{
  "best_metric": 0.14174455404281616,
  "best_model_checkpoint": "/home/ray/default/save/Orca_End/checkpoint-4000",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 4272,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0046816479400749065,
      "grad_norm": 0.14989347755908966,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.9233,
      "step": 10
    },
    {
      "epoch": 0.009363295880149813,
      "grad_norm": 0.16152486205101013,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.9302,
      "step": 20
    },
    {
      "epoch": 0.014044943820224719,
      "grad_norm": 0.14786499738693237,
      "learning_rate": 1.5e-06,
      "loss": 0.9066,
      "step": 30
    },
    {
      "epoch": 0.018726591760299626,
      "grad_norm": 0.17683209478855133,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.8958,
      "step": 40
    },
    {
      "epoch": 0.023408239700374533,
      "grad_norm": 0.18914400041103363,
      "learning_rate": 2.5e-06,
      "loss": 0.8957,
      "step": 50
    },
    {
      "epoch": 0.028089887640449437,
      "grad_norm": 0.17229343950748444,
      "learning_rate": 3e-06,
      "loss": 0.9006,
      "step": 60
    },
    {
      "epoch": 0.03277153558052434,
      "grad_norm": 0.19986914098262787,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 0.8843,
      "step": 70
    },
    {
      "epoch": 0.03745318352059925,
      "grad_norm": 0.23050718009471893,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.8824,
      "step": 80
    },
    {
      "epoch": 0.042134831460674156,
      "grad_norm": 0.2016942948102951,
      "learning_rate": 4.5e-06,
      "loss": 0.8842,
      "step": 90
    },
    {
      "epoch": 0.04681647940074907,
      "grad_norm": 0.24438633024692535,
      "learning_rate": 5e-06,
      "loss": 0.8783,
      "step": 100
    },
    {
      "epoch": 0.05149812734082397,
      "grad_norm": 0.33506831526756287,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.854,
      "step": 110
    },
    {
      "epoch": 0.056179775280898875,
      "grad_norm": 0.29310888051986694,
      "learning_rate": 6e-06,
      "loss": 0.8332,
      "step": 120
    },
    {
      "epoch": 0.060861423220973786,
      "grad_norm": 0.3248985707759857,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.8105,
      "step": 130
    },
    {
      "epoch": 0.06554307116104868,
      "grad_norm": 0.31335774064064026,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.7664,
      "step": 140
    },
    {
      "epoch": 0.0702247191011236,
      "grad_norm": 0.2475409060716629,
      "learning_rate": 7.5e-06,
      "loss": 0.7265,
      "step": 150
    },
    {
      "epoch": 0.0749063670411985,
      "grad_norm": 0.19633963704109192,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.694,
      "step": 160
    },
    {
      "epoch": 0.07958801498127341,
      "grad_norm": 0.18646350502967834,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.6975,
      "step": 170
    },
    {
      "epoch": 0.08426966292134831,
      "grad_norm": 0.17781223356723785,
      "learning_rate": 9e-06,
      "loss": 0.6493,
      "step": 180
    },
    {
      "epoch": 0.08895131086142322,
      "grad_norm": 0.15868598222732544,
      "learning_rate": 9.5e-06,
      "loss": 0.6194,
      "step": 190
    },
    {
      "epoch": 0.09363295880149813,
      "grad_norm": 0.14928707480430603,
      "learning_rate": 1e-05,
      "loss": 0.6004,
      "step": 200
    },
    {
      "epoch": 0.09831460674157304,
      "grad_norm": 0.18063120543956757,
      "learning_rate": 1.05e-05,
      "loss": 0.5937,
      "step": 210
    },
    {
      "epoch": 0.10299625468164794,
      "grad_norm": 0.1594826579093933,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.5582,
      "step": 220
    },
    {
      "epoch": 0.10767790262172285,
      "grad_norm": 0.15458233654499054,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.5495,
      "step": 230
    },
    {
      "epoch": 0.11235955056179775,
      "grad_norm": 0.14062783122062683,
      "learning_rate": 1.2e-05,
      "loss": 0.5155,
      "step": 240
    },
    {
      "epoch": 0.11704119850187265,
      "grad_norm": 0.16381382942199707,
      "learning_rate": 1.25e-05,
      "loss": 0.4931,
      "step": 250
    },
    {
      "epoch": 0.12172284644194757,
      "grad_norm": 0.17048245668411255,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.4842,
      "step": 260
    },
    {
      "epoch": 0.12640449438202248,
      "grad_norm": 0.16214580833911896,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.4565,
      "step": 270
    },
    {
      "epoch": 0.13108614232209737,
      "grad_norm": 0.18523240089416504,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.4303,
      "step": 280
    },
    {
      "epoch": 0.13576779026217228,
      "grad_norm": 0.16191580891609192,
      "learning_rate": 1.45e-05,
      "loss": 0.3903,
      "step": 290
    },
    {
      "epoch": 0.1404494382022472,
      "grad_norm": 0.17054912447929382,
      "learning_rate": 1.5e-05,
      "loss": 0.3958,
      "step": 300
    },
    {
      "epoch": 0.1451310861423221,
      "grad_norm": 0.15889020264148712,
      "learning_rate": 1.55e-05,
      "loss": 0.3679,
      "step": 310
    },
    {
      "epoch": 0.149812734082397,
      "grad_norm": 0.1658034473657608,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.3566,
      "step": 320
    },
    {
      "epoch": 0.1544943820224719,
      "grad_norm": 0.1820608377456665,
      "learning_rate": 1.65e-05,
      "loss": 0.3353,
      "step": 330
    },
    {
      "epoch": 0.15917602996254682,
      "grad_norm": 0.18187420070171356,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.3335,
      "step": 340
    },
    {
      "epoch": 0.16385767790262173,
      "grad_norm": 0.20398898422718048,
      "learning_rate": 1.75e-05,
      "loss": 0.3273,
      "step": 350
    },
    {
      "epoch": 0.16853932584269662,
      "grad_norm": 0.19581130146980286,
      "learning_rate": 1.8e-05,
      "loss": 0.2973,
      "step": 360
    },
    {
      "epoch": 0.17322097378277154,
      "grad_norm": 0.1903996765613556,
      "learning_rate": 1.85e-05,
      "loss": 0.2939,
      "step": 370
    },
    {
      "epoch": 0.17790262172284643,
      "grad_norm": 0.21215596795082092,
      "learning_rate": 1.9e-05,
      "loss": 0.2836,
      "step": 380
    },
    {
      "epoch": 0.18258426966292135,
      "grad_norm": 0.20406150817871094,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.2766,
      "step": 390
    },
    {
      "epoch": 0.18726591760299627,
      "grad_norm": 0.19281837344169617,
      "learning_rate": 2e-05,
      "loss": 0.2716,
      "step": 400
    },
    {
      "epoch": 0.19194756554307116,
      "grad_norm": 0.1925257295370102,
      "learning_rate": 2.05e-05,
      "loss": 0.2619,
      "step": 410
    },
    {
      "epoch": 0.19662921348314608,
      "grad_norm": 0.20710717141628265,
      "learning_rate": 2.1e-05,
      "loss": 0.2676,
      "step": 420
    },
    {
      "epoch": 0.20131086142322097,
      "grad_norm": 0.2195947766304016,
      "learning_rate": 2.15e-05,
      "loss": 0.2675,
      "step": 430
    },
    {
      "epoch": 0.20599250936329588,
      "grad_norm": 0.23015384376049042,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.24,
      "step": 440
    },
    {
      "epoch": 0.21067415730337077,
      "grad_norm": 0.23879607021808624,
      "learning_rate": 2.25e-05,
      "loss": 0.251,
      "step": 450
    },
    {
      "epoch": 0.2153558052434457,
      "grad_norm": 0.25906631350517273,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.2485,
      "step": 460
    },
    {
      "epoch": 0.2200374531835206,
      "grad_norm": 0.2271612286567688,
      "learning_rate": 2.35e-05,
      "loss": 0.2291,
      "step": 470
    },
    {
      "epoch": 0.2247191011235955,
      "grad_norm": 0.23806224763393402,
      "learning_rate": 2.4e-05,
      "loss": 0.2262,
      "step": 480
    },
    {
      "epoch": 0.22940074906367042,
      "grad_norm": 0.23863674700260162,
      "learning_rate": 2.45e-05,
      "loss": 0.2481,
      "step": 490
    },
    {
      "epoch": 0.2340823970037453,
      "grad_norm": 0.23864531517028809,
      "learning_rate": 2.5e-05,
      "loss": 0.2267,
      "step": 500
    },
    {
      "epoch": 0.2340823970037453,
      "eval_loss": 0.22292079031467438,
      "eval_runtime": 233.7266,
      "eval_samples_per_second": 2.986,
      "eval_steps_per_second": 1.493,
      "step": 500
    },
    {
      "epoch": 0.23876404494382023,
      "grad_norm": 0.229400172829628,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 0.2235,
      "step": 510
    },
    {
      "epoch": 0.24344569288389514,
      "grad_norm": 0.25184643268585205,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.2177,
      "step": 520
    },
    {
      "epoch": 0.24812734082397003,
      "grad_norm": 0.255626916885376,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.2283,
      "step": 530
    },
    {
      "epoch": 0.25280898876404495,
      "grad_norm": 0.2295573651790619,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.2203,
      "step": 540
    },
    {
      "epoch": 0.25749063670411987,
      "grad_norm": 0.24802301824092865,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.2218,
      "step": 550
    },
    {
      "epoch": 0.26217228464419473,
      "grad_norm": 0.26119089126586914,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.2073,
      "step": 560
    },
    {
      "epoch": 0.26685393258426965,
      "grad_norm": 0.23319612443447113,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.215,
      "step": 570
    },
    {
      "epoch": 0.27153558052434457,
      "grad_norm": 0.2370581328868866,
      "learning_rate": 2.9e-05,
      "loss": 0.2066,
      "step": 580
    },
    {
      "epoch": 0.2762172284644195,
      "grad_norm": 0.22631758451461792,
      "learning_rate": 2.95e-05,
      "loss": 0.2054,
      "step": 590
    },
    {
      "epoch": 0.2808988764044944,
      "grad_norm": 0.259056031703949,
      "learning_rate": 3e-05,
      "loss": 0.2162,
      "step": 600
    },
    {
      "epoch": 0.28558052434456926,
      "grad_norm": 0.24505385756492615,
      "learning_rate": 3.05e-05,
      "loss": 0.2033,
      "step": 610
    },
    {
      "epoch": 0.2902621722846442,
      "grad_norm": 0.22634395956993103,
      "learning_rate": 3.1e-05,
      "loss": 0.1967,
      "step": 620
    },
    {
      "epoch": 0.2949438202247191,
      "grad_norm": 0.3079870641231537,
      "learning_rate": 3.15e-05,
      "loss": 0.2125,
      "step": 630
    },
    {
      "epoch": 0.299625468164794,
      "grad_norm": 0.2437993735074997,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.2136,
      "step": 640
    },
    {
      "epoch": 0.30430711610486894,
      "grad_norm": 0.25136232376098633,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.2173,
      "step": 650
    },
    {
      "epoch": 0.3089887640449438,
      "grad_norm": 0.263988733291626,
      "learning_rate": 3.3e-05,
      "loss": 0.2245,
      "step": 660
    },
    {
      "epoch": 0.3136704119850187,
      "grad_norm": 0.3366784453392029,
      "learning_rate": 3.35e-05,
      "loss": 0.1881,
      "step": 670
    },
    {
      "epoch": 0.31835205992509363,
      "grad_norm": 0.22763469815254211,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.2009,
      "step": 680
    },
    {
      "epoch": 0.32303370786516855,
      "grad_norm": 0.28085455298423767,
      "learning_rate": 3.45e-05,
      "loss": 0.193,
      "step": 690
    },
    {
      "epoch": 0.32771535580524347,
      "grad_norm": 0.21730464696884155,
      "learning_rate": 3.5e-05,
      "loss": 0.1956,
      "step": 700
    },
    {
      "epoch": 0.33239700374531833,
      "grad_norm": 0.2521045506000519,
      "learning_rate": 3.55e-05,
      "loss": 0.2393,
      "step": 710
    },
    {
      "epoch": 0.33707865168539325,
      "grad_norm": 0.2687033414840698,
      "learning_rate": 3.6e-05,
      "loss": 0.193,
      "step": 720
    },
    {
      "epoch": 0.34176029962546817,
      "grad_norm": 0.23822762072086334,
      "learning_rate": 3.65e-05,
      "loss": 0.1897,
      "step": 730
    },
    {
      "epoch": 0.3464419475655431,
      "grad_norm": 0.2817654609680176,
      "learning_rate": 3.7e-05,
      "loss": 0.1874,
      "step": 740
    },
    {
      "epoch": 0.351123595505618,
      "grad_norm": 0.2669580578804016,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.1876,
      "step": 750
    },
    {
      "epoch": 0.35580524344569286,
      "grad_norm": 0.26698926091194153,
      "learning_rate": 3.8e-05,
      "loss": 0.189,
      "step": 760
    },
    {
      "epoch": 0.3604868913857678,
      "grad_norm": 0.24417920410633087,
      "learning_rate": 3.85e-05,
      "loss": 0.1904,
      "step": 770
    },
    {
      "epoch": 0.3651685393258427,
      "grad_norm": 0.3116750121116638,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.1894,
      "step": 780
    },
    {
      "epoch": 0.3698501872659176,
      "grad_norm": 0.269699364900589,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 0.1948,
      "step": 790
    },
    {
      "epoch": 0.37453183520599254,
      "grad_norm": 0.24539124965667725,
      "learning_rate": 4e-05,
      "loss": 0.1924,
      "step": 800
    },
    {
      "epoch": 0.3792134831460674,
      "grad_norm": 0.24850206077098846,
      "learning_rate": 4.05e-05,
      "loss": 0.1902,
      "step": 810
    },
    {
      "epoch": 0.3838951310861423,
      "grad_norm": 0.24830132722854614,
      "learning_rate": 4.1e-05,
      "loss": 0.1753,
      "step": 820
    },
    {
      "epoch": 0.38857677902621723,
      "grad_norm": 0.28210338950157166,
      "learning_rate": 4.15e-05,
      "loss": 0.183,
      "step": 830
    },
    {
      "epoch": 0.39325842696629215,
      "grad_norm": 0.2719564437866211,
      "learning_rate": 4.2e-05,
      "loss": 0.1682,
      "step": 840
    },
    {
      "epoch": 0.397940074906367,
      "grad_norm": 0.2886371612548828,
      "learning_rate": 4.25e-05,
      "loss": 0.1756,
      "step": 850
    },
    {
      "epoch": 0.40262172284644193,
      "grad_norm": 0.27633577585220337,
      "learning_rate": 4.3e-05,
      "loss": 0.1879,
      "step": 860
    },
    {
      "epoch": 0.40730337078651685,
      "grad_norm": 0.24454927444458008,
      "learning_rate": 4.35e-05,
      "loss": 0.1832,
      "step": 870
    },
    {
      "epoch": 0.41198501872659177,
      "grad_norm": 0.258569598197937,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1781,
      "step": 880
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.24131977558135986,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 0.1751,
      "step": 890
    },
    {
      "epoch": 0.42134831460674155,
      "grad_norm": 0.26839905977249146,
      "learning_rate": 4.5e-05,
      "loss": 0.1853,
      "step": 900
    },
    {
      "epoch": 0.42602996254681647,
      "grad_norm": 0.2593300938606262,
      "learning_rate": 4.55e-05,
      "loss": 0.208,
      "step": 910
    },
    {
      "epoch": 0.4307116104868914,
      "grad_norm": 0.25040218234062195,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1803,
      "step": 920
    },
    {
      "epoch": 0.4353932584269663,
      "grad_norm": 0.2636476159095764,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 0.1743,
      "step": 930
    },
    {
      "epoch": 0.4400749063670412,
      "grad_norm": 0.2569904625415802,
      "learning_rate": 4.7e-05,
      "loss": 0.1777,
      "step": 940
    },
    {
      "epoch": 0.4447565543071161,
      "grad_norm": 0.2652021050453186,
      "learning_rate": 4.75e-05,
      "loss": 0.1759,
      "step": 950
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 0.25217893719673157,
      "learning_rate": 4.8e-05,
      "loss": 0.183,
      "step": 960
    },
    {
      "epoch": 0.4541198501872659,
      "grad_norm": 0.26192978024482727,
      "learning_rate": 4.85e-05,
      "loss": 0.1773,
      "step": 970
    },
    {
      "epoch": 0.45880149812734083,
      "grad_norm": 0.2585752308368683,
      "learning_rate": 4.9e-05,
      "loss": 0.1811,
      "step": 980
    },
    {
      "epoch": 0.46348314606741575,
      "grad_norm": 0.2950770854949951,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 0.1742,
      "step": 990
    },
    {
      "epoch": 0.4681647940074906,
      "grad_norm": 0.24856804311275482,
      "learning_rate": 5e-05,
      "loss": 0.177,
      "step": 1000
    },
    {
      "epoch": 0.4681647940074906,
      "eval_loss": 0.17352555692195892,
      "eval_runtime": 236.2712,
      "eval_samples_per_second": 2.954,
      "eval_steps_per_second": 1.477,
      "step": 1000
    },
    {
      "epoch": 0.47284644194756553,
      "grad_norm": 0.2841172218322754,
      "learning_rate": 4.9998847662136905e-05,
      "loss": 0.1799,
      "step": 1010
    },
    {
      "epoch": 0.47752808988764045,
      "grad_norm": 0.24798014760017395,
      "learning_rate": 4.999539075477821e-05,
      "loss": 0.1837,
      "step": 1020
    },
    {
      "epoch": 0.48220973782771537,
      "grad_norm": 0.2636326551437378,
      "learning_rate": 4.998962959660594e-05,
      "loss": 0.1737,
      "step": 1030
    },
    {
      "epoch": 0.4868913857677903,
      "grad_norm": 0.27235954999923706,
      "learning_rate": 4.998156471872415e-05,
      "loss": 0.1649,
      "step": 1040
    },
    {
      "epoch": 0.49157303370786515,
      "grad_norm": 0.26181653141975403,
      "learning_rate": 4.9971196864609975e-05,
      "loss": 0.1786,
      "step": 1050
    },
    {
      "epoch": 0.49625468164794007,
      "grad_norm": 0.2724978029727936,
      "learning_rate": 4.995852699004508e-05,
      "loss": 0.1709,
      "step": 1060
    },
    {
      "epoch": 0.5009363295880149,
      "grad_norm": 0.2702656090259552,
      "learning_rate": 4.9943556263027556e-05,
      "loss": 0.1714,
      "step": 1070
    },
    {
      "epoch": 0.5056179775280899,
      "grad_norm": 0.33144789934158325,
      "learning_rate": 4.992628606366426e-05,
      "loss": 0.1653,
      "step": 1080
    },
    {
      "epoch": 0.5102996254681648,
      "grad_norm": 0.2642982006072998,
      "learning_rate": 4.990671798404355e-05,
      "loss": 0.1742,
      "step": 1090
    },
    {
      "epoch": 0.5149812734082397,
      "grad_norm": 0.2788497805595398,
      "learning_rate": 4.988485382808856e-05,
      "loss": 0.1717,
      "step": 1100
    },
    {
      "epoch": 0.5196629213483146,
      "grad_norm": 0.23453935980796814,
      "learning_rate": 4.986069561139086e-05,
      "loss": 0.1913,
      "step": 1110
    },
    {
      "epoch": 0.5243445692883895,
      "grad_norm": 0.2543826401233673,
      "learning_rate": 4.983424556102469e-05,
      "loss": 0.1753,
      "step": 1120
    },
    {
      "epoch": 0.5290262172284644,
      "grad_norm": 0.28962579369544983,
      "learning_rate": 4.98055061153416e-05,
      "loss": 0.1702,
      "step": 1130
    },
    {
      "epoch": 0.5337078651685393,
      "grad_norm": 0.23189175128936768,
      "learning_rate": 4.97744799237457e-05,
      "loss": 0.1643,
      "step": 1140
    },
    {
      "epoch": 0.5383895131086143,
      "grad_norm": 0.21799345314502716,
      "learning_rate": 4.974116984644943e-05,
      "loss": 0.1835,
      "step": 1150
    },
    {
      "epoch": 0.5430711610486891,
      "grad_norm": 0.24570296704769135,
      "learning_rate": 4.970557895420984e-05,
      "loss": 0.1802,
      "step": 1160
    },
    {
      "epoch": 0.547752808988764,
      "grad_norm": 0.2388901561498642,
      "learning_rate": 4.966771052804555e-05,
      "loss": 0.1701,
      "step": 1170
    },
    {
      "epoch": 0.552434456928839,
      "grad_norm": 0.2784363627433777,
      "learning_rate": 4.9627568058934274e-05,
      "loss": 0.1697,
      "step": 1180
    },
    {
      "epoch": 0.5571161048689138,
      "grad_norm": 0.27096298336982727,
      "learning_rate": 4.958515524749096e-05,
      "loss": 0.1749,
      "step": 1190
    },
    {
      "epoch": 0.5617977528089888,
      "grad_norm": 0.25467076897621155,
      "learning_rate": 4.95404760036267e-05,
      "loss": 0.1598,
      "step": 1200
    },
    {
      "epoch": 0.5664794007490637,
      "grad_norm": 0.24679610133171082,
      "learning_rate": 4.949353444618825e-05,
      "loss": 0.1769,
      "step": 1210
    },
    {
      "epoch": 0.5711610486891385,
      "grad_norm": 0.2690020799636841,
      "learning_rate": 4.9444334902578315e-05,
      "loss": 0.1676,
      "step": 1220
    },
    {
      "epoch": 0.5758426966292135,
      "grad_norm": 0.2583072781562805,
      "learning_rate": 4.9392881908356656e-05,
      "loss": 0.1625,
      "step": 1230
    },
    {
      "epoch": 0.5805243445692884,
      "grad_norm": 0.2679818868637085,
      "learning_rate": 4.9339180206821955e-05,
      "loss": 0.1708,
      "step": 1240
    },
    {
      "epoch": 0.5852059925093633,
      "grad_norm": 0.2509777247905731,
      "learning_rate": 4.9283234748574524e-05,
      "loss": 0.1803,
      "step": 1250
    },
    {
      "epoch": 0.5898876404494382,
      "grad_norm": 0.2875745892524719,
      "learning_rate": 4.922505069105995e-05,
      "loss": 0.1712,
      "step": 1260
    },
    {
      "epoch": 0.5945692883895131,
      "grad_norm": 0.25088638067245483,
      "learning_rate": 4.9164633398093624e-05,
      "loss": 0.1591,
      "step": 1270
    },
    {
      "epoch": 0.599250936329588,
      "grad_norm": 0.27187079191207886,
      "learning_rate": 4.9101988439366295e-05,
      "loss": 0.1621,
      "step": 1280
    },
    {
      "epoch": 0.6039325842696629,
      "grad_norm": 0.23108245432376862,
      "learning_rate": 4.90371215899306e-05,
      "loss": 0.173,
      "step": 1290
    },
    {
      "epoch": 0.6086142322097379,
      "grad_norm": 0.2536250352859497,
      "learning_rate": 4.897003882966866e-05,
      "loss": 0.1692,
      "step": 1300
    },
    {
      "epoch": 0.6132958801498127,
      "grad_norm": 0.2388300895690918,
      "learning_rate": 4.8900746342740854e-05,
      "loss": 0.1676,
      "step": 1310
    },
    {
      "epoch": 0.6179775280898876,
      "grad_norm": 0.23416273295879364,
      "learning_rate": 4.8829250517015684e-05,
      "loss": 0.1739,
      "step": 1320
    },
    {
      "epoch": 0.6226591760299626,
      "grad_norm": 0.24538950622081757,
      "learning_rate": 4.87555579434809e-05,
      "loss": 0.1622,
      "step": 1330
    },
    {
      "epoch": 0.6273408239700374,
      "grad_norm": 0.27460214495658875,
      "learning_rate": 4.867967541563594e-05,
      "loss": 0.1769,
      "step": 1340
    },
    {
      "epoch": 0.6320224719101124,
      "grad_norm": 0.23591691255569458,
      "learning_rate": 4.86016099288656e-05,
      "loss": 0.1471,
      "step": 1350
    },
    {
      "epoch": 0.6367041198501873,
      "grad_norm": 0.23421670496463776,
      "learning_rate": 4.8521368679795154e-05,
      "loss": 0.1751,
      "step": 1360
    },
    {
      "epoch": 0.6413857677902621,
      "grad_norm": 0.2746402323246002,
      "learning_rate": 4.843895906562699e-05,
      "loss": 0.1602,
      "step": 1370
    },
    {
      "epoch": 0.6460674157303371,
      "grad_norm": 0.2615332007408142,
      "learning_rate": 4.835438868345858e-05,
      "loss": 0.1627,
      "step": 1380
    },
    {
      "epoch": 0.650749063670412,
      "grad_norm": 0.24973183870315552,
      "learning_rate": 4.826766532958221e-05,
      "loss": 0.163,
      "step": 1390
    },
    {
      "epoch": 0.6554307116104869,
      "grad_norm": 0.3358960449695587,
      "learning_rate": 4.817879699876623e-05,
      "loss": 0.1732,
      "step": 1400
    },
    {
      "epoch": 0.6601123595505618,
      "grad_norm": 0.23806847631931305,
      "learning_rate": 4.808779188351803e-05,
      "loss": 0.1513,
      "step": 1410
    },
    {
      "epoch": 0.6647940074906367,
      "grad_norm": 0.30218422412872314,
      "learning_rate": 4.7994658373328804e-05,
      "loss": 0.1683,
      "step": 1420
    },
    {
      "epoch": 0.6694756554307116,
      "grad_norm": 0.28052952885627747,
      "learning_rate": 4.7899405053900174e-05,
      "loss": 0.1656,
      "step": 1430
    },
    {
      "epoch": 0.6741573033707865,
      "grad_norm": 0.2771318852901459,
      "learning_rate": 4.780204070635266e-05,
      "loss": 0.1683,
      "step": 1440
    },
    {
      "epoch": 0.6788389513108615,
      "grad_norm": 0.28290867805480957,
      "learning_rate": 4.770257430641619e-05,
      "loss": 0.1619,
      "step": 1450
    },
    {
      "epoch": 0.6835205992509363,
      "grad_norm": 0.2428758144378662,
      "learning_rate": 4.760101502360268e-05,
      "loss": 0.1677,
      "step": 1460
    },
    {
      "epoch": 0.6882022471910112,
      "grad_norm": 0.26167434453964233,
      "learning_rate": 4.749737222036067e-05,
      "loss": 0.1657,
      "step": 1470
    },
    {
      "epoch": 0.6928838951310862,
      "grad_norm": 0.27498582005500793,
      "learning_rate": 4.739165545121228e-05,
      "loss": 0.1678,
      "step": 1480
    },
    {
      "epoch": 0.697565543071161,
      "grad_norm": 0.2729678750038147,
      "learning_rate": 4.728387446187238e-05,
      "loss": 0.1586,
      "step": 1490
    },
    {
      "epoch": 0.702247191011236,
      "grad_norm": 0.28416207432746887,
      "learning_rate": 4.717403918835017e-05,
      "loss": 0.1592,
      "step": 1500
    },
    {
      "epoch": 0.702247191011236,
      "eval_loss": 0.15914995968341827,
      "eval_runtime": 230.5999,
      "eval_samples_per_second": 3.027,
      "eval_steps_per_second": 1.513,
      "step": 1500
    },
    {
      "epoch": 0.7069288389513109,
      "grad_norm": 0.2723831832408905,
      "learning_rate": 4.706215975603319e-05,
      "loss": 0.1603,
      "step": 1510
    },
    {
      "epoch": 0.7116104868913857,
      "grad_norm": 0.2503916025161743,
      "learning_rate": 4.694824647875391e-05,
      "loss": 0.1656,
      "step": 1520
    },
    {
      "epoch": 0.7162921348314607,
      "grad_norm": 0.2601824998855591,
      "learning_rate": 4.6832309857838946e-05,
      "loss": 0.15,
      "step": 1530
    },
    {
      "epoch": 0.7209737827715356,
      "grad_norm": 0.24705946445465088,
      "learning_rate": 4.6714360581140935e-05,
      "loss": 0.1538,
      "step": 1540
    },
    {
      "epoch": 0.7256554307116105,
      "grad_norm": 0.22847461700439453,
      "learning_rate": 4.6594409522053274e-05,
      "loss": 0.1633,
      "step": 1550
    },
    {
      "epoch": 0.7303370786516854,
      "grad_norm": 0.32371050119400024,
      "learning_rate": 4.647246773850773e-05,
      "loss": 0.1582,
      "step": 1560
    },
    {
      "epoch": 0.7350187265917603,
      "grad_norm": 0.24720498919487,
      "learning_rate": 4.6348546471955046e-05,
      "loss": 0.1741,
      "step": 1570
    },
    {
      "epoch": 0.7397003745318352,
      "grad_norm": 0.2715936601161957,
      "learning_rate": 4.6222657146328624e-05,
      "loss": 0.1564,
      "step": 1580
    },
    {
      "epoch": 0.7443820224719101,
      "grad_norm": 0.24092724919319153,
      "learning_rate": 4.609481136699138e-05,
      "loss": 0.1864,
      "step": 1590
    },
    {
      "epoch": 0.7490636704119851,
      "grad_norm": 0.2625042200088501,
      "learning_rate": 4.596502091966587e-05,
      "loss": 0.1863,
      "step": 1600
    },
    {
      "epoch": 0.7537453183520599,
      "grad_norm": 0.2534978687763214,
      "learning_rate": 4.583329776934787e-05,
      "loss": 0.1768,
      "step": 1610
    },
    {
      "epoch": 0.7584269662921348,
      "grad_norm": 0.26979076862335205,
      "learning_rate": 4.5699654059203225e-05,
      "loss": 0.1587,
      "step": 1620
    },
    {
      "epoch": 0.7631086142322098,
      "grad_norm": 0.21820847690105438,
      "learning_rate": 4.5564102109448554e-05,
      "loss": 0.1597,
      "step": 1630
    },
    {
      "epoch": 0.7677902621722846,
      "grad_norm": 0.22770282626152039,
      "learning_rate": 4.542665441621537e-05,
      "loss": 0.1639,
      "step": 1640
    },
    {
      "epoch": 0.7724719101123596,
      "grad_norm": 0.27316999435424805,
      "learning_rate": 4.5287323650398164e-05,
      "loss": 0.1546,
      "step": 1650
    },
    {
      "epoch": 0.7771535580524345,
      "grad_norm": 0.27699345350265503,
      "learning_rate": 4.51461226564863e-05,
      "loss": 0.1725,
      "step": 1660
    },
    {
      "epoch": 0.7818352059925093,
      "grad_norm": 0.24093034863471985,
      "learning_rate": 4.50030644513799e-05,
      "loss": 0.1711,
      "step": 1670
    },
    {
      "epoch": 0.7865168539325843,
      "grad_norm": 0.27285608649253845,
      "learning_rate": 4.485816222318986e-05,
      "loss": 0.1487,
      "step": 1680
    },
    {
      "epoch": 0.7911985018726592,
      "grad_norm": 0.2582044303417206,
      "learning_rate": 4.471142933002213e-05,
      "loss": 0.1582,
      "step": 1690
    },
    {
      "epoch": 0.795880149812734,
      "grad_norm": 0.25239214301109314,
      "learning_rate": 4.4562879298746165e-05,
      "loss": 0.155,
      "step": 1700
    },
    {
      "epoch": 0.800561797752809,
      "grad_norm": 0.2343159019947052,
      "learning_rate": 4.441252582374802e-05,
      "loss": 0.1591,
      "step": 1710
    },
    {
      "epoch": 0.8052434456928839,
      "grad_norm": 0.264161080121994,
      "learning_rate": 4.4260382765667875e-05,
      "loss": 0.1589,
      "step": 1720
    },
    {
      "epoch": 0.8099250936329588,
      "grad_norm": 0.25752800703048706,
      "learning_rate": 4.410646415012224e-05,
      "loss": 0.1648,
      "step": 1730
    },
    {
      "epoch": 0.8146067415730337,
      "grad_norm": 0.28275129199028015,
      "learning_rate": 4.395078416641099e-05,
      "loss": 0.1474,
      "step": 1740
    },
    {
      "epoch": 0.8192883895131086,
      "grad_norm": 0.260202556848526,
      "learning_rate": 4.3793357166209313e-05,
      "loss": 0.1626,
      "step": 1750
    },
    {
      "epoch": 0.8239700374531835,
      "grad_norm": 0.2819097340106964,
      "learning_rate": 4.363419766224465e-05,
      "loss": 0.1551,
      "step": 1760
    },
    {
      "epoch": 0.8286516853932584,
      "grad_norm": 0.24001863598823547,
      "learning_rate": 4.34733203269588e-05,
      "loss": 0.156,
      "step": 1770
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.22871051728725433,
      "learning_rate": 4.3310739991155365e-05,
      "loss": 0.154,
      "step": 1780
    },
    {
      "epoch": 0.8380149812734082,
      "grad_norm": 0.2900903522968292,
      "learning_rate": 4.314647164263247e-05,
      "loss": 0.1618,
      "step": 1790
    },
    {
      "epoch": 0.8426966292134831,
      "grad_norm": 0.25664201378822327,
      "learning_rate": 4.2980530424801146e-05,
      "loss": 0.1529,
      "step": 1800
    },
    {
      "epoch": 0.8473782771535581,
      "grad_norm": 0.2573324143886566,
      "learning_rate": 4.281293163528925e-05,
      "loss": 0.1531,
      "step": 1810
    },
    {
      "epoch": 0.8520599250936329,
      "grad_norm": 0.2515137493610382,
      "learning_rate": 4.264369072453126e-05,
      "loss": 0.176,
      "step": 1820
    },
    {
      "epoch": 0.8567415730337079,
      "grad_norm": 0.2710267901420593,
      "learning_rate": 4.247282329434394e-05,
      "loss": 0.1477,
      "step": 1830
    },
    {
      "epoch": 0.8614232209737828,
      "grad_norm": 0.2349739819765091,
      "learning_rate": 4.230034509648804e-05,
      "loss": 0.1583,
      "step": 1840
    },
    {
      "epoch": 0.8661048689138576,
      "grad_norm": 0.23130016028881073,
      "learning_rate": 4.212627203121617e-05,
      "loss": 0.1485,
      "step": 1850
    },
    {
      "epoch": 0.8707865168539326,
      "grad_norm": 0.27001139521598816,
      "learning_rate": 4.19506201458071e-05,
      "loss": 0.1498,
      "step": 1860
    },
    {
      "epoch": 0.8754681647940075,
      "grad_norm": 0.25574082136154175,
      "learning_rate": 4.177340563308625e-05,
      "loss": 0.1697,
      "step": 1870
    },
    {
      "epoch": 0.8801498127340824,
      "grad_norm": 0.2866358458995819,
      "learning_rate": 4.159464482993308e-05,
      "loss": 0.1584,
      "step": 1880
    },
    {
      "epoch": 0.8848314606741573,
      "grad_norm": 0.23652639985084534,
      "learning_rate": 4.141435421577493e-05,
      "loss": 0.1507,
      "step": 1890
    },
    {
      "epoch": 0.8895131086142322,
      "grad_norm": 0.2574213445186615,
      "learning_rate": 4.123255041106788e-05,
      "loss": 0.1612,
      "step": 1900
    },
    {
      "epoch": 0.8941947565543071,
      "grad_norm": 0.25920340418815613,
      "learning_rate": 4.104925017576458e-05,
      "loss": 0.1483,
      "step": 1910
    },
    {
      "epoch": 0.898876404494382,
      "grad_norm": 0.26211079955101013,
      "learning_rate": 4.0864470407769114e-05,
      "loss": 0.1652,
      "step": 1920
    },
    {
      "epoch": 0.903558052434457,
      "grad_norm": 0.27797001600265503,
      "learning_rate": 4.0678228141379336e-05,
      "loss": 0.1461,
      "step": 1930
    },
    {
      "epoch": 0.9082397003745318,
      "grad_norm": 0.2917916774749756,
      "learning_rate": 4.049054054571648e-05,
      "loss": 0.1581,
      "step": 1940
    },
    {
      "epoch": 0.9129213483146067,
      "grad_norm": 0.2321026623249054,
      "learning_rate": 4.030142492314236e-05,
      "loss": 0.1754,
      "step": 1950
    },
    {
      "epoch": 0.9176029962546817,
      "grad_norm": 0.22449631989002228,
      "learning_rate": 4.011089870766437e-05,
      "loss": 0.1423,
      "step": 1960
    },
    {
      "epoch": 0.9222846441947565,
      "grad_norm": 0.2413541078567505,
      "learning_rate": 3.991897946332829e-05,
      "loss": 0.1619,
      "step": 1970
    },
    {
      "epoch": 0.9269662921348315,
      "grad_norm": 0.23216186463832855,
      "learning_rate": 3.972568488259905e-05,
      "loss": 0.1549,
      "step": 1980
    },
    {
      "epoch": 0.9316479400749064,
      "grad_norm": 0.2707996070384979,
      "learning_rate": 3.953103278472979e-05,
      "loss": 0.1622,
      "step": 1990
    },
    {
      "epoch": 0.9363295880149812,
      "grad_norm": 0.25455233454704285,
      "learning_rate": 3.93350411141191e-05,
      "loss": 0.1572,
      "step": 2000
    },
    {
      "epoch": 0.9363295880149812,
      "eval_loss": 0.15146031975746155,
      "eval_runtime": 232.8082,
      "eval_samples_per_second": 2.998,
      "eval_steps_per_second": 1.499,
      "step": 2000
    },
    {
      "epoch": 0.9410112359550562,
      "grad_norm": 0.25234970450401306,
      "learning_rate": 3.913772793865683e-05,
      "loss": 0.1552,
      "step": 2010
    },
    {
      "epoch": 0.9456928838951311,
      "grad_norm": 0.2573593556880951,
      "learning_rate": 3.8939111448058404e-05,
      "loss": 0.163,
      "step": 2020
    },
    {
      "epoch": 0.950374531835206,
      "grad_norm": 0.2533682584762573,
      "learning_rate": 3.8739209952188016e-05,
      "loss": 0.1647,
      "step": 2030
    },
    {
      "epoch": 0.9550561797752809,
      "grad_norm": 0.2530724108219147,
      "learning_rate": 3.853804187937066e-05,
      "loss": 0.1455,
      "step": 2040
    },
    {
      "epoch": 0.9597378277153558,
      "grad_norm": 0.2659183144569397,
      "learning_rate": 3.8335625774693326e-05,
      "loss": 0.1556,
      "step": 2050
    },
    {
      "epoch": 0.9644194756554307,
      "grad_norm": 0.2551698386669159,
      "learning_rate": 3.813198029829532e-05,
      "loss": 0.1609,
      "step": 2060
    },
    {
      "epoch": 0.9691011235955056,
      "grad_norm": 0.24625395238399506,
      "learning_rate": 3.792712422364812e-05,
      "loss": 0.1491,
      "step": 2070
    },
    {
      "epoch": 0.9737827715355806,
      "grad_norm": 0.28316351771354675,
      "learning_rate": 3.772107643582459e-05,
      "loss": 0.1549,
      "step": 2080
    },
    {
      "epoch": 0.9784644194756554,
      "grad_norm": 0.2540896236896515,
      "learning_rate": 3.7513855929758146e-05,
      "loss": 0.1497,
      "step": 2090
    },
    {
      "epoch": 0.9831460674157303,
      "grad_norm": 0.29905152320861816,
      "learning_rate": 3.730548180849161e-05,
      "loss": 0.162,
      "step": 2100
    },
    {
      "epoch": 0.9878277153558053,
      "grad_norm": 0.26653796434402466,
      "learning_rate": 3.7095973281416144e-05,
      "loss": 0.1606,
      "step": 2110
    },
    {
      "epoch": 0.9925093632958801,
      "grad_norm": 0.22521516680717468,
      "learning_rate": 3.688534966250042e-05,
      "loss": 0.1494,
      "step": 2120
    },
    {
      "epoch": 0.9971910112359551,
      "grad_norm": 0.2566962242126465,
      "learning_rate": 3.667363036851012e-05,
      "loss": 0.1512,
      "step": 2130
    },
    {
      "epoch": 1.0018726591760299,
      "grad_norm": 0.25175896286964417,
      "learning_rate": 3.646083491721794e-05,
      "loss": 0.14,
      "step": 2140
    },
    {
      "epoch": 1.006554307116105,
      "grad_norm": 0.24953345954418182,
      "learning_rate": 3.624698292560434e-05,
      "loss": 0.1563,
      "step": 2150
    },
    {
      "epoch": 1.0112359550561798,
      "grad_norm": 0.26830434799194336,
      "learning_rate": 3.603209410804906e-05,
      "loss": 0.1476,
      "step": 2160
    },
    {
      "epoch": 1.0159176029962547,
      "grad_norm": 0.23503483831882477,
      "learning_rate": 3.5816188274513795e-05,
      "loss": 0.1487,
      "step": 2170
    },
    {
      "epoch": 1.0205992509363295,
      "grad_norm": 0.24020551145076752,
      "learning_rate": 3.559928532871587e-05,
      "loss": 0.1464,
      "step": 2180
    },
    {
      "epoch": 1.0252808988764044,
      "grad_norm": 0.2504616975784302,
      "learning_rate": 3.5381405266293464e-05,
      "loss": 0.152,
      "step": 2190
    },
    {
      "epoch": 1.0299625468164795,
      "grad_norm": 0.2568669319152832,
      "learning_rate": 3.516256817296222e-05,
      "loss": 0.1534,
      "step": 2200
    },
    {
      "epoch": 1.0346441947565543,
      "grad_norm": 0.2495841681957245,
      "learning_rate": 3.4942794222663614e-05,
      "loss": 0.148,
      "step": 2210
    },
    {
      "epoch": 1.0393258426966292,
      "grad_norm": 0.2616524398326874,
      "learning_rate": 3.472210367570518e-05,
      "loss": 0.1498,
      "step": 2220
    },
    {
      "epoch": 1.044007490636704,
      "grad_norm": 0.2961607277393341,
      "learning_rate": 3.4500516876892786e-05,
      "loss": 0.154,
      "step": 2230
    },
    {
      "epoch": 1.048689138576779,
      "grad_norm": 0.26527124643325806,
      "learning_rate": 3.427805425365509e-05,
      "loss": 0.1471,
      "step": 2240
    },
    {
      "epoch": 1.053370786516854,
      "grad_norm": 0.2571045458316803,
      "learning_rate": 3.405473631416041e-05,
      "loss": 0.15,
      "step": 2250
    },
    {
      "epoch": 1.0580524344569289,
      "grad_norm": 0.272733598947525,
      "learning_rate": 3.383058364542611e-05,
      "loss": 0.1503,
      "step": 2260
    },
    {
      "epoch": 1.0627340823970037,
      "grad_norm": 0.24832037091255188,
      "learning_rate": 3.3605616911420786e-05,
      "loss": 0.1462,
      "step": 2270
    },
    {
      "epoch": 1.0674157303370786,
      "grad_norm": 0.2702969014644623,
      "learning_rate": 3.3379856851159267e-05,
      "loss": 0.1433,
      "step": 2280
    },
    {
      "epoch": 1.0720973782771535,
      "grad_norm": 0.2763786315917969,
      "learning_rate": 3.315332427679079e-05,
      "loss": 0.1425,
      "step": 2290
    },
    {
      "epoch": 1.0767790262172285,
      "grad_norm": 0.3010675609111786,
      "learning_rate": 3.292604007168037e-05,
      "loss": 0.1525,
      "step": 2300
    },
    {
      "epoch": 1.0814606741573034,
      "grad_norm": 0.25726327300071716,
      "learning_rate": 3.269802518848364e-05,
      "loss": 0.1501,
      "step": 2310
    },
    {
      "epoch": 1.0861423220973783,
      "grad_norm": 0.24328449368476868,
      "learning_rate": 3.246930064721524e-05,
      "loss": 0.1352,
      "step": 2320
    },
    {
      "epoch": 1.0908239700374531,
      "grad_norm": 0.2798154056072235,
      "learning_rate": 3.22398875333111e-05,
      "loss": 0.1492,
      "step": 2330
    },
    {
      "epoch": 1.095505617977528,
      "grad_norm": 0.26386189460754395,
      "learning_rate": 3.200980699568463e-05,
      "loss": 0.1472,
      "step": 2340
    },
    {
      "epoch": 1.100187265917603,
      "grad_norm": 0.2722947895526886,
      "learning_rate": 3.177908024477703e-05,
      "loss": 0.1557,
      "step": 2350
    },
    {
      "epoch": 1.104868913857678,
      "grad_norm": 0.2719181180000305,
      "learning_rate": 3.154772855060198e-05,
      "loss": 0.1408,
      "step": 2360
    },
    {
      "epoch": 1.1095505617977528,
      "grad_norm": 0.27931663393974304,
      "learning_rate": 3.131577324078485e-05,
      "loss": 0.1485,
      "step": 2370
    },
    {
      "epoch": 1.1142322097378277,
      "grad_norm": 0.2354390025138855,
      "learning_rate": 3.1083235698596505e-05,
      "loss": 0.1402,
      "step": 2380
    },
    {
      "epoch": 1.1189138576779025,
      "grad_norm": 0.25632205605506897,
      "learning_rate": 3.0850137360982104e-05,
      "loss": 0.1428,
      "step": 2390
    },
    {
      "epoch": 1.1235955056179776,
      "grad_norm": 0.26264774799346924,
      "learning_rate": 3.061649971658488e-05,
      "loss": 0.1465,
      "step": 2400
    },
    {
      "epoch": 1.1282771535580525,
      "grad_norm": 0.2421567738056183,
      "learning_rate": 3.038234430376513e-05,
      "loss": 0.1495,
      "step": 2410
    },
    {
      "epoch": 1.1329588014981273,
      "grad_norm": 0.2859581410884857,
      "learning_rate": 3.01476927086147e-05,
      "loss": 0.1494,
      "step": 2420
    },
    {
      "epoch": 1.1376404494382022,
      "grad_norm": 0.24564145505428314,
      "learning_rate": 2.9912566562967014e-05,
      "loss": 0.15,
      "step": 2430
    },
    {
      "epoch": 1.142322097378277,
      "grad_norm": 0.23918387293815613,
      "learning_rate": 2.967698754240289e-05,
      "loss": 0.146,
      "step": 2440
    },
    {
      "epoch": 1.1470037453183521,
      "grad_norm": 0.2613247334957123,
      "learning_rate": 2.944097736425233e-05,
      "loss": 0.1549,
      "step": 2450
    },
    {
      "epoch": 1.151685393258427,
      "grad_norm": 0.3330237865447998,
      "learning_rate": 2.92045577855925e-05,
      "loss": 0.1526,
      "step": 2460
    },
    {
      "epoch": 1.1563670411985019,
      "grad_norm": 0.24248412251472473,
      "learning_rate": 2.896775060124195e-05,
      "loss": 0.1537,
      "step": 2470
    },
    {
      "epoch": 1.1610486891385767,
      "grad_norm": 0.2552735507488251,
      "learning_rate": 2.8730577641751476e-05,
      "loss": 0.1458,
      "step": 2480
    },
    {
      "epoch": 1.1657303370786516,
      "grad_norm": 0.24274031817913055,
      "learning_rate": 2.8493060771391567e-05,
      "loss": 0.1477,
      "step": 2490
    },
    {
      "epoch": 1.1704119850187267,
      "grad_norm": 0.251447468996048,
      "learning_rate": 2.825522188613686e-05,
      "loss": 0.142,
      "step": 2500
    },
    {
      "epoch": 1.1704119850187267,
      "eval_loss": 0.14719446003437042,
      "eval_runtime": 235.8012,
      "eval_samples_per_second": 2.96,
      "eval_steps_per_second": 1.48,
      "step": 2500
    },
    {
      "epoch": 1.1750936329588015,
      "grad_norm": 0.31371602416038513,
      "learning_rate": 2.8017082911647586e-05,
      "loss": 0.1435,
      "step": 2510
    },
    {
      "epoch": 1.1797752808988764,
      "grad_norm": 0.25653213262557983,
      "learning_rate": 2.7778665801248292e-05,
      "loss": 0.1445,
      "step": 2520
    },
    {
      "epoch": 1.1844569288389513,
      "grad_norm": 0.26454073190689087,
      "learning_rate": 2.7539992533904062e-05,
      "loss": 0.1472,
      "step": 2530
    },
    {
      "epoch": 1.1891385767790261,
      "grad_norm": 0.2490745633840561,
      "learning_rate": 2.730108511219433e-05,
      "loss": 0.1728,
      "step": 2540
    },
    {
      "epoch": 1.1938202247191012,
      "grad_norm": 0.2763822674751282,
      "learning_rate": 2.7061965560284512e-05,
      "loss": 0.1466,
      "step": 2550
    },
    {
      "epoch": 1.198501872659176,
      "grad_norm": 0.26875513792037964,
      "learning_rate": 2.6822655921895695e-05,
      "loss": 0.1548,
      "step": 2560
    },
    {
      "epoch": 1.203183520599251,
      "grad_norm": 0.2581767737865448,
      "learning_rate": 2.6583178258272468e-05,
      "loss": 0.1508,
      "step": 2570
    },
    {
      "epoch": 1.2078651685393258,
      "grad_norm": 0.24231292307376862,
      "learning_rate": 2.6343554646149154e-05,
      "loss": 0.1523,
      "step": 2580
    },
    {
      "epoch": 1.2125468164794007,
      "grad_norm": 0.2455371618270874,
      "learning_rate": 2.6103807175714645e-05,
      "loss": 0.1417,
      "step": 2590
    },
    {
      "epoch": 1.2172284644194757,
      "grad_norm": 0.21632102131843567,
      "learning_rate": 2.586395794857597e-05,
      "loss": 0.1471,
      "step": 2600
    },
    {
      "epoch": 1.2219101123595506,
      "grad_norm": 0.24456676840782166,
      "learning_rate": 2.5624029075720785e-05,
      "loss": 0.1411,
      "step": 2610
    },
    {
      "epoch": 1.2265917602996255,
      "grad_norm": 0.25774601101875305,
      "learning_rate": 2.538404267547908e-05,
      "loss": 0.1471,
      "step": 2620
    },
    {
      "epoch": 1.2312734082397003,
      "grad_norm": 0.26534947752952576,
      "learning_rate": 2.5144020871484088e-05,
      "loss": 0.1538,
      "step": 2630
    },
    {
      "epoch": 1.2359550561797752,
      "grad_norm": 0.2735100984573364,
      "learning_rate": 2.4903985790632835e-05,
      "loss": 0.1492,
      "step": 2640
    },
    {
      "epoch": 1.2406367041198503,
      "grad_norm": 0.2741684317588806,
      "learning_rate": 2.46639595610463e-05,
      "loss": 0.1464,
      "step": 2650
    },
    {
      "epoch": 1.2453183520599251,
      "grad_norm": 0.233806774020195,
      "learning_rate": 2.4423964310029458e-05,
      "loss": 0.1404,
      "step": 2660
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.25946834683418274,
      "learning_rate": 2.4184022162031502e-05,
      "loss": 0.1474,
      "step": 2670
    },
    {
      "epoch": 1.2546816479400749,
      "grad_norm": 0.2744329869747162,
      "learning_rate": 2.3944155236606196e-05,
      "loss": 0.1525,
      "step": 2680
    },
    {
      "epoch": 1.2593632958801497,
      "grad_norm": 0.27152568101882935,
      "learning_rate": 2.3704385646372772e-05,
      "loss": 0.1454,
      "step": 2690
    },
    {
      "epoch": 1.2640449438202248,
      "grad_norm": 0.23862288892269135,
      "learning_rate": 2.3464735494977392e-05,
      "loss": 0.1311,
      "step": 2700
    },
    {
      "epoch": 1.2687265917602997,
      "grad_norm": 0.27397608757019043,
      "learning_rate": 2.3225226875055535e-05,
      "loss": 0.1416,
      "step": 2710
    },
    {
      "epoch": 1.2734082397003745,
      "grad_norm": 0.2350151538848877,
      "learning_rate": 2.2985881866195307e-05,
      "loss": 0.1386,
      "step": 2720
    },
    {
      "epoch": 1.2780898876404494,
      "grad_norm": 0.26997503638267517,
      "learning_rate": 2.274672253290198e-05,
      "loss": 0.1412,
      "step": 2730
    },
    {
      "epoch": 1.2827715355805243,
      "grad_norm": 0.2512725591659546,
      "learning_rate": 2.2507770922563966e-05,
      "loss": 0.1583,
      "step": 2740
    },
    {
      "epoch": 1.2874531835205993,
      "grad_norm": 0.27003416419029236,
      "learning_rate": 2.2269049063420305e-05,
      "loss": 0.1553,
      "step": 2750
    },
    {
      "epoch": 1.2921348314606742,
      "grad_norm": 0.25606778264045715,
      "learning_rate": 2.2030578962529964e-05,
      "loss": 0.1358,
      "step": 2760
    },
    {
      "epoch": 1.296816479400749,
      "grad_norm": 0.2931346297264099,
      "learning_rate": 2.179238260374307e-05,
      "loss": 0.1461,
      "step": 2770
    },
    {
      "epoch": 1.301498127340824,
      "grad_norm": 0.2787070572376251,
      "learning_rate": 2.155448194567425e-05,
      "loss": 0.1412,
      "step": 2780
    },
    {
      "epoch": 1.3061797752808988,
      "grad_norm": 0.2558610737323761,
      "learning_rate": 2.131689891967839e-05,
      "loss": 0.1447,
      "step": 2790
    },
    {
      "epoch": 1.3108614232209739,
      "grad_norm": 0.27285927534103394,
      "learning_rate": 2.1079655427828807e-05,
      "loss": 0.1315,
      "step": 2800
    },
    {
      "epoch": 1.3155430711610487,
      "grad_norm": 0.2873709797859192,
      "learning_rate": 2.0842773340898187e-05,
      "loss": 0.1404,
      "step": 2810
    },
    {
      "epoch": 1.3202247191011236,
      "grad_norm": 0.2794095277786255,
      "learning_rate": 2.060627449634234e-05,
      "loss": 0.1473,
      "step": 2820
    },
    {
      "epoch": 1.3249063670411985,
      "grad_norm": 0.2690732777118683,
      "learning_rate": 2.0370180696287123e-05,
      "loss": 0.1446,
      "step": 2830
    },
    {
      "epoch": 1.3295880149812733,
      "grad_norm": 0.28744590282440186,
      "learning_rate": 2.0134513705518547e-05,
      "loss": 0.145,
      "step": 2840
    },
    {
      "epoch": 1.3342696629213484,
      "grad_norm": 0.2891230285167694,
      "learning_rate": 1.9899295249476337e-05,
      "loss": 0.1437,
      "step": 2850
    },
    {
      "epoch": 1.3389513108614233,
      "grad_norm": 0.28290775418281555,
      "learning_rate": 1.9664547012251122e-05,
      "loss": 0.1425,
      "step": 2860
    },
    {
      "epoch": 1.3436329588014981,
      "grad_norm": 0.2865421772003174,
      "learning_rate": 1.9430290634585475e-05,
      "loss": 0.1501,
      "step": 2870
    },
    {
      "epoch": 1.348314606741573,
      "grad_norm": 0.27473053336143494,
      "learning_rate": 1.9196547711878883e-05,
      "loss": 0.1486,
      "step": 2880
    },
    {
      "epoch": 1.3529962546816479,
      "grad_norm": 0.2682128846645355,
      "learning_rate": 1.8963339792196968e-05,
      "loss": 0.1429,
      "step": 2890
    },
    {
      "epoch": 1.357677902621723,
      "grad_norm": 0.2450461983680725,
      "learning_rate": 1.873068837428497e-05,
      "loss": 0.1539,
      "step": 2900
    },
    {
      "epoch": 1.3623595505617978,
      "grad_norm": 0.25303879380226135,
      "learning_rate": 1.8498614905585924e-05,
      "loss": 0.1465,
      "step": 2910
    },
    {
      "epoch": 1.3670411985018727,
      "grad_norm": 0.27826565504074097,
      "learning_rate": 1.8267140780263427e-05,
      "loss": 0.1455,
      "step": 2920
    },
    {
      "epoch": 1.3717228464419475,
      "grad_norm": 0.20451192557811737,
      "learning_rate": 1.8036287337229407e-05,
      "loss": 0.1572,
      "step": 2930
    },
    {
      "epoch": 1.3764044943820224,
      "grad_norm": 0.2710937261581421,
      "learning_rate": 1.7806075858176903e-05,
      "loss": 0.1427,
      "step": 2940
    },
    {
      "epoch": 1.3810861423220975,
      "grad_norm": 0.2961539626121521,
      "learning_rate": 1.757652756561822e-05,
      "loss": 0.1481,
      "step": 2950
    },
    {
      "epoch": 1.3857677902621723,
      "grad_norm": 0.2685253620147705,
      "learning_rate": 1.7347663620928495e-05,
      "loss": 0.1488,
      "step": 2960
    },
    {
      "epoch": 1.3904494382022472,
      "grad_norm": 0.29267454147338867,
      "learning_rate": 1.711950512239483e-05,
      "loss": 0.1441,
      "step": 2970
    },
    {
      "epoch": 1.395131086142322,
      "grad_norm": 0.25182071328163147,
      "learning_rate": 1.6892073103271355e-05,
      "loss": 0.1422,
      "step": 2980
    },
    {
      "epoch": 1.399812734082397,
      "grad_norm": 0.23954342305660248,
      "learning_rate": 1.6665388529840225e-05,
      "loss": 0.1465,
      "step": 2990
    },
    {
      "epoch": 1.404494382022472,
      "grad_norm": 0.2718564569950104,
      "learning_rate": 1.6439472299478803e-05,
      "loss": 0.1376,
      "step": 3000
    },
    {
      "epoch": 1.404494382022472,
      "eval_loss": 0.14412875473499298,
      "eval_runtime": 230.7877,
      "eval_samples_per_second": 3.024,
      "eval_steps_per_second": 1.512,
      "step": 3000
    },
    {
      "epoch": 1.4091760299625469,
      "grad_norm": 0.2633533477783203,
      "learning_rate": 1.621434523873318e-05,
      "loss": 0.1351,
      "step": 3010
    },
    {
      "epoch": 1.4138576779026217,
      "grad_norm": 0.26720061898231506,
      "learning_rate": 1.5990028101398234e-05,
      "loss": 0.1382,
      "step": 3020
    },
    {
      "epoch": 1.4185393258426966,
      "grad_norm": 0.26572489738464355,
      "learning_rate": 1.576654156660442e-05,
      "loss": 0.1462,
      "step": 3030
    },
    {
      "epoch": 1.4232209737827715,
      "grad_norm": 0.29276156425476074,
      "learning_rate": 1.5543906236911424e-05,
      "loss": 0.1518,
      "step": 3040
    },
    {
      "epoch": 1.4279026217228465,
      "grad_norm": 0.25354310870170593,
      "learning_rate": 1.5322142636408855e-05,
      "loss": 0.1727,
      "step": 3050
    },
    {
      "epoch": 1.4325842696629214,
      "grad_norm": 0.2621780037879944,
      "learning_rate": 1.5101271208824168e-05,
      "loss": 0.1414,
      "step": 3060
    },
    {
      "epoch": 1.4372659176029963,
      "grad_norm": 0.275648295879364,
      "learning_rate": 1.4881312315638097e-05,
      "loss": 0.1387,
      "step": 3070
    },
    {
      "epoch": 1.4419475655430711,
      "grad_norm": 0.25782111287117004,
      "learning_rate": 1.4662286234207512e-05,
      "loss": 0.1471,
      "step": 3080
    },
    {
      "epoch": 1.446629213483146,
      "grad_norm": 0.25094345211982727,
      "learning_rate": 1.4444213155896147e-05,
      "loss": 0.1523,
      "step": 3090
    },
    {
      "epoch": 1.451310861423221,
      "grad_norm": 0.2622791528701782,
      "learning_rate": 1.4227113184213198e-05,
      "loss": 0.1413,
      "step": 3100
    },
    {
      "epoch": 1.455992509363296,
      "grad_norm": 0.3153008222579956,
      "learning_rate": 1.4011006332960064e-05,
      "loss": 0.1458,
      "step": 3110
    },
    {
      "epoch": 1.4606741573033708,
      "grad_norm": 0.247295081615448,
      "learning_rate": 1.3795912524385323e-05,
      "loss": 0.1514,
      "step": 3120
    },
    {
      "epoch": 1.4653558052434457,
      "grad_norm": 0.266445517539978,
      "learning_rate": 1.358185158734816e-05,
      "loss": 0.1443,
      "step": 3130
    },
    {
      "epoch": 1.4700374531835205,
      "grad_norm": 0.28155022859573364,
      "learning_rate": 1.3368843255490383e-05,
      "loss": 0.1431,
      "step": 3140
    },
    {
      "epoch": 1.4747191011235956,
      "grad_norm": 0.25658926367759705,
      "learning_rate": 1.3156907165417271e-05,
      "loss": 0.1489,
      "step": 3150
    },
    {
      "epoch": 1.4794007490636705,
      "grad_norm": 0.29060763120651245,
      "learning_rate": 1.2946062854887313e-05,
      "loss": 0.161,
      "step": 3160
    },
    {
      "epoch": 1.4840823970037453,
      "grad_norm": 0.2649938464164734,
      "learning_rate": 1.273632976101111e-05,
      "loss": 0.1429,
      "step": 3170
    },
    {
      "epoch": 1.4887640449438202,
      "grad_norm": 0.28675931692123413,
      "learning_rate": 1.252772721845945e-05,
      "loss": 0.1519,
      "step": 3180
    },
    {
      "epoch": 1.493445692883895,
      "grad_norm": 0.2792210876941681,
      "learning_rate": 1.2320274457680997e-05,
      "loss": 0.1464,
      "step": 3190
    },
    {
      "epoch": 1.4981273408239701,
      "grad_norm": 0.29543542861938477,
      "learning_rate": 1.2113990603129433e-05,
      "loss": 0.1463,
      "step": 3200
    },
    {
      "epoch": 1.5028089887640448,
      "grad_norm": 0.2720608711242676,
      "learning_rate": 1.1908894671500459e-05,
      "loss": 0.1397,
      "step": 3210
    },
    {
      "epoch": 1.5074906367041199,
      "grad_norm": 0.2443593591451645,
      "learning_rate": 1.1705005569978664e-05,
      "loss": 0.1385,
      "step": 3220
    },
    {
      "epoch": 1.5121722846441947,
      "grad_norm": 0.26821592450141907,
      "learning_rate": 1.1502342094494584e-05,
      "loss": 0.1327,
      "step": 3230
    },
    {
      "epoch": 1.5168539325842696,
      "grad_norm": 0.2925826609134674,
      "learning_rate": 1.1300922927991913e-05,
      "loss": 0.143,
      "step": 3240
    },
    {
      "epoch": 1.5215355805243447,
      "grad_norm": 0.3027312457561493,
      "learning_rate": 1.1100766638705224e-05,
      "loss": 0.1481,
      "step": 3250
    },
    {
      "epoch": 1.5262172284644193,
      "grad_norm": 0.2656176686286926,
      "learning_rate": 1.0901891678448144e-05,
      "loss": 0.1357,
      "step": 3260
    },
    {
      "epoch": 1.5308988764044944,
      "grad_norm": 0.29944488406181335,
      "learning_rate": 1.0704316380912425e-05,
      "loss": 0.1478,
      "step": 3270
    },
    {
      "epoch": 1.5355805243445693,
      "grad_norm": 0.2791348695755005,
      "learning_rate": 1.0508058959977757e-05,
      "loss": 0.1482,
      "step": 3280
    },
    {
      "epoch": 1.5402621722846441,
      "grad_norm": 0.29027488827705383,
      "learning_rate": 1.0313137508032719e-05,
      "loss": 0.1389,
      "step": 3290
    },
    {
      "epoch": 1.5449438202247192,
      "grad_norm": 0.29706457257270813,
      "learning_rate": 1.0119569994306841e-05,
      "loss": 0.1531,
      "step": 3300
    },
    {
      "epoch": 1.5496254681647939,
      "grad_norm": 0.27216818928718567,
      "learning_rate": 9.927374263214143e-06,
      "loss": 0.146,
      "step": 3310
    },
    {
      "epoch": 1.554307116104869,
      "grad_norm": 0.258273184299469,
      "learning_rate": 9.736568032708069e-06,
      "loss": 0.1441,
      "step": 3320
    },
    {
      "epoch": 1.5589887640449438,
      "grad_norm": 0.2688770592212677,
      "learning_rate": 9.547168892648136e-06,
      "loss": 0.1551,
      "step": 3330
    },
    {
      "epoch": 1.5636704119850187,
      "grad_norm": 0.24575349688529968,
      "learning_rate": 9.359194303178371e-06,
      "loss": 0.1409,
      "step": 3340
    },
    {
      "epoch": 1.5683520599250937,
      "grad_norm": 0.2606309950351715,
      "learning_rate": 9.172661593117712e-06,
      "loss": 0.1535,
      "step": 3350
    },
    {
      "epoch": 1.5730337078651684,
      "grad_norm": 0.2664453387260437,
      "learning_rate": 8.987587958362517e-06,
      "loss": 0.1403,
      "step": 3360
    },
    {
      "epoch": 1.5777153558052435,
      "grad_norm": 0.24338924884796143,
      "learning_rate": 8.803990460301343e-06,
      "loss": 0.1454,
      "step": 3370
    },
    {
      "epoch": 1.5823970037453183,
      "grad_norm": 0.2727632522583008,
      "learning_rate": 8.621886024242057e-06,
      "loss": 0.1381,
      "step": 3380
    },
    {
      "epoch": 1.5870786516853932,
      "grad_norm": 0.27160879969596863,
      "learning_rate": 8.441291437851605e-06,
      "loss": 0.1386,
      "step": 3390
    },
    {
      "epoch": 1.5917602996254683,
      "grad_norm": 0.24165265262126923,
      "learning_rate": 8.262223349608366e-06,
      "loss": 0.1391,
      "step": 3400
    },
    {
      "epoch": 1.596441947565543,
      "grad_norm": 0.22898262739181519,
      "learning_rate": 8.084698267267408e-06,
      "loss": 0.1705,
      "step": 3410
    },
    {
      "epoch": 1.601123595505618,
      "grad_norm": 0.26720166206359863,
      "learning_rate": 7.908732556338628e-06,
      "loss": 0.1427,
      "step": 3420
    },
    {
      "epoch": 1.6058052434456929,
      "grad_norm": 0.2854466140270233,
      "learning_rate": 7.73434243857814e-06,
      "loss": 0.1405,
      "step": 3430
    },
    {
      "epoch": 1.6104868913857677,
      "grad_norm": 0.29229024052619934,
      "learning_rate": 7.561543990492803e-06,
      "loss": 0.1445,
      "step": 3440
    },
    {
      "epoch": 1.6151685393258428,
      "grad_norm": 0.2715863883495331,
      "learning_rate": 7.390353141858172e-06,
      "loss": 0.1363,
      "step": 3450
    },
    {
      "epoch": 1.6198501872659175,
      "grad_norm": 0.24178357422351837,
      "learning_rate": 7.2207856742499695e-06,
      "loss": 0.144,
      "step": 3460
    },
    {
      "epoch": 1.6245318352059925,
      "grad_norm": 0.25843772292137146,
      "learning_rate": 7.052857219589262e-06,
      "loss": 0.1458,
      "step": 3470
    },
    {
      "epoch": 1.6292134831460674,
      "grad_norm": 0.2684144079685211,
      "learning_rate": 6.886583258701382e-06,
      "loss": 0.1337,
      "step": 3480
    },
    {
      "epoch": 1.6338951310861423,
      "grad_norm": 0.2679165303707123,
      "learning_rate": 6.721979119888796e-06,
      "loss": 0.1814,
      "step": 3490
    },
    {
      "epoch": 1.6385767790262173,
      "grad_norm": 0.2938106954097748,
      "learning_rate": 6.559059977518017e-06,
      "loss": 0.1569,
      "step": 3500
    },
    {
      "epoch": 1.6385767790262173,
      "eval_loss": 0.14246584475040436,
      "eval_runtime": 232.5612,
      "eval_samples_per_second": 3.001,
      "eval_steps_per_second": 1.501,
      "step": 3500
    },
    {
      "epoch": 1.643258426966292,
      "grad_norm": 0.2740805447101593,
      "learning_rate": 6.397840850620762e-06,
      "loss": 0.1476,
      "step": 3510
    },
    {
      "epoch": 1.647940074906367,
      "grad_norm": 0.260959655046463,
      "learning_rate": 6.238336601509365e-06,
      "loss": 0.1472,
      "step": 3520
    },
    {
      "epoch": 1.652621722846442,
      "grad_norm": 0.2726864814758301,
      "learning_rate": 6.080561934406678e-06,
      "loss": 0.1474,
      "step": 3530
    },
    {
      "epoch": 1.6573033707865168,
      "grad_norm": 0.2838584780693054,
      "learning_rate": 5.92453139409051e-06,
      "loss": 0.136,
      "step": 3540
    },
    {
      "epoch": 1.6619850187265919,
      "grad_norm": 0.2548011243343353,
      "learning_rate": 5.770259364552816e-06,
      "loss": 0.1435,
      "step": 3550
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.25916817784309387,
      "learning_rate": 5.617760067673666e-06,
      "loss": 0.1413,
      "step": 3560
    },
    {
      "epoch": 1.6713483146067416,
      "grad_norm": 0.25346603989601135,
      "learning_rate": 5.46704756191018e-06,
      "loss": 0.1359,
      "step": 3570
    },
    {
      "epoch": 1.6760299625468165,
      "grad_norm": 0.2750667333602905,
      "learning_rate": 5.318135741000488e-06,
      "loss": 0.1542,
      "step": 3580
    },
    {
      "epoch": 1.6807116104868913,
      "grad_norm": 0.28603386878967285,
      "learning_rate": 5.171038332682959e-06,
      "loss": 0.1283,
      "step": 3590
    },
    {
      "epoch": 1.6853932584269664,
      "grad_norm": 0.2523970305919647,
      "learning_rate": 5.025768897430644e-06,
      "loss": 0.1376,
      "step": 3600
    },
    {
      "epoch": 1.690074906367041,
      "grad_norm": 0.27870383858680725,
      "learning_rate": 4.882340827201193e-06,
      "loss": 0.139,
      "step": 3610
    },
    {
      "epoch": 1.6947565543071161,
      "grad_norm": 0.2658084034919739,
      "learning_rate": 4.740767344202282e-06,
      "loss": 0.1457,
      "step": 3620
    },
    {
      "epoch": 1.699438202247191,
      "grad_norm": 0.2911655008792877,
      "learning_rate": 4.601061499672698e-06,
      "loss": 0.1353,
      "step": 3630
    },
    {
      "epoch": 1.7041198501872659,
      "grad_norm": 0.270166277885437,
      "learning_rate": 4.463236172679192e-06,
      "loss": 0.1525,
      "step": 3640
    },
    {
      "epoch": 1.708801498127341,
      "grad_norm": 0.2789149582386017,
      "learning_rate": 4.327304068929192e-06,
      "loss": 0.1604,
      "step": 3650
    },
    {
      "epoch": 1.7134831460674156,
      "grad_norm": 0.2603972554206848,
      "learning_rate": 4.193277719599484e-06,
      "loss": 0.1482,
      "step": 3660
    },
    {
      "epoch": 1.7181647940074907,
      "grad_norm": 0.28370609879493713,
      "learning_rate": 4.061169480181029e-06,
      "loss": 0.1326,
      "step": 3670
    },
    {
      "epoch": 1.7228464419475655,
      "grad_norm": 0.28793907165527344,
      "learning_rate": 3.9309915293399366e-06,
      "loss": 0.1483,
      "step": 3680
    },
    {
      "epoch": 1.7275280898876404,
      "grad_norm": 0.2687625288963318,
      "learning_rate": 3.802755867794744e-06,
      "loss": 0.1343,
      "step": 3690
    },
    {
      "epoch": 1.7322097378277155,
      "grad_norm": 0.2577783763408661,
      "learning_rate": 3.676474317210099e-06,
      "loss": 0.1381,
      "step": 3700
    },
    {
      "epoch": 1.7368913857677901,
      "grad_norm": 0.2647913992404938,
      "learning_rate": 3.552158519106977e-06,
      "loss": 0.1417,
      "step": 3710
    },
    {
      "epoch": 1.7415730337078652,
      "grad_norm": 0.2621007561683655,
      "learning_rate": 3.4298199337894685e-06,
      "loss": 0.1462,
      "step": 3720
    },
    {
      "epoch": 1.74625468164794,
      "grad_norm": 0.2610481083393097,
      "learning_rate": 3.309469839288301e-06,
      "loss": 0.1315,
      "step": 3730
    },
    {
      "epoch": 1.750936329588015,
      "grad_norm": 0.2840542197227478,
      "learning_rate": 3.1911193303211185e-06,
      "loss": 0.1362,
      "step": 3740
    },
    {
      "epoch": 1.75561797752809,
      "grad_norm": 0.2668418884277344,
      "learning_rate": 3.0747793172697326e-06,
      "loss": 0.1492,
      "step": 3750
    },
    {
      "epoch": 1.7602996254681647,
      "grad_norm": 0.2559303939342499,
      "learning_rate": 2.9604605251743136e-06,
      "loss": 0.1315,
      "step": 3760
    },
    {
      "epoch": 1.7649812734082397,
      "grad_norm": 0.2896023392677307,
      "learning_rate": 2.84817349274466e-06,
      "loss": 0.1631,
      "step": 3770
    },
    {
      "epoch": 1.7696629213483146,
      "grad_norm": 0.2894950211048126,
      "learning_rate": 2.7379285713886954e-06,
      "loss": 0.1472,
      "step": 3780
    },
    {
      "epoch": 1.7743445692883895,
      "grad_norm": 0.281281977891922,
      "learning_rate": 2.6297359242581865e-06,
      "loss": 0.138,
      "step": 3790
    },
    {
      "epoch": 1.7790262172284645,
      "grad_norm": 0.24921460449695587,
      "learning_rate": 2.5236055253118423e-06,
      "loss": 0.1753,
      "step": 3800
    },
    {
      "epoch": 1.7837078651685392,
      "grad_norm": 0.2553311884403229,
      "learning_rate": 2.4195471583958266e-06,
      "loss": 0.1375,
      "step": 3810
    },
    {
      "epoch": 1.7883895131086143,
      "grad_norm": 0.2462325245141983,
      "learning_rate": 2.3175704163418327e-06,
      "loss": 0.133,
      "step": 3820
    },
    {
      "epoch": 1.7930711610486891,
      "grad_norm": 0.2653181552886963,
      "learning_rate": 2.217684700082748e-06,
      "loss": 0.1432,
      "step": 3830
    },
    {
      "epoch": 1.797752808988764,
      "grad_norm": 0.2803376615047455,
      "learning_rate": 2.119899217785995e-06,
      "loss": 0.1458,
      "step": 3840
    },
    {
      "epoch": 1.802434456928839,
      "grad_norm": 0.2679455280303955,
      "learning_rate": 2.0242229840046726e-06,
      "loss": 0.1431,
      "step": 3850
    },
    {
      "epoch": 1.8071161048689137,
      "grad_norm": 0.2571849226951599,
      "learning_rate": 1.9306648188465252e-06,
      "loss": 0.142,
      "step": 3860
    },
    {
      "epoch": 1.8117977528089888,
      "grad_norm": 0.24948078393936157,
      "learning_rate": 1.8392333471608347e-06,
      "loss": 0.1468,
      "step": 3870
    },
    {
      "epoch": 1.8164794007490637,
      "grad_norm": 0.2405436486005783,
      "learning_rate": 1.7499369977433456e-06,
      "loss": 0.1338,
      "step": 3880
    },
    {
      "epoch": 1.8211610486891385,
      "grad_norm": 0.25163865089416504,
      "learning_rate": 1.662784002559209e-06,
      "loss": 0.1439,
      "step": 3890
    },
    {
      "epoch": 1.8258426966292136,
      "grad_norm": 0.2572096288204193,
      "learning_rate": 1.577782395984126e-06,
      "loss": 0.1447,
      "step": 3900
    },
    {
      "epoch": 1.8305243445692883,
      "grad_norm": 0.286592960357666,
      "learning_rate": 1.4949400140636727e-06,
      "loss": 0.1482,
      "step": 3910
    },
    {
      "epoch": 1.8352059925093633,
      "grad_norm": 0.268310546875,
      "learning_rate": 1.4142644937909206e-06,
      "loss": 0.1478,
      "step": 3920
    },
    {
      "epoch": 1.8398876404494382,
      "grad_norm": 0.25413018465042114,
      "learning_rate": 1.3357632724023933e-06,
      "loss": 0.14,
      "step": 3930
    },
    {
      "epoch": 1.844569288389513,
      "grad_norm": 0.25173068046569824,
      "learning_rate": 1.2594435866924686e-06,
      "loss": 0.1444,
      "step": 3940
    },
    {
      "epoch": 1.8492509363295881,
      "grad_norm": 0.2957823574542999,
      "learning_rate": 1.1853124723462311e-06,
      "loss": 0.1437,
      "step": 3950
    },
    {
      "epoch": 1.8539325842696628,
      "grad_norm": 0.2906501293182373,
      "learning_rate": 1.1133767632908799e-06,
      "loss": 0.141,
      "step": 3960
    },
    {
      "epoch": 1.8586142322097379,
      "grad_norm": 0.2822731137275696,
      "learning_rate": 1.043643091065702e-06,
      "loss": 0.1516,
      "step": 3970
    },
    {
      "epoch": 1.8632958801498127,
      "grad_norm": 0.2239583134651184,
      "learning_rate": 9.761178842107699e-07,
      "loss": 0.1289,
      "step": 3980
    },
    {
      "epoch": 1.8679775280898876,
      "grad_norm": 0.2751089632511139,
      "learning_rate": 9.108073676742935e-07,
      "loss": 0.1691,
      "step": 3990
    },
    {
      "epoch": 1.8726591760299627,
      "grad_norm": 0.26302212476730347,
      "learning_rate": 8.477175622387562e-07,
      "loss": 0.1413,
      "step": 4000
    },
    {
      "epoch": 1.8726591760299627,
      "eval_loss": 0.14174455404281616,
      "eval_runtime": 234.3945,
      "eval_samples_per_second": 2.978,
      "eval_steps_per_second": 1.489,
      "step": 4000
    },
    {
      "epoch": 1.8773408239700373,
      "grad_norm": 0.2844927906990051,
      "learning_rate": 7.868542839658793e-07,
      "loss": 0.1337,
      "step": 4010
    },
    {
      "epoch": 1.8820224719101124,
      "grad_norm": 0.2576298415660858,
      "learning_rate": 7.282231436604698e-07,
      "loss": 0.1378,
      "step": 4020
    },
    {
      "epoch": 1.8867041198501873,
      "grad_norm": 0.24751824140548706,
      "learning_rate": 6.718295463531587e-07,
      "loss": 0.1661,
      "step": 4030
    },
    {
      "epoch": 1.8913857677902621,
      "grad_norm": 0.2758065164089203,
      "learning_rate": 6.176786908021453e-07,
      "loss": 0.1506,
      "step": 4040
    },
    {
      "epoch": 1.8960674157303372,
      "grad_norm": 0.24978813529014587,
      "learning_rate": 5.657755690139155e-07,
      "loss": 0.1467,
      "step": 4050
    },
    {
      "epoch": 1.9007490636704119,
      "grad_norm": 0.2835918068885803,
      "learning_rate": 5.161249657830686e-07,
      "loss": 0.1582,
      "step": 4060
    },
    {
      "epoch": 1.905430711610487,
      "grad_norm": 0.2483970671892166,
      "learning_rate": 4.6873145825120926e-07,
      "loss": 0.1382,
      "step": 4070
    },
    {
      "epoch": 1.9101123595505618,
      "grad_norm": 0.24273446202278137,
      "learning_rate": 4.2359941548499037e-07,
      "loss": 0.1229,
      "step": 4080
    },
    {
      "epoch": 1.9147940074906367,
      "grad_norm": 0.2832223176956177,
      "learning_rate": 3.8073299807334926e-07,
      "loss": 0.138,
      "step": 4090
    },
    {
      "epoch": 1.9194756554307117,
      "grad_norm": 0.27922406792640686,
      "learning_rate": 3.4013615774395323e-07,
      "loss": 0.1433,
      "step": 4100
    },
    {
      "epoch": 1.9241573033707864,
      "grad_norm": 0.2603136897087097,
      "learning_rate": 3.0181263699890215e-07,
      "loss": 0.1406,
      "step": 4110
    },
    {
      "epoch": 1.9288389513108615,
      "grad_norm": 0.24417099356651306,
      "learning_rate": 2.657659687697156e-07,
      "loss": 0.1483,
      "step": 4120
    },
    {
      "epoch": 1.9335205992509363,
      "grad_norm": 0.2983071506023407,
      "learning_rate": 2.319994760916433e-07,
      "loss": 0.1497,
      "step": 4130
    },
    {
      "epoch": 1.9382022471910112,
      "grad_norm": 0.23415164649486542,
      "learning_rate": 2.0051627179733247e-07,
      "loss": 0.1601,
      "step": 4140
    },
    {
      "epoch": 1.9428838951310863,
      "grad_norm": 0.280741810798645,
      "learning_rate": 1.713192582298434e-07,
      "loss": 0.1432,
      "step": 4150
    },
    {
      "epoch": 1.947565543071161,
      "grad_norm": 0.3024005889892578,
      "learning_rate": 1.444111269751164e-07,
      "loss": 0.1537,
      "step": 4160
    },
    {
      "epoch": 1.952247191011236,
      "grad_norm": 0.26236817240715027,
      "learning_rate": 1.1979435861383392e-07,
      "loss": 0.1574,
      "step": 4170
    },
    {
      "epoch": 1.9569288389513109,
      "grad_norm": 0.28959619998931885,
      "learning_rate": 9.747122249273156e-08,
      "loss": 0.1396,
      "step": 4180
    },
    {
      "epoch": 1.9616104868913857,
      "grad_norm": 0.2532627582550049,
      "learning_rate": 7.74437765154068e-08,
      "loss": 0.1404,
      "step": 4190
    },
    {
      "epoch": 1.9662921348314608,
      "grad_norm": 0.2636815309524536,
      "learning_rate": 5.971386695260706e-08,
      "loss": 0.1445,
      "step": 4200
    },
    {
      "epoch": 1.9709737827715355,
      "grad_norm": 0.27110251784324646,
      "learning_rate": 4.4283128272018524e-08,
      "loss": 0.1436,
      "step": 4210
    },
    {
      "epoch": 1.9756554307116105,
      "grad_norm": 0.2736012935638428,
      "learning_rate": 3.1152982987595056e-08,
      "loss": 0.1689,
      "step": 4220
    },
    {
      "epoch": 1.9803370786516854,
      "grad_norm": 0.2657409608364105,
      "learning_rate": 2.0324641528421395e-08,
      "loss": 0.1399,
      "step": 4230
    },
    {
      "epoch": 1.9850187265917603,
      "grad_norm": 0.2698764204978943,
      "learning_rate": 1.1799102127130202e-08,
      "loss": 0.1302,
      "step": 4240
    },
    {
      "epoch": 1.9897003745318353,
      "grad_norm": 0.27992117404937744,
      "learning_rate": 5.5771507278645775e-09,
      "loss": 0.1438,
      "step": 4250
    },
    {
      "epoch": 1.99438202247191,
      "grad_norm": 0.2756190001964569,
      "learning_rate": 1.6593609138415567e-09,
      "loss": 0.1459,
      "step": 4260
    },
    {
      "epoch": 1.999063670411985,
      "grad_norm": 0.2640279233455658,
      "learning_rate": 4.6093854466633393e-11,
      "loss": 0.1339,
      "step": 4270
    },
    {
      "epoch": 2.0,
      "step": 4272,
      "total_flos": 2.778312922424672e+18,
      "train_loss": 0.20174445870440552,
      "train_runtime": 74295.9338,
      "train_samples_per_second": 0.92,
      "train_steps_per_second": 0.057
    }
  ],
  "logging_steps": 10,
  "max_steps": 4272,
  "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": 2.778312922424672e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}