369233,
      "loss": 5.477756500244141,
      "step": 2020
    },
    {
      "epoch": 0.3429633384017571,
      "grad_norm": 0.8275931477546692,
      "learning_rate": 0.0002999935680854744,
      "loss": 5.4467018127441404,
      "step": 2030
    },
    {
      "epoch": 0.3446528129751647,
      "grad_norm": 0.8972215056419373,
      "learning_rate": 0.00029998836755593,
      "loss": 5.415990829467773,
      "step": 2040
    },
    {
      "epoch": 0.3463422875485724,
      "grad_norm": 1.0727229118347168,
      "learning_rate": 0.00029998163754809044,
      "loss": 5.403407287597656,
      "step": 2050
    },
    {
      "epoch": 0.34803176212198006,
      "grad_norm": 1.0068520307540894,
      "learning_rate": 0.0002999733781305839,
      "loss": 5.4188987731933596,
      "step": 2060
    },
    {
      "epoch": 0.34972123669538774,
      "grad_norm": 0.9327341914176941,
      "learning_rate": 0.00029996358938763406,
      "loss": 5.406315612792969,
      "step": 2070
    },
    {
      "epoch": 0.3514107112687954,
      "grad_norm": 1.022828221321106,
      "learning_rate": 0.0002999522714190599,
      "loss": 5.410961532592774,
      "step": 2080
    },
    {
      "epoch": 0.3531001858422031,
      "grad_norm": 0.8379955887794495,
      "learning_rate": 0.0002999394243402743,
      "loss": 5.411350250244141,
      "step": 2090
    },
    {
      "epoch": 0.3547896604156107,
      "grad_norm": 0.8905497193336487,
      "learning_rate": 0.00029992504828228283,
      "loss": 5.384899520874024,
      "step": 2100
    },
    {
      "epoch": 0.3564791349890184,
      "grad_norm": 0.7869957685470581,
      "learning_rate": 0.00029990914339168286,
      "loss": 5.391331481933594,
      "step": 2110
    },
    {
      "epoch": 0.3581686095624261,
      "grad_norm": 0.7781967520713806,
      "learning_rate": 0.00029989170983066126,
      "loss": 5.365080261230469,
      "step": 2120
    },
    {
      "epoch": 0.35985808413583376,
      "grad_norm": 0.8611620664596558,
      "learning_rate": 0.0002998727477769937,
      "loss": 5.367116546630859,
      "step": 2130
    },
    {
      "epoch": 0.36154755870924143,
      "grad_norm": 0.8369846940040588,
      "learning_rate": 0.0002998522574240421,
      "loss": 5.361904525756836,
      "step": 2140
    },
    {
      "epoch": 0.3632370332826491,
      "grad_norm": 0.893395721912384,
      "learning_rate": 0.00029983023898075305,
      "loss": 5.338259887695313,
      "step": 2150
    },
    {
      "epoch": 0.3649265078560568,
      "grad_norm": 0.9806540012359619,
      "learning_rate": 0.00029980669267165545,
      "loss": 5.33393440246582,
      "step": 2160
    },
    {
      "epoch": 0.3666159824294644,
      "grad_norm": 0.789153516292572,
      "learning_rate": 0.0002997816187368584,
      "loss": 5.347314834594727,
      "step": 2170
    },
    {
      "epoch": 0.3683054570028721,
      "grad_norm": 0.731369137763977,
      "learning_rate": 0.00029975501743204866,
      "loss": 5.322664260864258,
      "step": 2180
    },
    {
      "epoch": 0.3699949315762798,
      "grad_norm": 0.6811886429786682,
      "learning_rate": 0.00029972688902848803,
      "loss": 5.326079177856445,
      "step": 2190
    },
    {
      "epoch": 0.37168440614968745,
      "grad_norm": 0.8143295645713806,
      "learning_rate": 0.0002996972338130106,
      "loss": 5.30379638671875,
      "step": 2200
    },
    {
      "epoch": 0.37337388072309513,
      "grad_norm": 0.8854978680610657,
      "learning_rate": 0.00029966605208801996,
      "loss": 5.301242828369141,
      "step": 2210
    },
    {
      "epoch": 0.3750633552965028,
      "grad_norm": 0.77631014585495,
      "learning_rate": 0.0002996333441714859,
      "loss": 5.294522476196289,
      "step": 2220
    },
    {
      "epoch": 0.37675282986991043,
      "grad_norm": 0.7743359208106995,
      "learning_rate": 0.00029959911039694127,
      "loss": 5.313030624389649,
      "step": 2230
    },
    {
      "epoch": 0.3784423044433181,
      "grad_norm": 0.8531479239463806,
      "learning_rate": 0.00029956335111347855,
      "loss": 5.275916671752929,
      "step": 2240
    },
    {
      "epoch": 0.3801317790167258,
      "grad_norm": 0.722363293170929,
      "learning_rate": 0.0002995260666857463,
      "loss": 5.2906639099121096,
      "step": 2250
    },
    {
      "epoch": 0.38182125359013347,
      "grad_norm": 0.7797225713729858,
      "learning_rate": 0.00029948725749394563,
      "loss": 5.2658641815185545,
      "step": 2260
    },
    {
      "epoch": 0.38351072816354115,
      "grad_norm": 0.8231165409088135,
      "learning_rate": 0.00029944692393382586,
      "loss": 5.2770263671875,
      "step": 2270
    },
    {
      "epoch": 0.3852002027369488,
      "grad_norm": 0.8083261847496033,
      "learning_rate": 0.000299405066416681,
      "loss": 5.277169799804687,
      "step": 2280
    },
    {
      "epoch": 0.3868896773103565,
      "grad_norm": 0.8675849437713623,
      "learning_rate": 0.0002993616853693452,
      "loss": 5.258210754394531,
      "step": 2290
    },
    {
      "epoch": 0.38857915188376413,
      "grad_norm": 0.7585932016372681,
      "learning_rate": 0.0002993167812341886,
      "loss": 5.252765655517578,
      "step": 2300
    },
    {
      "epoch": 0.3902686264571718,
      "grad_norm": 0.8213605284690857,
      "learning_rate": 0.0002992703544691127,
      "loss": 5.222419357299804,
      "step": 2310
    },
    {
      "epoch": 0.3919581010305795,
      "grad_norm": 0.7984234690666199,
      "learning_rate": 0.00029922240554754577,
      "loss": 5.227847671508789,
      "step": 2320
    },
    {
      "epoch": 0.39364757560398717,
      "grad_norm": 0.8216149806976318,
      "learning_rate": 0.00029917293495843793,
      "loss": 5.215268325805664,
      "step": 2330
    },
    {
      "epoch": 0.39533705017739484,
      "grad_norm": 0.7992113828659058,
      "learning_rate": 0.0002991219432062562,
      "loss": 5.251160049438477,
      "step": 2340
    },
    {
      "epoch": 0.3970265247508025,
      "grad_norm": 0.7669650316238403,
      "learning_rate": 0.0002990694308109795,
      "loss": 5.255714797973633,
      "step": 2350
    },
    {
      "epoch": 0.39871599932421015,
      "grad_norm": 0.7685340046882629,
      "learning_rate": 0.0002990153983080932,
      "loss": 5.2186332702636715,
      "step": 2360
    },
    {
      "epoch": 0.4004054738976178,
      "grad_norm": 0.8289806246757507,
      "learning_rate": 0.0002989598462485835,
      "loss": 5.2316020965576175,
      "step": 2370
    },
    {
      "epoch": 0.4020949484710255,
      "grad_norm": 0.7260857224464417,
      "learning_rate": 0.00029890277519893215,
      "loss": 5.210884857177734,
      "step": 2380
    },
    {
      "epoch": 0.4037844230444332,
      "grad_norm": 0.6450658440589905,
      "learning_rate": 0.0002988441857411106,
      "loss": 5.194115066528321,
      "step": 2390
    },
    {
      "epoch": 0.40547389761784086,
      "grad_norm": 0.723818838596344,
      "learning_rate": 0.0002987840784725737,
      "loss": 5.197711563110351,
      "step": 2400
    },
    {
      "epoch": 0.40716337219124854,
      "grad_norm": 0.8113153576850891,
      "learning_rate": 0.0002987224540062542,
      "loss": 5.196290588378906,
      "step": 2410
    },
    {
      "epoch": 0.4088528467646562,
      "grad_norm": 0.8224965929985046,
      "learning_rate": 0.00029865931297055605,
      "loss": 5.174480819702149,
      "step": 2420
    },
    {
      "epoch": 0.41054232133806384,
      "grad_norm": 0.9786369204521179,
      "learning_rate": 0.00029859465600934814,
      "loss": 5.19611701965332,
      "step": 2430
    },
    {
      "epoch": 0.4122317959114715,
      "grad_norm": 0.8020685911178589,
      "learning_rate": 0.0002985284837819577,
      "loss": 5.181368637084961,
      "step": 2440
    },
    {
      "epoch": 0.4139212704848792,
      "grad_norm": 0.7282792329788208,
      "learning_rate": 0.0002984607969631636,
      "loss": 5.1728168487548825,
      "step": 2450
    },
    {
      "epoch": 0.4156107450582869,
      "grad_norm": 0.6869542598724365,
      "learning_rate": 0.00029839159624318954,
      "loss": 5.172641372680664,
      "step": 2460
    },
    {
      "epoch": 0.41730021963169456,
      "grad_norm": 0.8235262632369995,
      "learning_rate": 0.00029832088232769694,
      "loss": 5.165771484375,
      "step": 2470
    },
    {
      "epoch": 0.41898969420510224,
      "grad_norm": 0.7626176476478577,
      "learning_rate": 0.0002982486559377776,
      "loss": 5.175928115844727,
      "step": 2480
    },
    {
      "epoch": 0.42067916877850986,
      "grad_norm": 0.636053740978241,
      "learning_rate": 0.0002981749178099467,
      "loss": 5.135253143310547,
      "step": 2490
    },
    {
      "epoch": 0.42236864335191754,
      "grad_norm": 0.6814470291137695,
      "learning_rate": 0.000298099668696135,
      "loss": 5.177354049682617,
      "step": 2500
    },
    {
      "epoch": 0.42236864335191754,
      "eval_loss": 5.138686656951904,
      "eval_runtime": 3.9981,
      "eval_samples_per_second": 250.119,
      "eval_steps_per_second": 5.253,
      "step": 2500
    },
    {
      "epoch": 0.4240581179253252,
      "grad_norm": 0.786521315574646,
      "learning_rate": 0.0002980229093636812,
      "loss": 5.136567687988281,
      "step": 2510
    },
    {
      "epoch": 0.4257475924987329,
      "grad_norm": 0.7561874389648438,
      "learning_rate": 0.00029794464059532426,
      "loss": 5.145055770874023,
      "step": 2520
    },
    {
      "epoch": 0.4274370670721406,
      "grad_norm": 0.6505213975906372,
      "learning_rate": 0.0002978648631891952,
      "loss": 5.145381164550781,
      "step": 2530
    },
    {
      "epoch": 0.42912654164554825,
      "grad_norm": 0.7278615832328796,
      "learning_rate": 0.0002977835779588093,
      "loss": 5.112863540649414,
      "step": 2540
    },
    {
      "epoch": 0.4308160162189559,
      "grad_norm": 0.6332527995109558,
      "learning_rate": 0.0002977007857330575,
      "loss": 5.129104995727539,
      "step": 2550
    },
    {
      "epoch": 0.43250549079236356,
      "grad_norm": 0.669188380241394,
      "learning_rate": 0.0002976164873561979,
      "loss": 5.100088500976563,
      "step": 2560
    },
    {
      "epoch": 0.43419496536577123,
      "grad_norm": 0.6842843294143677,
      "learning_rate": 0.0002975306836878474,
      "loss": 5.092770004272461,
      "step": 2570
    },
    {
      "epoch": 0.4358844399391789,
      "grad_norm": 0.7057438492774963,
      "learning_rate": 0.000297443375602973,
      "loss": 5.1130115509033205,
      "step": 2580
    },
    {
      "epoch": 0.4375739145125866,
      "grad_norm": 0.6845251321792603,
      "learning_rate": 0.0002973545639918824,
      "loss": 5.112728500366211,
      "step": 2590
    },
    {
      "epoch": 0.43926338908599427,
      "grad_norm": 0.6881667971611023,
      "learning_rate": 0.00029726424976021543,
      "loss": 5.095853042602539,
      "step": 2600
    },
    {
      "epoch": 0.44095286365940195,
      "grad_norm": 0.6859349608421326,
      "learning_rate": 0.0002971724338289346,
      "loss": 5.099851989746094,
      "step": 2610
    },
    {
      "epoch": 0.4426423382328096,
      "grad_norm": 0.6879841089248657,
      "learning_rate": 0.0002970791171343156,
      "loss": 5.113912582397461,
      "step": 2620
    },
    {
      "epoch": 0.44433181280621725,
      "grad_norm": 0.711805522441864,
      "learning_rate": 0.000296984300627938,
      "loss": 5.081494903564453,
      "step": 2630
    },
    {
      "epoch": 0.44602128737962493,
      "grad_norm": 0.675470232963562,
      "learning_rate": 0.00029688798527667537,
      "loss": 5.089406585693359,
      "step": 2640
    },
    {
      "epoch": 0.4477107619530326,
      "grad_norm": 0.6627302169799805,
      "learning_rate": 0.00029679017206268545,
      "loss": 5.071472930908203,
      "step": 2650
    },
    {
      "epoch": 0.4494002365264403,
      "grad_norm": 0.6572045087814331,
      "learning_rate": 0.00029669086198340014,
      "loss": 5.081936645507812,
      "step": 2660
    },
    {
      "epoch": 0.45108971109984797,
      "grad_norm": 0.8288828730583191,
      "learning_rate": 0.0002965900560515155,
      "loss": 5.082733535766602,
      "step": 2670
    },
    {
      "epoch": 0.4527791856732556,
      "grad_norm": 0.6581189036369324,
      "learning_rate": 0.00029648775529498103,
      "loss": 5.069281387329101,
      "step": 2680
    },
    {
      "epoch": 0.45446866024666327,
      "grad_norm": 0.737130880355835,
      "learning_rate": 0.00029638396075698953,
      "loss": 5.066775894165039,
      "step": 2690
    },
    {
      "epoch": 0.45615813482007095,
      "grad_norm": 0.7000970244407654,
      "learning_rate": 0.00029627867349596654,
      "loss": 5.027889251708984,
      "step": 2700
    },
    {
      "epoch": 0.4578476093934786,
      "grad_norm": 0.6418822407722473,
      "learning_rate": 0.000296171894585559,
      "loss": 5.060458374023438,
      "step": 2710
    },
    {
      "epoch": 0.4595370839668863,
      "grad_norm": 0.6689320802688599,
      "learning_rate": 0.00029606362511462494,
      "loss": 5.073564910888672,
      "step": 2720
    },
    {
      "epoch": 0.461226558540294,
      "grad_norm": 0.7149254083633423,
      "learning_rate": 0.000295953866187222,
      "loss": 5.058617782592774,
      "step": 2730
    },
    {
      "epoch": 0.46291603311370166,
      "grad_norm": 0.6966880559921265,
      "learning_rate": 0.00029584261892259627,
      "loss": 5.050143432617188,
      "step": 2740
    },
    {
      "epoch": 0.4646055076871093,
      "grad_norm": 0.6495580077171326,
      "learning_rate": 0.00029572988445517094,
      "loss": 5.034864807128907,
      "step": 2750
    },
    {
      "epoch": 0.46629498226051697,
      "grad_norm": 0.6543110609054565,
      "learning_rate": 0.0002956156639345346,
      "loss": 5.027247619628906,
      "step": 2760
    },
    {
      "epoch": 0.46798445683392464,
      "grad_norm": 0.6335380673408508,
      "learning_rate": 0.00029549995852542967,
      "loss": 5.0187946319580075,
      "step": 2770
    },
    {
      "epoch": 0.4696739314073323,
      "grad_norm": 0.6705760359764099,
      "learning_rate": 0.00029538276940774044,
      "loss": 5.034427261352539,
      "step": 2780
    },
    {
      "epoch": 0.47136340598074,
      "grad_norm": 0.6140398979187012,
      "learning_rate": 0.0002952640977764808,
      "loss": 5.027993011474609,
      "step": 2790
    },
    {
      "epoch": 0.4730528805541477,
      "grad_norm": 0.6979998350143433,
      "learning_rate": 0.00029514394484178266,
      "loss": 5.034260940551758,
      "step": 2800
    },
    {
      "epoch": 0.4747423551275553,
      "grad_norm": 0.6220052242279053,
      "learning_rate": 0.00029502231182888306,
      "loss": 5.024603652954101,
      "step": 2810
    },
    {
      "epoch": 0.476431829700963,
      "grad_norm": 0.6017596125602722,
      "learning_rate": 0.0002948991999781118,
      "loss": 5.012111663818359,
      "step": 2820
    },
    {
      "epoch": 0.47812130427437066,
      "grad_norm": 0.6071211695671082,
      "learning_rate": 0.000294774610544879,
      "loss": 5.029761886596679,
      "step": 2830
    },
    {
      "epoch": 0.47981077884777834,
      "grad_norm": 0.6241064071655273,
      "learning_rate": 0.0002946485447996621,
      "loss": 5.060077667236328,
      "step": 2840
    },
    {
      "epoch": 0.481500253421186,
      "grad_norm": 0.6929198503494263,
      "learning_rate": 0.0002945210040279928,
      "loss": 4.980299758911133,
      "step": 2850
    },
    {
      "epoch": 0.4831897279945937,
      "grad_norm": 0.6135720014572144,
      "learning_rate": 0.0002943919895304443,
      "loss": 4.994546508789062,
      "step": 2860
    },
    {
      "epoch": 0.4848792025680014,
      "grad_norm": 0.6381633281707764,
      "learning_rate": 0.0002942615026226179,
      "loss": 4.993935012817383,
      "step": 2870
    },
    {
      "epoch": 0.486568677141409,
      "grad_norm": 0.5814259648323059,
      "learning_rate": 0.0002941295446351292,
      "loss": 4.992059326171875,
      "step": 2880
    },
    {
      "epoch": 0.4882581517148167,
      "grad_norm": 0.5999816060066223,
      "learning_rate": 0.00029399611691359527,
      "loss": 4.977694320678711,
      "step": 2890
    },
    {
      "epoch": 0.48994762628822436,
      "grad_norm": 0.6875694990158081,
      "learning_rate": 0.0002938612208186202,
      "loss": 4.999196243286133,
      "step": 2900
    },
    {
      "epoch": 0.49163710086163204,
      "grad_norm": 0.6184036135673523,
      "learning_rate": 0.0002937248577257817,
      "loss": 5.010132217407227,
      "step": 2910
    },
    {
      "epoch": 0.4933265754350397,
      "grad_norm": 0.7426770329475403,
      "learning_rate": 0.0002935870290256169,
      "loss": 4.990754699707031,
      "step": 2920
    },
    {
      "epoch": 0.4950160500084474,
      "grad_norm": 0.6430733799934387,
      "learning_rate": 0.0002934477361236081,
      "loss": 4.980986404418945,
      "step": 2930
    },
    {
      "epoch": 0.496705524581855,
      "grad_norm": 0.6040016412734985,
      "learning_rate": 0.0002933069804401687,
      "loss": 5.0005535125732425,
      "step": 2940
    },
    {
      "epoch": 0.4983949991552627,
      "grad_norm": 0.6449369788169861,
      "learning_rate": 0.0002931647634106282,
      "loss": 4.974679946899414,
      "step": 2950
    },
    {
      "epoch": 0.5000844737286704,
      "grad_norm": 0.5843121409416199,
      "learning_rate": 0.0002930210864852184,
      "loss": 4.985787963867187,
      "step": 2960
    },
    {
      "epoch": 0.501773948302078,
      "grad_norm": 0.6128187775611877,
      "learning_rate": 0.00029287595112905773,
      "loss": 4.969168090820313,
      "step": 2970
    },
    {
      "epoch": 0.5034634228754857,
      "grad_norm": 0.6031991839408875,
      "learning_rate": 0.00029272935882213675,
      "loss": 4.946027374267578,
      "step": 2980
    },
    {
      "epoch": 0.5051528974488934,
      "grad_norm": 0.7001163959503174,
      "learning_rate": 0.00029258131105930314,
      "loss": 4.9540660858154295,
      "step": 2990
    },
    {
      "epoch": 0.5068423720223011,
      "grad_norm": 0.5723311305046082,
      "learning_rate": 0.0002924318093502462,
      "loss": 4.953271865844727,
      "step": 3000
    },
    {
      "epoch": 0.5068423720223011,
      "eval_loss": 4.9452409744262695,
      "eval_runtime": 3.6082,
      "eval_samples_per_second": 277.146,
      "eval_steps_per_second": 5.82,
      "step": 3000
    },
    {
      "epoch": 0.5085318465957087,
      "grad_norm": 0.6639596819877625,
      "learning_rate": 0.00029228085521948167,
      "loss": 4.964737319946289,
      "step": 3010
    },
    {
      "epoch": 0.5102213211691164,
      "grad_norm": 0.6372225284576416,
      "learning_rate": 0.000292128450206336,
      "loss": 4.964575576782226,
      "step": 3020
    },
    {
      "epoch": 0.5119107957425241,
      "grad_norm": 0.6287721991539001,
      "learning_rate": 0.00029197459586493077,
      "loss": 4.968457794189453,
      "step": 3030
    },
    {
      "epoch": 0.5136002703159317,
      "grad_norm": 0.596134603023529,
      "learning_rate": 0.0002918192937641668,
      "loss": 4.937327194213867,
      "step": 3040
    },
    {
      "epoch": 0.5152897448893394,
      "grad_norm": 0.6308782696723938,
      "learning_rate": 0.00029166254548770827,
      "loss": 4.955413055419922,
      "step": 3050
    },
    {
      "epoch": 0.516979219462747,
      "grad_norm": 0.6965638399124146,
      "learning_rate": 0.0002915043526339663,
      "loss": 4.951367568969727,
      "step": 3060
    },
    {
      "epoch": 0.5186686940361548,
      "grad_norm": 0.5890870690345764,
      "learning_rate": 0.00029134471681608286,
      "loss": 4.956174087524414,
      "step": 3070
    },
    {
      "epoch": 0.5203581686095624,
      "grad_norm": 0.6104860901832581,
      "learning_rate": 0.00029118363966191445,
      "loss": 4.943370056152344,
      "step": 3080
    },
    {
      "epoch": 0.5220476431829701,
      "grad_norm": 0.5911648869514465,
      "learning_rate": 0.00029102112281401507,
      "loss": 4.944694900512696,
      "step": 3090
    },
    {
      "epoch": 0.5237371177563778,
      "grad_norm": 0.565938413143158,
      "learning_rate": 0.0002908571679296199,
      "loss": 4.952177047729492,
      "step": 3100
    },
    {
      "epoch": 0.5254265923297854,
      "grad_norm": 0.5907305479049683,
      "learning_rate": 0.00029069177668062816,
      "loss": 4.945420837402343,
      "step": 3110
    },
    {
      "epoch": 0.5271160669031931,
      "grad_norm": 0.6035214066505432,
      "learning_rate": 0.00029052495075358617,
      "loss": 4.937542343139649,
      "step": 3120
    },
    {
      "epoch": 0.5288055414766007,
      "grad_norm": 0.6489094495773315,
      "learning_rate": 0.0002903566918496701,
      "loss": 4.9244636535644535,
      "step": 3130
    },
    {
      "epoch": 0.5304950160500085,
      "grad_norm": 0.6260411739349365,
      "learning_rate": 0.0002901870016846685,
      "loss": 4.8926746368408205,
      "step": 3140
    },
    {
      "epoch": 0.5321844906234161,
      "grad_norm": 0.6731085777282715,
      "learning_rate": 0.00029001588198896523,
      "loss": 4.926407241821289,
      "step": 3150
    },
    {
      "epoch": 0.5338739651968238,
      "grad_norm": 0.7428810596466064,
      "learning_rate": 0.0002898433345075212,
      "loss": 4.938412857055664,
      "step": 3160
    },
    {
      "epoch": 0.5355634397702315,
      "grad_norm": 0.6765902042388916,
      "learning_rate": 0.0002896693609998571,
      "loss": 4.913272094726563,
      "step": 3170
    },
    {
      "epoch": 0.5372529143436391,
      "grad_norm": 0.7018482089042664,
      "learning_rate": 0.0002894939632400352,
      "loss": 4.8967021942138675,
      "step": 3180
    },
    {
      "epoch": 0.5389423889170468,
      "grad_norm": 0.6272666454315186,
      "learning_rate": 0.0002893171430166413,
      "loss": 4.921631622314453,
      "step": 3190
    },
    {
      "epoch": 0.5406318634904544,
      "grad_norm": 0.6220508217811584,
      "learning_rate": 0.00028913890213276664,
      "loss": 4.878177642822266,
      "step": 3200
    },
    {
      "epoch": 0.5423213380638622,
      "grad_norm": 0.6996421813964844,
      "learning_rate": 0.0002889592424059891,
      "loss": 4.867180633544922,
      "step": 3210
    },
    {
      "epoch": 0.5440108126372698,
      "grad_norm": 0.6159099340438843,
      "learning_rate": 0.0002887781656683551,
      "loss": 4.913734436035156,
      "step": 3220
    },
    {
      "epoch": 0.5457002872106774,
      "grad_norm": 0.6269960403442383,
      "learning_rate": 0.0002885956737663609,
      "loss": 4.884205627441406,
      "step": 3230
    },
    {
      "epoch": 0.5473897617840852,
      "grad_norm": 0.6038281321525574,
      "learning_rate": 0.00028841176856093346,
      "loss": 4.866396713256836,
      "step": 3240
    },
    {
      "epoch": 0.5490792363574928,
      "grad_norm": 0.587129533290863,
      "learning_rate": 0.0002882264519274116,
      "loss": 4.902432632446289,
      "step": 3250
    },
    {
      "epoch": 0.5507687109309005,
      "grad_norm": 0.58821040391922,
      "learning_rate": 0.0002880397257555271,
      "loss": 4.8774364471435545,
      "step": 3260
    },
    {
      "epoch": 0.5524581855043081,
      "grad_norm": 0.5894278287887573,
      "learning_rate": 0.000287851591949385,
      "loss": 4.882402038574218,
      "step": 3270
    },
    {
      "epoch": 0.5541476600777159,
      "grad_norm": 0.6682536005973816,
      "learning_rate": 0.0002876620524274447,
      "loss": 4.875947570800781,
      "step": 3280
    },
    {
      "epoch": 0.5558371346511235,
      "grad_norm": 0.56892329454422,
      "learning_rate": 0.0002874711091224999,
      "loss": 4.8791545867919925,
      "step": 3290
    },
    {
      "epoch": 0.5575266092245311,
      "grad_norm": 0.6786117553710938,
      "learning_rate": 0.0002872787639816593,
      "loss": 4.891143417358398,
      "step": 3300
    },
    {
      "epoch": 0.5592160837979389,
      "grad_norm": 0.5622825622558594,
      "learning_rate": 0.00028708501896632636,
      "loss": 4.864010620117187,
      "step": 3310
    },
    {
      "epoch": 0.5609055583713465,
      "grad_norm": 0.5666744709014893,
      "learning_rate": 0.0002868898760521796,
      "loss": 4.874967956542969,
      "step": 3320
    },
    {
      "epoch": 0.5625950329447542,
      "grad_norm": 0.5970442295074463,
      "learning_rate": 0.00028669333722915245,
      "loss": 4.873799514770508,
      "step": 3330
    },
    {
      "epoch": 0.5642845075181618,
      "grad_norm": 0.6412243843078613,
      "learning_rate": 0.0002864954045014126,
      "loss": 4.852934646606445,
      "step": 3340
    },
    {
      "epoch": 0.5659739820915696,
      "grad_norm": 0.6244974732398987,
      "learning_rate": 0.00028629607988734214,
      "loss": 4.876249694824219,
      "step": 3350
    },
    {
      "epoch": 0.5676634566649772,
      "grad_norm": 0.6246965527534485,
      "learning_rate": 0.00028609536541951636,
      "loss": 4.86097412109375,
      "step": 3360
    },
    {
      "epoch": 0.5693529312383848,
      "grad_norm": 0.6178351044654846,
      "learning_rate": 0.00028589326314468357,
      "loss": 4.864706420898438,
      "step": 3370
    },
    {
      "epoch": 0.5710424058117926,
      "grad_norm": 0.5810765624046326,
      "learning_rate": 0.0002856897751237439,
      "loss": 4.842799377441406,
      "step": 3380
    },
    {
      "epoch": 0.5727318803852002,
      "grad_norm": 0.5823137760162354,
      "learning_rate": 0.0002854849034317282,
      "loss": 4.846402740478515,
      "step": 3390
    },
    {
      "epoch": 0.5744213549586079,
      "grad_norm": 0.5997573137283325,
      "learning_rate": 0.0002852786501577773,
      "loss": 4.83245849609375,
      "step": 3400
    },
    {
      "epoch": 0.5761108295320155,
      "grad_norm": 0.587899386882782,
      "learning_rate": 0.0002850710174051204,
      "loss": 4.871646118164063,
      "step": 3410
    },
    {
      "epoch": 0.5778003041054233,
      "grad_norm": 0.6415139436721802,
      "learning_rate": 0.0002848620072910535,
      "loss": 4.83592414855957,
      "step": 3420
    },
    {
      "epoch": 0.5794897786788309,
      "grad_norm": 0.5660437345504761,
      "learning_rate": 0.0002846516219469181,
      "loss": 4.855052947998047,
      "step": 3430
    },
    {
      "epoch": 0.5811792532522385,
      "grad_norm": 0.5609036087989807,
      "learning_rate": 0.0002844398635180794,
      "loss": 4.856648254394531,
      "step": 3440
    },
    {
      "epoch": 0.5828687278256462,
      "grad_norm": 0.5897738933563232,
      "learning_rate": 0.00028422673416390437,
      "loss": 4.862053298950196,
      "step": 3450
    },
    {
      "epoch": 0.5845582023990539,
      "grad_norm": 0.5655834674835205,
      "learning_rate": 0.00028401223605773964,
      "loss": 4.845742416381836,
      "step": 3460
    },
    {
      "epoch": 0.5862476769724616,
      "grad_norm": 0.604269802570343,
      "learning_rate": 0.00028379637138688945,
      "loss": 4.828603363037109,
      "step": 3470
    },
    {
      "epoch": 0.5879371515458692,
      "grad_norm": 0.5728313326835632,
      "learning_rate": 0.0002835791423525935,
      "loss": 4.821448516845703,
      "step": 3480
    },
    {
      "epoch": 0.5896266261192769,
      "grad_norm": 0.6384554505348206,
      "learning_rate": 0.00028336055117000403,
      "loss": 4.838501358032227,
      "step": 3490
    },
    {
      "epoch": 0.5913161006926846,
      "grad_norm": 0.5678889155387878,
      "learning_rate": 0.0002831406000681638,
      "loss": 4.827870178222656,
      "step": 3500
    },
    {
      "epoch": 0.5913161006926846,
      "eval_loss": 4.812247276306152,
      "eval_runtime": 3.7454,
      "eval_samples_per_second": 266.992,
      "eval_steps_per_second": 5.607,
      "step": 3500
    },
    {
      "epoch": 0.5930055752660922,
      "grad_norm": 0.6172225475311279,
      "learning_rate": 0.00028291929128998293,
      "loss": 4.818932342529297,
      "step": 3510
    },
    {
      "epoch": 0.5946950498395,
      "grad_norm": 0.6093600392341614,
      "learning_rate": 0.00028269662709221635,
      "loss": 4.821521759033203,
      "step": 3520
    },
    {
      "epoch": 0.5963845244129076,
      "grad_norm": 0.5978082418441772,
      "learning_rate": 0.00028247260974544037,
      "loss": 4.8287818908691404,
      "step": 3530
    },
    {
      "epoch": 0.5980739989863153,
      "grad_norm": 0.6068260073661804,
      "learning_rate": 0.00028224724153403015,
      "loss": 4.830426025390625,
      "step": 3540
    },
    {
      "epoch": 0.5997634735597229,
      "grad_norm": 0.5650802850723267,
      "learning_rate": 0.0002820205247561356,
      "loss": 4.82833251953125,
      "step": 3550
    },
    {
      "epoch": 0.6014529481331305,
      "grad_norm": 0.5627108812332153,
      "learning_rate": 0.0002817924617236587,
      "loss": 4.835605239868164,
      "step": 3560
    },
    {
      "epoch": 0.6031424227065383,
      "grad_norm": 0.5983113050460815,
      "learning_rate": 0.00028156305476222966,
      "loss": 4.8316596984863285,
      "step": 3570
    },
    {
      "epoch": 0.6048318972799459,
      "grad_norm": 0.6292333006858826,
      "learning_rate": 0.0002813323062111828,
      "loss": 4.8177978515625,
      "step": 3580
    },
    {
      "epoch": 0.6065213718533536,
      "grad_norm": 0.5784122943878174,
      "learning_rate": 0.0002811002184235334,
      "loss": 4.801444625854492,
      "step": 3590
    },
    {
      "epoch": 0.6082108464267613,
      "grad_norm": 0.598548173904419,
      "learning_rate": 0.00028086679376595314,
      "loss": 4.825639343261718,
      "step": 3600
    },
    {
      "epoch": 0.609900321000169,
      "grad_norm": 0.6459059715270996,
      "learning_rate": 0.00028063203461874635,
      "loss": 4.819043350219727,
      "step": 3610
    },
    {
      "epoch": 0.6115897955735766,
      "grad_norm": 0.6150253415107727,
      "learning_rate": 0.0002803959433758254,
      "loss": 4.791939926147461,
      "step": 3620
    },
    {
      "epoch": 0.6132792701469842,
      "grad_norm": 0.6212024092674255,
      "learning_rate": 0.0002801585224446866,
      "loss": 4.81633415222168,
      "step": 3630
    },
    {
      "epoch": 0.614968744720392,
      "grad_norm": 0.677480936050415,
      "learning_rate": 0.0002799197742463854,
      "loss": 4.800606918334961,
      "step": 3640
    },
    {
      "epoch": 0.6166582192937996,
      "grad_norm": 0.5394392609596252,
      "learning_rate": 0.0002796797012155118,
      "loss": 4.792086791992188,
      "step": 3650
    },
    {
      "epoch": 0.6183476938672073,
      "grad_norm": 0.6263740658760071,
      "learning_rate": 0.0002794383058001657,
      "loss": 4.8161579132080075,
      "step": 3660
    },
    {
      "epoch": 0.620037168440615,
      "grad_norm": 0.5984665751457214,
      "learning_rate": 0.00027919559046193156,
      "loss": 4.80903434753418,
      "step": 3670
    },
    {
      "epoch": 0.6217266430140226,
      "grad_norm": 0.570600688457489,
      "learning_rate": 0.0002789515576758536,
      "loss": 4.814385986328125,
      "step": 3680
    },
    {
      "epoch": 0.6234161175874303,
      "grad_norm": 0.6285440325737,
      "learning_rate": 0.00027870620993041055,
      "loss": 4.766680908203125,
      "step": 3690
    },
    {
      "epoch": 0.6251055921608379,
      "grad_norm": 0.6015235185623169,
      "learning_rate": 0.00027845954972749004,
      "loss": 4.784580230712891,
      "step": 3700
    },
    {
      "epoch": 0.6267950667342457,
      "grad_norm": 0.5941788554191589,
      "learning_rate": 0.0002782115795823633,
      "loss": 4.8005828857421875,
      "step": 3710
    },
    {
      "epoch": 0.6284845413076533,
      "grad_norm": 0.5614147782325745,
      "learning_rate": 0.0002779623020236594,
      "loss": 4.785182952880859,
      "step": 3720
    },
    {
      "epoch": 0.630174015881061,
      "grad_norm": 0.6022568941116333,
      "learning_rate": 0.00027771171959333976,
      "loss": 4.825439453125,
      "step": 3730
    },
    {
      "epoch": 0.6318634904544687,
      "grad_norm": 0.5763116478919983,
      "learning_rate": 0.00027745983484667164,
      "loss": 4.781736373901367,
      "step": 3740
    },
    {
      "epoch": 0.6335529650278763,
      "grad_norm": 0.5323732495307922,
      "learning_rate": 0.0002772066503522026,
      "loss": 4.793933868408203,
      "step": 3750
    },
    {
      "epoch": 0.635242439601284,
      "grad_norm": 0.6131083369255066,
      "learning_rate": 0.00027695216869173415,
      "loss": 4.761587905883789,
      "step": 3760
    },
    {
      "epoch": 0.6369319141746916,
      "grad_norm": 0.6051084399223328,
      "learning_rate": 0.0002766963924602953,
      "loss": 4.781452941894531,
      "step": 3770
    },
    {
      "epoch": 0.6386213887480994,
      "grad_norm": 0.5932102203369141,
      "learning_rate": 0.00027643932426611647,
      "loss": 4.7832183837890625,
      "step": 3780
    },
    {
      "epoch": 0.640310863321507,
      "grad_norm": 0.5552225708961487,
      "learning_rate": 0.0002761809667306022,
      "loss": 4.767021560668946,
      "step": 3790
    },
    {
      "epoch": 0.6420003378949147,
      "grad_norm": 0.6563674211502075,
      "learning_rate": 0.00027592132248830526,
      "loss": 4.7745105743408205,
      "step": 3800
    },
    {
      "epoch": 0.6436898124683224,
      "grad_norm": 0.5460559725761414,
      "learning_rate": 0.00027566039418689905,
      "loss": 4.758267593383789,
      "step": 3810
    },
    {
      "epoch": 0.64537928704173,
      "grad_norm": 0.5483755469322205,
      "learning_rate": 0.00027539818448715124,
      "loss": 4.763653182983399,
      "step": 3820
    },
    {
      "epoch": 0.6470687616151377,
      "grad_norm": 0.5985057950019836,
      "learning_rate": 0.000275134696062896,
      "loss": 4.773738098144531,
      "step": 3830
    },
    {
      "epoch": 0.6487582361885453,
      "grad_norm": 0.5881696939468384,
      "learning_rate": 0.0002748699316010073,
      "loss": 4.750997161865234,
      "step": 3840
    },
    {
      "epoch": 0.6504477107619531,
      "grad_norm": 0.5993970632553101,
      "learning_rate": 0.000274603893801371,
      "loss": 4.757858657836914,
      "step": 3850
    },
    {
      "epoch": 0.6521371853353607,
      "grad_norm": 0.5355719327926636,
      "learning_rate": 0.000274336585376858,
      "loss": 4.756174468994141,
      "step": 3860
    },
    {
      "epoch": 0.6538266599087684,
      "grad_norm": 0.5604658722877502,
      "learning_rate": 0.0002740680090532958,
      "loss": 4.771471786499023,
      "step": 3870
    },
    {
      "epoch": 0.655516134482176,
      "grad_norm": 0.5580816268920898,
      "learning_rate": 0.0002737981675694411,
      "loss": 4.767171859741211,
      "step": 3880
    },
    {
      "epoch": 0.6572056090555837,
      "grad_norm": 0.5741564035415649,
      "learning_rate": 0.00027352706367695203,
      "loss": 4.755613708496094,
      "step": 3890
    },
    {
      "epoch": 0.6588950836289914,
      "grad_norm": 0.5679869055747986,
      "learning_rate": 0.00027325470014035965,
      "loss": 4.768374252319336,
      "step": 3900
    },
    {
      "epoch": 0.660584558202399,
      "grad_norm": 0.5516665577888489,
      "learning_rate": 0.0002729810797370402,
      "loss": 4.754274368286133,
      "step": 3910
    },
    {
      "epoch": 0.6622740327758068,
      "grad_norm": 0.572651743888855,
      "learning_rate": 0.00027270620525718647,
      "loss": 4.740098190307617,
      "step": 3920
    },
    {
      "epoch": 0.6639635073492144,
      "grad_norm": 0.5252317190170288,
      "learning_rate": 0.0002724300795037796,
      "loss": 4.781079864501953,
      "step": 3930
    },
    {
      "epoch": 0.665652981922622,
      "grad_norm": 0.578183114528656,
      "learning_rate": 0.00027215270529256015,
      "loss": 4.738787460327148,
      "step": 3940
    },
    {
      "epoch": 0.6673424564960297,
      "grad_norm": 0.6071267127990723,
      "learning_rate": 0.00027187408545199977,
      "loss": 4.73607177734375,
      "step": 3950
    },
    {
      "epoch": 0.6690319310694374,
      "grad_norm": 0.5916706919670105,
      "learning_rate": 0.00027159422282327204,
      "loss": 4.747200775146484,
      "step": 3960
    },
    {
      "epoch": 0.6707214056428451,
      "grad_norm": 0.5538118481636047,
      "learning_rate": 0.0002713131202602238,
      "loss": 4.765713119506836,
      "step": 3970
    },
    {
      "epoch": 0.6724108802162527,
      "grad_norm": 0.5321049094200134,
      "learning_rate": 0.0002710307806293458,
      "loss": 4.7143207550048825,
      "step": 3980
    },
    {
      "epoch": 0.6741003547896605,
      "grad_norm": 0.5859444737434387,
      "learning_rate": 0.0002707472068097435,
      "loss": 4.749985122680664,
      "step": 3990
    },
    {
      "epoch": 0.6757898293630681,
      "grad_norm": 0.520622730255127,
      "learning_rate": 0.0002704624016931079,
      "loss": 4.7440532684326175,
      "step": 4000
    },
    {
      "epoch": 0.6757898293630681,
      "eval_loss": 4.7194108963012695,
      "eval_runtime": 3.7971,
      "eval_samples_per_second": 263.356,
      "eval_steps_per_second": 5.53,
      "step": 4000
    },
    {
      "epoch": 0.6774793039364757,
      "grad_norm": 0.5953722596168518,
      "learning_rate": 0.00027017636818368575,
      "loss": 4.737245559692383,
      "step": 4010
    },
    {
      "epoch": 0.6791687785098834,
      "grad_norm": 0.6203189492225647,
      "learning_rate": 0.0002698891091982504,
      "loss": 4.716173934936523,
      "step": 4020
    },
    {
      "epoch": 0.6808582530832911,
      "grad_norm": 0.5239487886428833,
      "learning_rate": 0.00026960062766607135,
      "loss": 4.735467529296875,
      "step": 4030
    },
    {
      "epoch": 0.6825477276566988,
      "grad_norm": 0.5474298000335693,
      "learning_rate": 0.0002693109265288851,
      "loss": 4.725514984130859,
      "step": 4040
    },
    {
      "epoch": 0.6842372022301064,
      "grad_norm": 0.5452102422714233,
      "learning_rate": 0.0002690200087408648,
      "loss": 4.726776885986328,
      "step": 4050
    },
    {
      "epoch": 0.6859266768035142,
      "grad_norm": 0.6271504759788513,
      "learning_rate": 0.00026872787726859004,
      "loss": 4.71842041015625,
      "step": 4060
    },
    {
      "epoch": 0.6876161513769218,
      "grad_norm": 0.5585569143295288,
      "learning_rate": 0.0002684345350910169,
      "loss": 4.728883361816406,
      "step": 4070
    },
    {
      "epoch": 0.6893056259503294,
      "grad_norm": 0.544662594795227,
      "learning_rate": 0.0002681399851994472,
      "loss": 4.729270553588867,
      "step": 4080
    },
    {
      "epoch": 0.6909951005237371,
      "grad_norm": 0.5363122820854187,
      "learning_rate": 0.00026784423059749845,
      "loss": 4.726214599609375,
      "step": 4090
    },
    {
      "epoch": 0.6926845750971448,
      "grad_norm": 0.5298801064491272,
      "learning_rate": 0.0002675472743010727,
      "loss": 4.697872924804687,
      "step": 4100
    },
    {
      "epoch": 0.6943740496705525,
      "grad_norm": 0.5710757374763489,
      "learning_rate": 0.0002672491193383263,
      "loss": 4.723146438598633,
      "step": 4110
    },
    {
      "epoch": 0.6960635242439601,
      "grad_norm": 0.5484883785247803,
      "learning_rate": 0.00026694976874963854,
      "loss": 4.738557052612305,
      "step": 4120
    },
    {
      "epoch": 0.6977529988173679,
      "grad_norm": 0.5273333191871643,
      "learning_rate": 0.00026664922558758105,
      "loss": 4.700592803955078,
      "step": 4130
    },
    {
      "epoch": 0.6994424733907755,
      "grad_norm": 0.5574657320976257,
      "learning_rate": 0.00026634749291688646,
      "loss": 4.729513168334961,
      "step": 4140
    },
    {
      "epoch": 0.7011319479641831,
      "grad_norm": 0.5571582317352295,
      "learning_rate": 0.00026604457381441715,
      "loss": 4.706221389770508,
      "step": 4150
    },
    {
      "epoch": 0.7028214225375908,
      "grad_norm": 0.6286988258361816,
      "learning_rate": 0.00026574047136913403,
      "loss": 4.701080322265625,
      "step": 4160
    },
    {
      "epoch": 0.7045108971109985,
      "grad_norm": 0.5314433574676514,
      "learning_rate": 0.0002654351886820648,
      "loss": 4.714921188354492,
      "step": 4170
    },
    {
      "epoch": 0.7062003716844062,
      "grad_norm": 0.539644718170166,
      "learning_rate": 0.0002651287288662724,
      "loss": 4.722955703735352,
      "step": 4180
    },
    {
      "epoch": 0.7078898462578138,
      "grad_norm": 0.5164220333099365,
      "learning_rate": 0.0002648210950468236,
      "loss": 4.7029579162597654,
      "step": 4190
    },
    {
      "epoch": 0.7095793208312214,
      "grad_norm": 0.5345500111579895,
      "learning_rate": 0.0002645122903607566,
      "loss": 4.696025085449219,
      "step": 4200
    },
    {
      "epoch": 0.7112687954046292,
      "grad_norm": 0.5561880469322205,
      "learning_rate": 0.0002642023179570493,
      "loss": 4.696010971069336,
      "step": 4210
    },
    {
      "epoch": 0.7129582699780368,
      "grad_norm": 0.5260653495788574,
      "learning_rate": 0.0002638911809965874,
      "loss": 4.705658721923828,
      "step": 4220
    },
    {
      "epoch": 0.7146477445514445,
      "grad_norm": 0.517846941947937,
      "learning_rate": 0.0002635788826521316,
      "loss": 4.690948104858398,
      "step": 4230
    },
    {
      "epoch": 0.7163372191248522,
      "grad_norm": 0.5815365314483643,
      "learning_rate": 0.00026326542610828597,
      "loss": 4.702710723876953,
      "step": 4240
    },
    {
      "epoch": 0.7180266936982599,
      "grad_norm": 0.5511707067489624,
      "learning_rate": 0.00026295081456146485,
      "loss": 4.713930130004883,
      "step": 4250
    },
    {
      "epoch": 0.7197161682716675,
      "grad_norm": 0.5390937924385071,
      "learning_rate": 0.0002626350512198606,
      "loss": 4.694212341308594,
      "step": 4260
    },
    {
      "epoch": 0.7214056428450751,
      "grad_norm": 0.5410081744194031,
      "learning_rate": 0.0002623181393034108,
      "loss": 4.696395492553711,
      "step": 4270
    },
    {
      "epoch": 0.7230951174184829,
      "grad_norm": 0.5272055268287659,
      "learning_rate": 0.00026200008204376525,
      "loss": 4.715652847290039,
      "step": 4280
    },
    {
      "epoch": 0.7247845919918905,
      "grad_norm": 0.5485383868217468,
      "learning_rate": 0.00026168088268425346,
      "loss": 4.689223861694336,
      "step": 4290
    },
    {
      "epoch": 0.7264740665652982,
      "grad_norm": 0.4974030554294586,
      "learning_rate": 0.00026136054447985105,
      "loss": 4.6958671569824215,
      "step": 4300
    },
    {
      "epoch": 0.7281635411387058,
      "grad_norm": 0.5421763062477112,
      "learning_rate": 0.00026103907069714694,
      "loss": 4.706072235107422,
      "step": 4310
    },
    {
      "epoch": 0.7298530157121136,
      "grad_norm": 0.5402170419692993,
      "learning_rate": 0.0002607164646143098,
      "loss": 4.684348297119141,
      "step": 4320
    },
    {
      "epoch": 0.7315424902855212,
      "grad_norm": 0.5388095378875732,
      "learning_rate": 0.0002603927295210547,
      "loss": 4.681607818603515,
      "step": 4330
    },
    {
      "epoch": 0.7332319648589288,
      "grad_norm": 0.5691295266151428,
      "learning_rate": 0.00026006786871860975,
      "loss": 4.659119033813477,
      "step": 4340
    },
    {
      "epoch": 0.7349214394323366,
      "grad_norm": 0.5657386183738708,
      "learning_rate": 0.00025974188551968207,
      "loss": 4.707662963867188,
      "step": 4350
    },
    {
      "epoch": 0.7366109140057442,
      "grad_norm": 0.5887618660926819,
      "learning_rate": 0.0002594147832484243,
      "loss": 4.678396606445313,
      "step": 4360
    },
    {
      "epoch": 0.7383003885791519,
      "grad_norm": 0.5618587136268616,
      "learning_rate": 0.0002590865652404007,
      "loss": 4.6809638977050785,
      "step": 4370
    },
    {
      "epoch": 0.7399898631525595,
      "grad_norm": 0.5673303604125977,
      "learning_rate": 0.0002587572348425529,
      "loss": 4.683576583862305,
      "step": 4380
    },
    {
      "epoch": 0.7416793377259672,
      "grad_norm": 0.5109097361564636,
      "learning_rate": 0.0002584267954131659,
      "loss": 4.674320983886719,
      "step": 4390
    },
    {
      "epoch": 0.7433688122993749,
      "grad_norm": 0.5133926272392273,
      "learning_rate": 0.000258095250321834,
      "loss": 4.676524353027344,
      "step": 4400
    },
    {
      "epoch": 0.7450582868727825,
      "grad_norm": 0.5628970265388489,
      "learning_rate": 0.00025776260294942615,
      "loss": 4.688607025146484,
      "step": 4410
    },
    {
      "epoch": 0.7467477614461903,
      "grad_norm": 0.5761396884918213,
      "learning_rate": 0.0002574288566880517,
      "loss": 4.666116333007812,
      "step": 4420
    },
    {
      "epoch": 0.7484372360195979,
      "grad_norm": 0.5518139004707336,
      "learning_rate": 0.0002570940149410256,
      "loss": 4.665610504150391,
      "step": 4430
    },
    {
      "epoch": 0.7501267105930056,
      "grad_norm": 0.5176488757133484,
      "learning_rate": 0.00025675808112283387,
      "loss": 4.673014831542969,
      "step": 4440
    },
    {
      "epoch": 0.7518161851664132,
      "grad_norm": 0.5482094287872314,
      "learning_rate": 0.00025642105865909874,
      "loss": 4.665557098388672,
      "step": 4450
    },
    {
      "epoch": 0.7535056597398209,
      "grad_norm": 0.7407347559928894,
      "learning_rate": 0.0002560829509865437,
      "loss": 4.660655975341797,
      "step": 4460
    },
    {
      "epoch": 0.7551951343132286,
      "grad_norm": 0.5341119766235352,
      "learning_rate": 0.00025574376155295845,
      "loss": 4.670759582519532,
      "step": 4470
    },
    {
      "epoch": 0.7568846088866362,
      "grad_norm": 0.5163617134094238,
      "learning_rate": 0.00025540349381716367,
      "loss": 4.689437484741211,
      "step": 4480
    },
    {
      "epoch": 0.758574083460044,
      "grad_norm": 0.6329180598258972,
      "learning_rate": 0.00025506215124897593,
      "loss": 4.6677288055419925,
      "step": 4490
    },
    {
      "epoch": 0.7602635580334516,
      "grad_norm": 0.5871708393096924,
      "learning_rate": 0.0002547197373291721,
      "loss": 4.678330993652343,
      "step": 4500
    },
    {
      "epoch": 0.7602635580334516,
      "eval_loss": 4.647042751312256,
      "eval_runtime": 3.6169,
      "eval_samples_per_second": 276.482,
      "eval_steps_per_second": 5.806,
      "step": 4500
    },
    {
      "epoch": 0.7619530326068593,
      "grad_norm": 0.5751690864562988,
      "learning_rate": 0.0002543762555494541,
      "loss": 4.659806823730468,
      "step": 4510
    },
    {
      "epoch": 0.7636425071802669,
      "grad_norm": 0.5488387942314148,
      "learning_rate": 0.0002540317094124131,
      "loss": 4.675619888305664,
      "step": 4520
    },
    {
      "epoch": 0.7653319817536746,
      "grad_norm": 0.5706210136413574,
      "learning_rate": 0.0002536861024314936,
      "loss": 4.647731018066406,
      "step": 4530
    },
    {
      "epoch": 0.7670214563270823,
      "grad_norm": 0.5262100100517273,
      "learning_rate": 0.0002533394381309583,
      "loss": 4.629973220825195,
      "step": 4540
    },
    {
      "epoch": 0.7687109309004899,
      "grad_norm": 0.5438910126686096,
      "learning_rate": 0.00025299172004585144,
      "loss": 4.680305099487304,
      "step": 4550
    },
    {
      "epoch": 0.7704004054738977,
      "grad_norm": 0.5125553011894226,
      "learning_rate": 0.00025264295172196304,
      "loss": 4.6679943084716795,
      "step": 4560
    },
    {
      "epoch": 0.7720898800473053,
      "grad_norm": 0.5525355339050293,
      "learning_rate": 0.0002522931367157928,
      "loss": 4.6561134338378904,
      "step": 4570
    },
    {
      "epoch": 0.773779354620713,
      "grad_norm": 0.5133577585220337,
      "learning_rate": 0.00025194227859451384,
      "loss": 4.66561279296875,
      "step": 4580
    },
    {
      "epoch": 0.7754688291941206,
      "grad_norm": 0.5095699429512024,
      "learning_rate": 0.00025159038093593606,
      "loss": 4.678707122802734,
      "step": 4590
    },
    {
      "epoch": 0.7771583037675283,
      "grad_norm": 0.5241293907165527,
      "learning_rate": 0.0002512374473284699,
      "loss": 4.642659759521484,
      "step": 4600
    },
    {
      "epoch": 0.778847778340936,
      "grad_norm": 0.557011067867279,
      "learning_rate": 0.00025088348137108983,
      "loss": 4.642984771728516,
      "step": 4610
    },
    {
      "epoch": 0.7805372529143436,
      "grad_norm": 0.5290088653564453,
      "learning_rate": 0.0002505284866732974,
      "loss": 4.668995666503906,
      "step": 4620
    },
    {
      "epoch": 0.7822267274877513,
      "grad_norm": 0.519223153591156,
      "learning_rate": 0.0002501724668550846,
      "loss": 4.627962112426758,
      "step": 4630
    },
    {
      "epoch": 0.783916202061159,
      "grad_norm": 0.5338088274002075,
      "learning_rate": 0.00024981542554689684,
      "loss": 4.67579231262207,
      "step": 4640
    },
    {
      "epoch": 0.7856056766345666,
      "grad_norm": 0.5252251625061035,
      "learning_rate": 0.000249457366389596,
      "loss": 4.656952285766602,
      "step": 4650
    },
    {
      "epoch": 0.7872951512079743,
      "grad_norm": 0.5428206324577332,
      "learning_rate": 0.0002490982930344233,
      "loss": 4.646731185913086,
      "step": 4660
    },
    {
      "epoch": 0.788984625781382,
      "grad_norm": 0.5392381548881531,
      "learning_rate": 0.0002487382091429621,
      "loss": 4.663632583618164,
      "step": 4670
    },
    {
      "epoch": 0.7906741003547897,
      "grad_norm": 0.51649409532547,
      "learning_rate": 0.00024837711838710035,
      "loss": 4.620084762573242,
      "step": 4680
    },
    {
      "epoch": 0.7923635749281973,
      "grad_norm": 0.5248917937278748,
      "learning_rate": 0.00024801502444899353,
      "loss": 4.66024169921875,
      "step": 4690
    },
    {
      "epoch": 0.794053049501605,
      "grad_norm": 0.5321633219718933,
      "learning_rate": 0.00024765193102102676,
      "loss": 4.647469329833984,
      "step": 4700
    },
    {
      "epoch": 0.7957425240750127,
      "grad_norm": 0.5236574411392212,
      "learning_rate": 0.0002472878418057772,
      "loss": 4.6667522430419925,
      "step": 4710
    },
    {
      "epoch": 0.7974319986484203,
      "grad_norm": 0.5166000127792358,
      "learning_rate": 0.0002469227605159766,
      "loss": 4.6316486358642575,
      "step": 4720
    },
    {
      "epoch": 0.799121473221828,
      "grad_norm": 0.6069431304931641,
      "learning_rate": 0.0002465566908744729,
      "loss": 4.614125442504883,
      "step": 4730
    },
    {
      "epoch": 0.8008109477952357,
      "grad_norm": 0.5319153666496277,
      "learning_rate": 0.00024618963661419285,
      "loss": 4.649255752563477,
      "step": 4740
    },
    {
      "epoch": 0.8025004223686434,
      "grad_norm": 0.4894997477531433,
      "learning_rate": 0.0002458216014781035,
      "loss": 4.621485900878906,
      "step": 4750
    },
    {
      "epoch": 0.804189896942051,
      "grad_norm": 0.516018807888031,
      "learning_rate": 0.00024545258921917416,
      "loss": 4.630000305175781,
      "step": 4760
    },
    {
      "epoch": 0.8058793715154587,
      "grad_norm": 0.5458150506019592,
      "learning_rate": 0.0002450826036003384,
      "loss": 4.635307312011719,
      "step": 4770
    },
    {
      "epoch": 0.8075688460888664,
      "grad_norm": 0.5067882537841797,
      "learning_rate": 0.00024471164839445526,
      "loss": 4.636883163452149,
      "step": 4780
    },
    {
      "epoch": 0.809258320662274,
      "grad_norm": 0.4767204523086548,
      "learning_rate": 0.0002443397273842709,
      "loss": 4.645626831054687,
      "step": 4790
    },
    {
      "epoch": 0.8109477952356817,
      "grad_norm": 0.5159788727760315,
      "learning_rate": 0.00024396684436238025,
      "loss": 4.605623626708985,
      "step": 4800
    },
    {
      "epoch": 0.8126372698090893,
      "grad_norm": 0.5320490598678589,
      "learning_rate": 0.00024359300313118814,
      "loss": 4.638732147216797,
      "step": 4810
    },
    {
      "epoch": 0.8143267443824971,
      "grad_norm": 0.5451418161392212,
      "learning_rate": 0.00024321820750287045,
      "loss": 4.6449028015136715,
      "step": 4820
    },
    {
      "epoch": 0.8160162189559047,
      "grad_norm": 0.5369979739189148,
      "learning_rate": 0.00024284246129933543,
      "loss": 4.602875518798828,
      "step": 4830
    },
    {
      "epoch": 0.8177056935293124,
      "grad_norm": 0.5349618196487427,
      "learning_rate": 0.0002424657683521847,
      "loss": 4.624568939208984,
      "step": 4840
    },
    {
      "epoch": 0.8193951681027201,
      "grad_norm": 0.5187742114067078,
      "learning_rate": 0.00024208813250267404,
      "loss": 4.621414566040039,
      "step": 4850
    },
    {
      "epoch": 0.8210846426761277,
      "grad_norm": 0.49689674377441406,
      "learning_rate": 0.00024170955760167436,
      "loss": 4.63438606262207,
      "step": 4860
    },
    {
      "epoch": 0.8227741172495354,
      "grad_norm": 0.5191966891288757,
      "learning_rate": 0.0002413300475096322,
      "loss": 4.629247665405273,
      "step": 4870
    },
    {
      "epoch": 0.824463591822943,
      "grad_norm": 0.5321470499038696,
      "learning_rate": 0.00024094960609653078,
      "loss": 4.630535507202149,
      "step": 4880
    },
    {
      "epoch": 0.8261530663963508,
      "grad_norm": 0.577171802520752,
      "learning_rate": 0.00024056823724185014,
      "loss": 4.614957809448242,
      "step": 4890
    },
    {
      "epoch": 0.8278425409697584,
      "grad_norm": 0.5203391313552856,
      "learning_rate": 0.00024018594483452783,
      "loss": 4.597796630859375,
      "step": 4900
    },
    {
      "epoch": 0.829532015543166,
      "grad_norm": 0.568663477897644,
      "learning_rate": 0.00023980273277291893,
      "loss": 4.630698394775391,
      "step": 4910
    },
    {
      "epoch": 0.8312214901165738,
      "grad_norm": 0.5214170813560486,
      "learning_rate": 0.00023941860496475687,
      "loss": 4.6348930358886715,
      "step": 4920
    },
    {
      "epoch": 0.8329109646899814,
      "grad_norm": 0.5391976237297058,
      "learning_rate": 0.00023903356532711296,
      "loss": 4.6155132293701175,
      "step": 4930
    },
    {
      "epoch": 0.8346004392633891,
      "grad_norm": 0.4739229381084442,
      "learning_rate": 0.0002386476177863568,
      "loss": 4.622202301025391,
      "step": 4940
    },
    {
      "epoch": 0.8362899138367967,
      "grad_norm": 0.5011942386627197,
      "learning_rate": 0.00023826076627811628,
      "loss": 4.608601379394531,
      "step": 4950
    },
    {
      "epoch": 0.8379793884102045,
      "grad_norm": 0.5716709494590759,
      "learning_rate": 0.0002378730147472371,
      "loss": 4.581511306762695,
      "step": 4960
    },
    {
      "epoch": 0.8396688629836121,
      "grad_norm": 0.5052880644798279,
      "learning_rate": 0.00023748436714774294,
      "loss": 4.649203491210938,
      "step": 4970
    },
    {
      "epoch": 0.8413583375570197,
      "grad_norm": 0.512668788433075,
      "learning_rate": 0.00023709482744279492,
      "loss": 4.621175765991211,
      "step": 4980
    },
    {
      "epoch": 0.8430478121304275,
      "grad_norm": 0.5231815576553345,
      "learning_rate": 0.00023670439960465128,
      "loss": 4.606881713867187,
      "step": 4990
    },
    {
      "epoch": 0.8447372867038351,
      "grad_norm": 0.5233691930770874,
      "learning_rate": 0.00023631308761462677,
      "loss": 4.614410018920898,
      "step": 5000
    },
    {
      "epoch": 0.8447372867038351,
      "eval_loss": 4.584611415863037,
      "eval_runtime": 3.6357,
      "eval_samples_per_second": 275.05,
      "eval_steps_per_second": 5.776,
      "step": 5000
    },
    {
      "epoch": 0.8464267612772428,
      "grad_norm": 0.5504565238952637,
      "learning_rate": 0.00023592089546305216,
      "loss": 4.576148986816406,
      "step": 5010
    },
    {
      "epoch": 0.8481162358506504,
      "grad_norm": 0.5207253098487854,
      "learning_rate": 0.00023552782714923343,
      "loss": 4.615359497070313,
      "step": 5020
    },
    {
      "epoch": 0.8498057104240582,
      "grad_norm": 0.5456526875495911,
      "learning_rate": 0.00023513388668141118,
      "loss": 4.583608627319336,
      "step": 5030
    },
    {
      "epoch": 0.8514951849974658,
      "grad_norm": 0.5371212959289551,
      "learning_rate": 0.00023473907807671952,
      "loss": 4.605810546875,
      "step": 5040
    },
    {
      "epoch": 0.8531846595708734,
      "grad_norm": 0.5273321270942688,
      "learning_rate": 0.00023434340536114531,
      "loss": 4.596974945068359,
      "step": 5050
    },
    {
      "epoch": 0.8548741341442812,
      "grad_norm": 0.5454714894294739,
      "learning_rate": 0.00023394687256948697,
      "loss": 4.595716094970703,
      "step": 5060
    },
    {
      "epoch": 0.8565636087176888,
      "grad_norm": 0.6011702418327332,
      "learning_rate": 0.00023354948374531344,
      "loss": 4.590705108642578,
      "step": 5070
    },
    {
      "epoch": 0.8582530832910965,
      "grad_norm": 0.5225823521614075,
      "learning_rate": 0.00023315124294092277,
      "loss": 4.578453063964844,
      "step": 5080
    },
    {
      "epoch": 0.8599425578645041,
      "grad_norm": 0.5181743502616882,
      "learning_rate": 0.000232752154217301,
      "loss": 4.5722908020019535,
      "step": 5090
    },
    {
      "epoch": 0.8616320324379118,
      "grad_norm": 0.5235112309455872,
      "learning_rate": 0.00023235222164408076,
      "loss": 4.600410461425781,
      "step": 5100
    },
    {
      "epoch": 0.8633215070113195,
      "grad_norm": 0.5427247881889343,
      "learning_rate": 0.00023195144929949953,
      "loss": 4.576435089111328,
      "step": 5110
    },
    {
      "epoch": 0.8650109815847271,
      "grad_norm": 0.5017905235290527,
      "learning_rate": 0.00023154984127035823,
      "loss": 4.6031841278076175,
      "step": 5120
    },
    {
      "epoch": 0.8667004561581348,
      "grad_norm": 0.5279256105422974,
      "learning_rate": 0.00023114740165197957,
      "loss": 4.570458221435547,
      "step": 5130
    },
    {
      "epoch": 0.8683899307315425,
      "grad_norm": 0.5026883482933044,
      "learning_rate": 0.00023074413454816619,
      "loss": 4.587477493286133,
      "step": 5140
    },
    {
      "epoch": 0.8700794053049502,
      "grad_norm": 0.5021783709526062,
      "learning_rate": 0.0002303400440711589,
      "loss": 4.580776977539062,
      "step": 5150
    },
    {
      "epoch": 0.8717688798783578,
      "grad_norm": 0.5208005309104919,
      "learning_rate": 0.00022993513434159464,
      "loss": 4.606272125244141,
      "step": 5160
    },
    {
      "epoch": 0.8734583544517655,
      "grad_norm": 0.4933724105358124,
      "learning_rate": 0.0002295294094884646,
      "loss": 4.598735046386719,
      "step": 5170
    },
    {
      "epoch": 0.8751478290251732,
      "grad_norm": 0.4844622015953064,
      "learning_rate": 0.00022912287364907204,
      "loss": 4.577612686157226,
      "step": 5180
    },
    {
      "epoch": 0.8768373035985808,
      "grad_norm": 0.49681806564331055,
      "learning_rate": 0.00022871553096899,
      "loss": 4.6206306457519535,
      "step": 5190
    },
    {
      "epoch": 0.8785267781719885,
      "grad_norm": 0.5069138407707214,
      "learning_rate": 0.00022830738560201911,
      "loss": 4.576866149902344,
      "step": 5200
    },
    {
      "epoch": 0.8802162527453962,
      "grad_norm": 0.49277958273887634,
      "learning_rate": 0.00022789844171014557,
      "loss": 4.570761489868164,
      "step": 5210
    },
    {
      "epoch": 0.8819057273188039,
      "grad_norm": 0.5152326822280884,
      "learning_rate": 0.00022748870346349796,
      "loss": 4.591669082641602,
      "step": 5220
    },
    {
      "epoch": 0.8835952018922115,
      "grad_norm": 0.5280734896659851,
      "learning_rate": 0.00022707817504030538,
      "loss": 4.600007629394531,
      "step": 5230
    },
    {
      "epoch": 0.8852846764656191,
      "grad_norm": 0.5109785795211792,
      "learning_rate": 0.0002266668606268545,
      "loss": 4.551007461547852,
      "step": 5240
    },
    {
      "epoch": 0.8869741510390269,
      "grad_norm": 0.511035144329071,
      "learning_rate": 0.00022625476441744715,
      "loss": 4.596772766113281,
      "step": 5250
    },
    {
      "epoch": 0.8886636256124345,
      "grad_norm": 0.5007238984107971,
      "learning_rate": 0.00022584189061435725,
      "loss": 4.5646717071533205,
      "step": 5260
    },
    {
      "epoch": 0.8903531001858422,
      "grad_norm": 0.517419159412384,
      "learning_rate": 0.00022542824342778806,
      "loss": 4.561199188232422,
      "step": 5270
    },
    {
      "epoch": 0.8920425747592499,
      "grad_norm": 0.5943387746810913,
      "learning_rate": 0.0002250138270758293,
      "loss": 4.576548385620117,
      "step": 5280
    },
    {
      "epoch": 0.8937320493326576,
      "grad_norm": 0.5131561160087585,
      "learning_rate": 0.00022459864578441415,
      "loss": 4.587300109863281,
      "step": 5290
    },
    {
      "epoch": 0.8954215239060652,
      "grad_norm": 0.5333006381988525,
      "learning_rate": 0.0002241827037872761,
      "loss": 4.5638988494873045,
      "step": 5300
    },
    {
      "epoch": 0.8971109984794728,
      "grad_norm": 0.46661046147346497,
      "learning_rate": 0.00022376600532590578,
      "loss": 4.5343585968017575,
      "step": 5310
    },
    {
      "epoch": 0.8988004730528806,
      "grad_norm": 0.4886866509914398,
      "learning_rate": 0.00022334855464950775,
      "loss": 4.5834095001220705,
      "step": 5320
    },
    {
      "epoch": 0.9004899476262882,
      "grad_norm": 0.5262774229049683,
      "learning_rate": 0.00022293035601495708,
      "loss": 4.579534912109375,
      "step": 5330
    },
    {
      "epoch": 0.9021794221996959,
      "grad_norm": 0.5163218975067139,
      "learning_rate": 0.00022251141368675607,
      "loss": 4.577048492431641,
      "step": 5340
    },
    {
      "epoch": 0.9038688967731036,
      "grad_norm": 0.5345433950424194,
      "learning_rate": 0.00022209173193699067,
      "loss": 4.582790374755859,
      "step": 5350
    },
    {
      "epoch": 0.9055583713465112,
      "grad_norm": 0.5151252150535583,
      "learning_rate": 0.00022167131504528695,
      "loss": 4.594097900390625,
      "step": 5360
    },
    {
      "epoch": 0.9072478459199189,
      "grad_norm": 0.47062498331069946,
      "learning_rate": 0.00022125016729876743,
      "loss": 4.574803161621094,
      "step": 5370
    },
    {
      "epoch": 0.9089373204933265,
      "grad_norm": 0.49667978286743164,
      "learning_rate": 0.00022082829299200743,
      "loss": 4.580567932128906,
      "step": 5380
    },
    {
      "epoch": 0.9106267950667343,
      "grad_norm": 0.48394060134887695,
      "learning_rate": 0.00022040569642699112,
      "loss": 4.555598449707031,
      "step": 5390
    },
    {
      "epoch": 0.9123162696401419,
      "grad_norm": 0.48837390542030334,
      "learning_rate": 0.00021998238191306798,
      "loss": 4.534821319580078,
      "step": 5400
    },
    {
      "epoch": 0.9140057442135496,
      "grad_norm": 0.5261453986167908,
      "learning_rate": 0.00021955835376690841,
      "loss": 4.546956634521484,
      "step": 5410
    },
    {
      "epoch": 0.9156952187869573,
      "grad_norm": 0.5199710130691528,
      "learning_rate": 0.00021913361631246004,
      "loss": 4.561407852172851,
      "step": 5420
    },
    {
      "epoch": 0.9173846933603649,
      "grad_norm": 0.5369474291801453,
      "learning_rate": 0.0002187081738809036,
      "loss": 4.550098419189453,
      "step": 5430
    },
    {
      "epoch": 0.9190741679337726,
      "grad_norm": 0.5480945110321045,
      "learning_rate": 0.00021828203081060858,
      "loss": 4.559786224365235,
      "step": 5440
    },
    {
      "epoch": 0.9207636425071802,
      "grad_norm": 0.5149338245391846,
      "learning_rate": 0.00021785519144708912,
      "loss": 4.534018325805664,
      "step": 5450
    },
    {
      "epoch": 0.922453117080588,
      "grad_norm": 0.5365586280822754,
      "learning_rate": 0.00021742766014295976,
      "loss": 4.546533584594727,
      "step": 5460
    },
    {
      "epoch": 0.9241425916539956,
      "grad_norm": 0.5260055661201477,
      "learning_rate": 0.00021699944125789096,
      "loss": 4.534712600708008,
      "step": 5470
    },
    {
      "epoch": 0.9258320662274033,
      "grad_norm": 0.4802268147468567,
      "learning_rate": 0.00021657053915856455,
      "loss": 4.560755920410156,
      "step": 5480
    },
    {
      "epoch": 0.927521540800811,
      "grad_norm": 0.4982668459415436,
      "learning_rate": 0.0002161409582186294,
      "loss": 4.584963989257813,
      "step": 5490
    },
    {
      "epoch": 0.9292110153742186,
      "grad_norm": 0.49544209241867065,
      "learning_rate": 0.0002157107028186567,
      "loss": 4.547665786743164,
      "step": 5500
    },
    {
      "epoch": 0.9292110153742186,
      "eval_loss": 4.52970552444458,
      "eval_runtime": 3.6346,
      "eval_samples_per_second": 275.136,
      "eval_steps_per_second": 5.778,
      "step": 5500
    },
    {
      "epoch": 0.9309004899476263,
      "grad_norm": 0.4983241558074951,
      "learning_rate": 0.00021527977734609537,
      "loss": 4.547625732421875,
      "step": 5510
    },
    {
      "epoch": 0.9325899645210339,
      "grad_norm": 0.5012770295143127,
      "learning_rate": 0.00021484818619522722,
      "loss": 4.557040023803711,
      "step": 5520
    },
    {
      "epoch": 0.9342794390944417,
      "grad_norm": 0.5078200101852417,
      "learning_rate": 0.00021441593376712224,
      "loss": 4.553184890747071,
      "step": 5530
    },
    {
      "epoch": 0.9359689136678493,
      "grad_norm": 0.48705384135246277,
      "learning_rate": 0.0002139830244695935,
      "loss": 4.5813232421875,
      "step": 5540
    },
    {
      "epoch": 0.937658388241257,
      "grad_norm": 0.5023474097251892,
      "learning_rate": 0.00021354946271715265,
      "loss": 4.552815628051758,
      "step": 5550
    },
    {
      "epoch": 0.9393478628146646,
      "grad_norm": 0.5058281421661377,
      "learning_rate": 0.00021311525293096444,
      "loss": 4.541165924072265,
      "step": 5560
    },
    {
      "epoch": 0.9410373373880723,
      "grad_norm": 0.5129496455192566,
      "learning_rate": 0.00021268039953880184,
      "loss": 4.529154968261719,
      "step": 5570
    },
    {
      "epoch": 0.94272681196148,
      "grad_norm": 0.5097109079360962,
      "learning_rate": 0.00021224490697500088,
      "loss": 4.535088348388672,
      "step": 5580
    },
    {
      "epoch": 0.9444162865348876,
      "grad_norm": 0.5103420615196228,
      "learning_rate": 0.00021180877968041552,
      "loss": 4.553527069091797,
      "step": 5590
    },
    {
      "epoch": 0.9461057611082954,
      "grad_norm": 0.4936409294605255,
      "learning_rate": 0.00021137202210237213,
      "loss": 4.54007568359375,
      "step": 5600
    },
    {
      "epoch": 0.947795235681703,
      "grad_norm": 0.5701144933700562,
      "learning_rate": 0.0002109346386946243,
      "loss": 4.558887100219726,
      "step": 5610
    },
    {
      "epoch": 0.9494847102551106,
      "grad_norm": 0.4890182912349701,
      "learning_rate": 0.00021049663391730752,
      "loss": 4.543179702758789,
      "step": 5620
    },
    {
      "epoch": 0.9511741848285183,
      "grad_norm": 0.5074143409729004,
      "learning_rate": 0.00021005801223689344,
      "loss": 4.5704292297363285,
      "step": 5630
    },
    {
      "epoch": 0.952863659401926,
      "grad_norm": 0.4767675995826721,
      "learning_rate": 0.00020961877812614458,
      "loss": 4.569948196411133,
      "step": 5640
    },
    {
      "epoch": 0.9545531339753337,
      "grad_norm": 0.5034293532371521,
      "learning_rate": 0.00020917893606406843,
      "loss": 4.524322128295898,
      "step": 5650
    },
    {
      "epoch": 0.9562426085487413,
      "grad_norm": 0.5619840621948242,
      "learning_rate": 0.0002087384905358722,
      "loss": 4.528865051269531,
      "step": 5660
    },
    {
      "epoch": 0.9579320831221491,
      "grad_norm": 0.5692474842071533,
      "learning_rate": 0.00020829744603291663,
      "loss": 4.5155292510986325,
      "step": 5670
    },
    {
      "epoch": 0.9596215576955567,
      "grad_norm": 0.504224419593811,
      "learning_rate": 0.00020785580705267047,
      "loss": 4.559905624389648,
      "step": 5680
    },
    {
      "epoch": 0.9613110322689643,
      "grad_norm": 0.563014805316925,
      "learning_rate": 0.00020741357809866447,
      "loss": 4.556017303466797,
      "step": 5690
    },
    {
      "epoch": 0.963000506842372,
      "grad_norm": 0.4872301518917084,
      "learning_rate": 0.0002069707636804457,
      "loss": 4.550839233398437,
      "step": 5700
    },
    {
      "epoch": 0.9646899814157797,
      "grad_norm": 0.5135483145713806,
      "learning_rate": 0.0002065273683135312,
      "loss": 4.550697708129883,
      "step": 5710
    },
    {
      "epoch": 0.9663794559891874,
      "grad_norm": 0.4852290451526642,
      "learning_rate": 0.00020608339651936224,
      "loss": 4.531842422485352,
      "step": 5720
    },
    {
      "epoch": 0.968068930562595,
      "grad_norm": 0.5045028924942017,
      "learning_rate": 0.00020563885282525802,
      "loss": 4.532521057128906,
      "step": 5730
    },
    {
      "epoch": 0.9697584051360028,
      "grad_norm": 0.530616044998169,
      "learning_rate": 0.00020519374176436968,
      "loss": 4.546891403198242,
      "step": 5740
    },
    {
      "epoch": 0.9714478797094104,
      "grad_norm": 0.49565091729164124,
      "learning_rate": 0.00020474806787563392,
      "loss": 4.533766555786133,
      "step": 5750
    },
    {
      "epoch": 0.973137354282818,
      "grad_norm": 0.5225724577903748,
      "learning_rate": 0.0002043018357037267,
      "loss": 4.542942810058594,
      "step": 5760
    },
    {
      "epoch": 0.9748268288562257,
      "grad_norm": 0.49189162254333496,
      "learning_rate": 0.00020385504979901712,
      "loss": 4.545899200439453,
      "step": 5770
    },
    {
      "epoch": 0.9765163034296334,
      "grad_norm": 0.5116291642189026,
      "learning_rate": 0.00020340771471752078,
      "loss": 4.532541656494141,
      "step": 5780
    },
    {
      "epoch": 0.9782057780030411,
      "grad_norm": 0.5132644772529602,
      "learning_rate": 0.0002029598350208534,
      "loss": 4.524928283691406,
      "step": 5790
    },
    {
      "epoch": 0.9798952525764487,
      "grad_norm": 0.4904372990131378,
      "learning_rate": 0.00020251141527618434,
      "loss": 4.532776641845703,
      "step": 5800
    },
    {
      "epoch": 0.9815847271498563,
      "grad_norm": 0.48598089814186096,
      "learning_rate": 0.00020206246005618998,
      "loss": 4.519465637207031,
      "step": 5810
    },
    {
      "epoch": 0.9832742017232641,
      "grad_norm": 0.5415476560592651,
      "learning_rate": 0.00020161297393900713,
      "loss": 4.512179565429688,
      "step": 5820
    },
    {
      "epoch": 0.9849636762966717,
      "grad_norm": 0.5061231255531311,
      "learning_rate": 0.00020116296150818623,
      "loss": 4.534863662719727,
      "step": 5830
    },
    {
      "epoch": 0.9866531508700794,
      "grad_norm": 0.5157834887504578,
      "learning_rate": 0.0002007124273526449,
      "loss": 4.50738639831543,
      "step": 5840
    },
    {
      "epoch": 0.988342625443487,
      "grad_norm": 0.509292483329773,
      "learning_rate": 0.00020026137606662077,
      "loss": 4.5266845703125,
      "step": 5850
    },
    {
      "epoch": 0.9900321000168948,
      "grad_norm": 0.5107020139694214,
      "learning_rate": 0.0001998098122496249,
      "loss": 4.533035659790039,
      "step": 5860
    },
    {
      "epoch": 0.9917215745903024,
      "grad_norm": 0.5432437062263489,
      "learning_rate": 0.00019935774050639472,
      "loss": 4.518278884887695,
      "step": 5870
    },
    {
      "epoch": 0.99341104916371,
      "grad_norm": 0.5360410213470459,
      "learning_rate": 0.0001989051654468473,
      "loss": 4.502675628662109,
      "step": 5880
    },
    {
      "epoch": 0.9951005237371178,
      "grad_norm": 0.5418276786804199,
      "learning_rate": 0.00019845209168603195,
      "loss": 4.5235343933105465,
      "step": 5890
    },
    {
      "epoch": 0.9967899983105254,
      "grad_norm": 0.5157185792922974,
      "learning_rate": 0.00019799852384408355,
      "loss": 4.524081420898438,
      "step": 5900
    },
    {
      "epoch": 0.9984794728839331,
      "grad_norm": 0.5043293237686157,
      "learning_rate": 0.00019754446654617527,
      "loss": 4.508223342895508,
      "step": 5910
    },
    {
      "epoch": 1.0001689474573408,
      "grad_norm": 0.5386601090431213,
      "learning_rate": 0.00019708992442247136,
      "loss": 4.5236083984375,
      "step": 5920
    },
    {
      "epoch": 1.0018584220307485,
      "grad_norm": 0.5341511368751526,
      "learning_rate": 0.0001966349021080801,
      "loss": 4.459320068359375,
      "step": 5930
    },
    {
      "epoch": 1.003547896604156,
      "grad_norm": 0.5038416981697083,
      "learning_rate": 0.0001961794042430062,
      "loss": 4.505880355834961,
      "step": 5940
    },
    {
      "epoch": 1.0052373711775637,
      "grad_norm": 0.47585076093673706,
      "learning_rate": 0.000195723435472104,
      "loss": 4.477125930786133,
      "step": 5950
    },
    {
      "epoch": 1.0069268457509715,
      "grad_norm": 0.49405696988105774,
      "learning_rate": 0.00019526700044502956,
      "loss": 4.483388137817383,
      "step": 5960
    },
    {
      "epoch": 1.0086163203243792,
      "grad_norm": 0.47832658886909485,
      "learning_rate": 0.0001948101038161937,
      "loss": 4.474266052246094,
      "step": 5970
    },
    {
      "epoch": 1.0103057948977867,
      "grad_norm": 0.470113068819046,
      "learning_rate": 0.0001943527502447141,
      "loss": 4.483303833007812,
      "step": 5980
    },
    {
      "epoch": 1.0119952694711944,
      "grad_norm": 0.4839136004447937,
      "learning_rate": 0.00019389494439436836,
      "loss": 4.453615188598633,
      "step": 5990
    },
    {
      "epoch": 1.0136847440446022,
      "grad_norm": 0.482327401638031,
      "learning_rate": 0.0001934366909335458,
      "loss": 4.491983413696289,
      "step": 6000
    },
    {
      "epoch": 1.0136847440446022,
      "eval_loss": 4.487085819244385,
      "eval_runtime": 4.7973,
      "eval_samples_per_second": 208.452,
      "eval_steps_per_second": 4.377,
      "step": 6000
    },
    {
      "epoch": 1.0153742186180097,
      "grad_norm": 0.5339483022689819,
      "learning_rate": 0.00019297799453520028,
      "loss": 4.500830459594726,
      "step": 6010
    },
    {
      "epoch": 1.0170636931914174,
      "grad_norm": 0.5642256736755371,
      "learning_rate": 0.00019251885987680252,
      "loss": 4.485746002197265,
      "step": 6020
    },
    {
      "epoch": 1.0187531677648252,
      "grad_norm": 0.5060975551605225,
      "learning_rate": 0.00019205929164029217,
      "loss": 4.475402450561523,
      "step": 6030
    },
    {
      "epoch": 1.020442642338233,
      "grad_norm": 0.49786120653152466,
      "learning_rate": 0.00019159929451203033,
      "loss": 4.486777114868164,
      "step": 6040
    },
    {
      "epoch": 1.0221321169116404,
      "grad_norm": 0.506598949432373,
      "learning_rate": 0.00019113887318275149,
      "loss": 4.489146041870117,
      "step": 6050
    },
    {
      "epoch": 1.0238215914850481,
      "grad_norm": 0.48270103335380554,
      "learning_rate": 0.00019067803234751603,
      "loss": 4.474691009521484,
      "step": 6060
    },
    {
      "epoch": 1.0255110660584559,
      "grad_norm": 0.48239970207214355,
      "learning_rate": 0.00019021677670566208,
      "loss": 4.4708606719970705,
      "step": 6070
    },
    {
      "epoch": 1.0272005406318634,
      "grad_norm": 0.4966093599796295,
      "learning_rate": 0.00018975511096075762,
      "loss": 4.505655670166016,
      "step": 6080
    },
    {
      "epoch": 1.0288900152052711,
      "grad_norm": 0.5429375767707825,
      "learning_rate": 0.00018929303982055272,
      "loss": 4.499347305297851,
      "step": 6090
    },
    {
      "epoch": 1.0305794897786789,
      "grad_norm": 0.4981507360935211,
      "learning_rate": 0.00018883056799693125,
      "loss": 4.461819839477539,
      "step": 6100
    },
    {
      "epoch": 1.0322689643520866,
      "grad_norm": 0.5121614336967468,
      "learning_rate": 0.00018836770020586315,
      "loss": 4.478689956665039,
      "step": 6110
    },
    {
      "epoch": 1.033958438925494,
      "grad_norm": 0.4835728406906128,
      "learning_rate": 0.00018790444116735595,
      "loss": 4.47772216796875,
      "step": 6120
    },
    {
      "epoch": 1.0356479134989018,
      "grad_norm": 0.4881206154823303,
      "learning_rate": 0.00018744079560540695,
      "loss": 4.479801177978516,
      "step": 6130
    },
    {
      "epoch": 1.0373373880723096,
      "grad_norm": 0.47434431314468384,
      "learning_rate": 0.000186976768247955,
      "loss": 4.480235290527344,
      "step": 6140
    },
    {
      "epoch": 1.039026862645717,
      "grad_norm": 0.48258504271507263,
      "learning_rate": 0.00018651236382683225,
      "loss": 4.469864273071289,
      "step": 6150
    },
    {
      "epoch": 1.0407163372191248,
      "grad_norm": 0.5025637745857239,
      "learning_rate": 0.0001860475870777157,
      "loss": 4.472750091552735,
      "step": 6160
    },
    {
      "epoch": 1.0424058117925326,
      "grad_norm": 0.4636594355106354,
      "learning_rate": 0.0001855824427400793,
      "loss": 4.450835418701172,
      "step": 6170
    },
    {
      "epoch": 1.0440952863659403,
      "grad_norm": 0.4901501536369324,
      "learning_rate": 0.00018511693555714535,
      "loss": 4.490735626220703,
      "step": 6180
    },
    {
      "epoch": 1.0457847609393478,
      "grad_norm": 0.5198561549186707,
      "learning_rate": 0.00018465107027583615,
      "loss": 4.474180221557617,
      "step": 6190
    },
    {
      "epoch": 1.0474742355127555,
      "grad_norm": 0.4723539352416992,
      "learning_rate": 0.00018418485164672574,
      "loss": 4.4745361328125,
      "step": 6200
    },
    {
      "epoch": 1.0491637100861633,
      "grad_norm": 0.5074954628944397,
      "learning_rate": 0.00018371828442399128,
      "loss": 4.469810485839844,
      "step": 6210
    },
    {
      "epoch": 1.0508531846595708,
      "grad_norm": 0.49918699264526367,
      "learning_rate": 0.00018325137336536464,
      "loss": 4.442096710205078,
      "step": 6220
    },
    {
      "epoch": 1.0525426592329785,
      "grad_norm": 0.5088530778884888,
      "learning_rate": 0.00018278412323208392,
      "loss": 4.484762573242188,
      "step": 6230
    },
    {
      "epoch": 1.0542321338063863,
      "grad_norm": 0.506341814994812,
      "learning_rate": 0.00018231653878884486,
      "loss": 4.486656188964844,
      "step": 6240
    },
    {
      "epoch": 1.055921608379794,
      "grad_norm": 0.5262649059295654,
      "learning_rate": 0.00018184862480375233,
      "loss": 4.455668640136719,
      "step": 6250
    },
    {
      "epoch": 1.0576110829532015,
      "grad_norm": 0.5115051865577698,
      "learning_rate": 0.00018138038604827153,
      "loss": 4.479043960571289,
      "step": 6260
    },
    {
      "epoch": 1.0593005575266092,
      "grad_norm": 0.50110924243927,
      "learning_rate": 0.0001809118272971795,
      "loss": 4.446685409545898,
      "step": 6270
    },
    {
      "epoch": 1.060990032100017,
      "grad_norm": 0.5022484660148621,
      "learning_rate": 0.0001804429533285164,
      "loss": 4.4593353271484375,
      "step": 6280
    },
    {
      "epoch": 1.0626795066734245,
      "grad_norm": 0.492165744304657,
      "learning_rate": 0.00017997376892353668,
      "loss": 4.496971511840821,
      "step": 6290
    },
    {
      "epoch": 1.0643689812468322,
      "grad_norm": 0.5134599208831787,
      "learning_rate": 0.0001795042788666605,
      "loss": 4.465629196166992,
      "step": 6300
    },
    {
      "epoch": 1.06605845582024,
      "grad_norm": 0.5151488184928894,
      "learning_rate": 0.00017903448794542488,
      "loss": 4.454899597167969,
      "step": 6310
    },
    {
      "epoch": 1.0677479303936477,
      "grad_norm": 0.5240500569343567,
      "learning_rate": 0.00017856440095043464,
      "loss": 4.481625747680664,
      "step": 6320
    },
    {
      "epoch": 1.0694374049670552,
      "grad_norm": 0.5187123417854309,
      "learning_rate": 0.00017809402267531405,
      "loss": 4.437789535522461,
      "step": 6330
    },
    {
      "epoch": 1.071126879540463,
      "grad_norm": 0.4693409502506256,
      "learning_rate": 0.00017762335791665735,
      "loss": 4.450423812866211,
      "step": 6340
    },
    {
      "epoch": 1.0728163541138707,
      "grad_norm": 0.5061246752738953,
      "learning_rate": 0.00017715241147398035,
      "loss": 4.46313705444336,
      "step": 6350
    },
    {
      "epoch": 1.0745058286872782,
      "grad_norm": 0.47927796840667725,
      "learning_rate": 0.00017668118814967126,
      "loss": 4.446915817260742,
      "step": 6360
    },
    {
      "epoch": 1.076195303260686,
      "grad_norm": 0.47587907314300537,
      "learning_rate": 0.00017620969274894163,
      "loss": 4.461398696899414,
      "step": 6370
    },
    {
      "epoch": 1.0778847778340936,
      "grad_norm": 0.5091392397880554,
      "learning_rate": 0.00017573793007977763,
      "loss": 4.450970458984375,
      "step": 6380
    },
    {
      "epoch": 1.0795742524075012,
      "grad_norm": 0.5105127692222595,
      "learning_rate": 0.0001752659049528906,
      "loss": 4.458657455444336,
      "step": 6390
    },
    {
      "epoch": 1.081263726980909,
      "grad_norm": 0.5196726322174072,
      "learning_rate": 0.00017479362218166854,
      "loss": 4.444008636474609,
      "step": 6400
    },
    {
      "epoch": 1.0829532015543166,
      "grad_norm": 0.4891359210014343,
      "learning_rate": 0.0001743210865821265,
      "loss": 4.436445236206055,
      "step": 6410
    },
    {
      "epoch": 1.0846426761277244,
      "grad_norm": 0.5141095519065857,
      "learning_rate": 0.0001738483029728578,
      "loss": 4.455481338500976,
      "step": 6420
    },
    {
      "epoch": 1.0863321507011319,
      "grad_norm": 0.5223525166511536,
      "learning_rate": 0.00017337527617498474,
      "loss": 4.485405731201172,
      "step": 6430
    },
    {
      "epoch": 1.0880216252745396,
      "grad_norm": 0.4939091205596924,
      "learning_rate": 0.0001729020110121096,
      "loss": 4.448784255981446,
      "step": 6440
    },
    {
      "epoch": 1.0897110998479473,
      "grad_norm": 0.49695253372192383,
      "learning_rate": 0.0001724285123102652,
      "loss": 4.4587146759033205,
      "step": 6450
    },
    {
      "epoch": 1.091400574421355,
      "grad_norm": 0.4882517158985138,
      "learning_rate": 0.00017195478489786593,
      "loss": 4.43580207824707,
      "step": 6460
    },
    {
      "epoch": 1.0930900489947626,
      "grad_norm": 0.4971882998943329,
      "learning_rate": 0.00017148083360565836,
      "loss": 4.436479949951172,
      "step": 6470
    },
    {
      "epoch": 1.0947795235681703,
      "grad_norm": 0.4835260808467865,
      "learning_rate": 0.00017100666326667202,
      "loss": 4.476963043212891,
      "step": 6480
    },
    {
      "epoch": 1.096468998141578,
      "grad_norm": 0.4847490191459656,
      "learning_rate": 0.00017053227871617027,
      "loss": 4.449015426635742,
      "step": 6490
    },
    {
      "epoch": 1.0981584727149856,
      "grad_norm": 0.5305824279785156,
      "learning_rate": 0.00017005768479160064,
      "loss": 4.452330780029297,
      "step": 6500
    },
    {
      "epoch": 1.0981584727149856,
      "eval_loss": 4.447469711303711,
      "eval_runtime": 4.0239,
      "eval_samples_per_second": 248.518,
      "eval_steps_per_second": 5.219,
      "step": 6500
    },
    {
      "epoch": 1.0998479472883933,
      "grad_norm": 0.4943171739578247,
      "learning_rate": 0.0001695828863325459,
      "loss": 4.467470932006836,
      "step": 6510
    },
    {
      "epoch": 1.101537421861801,
      "grad_norm": 0.474933385848999,
      "learning_rate": 0.00016910788818067434,
      "loss": 4.4371185302734375,
      "step": 6520
    },
    {
      "epoch": 1.1032268964352085,
      "grad_norm": 0.5118041634559631,
      "learning_rate": 0.0001686326951796907,
      "loss": 4.451096725463867,
      "step": 6530
    },
    {
      "epoch": 1.1049163710086163,
      "grad_norm": 0.5289651155471802,
      "learning_rate": 0.00016815731217528667,
      "loss": 4.448075485229492,
      "step": 6540
    },
    {
      "epoch": 1.106605845582024,
      "grad_norm": 0.5182890295982361,
      "learning_rate": 0.00016768174401509143,
      "loss": 4.467396926879883,
      "step": 6550
    },
    {
      "epoch": 1.1082953201554318,
      "grad_norm": 0.5209820866584778,
      "learning_rate": 0.0001672059955486223,
      "loss": 4.459186172485351,
      "step": 6560
    },
    {
      "epoch": 1.1099847947288393,
      "grad_norm": 0.48584309220314026,
      "learning_rate": 0.000166730071627235,
      "loss": 4.46546516418457,
      "step": 6570
    },
    {
      "epoch": 1.111674269302247,
      "grad_norm": 0.5017306804656982,
      "learning_rate": 0.00016625397710407487,
      "loss": 4.452592086791992,
      "step": 6580
    },
    {
      "epoch": 1.1133637438756547,
      "grad_norm": 0.46485376358032227,
      "learning_rate": 0.00016577771683402647,
      "loss": 4.46324348449707,
      "step": 6590
    },
    {
      "epoch": 1.1150532184490622,
      "grad_norm": 0.5154596567153931,
      "learning_rate": 0.00016530129567366483,
      "loss": 4.457768249511719,
      "step": 6600
    },
    {
      "epoch": 1.11674269302247,
      "grad_norm": 0.49490463733673096,
      "learning_rate": 0.0001648247184812054,
      "loss": 4.427638244628906,
      "step": 6610
    },
    {
      "epoch": 1.1184321675958777,
      "grad_norm": 0.4721022844314575,
      "learning_rate": 0.00016434799011645507,
      "loss": 4.4389793395996096,
      "step": 6620
    },
    {
      "epoch": 1.1201216421692854,
      "grad_norm": 0.4648183286190033,
      "learning_rate": 0.00016387111544076193,
      "loss": 4.460124969482422,
      "step": 6630
    },
    {
      "epoch": 1.121811116742693,
      "grad_norm": 0.5035665035247803,
      "learning_rate": 0.00016339409931696625,
      "loss": 4.439287185668945,
      "step": 6640
    },
    {
      "epoch": 1.1235005913161007,
      "grad_norm": 0.4910880923271179,
      "learning_rate": 0.00016291694660935065,
      "loss": 4.456634140014648,
      "step": 6650
    },
    {
      "epoch": 1.1251900658895084,
      "grad_norm": 0.48906245827674866,
      "learning_rate": 0.00016243966218359047,
      "loss": 4.428804016113281,
      "step": 6660
    },
    {
      "epoch": 1.126879540462916,
      "grad_norm": 0.5756556391716003,
      "learning_rate": 0.00016196225090670435,
      "loss": 4.411157608032227,
      "step": 6670
    },
    {
      "epoch": 1.1285690150363237,
      "grad_norm": 0.49011167883872986,
      "learning_rate": 0.0001614847176470043,
      "loss": 4.435109329223633,
      "step": 6680
    },
    {
      "epoch": 1.1302584896097314,
      "grad_norm": 0.4775542616844177,
      "learning_rate": 0.00016100706727404645,
      "loss": 4.428675842285156,
      "step": 6690
    },
    {
      "epoch": 1.131947964183139,
      "grad_norm": 0.5201391577720642,
      "learning_rate": 0.00016052930465858094,
      "loss": 4.4389808654785154,
      "step": 6700
    },
    {
      "epoch": 1.1336374387565467,
      "grad_norm": 0.49004724621772766,
      "learning_rate": 0.00016005143467250267,
      "loss": 4.459021377563476,
      "step": 6710
    },
    {
      "epoch": 1.1353269133299544,
      "grad_norm": 0.49011871218681335,
      "learning_rate": 0.00015957346218880124,
      "loss": 4.455972290039062,
      "step": 6720
    },
    {
      "epoch": 1.1370163879033621,
      "grad_norm": 0.5173168182373047,
      "learning_rate": 0.0001590953920815117,
      "loss": 4.443459701538086,
      "step": 6730
    },
    {
      "epoch": 1.1387058624767696,
      "grad_norm": 0.47700756788253784,
      "learning_rate": 0.00015861722922566436,
      "loss": 4.435110473632813,
      "step": 6740
    },
    {
      "epoch": 1.1403953370501774,
      "grad_norm": 0.5626063942909241,
      "learning_rate": 0.00015813897849723544,
      "loss": 4.432453536987305,
      "step": 6750
    },
    {
      "epoch": 1.142084811623585,
      "grad_norm": 0.49542316794395447,
      "learning_rate": 0.0001576606447730972,
      "loss": 4.4374950408935545,
      "step": 6760
    },
    {
      "epoch": 1.1437742861969928,
      "grad_norm": 0.5116281509399414,
      "learning_rate": 0.0001571822329309682,
      "loss": 4.423119354248047,
      "step": 6770
    },
    {
      "epoch": 1.1454637607704004,
      "grad_norm": 0.4868847131729126,
      "learning_rate": 0.00015670374784936371,
      "loss": 4.4402107238769535,
      "step": 6780
    },
    {
      "epoch": 1.147153235343808,
      "grad_norm": 0.4938635230064392,
      "learning_rate": 0.00015622519440754566,
      "loss": 4.424631881713867,
      "step": 6790
    },
    {
      "epoch": 1.1488427099172158,
      "grad_norm": 0.5740174651145935,
      "learning_rate": 0.0001557465774854732,
      "loss": 4.450838470458985,
      "step": 6800
    },
    {
      "epoch": 1.1505321844906233,
      "grad_norm": 0.4828670918941498,
      "learning_rate": 0.0001552679019637528,
      "loss": 4.438276290893555,
      "step": 6810
    },
    {
      "epoch": 1.152221659064031,
      "grad_norm": 0.4659689664840698,
      "learning_rate": 0.00015478917272358848,
      "loss": 4.426282501220703,
      "step": 6820
    },
    {
      "epoch": 1.1539111336374388,
      "grad_norm": 0.4927656352519989,
      "learning_rate": 0.000154310394646732,
      "loss": 4.464373016357422,
      "step": 6830
    },
    {
      "epoch": 1.1556006082108463,
      "grad_norm": 0.5161291360855103,
      "learning_rate": 0.00015383157261543318,
      "loss": 4.416297531127929,
      "step": 6840
    },
    {
      "epoch": 1.157290082784254,
      "grad_norm": 0.4933563768863678,
      "learning_rate": 0.00015335271151239,
      "loss": 4.420982742309571,
      "step": 6850
    },
    {
      "epoch": 1.1589795573576618,
      "grad_norm": 0.4847005307674408,
      "learning_rate": 0.00015287381622069892,
      "loss": 4.416022872924804,
      "step": 6860
    },
    {
      "epoch": 1.1606690319310695,
      "grad_norm": 0.4981960654258728,
      "learning_rate": 0.00015239489162380504,
      "loss": 4.422767639160156,
      "step": 6870
    },
    {
      "epoch": 1.162358506504477,
      "grad_norm": 0.5001937747001648,
      "learning_rate": 0.0001519159426054522,
      "loss": 4.4368339538574215,
      "step": 6880
    },
    {
      "epoch": 1.1640479810778848,
      "grad_norm": 0.5044972896575928,
      "learning_rate": 0.0001514369740496334,
      "loss": 4.411078643798828,
      "step": 6890
    },
    {
      "epoch": 1.1657374556512925,
      "grad_norm": 0.4734691083431244,
      "learning_rate": 0.00015095799084054073,
      "loss": 4.438079071044922,
      "step": 6900
    },
    {
      "epoch": 1.1674269302247002,
      "grad_norm": 0.49377161264419556,
      "learning_rate": 0.00015047899786251587,
      "loss": 4.442370986938476,
      "step": 6910
    },
    {
      "epoch": 1.1691164047981077,
      "grad_norm": 0.5010132193565369,
      "learning_rate": 0.00015,
      "loss": 4.442108917236328,
      "step": 6920
    },
    {
      "epoch": 1.1708058793715155,
      "grad_norm": 0.5035766959190369,
      "learning_rate": 0.0001495210021374841,
      "loss": 4.430604553222656,
      "step": 6930
    },
    {
      "epoch": 1.1724953539449232,
      "grad_norm": 0.4899141788482666,
      "learning_rate": 0.00014904200915945927,
      "loss": 4.435578918457031,
      "step": 6940
    },
    {
      "epoch": 1.1741848285183307,
      "grad_norm": 0.4718686044216156,
      "learning_rate": 0.00014856302595036663,
      "loss": 4.429093551635742,
      "step": 6950
    },
    {
      "epoch": 1.1758743030917385,
      "grad_norm": 0.4881162941455841,
      "learning_rate": 0.00014808405739454776,
      "loss": 4.408749008178711,
      "step": 6960
    },
    {
      "epoch": 1.1775637776651462,
      "grad_norm": 0.46740713715553284,
      "learning_rate": 0.00014760510837619493,
      "loss": 4.419464492797852,
      "step": 6970
    },
    {
      "epoch": 1.1792532522385537,
      "grad_norm": 0.4737609922885895,
      "learning_rate": 0.00014712618377930105,
      "loss": 4.421468353271484,
      "step": 6980
    },
    {
      "epoch": 1.1809427268119614,
      "grad_norm": 0.4975055754184723,
      "learning_rate": 0.00014664728848760996,
      "loss": 4.422280502319336,
      "step": 6990
    },
    {
      "epoch": 1.1826322013853692,
      "grad_norm": 0.4839191734790802,
      "learning_rate": 0.00014616842738456682,
      "loss": 4.395424652099609,
      "step": 7000
    },
    {
      "epoch": 1.1826322013853692,
      "eval_loss": 4.412718772888184,
      "eval_runtime": 4.0717,
      "eval_samples_per_second": 245.6,
      "eval_steps_per_second": 5.158,
      "step": 7000
    },
    {
      "epoch": 1.184321675958777,
      "grad_norm": 0.49278581142425537,
      "learning_rate": 0.000145689605353268,
      "loss": 4.424203109741211,
      "step": 7010
    },
    {
      "epoch": 1.1860111505321844,
      "grad_norm": 0.49644234776496887,
      "learning_rate": 0.00014521082727641152,
      "loss": 4.395336151123047,
      "step": 7020
    },
    {
      "epoch": 1.1877006251055922,
      "grad_norm": 0.483456552028656,
      "learning_rate": 0.0001447320980362472,
      "loss": 4.440347671508789,
      "step": 7030
    },
    {
      "epoch": 1.189390099679,
      "grad_norm": 0.5150992274284363,
      "learning_rate": 0.00014425342251452679,
      "loss": 4.393960571289062,
      "step": 7040
    },
    {
      "epoch": 1.1910795742524076,
      "grad_norm": 0.47316014766693115,
      "learning_rate": 0.00014377480559245434,
      "loss": 4.433261108398438,
      "step": 7050
    },
    {
      "epoch": 1.1927690488258151,
      "grad_norm": 0.5043189525604248,
      "learning_rate": 0.00014329625215063629,
      "loss": 4.437650680541992,
      "step": 7060
    },
    {
      "epoch": 1.1944585233992229,
      "grad_norm": 0.49998390674591064,
      "learning_rate": 0.00014281776706903177,
      "loss": 4.40019416809082,
      "step": 7070
    },
    {
      "epoch": 1.1961479979726306,
      "grad_norm": 0.5133141279220581,
      "learning_rate": 0.0001423393552269028,
      "loss": 4.417116928100586,
      "step": 7080
    },
    {
      "epoch": 1.1978374725460381,
      "grad_norm": 0.513031005859375,
      "learning_rate": 0.00014186102150276454,
      "loss": 4.438409805297852,
      "step": 7090
    },
    {
      "epoch": 1.1995269471194459,
      "grad_norm": 0.4915519058704376,
      "learning_rate": 0.00014138277077433567,
      "loss": 4.4253074645996096,
      "step": 7100
    },
    {
      "epoch": 1.2012164216928536,
      "grad_norm": 0.5202800035476685,
      "learning_rate": 0.00014090460791848827,
      "loss": 4.41809310913086,
      "step": 7110
    },
    {
      "epoch": 1.202905896266261,
      "grad_norm": 0.49077826738357544,
      "learning_rate": 0.00014042653781119868,
      "loss": 4.397499465942383,
      "step": 7120
    },
    {
      "epoch": 1.2045953708396688,
      "grad_norm": 0.4648706912994385,
      "learning_rate": 0.0001399485653274973,
      "loss": 4.408271026611328,
      "step": 7130
    },
    {
      "epoch": 1.2062848454130766,
      "grad_norm": 0.4614482820034027,
      "learning_rate": 0.00013947069534141904,
      "loss": 4.425214004516602,
      "step": 7140
    },
    {
      "epoch": 1.207974319986484,
      "grad_norm": 0.4744400084018707,
      "learning_rate": 0.00013899293272595355,
      "loss": 4.440077590942383,
      "step": 7150
    },
    {
      "epoch": 1.2096637945598918,
      "grad_norm": 0.46556323766708374,
      "learning_rate": 0.0001385152823529957,
      "loss": 4.412957382202149,
      "step": 7160
    },
    {
      "epoch": 1.2113532691332995,
      "grad_norm": 0.4939349591732025,
      "learning_rate": 0.00013803774909329567,
      "loss": 4.405846023559571,
      "step": 7170
    },
    {
      "epoch": 1.2130427437067073,
      "grad_norm": 0.47055721282958984,
      "learning_rate": 0.0001375603378164095,
      "loss": 4.382000350952149,
      "step": 7180
    },
    {
      "epoch": 1.2147322182801148,
      "grad_norm": 0.47987523674964905,
      "learning_rate": 0.00013708305339064933,
      "loss": 4.415153121948242,
      "step": 7190
    },
    {
      "epoch": 1.2164216928535225,
      "grad_norm": 0.4784037470817566,
      "learning_rate": 0.00013660590068303373,
      "loss": 4.4415229797363285,
      "step": 7200
    },
    {
      "epoch": 1.2181111674269303,
      "grad_norm": 0.500056803226471,
      "learning_rate": 0.00013612888455923804,
      "loss": 4.416479873657226,
      "step": 7210
    },
    {
      "epoch": 1.219800642000338,
      "grad_norm": 0.4778987169265747,
      "learning_rate": 0.0001356520098835449,
      "loss": 4.442354583740235,
      "step": 7220
    },
    {
      "epoch": 1.2214901165737455,
      "grad_norm": 0.5005702376365662,
      "learning_rate": 0.00013517528151879457,
      "loss": 4.411639404296875,
      "step": 7230
    },
    {
      "epoch": 1.2231795911471532,
      "grad_norm": 0.4689568281173706,
      "learning_rate": 0.0001346987043263352,
      "loss": 4.414199447631836,
      "step": 7240
    },
    {
      "epoch": 1.224869065720561,
      "grad_norm": 0.4993502199649811,
      "learning_rate": 0.00013422228316597356,
      "loss": 4.432155609130859,
      "step": 7250
    },
    {
      "epoch": 1.2265585402939685,
      "grad_norm": 0.4782608151435852,
      "learning_rate": 0.00013374602289592508,
      "loss": 4.431560897827149,
      "step": 7260
    },
    {
      "epoch": 1.2282480148673762,
      "grad_norm": 0.5125144124031067,
      "learning_rate": 0.00013326992837276494,
      "loss": 4.405394744873047,
      "step": 7270
    },
    {
      "epoch": 1.229937489440784,
      "grad_norm": 0.48408523201942444,
      "learning_rate": 0.0001327940044513777,
      "loss": 4.4137004852294925,
      "step": 7280
    },
    {
      "epoch": 1.2316269640141915,
      "grad_norm": 0.4888753294944763,
      "learning_rate": 0.00013231825598490854,
      "loss": 4.409386062622071,
      "step": 7290
    },
    {
      "epoch": 1.2333164385875992,
      "grad_norm": 0.47923538088798523,
      "learning_rate": 0.0001318426878247133,
      "loss": 4.4191631317138675,
      "step": 7300
    },
    {
      "epoch": 1.235005913161007,
      "grad_norm": 0.46775540709495544,
      "learning_rate": 0.00013136730482030928,
      "loss": 4.423541259765625,
      "step": 7310
    },
    {
      "epoch": 1.2366953877344147,
      "grad_norm": 0.48620909452438354,
      "learning_rate": 0.0001308921118193257,
      "loss": 4.431262969970703,
      "step": 7320
    },
    {
      "epoch": 1.2383848623078222,
      "grad_norm": 0.5028111338615417,
      "learning_rate": 0.00013041711366745408,
      "loss": 4.423612976074219,
      "step": 7330
    },
    {
      "epoch": 1.24007433688123,
      "grad_norm": 0.4982888996601105,
      "learning_rate": 0.00012994231520839934,
      "loss": 4.428596878051758,
      "step": 7340
    },
    {
      "epoch": 1.2417638114546377,
      "grad_norm": 0.5141102075576782,
      "learning_rate": 0.0001294677212838297,
      "loss": 4.398578262329101,
      "step": 7350
    },
    {
      "epoch": 1.2434532860280454,
      "grad_norm": 0.48103561997413635,
      "learning_rate": 0.00012899333673332795,
      "loss": 4.439675140380859,
      "step": 7360
    },
    {
      "epoch": 1.245142760601453,
      "grad_norm": 0.5084096789360046,
      "learning_rate": 0.00012851916639434164,
      "loss": 4.3824302673339846,
      "step": 7370
    },
    {
      "epoch": 1.2468322351748606,
      "grad_norm": 0.4776511788368225,
      "learning_rate": 0.00012804521510213407,
      "loss": 4.402749633789062,
      "step": 7380
    },
    {
      "epoch": 1.2485217097482684,
      "grad_norm": 0.499318391084671,
      "learning_rate": 0.00012757148768973483,
      "loss": 4.405498886108399,
      "step": 7390
    },
    {
      "epoch": 1.2502111843216759,
      "grad_norm": 0.4898117184638977,
      "learning_rate": 0.00012709798898789042,
      "loss": 4.4396411895751955,
      "step": 7400
    },
    {
      "epoch": 1.2519006588950836,
      "grad_norm": 0.4777224361896515,
      "learning_rate": 0.00012662472382501524,
      "loss": 4.409711074829102,
      "step": 7410
    },
    {
      "epoch": 1.2535901334684914,
      "grad_norm": 0.48530757427215576,
      "learning_rate": 0.0001261516970271422,
      "loss": 4.4214935302734375,
      "step": 7420
    },
    {
      "epoch": 1.2552796080418989,
      "grad_norm": 0.48434415459632874,
      "learning_rate": 0.0001256789134178735,
      "loss": 4.438081741333008,
      "step": 7430
    },
    {
      "epoch": 1.2569690826153066,
      "grad_norm": 0.4974631071090698,
      "learning_rate": 0.00012520637781833144,
      "loss": 4.407797622680664,
      "step": 7440
    },
    {
      "epoch": 1.2586585571887143,
      "grad_norm": 0.4732743799686432,
      "learning_rate": 0.0001247340950471094,
      "loss": 4.418028259277344,
      "step": 7450
    },
    {
      "epoch": 1.2603480317621218,
      "grad_norm": 0.5003547072410583,
      "learning_rate": 0.0001242620699202224,
      "loss": 4.400883483886719,
      "step": 7460
    },
    {
      "epoch": 1.2620375063355296,
      "grad_norm": 0.49987900257110596,
      "learning_rate": 0.00012379030725105837,
      "loss": 4.402442169189453,
      "step": 7470
    },
    {
      "epoch": 1.2637269809089373,
      "grad_norm": 0.49416637420654297,
      "learning_rate": 0.00012331881185032872,
      "loss": 4.388990020751953,
      "step": 7480
    },
    {
      "epoch": 1.265416455482345,
      "grad_norm": 0.5343226194381714,
      "learning_rate": 0.00012284758852601962,
      "loss": 4.411848449707032,
      "step": 7490
    },
    {
      "epoch": 1.2671059300557528,
      "grad_norm": 0.5128340125083923,
      "learning_rate": 0.00012237664208334263,
      "loss": 4.403173446655273,
      "step": 7500
    },
    {
      "epoch": 1.2671059300557528,
      "eval_loss": 4.382744789123535,
      "eval_runtime": 3.7472,
      "eval_samples_per_second": 266.869,
      "eval_steps_per_second": 5.604,
      "step": 7500
    },
    {
      "epoch": 1.2687954046291603,
      "grad_norm": 0.48278650641441345,
      "learning_rate": 0.00012190597732468595,
      "loss": 4.407323837280273,
      "step": 7510
    },
    {
      "epoch": 1.270484879202568,
      "grad_norm": 0.48528528213500977,
      "learning_rate": 0.00012143559904956533,
      "loss": 4.389751815795899,
      "step": 7520
    },
    {
      "epoch": 1.2721743537759758,
      "grad_norm": 0.4944697320461273,
      "learning_rate": 0.00012096551205457511,
      "loss": 4.385165786743164,
      "step": 7530
    },
    {
      "epoch": 1.2738638283493833,
      "grad_norm": 0.5002730488777161,
      "learning_rate": 0.00012049572113333949,
      "loss": 4.374062347412109,
      "step": 7540
    },
    {
      "epoch": 1.275553302922791,
      "grad_norm": 0.46715047955513,
      "learning_rate": 0.00012002623107646327,
      "loss": 4.394298553466797,
      "step": 7550
    },
    {
      "epoch": 1.2772427774961987,
      "grad_norm": 0.4903099834918976,
      "learning_rate": 0.00011955704667148361,
      "loss": 4.400055694580078,
      "step": 7560
    },
    {
      "epoch": 1.2789322520696063,
      "grad_norm": 0.5333164930343628,
      "learning_rate": 0.00011908817270282048,
      "loss": 4.424139404296875,
      "step": 7570
    },
    {
      "epoch": 1.280621726643014,
      "grad_norm": 0.47946473956108093,
      "learning_rate": 0.00011861961395172844,
      "loss": 4.419405746459961,
      "step": 7580
    },
    {
      "epoch": 1.2823112012164217,
      "grad_norm": 0.4778226315975189,
      "learning_rate": 0.00011815137519624767,
      "loss": 4.414478302001953,
      "step": 7590
    },
    {
      "epoch": 1.2840006757898292,
      "grad_norm": 0.4878886342048645,
      "learning_rate": 0.0001176834612111551,
      "loss": 4.384803009033203,
      "step": 7600
    },
    {
      "epoch": 1.285690150363237,
      "grad_norm": 0.4819967746734619,
      "learning_rate": 0.0001172158767679161,
      "loss": 4.3730110168457035,
      "step": 7610
    },
    {
      "epoch": 1.2873796249366447,
      "grad_norm": 0.4928823411464691,
      "learning_rate": 0.00011674862663463538,
      "loss": 4.3778236389160154,
      "step": 7620
    },
    {
      "epoch": 1.2890690995100524,
      "grad_norm": 0.4724312722682953,
      "learning_rate": 0.00011628171557600869,
      "loss": 4.387655639648438,
      "step": 7630
    },
    {
      "epoch": 1.2907585740834602,
      "grad_norm": 0.5023632049560547,
      "learning_rate": 0.0001158151483532742,
      "loss": 4.366682052612305,
      "step": 7640
    },
    {
      "epoch": 1.2924480486568677,
      "grad_norm": 0.47042906284332275,
      "learning_rate": 0.00011534892972416382,
      "loss": 4.3992149353027346,
      "step": 7650
    },
    {
      "epoch": 1.2941375232302754,
      "grad_norm": 0.5019961595535278,
      "learning_rate": 0.00011488306444285465,
      "loss": 4.408546829223633,
      "step": 7660
    },
    {
      "epoch": 1.2958269978036832,
      "grad_norm": 0.4686186909675598,
      "learning_rate": 0.0001144175572599207,
      "loss": 4.392362976074219,
      "step": 7670
    },
    {
      "epoch": 1.2975164723770907,
      "grad_norm": 0.5097217559814453,
      "learning_rate": 0.00011395241292228435,
      "loss": 4.350882339477539,
      "step": 7680
    },
    {
      "epoch": 1.2992059469504984,
      "grad_norm": 0.5009888410568237,
      "learning_rate": 0.00011348763617316781,
      "loss": 4.407807159423828,
      "step": 7690
    },
    {
      "epoch": 1.3008954215239061,
      "grad_norm": 0.4623536765575409,
      "learning_rate": 0.00011302323175204497,
      "loss": 4.383738708496094,
      "step": 7700
    },
    {
      "epoch": 1.3025848960973136,
      "grad_norm": 0.49098923802375793,
      "learning_rate": 0.00011255920439459302,
      "loss": 4.3777015686035154,
      "step": 7710
    },
    {
      "epoch": 1.3042743706707214,
      "grad_norm": 0.47158893942832947,
      "learning_rate": 0.00011209555883264406,
      "loss": 4.398603439331055,
      "step": 7720
    },
    {
      "epoch": 1.3059638452441291,
      "grad_norm": 0.4723564684391022,
      "learning_rate": 0.00011163229979413685,
      "loss": 4.379953384399414,
      "step": 7730
    },
    {
      "epoch": 1.3076533198175366,
      "grad_norm": 0.478575199842453,
      "learning_rate": 0.00011116943200306871,
      "loss": 4.369690322875977,
      "step": 7740
    },
    {
      "epoch": 1.3093427943909444,
      "grad_norm": 0.4801791310310364,
      "learning_rate": 0.00011070696017944728,
      "loss": 4.421099853515625,
      "step": 7750
    },
    {
      "epoch": 1.311032268964352,
      "grad_norm": 0.5147274732589722,
      "learning_rate": 0.00011024488903924235,
      "loss": 4.396934127807617,
      "step": 7760
    },
    {
      "epoch": 1.3127217435377598,
      "grad_norm": 0.4905327558517456,
      "learning_rate": 0.00010978322329433796,
      "loss": 4.368836975097656,
      "step": 7770
    },
    {
      "epoch": 1.3144112181111673,
      "grad_norm": 0.47583821415901184,
      "learning_rate": 0.00010932196765248396,
      "loss": 4.351024627685547,
      "step": 7780
    },
    {
      "epoch": 1.316100692684575,
      "grad_norm": 0.4749636650085449,
      "learning_rate": 0.0001088611268172485,
      "loss": 4.381603622436524,
      "step": 7790
    },
    {
      "epoch": 1.3177901672579828,
      "grad_norm": 0.47106119990348816,
      "learning_rate": 0.00010840070548796967,
      "loss": 4.386127471923828,
      "step": 7800
    },
    {
      "epoch": 1.3194796418313905,
      "grad_norm": 0.49278977513313293,
      "learning_rate": 0.00010794070835970782,
      "loss": 4.393439865112304,
      "step": 7810
    },
    {
      "epoch": 1.321169116404798,
      "grad_norm": 0.49596497416496277,
      "learning_rate": 0.00010748114012319747,
      "loss": 4.369705581665039,
      "step": 7820
    },
    {
      "epoch": 1.3228585909782058,
      "grad_norm": 0.48959940671920776,
      "learning_rate": 0.0001070220054647997,
      "loss": 4.353339767456054,
      "step": 7830
    },
    {
      "epoch": 1.3245480655516135,
      "grad_norm": 0.4975447952747345,
      "learning_rate": 0.00010656330906645422,
      "loss": 4.378279113769532,
      "step": 7840
    },
    {
      "epoch": 1.326237540125021,
      "grad_norm": 0.48734408617019653,
      "learning_rate": 0.00010610505560563163,
      "loss": 4.365981674194336,
      "step": 7850
    },
    {
      "epoch": 1.3279270146984288,
      "grad_norm": 0.4985700845718384,
      "learning_rate": 0.00010564724975528584,
      "loss": 4.384627151489258,
      "step": 7860
    },
    {
      "epoch": 1.3296164892718365,
      "grad_norm": 0.48617759346961975,
      "learning_rate": 0.00010518989618380632,
      "loss": 4.387208938598633,
      "step": 7870
    },
    {
      "epoch": 1.331305963845244,
      "grad_norm": 0.479184091091156,
      "learning_rate": 0.00010473299955497044,
      "loss": 4.39497184753418,
      "step": 7880
    },
    {
      "epoch": 1.3329954384186518,
      "grad_norm": 0.5024631023406982,
      "learning_rate": 0.000104276564527896,
      "loss": 4.341180801391602,
      "step": 7890
    },
    {
      "epoch": 1.3346849129920595,
      "grad_norm": 0.5147078633308411,
      "learning_rate": 0.0001038205957569938,
      "loss": 4.36151008605957,
      "step": 7900
    },
    {
      "epoch": 1.336374387565467,
      "grad_norm": 0.4864480197429657,
      "learning_rate": 0.00010336509789191994,
      "loss": 4.3700817108154295,
      "step": 7910
    },
    {
      "epoch": 1.3380638621388747,
      "grad_norm": 0.48009052872657776,
      "learning_rate": 0.00010291007557752861,
      "loss": 4.372967910766602,
      "step": 7920
    },
    {
      "epoch": 1.3397533367122825,
      "grad_norm": 0.4770645499229431,
      "learning_rate": 0.00010245553345382467,
      "loss": 4.361065673828125,
      "step": 7930
    },
    {
      "epoch": 1.3414428112856902,
      "grad_norm": 0.47222378849983215,
      "learning_rate": 0.00010200147615591643,
      "loss": 4.3356986999511715,
      "step": 7940
    },
    {
      "epoch": 1.343132285859098,
      "grad_norm": 0.513080894947052,
      "learning_rate": 0.00010154790831396805,
      "loss": 4.402030181884766,
      "step": 7950
    },
    {
      "epoch": 1.3448217604325055,
      "grad_norm": 0.48416030406951904,
      "learning_rate": 0.00010109483455315269,
      "loss": 4.381985855102539,
      "step": 7960
    },
    {
      "epoch": 1.3465112350059132,
      "grad_norm": 0.46342408657073975,
      "learning_rate": 0.00010064225949360525,
      "loss": 4.364437103271484,
      "step": 7970
    },
    {
      "epoch": 1.348200709579321,
      "grad_norm": 0.4690420925617218,
      "learning_rate": 0.00010019018775037509,
      "loss": 4.399689102172852,
      "step": 7980
    },
    {
      "epoch": 1.3498901841527284,
      "grad_norm": 0.47876372933387756,
      "learning_rate": 9.973862393337925e-05,
      "loss": 4.388835144042969,
      "step": 7990
    },
    {
      "epoch": 1.3515796587261362,
      "grad_norm": 0.48350629210472107,
      "learning_rate": 9.928757264735506e-05,
      "loss": 4.405188751220703,
      "step": 8000
    },
    {
      "epoch": 1.3515796587261362,
      "eval_loss": 4.357097148895264,
      "eval_runtime": 3.734,
      "eval_samples_per_second": 267.812,
      "eval_steps_per_second": 5.624,
      "step": 8000
    },
    {
      "epoch": 1.353269133299544,
      "grad_norm": 0.463106632232666,
      "learning_rate": 9.883703849181374e-05,
      "loss": 4.368831634521484,
      "step": 8010
    },
    {
      "epoch": 1.3549586078729514,
      "grad_norm": 0.4774092137813568,
      "learning_rate": 9.838702606099289e-05,
      "loss": 4.350126647949219,
      "step": 8020
    },
    {
      "epoch": 1.3566480824463591,
      "grad_norm": 0.5083175897598267,
      "learning_rate": 9.793753994381003e-05,
      "loss": 4.375761032104492,
      "step": 8030
    },
    {
      "epoch": 1.3583375570197669,
      "grad_norm": 0.493473619222641,
      "learning_rate": 9.748858472381567e-05,
      "loss": 4.382857894897461,
      "step": 8040
    },
    {
      "epoch": 1.3600270315931744,
      "grad_norm": 0.47200217843055725,
      "learning_rate": 9.704016497914657e-05,
      "loss": 4.363901901245117,
      "step": 8050
    },
    {
      "epoch": 1.3617165061665821,
      "grad_norm": 0.47423017024993896,
      "learning_rate": 9.659228528247923e-05,
      "loss": 4.352508544921875,
      "step": 8060
    },
    {
      "epoch": 1.3634059807399899,
      "grad_norm": 0.49032631516456604,
      "learning_rate": 9.614495020098284e-05,
      "loss": 4.386605834960937,
      "step": 8070
    },
    {
      "epoch": 1.3650954553133976,
      "grad_norm": 0.5129415392875671,
      "learning_rate": 9.569816429627329e-05,
      "loss": 4.370170211791992,
      "step": 8080
    },
    {
      "epoch": 1.3667849298868053,
      "grad_norm": 0.47328782081604004,
      "learning_rate": 9.525193212436607e-05,
      "loss": 4.394309616088867,
      "step": 8090
    },
    {
      "epoch": 1.3684744044602128,
      "grad_norm": 0.5091307759284973,
      "learning_rate": 9.480625823563032e-05,
      "loss": 4.353821182250977,
      "step": 8100
    },
    {
      "epoch": 1.3701638790336206,
      "grad_norm": 0.49530673027038574,
      "learning_rate": 9.436114717474197e-05,
      "loss": 4.374178314208985,
      "step": 8110
    },
    {
      "epoch": 1.3718533536070283,
      "grad_norm": 0.5062808394432068,
      "learning_rate": 9.391660348063778e-05,
      "loss": 4.366446685791016,
      "step": 8120
    },
    {
      "epoch": 1.3735428281804358,
      "grad_norm": 0.4893403947353363,
      "learning_rate": 9.347263168646881e-05,
      "loss": 4.377128601074219,
      "step": 8130
    },
    {
      "epoch": 1.3752323027538436,
      "grad_norm": 0.49352315068244934,
      "learning_rate": 9.30292363195543e-05,
      "loss": 4.390756988525391,
      "step": 8140
    },
    {
      "epoch": 1.3769217773272513,
      "grad_norm": 0.4956866502761841,
      "learning_rate": 9.258642190133548e-05,
      "loss": 4.364201354980469,
      "step": 8150
    },
    {
      "epoch": 1.3786112519006588,
      "grad_norm": 0.4806705415248871,
      "learning_rate": 9.21441929473295e-05,
      "loss": 4.336410140991211,
      "step": 8160
    },
    {
      "epoch": 1.3803007264740665,
      "grad_norm": 0.503070056438446,
      "learning_rate": 9.170255396708336e-05,
      "loss": 4.363087463378906,
      "step": 8170
    },
    {
      "epoch": 1.3819902010474743,
      "grad_norm": 0.4839601218700409,
      "learning_rate": 9.126150946412775e-05,
      "loss": 4.353903961181641,
      "step": 8180
    },
    {
      "epoch": 1.3836796756208818,
      "grad_norm": 0.4867366552352905,
      "learning_rate": 9.082106393593153e-05,
      "loss": 4.347708892822266,
      "step": 8190
    },
    {
      "epoch": 1.3853691501942895,
      "grad_norm": 0.4875339865684509,
      "learning_rate": 9.038122187385543e-05,
      "loss": 4.371865844726562,
      "step": 8200
    },
    {
      "epoch": 1.3870586247676973,
      "grad_norm": 0.49478384852409363,
      "learning_rate": 8.994198776310652e-05,
      "loss": 4.368743133544922,
      "step": 8210
    },
    {
      "epoch": 1.388748099341105,
      "grad_norm": 0.4815446734428406,
      "learning_rate": 8.950336608269243e-05,
      "loss": 4.383320999145508,
      "step": 8220
    },
    {
      "epoch": 1.3904375739145125,
      "grad_norm": 0.4883415997028351,
      "learning_rate": 8.906536130537566e-05,
      "loss": 4.368521881103516,
      "step": 8230
    },
    {
      "epoch": 1.3921270484879202,
      "grad_norm": 0.5107654929161072,
      "learning_rate": 8.862797789762785e-05,
      "loss": 4.353972244262695,
      "step": 8240
    },
    {
      "epoch": 1.393816523061328,
      "grad_norm": 0.46853381395339966,
      "learning_rate": 8.819122031958446e-05,
      "loss": 4.374198150634766,
      "step": 8250
    },
    {
      "epoch": 1.3955059976347357,
      "grad_norm": 0.49264970421791077,
      "learning_rate": 8.77550930249991e-05,
      "loss": 4.353750228881836,
      "step": 8260
    },
    {
      "epoch": 1.3971954722081432,
      "grad_norm": 0.49197956919670105,
      "learning_rate": 8.731960046119819e-05,
      "loss": 4.378075408935547,
      "step": 8270
    },
    {
      "epoch": 1.398884946781551,
      "grad_norm": 0.48225274682044983,
      "learning_rate": 8.688474706903554e-05,
      "loss": 4.360022735595703,
      "step": 8280
    },
    {
      "epoch": 1.4005744213549587,
      "grad_norm": 0.4796869456768036,
      "learning_rate": 8.645053728284734e-05,
      "loss": 4.351276779174805,
      "step": 8290
    },
    {
      "epoch": 1.4022638959283662,
      "grad_norm": 0.46706125140190125,
      "learning_rate": 8.601697553040645e-05,
      "loss": 4.367401885986328,
      "step": 8300
    },
    {
      "epoch": 1.403953370501774,
      "grad_norm": 0.4695565104484558,
      "learning_rate": 8.55840662328778e-05,
      "loss": 4.338150405883789,
      "step": 8310
    },
    {
      "epoch": 1.4056428450751817,
      "grad_norm": 0.4987981915473938,
      "learning_rate": 8.515181380477273e-05,
      "loss": 4.369682693481446,
      "step": 8320
    },
    {
      "epoch": 1.4073323196485892,
      "grad_norm": 0.4853006899356842,
      "learning_rate": 8.47202226539046e-05,
      "loss": 4.392825698852539,
      "step": 8330
    },
    {
      "epoch": 1.409021794221997,
      "grad_norm": 0.48891976475715637,
      "learning_rate": 8.428929718134331e-05,
      "loss": 4.3820442199707035,
      "step": 8340
    },
    {
      "epoch": 1.4107112687954046,
      "grad_norm": 0.48374229669570923,
      "learning_rate": 8.385904178137061e-05,
      "loss": 4.367736053466797,
      "step": 8350
    },
    {
      "epoch": 1.4124007433688122,
      "grad_norm": 0.4966294765472412,
      "learning_rate": 8.342946084143546e-05,
      "loss": 4.336813354492188,
      "step": 8360
    },
    {
      "epoch": 1.41409021794222,
      "grad_norm": 0.4939606487751007,
      "learning_rate": 8.300055874210903e-05,
      "loss": 4.390798568725586,
      "step": 8370
    },
    {
      "epoch": 1.4157796925156276,
      "grad_norm": 0.48403191566467285,
      "learning_rate": 8.257233985704021e-05,
      "loss": 4.3521678924560545,
      "step": 8380
    },
    {
      "epoch": 1.4174691670890354,
      "grad_norm": 0.4766407012939453,
      "learning_rate": 8.214480855291084e-05,
      "loss": 4.337980651855469,
      "step": 8390
    },
    {
      "epoch": 1.419158641662443,
      "grad_norm": 0.469018816947937,
      "learning_rate": 8.171796918939142e-05,
      "loss": 4.341955184936523,
      "step": 8400
    },
    {
      "epoch": 1.4208481162358506,
      "grad_norm": 0.4855271875858307,
      "learning_rate": 8.129182611909642e-05,
      "loss": 4.353343963623047,
      "step": 8410
    },
    {
      "epoch": 1.4225375908092583,
      "grad_norm": 0.4870193898677826,
      "learning_rate": 8.086638368753993e-05,
      "loss": 4.374142074584961,
      "step": 8420
    },
    {
      "epoch": 1.424227065382666,
      "grad_norm": 0.4896891415119171,
      "learning_rate": 8.04416462330916e-05,
      "loss": 4.367203140258789,
      "step": 8430
    },
    {
      "epoch": 1.4259165399560736,
      "grad_norm": 0.46844348311424255,
      "learning_rate": 8.0017618086932e-05,
      "loss": 4.35595817565918,
      "step": 8440
    },
    {
      "epoch": 1.4276060145294813,
      "grad_norm": 0.4512944519519806,
      "learning_rate": 7.959430357300885e-05,
      "loss": 4.3400733947753904,
      "step": 8450
    },
    {
      "epoch": 1.429295489102889,
      "grad_norm": 0.4732443392276764,
      "learning_rate": 7.917170700799256e-05,
      "loss": 4.333652114868164,
      "step": 8460
    },
    {
      "epoch": 1.4309849636762966,
      "grad_norm": 0.4684848487377167,
      "learning_rate": 7.874983270123254e-05,
      "loss": 4.352918243408203,
      "step": 8470
    },
    {
      "epoch": 1.4326744382497043,
      "grad_norm": 0.506878137588501,
      "learning_rate": 7.832868495471306e-05,
      "loss": 4.357436752319336,
      "step": 8480
    },
    {
      "epoch": 1.434363912823112,
      "grad_norm": 0.5020336508750916,
      "learning_rate": 7.790826806300928e-05,
      "loss": 4.359925079345703,
      "step": 8490
    },
    {
      "epoch": 1.4360533873965196,
      "grad_norm": 0.4732269048690796,
      "learning_rate": 7.748858631324393e-05,
      "loss": 4.356634902954101,
      "step": 8500
    },
    {
      "epoch": 1.4360533873965196,
      "eval_loss": 4.3328938484191895,
      "eval_runtime": 3.6888,
      "eval_samples_per_second": 271.089,
      "eval_steps_per_second": 5.693,
      "step": 8500
    },
    {
      "epoch": 1.4377428619699273,
      "grad_norm": 0.4925293028354645,
      "learning_rate": 7.706964398504293e-05,
      "loss": 4.376210403442383,
      "step": 8510
    },
    {
      "epoch": 1.439432336543335,
      "grad_norm": 0.4719123840332031,
      "learning_rate": 7.665144535049224e-05,
      "loss": 4.338931274414063,
      "step": 8520
    },
    {
      "epoch": 1.4411218111167428,
      "grad_norm": 0.4722173511981964,
      "learning_rate": 7.623399467409416e-05,
      "loss": 4.352537536621094,
      "step": 8530
    },
    {
      "epoch": 1.4428112856901505,
      "grad_norm": 0.4844585955142975,
      "learning_rate": 7.581729621272386e-05,
      "loss": 4.332356262207031,
      "step": 8540
    },
    {
      "epoch": 1.444500760263558,
      "grad_norm": 0.49630841612815857,
      "learning_rate": 7.540135421558585e-05,
      "loss": 4.3133392333984375,
      "step": 8550
    },
    {
      "epoch": 1.4461902348369657,
      "grad_norm": 0.472133994102478,
      "learning_rate": 7.498617292417074e-05,
      "loss": 4.3697349548339846,
      "step": 8560
    },
    {
      "epoch": 1.4478797094103735,
      "grad_norm": 0.48327624797821045,
      "learning_rate": 7.457175657221194e-05,
      "loss": 4.366666030883789,
      "step": 8570
    },
    {
      "epoch": 1.449569183983781,
      "grad_norm": 0.4768034815788269,
      "learning_rate": 7.415810938564277e-05,
      "loss": 4.33704719543457,
      "step": 8580
    },
    {
      "epoch": 1.4512586585571887,
      "grad_norm": 0.4592680037021637,
      "learning_rate": 7.37452355825528e-05,
      "loss": 4.343940734863281,
      "step": 8590
    },
    {
      "epoch": 1.4529481331305965,
      "grad_norm": 0.4643280804157257,
      "learning_rate": 7.333313937314548e-05,
      "loss": 4.346873474121094,
      "step": 8600
    },
    {
      "epoch": 1.454637607704004,
      "grad_norm": 0.4980602264404297,
      "learning_rate": 7.292182495969462e-05,
      "loss": 4.370085525512695,
      "step": 8610
    },
    {
      "epoch": 1.4563270822774117,
      "grad_norm": 0.4845782518386841,
      "learning_rate": 7.251129653650206e-05,
      "loss": 4.3420463562011715,
      "step": 8620
    },
    {
      "epoch": 1.4580165568508194,
      "grad_norm": 0.47701558470726013,
      "learning_rate": 7.210155828985447e-05,
      "loss": 4.333865356445313,
      "step": 8630
    },
    {
      "epoch": 1.459706031424227,
      "grad_norm": 0.4681967794895172,
      "learning_rate": 7.169261439798083e-05,
      "loss": 4.315822982788086,
      "step": 8640
    },
    {
      "epoch": 1.4613955059976347,
      "grad_norm": 0.48438313603401184,
      "learning_rate": 7.128446903101004e-05,
      "loss": 4.31340446472168,
      "step": 8650
    },
    {
      "epoch": 1.4630849805710424,
      "grad_norm": 0.4675985872745514,
      "learning_rate": 7.087712635092802e-05,
      "loss": 4.347599792480469,
      "step": 8660
    },
    {
      "epoch": 1.4647744551444501,
      "grad_norm": 0.5026019215583801,
      "learning_rate": 7.047059051153538e-05,
      "loss": 4.3385356903076175,
      "step": 8670
    },
    {
      "epoch": 1.4664639297178579,
      "grad_norm": 0.4908424913883209,
      "learning_rate": 7.006486565840532e-05,
      "loss": 4.337771224975586,
      "step": 8680
    },
    {
      "epoch": 1.4681534042912654,
      "grad_norm": 0.47692814469337463,
      "learning_rate": 6.96599559288411e-05,
      "loss": 4.350002288818359,
      "step": 8690
    },
    {
      "epoch": 1.4698428788646731,
      "grad_norm": 0.4985916316509247,
      "learning_rate": 6.925586545183383e-05,
      "loss": 4.357270812988281,
      "step": 8700
    },
    {
      "epoch": 1.4715323534380809,
      "grad_norm": 0.4779921770095825,
      "learning_rate": 6.885259834802042e-05,
      "loss": 4.3343353271484375,
      "step": 8710
    },
    {
      "epoch": 1.4732218280114884,
      "grad_norm": 0.4964430630207062,
      "learning_rate": 6.845015872964179e-05,
      "loss": 4.345649337768554,
      "step": 8720
    },
    {
      "epoch": 1.4749113025848961,
      "grad_norm": 0.4816732108592987,
      "learning_rate": 6.80485507005005e-05,
      "loss": 4.349812316894531,
      "step": 8730
    },
    {
      "epoch": 1.4766007771583038,
      "grad_norm": 0.4839925765991211,
      "learning_rate": 6.764777835591921e-05,
      "loss": 4.342644119262696,
      "step": 8740
    },
    {
      "epoch": 1.4782902517317114,
      "grad_norm": 0.5161303877830505,
      "learning_rate": 6.724784578269892e-05,
      "loss": 4.322945022583008,
      "step": 8750
    },
    {
      "epoch": 1.479979726305119,
      "grad_norm": 0.4845769703388214,
      "learning_rate": 6.684875705907722e-05,
      "loss": 4.33643798828125,
      "step": 8760
    },
    {
      "epoch": 1.4816692008785268,
      "grad_norm": 0.48371464014053345,
      "learning_rate": 6.645051625468657e-05,
      "loss": 4.319810104370117,
      "step": 8770
    },
    {
      "epoch": 1.4833586754519343,
      "grad_norm": 0.4810192286968231,
      "learning_rate": 6.605312743051297e-05,
      "loss": 4.350659561157227,
      "step": 8780
    },
    {
      "epoch": 1.485048150025342,
      "grad_norm": 0.4886019825935364,
      "learning_rate": 6.565659463885467e-05,
      "loss": 4.340823364257813,
      "step": 8790
    },
    {
      "epoch": 1.4867376245987498,
      "grad_norm": 0.4922144114971161,
      "learning_rate": 6.526092192328048e-05,
      "loss": 4.337167358398437,
      "step": 8800
    },
    {
      "epoch": 1.4884270991721575,
      "grad_norm": 0.47720760107040405,
      "learning_rate": 6.486611331858879e-05,
      "loss": 4.330669403076172,
      "step": 8810
    },
    {
      "epoch": 1.490116573745565,
      "grad_norm": 0.45629069209098816,
      "learning_rate": 6.447217285076651e-05,
      "loss": 4.354007339477539,
      "step": 8820
    },
    {
      "epoch": 1.4918060483189728,
      "grad_norm": 0.4794461727142334,
      "learning_rate": 6.407910453694782e-05,
      "loss": 4.356667327880859,
      "step": 8830
    },
    {
      "epoch": 1.4934955228923805,
      "grad_norm": 0.4836932420730591,
      "learning_rate": 6.368691238537321e-05,
      "loss": 4.3167163848876955,
      "step": 8840
    },
    {
      "epoch": 1.4951849974657883,
      "grad_norm": 0.5060141086578369,
      "learning_rate": 6.329560039534874e-05,
      "loss": 4.362548828125,
      "step": 8850
    },
    {
      "epoch": 1.4968744720391958,
      "grad_norm": 0.48216700553894043,
      "learning_rate": 6.290517255720505e-05,
      "loss": 4.3512012481689455,
      "step": 8860
    },
    {
      "epoch": 1.4985639466126035,
      "grad_norm": 0.46019911766052246,
      "learning_rate": 6.251563285225707e-05,
      "loss": 4.32593002319336,
      "step": 8870
    },
    {
      "epoch": 1.5002534211860112,
      "grad_norm": 0.4773600697517395,
      "learning_rate": 6.212698525276294e-05,
      "loss": 4.345823287963867,
      "step": 8880
    },
    {
      "epoch": 1.5019428957594188,
      "grad_norm": 0.4903421401977539,
      "learning_rate": 6.173923372188372e-05,
      "loss": 4.330167770385742,
      "step": 8890
    },
    {
      "epoch": 1.5036323703328265,
      "grad_norm": 0.47027841210365295,
      "learning_rate": 6.135238221364313e-05,
      "loss": 4.352994155883789,
      "step": 8900
    },
    {
      "epoch": 1.5053218449062342,
      "grad_norm": 0.4893588125705719,
      "learning_rate": 6.096643467288703e-05,
      "loss": 4.3315269470214846,
      "step": 8910
    },
    {
      "epoch": 1.5070113194796417,
      "grad_norm": 0.4835808277130127,
      "learning_rate": 6.058139503524314e-05,
      "loss": 4.349056625366211,
      "step": 8920
    },
    {
      "epoch": 1.5087007940530495,
      "grad_norm": 0.4750809967517853,
      "learning_rate": 6.019726722708104e-05,
      "loss": 4.325545120239258,
      "step": 8930
    },
    {
      "epoch": 1.5103902686264572,
      "grad_norm": 0.4945700466632843,
      "learning_rate": 5.981405516547222e-05,
      "loss": 4.312815093994141,
      "step": 8940
    },
    {
      "epoch": 1.5120797431998647,
      "grad_norm": 0.4704221487045288,
      "learning_rate": 5.9431762758149875e-05,
      "loss": 4.328189849853516,
      "step": 8950
    },
    {
      "epoch": 1.5137692177732727,
      "grad_norm": 0.48752453923225403,
      "learning_rate": 5.9050393903469215e-05,
      "loss": 4.324124145507812,
      "step": 8960
    },
    {
      "epoch": 1.5154586923466802,
      "grad_norm": 0.5149093270301819,
      "learning_rate": 5.866995249036775e-05,
      "loss": 4.334346771240234,
      "step": 8970
    },
    {
      "epoch": 1.5171481669200877,
      "grad_norm": 0.49064958095550537,
      "learning_rate": 5.829044239832564e-05,
      "loss": 4.324323654174805,
      "step": 8980
    },
    {
      "epoch": 1.5188376414934956,
      "grad_norm": 0.486092746257782,
      "learning_rate": 5.791186749732594e-05,
      "loss": 4.346895599365235,
      "step": 8990
    },
    {
      "epoch": 1.5205271160669032,
      "grad_norm": 0.48512768745422363,
      "learning_rate": 5.7534231647815244e-05,
      "loss": 4.350548934936524,
      "step": 9000
    },
    {
      "epoch": 1.5205271160669032,
      "eval_loss": 4.312350273132324,
      "eval_runtime": 4.1596,
      "eval_samples_per_second": 240.409,
      "eval_steps_per_second": 5.049,
      "step": 9000
    },
    {
      "epoch": 1.522216590640311,
      "grad_norm": 0.48225322365760803,
      "learning_rate": 5.715753870066455e-05,
      "loss": 4.3221698760986325,
      "step": 9010
    },
    {
      "epoch": 1.5239060652137186,
      "grad_norm": 0.5018514394760132,
      "learning_rate": 5.67817924971296e-05,
      "loss": 4.321614456176758,
      "step": 9020
    },
    {
      "epoch": 1.5255955397871261,
      "grad_norm": 0.5176340341567993,
      "learning_rate": 5.6406996868811885e-05,
      "loss": 4.335358810424805,
      "step": 9030
    },
    {
      "epoch": 1.5272850143605339,
      "grad_norm": 0.48670732975006104,
      "learning_rate": 5.60331556376197e-05,
      "loss": 4.33309326171875,
      "step": 9040
    },
    {
      "epoch": 1.5289744889339416,
      "grad_norm": 0.49554112553596497,
      "learning_rate": 5.566027261572907e-05,
      "loss": 4.316144943237305,
      "step": 9050
    },
    {
      "epoch": 1.5306639635073491,
      "grad_norm": 0.5128636956214905,
      "learning_rate": 5.528835160554475e-05,
      "loss": 4.335286712646484,
      "step": 9060
    },
    {
      "epoch": 1.5323534380807569,
      "grad_norm": 0.4977918863296509,
      "learning_rate": 5.491739639966153e-05,
      "loss": 4.342447662353516,
      "step": 9070
    },
    {
      "epoch": 1.5340429126541646,
      "grad_norm": 0.5133760571479797,
      "learning_rate": 5.454741078082578e-05,
      "loss": 4.342383956909179,
      "step": 9080
    },
    {
      "epoch": 1.535732387227572,
      "grad_norm": 0.4748549461364746,
      "learning_rate": 5.417839852189653e-05,
      "loss": 4.369438171386719,
      "step": 9090
    },
    {
      "epoch": 1.53742186180098,
      "grad_norm": 0.4682454466819763,
      "learning_rate": 5.381036338580718e-05,
      "loss": 4.323982238769531,
      "step": 9100
    },
    {
      "epoch": 1.5391113363743876,
      "grad_norm": 0.48015832901000977,
      "learning_rate": 5.344330912552703e-05,
      "loss": 4.3247119903564455,
      "step": 9110
    },
    {
      "epoch": 1.540800810947795,
      "grad_norm": 0.4660989046096802,
      "learning_rate": 5.3077239484023385e-05,
      "loss": 4.338339614868164,
      "step": 9120
    },
    {
      "epoch": 1.542490285521203,
      "grad_norm": 0.48378968238830566,
      "learning_rate": 5.271215819422277e-05,
      "loss": 4.342069244384765,
      "step": 9130
    },
    {
      "epoch": 1.5441797600946106,
      "grad_norm": 0.4721354842185974,
      "learning_rate": 5.234806897897328e-05,
      "loss": 4.35260009765625,
      "step": 9140
    },
    {
      "epoch": 1.5458692346680183,
      "grad_norm": 0.47296905517578125,
      "learning_rate": 5.1984975551006434e-05,
      "loss": 4.3343055725097654,
      "step": 9150
    },
    {
      "epoch": 1.547558709241426,
      "grad_norm": 0.49029457569122314,
      "learning_rate": 5.1622881612899635e-05,
      "loss": 4.331478881835937,
      "step": 9160
    },
    {
      "epoch": 1.5492481838148335,
      "grad_norm": 0.4729316234588623,
      "learning_rate": 5.126179085703794e-05,
      "loss": 4.309265899658203,
      "step": 9170
    },
    {
      "epoch": 1.5509376583882413,
      "grad_norm": 0.4636003375053406,
      "learning_rate": 5.090170696557667e-05,
      "loss": 4.332284164428711,
      "step": 9180
    },
    {
      "epoch": 1.552627132961649,
      "grad_norm": 0.4683416783809662,
      "learning_rate": 5.054263361040395e-05,
      "loss": 4.323814392089844,
      "step": 9190
    },
    {
      "epoch": 1.5543166075350565,
      "grad_norm": 0.48071300983428955,
      "learning_rate": 5.018457445310313e-05,
      "loss": 4.331411743164063,
      "step": 9200
    },
    {
      "epoch": 1.5560060821084643,
      "grad_norm": 0.48741987347602844,
      "learning_rate": 4.9827533144915384e-05,
      "loss": 4.315482711791992,
      "step": 9210
    },
    {
      "epoch": 1.557695556681872,
      "grad_norm": 0.47064927220344543,
      "learning_rate": 4.9471513326702544e-05,
      "loss": 4.333251571655273,
      "step": 9220
    },
    {
      "epoch": 1.5593850312552795,
      "grad_norm": 0.48281940817832947,
      "learning_rate": 4.911651862891014e-05,
      "loss": 4.332812118530273,
      "step": 9230
    },
    {
      "epoch": 1.5610745058286872,
      "grad_norm": 0.4713364541530609,
      "learning_rate": 4.876255267153011e-05,
      "loss": 4.334049224853516,
      "step": 9240
    },
    {
      "epoch": 1.562763980402095,
      "grad_norm": 0.47604429721832275,
      "learning_rate": 4.8409619064063965e-05,
      "loss": 4.322870254516602,
      "step": 9250
    },
    {
      "epoch": 1.5644534549755025,
      "grad_norm": 0.4858945608139038,
      "learning_rate": 4.805772140548613e-05,
      "loss": 4.333293914794922,
      "step": 9260
    },
    {
      "epoch": 1.5661429295489104,
      "grad_norm": 0.4747396409511566,
      "learning_rate": 4.770686328420713e-05,
      "loss": 4.309678649902343,
      "step": 9270
    },
    {
      "epoch": 1.567832404122318,
      "grad_norm": 0.46066993474960327,
      "learning_rate": 4.7357048278036944e-05,
      "loss": 4.335137176513672,
      "step": 9280
    },
    {
      "epoch": 1.5695218786957257,
      "grad_norm": 0.48828113079071045,
      "learning_rate": 4.700827995414853e-05,
      "loss": 4.319439315795899,
      "step": 9290
    },
    {
      "epoch": 1.5712113532691334,
      "grad_norm": 0.48410648107528687,
      "learning_rate": 4.666056186904168e-05,
      "loss": 4.3514057159423825,
      "step": 9300
    },
    {
      "epoch": 1.572900827842541,
      "grad_norm": 0.4797396957874298,
      "learning_rate": 4.63138975685064e-05,
      "loss": 4.323817443847656,
      "step": 9310
    },
    {
      "epoch": 1.5745903024159487,
      "grad_norm": 0.46725404262542725,
      "learning_rate": 4.596829058758694e-05,
      "loss": 4.341088104248047,
      "step": 9320
    },
    {
      "epoch": 1.5762797769893564,
      "grad_norm": 0.4728842079639435,
      "learning_rate": 4.5623744450545846e-05,
      "loss": 4.356753540039063,
      "step": 9330
    },
    {
      "epoch": 1.577969251562764,
      "grad_norm": 0.4818381071090698,
      "learning_rate": 4.528026267082786e-05,
      "loss": 4.344687652587891,
      "step": 9340
    },
    {
      "epoch": 1.5796587261361716,
      "grad_norm": 0.47536230087280273,
      "learning_rate": 4.493784875102409e-05,
      "loss": 4.327443695068359,
      "step": 9350
    },
    {
      "epoch": 1.5813482007095794,
      "grad_norm": 0.5105261206626892,
      "learning_rate": 4.45965061828363e-05,
      "loss": 4.336804962158203,
      "step": 9360
    },
    {
      "epoch": 1.583037675282987,
      "grad_norm": 0.46770450472831726,
      "learning_rate": 4.4256238447041556e-05,
      "loss": 4.3366447448730465,
      "step": 9370
    },
    {
      "epoch": 1.5847271498563946,
      "grad_norm": 0.508904218673706,
      "learning_rate": 4.39170490134563e-05,
      "loss": 4.325738143920899,
      "step": 9380
    },
    {
      "epoch": 1.5864166244298024,
      "grad_norm": 0.46618375182151794,
      "learning_rate": 4.3578941340901274e-05,
      "loss": 4.311971282958984,
      "step": 9390
    },
    {
      "epoch": 1.5881060990032099,
      "grad_norm": 0.4693259596824646,
      "learning_rate": 4.324191887716612e-05,
      "loss": 4.320106124877929,
      "step": 9400
    },
    {
      "epoch": 1.5897955735766178,
      "grad_norm": 0.4754733145236969,
      "learning_rate": 4.290598505897439e-05,
      "loss": 4.334828948974609,
      "step": 9410
    },
    {
      "epoch": 1.5914850481500253,
      "grad_norm": 0.4678189158439636,
      "learning_rate": 4.25711433119483e-05,
      "loss": 4.344146347045898,
      "step": 9420
    },
    {
      "epoch": 1.5931745227234329,
      "grad_norm": 0.48414650559425354,
      "learning_rate": 4.223739705057384e-05,
      "loss": 4.333245849609375,
      "step": 9430
    },
    {
      "epoch": 1.5948639972968408,
      "grad_norm": 0.4806137979030609,
      "learning_rate": 4.1904749678165965e-05,
      "loss": 4.320773315429688,
      "step": 9440
    },
    {
      "epoch": 1.5965534718702483,
      "grad_norm": 0.46673110127449036,
      "learning_rate": 4.157320458683409e-05,
      "loss": 4.282149887084961,
      "step": 9450
    },
    {
      "epoch": 1.598242946443656,
      "grad_norm": 0.48379746079444885,
      "learning_rate": 4.124276515744713e-05,
      "loss": 4.316770935058594,
      "step": 9460
    },
    {
      "epoch": 1.5999324210170638,
      "grad_norm": 0.467012882232666,
      "learning_rate": 4.091343475959928e-05,
      "loss": 4.311006164550781,
      "step": 9470
    },
    {
      "epoch": 1.6016218955904713,
      "grad_norm": 0.45999497175216675,
      "learning_rate": 4.058521675157563e-05,
      "loss": 4.31392822265625,
      "step": 9480
    },
    {
      "epoch": 1.603311370163879,
      "grad_norm": 0.49554598331451416,
      "learning_rate": 4.025811448031792e-05,
      "loss": 4.317913818359375,
      "step": 9490
    },
    {
      "epoch": 1.6050008447372868,
      "grad_norm": 0.4512559771537781,
      "learning_rate": 3.993213128139027e-05,
      "loss": 4.320844650268555,
      "step": 9500
    },
    {
      "epoch": 1.6050008447372868,
      "eval_loss": 4.294473648071289,
      "eval_runtime": 7.7128,
      "eval_samples_per_second": 129.654,
      "eval_steps_per_second": 2.723,
      "step": 9500
    },
    {
      "epoch": 1.6066903193106943,
      "grad_norm": 0.4606820344924927,
      "learning_rate": 3.960727047894527e-05,
      "loss": 4.359199523925781,
      "step": 9510
    },
    {
      "epoch": 1.608379793884102,
      "grad_norm": 0.48804476857185364,
      "learning_rate": 3.928353538569023e-05,
      "loss": 4.32340087890625,
      "step": 9520
    },
    {
      "epoch": 1.6100692684575098,
      "grad_norm": 0.4648666977882385,
      "learning_rate": 3.8960929302853074e-05,
      "loss": 4.31898078918457,
      "step": 9530
    },
    {
      "epoch": 1.6117587430309173,
      "grad_norm": 0.48212724924087524,
      "learning_rate": 3.863945552014892e-05,
      "loss": 4.320017242431641,
      "step": 9540
    },
    {
      "epoch": 1.6134482176043252,
      "grad_norm": 0.46979817748069763,
      "learning_rate": 3.831911731574648e-05,
      "loss": 4.365304946899414,
      "step": 9550
    },
    {
      "epoch": 1.6151376921777327,
      "grad_norm": 0.47188496589660645,
      "learning_rate": 3.799991795623471e-05,
      "loss": 4.329359817504883,
      "step": 9560
    },
    {
      "epoch": 1.6168271667511402,
      "grad_norm": 0.47442197799682617,
      "learning_rate": 3.7681860696589216e-05,
      "loss": 4.333200836181641,
      "step": 9570
    },
    {
      "epoch": 1.6185166413245482,
      "grad_norm": 0.46460849046707153,
      "learning_rate": 3.7364948780139344e-05,
      "loss": 4.2955772399902346,
      "step": 9580
    },
    {
      "epoch": 1.6202061158979557,
      "grad_norm": 0.4687038064002991,
      "learning_rate": 3.70491854385351e-05,
      "loss": 4.287596893310547,
      "step": 9590
    },
    {
      "epoch": 1.6218955904713634,
      "grad_norm": 0.4717998802661896,
      "learning_rate": 3.673457389171401e-05,
      "loss": 4.3026374816894535,
      "step": 9600
    },
    {
      "epoch": 1.6235850650447712,
      "grad_norm": 0.47237226366996765,
      "learning_rate": 3.642111734786833e-05,
      "loss": 4.3385662078857425,
      "step": 9610
    },
    {
      "epoch": 1.6252745396181787,
      "grad_norm": 0.48337623476982117,
      "learning_rate": 3.610881900341261e-05,
      "loss": 4.29266357421875,
      "step": 9620
    },
    {
      "epoch": 1.6269640141915864,
      "grad_norm": 0.46639102697372437,
      "learning_rate": 3.579768204295063e-05,
      "loss": 4.3327476501464846,
      "step": 9630
    },
    {
      "epoch": 1.6286534887649942,
      "grad_norm": 0.4697898030281067,
      "learning_rate": 3.54877096392434e-05,
      "loss": 4.336753463745117,
      "step": 9640
    },
    {
      "epoch": 1.6303429633384017,
      "grad_norm": 0.46316251158714294,
      "learning_rate": 3.5178904953176354e-05,
      "loss": 4.306925964355469,
      "step": 9650
    },
    {
      "epoch": 1.6320324379118094,
      "grad_norm": 0.4708452820777893,
      "learning_rate": 3.487127113372755e-05,
      "loss": 4.326674270629883,
      "step": 9660
    },
    {
      "epoch": 1.6337219124852171,
      "grad_norm": 0.4727766811847687,
      "learning_rate": 3.4564811317935235e-05,
      "loss": 4.304772186279297,
      "step": 9670
    },
    {
      "epoch": 1.6354113870586247,
      "grad_norm": 0.47584787011146545,
      "learning_rate": 3.4259528630865995e-05,
      "loss": 4.3285400390625,
      "step": 9680
    },
    {
      "epoch": 1.6371008616320324,
      "grad_norm": 0.4718579947948456,
      "learning_rate": 3.3955426185582826e-05,
      "loss": 4.310879135131836,
      "step": 9690
    },
    {
      "epoch": 1.6387903362054401,
      "grad_norm": 0.466880738735199,
      "learning_rate": 3.365250708311352e-05,
      "loss": 4.325877380371094,
      "step": 9700
    },
    {
      "epoch": 1.6404798107788476,
      "grad_norm": 0.46377378702163696,
      "learning_rate": 3.335077441241895e-05,
      "loss": 4.307848358154297,
      "step": 9710
    },
    {
      "epoch": 1.6421692853522556,
      "grad_norm": 0.718170166015625,
      "learning_rate": 3.305023125036148e-05,
      "loss": 4.313734436035157,
      "step": 9720
    },
    {
      "epoch": 1.643858759925663,
      "grad_norm": 0.463375985622406,
      "learning_rate": 3.275088066167369e-05,
      "loss": 4.3089752197265625,
      "step": 9730
    },
    {
      "epoch": 1.6455482344990708,
      "grad_norm": 0.47580841183662415,
      "learning_rate": 3.245272569892727e-05,
      "loss": 4.3522186279296875,
      "step": 9740
    },
    {
      "epoch": 1.6472377090724786,
      "grad_norm": 0.46081092953681946,
      "learning_rate": 3.215576940250155e-05,
      "loss": 4.3113548278808596,
      "step": 9750
    },
    {
      "epoch": 1.648927183645886,
      "grad_norm": 0.47329118847846985,
      "learning_rate": 3.1860014800552734e-05,
      "loss": 4.3111930847167965,
      "step": 9760
    },
    {
      "epoch": 1.6506166582192938,
      "grad_norm": 0.4813630282878876,
      "learning_rate": 3.15654649089831e-05,
      "loss": 4.312236404418945,
      "step": 9770
    },
    {
      "epoch": 1.6523061327927016,
      "grad_norm": 0.5134222507476807,
      "learning_rate": 3.1272122731409916e-05,
      "loss": 4.3267356872558596,
      "step": 9780
    },
    {
      "epoch": 1.653995607366109,
      "grad_norm": 0.4687715768814087,
      "learning_rate": 3.097999125913518e-05,
      "loss": 4.311066055297852,
      "step": 9790
    },
    {
      "epoch": 1.6556850819395168,
      "grad_norm": 0.4736403524875641,
      "learning_rate": 3.068907347111485e-05,
      "loss": 4.3107654571533205,
      "step": 9800
    },
    {
      "epoch": 1.6573745565129245,
      "grad_norm": 0.4813496172428131,
      "learning_rate": 3.0399372333928644e-05,
      "loss": 4.314376449584961,
      "step": 9810
    },
    {
      "epoch": 1.659064031086332,
      "grad_norm": 0.49036741256713867,
      "learning_rate": 3.0110890801749627e-05,
      "loss": 4.307826995849609,
      "step": 9820
    },
    {
      "epoch": 1.6607535056597398,
      "grad_norm": 0.4669703543186188,
      "learning_rate": 2.982363181631418e-05,
      "loss": 4.303530883789063,
      "step": 9830
    },
    {
      "epoch": 1.6624429802331475,
      "grad_norm": 0.4788713753223419,
      "learning_rate": 2.9537598306892103e-05,
      "loss": 4.308844375610351,
      "step": 9840
    },
    {
      "epoch": 1.664132454806555,
      "grad_norm": 0.5307414531707764,
      "learning_rate": 2.9252793190256447e-05,
      "loss": 4.285565567016602,
      "step": 9850
    },
    {
      "epoch": 1.665821929379963,
      "grad_norm": 0.4659578502178192,
      "learning_rate": 2.896921937065419e-05,
      "loss": 4.313910675048828,
      "step": 9860
    },
    {
      "epoch": 1.6675114039533705,
      "grad_norm": 0.46300381422042847,
      "learning_rate": 2.8686879739776137e-05,
      "loss": 4.31811408996582,
      "step": 9870
    },
    {
      "epoch": 1.669200878526778,
      "grad_norm": 0.4717971086502075,
      "learning_rate": 2.8405777176727924e-05,
      "loss": 4.318044662475586,
      "step": 9880
    },
    {
      "epoch": 1.670890353100186,
      "grad_norm": 0.45347994565963745,
      "learning_rate": 2.8125914548000243e-05,
      "loss": 4.295824432373047,
      "step": 9890
    },
    {
      "epoch": 1.6725798276735935,
      "grad_norm": 0.4703952670097351,
      "learning_rate": 2.7847294707439828e-05,
      "loss": 4.28874626159668,
      "step": 9900
    },
    {
      "epoch": 1.6742693022470012,
      "grad_norm": 0.4726548194885254,
      "learning_rate": 2.7569920496220398e-05,
      "loss": 4.304931259155273,
      "step": 9910
    },
    {
      "epoch": 1.675958776820409,
      "grad_norm": 0.47394225001335144,
      "learning_rate": 2.729379474281352e-05,
      "loss": 4.3050182342529295,
      "step": 9920
    },
    {
      "epoch": 1.6776482513938165,
      "grad_norm": 0.49833500385284424,
      "learning_rate": 2.701892026295979e-05,
      "loss": 4.331858062744141,
      "step": 9930
    },
    {
      "epoch": 1.6793377259672242,
      "grad_norm": 0.4709710478782654,
      "learning_rate": 2.6745299859640318e-05,
      "loss": 4.332807159423828,
      "step": 9940
    },
    {
      "epoch": 1.681027200540632,
      "grad_norm": 0.48379939794540405,
      "learning_rate": 2.6472936323047972e-05,
      "loss": 4.311476516723633,
      "step": 9950
    },
    {
      "epoch": 1.6827166751140394,
      "grad_norm": 0.475941926240921,
      "learning_rate": 2.6201832430558866e-05,
      "loss": 4.314311599731445,
      "step": 9960
    },
    {
      "epoch": 1.6844061496874472,
      "grad_norm": 0.4633561372756958,
      "learning_rate": 2.5931990946704206e-05,
      "loss": 4.312783050537109,
      "step": 9970
    },
    {
      "epoch": 1.686095624260855,
      "grad_norm": 0.4624374806880951,
      "learning_rate": 2.5663414623141943e-05,
      "loss": 4.315936279296875,
      "step": 9980
    },
    {
      "epoch": 1.6877850988342624,
      "grad_norm": 0.46104687452316284,
      "learning_rate": 2.5396106198628947e-05,
      "loss": 4.317576217651367,
      "step": 9990
    },
    {
      "epoch": 1.6894745734076704,
      "grad_norm": 0.46486878395080566,
      "learning_rate": 2.5130068398992716e-05,
      "loss": 4.3148681640625,
      "step": 10000
    },
    {
      "epoch": 1.6894745734076704,
      "eval_loss": 4.282918930053711,
      "eval_runtime": 3.8826,
      "eval_samples_per_second": 257.563,
      "eval_steps_per_second": 5.409,
      "step": 10000
    },
    {
      "epoch": 1.691164047981078,
      "grad_norm": 0.4754472076892853,
      "learning_rate": 2.4865303937104007e-05,
      "loss": 4.285601425170898,
      "step": 10010
    },
    {
      "epoch": 1.6928535225544854,
      "grad_norm": 0.4589325785636902,
      "learning_rate": 2.460181551284876e-05,
      "loss": 4.316444396972656,
      "step": 10020
    },
    {
      "epoch": 1.6945429971278934,
      "grad_norm": 0.4686416685581207,
      "learning_rate": 2.433960581310091e-05,
      "loss": 4.296805191040039,
      "step": 10030
    },
    {
      "epoch": 1.6962324717013009,
      "grad_norm": 0.44992297887802124,
      "learning_rate": 2.4078677511694776e-05,
      "loss": 4.326528930664063,
      "step": 10040
    },
    {
      "epoch": 1.6979219462747086,
      "grad_norm": 0.460001677274704,
      "learning_rate": 2.381903326939777e-05,
      "loss": 4.270325088500977,
      "step": 10050
    },
    {
      "epoch": 1.6996114208481163,
      "grad_norm": 0.45742112398147583,
      "learning_rate": 2.356067573388355e-05,
      "loss": 4.311310958862305,
      "step": 10060
    },
    {
      "epoch": 1.7013008954215239,
      "grad_norm": 0.5051885843276978,
      "learning_rate": 2.3303607539704628e-05,
      "loss": 4.305488586425781,
      "step": 10070
    },
    {
      "epoch": 1.7029903699949316,
      "grad_norm": 0.460809588432312,
      "learning_rate": 2.3047831308265845e-05,
      "loss": 4.284737777709961,
      "step": 10080
    },
    {
      "epoch": 1.7046798445683393,
      "grad_norm": 0.48899003863334656,
      "learning_rate": 2.2793349647797372e-05,
      "loss": 4.308148956298828,
      "step": 10090
    },
    {
      "epoch": 1.7063693191417468,
      "grad_norm": 0.47210270166397095,
      "learning_rate": 2.2540165153328345e-05,
      "loss": 4.300167465209961,
      "step": 10100
    },
    {
      "epoch": 1.7080587937151546,
      "grad_norm": 0.4561355710029602,
      "learning_rate": 2.2288280406660237e-05,
      "loss": 4.295189285278321,
      "step": 10110
    },
    {
      "epoch": 1.7097482682885623,
      "grad_norm": 0.4685342013835907,
      "learning_rate": 2.2037697976340525e-05,
      "loss": 4.32569465637207,
      "step": 10120
    },
    {
      "epoch": 1.7114377428619698,
      "grad_norm": 0.4782038629055023,
      "learning_rate": 2.1788420417636704e-05,
      "loss": 4.281776046752929,
      "step": 10130
    },
    {
      "epoch": 1.7131272174353775,
      "grad_norm": 0.45496320724487305,
      "learning_rate": 2.1540450272509986e-05,
      "loss": 4.289628219604492,
      "step": 10140
    },
    {
      "epoch": 1.7148166920087853,
      "grad_norm": 0.4686676263809204,
      "learning_rate": 2.129379006958944e-05,
      "loss": 4.304840087890625,
      "step": 10150
    },
    {
      "epoch": 1.7165061665821928,
      "grad_norm": 0.45078393816947937,
      "learning_rate": 2.104844232414634e-05,
      "loss": 4.333132934570313,
      "step": 10160
    },
    {
      "epoch": 1.7181956411556008,
      "grad_norm": 0.4551495313644409,
      "learning_rate": 2.080440953806844e-05,
      "loss": 4.313465118408203,
      "step": 10170
    },
    {
      "epoch": 1.7198851157290083,
      "grad_norm": 0.46409592032432556,
      "learning_rate": 2.056169419983432e-05,
      "loss": 4.303678131103515,
      "step": 10180
    },
    {
      "epoch": 1.721574590302416,
      "grad_norm": 0.46051809191703796,
      "learning_rate": 2.0320298784488177e-05,
      "loss": 4.297393798828125,
      "step": 10190
    },
    {
      "epoch": 1.7232640648758237,
      "grad_norm": 0.541107714176178,
      "learning_rate": 2.008022575361464e-05,
      "loss": 4.302070617675781,
      "step": 10200
    },
    {
      "epoch": 1.7249535394492312,
      "grad_norm": 0.46840059757232666,
      "learning_rate": 1.9841477555313428e-05,
      "loss": 4.290169143676758,
      "step": 10210
    },
    {
      "epoch": 1.726643014022639,
      "grad_norm": 0.46939900517463684,
      "learning_rate": 1.960405662417458e-05,
      "loss": 4.315706634521485,
      "step": 10220
    },
    {
      "epoch": 1.7283324885960467,
      "grad_norm": 0.4771457016468048,
      "learning_rate": 1.9367965381253632e-05,
      "loss": 4.289263534545898,
      "step": 10230
    },
    {
      "epoch": 1.7300219631694542,
      "grad_norm": 0.48085805773735046,
      "learning_rate": 1.9133206234046833e-05,
      "loss": 4.3228507995605465,
      "step": 10240
    },
    {
      "epoch": 1.731711437742862,
      "grad_norm": 0.4604587256908417,
      "learning_rate": 1.8899781576466605e-05,
      "loss": 4.296081924438477,
      "step": 10250
    },
    {
      "epoch": 1.7334009123162697,
      "grad_norm": 0.4865635633468628,
      "learning_rate": 1.86676937888172e-05,
      "loss": 4.302744674682617,
      "step": 10260
    },
    {
      "epoch": 1.7350903868896772,
      "grad_norm": 0.4594942033290863,
      "learning_rate": 1.8436945237770347e-05,
      "loss": 4.3057910919189455,
      "step": 10270
    },
    {
      "epoch": 1.736779861463085,
      "grad_norm": 0.4511856734752655,
      "learning_rate": 1.8207538276341255e-05,
      "loss": 4.311210632324219,
      "step": 10280
    },
    {
      "epoch": 1.7384693360364927,
      "grad_norm": 0.46823564171791077,
      "learning_rate": 1.7979475243864422e-05,
      "loss": 4.291423797607422,
      "step": 10290
    },
    {
      "epoch": 1.7401588106099002,
      "grad_norm": 0.456841379404068,
      "learning_rate": 1.7752758465969835e-05,
      "loss": 4.291481781005859,
      "step": 10300
    },
    {
      "epoch": 1.7418482851833081,
      "grad_norm": 0.464433491230011,
      "learning_rate": 1.7527390254559564e-05,
      "loss": 4.306121826171875,
      "step": 10310
    },
    {
      "epoch": 1.7435377597567157,
      "grad_norm": 0.43991556763648987,
      "learning_rate": 1.7303372907783646e-05,
      "loss": 4.288319778442383,
      "step": 10320
    },
    {
      "epoch": 1.7452272343301232,
      "grad_norm": 0.4612221121788025,
      "learning_rate": 1.708070871001704e-05,
      "loss": 4.296160125732422,
      "step": 10330
    },
    {
      "epoch": 1.7469167089035311,
      "grad_norm": 0.4536151587963104,
      "learning_rate": 1.6859399931836182e-05,
      "loss": 4.302063751220703,
      "step": 10340
    },
    {
      "epoch": 1.7486061834769386,
      "grad_norm": 0.47430509328842163,
      "learning_rate": 1.663944882999596e-05,
      "loss": 4.320109176635742,
      "step": 10350
    },
    {
      "epoch": 1.7502956580503464,
      "grad_norm": 0.4570671021938324,
      "learning_rate": 1.6420857647406533e-05,
      "loss": 4.309846115112305,
      "step": 10360
    },
    {
      "epoch": 1.751985132623754,
      "grad_norm": 0.46541541814804077,
      "learning_rate": 1.6203628613110513e-05,
      "loss": 4.321808242797852,
      "step": 10370
    },
    {
      "epoch": 1.7536746071971616,
      "grad_norm": 0.4689694941043854,
      "learning_rate": 1.598776394226035e-05,
      "loss": 4.342444992065429,
      "step": 10380
    },
    {
      "epoch": 1.7553640817705694,
      "grad_norm": 0.4656012952327728,
      "learning_rate": 1.5773265836095615e-05,
      "loss": 4.284120178222656,
      "step": 10390
    },
    {
      "epoch": 1.757053556343977,
      "grad_norm": 0.44993332028388977,
      "learning_rate": 1.5560136481920583e-05,
      "loss": 4.305658340454102,
      "step": 10400
    },
    {
      "epoch": 1.7587430309173846,
      "grad_norm": 0.45734935998916626,
      "learning_rate": 1.5348378053081885e-05,
      "loss": 4.28479232788086,
      "step": 10410
    },
    {
      "epoch": 1.7604325054907923,
      "grad_norm": 0.48875826597213745,
      "learning_rate": 1.5137992708946522e-05,
      "loss": 4.30067024230957,
      "step": 10420
    },
    {
      "epoch": 1.7621219800642,
      "grad_norm": 0.4599165618419647,
      "learning_rate": 1.4928982594879602e-05,
      "loss": 4.302487564086914,
      "step": 10430
    },
    {
      "epoch": 1.7638114546376076,
      "grad_norm": 0.45845454931259155,
      "learning_rate": 1.4721349842222623e-05,
      "loss": 4.285428619384765,
      "step": 10440
    },
    {
      "epoch": 1.7655009292110155,
      "grad_norm": 0.4705585539340973,
      "learning_rate": 1.4515096568271728e-05,
      "loss": 4.30066032409668,
      "step": 10450
    },
    {
      "epoch": 1.767190403784423,
      "grad_norm": 0.47239530086517334,
      "learning_rate": 1.4310224876256071e-05,
      "loss": 4.3203174591064455,
      "step": 10460
    },
    {
      "epoch": 1.7688798783578306,
      "grad_norm": 0.4652308225631714,
      "learning_rate": 1.410673685531638e-05,
      "loss": 4.307133483886719,
      "step": 10470
    },
    {
      "epoch": 1.7705693529312385,
      "grad_norm": 0.4624398946762085,
      "learning_rate": 1.390463458048357e-05,
      "loss": 4.315113830566406,
      "step": 10480
    },
    {
      "epoch": 1.772258827504646,
      "grad_norm": 0.46324899792671204,
      "learning_rate": 1.3703920112657856e-05,
      "loss": 4.304290771484375,
      "step": 10490
    },
    {
      "epoch": 1.7739483020780538,
      "grad_norm": 0.4670204222202301,
      "learning_rate": 1.3504595498587378e-05,
      "loss": 4.301520919799804,
      "step": 10500
    },
    {
      "epoch": 1.7739483020780538,
      "eval_loss": 4.273873805999756,
      "eval_runtime": 4.0508,
      "eval_samples_per_second": 246.863,
      "eval_steps_per_second": 5.184,
      "step": 10500
    },
    {
      "epoch": 1.7756377766514615,
      "grad_norm": 0.45778968930244446,
      "learning_rate": 1.330666277084756e-05,
      "loss": 4.324074172973633,
      "step": 10510
    },
    {
      "epoch": 1.777327251224869,
      "grad_norm": 0.44530189037323,
      "learning_rate": 1.3110123947820345e-05,
      "loss": 4.296671295166016,
      "step": 10520
    },
    {
      "epoch": 1.7790167257982767,
      "grad_norm": 0.4516686797142029,
      "learning_rate": 1.2914981033673616e-05,
      "loss": 4.3019359588623045,
      "step": 10530
    },
    {
      "epoch": 1.7807062003716845,
      "grad_norm": 0.4497534930706024,
      "learning_rate": 1.2721236018340675e-05,
      "loss": 4.252984237670899,
      "step": 10540
    },
    {
      "epoch": 1.782395674945092,
      "grad_norm": 0.4479978680610657,
      "learning_rate": 1.2528890877500025e-05,
      "loss": 4.305055618286133,
      "step": 10550
    },
    {
      "epoch": 1.7840851495184997,
      "grad_norm": 0.462827205657959,
      "learning_rate": 1.2337947572555257e-05,
      "loss": 4.314754486083984,
      "step": 10560
    },
    {
      "epoch": 1.7857746240919075,
      "grad_norm": 0.4561219811439514,
      "learning_rate": 1.2148408050614961e-05,
      "loss": 4.2755790710449215,
      "step": 10570
    },
    {
      "epoch": 1.787464098665315,
      "grad_norm": 0.4636087119579315,
      "learning_rate": 1.1960274244472928e-05,
      "loss": 4.280724716186524,
      "step": 10580
    },
    {
      "epoch": 1.7891535732387227,
      "grad_norm": 0.4560607373714447,
      "learning_rate": 1.1773548072588352e-05,
      "loss": 4.296182632446289,
      "step": 10590
    },
    {
      "epoch": 1.7908430478121304,
      "grad_norm": 0.46516045928001404,
      "learning_rate": 1.158823143906652e-05,
      "loss": 4.301852416992188,
      "step": 10600
    },
    {
      "epoch": 1.792532522385538,
      "grad_norm": 0.4671533703804016,
      "learning_rate": 1.1404326233639056e-05,
      "loss": 4.321551132202148,
      "step": 10610
    },
    {
      "epoch": 1.794221996958946,
      "grad_norm": 0.46711355447769165,
      "learning_rate": 1.1221834331644857e-05,
      "loss": 4.292984390258789,
      "step": 10620
    },
    {
      "epoch": 1.7959114715323534,
      "grad_norm": 0.46830058097839355,
      "learning_rate": 1.1040757594010908e-05,
      "loss": 4.294471740722656,
      "step": 10630
    },
    {
      "epoch": 1.7976009461057612,
      "grad_norm": 0.45422518253326416,
      "learning_rate": 1.0861097867233375e-05,
      "loss": 4.302399444580078,
      "step": 10640
    },
    {
      "epoch": 1.799290420679169,
      "grad_norm": 0.46243947744369507,
      "learning_rate": 1.0682856983358645e-05,
      "loss": 4.300415420532227,
      "step": 10650
    },
    {
      "epoch": 1.8009798952525764,
      "grad_norm": 0.45327311754226685,
      "learning_rate": 1.050603675996477e-05,
      "loss": 4.294659042358399,
      "step": 10660
    },
    {
      "epoch": 1.8026693698259841,
      "grad_norm": 0.4481427073478699,
      "learning_rate": 1.0330639000142877e-05,
      "loss": 4.29761962890625,
      "step": 10670
    },
    {
      "epoch": 1.8043588443993919,
      "grad_norm": 0.45235884189605713,
      "learning_rate": 1.0156665492478794e-05,
      "loss": 4.2950092315673825,
      "step": 10680
    },
    {
      "epoch": 1.8060483189727994,
      "grad_norm": 0.4574648439884186,
      "learning_rate": 9.984118011034787e-06,
      "loss": 4.296451187133789,
      "step": 10690
    },
    {
      "epoch": 1.8077377935462071,
      "grad_norm": 0.44989126920700073,
      "learning_rate": 9.812998315331449e-06,
      "loss": 4.295338821411133,
      "step": 10700
    },
    {
      "epoch": 1.8094272681196149,
      "grad_norm": 0.44825267791748047,
      "learning_rate": 9.64330815032991e-06,
      "loss": 4.29632682800293,
      "step": 10710
    },
    {
      "epoch": 1.8111167426930224,
      "grad_norm": 0.44391629099845886,
      "learning_rate": 9.475049246413801e-06,
      "loss": 4.282930374145508,
      "step": 10720
    },
    {
      "epoch": 1.81280621726643,
      "grad_norm": 0.4501837491989136,
      "learning_rate": 9.308223319371789e-06,
      "loss": 4.3113666534423825,
      "step": 10730
    },
    {
      "epoch": 1.8144956918398378,
      "grad_norm": 0.45159661769866943,
      "learning_rate": 9.142832070380051e-06,
      "loss": 4.275300979614258,
      "step": 10740
    },
    {
      "epoch": 1.8161851664132453,
      "grad_norm": 0.4615607261657715,
      "learning_rate": 8.978877185984895e-06,
      "loss": 4.27879753112793,
      "step": 10750
    },
    {
      "epoch": 1.8178746409866533,
      "grad_norm": 0.4554959535598755,
      "learning_rate": 8.816360338085537e-06,
      "loss": 4.320524597167969,
      "step": 10760
    },
    {
      "epoch": 1.8195641155600608,
      "grad_norm": 0.4588150978088379,
      "learning_rate": 8.655283183917094e-06,
      "loss": 4.305972671508789,
      "step": 10770
    },
    {
      "epoch": 1.8212535901334683,
      "grad_norm": 0.4642908275127411,
      "learning_rate": 8.495647366033708e-06,
      "loss": 4.303414154052734,
      "step": 10780
    },
    {
      "epoch": 1.8229430647068763,
      "grad_norm": 0.44532260298728943,
      "learning_rate": 8.33745451229173e-06,
      "loss": 4.2890056610107425,
      "step": 10790
    },
    {
      "epoch": 1.8246325392802838,
      "grad_norm": 0.44421857595443726,
      "learning_rate": 8.180706235833162e-06,
      "loss": 4.27965087890625,
      "step": 10800
    },
    {
      "epoch": 1.8263220138536915,
      "grad_norm": 0.4534235894680023,
      "learning_rate": 8.025404135069207e-06,
      "loss": 4.3062583923339846,
      "step": 10810
    },
    {
      "epoch": 1.8280114884270993,
      "grad_norm": 0.44868797063827515,
      "learning_rate": 7.871549793663985e-06,
      "loss": 4.286159896850586,
      "step": 10820
    },
    {
      "epoch": 1.8297009630005068,
      "grad_norm": 0.4559250771999359,
      "learning_rate": 7.719144780518315e-06,
      "loss": 4.280204391479492,
      "step": 10830
    },
    {
      "epoch": 1.8313904375739145,
      "grad_norm": 0.4582137167453766,
      "learning_rate": 7.568190649753753e-06,
      "loss": 4.293819427490234,
      "step": 10840
    },
    {
      "epoch": 1.8330799121473222,
      "grad_norm": 0.44784441590309143,
      "learning_rate": 7.418688940696843e-06,
      "loss": 4.301911163330078,
      "step": 10850
    },
    {
      "epoch": 1.8347693867207298,
      "grad_norm": 0.4547264575958252,
      "learning_rate": 7.270641177863251e-06,
      "loss": 4.318780136108399,
      "step": 10860
    },
    {
      "epoch": 1.8364588612941375,
      "grad_norm": 0.44876977801322937,
      "learning_rate": 7.124048870942301e-06,
      "loss": 4.305691528320312,
      "step": 10870
    },
    {
      "epoch": 1.8381483358675452,
      "grad_norm": 0.4435437321662903,
      "learning_rate": 6.97891351478157e-06,
      "loss": 4.285098648071289,
      "step": 10880
    },
    {
      "epoch": 1.8398378104409527,
      "grad_norm": 0.4529848098754883,
      "learning_rate": 6.83523658937174e-06,
      "loss": 4.30163345336914,
      "step": 10890
    },
    {
      "epoch": 1.8415272850143607,
      "grad_norm": 0.44488754868507385,
      "learning_rate": 6.693019559831319e-06,
      "loss": 4.272104644775391,
      "step": 10900
    },
    {
      "epoch": 1.8432167595877682,
      "grad_norm": 0.44506925344467163,
      "learning_rate": 6.552263876391878e-06,
      "loss": 4.296164703369141,
      "step": 10910
    },
    {
      "epoch": 1.8449062341611757,
      "grad_norm": 0.453124076128006,
      "learning_rate": 6.412970974383069e-06,
      "loss": 4.268503189086914,
      "step": 10920
    },
    {
      "epoch": 1.8465957087345837,
      "grad_norm": 0.45630943775177,
      "learning_rate": 6.275142274218264e-06,
      "loss": 4.276957702636719,
      "step": 10930
    },
    {
      "epoch": 1.8482851833079912,
      "grad_norm": 0.4438062012195587,
      "learning_rate": 6.138779181379777e-06,
      "loss": 4.31237564086914,
      "step": 10940
    },
    {
      "epoch": 1.849974657881399,
      "grad_norm": 0.4586540460586548,
      "learning_rate": 6.003883086404709e-06,
      "loss": 4.296250915527343,
      "step": 10950
    },
    {
      "epoch": 1.8516641324548067,
      "grad_norm": 0.4555058777332306,
      "learning_rate": 5.870455364870747e-06,
      "loss": 4.289797973632813,
      "step": 10960
    },
    {
      "epoch": 1.8533536070282142,
      "grad_norm": 0.4580257833003998,
      "learning_rate": 5.738497377382117e-06,
      "loss": 4.288161849975586,
      "step": 10970
    },
    {
      "epoch": 1.855043081601622,
      "grad_norm": 0.44520384073257446,
      "learning_rate": 5.608010469555674e-06,
      "loss": 4.309579467773437,
      "step": 10980
    },
    {
      "epoch": 1.8567325561750296,
      "grad_norm": 0.4458165168762207,
      "learning_rate": 5.4789959720071995e-06,
      "loss": 4.300251007080078,
      "step": 10990
    },
    {
      "epoch": 1.8584220307484371,
      "grad_norm": 0.4537349343299866,
      "learning_rate": 5.3514552003379395e-06,
      "loss": 4.293206024169922,
      "step": 11000
    },
    {
      "epoch": 1.8584220307484371,
      "eval_loss": 4.268224716186523,
      "eval_runtime": 3.8046,
      "eval_samples_per_second": 262.842,
      "eval_steps_per_second": 5.52,
      "step": 11000
    },
    {
      "epoch": 1.8601115053218449,
      "grad_norm": 0.455091655254364,
      "learning_rate": 5.225389455120976e-06,
      "loss": 4.29366455078125,
      "step": 11010
    },
    {
      "epoch": 1.8618009798952526,
      "grad_norm": 0.46570661664009094,
      "learning_rate": 5.1008000218881576e-06,
      "loss": 4.286912536621093,
      "step": 11020
    },
    {
      "epoch": 1.8634904544686601,
      "grad_norm": 0.4428755044937134,
      "learning_rate": 4.977688171116923e-06,
      "loss": 4.300152206420899,
      "step": 11030
    },
    {
      "epoch": 1.8651799290420679,
      "grad_norm": 0.4450303316116333,
      "learning_rate": 4.856055158217298e-06,
      "loss": 4.289414978027343,
      "step": 11040
    },
    {
      "epoch": 1.8668694036154756,
      "grad_norm": 0.45025452971458435,
      "learning_rate": 4.735902223519173e-06,
      "loss": 4.273600006103516,
      "step": 11050
    },
    {
      "epoch": 1.868558878188883,
      "grad_norm": 0.4395146071910858,
      "learning_rate": 4.6172305922595746e-06,
      "loss": 4.288070297241211,
      "step": 11060
    },
    {
      "epoch": 1.870248352762291,
      "grad_norm": 0.45017367601394653,
      "learning_rate": 4.500041474570265e-06,
      "loss": 4.298558044433594,
      "step": 11070
    },
    {
      "epoch": 1.8719378273356986,
      "grad_norm": 0.45083948969841003,
      "learning_rate": 4.384336065465349e-06,
      "loss": 4.278664398193359,
      "step": 11080
    },
    {
      "epoch": 1.8736273019091063,
      "grad_norm": 0.4492949843406677,
      "learning_rate": 4.270115544829017e-06,
      "loss": 4.304440307617187,
      "step": 11090
    },
    {
      "epoch": 1.875316776482514,
      "grad_norm": 0.4543094336986542,
      "learning_rate": 4.1573810774037044e-06,
      "loss": 4.284811401367188,
      "step": 11100
    },
    {
      "epoch": 1.8770062510559216,
      "grad_norm": 0.44232332706451416,
      "learning_rate": 4.046133812777985e-06,
      "loss": 4.275522232055664,
      "step": 11110
    },
    {
      "epoch": 1.8786957256293293,
      "grad_norm": 0.4532018005847931,
      "learning_rate": 3.936374885375049e-06,
      "loss": 4.325132751464844,
      "step": 11120
    },
    {
      "epoch": 1.880385200202737,
      "grad_norm": 0.44870230555534363,
      "learning_rate": 3.828105414440974e-06,
      "loss": 4.293384170532226,
      "step": 11130
    },
    {
      "epoch": 1.8820746747761445,
      "grad_norm": 0.4408150017261505,
      "learning_rate": 3.7213265040334394e-06,
      "loss": 4.296081161499023,
      "step": 11140
    },
    {
      "epoch": 1.8837641493495523,
      "grad_norm": 0.44336998462677,
      "learning_rate": 3.616039243010399e-06,
      "loss": 4.299095916748047,
      "step": 11150
    },
    {
      "epoch": 1.88545362392296,
      "grad_norm": 0.4412024915218353,
      "learning_rate": 3.5122447050189573e-06,
      "loss": 4.285486221313477,
      "step": 11160
    },
    {
      "epoch": 1.8871430984963675,
      "grad_norm": 0.45283156633377075,
      "learning_rate": 3.4099439484844947e-06,
      "loss": 4.294749069213867,
      "step": 11170
    },
    {
      "epoch": 1.8888325730697753,
      "grad_norm": 0.4460100829601288,
      "learning_rate": 3.3091380165998103e-06,
      "loss": 4.310376739501953,
      "step": 11180
    },
    {
      "epoch": 1.890522047643183,
      "grad_norm": 0.44468414783477783,
      "learning_rate": 3.2098279373145463e-06,
      "loss": 4.327771377563477,
      "step": 11190
    },
    {
      "epoch": 1.8922115222165905,
      "grad_norm": 0.4378024637699127,
      "learning_rate": 3.1120147233246463e-06,
      "loss": 4.273694610595703,
      "step": 11200
    },
    {
      "epoch": 1.8939009967899985,
      "grad_norm": 0.44141000509262085,
      "learning_rate": 3.0156993720619804e-06,
      "loss": 4.287034606933593,
      "step": 11210
    },
    {
      "epoch": 1.895590471363406,
      "grad_norm": 0.4455374479293823,
      "learning_rate": 2.9208828656843876e-06,
      "loss": 4.320920181274414,
      "step": 11220
    },
    {
      "epoch": 1.8972799459368137,
      "grad_norm": 0.45368343591690063,
      "learning_rate": 2.827566171065415e-06,
      "loss": 4.285198974609375,
      "step": 11230
    },
    {
      "epoch": 1.8989694205102214,
      "grad_norm": 0.44222062826156616,
      "learning_rate": 2.7357502397845454e-06,
      "loss": 4.296764755249024,
      "step": 11240
    },
    {
      "epoch": 1.900658895083629,
      "grad_norm": 0.45191657543182373,
      "learning_rate": 2.645436008117602e-06,
      "loss": 4.27384033203125,
      "step": 11250
    },
    {
      "epoch": 1.9023483696570367,
      "grad_norm": 0.4548667371273041,
      "learning_rate": 2.5566243970270073e-06,
      "loss": 4.297956085205078,
      "step": 11260
    },
    {
      "epoch": 1.9040378442304444,
      "grad_norm": 0.4443969130516052,
      "learning_rate": 2.469316312152575e-06,
      "loss": 4.291641616821289,
      "step": 11270
    },
    {
      "epoch": 1.905727318803852,
      "grad_norm": 0.4367770850658417,
      "learning_rate": 2.3835126438021156e-06,
      "loss": 4.266088485717773,
      "step": 11280
    },
    {
      "epoch": 1.9074167933772597,
      "grad_norm": 0.4365804195404053,
      "learning_rate": 2.299214266942495e-06,
      "loss": 4.263021850585938,
      "step": 11290
    },
    {
      "epoch": 1.9091062679506674,
      "grad_norm": 0.4369988441467285,
      "learning_rate": 2.2164220411906407e-06,
      "loss": 4.288222122192383,
      "step": 11300
    },
    {
      "epoch": 1.910795742524075,
      "grad_norm": 0.44547080993652344,
      "learning_rate": 2.1351368108047495e-06,
      "loss": 4.26991081237793,
      "step": 11310
    },
    {
      "epoch": 1.9124852170974826,
      "grad_norm": 0.45165297389030457,
      "learning_rate": 2.0553594046757438e-06,
      "loss": 4.2671764373779295,
      "step": 11320
    },
    {
      "epoch": 1.9141746916708904,
      "grad_norm": 0.4523044526576996,
      "learning_rate": 1.9770906363187787e-06,
      "loss": 4.28791618347168,
      "step": 11330
    },
    {
      "epoch": 1.915864166244298,
      "grad_norm": 0.43898409605026245,
      "learning_rate": 1.9003313038649826e-06,
      "loss": 4.301726150512695,
      "step": 11340
    },
    {
      "epoch": 1.9175536408177059,
      "grad_norm": 0.44454851746559143,
      "learning_rate": 1.825082190053262e-06,
      "loss": 4.280124664306641,
      "step": 11350
    },
    {
      "epoch": 1.9192431153911134,
      "grad_norm": 0.44016656279563904,
      "learning_rate": 1.7513440622223762e-06,
      "loss": 4.312954330444336,
      "step": 11360
    },
    {
      "epoch": 1.9209325899645209,
      "grad_norm": 0.439481645822525,
      "learning_rate": 1.6791176723030763e-06,
      "loss": 4.291484069824219,
      "step": 11370
    },
    {
      "epoch": 1.9226220645379288,
      "grad_norm": 0.44403141736984253,
      "learning_rate": 1.608403756810428e-06,
      "loss": 4.297753524780274,
      "step": 11380
    },
    {
      "epoch": 1.9243115391113363,
      "grad_norm": 0.4380677342414856,
      "learning_rate": 1.5392030368363839e-06,
      "loss": 4.311534881591797,
      "step": 11390
    },
    {
      "epoch": 1.926001013684744,
      "grad_norm": 0.4598468542098999,
      "learning_rate": 1.4715162180422902e-06,
      "loss": 4.272250747680664,
      "step": 11400
    },
    {
      "epoch": 1.9276904882581518,
      "grad_norm": 0.44107797741889954,
      "learning_rate": 1.405343990651825e-06,
      "loss": 4.283835601806641,
      "step": 11410
    },
    {
      "epoch": 1.9293799628315593,
      "grad_norm": 0.43588972091674805,
      "learning_rate": 1.3406870294438876e-06,
      "loss": 4.273925399780273,
      "step": 11420
    },
    {
      "epoch": 1.931069437404967,
      "grad_norm": 0.4393414258956909,
      "learning_rate": 1.2775459937457544e-06,
      "loss": 4.295301055908203,
      "step": 11430
    },
    {
      "epoch": 1.9327589119783748,
      "grad_norm": 0.44228672981262207,
      "learning_rate": 1.2159215274262834e-06,
      "loss": 4.273171997070312,
      "step": 11440
    },
    {
      "epoch": 1.9344483865517823,
      "grad_norm": 0.4422619640827179,
      "learning_rate": 1.155814258889437e-06,
      "loss": 4.285517883300781,
      "step": 11450
    },
    {
      "epoch": 1.93613786112519,
      "grad_norm": 0.4427924156188965,
      "learning_rate": 1.0972248010678365e-06,
      "loss": 4.312974548339843,
      "step": 11460
    },
    {
      "epoch": 1.9378273356985978,
      "grad_norm": 0.44030192494392395,
      "learning_rate": 1.040153751416517e-06,
      "loss": 4.302379989624024,
      "step": 11470
    },
    {
      "epoch": 1.9395168102720053,
      "grad_norm": 0.44400596618652344,
      "learning_rate": 9.846016919068167e-07,
      "loss": 4.280198287963867,
      "step": 11480
    },
    {
      "epoch": 1.941206284845413,
      "grad_norm": 0.4478650689125061,
      "learning_rate": 9.305691890204469e-07,
      "loss": 4.281633758544922,
      "step": 11490
    },
    {
      "epoch": 1.9428957594188208,
      "grad_norm": 0.4458984434604645,
      "learning_rate": 8.780567937437644e-07,
      "loss": 4.278944396972657,
      "step": 11500
    },
    {
      "epoch": 1.9428957594188208,
      "eval_loss": 4.265942573547363,
      "eval_runtime": 4.0629,
      "eval_samples_per_second": 246.128,
      "eval_steps_per_second": 5.169,
      "step": 11500
    },
    {
      "epoch": 1.9445852339922283,
      "grad_norm": 0.4352650046348572,
      "learning_rate": 8.270650415620584e-07,
      "loss": 4.2965538024902346,
      "step": 11510
    },
    {
      "epoch": 1.9462747085656362,
      "grad_norm": 0.43585142493247986,
      "learning_rate": 7.775944524542055e-07,
      "loss": 4.270129776000976,
      "step": 11520
    },
    {
      "epoch": 1.9479641831390437,
      "grad_norm": 0.4469541907310486,
      "learning_rate": 7.296455308872406e-07,
      "loss": 4.283909606933594,
      "step": 11530
    },
    {
      "epoch": 1.9496536577124515,
      "grad_norm": 0.4361380636692047,
      "learning_rate": 6.832187658113441e-07,
      "loss": 4.296160125732422,
      "step": 11540
    },
    {
      "epoch": 1.9513431322858592,
      "grad_norm": 0.44409504532814026,
      "learning_rate": 6.383146306547626e-07,
      "loss": 4.304541778564453,
      "step": 11550
    },
    {
      "epoch": 1.9530326068592667,
      "grad_norm": 0.45062074065208435,
      "learning_rate": 5.949335833189628e-07,
      "loss": 4.3281913757324215,
      "step": 11560
    },
    {
      "epoch": 1.9547220814326745,
      "grad_norm": 0.45208507776260376,
      "learning_rate": 5.530760661741018e-07,
      "loss": 4.3035846710205075,
      "step": 11570
    },
    {
      "epoch": 1.9564115560060822,
      "grad_norm": 0.44333794713020325,
      "learning_rate": 5.127425060543478e-07,
      "loss": 4.278887939453125,
      "step": 11580
    },
    {
      "epoch": 1.9581010305794897,
      "grad_norm": 0.44367748498916626,
      "learning_rate": 4.7393331425364943e-07,
      "loss": 4.281793594360352,
      "step": 11590
    },
    {
      "epoch": 1.9597905051528974,
      "grad_norm": 0.4411092698574066,
      "learning_rate": 4.3664888652144017e-07,
      "loss": 4.278807067871094,
      "step": 11600
    },
    {
      "epoch": 1.9614799797263052,
      "grad_norm": 0.44609910249710083,
      "learning_rate": 4.008896030587072e-07,
      "loss": 4.270274353027344,
      "step": 11610
    },
    {
      "epoch": 1.9631694542997127,
      "grad_norm": 0.43740522861480713,
      "learning_rate": 3.6665582851406195e-07,
      "loss": 4.296014785766602,
      "step": 11620
    },
    {
      "epoch": 1.9648589288731204,
      "grad_norm": 0.44448962807655334,
      "learning_rate": 3.3394791198000927e-07,
      "loss": 4.282284927368164,
      "step": 11630
    },
    {
      "epoch": 1.9665484034465281,
      "grad_norm": 0.45065152645111084,
      "learning_rate": 3.027661869893672e-07,
      "loss": 4.2820892333984375,
      "step": 11640
    },
    {
      "epoch": 1.9682378780199357,
      "grad_norm": 0.4398045539855957,
      "learning_rate": 2.731109715119861e-07,
      "loss": 4.281244277954102,
      "step": 11650
    },
    {
      "epoch": 1.9699273525933436,
      "grad_norm": 0.4467960000038147,
      "learning_rate": 2.4498256795135173e-07,
      "loss": 4.307322311401367,
      "step": 11660
    },
    {
      "epoch": 1.9716168271667511,
      "grad_norm": 0.4327242970466614,
      "learning_rate": 2.183812631415871e-07,
      "loss": 4.275672149658203,
      "step": 11670
    },
    {
      "epoch": 1.9733063017401589,
      "grad_norm": 0.43306484818458557,
      "learning_rate": 1.933073283445219e-07,
      "loss": 4.291437149047852,
      "step": 11680
    },
    {
      "epoch": 1.9749957763135666,
      "grad_norm": 0.4464097023010254,
      "learning_rate": 1.697610192469112e-07,
      "loss": 4.312542343139649,
      "step": 11690
    },
    {
      "epoch": 1.976685250886974,
      "grad_norm": 0.4436480700969696,
      "learning_rate": 1.4774257595783766e-07,
      "loss": 4.300673294067383,
      "step": 11700
    },
    {
      "epoch": 1.9783747254603818,
      "grad_norm": 0.44450485706329346,
      "learning_rate": 1.272522230062467e-07,
      "loss": 4.290340805053711,
      "step": 11710
    },
    {
      "epoch": 1.9800642000337896,
      "grad_norm": 0.4362986981868744,
      "learning_rate": 1.0829016933869838e-07,
      "loss": 4.2894245147705075,
      "step": 11720
    },
    {
      "epoch": 1.981753674607197,
      "grad_norm": 0.43450725078582764,
      "learning_rate": 9.085660831715247e-08,
      "loss": 4.298795700073242,
      "step": 11730
    },
    {
      "epoch": 1.9834431491806048,
      "grad_norm": 0.44246765971183777,
      "learning_rate": 7.495171771710328e-08,
      "loss": 4.293585968017578,
      "step": 11740
    },
    {
      "epoch": 1.9851326237540126,
      "grad_norm": 0.43929263949394226,
      "learning_rate": 6.057565972568123e-08,
      "loss": 4.293174743652344,
      "step": 11750
    },
    {
      "epoch": 1.98682209832742,
      "grad_norm": 0.4450415372848511,
      "learning_rate": 4.772858094005405e-08,
      "loss": 4.3004913330078125,
      "step": 11760
    },
    {
      "epoch": 1.9885115729008278,
      "grad_norm": 0.4472520053386688,
      "learning_rate": 3.641061236591136e-08,
      "loss": 4.2836250305175785,
      "step": 11770
    },
    {
      "epoch": 1.9902010474742355,
      "grad_norm": 0.44302183389663696,
      "learning_rate": 2.6621869416099118e-08,
      "loss": 4.290175247192383,
      "step": 11780
    },
    {
      "epoch": 1.991890522047643,
      "grad_norm": 0.4414844512939453,
      "learning_rate": 1.8362451909520458e-08,
      "loss": 4.286873245239258,
      "step": 11790
    },
    {
      "epoch": 1.993579996621051,
      "grad_norm": 0.44598934054374695,
      "learning_rate": 1.16324440700033e-08,
      "loss": 4.297615051269531,
      "step": 11800
    },
    {
      "epoch": 1.9952694711944585,
      "grad_norm": 0.4523853361606598,
      "learning_rate": 6.431914525567572e-09,
      "loss": 4.289733505249023,
      "step": 11810
    },
    {
      "epoch": 1.996958945767866,
      "grad_norm": 0.4494129419326782,
      "learning_rate": 2.760916307625871e-09,
      "loss": 4.304800415039063,
      "step": 11820
    },
    {
      "epoch": 1.998648420341274,
      "grad_norm": 0.4344528913497925,
      "learning_rate": 6.194868504838524e-10,
      "loss": 4.279055786132813,
      "step": 11830
    }
  ],
  "logging_steps": 10,
  "max_steps": 11838,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.959258038224814e+17,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": null
}